vapor-docs/build/2.0/mkdocs/search_index.json

{
    "docs": [
        {
            "location": "/", 
            "text": "Vapor Documentation\n\n\nThis is the documentation for Vapor, a Web Framework for Swift that works on iOS, macOS, and Ubuntu; and all of the packages that Vapor offers.\n\n\nVapor is the most used web framework for Swift. It provides a beautifully expressive and easy to use foundation for your next website or API.\n\n\nGetting Started\n\n\nIf this is your first time using Vapor, head to the \nGetting Started\n section to install Swift and create your first app.\n\n\nViewing Mediums\n\n\nYou can read this guide by clicking through the folders and markdown files on \nGitHub\n or through the rendered \nwebsite\n.\n\n\nOther Sources\n\n\nHere are some other great places to find information about Vapor.\n\n\nAPI\n\n\nAuto-generated API documentation is located at \napi.vapor.codes\n.\n\n\nStack Overflow\n\n\nView or ask questions related to Vapor on Stack Overflow using the \nvapor\n tag.\n\n\nGitHub\n\n\nSource Code\n\n\nTo view the framework's source code and code documentation, visit \nVapor's GitHub\n.\n\n\nIssues\n\n\nTo view open bug reports and feature requests, or to create one, visit the \nissues\n tab on \nVapor's GitHub\n.\n\n\nPackages\n\n\nVapor is a modular framework built for a modular language. Code is split up into modules which are grouped to form packages. Packages can be added to your project by adding the package's Git url to your \nPackage.swift\n file. Once a package is included, all of its modules will be available to \nimport\n. You can read more about packages and modules in the Swift Package Manager \nconceptual overview\n. \n\n\nBelow is a list of packages and modules that come with or can be used by Vapor projects. Packages will have a link to their respective GitHub page.\n\n\nIncluded\n\n\nHere is a list of all the packages and modules included with Vapor. \n\n\n\n\nTip\n\n\nWhile these packages are included in Vapor by default, they can also be used individually.\n\n\n\n\n\n\nVapor\n: Swift's most used web framework.\n\n\nAuth: User authentication and persistance.\n\n\nSessions: Secure, ephemeral cookie based data storage.\n\n\nCookies: HTTP cookies.\n\n\nRouting: Advanced router with type-safe parameterization.\n\n\n\n\n\n\nEngine\n: Core transport layers.\n\n\nHTTP: Pure Swift HTTP client and server.\n\n\nURI: Pure Swift URI parsing and serializing.\n\n\nWebSockets: Full-duplex communication channels over a single TCP connection.\n\n\nSMTP: Send email using Sendgrid and Gmail.\n\n\n\n\n\n\nMultipart\n: Fast, streaming, non-blocking multipart parser and serializer.\n\n\nMultipart: Parses and serializes \nmultipart/mixed\n.\n\n\nFormData: Parses and serializes \nmultipart/form-data\n.\n\n\n\n\n\n\nJSON\n: Conveniences for working with JSON in Swift.\n\n\nConsole\n: Swift wrapper for console IO and commands.\n\n\nTLS\n: Swift wrapper for CLibreSSL's new TLS.\n\n\nCrypto\n: Cryptography from LibreSSL and Swift.\n\n\nDigests: Hashing with and without authentication.\n\n\nCiphers: Encryption and decryption\n\n\nRandom: Pseudo and cryptographically secure randomness.\n\n\nBCrypt: Pure Swift implementation.\n\n\n\n\n\n\nNode\n: Data structure for easy type conversions.\n\n\nPolymorphic\n: Syntax for easily accessing values from common types like JSON.\n\n\nPath Indexable\n: A protocol for powerful subscript access of common types like JSON.\n\n\n\n\n\n\nCore\n: Core extensions, type-aliases, and functions that facilitate common tasks.\n\n\nSocks\n: Swift C Socket API wrapper.\n\n\nBits\n: Low level byte manipulation helpers\n\n\n\n\nExtras\n\n\nThese are officially supported packages for Vapor that are not included by default.\n\n\n\n\nFluent\n: Models, relationships, and querying for NoSQL and SQL databases.\n\n\nFluent Provider\n: Fluent provider for Vapor.\n\n\n\n\n\n\nMySQL\n: Robust MySQL interface for Swift.\n\n\nMySQL Driver\n: MySQL driver for Fluent.\n\n\nMySQL Provider\n: MySQL provider for Vapor.\n\n\n\n\n\n\nLeaf\n: An extensible templating language.\n\n\nLeaf Provider\n: Leaf provider for Vapor.\n\n\n\n\n\n\nRedbird\n: Pure-Swift Redis client implemented from the original protocol spec..\n\n\nRedis Provider\n: Redis cache provider for Vapor.\n\n\n\n\n\n\nJWT\n: JSON Web Tokens in Swift.\n\n\nJWT Provider\n: JWT conveniences for Vapor.\n\n\n\n\n\n\n\n\nThird Party\n\n\nThese are packages created by community members that work great with Vapor.\n\n\n\n\nPostgreSQL\n: Robust PostgreSQL interface for Swift.\n\n\nPostgreSQL Driver\n: PostgreSQL driver for Fluent.\n\n\nPostgreSQL Provider\n: PostgreSQL provider for Vapor.\n\n\n\n\n\n\nMongoKitten*\n: Native MongoDB driver for Swift, written in Swift\n\n\nMongo Driver\n: MongoKitten driver for Fluent.\n\n\nMongo Provider\n: MongoKitten provider for Vapor.\n\n\nMainecoonVapor\n: MongoKitten ORM for Vapor.\n\n\n\n\n\n\nKitura Provider\n: Use IBM's Kitura HTTP server in Vapor.\n\n\nSwiftyBeaver\n: Adds the powerful logging of SwiftyBeaver to Vapor.\n\n\nAPNS\n: Simple APNS Library for Vapor (Swift).\n\n\nVaporFCM\n: Simple FCM (iOS + Android Push Notifications) library built for Vapor in Swift.\n\n\nVaporS3Signer\n: Generate V4 Auth Header/Pre-Signed URL for AWS S3 REST API\n\n\nFlock\n: Automated deployment of Swift projects to servers\n\n\nVaporFlock\n: Use Flock to deploy Vapor applications\n\n\n\n\n\n\nVaporForms\n: Brings simple, dynamic and re-usable web form handling to Vapor.\n\n\nJobs\n: A minimalistic job/background-task system for Swift.\n\n\nHeimdall\n: An easy to use HTTP request logger.\n\n\n\n\nProviders\n\n\nVapor providers are a convenient way to add functionality to your Vapor projects. For a full list of providers, check out the \nvapor-provider\n tag on GitHub.\n\n\nAuthors\n\n\nTanner Nelson\n, \nLogan Wright\n, and the hundreds of members of Vapor.", 
            "title": "Overview"
        }, 
        {
            "location": "/#vapor-documentation", 
            "text": "This is the documentation for Vapor, a Web Framework for Swift that works on iOS, macOS, and Ubuntu; and all of the packages that Vapor offers.  Vapor is the most used web framework for Swift. It provides a beautifully expressive and easy to use foundation for your next website or API.", 
            "title": "Vapor Documentation"
        }, 
        {
            "location": "/#getting-started", 
            "text": "If this is your first time using Vapor, head to the  Getting Started  section to install Swift and create your first app.", 
            "title": "Getting Started"
        }, 
        {
            "location": "/#viewing-mediums", 
            "text": "You can read this guide by clicking through the folders and markdown files on  GitHub  or through the rendered  website .", 
            "title": "Viewing Mediums"
        }, 
        {
            "location": "/#other-sources", 
            "text": "Here are some other great places to find information about Vapor.", 
            "title": "Other Sources"
        }, 
        {
            "location": "/#api", 
            "text": "Auto-generated API documentation is located at  api.vapor.codes .", 
            "title": "API"
        }, 
        {
            "location": "/#stack-overflow", 
            "text": "View or ask questions related to Vapor on Stack Overflow using the  vapor  tag.", 
            "title": "Stack Overflow"
        }, 
        {
            "location": "/#github", 
            "text": "", 
            "title": "GitHub"
        }, 
        {
            "location": "/#source-code", 
            "text": "To view the framework's source code and code documentation, visit  Vapor's GitHub .", 
            "title": "Source Code"
        }, 
        {
            "location": "/#issues", 
            "text": "To view open bug reports and feature requests, or to create one, visit the  issues  tab on  Vapor's GitHub .", 
            "title": "Issues"
        }, 
        {
            "location": "/#packages", 
            "text": "Vapor is a modular framework built for a modular language. Code is split up into modules which are grouped to form packages. Packages can be added to your project by adding the package's Git url to your  Package.swift  file. Once a package is included, all of its modules will be available to  import . You can read more about packages and modules in the Swift Package Manager  conceptual overview .   Below is a list of packages and modules that come with or can be used by Vapor projects. Packages will have a link to their respective GitHub page.", 
            "title": "Packages"
        }, 
        {
            "location": "/#included", 
            "text": "Here is a list of all the packages and modules included with Vapor.    Tip  While these packages are included in Vapor by default, they can also be used individually.    Vapor : Swift's most used web framework.  Auth: User authentication and persistance.  Sessions: Secure, ephemeral cookie based data storage.  Cookies: HTTP cookies.  Routing: Advanced router with type-safe parameterization.    Engine : Core transport layers.  HTTP: Pure Swift HTTP client and server.  URI: Pure Swift URI parsing and serializing.  WebSockets: Full-duplex communication channels over a single TCP connection.  SMTP: Send email using Sendgrid and Gmail.    Multipart : Fast, streaming, non-blocking multipart parser and serializer.  Multipart: Parses and serializes  multipart/mixed .  FormData: Parses and serializes  multipart/form-data .    JSON : Conveniences for working with JSON in Swift.  Console : Swift wrapper for console IO and commands.  TLS : Swift wrapper for CLibreSSL's new TLS.  Crypto : Cryptography from LibreSSL and Swift.  Digests: Hashing with and without authentication.  Ciphers: Encryption and decryption  Random: Pseudo and cryptographically secure randomness.  BCrypt: Pure Swift implementation.    Node : Data structure for easy type conversions.  Polymorphic : Syntax for easily accessing values from common types like JSON.  Path Indexable : A protocol for powerful subscript access of common types like JSON.    Core : Core extensions, type-aliases, and functions that facilitate common tasks.  Socks : Swift C Socket API wrapper.  Bits : Low level byte manipulation helpers", 
            "title": "Included"
        }, 
        {
            "location": "/#extras", 
            "text": "These are officially supported packages for Vapor that are not included by default.   Fluent : Models, relationships, and querying for NoSQL and SQL databases.  Fluent Provider : Fluent provider for Vapor.    MySQL : Robust MySQL interface for Swift.  MySQL Driver : MySQL driver for Fluent.  MySQL Provider : MySQL provider for Vapor.    Leaf : An extensible templating language.  Leaf Provider : Leaf provider for Vapor.    Redbird : Pure-Swift Redis client implemented from the original protocol spec..  Redis Provider : Redis cache provider for Vapor.    JWT : JSON Web Tokens in Swift.  JWT Provider : JWT conveniences for Vapor.", 
            "title": "Extras"
        }, 
        {
            "location": "/#third-party", 
            "text": "These are packages created by community members that work great with Vapor.   PostgreSQL : Robust PostgreSQL interface for Swift.  PostgreSQL Driver : PostgreSQL driver for Fluent.  PostgreSQL Provider : PostgreSQL provider for Vapor.    MongoKitten* : Native MongoDB driver for Swift, written in Swift  Mongo Driver : MongoKitten driver for Fluent.  Mongo Provider : MongoKitten provider for Vapor.  MainecoonVapor : MongoKitten ORM for Vapor.    Kitura Provider : Use IBM's Kitura HTTP server in Vapor.  SwiftyBeaver : Adds the powerful logging of SwiftyBeaver to Vapor.  APNS : Simple APNS Library for Vapor (Swift).  VaporFCM : Simple FCM (iOS + Android Push Notifications) library built for Vapor in Swift.  VaporS3Signer : Generate V4 Auth Header/Pre-Signed URL for AWS S3 REST API  Flock : Automated deployment of Swift projects to servers  VaporFlock : Use Flock to deploy Vapor applications    VaporForms : Brings simple, dynamic and re-usable web form handling to Vapor.  Jobs : A minimalistic job/background-task system for Swift.  Heimdall : An easy to use HTTP request logger.", 
            "title": "Third Party"
        }, 
        {
            "location": "/#providers", 
            "text": "Vapor providers are a convenient way to add functionality to your Vapor projects. For a full list of providers, check out the  vapor-provider  tag on GitHub.", 
            "title": "Providers"
        }, 
        {
            "location": "/#authors", 
            "text": "Tanner Nelson ,  Logan Wright , and the hundreds of members of Vapor.", 
            "title": "Authors"
        }, 
        {
            "location": "/getting-started/install-on-macos/", 
            "text": "Install on macOS\n\n\nTo use Vapor on macOS, you just need to have Xcode 8 installed.\n\n\nInstall Xcode\n\n\nInstall \nXcode 8\n from the Mac App Store.\n\n\n\n\nOpen Xcode\n\n\nAfter Xcode 8 has been downloaded, you must open it to finish the installation. This may take a while.\n\n\nVerify Swift Installation\n\n\nDouble check the installation was successful by running:\n\n\neval\n \n$(\ncurl -sL check2.vapor.sh\n)\n\n\n\n\n\n\nInstall Vapor\n\n\nNow that you have Swift 3.1, let's install the Vapor toolbox. \n\n\nThe toolbox includes all of Vapor's dependencies as well as a handy CLI tool for creating new projects.\n\n\nInstall Homebrew\n\n\nIf you don't already have Homebrew installed, install it! It's incredibly useful for installing software dependencies like OpenSSL, MySQL, Postgres, Redis, SQLite, and more.\n\n\n/usr/bin/ruby -e \n$(\ncurl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install\n)\n\n\n\n\n\n\nFor more information on installing Homebrew, visit \nbrew.sh\n.\n\n\nAdd Homebrew Tap\n\n\nVapor's Homebrew tap will give your Homebrew installation access to all of Vapor's macOS packages.\n\n\nbrew tap vapor/homebrew-tap\nbrew update\n\n\n\n\n\nInstall\n\n\nNow that you've added Vapor's tap, you can install Vapor's toolbox and dependencies.\n\n\nbrew install vapor\n\n\n\n\n\nNext\n\n\nLearn more about the Vapor toolbox CLI in the \nToolbox section\n of the Getting Started section.\n\n\nSwift.org\n\n\nCheck out \nSwift.org\n's extensive guides if you need more detailed instructions for installing Swift 3.1.", 
            "title": "Install: macOS"
        }, 
        {
            "location": "/getting-started/install-on-macos/#install-on-macos", 
            "text": "To use Vapor on macOS, you just need to have Xcode 8 installed.", 
            "title": "Install on macOS"
        }, 
        {
            "location": "/getting-started/install-on-macos/#install-xcode", 
            "text": "Install  Xcode 8  from the Mac App Store.", 
            "title": "Install Xcode"
        }, 
        {
            "location": "/getting-started/install-on-macos/#open-xcode", 
            "text": "After Xcode 8 has been downloaded, you must open it to finish the installation. This may take a while.", 
            "title": "Open Xcode"
        }, 
        {
            "location": "/getting-started/install-on-macos/#verify-swift-installation", 
            "text": "Double check the installation was successful by running:  eval   $( curl -sL check2.vapor.sh )", 
            "title": "Verify Swift Installation"
        }, 
        {
            "location": "/getting-started/install-on-macos/#install-vapor", 
            "text": "Now that you have Swift 3.1, let's install the Vapor toolbox.   The toolbox includes all of Vapor's dependencies as well as a handy CLI tool for creating new projects.", 
            "title": "Install Vapor"
        }, 
        {
            "location": "/getting-started/install-on-macos/#install-homebrew", 
            "text": "If you don't already have Homebrew installed, install it! It's incredibly useful for installing software dependencies like OpenSSL, MySQL, Postgres, Redis, SQLite, and more.  /usr/bin/ruby -e  $( curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install )   For more information on installing Homebrew, visit  brew.sh .", 
            "title": "Install Homebrew"
        }, 
        {
            "location": "/getting-started/install-on-macos/#add-homebrew-tap", 
            "text": "Vapor's Homebrew tap will give your Homebrew installation access to all of Vapor's macOS packages.  brew tap vapor/homebrew-tap\nbrew update", 
            "title": "Add Homebrew Tap"
        }, 
        {
            "location": "/getting-started/install-on-macos/#install", 
            "text": "Now that you've added Vapor's tap, you can install Vapor's toolbox and dependencies.  brew install vapor", 
            "title": "Install"
        }, 
        {
            "location": "/getting-started/install-on-macos/#next", 
            "text": "Learn more about the Vapor toolbox CLI in the  Toolbox section  of the Getting Started section.", 
            "title": "Next"
        }, 
        {
            "location": "/getting-started/install-on-macos/#swiftorg", 
            "text": "Check out  Swift.org 's extensive guides if you need more detailed instructions for installing Swift 3.1.", 
            "title": "Swift.org"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/", 
            "text": "Install on Ubuntu\n\n\nInstalling Vapor on Ubuntu only takes a couple of minutes. \n\n\nAPT Repo\n\n\nAdd Vapor's APT repo to get access to all of Vapor's system packages.\n\n\nQuick Script\n\n\nEasily add Vapor's APT repo with this handy script.\n\n\neval\n \n$(\ncurl -sL https://apt.vapor.sh\n)\n\n\n\n\n\n\nManual\n\n\nOr add the repo manually.\n\n\nwget -q https://repo.vapor.codes/apt/keyring.gpg -O- \n|\n sudo apt-key add -\n\necho\n \ndeb https://repo.vapor.codes/apt \n$(\nlsb_release -sc\n)\n main\n \n|\n sudo tee /etc/apt/sources.list.d/vapor.list\nsudo apt-get update\n\n\n\n\n\nInstall Vapor\n\n\nNow that you have added Vapor's APT repo, you can install the required dependencies.\n\n\nsudo apt-get install swift vapor\n\n\n\n\n\nVerify Installation\n\n\nDouble check the installation was successful by running:\n\n\neval\n \n$(\ncurl -sL check2.vapor.sh\n)\n\n\n\n\n\n\nNext\n\n\nLearn more about the Vapor toolbox CLI in the \nToolbox section\n of the Getting Started section.\n\n\nSwift.org\n\n\nCheck out \nSwift.org\n's guide to \nusing downloads\n if you need more detailed instructions for installing Swift 3.1.", 
            "title": "Install: Ubuntu"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#install-on-ubuntu", 
            "text": "Installing Vapor on Ubuntu only takes a couple of minutes.", 
            "title": "Install on Ubuntu"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#apt-repo", 
            "text": "Add Vapor's APT repo to get access to all of Vapor's system packages.", 
            "title": "APT Repo"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#quick-script", 
            "text": "Easily add Vapor's APT repo with this handy script.  eval   $( curl -sL https://apt.vapor.sh )", 
            "title": "Quick Script"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#manual", 
            "text": "Or add the repo manually.  wget -q https://repo.vapor.codes/apt/keyring.gpg -O-  |  sudo apt-key add - echo   deb https://repo.vapor.codes/apt  $( lsb_release -sc )  main   |  sudo tee /etc/apt/sources.list.d/vapor.list\nsudo apt-get update", 
            "title": "Manual"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#install-vapor", 
            "text": "Now that you have added Vapor's APT repo, you can install the required dependencies.  sudo apt-get install swift vapor", 
            "title": "Install Vapor"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#verify-installation", 
            "text": "Double check the installation was successful by running:  eval   $( curl -sL check2.vapor.sh )", 
            "title": "Verify Installation"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#next", 
            "text": "Learn more about the Vapor toolbox CLI in the  Toolbox section  of the Getting Started section.", 
            "title": "Next"
        }, 
        {
            "location": "/getting-started/install-on-ubuntu/#swiftorg", 
            "text": "Check out  Swift.org 's guide to  using downloads  if you need more detailed instructions for installing Swift 3.1.", 
            "title": "Swift.org"
        }, 
        {
            "location": "/getting-started/toolbox/", 
            "text": "Install Toolbox\n\n\nVapor's command line interface provides shortcuts and assistance for common tasks.\n\n\n\n\n\n\nTip\n\n\nIf you do not want to use the Toolbox or templates, checkout the \nManual\n quickstart.\n\n\n\n\nHelp\n\n\nHelp prints useful information about available commands and flags. You can also run the \n--help\n option on any Toolbox command.\n\n\nvapor --help\n\n\n\n\n\nUpdating\n\n\nThe toolbox should be updated by the package manager it was installed with.\n\n\nHomebrew\n\n\nbrew upgrade vapor\n\n\n\n\n\nAPT\n\n\napt-get update\napt-get install vapor\n\n\n\n\n\nTemplates\n\n\nThe toolbox can create a project from the Vapor basic-template or any other git repo.\n\n\nvapor new \nname\n \n[\n--template\n=\nrepo-url-or-github-path\n]\n\n\n\n\n\n\n\n\nWarning\n\n\nUse \nvapor new --template=api --branch=beta\n while Vapor 2 is in beta\n\n\n\n\nOptions\n\n\nThe toolbox will build an absolute URL based on what you pass as the template option. \n\n\n\n\n--template=light\n clones \nhttp://github.com/vapor/light-template\n.\n\n\n--template=user/repo\n clones \nhttp://github.com/user/repo\n.\n\n\n--template=http://example.com/repo-path\n clones the full url given.\n\n\n\n\n\n\nNote\n\n\nIf you do not specify a template option, the project will be built from Vapor's \nbasic template\n.", 
            "title": "Toolbox"
        }, 
        {
            "location": "/getting-started/toolbox/#install-toolbox", 
            "text": "Vapor's command line interface provides shortcuts and assistance for common tasks.    Tip  If you do not want to use the Toolbox or templates, checkout the  Manual  quickstart.", 
            "title": "Install Toolbox"
        }, 
        {
            "location": "/getting-started/toolbox/#help", 
            "text": "Help prints useful information about available commands and flags. You can also run the  --help  option on any Toolbox command.  vapor --help", 
            "title": "Help"
        }, 
        {
            "location": "/getting-started/toolbox/#updating", 
            "text": "The toolbox should be updated by the package manager it was installed with.", 
            "title": "Updating"
        }, 
        {
            "location": "/getting-started/toolbox/#homebrew", 
            "text": "brew upgrade vapor", 
            "title": "Homebrew"
        }, 
        {
            "location": "/getting-started/toolbox/#apt", 
            "text": "apt-get update\napt-get install vapor", 
            "title": "APT"
        }, 
        {
            "location": "/getting-started/toolbox/#templates", 
            "text": "The toolbox can create a project from the Vapor basic-template or any other git repo.  vapor new  name   [ --template = repo-url-or-github-path ]    Warning  Use  vapor new --template=api --branch=beta  while Vapor 2 is in beta", 
            "title": "Templates"
        }, 
        {
            "location": "/getting-started/toolbox/#options", 
            "text": "The toolbox will build an absolute URL based on what you pass as the template option.    --template=light  clones  http://github.com/vapor/light-template .  --template=user/repo  clones  http://github.com/user/repo .  --template=http://example.com/repo-path  clones the full url given.    Note  If you do not specify a template option, the project will be built from Vapor's  basic template .", 
            "title": "Options"
        }, 
        {
            "location": "/getting-started/hello-world/", 
            "text": "Hello, World\n\n\nThis section assumes you have installed Swift 3.1 and the Vapor Toolbox and have verified they are working.\n\n\n\n\nTip\n\n\nNote: If you don't want to use the Toolbox, follow the \nmanual guide\n.\n\n\n\n\nNew Project\n\n\nLet's start by creating a new project called \"Hello, World\".\n\n\nvapor new Hello --template\n=\napi\n\n\n\n\n\n\n\nWarning\n\n\nUse \nvapor new Hello --template=api --branch=beta\n while Vapor 2 is in beta\n\n\n\n\nVapor's folder structure will probably look familiar to you if you have worked with other web frameworks.\n\n\nHello\n\u251c\u2500\u2500 Sources\n\u2502   \u2514\u2500\u2500 App\n\u2502       \u2514\u2500\u2500 Controllers\n\u2502       \u2514\u2500\u2500 Middleware\n\u2502       \u2514\u2500\u2500 Models\n\u2502       \u2514\u2500\u2500 main.swift\n\u251c\u2500\u2500 Public\n\u251c\u2500\u2500 Resources\n\u2502   \u2514\u2500\u2500 Views\n\u2514\u2500\u2500 Package.swift\n\n\n\n\n\nFor our Hello, World project, we will be focusing on the \nmain.swift\n file.\n\n\nHello\n\u2514\u2500\u2500 Sources\n    \u2514\u2500\u2500 App\n        \u2514\u2500\u2500 main.swift\n\n\n\n\n\n\n\nTip\n\n\nThe \nvapor new\n command creates a new project with examples and comments about how to use the framework. You can delete these if you want.\n\n\n\n\nCode\n\n\nDroplet\n\n\nLook for the following line in the \nmain.swift\n file.\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\n\n\n\nThis is where the one and only \nDroplet\nfor this example will be created. The \nDroplet\n class has a plethora of useful functions on it, and is used extensively.\n\n\nRouting\n\n\nRight after the creation of \ndrop\n, add the following code snippet.\n\n\ndrop\n.\nget\n(\nhello\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \nHello, world!\n\n\n}\n\n\n\n\n\n\nThis creates a new route on the \nDroplet\n that will match all \nGET\n requests to \n/hello\n.\n\n\nAll route closures are passed an instance of \nRequest\n that contains information such as the URI requested and data sent.\n\n\nThis route simply returns a string, but anything that is \nResponseRepresentable\n can be returned. Learn more in the \nRouting\n section of the guide.\n\n\n\n\nTip\n\n\nXcode autocomplete may add extraneous type information to your closure's input arguments. This can be deleted to keep the code clean. If you'd like to keep the type information add \nimport HTTP\n to the top of the file.\n\n\n\n\nServing\n\n\nAt the bottom of the main file, make sure to run your \nDroplet\n.\n\n\ndrop\n.\nrun\n()\n\n\n\n\n\n\nSave the file, and switch back to the terminal.\n\n\nCompile \n Run\n\n\nBuilding\n\n\nA big part of what makes Vapor so great is Swift's state of the art compiler. Let's fire it up. Make sure you are in the root directory of the project and run the following command.\n\n\nvapor\n \nbuild\n\n\n\n\n\n\n\n\nNote\n\n\nvapor build\n runs \nswift build\n in the background.\n\n\n\n\nThe Swift Package Manager will first start by downloading the appropriate dependencies from git. It will then compile and link these dependencies together.\n\n\nWhen the process has completed, you will see \nBuilding Project [Done]\n\n\n\n\nTip\n\n\nIf you see a message like \nunable to execute command: Killed\n, you need to increase your swap space. This can happen if you are running on a machine with limited memory.\n\n\n\n\nRelease\n\n\nBuilding your application in release mode takes longer, but increases performance.\n\n\nvapor build --release\n\n\n\n\n\nServing\n\n\nBoot up the server by running the following command.\n\n\nvapor run serve\n\n\n\n\n\nYou should see a message \nServer starting...\n. You can now visit \nhttp://localhost:8080/hello\n in your browser.\n\n\n\n\nNote\n\n\nCertain port numbers require super user access to bind. Simply run \nsudo vapor run\n to allow access. If you decide to run on a port besides \n80\n, make sure to direct your browser accordingly.\n\n\n\n\nProduction\n\n\nServing your application in the production environment increases its security and performance.\n\n\nvapor run serve --env\n=\nproduction\n\n\n\n\n\nDebug errors will be silenced while in the production environment, so make sure to check your logs for errors.\n\n\n\n\nWarning\n\n\nIf you compiled your application with \n--release\n, make sure to add that flag to the \nvapor run\n command as well. e.g., \nvapor run serve --env=production --release\n.\n\n\n\n\nHello, World\n\n\nYou should see the following output in your browser window.\n\n\nHello, world!\n\n\n\n\n\n\n\nSuccess\n\n\nLike Vapor so far? Click the button below and star the repo to help spread the word!", 
            "title": "Hello, World"
        }, 
        {
            "location": "/getting-started/hello-world/#hello-world", 
            "text": "This section assumes you have installed Swift 3.1 and the Vapor Toolbox and have verified they are working.   Tip  Note: If you don't want to use the Toolbox, follow the  manual guide .", 
            "title": "Hello, World"
        }, 
        {
            "location": "/getting-started/hello-world/#new-project", 
            "text": "Let's start by creating a new project called \"Hello, World\".  vapor new Hello --template = api   Warning  Use  vapor new Hello --template=api --branch=beta  while Vapor 2 is in beta   Vapor's folder structure will probably look familiar to you if you have worked with other web frameworks.  Hello\n\u251c\u2500\u2500 Sources\n\u2502   \u2514\u2500\u2500 App\n\u2502       \u2514\u2500\u2500 Controllers\n\u2502       \u2514\u2500\u2500 Middleware\n\u2502       \u2514\u2500\u2500 Models\n\u2502       \u2514\u2500\u2500 main.swift\n\u251c\u2500\u2500 Public\n\u251c\u2500\u2500 Resources\n\u2502   \u2514\u2500\u2500 Views\n\u2514\u2500\u2500 Package.swift  For our Hello, World project, we will be focusing on the  main.swift  file.  Hello\n\u2514\u2500\u2500 Sources\n    \u2514\u2500\u2500 App\n        \u2514\u2500\u2500 main.swift   Tip  The  vapor new  command creates a new project with examples and comments about how to use the framework. You can delete these if you want.", 
            "title": "New Project"
        }, 
        {
            "location": "/getting-started/hello-world/#code", 
            "text": "", 
            "title": "Code"
        }, 
        {
            "location": "/getting-started/hello-world/#droplet", 
            "text": "Look for the following line in the  main.swift  file.  let   drop   =   try   Droplet ()   This is where the one and only  Droplet for this example will be created. The  Droplet  class has a plethora of useful functions on it, and is used extensively.", 
            "title": "Droplet"
        }, 
        {
            "location": "/getting-started/hello-world/#routing", 
            "text": "Right after the creation of  drop , add the following code snippet.  drop . get ( hello )   {   request   in \n     return   Hello, world!  }   This creates a new route on the  Droplet  that will match all  GET  requests to  /hello .  All route closures are passed an instance of  Request  that contains information such as the URI requested and data sent.  This route simply returns a string, but anything that is  ResponseRepresentable  can be returned. Learn more in the  Routing  section of the guide.   Tip  Xcode autocomplete may add extraneous type information to your closure's input arguments. This can be deleted to keep the code clean. If you'd like to keep the type information add  import HTTP  to the top of the file.", 
            "title": "Routing"
        }, 
        {
            "location": "/getting-started/hello-world/#serving", 
            "text": "At the bottom of the main file, make sure to run your  Droplet .  drop . run ()   Save the file, and switch back to the terminal.", 
            "title": "Serving"
        }, 
        {
            "location": "/getting-started/hello-world/#compile-run", 
            "text": "", 
            "title": "Compile & Run"
        }, 
        {
            "location": "/getting-started/hello-world/#building", 
            "text": "A big part of what makes Vapor so great is Swift's state of the art compiler. Let's fire it up. Make sure you are in the root directory of the project and run the following command.  vapor   build    Note  vapor build  runs  swift build  in the background.   The Swift Package Manager will first start by downloading the appropriate dependencies from git. It will then compile and link these dependencies together.  When the process has completed, you will see  Building Project [Done]   Tip  If you see a message like  unable to execute command: Killed , you need to increase your swap space. This can happen if you are running on a machine with limited memory.", 
            "title": "Building"
        }, 
        {
            "location": "/getting-started/hello-world/#release", 
            "text": "Building your application in release mode takes longer, but increases performance.  vapor build --release", 
            "title": "Release"
        }, 
        {
            "location": "/getting-started/hello-world/#serving_1", 
            "text": "Boot up the server by running the following command.  vapor run serve  You should see a message  Server starting... . You can now visit  http://localhost:8080/hello  in your browser.   Note  Certain port numbers require super user access to bind. Simply run  sudo vapor run  to allow access. If you decide to run on a port besides  80 , make sure to direct your browser accordingly.", 
            "title": "Serving"
        }, 
        {
            "location": "/getting-started/hello-world/#production", 
            "text": "Serving your application in the production environment increases its security and performance.  vapor run serve --env = production  Debug errors will be silenced while in the production environment, so make sure to check your logs for errors.   Warning  If you compiled your application with  --release , make sure to add that flag to the  vapor run  command as well. e.g.,  vapor run serve --env=production --release .", 
            "title": "Production"
        }, 
        {
            "location": "/getting-started/hello-world/#hello-world_1", 
            "text": "You should see the following output in your browser window.  Hello, world!   Success  Like Vapor so far? Click the button below and star the repo to help spread the word!", 
            "title": "Hello, World"
        }, 
        {
            "location": "/getting-started/manual/", 
            "text": "Manual Quickstart\n\n\nLearn how to create a Vapor project \nwithout\n the Toolbox using just Swift 3 and the Swift Package Manager.\n\n\nThis document assumes that you have Swift 3.1 installed.\n\n\n\n\nTip\n\n\nIf you'd prefer to use the Toolbox, follow the toolbox guide \nhere\n.\n\n\n\n\nMake new project using SwiftPM\n\n\nOpen your terminal\n\n\n\n\nNote\n\n\nFor our example, we'll be using the Desktop folder.\n\n\n\n\ncd\n ~/Desktop\nmkdir Hello\n\ncd\n Hello\nswift package init --type executable\n\n\n\n\n\nYour folder should look like this:\n\n\n\u251c\u2500\u2500 Package.swift\n\u251c\u2500\u2500 Sources\n\u2502\u00a0\u00a0 \u2514\u2500\u2500 main.swift\n\u2514\u2500\u2500 Tests\n\n\n\n\n\nEdit \nPackage.swift\n\n\nOpen your \nPackage.swift\n file:\n\n\nopen Package.swift\n\n\n\n\n\nAnd add Vapor as a dependency. Here's how your file will look.\n\n\nPackage.swift\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nHello\n,\n\n    \ndependencies\n:\n \n[\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/vapor.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n]\n\n\n)\n\n\n\n\n\n\n\n\nWarning\n\n\nWe try to keep this document up to date, however, you can view latest releases \nhere\n.\n\n\n\n\nEdit \nmain.swift\n\n\nA simple hello world:\n\n\nimport\n \nVapor\n\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\n\ndrop\n.\nget\n(\nhello\n)\n \n{\n \nreq\n \nin\n\n  \nreturn\n \nHello Vapor\n\n\n}\n\n\n\ndrop\n.\nrun\n()\n\n\n\n\n\n\nCompile \n Run\n\n\nThe first \nbuild\n command can take a while to fetch dependencies.\n\n\nswift build\n.build/debug/Hello serve\n\n\n\n\n\n\n\nWarning\n\n\nIf different, replace \nHello\n above with the name of your executable (as defined in \nPackage.swift\n).\n\n\n\n\nProduction\n\n\nCompiling in Swift's release mode and setting Vapor's environment to production will make your app more secure and performant.\n\n\nswift build --configuration release\n.build/release/Hello serve --env\n=\nproduction\n\n\n\n\n\nView\n\n\nGo to your favorite browser and visit \nhttp://localhost:8080/hello", 
            "title": "Manual"
        }, 
        {
            "location": "/getting-started/manual/#manual-quickstart", 
            "text": "Learn how to create a Vapor project  without  the Toolbox using just Swift 3 and the Swift Package Manager.  This document assumes that you have Swift 3.1 installed.   Tip  If you'd prefer to use the Toolbox, follow the toolbox guide  here .", 
            "title": "Manual Quickstart"
        }, 
        {
            "location": "/getting-started/manual/#make-new-project-using-swiftpm", 
            "text": "Open your terminal   Note  For our example, we'll be using the Desktop folder.   cd  ~/Desktop\nmkdir Hello cd  Hello\nswift package init --type executable  Your folder should look like this:  \u251c\u2500\u2500 Package.swift\n\u251c\u2500\u2500 Sources\n\u2502\u00a0\u00a0 \u2514\u2500\u2500 main.swift\n\u2514\u2500\u2500 Tests", 
            "title": "Make new project using SwiftPM"
        }, 
        {
            "location": "/getting-started/manual/#edit-packageswift", 
            "text": "Open your  Package.swift  file:  open Package.swift  And add Vapor as a dependency. Here's how your file will look.", 
            "title": "Edit Package.swift"
        }, 
        {
            "location": "/getting-started/manual/#packageswift", 
            "text": "import   PackageDescription  let   package   =   Package ( \n     name :   Hello , \n     dependencies :   [ \n         . Package ( url :   https://github.com/vapor/vapor.git ,   majorVersion :   2 ) \n     ]  )    Warning  We try to keep this document up to date, however, you can view latest releases  here .", 
            "title": "Package.swift"
        }, 
        {
            "location": "/getting-started/manual/#edit-mainswift", 
            "text": "A simple hello world:  import   Vapor  let   drop   =   Droplet ()  drop . get ( hello )   {   req   in \n   return   Hello Vapor  }  drop . run ()", 
            "title": "Edit main.swift"
        }, 
        {
            "location": "/getting-started/manual/#compile-run", 
            "text": "The first  build  command can take a while to fetch dependencies.  swift build\n.build/debug/Hello serve   Warning  If different, replace  Hello  above with the name of your executable (as defined in  Package.swift ).", 
            "title": "Compile & Run"
        }, 
        {
            "location": "/getting-started/manual/#production", 
            "text": "Compiling in Swift's release mode and setting Vapor's environment to production will make your app more secure and performant.  swift build --configuration release\n.build/release/Hello serve --env = production", 
            "title": "Production"
        }, 
        {
            "location": "/getting-started/manual/#view", 
            "text": "Go to your favorite browser and visit  http://localhost:8080/hello", 
            "title": "View"
        }, 
        {
            "location": "/getting-started/xcode/", 
            "text": "Xcode\n\n\nIf you're on a Mac, you can develop your Vapor project using Xcode. \nYou can build, run, and stop your server from within Xcode, as well as use breakpoints to debug your code.\n\n\n\n\nTo use Xcode, you will first need to generate a \n*.xcodeproj\n file.\n\n\nGenerate Project\n\n\nVapor Toolbox\n\n\nTo generate a new Xcode project for a project, use:\n\n\nvapor xcode\n\n\n\n\n\n\n\nTip\n\n\nIf you'd like to automatically open the Xcode project, use \nvapor xcode -y\n\n\n\n\nManual\n\n\nTo generate a new Xcode project manually.\n\n\nswift package generate-xcodeproj\n\n\n\n\n\nOpen the project and continue normally.", 
            "title": "Xcode"
        }, 
        {
            "location": "/getting-started/xcode/#xcode", 
            "text": "If you're on a Mac, you can develop your Vapor project using Xcode. \nYou can build, run, and stop your server from within Xcode, as well as use breakpoints to debug your code.   To use Xcode, you will first need to generate a  *.xcodeproj  file.", 
            "title": "Xcode"
        }, 
        {
            "location": "/getting-started/xcode/#generate-project", 
            "text": "", 
            "title": "Generate Project"
        }, 
        {
            "location": "/getting-started/xcode/#vapor-toolbox", 
            "text": "To generate a new Xcode project for a project, use:  vapor xcode   Tip  If you'd like to automatically open the Xcode project, use  vapor xcode -y", 
            "title": "Vapor Toolbox"
        }, 
        {
            "location": "/getting-started/xcode/#manual", 
            "text": "To generate a new Xcode project manually.  swift package generate-xcodeproj  Open the project and continue normally.", 
            "title": "Manual"
        }, 
        {
            "location": "/vapor/folder-structure/", 
            "text": "Folder Structure\n\n\nThe first step to creating an awesome application is knowing where things are. If you created your project using the \nToolbox\n or from a template, you will already have the folder structure created.\n\n\nIf you are making a Vapor application from scratch, this will show you exactly how to set it up.\n\n\nMinimum Folder Structure\n\n\nWe recommend putting all of your Swift code inside of the \nApp/\n folder. This will allow you to create subfolders in \nApp/\n to organize your models and resources.\n\n\nThis works best with the Swift package manager's restrictions on how packages should be structured.\n\n\n.\n\u251c\u2500\u2500 App\n\u2502   \u2514\u2500\u2500 main.swift\n\u251c\u2500\u2500 Public\n\u2514\u2500\u2500 Package.swift\n\n\n\n\n\nThe \nPublic\n folder is where all publicly accessible files should go. This folder will be automatically checked every time a URL is requested that is not found in your routes.\n\n\n\n\nNote\n\n\nThe \nFileMiddleware\n is responsible for accessing files from the \nPublic\n folder.\n\n\n\n\nModels\n\n\nThe \nModels\n folder is a recommendation of where you can put your database and other models.\n\n\n.\n\u251c\u2500\u2500 App\n.   \u2514\u2500\u2500 Models\n.       \u2514\u2500\u2500 User.swift\n\n\n\n\n\nControllers\n\n\nThe \nControllers\n folder is a recommendation of where you can put your route controllers.\n\n\n.\n\u251c\u2500\u2500 App\n.   \u2514\u2500\u2500 Controllers\n.       \u2514\u2500\u2500 UserController.swift\n\n\n\n\n\nViews\n\n\nThe \nViews\n folder in \nResources\n is where Vapor will look when you render views.\n\n\n.\n\u251c\u2500\u2500 App\n\u2514\u2500\u2500 Resources\n    \u2514\u2500\u2500 Views\n         \u2514\u2500\u2500 user.html\n\n\n\n\n\nThe following code would load the \nuser.html\n file.\n\n\ndrop\n.\nview\n.\nmake\n(\nuser.html\n)\n\n\n\n\n\n\nConfig\n\n\nVapor has a sophisticated configuration system that involves a hierarchy of configuration importance.\n\n\n.\n\u251c\u2500\u2500 App\n\u2514\u2500\u2500 Config\n  \u2514\u2500\u2500 app.json         // default app.json\n    \u2514\u2500\u2500 development\n         \u2514\u2500\u2500 app.json  // overrides app.json when in development environment\n    \u2514\u2500\u2500 production\n         \u2514\u2500\u2500 app.json  // overrides app.json when in production environment\n    \u2514\u2500\u2500 secrets\n         \u2514\u2500\u2500 app.json  // overrides app.json in all environments, ignored by git\n\n\n\n\n\n.json\n files are structured in the \nConfig\n folder as shown above. The configuration will be applied dependant on where the \n.json\n file exists in the hierarchy. Learn more in \nConfig\n.\n\n\nLearn about changing environments (the \n--env=\n flag) in the \nDroplet\n section.", 
            "title": "Folder Structure"
        }, 
        {
            "location": "/vapor/folder-structure/#folder-structure", 
            "text": "The first step to creating an awesome application is knowing where things are. If you created your project using the  Toolbox  or from a template, you will already have the folder structure created.  If you are making a Vapor application from scratch, this will show you exactly how to set it up.", 
            "title": "Folder Structure"
        }, 
        {
            "location": "/vapor/folder-structure/#minimum-folder-structure", 
            "text": "We recommend putting all of your Swift code inside of the  App/  folder. This will allow you to create subfolders in  App/  to organize your models and resources.  This works best with the Swift package manager's restrictions on how packages should be structured.  .\n\u251c\u2500\u2500 App\n\u2502   \u2514\u2500\u2500 main.swift\n\u251c\u2500\u2500 Public\n\u2514\u2500\u2500 Package.swift  The  Public  folder is where all publicly accessible files should go. This folder will be automatically checked every time a URL is requested that is not found in your routes.   Note  The  FileMiddleware  is responsible for accessing files from the  Public  folder.", 
            "title": "Minimum Folder Structure"
        }, 
        {
            "location": "/vapor/folder-structure/#models", 
            "text": "The  Models  folder is a recommendation of where you can put your database and other models.  .\n\u251c\u2500\u2500 App\n.   \u2514\u2500\u2500 Models\n.       \u2514\u2500\u2500 User.swift", 
            "title": "Models"
        }, 
        {
            "location": "/vapor/folder-structure/#controllers", 
            "text": "The  Controllers  folder is a recommendation of where you can put your route controllers.  .\n\u251c\u2500\u2500 App\n.   \u2514\u2500\u2500 Controllers\n.       \u2514\u2500\u2500 UserController.swift", 
            "title": "Controllers"
        }, 
        {
            "location": "/vapor/folder-structure/#views", 
            "text": "The  Views  folder in  Resources  is where Vapor will look when you render views.  .\n\u251c\u2500\u2500 App\n\u2514\u2500\u2500 Resources\n    \u2514\u2500\u2500 Views\n         \u2514\u2500\u2500 user.html  The following code would load the  user.html  file.  drop . view . make ( user.html )", 
            "title": "Views"
        }, 
        {
            "location": "/vapor/folder-structure/#config", 
            "text": "Vapor has a sophisticated configuration system that involves a hierarchy of configuration importance.  .\n\u251c\u2500\u2500 App\n\u2514\u2500\u2500 Config\n  \u2514\u2500\u2500 app.json         // default app.json\n    \u2514\u2500\u2500 development\n         \u2514\u2500\u2500 app.json  // overrides app.json when in development environment\n    \u2514\u2500\u2500 production\n         \u2514\u2500\u2500 app.json  // overrides app.json when in production environment\n    \u2514\u2500\u2500 secrets\n         \u2514\u2500\u2500 app.json  // overrides app.json in all environments, ignored by git  .json  files are structured in the  Config  folder as shown above. The configuration will be applied dependant on where the  .json  file exists in the hierarchy. Learn more in  Config .  Learn about changing environments (the  --env=  flag) in the  Droplet  section.", 
            "title": "Config"
        }, 
        {
            "location": "/vapor/droplet/", 
            "text": "Droplet\n\n\nThe \nDroplet\n is a service container that gives you access to many of Vapor's facilities. It is responsible for registering routes, starting the server, appending middleware, and more.\n\n\n\n\nTip\n\n\nNormally applications will only have one Droplet. However, for advanced use cases, it is possible to create more than one.\n\n\n\n\nInitialization\n\n\nAs you have probably already seen, the only thing required to create an instance of \nDroplet\n is to import Vapor.\n\n\nimport\n \nVapor\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\n// your magic here\n\n\n\ntry\n \ndrop\n.\nrun\n()\n\n\n\n\n\n\nCreation of the \nDroplet\n normally happens in the \nmain.swift\n file. \n\n\n\n\nNote\n\n\nFor the sake of simplicity, most of the documentations sample code uses just the \nmain.swift\n file. See the \nmodules\n section in Advanced to learn more about using multiple modules.\n\n\n\n\nEnvironment\n\n\nThe \nenvironment\n property contains the current environment your application is running in. Usually development, testing, or production.\n\n\nif\n \ndrop\n.\nenvironment\n \n==\n \n.\nproduction\n \n{\n\n    \n...\n\n\n}\n\n\n\n\n\n\nThe environment affects \nConfig\n and \nLogging\n. The environment is \ndevelopment\n by default. To change it, pass the \n--env=\n flag as an argument.\n\n\nvapor run serve --env\n=\nproduction\n\n\n\n\n\nIf you are in Xcode, you can pass arguments through the scheme editor.\n\n\n\n\nWarning\n\n\nDebug logs can reduce the number of requests your application can handle per second. Enable production mode to silence non-critical logs.\n\n\n\n\nWorking Directory\n\n\nThe \nworkDir\n property contains a path to the current working directory of the application. Vapor uses this property to find the folders related to your project, such as \nResources\n, \nPublic\n, and \nConfig\n.\n\n\nprint\n(\ndrop\n.\nworkDir\n)\n \n// /var/www/my-project/\n\n\n\n\n\n\nVapor automatically determines the working directory in most situations. However, you may need to manually set it for advanced use cases.\n\n\nYou can override the working directory through the \nDroplet\n's initializer, or by passing the \n--workdir\n argument.\n\n\nvapor run serve --workdir\n=\n/var/www/my-project\n\n\n\n\n\n\nModifying Properties\n\n\nThe \nDroplet\n's properties can be changed programmatically or through configuration.\n\n\nProgrammatic\n\n\nProperties on the \nDroplet\n are constant and can be overridden through the init method.\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nserver\n:\n \nMyServerType\n.\nself\n)\n\n\n\n\n\n\nHere the type of server the \nDroplet\n uses is changed to a custom type. When the \nDroplet\n is run, this custom server type will be booted instead of the default server.\n\n\n\n\nWarning\n\n\nUsing the init method manually can override configured properties.\n\n\n\n\nConfigurable\n\n\nIf you want to modify a property of the \nDroplet\n only in certain cases, you can use \naddConfigurable\n. Say for example you want to email error logs to yourself in production, but you don't want to spam your inbox while developing.\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\nconfig\n.\naddConfigurable\n(\nlog\n:\n \nMyEmailLogger\n.\nself\n,\n \nname\n:\n \nemail\n)\n\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n(\nconfig\n)\n\n\n\n\n\n\nThe \nDroplet\n will continue to use the default logger until you modify the \nConfig/droplet.json\n file to point to your email logger. If this is done in \nConfig/production/droplet.json\n, then your logger will only be used in production.\n\n\n{\n\n    \nlog\n:\n \nemail\n\n\n}\n\n\n\n\n\n\nSupported Properties\n\n\n\n\n\n\n\n\nProperty\n\n\nType\n\n\ndroplet.json\n key\n\n\nConfig Initializable\n\n\n\n\n\n\n\n\n\n\nserver\n\n\nServerProtocol.Type\n\n\nserver\n\n\nno\n\n\n\n\n\n\nclient\n\n\nClientProtocol.Type\n\n\nclient\n\n\nno\n\n\n\n\n\n\nlog\n\n\nLogProtocol\n\n\nlog\n\n\nyes\n\n\n\n\n\n\nhash\n\n\nHashProtocol\n\n\nhash\n\n\nyes\n\n\n\n\n\n\ncipher\n\n\nCipherProtocol\n\n\ncipher\n\n\nyes\n\n\n\n\n\n\nmiddleware\n\n\nMiddleware\n\n\nmiddleware.[server,client]\n\n\nno\n\n\n\n\n\n\nconsole\n\n\nConsoleProtocol\n\n\nconsole\n\n\nyes\n\n\n\n\n\n\ncache\n\n\nCacheProtocol\n\n\ncache\n\n\nyes\n\n\n\n\n\n\n\n\nExample\n\n\nLet's create a custom logger to demonstrate Vapor's configurable properties.\n\n\nAllCapsLogger.swift\n\n\nfinal\n \nclass\n \nAllCapsLogger\n:\n \nLogProtocol\n \n{\n\n    \nvar\n \nenabled\n:\n \n[\nLogLevel\n]\n \n=\n \n[]\n\n    \nfunc\n \nlog\n(\n_\n \nlevel\n:\n \nLogLevel\n,\n \nmessage\n:\n \nString\n,\n \nfile\n:\n \nString\n,\n \nfunction\n:\n \nString\n,\n \nline\n:\n \nInt\n)\n \n{\n\n        \nprint\n(\nmessage\n.\nuppercased\n \n+\n \n!!!\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nNow add the logger to the Droplet using the \naddConfigurable\n method for logs.\n\n\nmain.swift\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\nconfig\n.\naddConfigurable\n(\nlog\n:\n \nAllCapsLogger\n(),\n \nname\n:\n \nall-caps\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n\n\n\nWhenever the \n\"log\"\n property is set to \n\"all-caps\"\n in the \ndroplet.json\n, our new logger will be used. \n\n\nConfig/development/droplet.json\n\n\n{\n\n    \nlog\n:\n \nall-caps\n\n\n}\n\n\n\n\n\n\nHere we are setting our logger only in the \ndevelopment\n environment. All other environments will use Vapor's default logger.\n\n\nConfig Initializable\n\n\nFor an added layer of convenience, you can allow your custom types to be initialized from configuration files.\n\n\nIn our previous example, we initialized an \nAllCapsLogger\n before adding it to the Droplet.\n\n\nLet's say we want to allow our project to configure how many exclamation points get added with each log message.\n\n\nAllCapsLogger.swift\n\n\nfinal\n \nclass\n \nAllCapsLogger\n:\n \nLogProtocol\n \n{\n\n    \nvar\n \nenabled\n:\n \n[\nLogLevel\n]\n \n=\n \n[]\n\n    \nlet\n \nexclamationCount\n:\n \nInt\n\n\n    \ninit\n(\nexclamationCount\n:\n \nInt\n)\n \n{\n\n        \nself\n.\nexclamationCount\n \n=\n \nexclamationCount\n\n    \n}\n\n\n    \nfunc\n \nlog\n(\n_\n \nlevel\n:\n \nLogLevel\n,\n \nmessage\n:\n \nString\n,\n \nfile\n:\n \nString\n,\n \nfunction\n:\n \nString\n,\n \nline\n:\n \nInt\n)\n \n{\n\n        \nprint\n(\nmessage\n.\nuppercased\n \n+\n \nString\n(\nrepeating\n:\n \n!\n,\n \ncount\n:\n \nexclamationCount\n))\n\n    \n}\n\n\n}\n\n\n\nextension\n \nAllCapsLogger\n:\n \nConfigInitializable\n \n{\n\n   \ninit\n(\nconfig\n:\n \nConfig\n)\n \nthrows\n \n{\n\n        \nlet\n \ncount\n \n=\n \nconfig\n[\nallCaps\n,\n \nexclamationCount\n]?.\nint\n \n??\n \n3\n\n        \nself\n.\ninit\n(\nexclamationCount\n:\n \ncount\n)\n\n   \n}\n \n\n}\n\n\n\n\n\n\n\n\nNote\n\n\nThe first parameter to \nconfig\n is the name of the file.\n\n\n\n\nNow that we have conformed our logger to \nConfigInitializable\n, we can pass just the type name to \naddConfigurable\n.\n\n\nmain.swift\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\nconfig\n.\naddConfigurable\n(\nlog\n:\n \nAllCapsLogger\n.\nself\n,\n \nname\n:\n \nall-caps\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n\n\n\nNow if you add a file named \nallCaps.json\n to the \nConfig\n folder, you can configure the logger.\n\n\nallCaps.json\n\n\n{\n\n    \nexclamationCount\n:\n \n5\n\n\n}\n\n\n\n\n\n\nWith this configurable abstraction, you can easily change how your application functions in different environments without needing to hard code these values into your source code.", 
            "title": "Droplet"
        }, 
        {
            "location": "/vapor/droplet/#droplet", 
            "text": "The  Droplet  is a service container that gives you access to many of Vapor's facilities. It is responsible for registering routes, starting the server, appending middleware, and more.   Tip  Normally applications will only have one Droplet. However, for advanced use cases, it is possible to create more than one.", 
            "title": "Droplet"
        }, 
        {
            "location": "/vapor/droplet/#initialization", 
            "text": "As you have probably already seen, the only thing required to create an instance of  Droplet  is to import Vapor.  import   Vapor  let   drop   =   try   Droplet ()  // your magic here  try   drop . run ()   Creation of the  Droplet  normally happens in the  main.swift  file.    Note  For the sake of simplicity, most of the documentations sample code uses just the  main.swift  file. See the  modules  section in Advanced to learn more about using multiple modules.", 
            "title": "Initialization"
        }, 
        {
            "location": "/vapor/droplet/#environment", 
            "text": "The  environment  property contains the current environment your application is running in. Usually development, testing, or production.  if   drop . environment   ==   . production   { \n     ...  }   The environment affects  Config  and  Logging . The environment is  development  by default. To change it, pass the  --env=  flag as an argument.  vapor run serve --env = production  If you are in Xcode, you can pass arguments through the scheme editor.   Warning  Debug logs can reduce the number of requests your application can handle per second. Enable production mode to silence non-critical logs.", 
            "title": "Environment"
        }, 
        {
            "location": "/vapor/droplet/#working-directory", 
            "text": "The  workDir  property contains a path to the current working directory of the application. Vapor uses this property to find the folders related to your project, such as  Resources ,  Public , and  Config .  print ( drop . workDir )   // /var/www/my-project/   Vapor automatically determines the working directory in most situations. However, you may need to manually set it for advanced use cases.  You can override the working directory through the  Droplet 's initializer, or by passing the  --workdir  argument.  vapor run serve --workdir = /var/www/my-project", 
            "title": "Working Directory"
        }, 
        {
            "location": "/vapor/droplet/#modifying-properties", 
            "text": "The  Droplet 's properties can be changed programmatically or through configuration.", 
            "title": "Modifying Properties"
        }, 
        {
            "location": "/vapor/droplet/#programmatic", 
            "text": "Properties on the  Droplet  are constant and can be overridden through the init method.  let   drop   =   try   Droplet ( server :   MyServerType . self )   Here the type of server the  Droplet  uses is changed to a custom type. When the  Droplet  is run, this custom server type will be booted instead of the default server.   Warning  Using the init method manually can override configured properties.", 
            "title": "Programmatic"
        }, 
        {
            "location": "/vapor/droplet/#configurable", 
            "text": "If you want to modify a property of the  Droplet  only in certain cases, you can use  addConfigurable . Say for example you want to email error logs to yourself in production, but you don't want to spam your inbox while developing.  let   config   =   try   Config ()  config . addConfigurable ( log :   MyEmailLogger . self ,   name :   email )  let   drop   =   Droplet ( config )   The  Droplet  will continue to use the default logger until you modify the  Config/droplet.json  file to point to your email logger. If this is done in  Config/production/droplet.json , then your logger will only be used in production.  { \n     log :   email  }", 
            "title": "Configurable"
        }, 
        {
            "location": "/vapor/droplet/#supported-properties", 
            "text": "Property  Type  droplet.json  key  Config Initializable      server  ServerProtocol.Type  server  no    client  ClientProtocol.Type  client  no    log  LogProtocol  log  yes    hash  HashProtocol  hash  yes    cipher  CipherProtocol  cipher  yes    middleware  Middleware  middleware.[server,client]  no    console  ConsoleProtocol  console  yes    cache  CacheProtocol  cache  yes", 
            "title": "Supported Properties"
        }, 
        {
            "location": "/vapor/droplet/#example", 
            "text": "Let's create a custom logger to demonstrate Vapor's configurable properties.  AllCapsLogger.swift  final   class   AllCapsLogger :   LogProtocol   { \n     var   enabled :   [ LogLevel ]   =   [] \n     func   log ( _   level :   LogLevel ,   message :   String ,   file :   String ,   function :   String ,   line :   Int )   { \n         print ( message . uppercased   +   !!! ) \n     }  }   Now add the logger to the Droplet using the  addConfigurable  method for logs.  main.swift  let   config   =   try   Config ()  config . addConfigurable ( log :   AllCapsLogger (),   name :   all-caps )  let   drop   =   try   Droplet ( config )   Whenever the  \"log\"  property is set to  \"all-caps\"  in the  droplet.json , our new logger will be used.   Config/development/droplet.json  { \n     log :   all-caps  }   Here we are setting our logger only in the  development  environment. All other environments will use Vapor's default logger.", 
            "title": "Example"
        }, 
        {
            "location": "/vapor/droplet/#config-initializable", 
            "text": "For an added layer of convenience, you can allow your custom types to be initialized from configuration files.  In our previous example, we initialized an  AllCapsLogger  before adding it to the Droplet.  Let's say we want to allow our project to configure how many exclamation points get added with each log message.  AllCapsLogger.swift  final   class   AllCapsLogger :   LogProtocol   { \n     var   enabled :   [ LogLevel ]   =   [] \n     let   exclamationCount :   Int \n\n     init ( exclamationCount :   Int )   { \n         self . exclamationCount   =   exclamationCount \n     } \n\n     func   log ( _   level :   LogLevel ,   message :   String ,   file :   String ,   function :   String ,   line :   Int )   { \n         print ( message . uppercased   +   String ( repeating :   ! ,   count :   exclamationCount )) \n     }  }  extension   AllCapsLogger :   ConfigInitializable   { \n    init ( config :   Config )   throws   { \n         let   count   =   config [ allCaps ,   exclamationCount ]?. int   ??   3 \n         self . init ( exclamationCount :   count ) \n    }   }    Note  The first parameter to  config  is the name of the file.   Now that we have conformed our logger to  ConfigInitializable , we can pass just the type name to  addConfigurable .  main.swift  let   config   =   try   Config ()  config . addConfigurable ( log :   AllCapsLogger . self ,   name :   all-caps )  let   drop   =   try   Droplet ( config )   Now if you add a file named  allCaps.json  to the  Config  folder, you can configure the logger.  allCaps.json  { \n     exclamationCount :   5  }   With this configurable abstraction, you can easily change how your application functions in different environments without needing to hard code these values into your source code.", 
            "title": "Config Initializable"
        }, 
        {
            "location": "/vapor/views/", 
            "text": "Views\n\n\nViews return HTML data from your application. They can be created from pure HTML documents or passed through renderers such as \nLeaf\n.\n\n\nViews Directory\n\n\nViews are stored in \nResources/Views\n. They are created by calling the \nview\n method on \nDroplet\n.\n\n\nHTML\n\n\nReturning HTML, or any other non-rendered document, is simple. Just use the path of the document relative to the views directory.\n\n\ndrop\n.\nget\n(\nhtml\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \ntry\n \ndrop\n.\nview\n.\nmake\n(\nindex.html\n)\n\n\n}\n\n\n\n\n\n\nTemplating\n\n\nTemplated documents like \nLeaf\n can take a \nContext\n.\n\n\ndrop\n.\nget\n(\ntemplate\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \ntry\n \ndrop\n.\nview\n.\nmake\n(\nwelcome\n,\n \n[\n\n        \nmessage\n:\n \nHello, world!\n\n    \n])\n\n\n}\n\n\n\n\n\n\nThis context will be rendered in the view dynamically based on the \nViewRenderer\n used.\n\n\nPublic Resources\n\n\nAny resources that your views need, such as images, styles, and scripts, should be placed in the \nPublic\n folder at the root of your application.\n\n\nView Renderer\n\n\nAny class that conforms to \nViewRenderer\n can be added to our droplet. \n\n\nimport\n \nVapor\n\n\nimport\n \nVaporLeaf\n\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\n\ndrop\n.\nview\n \n=\n \nLeafRenderer\n(\nviewsDir\n:\n \ndrop\n.\nviewsDir\n)\n\n\n\n\n\n\nAvailable Renderers\n\n\nSearch GitHub\n for Vapor view \nproviders\n that can be added to your application.", 
            "title": "Views"
        }, 
        {
            "location": "/vapor/views/#views", 
            "text": "Views return HTML data from your application. They can be created from pure HTML documents or passed through renderers such as  Leaf .", 
            "title": "Views"
        }, 
        {
            "location": "/vapor/views/#views-directory", 
            "text": "Views are stored in  Resources/Views . They are created by calling the  view  method on  Droplet .", 
            "title": "Views Directory"
        }, 
        {
            "location": "/vapor/views/#html", 
            "text": "Returning HTML, or any other non-rendered document, is simple. Just use the path of the document relative to the views directory.  drop . get ( html )   {   request   in \n     return   try   drop . view . make ( index.html )  }", 
            "title": "HTML"
        }, 
        {
            "location": "/vapor/views/#templating", 
            "text": "Templated documents like  Leaf  can take a  Context .  drop . get ( template )   {   request   in \n     return   try   drop . view . make ( welcome ,   [ \n         message :   Hello, world! \n     ])  }   This context will be rendered in the view dynamically based on the  ViewRenderer  used.", 
            "title": "Templating"
        }, 
        {
            "location": "/vapor/views/#public-resources", 
            "text": "Any resources that your views need, such as images, styles, and scripts, should be placed in the  Public  folder at the root of your application.", 
            "title": "Public Resources"
        }, 
        {
            "location": "/vapor/views/#view-renderer", 
            "text": "Any class that conforms to  ViewRenderer  can be added to our droplet.   import   Vapor  import   VaporLeaf  let   drop   =   Droplet ()  drop . view   =   LeafRenderer ( viewsDir :   drop . viewsDir )", 
            "title": "View Renderer"
        }, 
        {
            "location": "/vapor/views/#available-renderers", 
            "text": "Search GitHub  for Vapor view  providers  that can be added to your application.", 
            "title": "Available Renderers"
        }, 
        {
            "location": "/vapor/controllers/", 
            "text": "Controllers\n\n\nControllers help you organize related functionality into a single place. They can also be used to create RESTful resources.\n\n\nBasic\n\n\nA basic controller looks like the following:\n\n\nimport\n \nVapor\n\n\nimport\n \nHTTP\n\n\n\nfinal\n \nclass\n \nHelloController\n \n{\n\n    \nfunc\n \nsayHello\n(\n_\n \nreq\n:\n \nRequest\n)\n \nthrows\n \n-\n \nResponseRepresentable\n \n{\n\n        \nguard\n \nlet\n \nname\n \n=\n \nreq\n.\ndata\n[\nname\n]?.\nstring\n \nelse\n \n{\n \n            \nthrow\n \nAbort\n(.\nbadRequest\n)\n\n        \n}\n\n\n        \nreturn\n \nHello, \n\\(\nname\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nSimple controllers don't need to conform to any protocols. You are free to design them however you see fit.\n\n\nRegistering\n\n\nThe only required structure is the signature of each method in the controller. In order to register this method into the router, it must have a signature like \n(Request) throws -\n ResponseRepresentable\n. \nRequest\n and \nResponseRepresentable\n are made available by importing the \nHTTP\n module.\n\n\nimport\n \nVapor\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\nlet\n \nhc\n \n=\n \nHelloController\n()\n\n\ndrop\n.\nget\n(\nhello\n,\n \nhandler\n:\n \nhc\n.\nsayHello\n)\n\n\n\n\n\n\nSince the signature of our \nsayHello\n method matches the signature of the closure for the \ndrop.get\n method, we can pass it directly.\n\n\nType Safe\n\n\nYou can also use controller methods with type-safe routing.\n\n\nfinal\n \nclass\n \nHelloController\n \n{\n\n    \n...\n\n\n    \nfunc\n \nsayHelloAlternate\n(\n_\n \nreq\n:\n \nRequest\n,\n \n_\n \nname\n:\n \nString\n)\n \n-\n \nResponseRepresentable\n \n{\n\n        \nreturn\n \nHello, \n\\(\nname\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nWe add a new method called \nsayHelloAlternate\n to the \nHelloController\n that accepts a second parameter \nname: String\n.\n\n\nlet\n \nhc\n \n=\n \nHelloController\n()\n\n\ndrop\n.\nget\n(\nhello\n,\n \nString\n.\nself\n,\n \nhandler\n:\n \nhc\n.\nsayHelloAlternate\n)\n\n\n\n\n\n\nSince this type-safe \ndrop.get\n accepts a signature \n(Request, String) throws -\n ResponseRepresentable\n, our method can now be used as the closure for this route. \n\n\nResources\n\n\nControllers that conform to \nResourceRepresentable\n can be easily registered into a router as a RESTful resource. Let's look at an example of a \nUserController\n.\n\n\nfinal\n \nclass\n \nUserController\n \n{\n\n    \nfunc\n \nindex\n(\n_\n \nrequest\n:\n \nRequest\n)\n \nthrows\n \n-\n \nResponseRepresentable\n \n{\n\n        \nreturn\n \ntry\n \nUser\n.\nall\n().\nmakeNode\n().\nconverted\n(\nto\n:\n \nJSON\n.\nself\n)\n\n    \n}\n\n\n    \nfunc\n \nshow\n(\n_\n \nrequest\n:\n \nRequest\n,\n \n_\n \nuser\n:\n \nUser\n)\n \n-\n \nResponseRepresentable\n \n{\n\n        \nreturn\n \nuser\n\n    \n}\n\n\n}\n\n\n\n\n\n\nHere is a typical user controller with an \nindex\n and \nshow\n route. Indexing returns a JSON list of all users and showing returns a JSON representation of a single user.\n\n\nWe \ncould\n register the controller like so:\n\n\nlet\n \nusers\n \n=\n \nUserController\n()\n\n\ndrop\n.\nget\n(\nusers\n,\n \nhandler\n:\n \nusers\n.\nindex\n)\n\n\ndrop\n.\nget\n(\nusers\n,\n \nUser\n.\nself\n,\n \nhandler\n:\n \nusers\n.\nshow\n)\n\n\n\n\n\n\nBut \nResourceRepresentable\n makes this standard RESTful structure easy.\n\n\nextension\n \nUserController\n:\n \nResourceRepresentable\n \n{\n\n    \nfunc\n \nmakeResource\n()\n \n-\n \nResource\nUser\n \n{\n\n        \nreturn\n \nResource\n(\n\n            \nindex\n:\n \nindex\n,\n\n            \nshow\n:\n \nshow\n\n        \n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nConforming \nUserController\n to \nResourceRepresentable\n requires that the signatures of the \nindex\n and \nshow\n methods match what the \nResource\nUser\n is expecting.\n\n\nHere is a peek into the \nResource\n class.\n\n\nfinal\n \nclass\n \nResource\nModel\n:\n \nStringInitializable\n \n{\n\n    \ntypealias\n \nMultiple\n \n=\n \n(\nRequest\n)\n \nthrows\n \n-\n \nResponseRepresentable\n\n    \ntypealias\n \nItem\n \n=\n \n(\nRequest\n,\n \nModel\n)\n \nthrows\n \n-\n \nResponseRepresentable\n\n\n    \nvar\n \nindex\n:\n \nMultiple\n?\n\n    \nvar\n \nstore\n:\n \nMultiple\n?\n\n    \nvar\n \nshow\n:\n \nItem\n?\n\n    \nvar\n \nreplace\n:\n \nItem\n?\n\n    \nvar\n \nmodify\n:\n \nItem\n?\n\n    \nvar\n \ndestroy\n:\n \nItem\n?\n\n    \nvar\n \nclear\n:\n \nMultiple\n?\n\n    \nvar\n \naboutItem\n:\n \nItem\n?\n\n    \nvar\n \naboutMultiple\n:\n \nMultiple\n?\n\n\n    \n...\n\n\n}\n\n\n\n\n\n\nNow that \nUserController\n conforms to \nResourceRepresentable\n, registering the routes is easy.\n\n\nlet\n \nusers\n \n=\n \nUserController\n()\n\n\ndrop\n.\nresource\n(\nusers\n,\n \nusers\n)\n\n\n\n\n\n\ndrop.resource\n will take care of registering only the routes that have been supplied by the call to \nmakeResource()\n. In this case, only the \nindex\n and \nshow\n routes will be supplied.\n\n\n\n\nNote\n\n\ndrop.resource\n also adds useful defaults for OPTIONS requests. These can be overriden.  \n\n\n\n\nFolder\n\n\nControllers can go anywhere in your application, but they are most often stored in the \nApp/Controllers/\n directory. \n\n\n\n\nTip\n\n\nIf you are building a large application, you may want to create your controllers in a separate module. This will allow you to perform unit tests on your controllers. For more information on creating modules, see the \nmodules\n section in Advanced.", 
            "title": "Controllers"
        }, 
        {
            "location": "/vapor/controllers/#controllers", 
            "text": "Controllers help you organize related functionality into a single place. They can also be used to create RESTful resources.", 
            "title": "Controllers"
        }, 
        {
            "location": "/vapor/controllers/#basic", 
            "text": "A basic controller looks like the following:  import   Vapor  import   HTTP  final   class   HelloController   { \n     func   sayHello ( _   req :   Request )   throws   -   ResponseRepresentable   { \n         guard   let   name   =   req . data [ name ]?. string   else   {  \n             throw   Abort (. badRequest ) \n         } \n\n         return   Hello,  \\( name ) \n     }  }   Simple controllers don't need to conform to any protocols. You are free to design them however you see fit.", 
            "title": "Basic"
        }, 
        {
            "location": "/vapor/controllers/#registering", 
            "text": "The only required structure is the signature of each method in the controller. In order to register this method into the router, it must have a signature like  (Request) throws -  ResponseRepresentable .  Request  and  ResponseRepresentable  are made available by importing the  HTTP  module.  import   Vapor  let   drop   =   try   Droplet ()  let   hc   =   HelloController ()  drop . get ( hello ,   handler :   hc . sayHello )   Since the signature of our  sayHello  method matches the signature of the closure for the  drop.get  method, we can pass it directly.", 
            "title": "Registering"
        }, 
        {
            "location": "/vapor/controllers/#type-safe", 
            "text": "You can also use controller methods with type-safe routing.  final   class   HelloController   { \n     ... \n\n     func   sayHelloAlternate ( _   req :   Request ,   _   name :   String )   -   ResponseRepresentable   { \n         return   Hello,  \\( name ) \n     }  }   We add a new method called  sayHelloAlternate  to the  HelloController  that accepts a second parameter  name: String .  let   hc   =   HelloController ()  drop . get ( hello ,   String . self ,   handler :   hc . sayHelloAlternate )   Since this type-safe  drop.get  accepts a signature  (Request, String) throws -  ResponseRepresentable , our method can now be used as the closure for this route.", 
            "title": "Type Safe"
        }, 
        {
            "location": "/vapor/controllers/#resources", 
            "text": "Controllers that conform to  ResourceRepresentable  can be easily registered into a router as a RESTful resource. Let's look at an example of a  UserController .  final   class   UserController   { \n     func   index ( _   request :   Request )   throws   -   ResponseRepresentable   { \n         return   try   User . all (). makeNode (). converted ( to :   JSON . self ) \n     } \n\n     func   show ( _   request :   Request ,   _   user :   User )   -   ResponseRepresentable   { \n         return   user \n     }  }   Here is a typical user controller with an  index  and  show  route. Indexing returns a JSON list of all users and showing returns a JSON representation of a single user.  We  could  register the controller like so:  let   users   =   UserController ()  drop . get ( users ,   handler :   users . index )  drop . get ( users ,   User . self ,   handler :   users . show )   But  ResourceRepresentable  makes this standard RESTful structure easy.  extension   UserController :   ResourceRepresentable   { \n     func   makeResource ()   -   Resource User   { \n         return   Resource ( \n             index :   index , \n             show :   show \n         ) \n     }  }   Conforming  UserController  to  ResourceRepresentable  requires that the signatures of the  index  and  show  methods match what the  Resource User  is expecting.  Here is a peek into the  Resource  class.  final   class   Resource Model :   StringInitializable   { \n     typealias   Multiple   =   ( Request )   throws   -   ResponseRepresentable \n     typealias   Item   =   ( Request ,   Model )   throws   -   ResponseRepresentable \n\n     var   index :   Multiple ? \n     var   store :   Multiple ? \n     var   show :   Item ? \n     var   replace :   Item ? \n     var   modify :   Item ? \n     var   destroy :   Item ? \n     var   clear :   Multiple ? \n     var   aboutItem :   Item ? \n     var   aboutMultiple :   Multiple ? \n\n     ...  }   Now that  UserController  conforms to  ResourceRepresentable , registering the routes is easy.  let   users   =   UserController ()  drop . resource ( users ,   users )   drop.resource  will take care of registering only the routes that have been supplied by the call to  makeResource() . In this case, only the  index  and  show  routes will be supplied.   Note  drop.resource  also adds useful defaults for OPTIONS requests. These can be overriden.", 
            "title": "Resources"
        }, 
        {
            "location": "/vapor/controllers/#folder", 
            "text": "Controllers can go anywhere in your application, but they are most often stored in the  App/Controllers/  directory.    Tip  If you are building a large application, you may want to create your controllers in a separate module. This will allow you to perform unit tests on your controllers. For more information on creating modules, see the  modules  section in Advanced.", 
            "title": "Folder"
        }, 
        {
            "location": "/vapor/provider/", 
            "text": "Provider\n\n\nThe \nProvider\n protocol creates a simple and predictable way for adding functionality and third party packages to your Vapor project.\n\n\nAdding a Provider\n\n\nAdding a provider to your application takes 2-3 steps.\n\n\nAdd Package\n\n\nAll of Vapor's providers end with the \n-provider\n syntax. You can see a list of \navailable providers\n by searching on our GitHub.\n\n\nTo add the provider to your package, add it as a dependency in your \nPackage.swift\n file.\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nMyApp\n,\n\n    \ndependencies\n:\n \n[\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/vapor.git\n,\n \nmajorVersion\n:\n \n2\n),\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/mysql-provider.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n]\n\n\n)\n\n\n\n\n\n\n\n\nWarning\n\n\nAlways run \nvapor update\n or \nvapor clean\n after editing your \nPackage.swift\n file.\n\n\n\n\nImport\n\n\nOnce the provider has been added, you can import it using \nimport VaporFoo\n where \nFoo\n is the name of the provider.\n\n\nHere is what importing the MySQL provider looks like:\n\n\nimport\n \nVapor\n\n\nimport\n \nVaporMySQL\n\n\n\n\n\n\nAdd to Droplet\n\n\nEvery provider comes with a class named \nProvider\n. Add this class to your Droplet using the \naddProvider\n method. \n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\ntry\n \nconfig\n.\naddProvider\n(\nVaporMySQL\n.\nProvider\n.\nself\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n// ...\n\n\n\ndrop\n.\nrun\n()\n\n\n\n\n\n\nConfiguration\n\n\nSome drivers may require a configuration file. For example, \nVaporMySQL\n requires a \nConfig/mysql.json\n file like the following:\n\n\n{\n\n    \nhostname\n:\n \nlocalhost\n,\n\n    \nuser\n:\n \nroot\n,\n\n    \npassword\n:\n \n,\n\n    \ndatabase\n:\n \nvapor\n\n\n}\n\n\n\n\n\n\nYou will receive an error during the \nDroplet\n's initialization if a configuration file is required.\n\n\n\n\nTip\n\n\nStoring sensitive configuration files (ones that contain passwords) in the \nConfig/secrets\n folder will prevent them from being tracked by git.\n\n\n\n\nManual\n\n\nSome providers can be configured manually by using the Provider's init method. This method can be used instead of configuration files.\n\n\nlet\n \nmysqlProvider\n \n=\n \nVaporMySQL\n.\nProvider\n(\nhost\n:\n \nlocalhost\n,\n \nuser\n:\n \nroot\n,\n \npassword\n:\n \n,\n \ndatabase\n:\n \nvapor\n)\n\n\ntry\n \nconfig\n.\naddProvider\n(\nmysqlProvider\n)\n\n\n\n\n\n\nCreate a Provider\n\n\nCreating a provider is easy, you just need to create a package with a class \nProvider\n that conforms to \nVapor.Provider\n.\n\n\nExample\n\n\nHere is what a provider for an example \nFoo\n package would look like. All the provider does is take a message, then print the message when the \nDroplet\n starts.\n\n\nimport\n \nVapor\n\n\n\npublic\n \nfinal\n \nclass\n \nProvider\n:\n \nVapor\n.\nProvider\n \n{\n\n    \npublic\n \nlet\n \nmessage\n:\n \nString\n\n\n    \npublic\n \nconvenience\n \ninit\n(\nconfig\n:\n \nConfig\n)\n \nthrows\n \n{\n\n        \nguard\n \nlet\n \nmessage\n \n=\n \nconfig\n[\nfoo\n,\n \nmessage\n].\nstring\n \nelse\n \n{\n\n            \nthrow\n \nConfigError\n.\nmissing\n(\nkey\n:\n \n[\nmessage\n],\n \nfile\n:\n \nfoo\n,\n \ndesiredType\n:\n \nString\n.\nself\n)\n\n        \n}\n\n\n        \nself\n.\ninit\n(\nmessage\n:\n \nmessage\n)\n\n    \n}\n\n\n    \npublic\n \ninit\n(\nmessage\n:\n \nString\n)\n \n{\n\n        \nself\n.\nmessage\n \n=\n \nmessage\n\n    \n}\n\n\n    \npublic\n \nfunc\n \nboot\n(\n_\n \ndrop\n:\n \nDroplet\n)\n \n{\n \n}\n\n\n    \npublic\n \nfunc\n \nbeforeRun\n(\n_\n \ndrop\n:\n \nDroplet\n)\n \n{\n\n        \ndrop\n.\nconsole\n.\ninfo\n(\nmessage\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nThis provider wil require a \nConfig/foo.json\n file that looks like:\n\n\n{\n\n    \nmessage\n:\n \nThe message to output\n\n\n}\n\n\n\n\n\n\nThe provider can also be initialized manually with the \ninit(message: String)\n init.", 
            "title": "Provider"
        }, 
        {
            "location": "/vapor/provider/#provider", 
            "text": "The  Provider  protocol creates a simple and predictable way for adding functionality and third party packages to your Vapor project.", 
            "title": "Provider"
        }, 
        {
            "location": "/vapor/provider/#adding-a-provider", 
            "text": "Adding a provider to your application takes 2-3 steps.", 
            "title": "Adding a Provider"
        }, 
        {
            "location": "/vapor/provider/#add-package", 
            "text": "All of Vapor's providers end with the  -provider  syntax. You can see a list of  available providers  by searching on our GitHub.  To add the provider to your package, add it as a dependency in your  Package.swift  file.  let   package   =   Package ( \n     name :   MyApp , \n     dependencies :   [ \n         . Package ( url :   https://github.com/vapor/vapor.git ,   majorVersion :   2 ), \n         . Package ( url :   https://github.com/vapor/mysql-provider.git ,   majorVersion :   2 ) \n     ]  )    Warning  Always run  vapor update  or  vapor clean  after editing your  Package.swift  file.", 
            "title": "Add Package"
        }, 
        {
            "location": "/vapor/provider/#import", 
            "text": "Once the provider has been added, you can import it using  import VaporFoo  where  Foo  is the name of the provider.  Here is what importing the MySQL provider looks like:  import   Vapor  import   VaporMySQL", 
            "title": "Import"
        }, 
        {
            "location": "/vapor/provider/#add-to-droplet", 
            "text": "Every provider comes with a class named  Provider . Add this class to your Droplet using the  addProvider  method.   let   config   =   try   Config ()  try   config . addProvider ( VaporMySQL . Provider . self )  let   drop   =   try   Droplet ( config )  // ...  drop . run ()", 
            "title": "Add to Droplet"
        }, 
        {
            "location": "/vapor/provider/#configuration", 
            "text": "Some drivers may require a configuration file. For example,  VaporMySQL  requires a  Config/mysql.json  file like the following:  { \n     hostname :   localhost , \n     user :   root , \n     password :   , \n     database :   vapor  }   You will receive an error during the  Droplet 's initialization if a configuration file is required.   Tip  Storing sensitive configuration files (ones that contain passwords) in the  Config/secrets  folder will prevent them from being tracked by git.", 
            "title": "Configuration"
        }, 
        {
            "location": "/vapor/provider/#manual", 
            "text": "Some providers can be configured manually by using the Provider's init method. This method can be used instead of configuration files.  let   mysqlProvider   =   VaporMySQL . Provider ( host :   localhost ,   user :   root ,   password :   ,   database :   vapor )  try   config . addProvider ( mysqlProvider )", 
            "title": "Manual"
        }, 
        {
            "location": "/vapor/provider/#create-a-provider", 
            "text": "Creating a provider is easy, you just need to create a package with a class  Provider  that conforms to  Vapor.Provider .", 
            "title": "Create a Provider"
        }, 
        {
            "location": "/vapor/provider/#example", 
            "text": "Here is what a provider for an example  Foo  package would look like. All the provider does is take a message, then print the message when the  Droplet  starts.  import   Vapor  public   final   class   Provider :   Vapor . Provider   { \n     public   let   message :   String \n\n     public   convenience   init ( config :   Config )   throws   { \n         guard   let   message   =   config [ foo ,   message ]. string   else   { \n             throw   ConfigError . missing ( key :   [ message ],   file :   foo ,   desiredType :   String . self ) \n         } \n\n         self . init ( message :   message ) \n     } \n\n     public   init ( message :   String )   { \n         self . message   =   message \n     } \n\n     public   func   boot ( _   drop :   Droplet )   {   } \n\n     public   func   beforeRun ( _   drop :   Droplet )   { \n         drop . console . info ( message ) \n     }  }   This provider wil require a  Config/foo.json  file that looks like:  { \n     message :   The message to output  }   The provider can also be initialized manually with the  init(message: String)  init.", 
            "title": "Example"
        }, 
        {
            "location": "/vapor/hash/", 
            "text": "Hash\n\n\nHashing is a one way method of converting arbitrary data into a fixed size format. Unlike ciphers, data that is hashed cannot be retrieved from the resulting digest. Hashes can be used to create keys, file identifiers, or store credentials.\n\n\n\n\n\n\nHash function diagram from \nWikipedia\n.\n\n\n\n\n\n\nWarning\n\n\nAvoid storing password hashes if possible. If you must, please make sure to research the state of the art before continuing.\n\n\n\n\nMake\n\n\nTo hash a string, use the \nhash\n property on \nDroplet\n.\n\n\nlet\n \ndigest\n \n=\n \ntry\n \ndrop\n.\nhash\n.\nmake\n(\nvapor\n)\n\n\nprint\n(\ndigest\n.\nstring\n)\n\n\n\n\n\n\nChecking\n\n\nSome hashing algorithms create different hash digests for the same message. Because of this, it is necessary to check your hashes using the \ncheck\n method.\n\n\nlet\n \nmatches\n \n=\n \ntry\n \ndrop\n.\nhash\n.\ncheck\n(\nvapor\n,\n \nmatchesHash\n:\n \ndigest\n)\n\n\n\n\n\n\nCryptoHasher\n\n\nBy default, Vapor uses a SHA-256 hasher. You can change this in the configuration files or by giving the \nDroplet\n a different hasher.\n\n\nConfiguration\n\n\nConfig/droplet.json\n\n\n{\n\n    \n...,\n\n    \nhash\n:\n \ncrypto\n,\n\n    \n...\n\n\n}\n\n\n\n\n\n\nConfig/crypto.json\n\n\n{\n\n    \nhash\n:\n \n{\n\n        \nmethod\n:\n \nsha256\n,\n\n        \nencoding\n:\n \nhex\n,\n\n        \nkey\n:\n \npassword\n\n    \n},\n\n    \n...\n\n\n}\n\n\n\n\n\n\nEncoding\n\n\nThe CryptoHasher supports three methods of encoding.\n\n\n\n\nhex\n\n\nbase64\n\n\nplain\n\n\n\n\nKey\n\n\nSupplying a key will cause the hasher to produce \nkeyed\n hashes using HMAC. Some hashers require a key.\n\n\nSupported\n\n\n\n\n\n\n\n\nName\n\n\nMethod\n\n\nRequires Key\n\n\n\n\n\n\n\n\n\n\nSHA-1\n\n\nsha1\n\n\nno\n\n\n\n\n\n\nSHA-224\n\n\nsha224\n\n\nno\n\n\n\n\n\n\nSHA-256\n\n\nsha256\n\n\nno\n\n\n\n\n\n\nSHA-384\n\n\nsha384\n\n\nno\n\n\n\n\n\n\nSHA-512\n\n\nsha512\n\n\nno\n\n\n\n\n\n\nMD4\n\n\nmd4\n\n\nno\n\n\n\n\n\n\nMD5\n\n\nmd5\n\n\nno\n\n\n\n\n\n\nRIPEMD-160\n\n\nripemd160\n\n\nno\n\n\n\n\n\n\nWhirlpool\n\n\nwhirlpool\n\n\nyes\n\n\n\n\n\n\nStreebog-256\n\n\nstreebog256\n\n\nyes\n\n\n\n\n\n\nStreebog-512\n\n\nstreebog512\n\n\nyes\n\n\n\n\n\n\nGostR341194\n\n\ngostr341194\n\n\nyes\n\n\n\n\n\n\n\n\nManual\n\n\nHashers can be swapped without the use of configuration files.\n\n\nHash\n\n\nlet\n \nhash\n \n=\n \nCryptoHasher\n(\n\n    \nhash\n:\n \n.\nsha256\n,\n\n    \nencoding\n:\n \n.\nhex\n\n\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nhash\n:\n \nhash\n)\n\n\n\n\n\n\nHMAC\n\n\nlet\n \nhash\n \n=\n \nCryptoHasher\n(\n\n    \nhmac\n:\n \n.\nsha256\n,\n\n    \nencoding\n:\n \n.\nhex\n,\n\n    \nkey\n:\n \npassword\n.\nmakeBytes\n()\n\n\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nhash\n:\n \nhash\n)\n\n\n\n\n\n\nBCryptHasher\n\n\nBCrypt is a password hashing function that automatically incorporates salts and offers a configurable work factor. The work factor can be used to increase the computation required to generate a hash.\n\n\n\n\nSeealso\n\n\nLearn more about \nkey stretching\n on Wikipedia.\n\n\n\n\nConfiguration\n\n\nTo use the BCryptHasher change the \n\"hash\"\n key in the \ndroplet.json\n configuration file.\n\n\nConfig/droplet.json\n\n\n{\n\n    \n...,\n\n    \nhash\n:\n \nbcrypt\n,\n\n    \n...\n\n\n}\n\n\n\n\n\n\nTo configure the work factor, add a \nbcrypt.json\n file.\n\n\nConfig/bcrypt.json\n\n\n{\n\n    \nworkFactor\n:\n \n8\n\n\n}\n\n\n\n\n\n\n\n\nTip\n\n\nYou can use different BCrypt work factors for production and development modes to keep password hashing fast while working on a project.\n\n\n\n\nManual\n\n\nYou can manually assign a \nBCryptHasher\n to \ndrop.hash\n.\n\n\nlet\n \nhash\n \n=\n \nBCryptHasher\n(\nworkFactor\n:\n \n8\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nhash\n:\n \nhash\n)\n\n\n\n\n\n\nAdvanced\n\n\nCustom\n\n\nYou can also create your own hasher. You just need to conform to the \nHash\n protocol.\n\n\n/// Creates hash digests\n\n\npublic\n \nprotocol\n \nHashProtocol\n \n{\n\n    \n/// Given a message, this method\n\n    \n/// returns a hashed digest of that message.\n\n    \nfunc\n \nmake\n(\n_\n \nmessage\n:\n \nBytes\n)\n \nthrows\n \n-\n \nBytes\n\n\n    \n/// Checks whether a given digest was created\n\n    \n/// by the supplied message.\n\n    \n///\n\n    \n/// Returns true if the digest was created\n\n    \n/// by the supplied message, false otherwise.\n\n    \nfunc\n \ncheck\n(\n_\n \nmessage\n:\n \nBytes\n,\n \nmatchesHash\n:\n \nBytes\n)\n \nthrows\n \n-\n \nBool\n\n\n}", 
            "title": "Hash"
        }, 
        {
            "location": "/vapor/hash/#hash", 
            "text": "Hashing is a one way method of converting arbitrary data into a fixed size format. Unlike ciphers, data that is hashed cannot be retrieved from the resulting digest. Hashes can be used to create keys, file identifiers, or store credentials.    Hash function diagram from  Wikipedia .    Warning  Avoid storing password hashes if possible. If you must, please make sure to research the state of the art before continuing.", 
            "title": "Hash"
        }, 
        {
            "location": "/vapor/hash/#make", 
            "text": "To hash a string, use the  hash  property on  Droplet .  let   digest   =   try   drop . hash . make ( vapor )  print ( digest . string )", 
            "title": "Make"
        }, 
        {
            "location": "/vapor/hash/#checking", 
            "text": "Some hashing algorithms create different hash digests for the same message. Because of this, it is necessary to check your hashes using the  check  method.  let   matches   =   try   drop . hash . check ( vapor ,   matchesHash :   digest )", 
            "title": "Checking"
        }, 
        {
            "location": "/vapor/hash/#cryptohasher", 
            "text": "By default, Vapor uses a SHA-256 hasher. You can change this in the configuration files or by giving the  Droplet  a different hasher.", 
            "title": "CryptoHasher"
        }, 
        {
            "location": "/vapor/hash/#configuration", 
            "text": "Config/droplet.json  { \n     ..., \n     hash :   crypto , \n     ...  }   Config/crypto.json  { \n     hash :   { \n         method :   sha256 , \n         encoding :   hex , \n         key :   password \n     }, \n     ...  }", 
            "title": "Configuration"
        }, 
        {
            "location": "/vapor/hash/#encoding", 
            "text": "The CryptoHasher supports three methods of encoding.   hex  base64  plain", 
            "title": "Encoding"
        }, 
        {
            "location": "/vapor/hash/#key", 
            "text": "Supplying a key will cause the hasher to produce  keyed  hashes using HMAC. Some hashers require a key.", 
            "title": "Key"
        }, 
        {
            "location": "/vapor/hash/#supported", 
            "text": "Name  Method  Requires Key      SHA-1  sha1  no    SHA-224  sha224  no    SHA-256  sha256  no    SHA-384  sha384  no    SHA-512  sha512  no    MD4  md4  no    MD5  md5  no    RIPEMD-160  ripemd160  no    Whirlpool  whirlpool  yes    Streebog-256  streebog256  yes    Streebog-512  streebog512  yes    GostR341194  gostr341194  yes", 
            "title": "Supported"
        }, 
        {
            "location": "/vapor/hash/#manual", 
            "text": "Hashers can be swapped without the use of configuration files.", 
            "title": "Manual"
        }, 
        {
            "location": "/vapor/hash/#hash_1", 
            "text": "let   hash   =   CryptoHasher ( \n     hash :   . sha256 , \n     encoding :   . hex  )  let   drop   =   try   Droplet ( hash :   hash )", 
            "title": "Hash"
        }, 
        {
            "location": "/vapor/hash/#hmac", 
            "text": "let   hash   =   CryptoHasher ( \n     hmac :   . sha256 , \n     encoding :   . hex , \n     key :   password . makeBytes ()  )  let   drop   =   try   Droplet ( hash :   hash )", 
            "title": "HMAC"
        }, 
        {
            "location": "/vapor/hash/#bcrypthasher", 
            "text": "BCrypt is a password hashing function that automatically incorporates salts and offers a configurable work factor. The work factor can be used to increase the computation required to generate a hash.   Seealso  Learn more about  key stretching  on Wikipedia.", 
            "title": "BCryptHasher"
        }, 
        {
            "location": "/vapor/hash/#configuration_1", 
            "text": "To use the BCryptHasher change the  \"hash\"  key in the  droplet.json  configuration file.  Config/droplet.json  { \n     ..., \n     hash :   bcrypt , \n     ...  }   To configure the work factor, add a  bcrypt.json  file.  Config/bcrypt.json  { \n     workFactor :   8  }    Tip  You can use different BCrypt work factors for production and development modes to keep password hashing fast while working on a project.", 
            "title": "Configuration"
        }, 
        {
            "location": "/vapor/hash/#manual_1", 
            "text": "You can manually assign a  BCryptHasher  to  drop.hash .  let   hash   =   BCryptHasher ( workFactor :   8 )  let   drop   =   try   Droplet ( hash :   hash )", 
            "title": "Manual"
        }, 
        {
            "location": "/vapor/hash/#advanced", 
            "text": "", 
            "title": "Advanced"
        }, 
        {
            "location": "/vapor/hash/#custom", 
            "text": "You can also create your own hasher. You just need to conform to the  Hash  protocol.  /// Creates hash digests  public   protocol   HashProtocol   { \n     /// Given a message, this method \n     /// returns a hashed digest of that message. \n     func   make ( _   message :   Bytes )   throws   -   Bytes \n\n     /// Checks whether a given digest was created \n     /// by the supplied message. \n     /// \n     /// Returns true if the digest was created \n     /// by the supplied message, false otherwise. \n     func   check ( _   message :   Bytes ,   matchesHash :   Bytes )   throws   -   Bool  }", 
            "title": "Custom"
        }, 
        {
            "location": "/vapor/log/", 
            "text": "Logging...", 
            "title": "Log"
        }, 
        {
            "location": "/vapor/commands/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\nCommands\n\n\nCustom console commands on Vapor are a breeze.\n\n\nExample\n\n\nTo make a custom console command we must first create a new \n.swift\n file, import \nVapor\n and \nConsole\n, and implement the \nCommand\n protocol.\n\n\nimport\n \nVapor\n\n\nimport\n \nConsole\n\n\n\nfinal\n \nclass\n \nMyCustomCommand\n:\n \nCommand\n \n{\n\n    \npublic\n \nlet\n \nid\n \n=\n \ncommand\n\n    \npublic\n \nlet\n \nhelp\n \n=\n \n[\nThis command does things, like foo, and bar.\n]\n\n    \npublic\n \nlet\n \nconsole\n:\n \nConsoleProtocol\n\n\n    \npublic\n \ninit\n(\nconsole\n:\n \nConsoleProtocol\n)\n \n{\n\n        \nself\n.\nconsole\n \n=\n \nconsole\n\n    \n}\n\n\n    \npublic\n \nfunc\n \nrun\n(\narguments\n:\n \n[\nString\n])\n \nthrows\n \n{\n\n        \nconsole\n.\nprint\n(\nrunning custom command...\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\n\n\nThe \nid\n property is the string you will type in the console to access the command. \n.build/debug/App command\n will run the Custom Command.\n\n\nThe \nhelp\n property is the help message that will give your custom command's users some idea of how to access it.\n\n\nThe \nconsole\n property is the object passed to your custom command that adheres to the console protocol, allowing manipulation of the console.\n\n\nThe \nrun\n method is where you put the logic relating to your command.\n\n\n\n\nAfter we work our magic in the Custom Command file, we switch over to our \nmain.swift\n file and add the custom command to the droplet like so.\n\n\ndrop\n.\ncommands\n.\nappend\n(\nMyCustomCommand\n(\nconsole\n:\n \ndrop\n.\nconsole\n))\n\n\n\n\n\n\nThis allows Vapor access to our custom command and lets it know to display it in the \n--help\n section of the program.\n\n\nAfter compiling the application we can run our custom command like so.\n\n\n.build/debug/App command", 
            "title": "Commands"
        }, 
        {
            "location": "/vapor/commands/#commands", 
            "text": "Custom console commands on Vapor are a breeze.", 
            "title": "Commands"
        }, 
        {
            "location": "/vapor/commands/#example", 
            "text": "To make a custom console command we must first create a new  .swift  file, import  Vapor  and  Console , and implement the  Command  protocol.  import   Vapor  import   Console  final   class   MyCustomCommand :   Command   { \n     public   let   id   =   command \n     public   let   help   =   [ This command does things, like foo, and bar. ] \n     public   let   console :   ConsoleProtocol \n\n     public   init ( console :   ConsoleProtocol )   { \n         self . console   =   console \n     } \n\n     public   func   run ( arguments :   [ String ])   throws   { \n         console . print ( running custom command... ) \n     }  }    The  id  property is the string you will type in the console to access the command.  .build/debug/App command  will run the Custom Command.  The  help  property is the help message that will give your custom command's users some idea of how to access it.  The  console  property is the object passed to your custom command that adheres to the console protocol, allowing manipulation of the console.  The  run  method is where you put the logic relating to your command.   After we work our magic in the Custom Command file, we switch over to our  main.swift  file and add the custom command to the droplet like so.  drop . commands . append ( MyCustomCommand ( console :   drop . console ))   This allows Vapor access to our custom command and lets it know to display it in the  --help  section of the program.  After compiling the application we can run our custom command like so.  .build/debug/App command", 
            "title": "Example"
        }, 
        {
            "location": "/settings/config/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\nConfig\n\n\nAn application's configuration settings. Cloud applications generally require complex configurations that can adjust based on their environment. Vapor intends to provide a flexible configuration interaction that can be customized for a given user.\n\n\nQuickStart\n\n\nFor Vapor applications, configuration files are expected to be nested under a top level folder named \nConfig\n. Here's an example of a basic config featuring a single \nservers\n configuration.\n\n\n./\n\u251c\u2500\u2500 Config/\n\u2502   \u251c\u2500\u2500 servers.json\n\n\n\n\n\nAnd an example of how this might look:\n\n\n{\n\n  \nhttp\n:\n \n{\n\n    \nhost\n:\n \n0.0.0.0\n,\n\n    \nport\n:\n \n8080\n\n  \n}\n\n\n}\n\n\n\n\n\n\nWhat that's saying, is that our application should start a single server named 'http' serving port \n8080\n on host \n0.0.0.0\n. This represents the following url: \nhttp://localhost:8080\n.\n\n\nCustom Keys\n\n\nLet's add a custom key to the \nservers.json\n file:\n\n\n{\n\n  \nhttp\n:\n \n{\n\n    \nhost\n:\n \n0.0.0.0\n,\n\n    \nport\n:\n \n8080\n,\n\n    \ncustom-key\n:\n \ncustom value\n\n  \n}\n\n\n}\n\n\n\n\n\n\nThis can be accessed from your application's config using the following.\n\n\nlet\n \ncustomValue\n \n=\n \ndrop\n.\nconfig\n[\nservers\n,\n \nhttp\n,\n \ncustom-key\n]?.\nstring\n \n??\n \ndefault\n\n\n\n\n\n\nThat's it, feel free to add and utilize keys as necessary to make your application configuration easier.\n\n\nConfig Syntax\n\n\nYou can access your config directory with the following syntax. \napp.config[\n#file-name#\n, \n#path#\n, \n#to#\n, \n#file#\n]\n. For example, let's hypothesize that in addition to the \nservers.json\n file we mentioned earlier, there is also a \nkeys.json\n that looks like this:\n\n\n{\n\n  \ntest-names\n:\n \n[\n\n    \njoe\n,\n\n    \njane\n,\n\n    \nsara\n\n  \n],\n\n  \nmongo\n:\n \n{\n\n    \nurl\n \n:\n \nwww.customMongoUrl.com\n\n  \n}\n\n\n}\n\n\n\n\n\n\nWe can access this file by making sure the first argument in our subscript is keys. To get the first name in our list:\n\n\nlet\n \nname\n \n=\n \ndrop\n.\nconfig\n[\nkeys\n,\n \ntest-names\n,\n \n0\n]?.\nstring\n \n??\n \ndefault\n\n\n\n\n\n\nOr our mongo url:\n\n\nlet\n \nmongoUrl\n \n=\n \ndrop\n.\nconfig\n[\nkeys\n,\n \nmongo\n,\n \nurl\n]?.\nstring\n \n??\n \ndefault\n\n\n\n\n\n\nAdvanced Configurations\n\n\nHaving the default \nservers.json\n is great, but what about more complex scenarios. For example, what if we want a different host in production and in development? These complex scenarios can be achieved by adding additional folders to our \nConfig/\n directory. Here's an example of a folder structure that's setup for production and development environments.\n\n\nWorkingDirectory/\n\u251c\u2500\u2500 Config/\n\u2502   \u251c\u2500\u2500 servers.json\n\u2502   \u251c\u2500\u2500 production/\n\u2502   \u2502   \u2514\u2500\u2500 servers.json\n\u2502   \u251c\u2500\u2500 development/\n\u2502   \u2502   \u2514\u2500\u2500 servers.json\n\u2502   \u2514\u2500\u2500 secrets/\n\u2502       \u2514\u2500\u2500 servers.json\n\n\n\n\n\n\n\nYou can specify the environment through the command line by using --env=. Custom environments are also available, a few are provided by default: production, development, and testing.\n\n\n\n\nvapor run --env\n=\nproduction\n\n\n\n\n\nPRIORITY\n\n\nConfig files will be accessed in the following priority.\n\n\n\n\nCLI (see below)\n\n\nConfig/secrets/\n\n\nConfig/name-of-environment/\n\n\nConfig/\n\n\n\n\nWhat this means is that if a user calls \napp.config[\"servers\", \"host\"]\n, the key will be searched in the CLI first, then the \nsecrets/\n directory, then the top level default configs.\n\n\n\n\nsecrets/\n directory should very likely be added to the gitignore.\n\n\n\n\nEXAMPLE\n\n\nLet's start with the following JSON files.\n\n\nservers.json\n\n\n{\n\n  \nhttp\n:\n \n{\n\n    \nhost\n:\n \n0.0.0.0\n,\n\n    \nport\n:\n \n9000\n\n  \n}\n\n\n}\n\n\n\n\n\n\nproduction/servers.json\n\n\n{\n\n  \nhttp\n:\n \n{\n\n    \nhost\n:\n \n127.0.0.1\n,\n\n    \nport\n:\n \n$PORT\n\n  \n}\n\n\n}\n\n\n\n\n\n\n\n\nThe \n\"$NAME\"\n syntax is available for all values to access environment variables.\n\n\n\n\nPlease notice that \nservers.json\n, and \nproduction/servers.json\n both declare the same keys: \nhost\n, and \nport\n. In our application, we'll call:\n\n\n// will load 0.0.0.0 or 127.0.0.1 based on above config\n\n\nlet\n \nhost\n \n=\n \ndrop\n.\nconfig\n[\nservers\n,\n \nhttp\n,\n \nhost\n]?.\nstring\n \n??\n \n0.0.0.0\n\n\n// will load 9000, or environment variable port.\n\n\nlet\n \nport\n \n=\n \ndrop\n.\nconfig\n[\nservers\n,\n \nhttp\n,\n \nport\n]?.\nint\n \n??\n \n9000\n\n\n\n\n\n\nCOMMAND LINE\n\n\nIn addition to json files nested within the \nConfig/\n directory, we can also use the command line to pass arguments into our config. By default, these values will be set as the \"cli\" file, but more complex options are also available.\n\n\n1. \n--KEY=VALUE\n\n\nArguments set through the command line can be accessed through config's cli file. For example, the following CLI command:\n\n\n--mongo-password\n=\n$MONGO_PASSWORD\n\n\n\n\n\n\nwould be accessible within your application by using the following:\n\n\nlet\n \nmongoPassword\n \n=\n \ndrop\n.\nconfig\n[\ncli\n,\n \nmongo-password\n]?.\nstring\n\n\n\n\n\n\n2. \n--CONFIG:FILE-NAME.KEY=CUSTOM-VALUE\n\n\nIf you want command line arguments set to a file besides \"cli\", you can use this more advanced specification. For example, the following CLI command:\n\n\n--config:keys.analytics\n=\n124ZH61F\n\n\n\n\n\nwould be accessible within your application by using the following:\n\n\nlet\n \nanalyticsKey\n \n=\n \ndrop\n.\nconfig\n[\nkeys\n,\n \nanalytics\n]?.\nstring", 
            "title": "Config"
        }, 
        {
            "location": "/settings/config/#config", 
            "text": "An application's configuration settings. Cloud applications generally require complex configurations that can adjust based on their environment. Vapor intends to provide a flexible configuration interaction that can be customized for a given user.", 
            "title": "Config"
        }, 
        {
            "location": "/settings/config/#quickstart", 
            "text": "For Vapor applications, configuration files are expected to be nested under a top level folder named  Config . Here's an example of a basic config featuring a single  servers  configuration.  ./\n\u251c\u2500\u2500 Config/\n\u2502   \u251c\u2500\u2500 servers.json  And an example of how this might look:  { \n   http :   { \n     host :   0.0.0.0 , \n     port :   8080 \n   }  }   What that's saying, is that our application should start a single server named 'http' serving port  8080  on host  0.0.0.0 . This represents the following url:  http://localhost:8080 .", 
            "title": "QuickStart"
        }, 
        {
            "location": "/settings/config/#custom-keys", 
            "text": "Let's add a custom key to the  servers.json  file:  { \n   http :   { \n     host :   0.0.0.0 , \n     port :   8080 , \n     custom-key :   custom value \n   }  }   This can be accessed from your application's config using the following.  let   customValue   =   drop . config [ servers ,   http ,   custom-key ]?. string   ??   default   That's it, feel free to add and utilize keys as necessary to make your application configuration easier.", 
            "title": "Custom Keys"
        }, 
        {
            "location": "/settings/config/#config-syntax", 
            "text": "You can access your config directory with the following syntax.  app.config[ #file-name# ,  #path# ,  #to# ,  #file# ] . For example, let's hypothesize that in addition to the  servers.json  file we mentioned earlier, there is also a  keys.json  that looks like this:  { \n   test-names :   [ \n     joe , \n     jane , \n     sara \n   ], \n   mongo :   { \n     url   :   www.customMongoUrl.com \n   }  }   We can access this file by making sure the first argument in our subscript is keys. To get the first name in our list:  let   name   =   drop . config [ keys ,   test-names ,   0 ]?. string   ??   default   Or our mongo url:  let   mongoUrl   =   drop . config [ keys ,   mongo ,   url ]?. string   ??   default", 
            "title": "Config Syntax"
        }, 
        {
            "location": "/settings/config/#advanced-configurations", 
            "text": "Having the default  servers.json  is great, but what about more complex scenarios. For example, what if we want a different host in production and in development? These complex scenarios can be achieved by adding additional folders to our  Config/  directory. Here's an example of a folder structure that's setup for production and development environments.  WorkingDirectory/\n\u251c\u2500\u2500 Config/\n\u2502   \u251c\u2500\u2500 servers.json\n\u2502   \u251c\u2500\u2500 production/\n\u2502   \u2502   \u2514\u2500\u2500 servers.json\n\u2502   \u251c\u2500\u2500 development/\n\u2502   \u2502   \u2514\u2500\u2500 servers.json\n\u2502   \u2514\u2500\u2500 secrets/\n\u2502       \u2514\u2500\u2500 servers.json   You can specify the environment through the command line by using --env=. Custom environments are also available, a few are provided by default: production, development, and testing.   vapor run --env = production", 
            "title": "Advanced Configurations"
        }, 
        {
            "location": "/settings/config/#priority", 
            "text": "Config files will be accessed in the following priority.   CLI (see below)  Config/secrets/  Config/name-of-environment/  Config/   What this means is that if a user calls  app.config[\"servers\", \"host\"] , the key will be searched in the CLI first, then the  secrets/  directory, then the top level default configs.   secrets/  directory should very likely be added to the gitignore.", 
            "title": "PRIORITY"
        }, 
        {
            "location": "/settings/config/#example", 
            "text": "Let's start with the following JSON files.", 
            "title": "EXAMPLE"
        }, 
        {
            "location": "/settings/config/#serversjson", 
            "text": "{ \n   http :   { \n     host :   0.0.0.0 , \n     port :   9000 \n   }  }", 
            "title": "servers.json"
        }, 
        {
            "location": "/settings/config/#productionserversjson", 
            "text": "{ \n   http :   { \n     host :   127.0.0.1 , \n     port :   $PORT \n   }  }    The  \"$NAME\"  syntax is available for all values to access environment variables.   Please notice that  servers.json , and  production/servers.json  both declare the same keys:  host , and  port . In our application, we'll call:  // will load 0.0.0.0 or 127.0.0.1 based on above config  let   host   =   drop . config [ servers ,   http ,   host ]?. string   ??   0.0.0.0  // will load 9000, or environment variable port.  let   port   =   drop . config [ servers ,   http ,   port ]?. int   ??   9000", 
            "title": "production/servers.json"
        }, 
        {
            "location": "/settings/config/#command-line", 
            "text": "In addition to json files nested within the  Config/  directory, we can also use the command line to pass arguments into our config. By default, these values will be set as the \"cli\" file, but more complex options are also available.", 
            "title": "COMMAND LINE"
        }, 
        {
            "location": "/settings/config/#1-keyvalue", 
            "text": "Arguments set through the command line can be accessed through config's cli file. For example, the following CLI command:  --mongo-password = $MONGO_PASSWORD   would be accessible within your application by using the following:  let   mongoPassword   =   drop . config [ cli ,   mongo-password ]?. string", 
            "title": "1. --KEY=VALUE"
        }, 
        {
            "location": "/settings/config/#2-configfile-namekeycustom-value", 
            "text": "If you want command line arguments set to a file besides \"cli\", you can use this more advanced specification. For example, the following CLI command:  --config:keys.analytics = 124ZH61F  would be accessible within your application by using the following:  let   analyticsKey   =   drop . config [ keys ,   analytics ]?. string", 
            "title": "2. --CONFIG:FILE-NAME.KEY=CUSTOM-VALUE"
        }, 
        {
            "location": "/json/package/", 
            "text": "Using JSON\n\n\nThis package is included with the Vapor dependency, use\n\n\nimport\n \nJSON", 
            "title": "Package"
        }, 
        {
            "location": "/json/package/#using-json", 
            "text": "This package is included with the Vapor dependency, use  import   JSON", 
            "title": "Using JSON"
        }, 
        {
            "location": "/json/overview/", 
            "text": "JSON\n\n\nJSON is an integral part of Vapor. It powers Vapor's \nConfig\n and is easy to use in both requests and responses.\n\n\nRequest\n\n\nJSON is automatically available in \nrequest.data\n alongside Form URL Encoded, Form Data, and Query data. This allows you to focus on making a great API, not worrying about what content types data will be sent in.\n\n\ndrop\n.\nget\n(\nhello\n)\n \n{\n \nrequest\n \nin\n\n    \nguard\n \nlet\n \nname\n \n=\n \nrequest\n.\ndata\n[\nname\n]?.\nstring\n \nelse\n \n{\n\n        \nthrow\n \nAbort\n(.\nbadRequest\n)\n\n    \n}\n\n    \nreturn\n \nHello, \n\\(\nname\n)\n!\n\n\n}\n\n\n\n\n\n\nThis will return a greeting for any HTTP method or content type that the \nname\n is sent as, including JSON.\n\n\nJSON Only\n\n\nTo specifically target JSON, use the \nrequest.json\n property.\n\n\ndrop\n.\npost\n(\njson\n)\n \n{\n \nrequest\n \nin\n\n    \nguard\n \nlet\n \nname\n \n=\n \nrequest\n.\njson\n?[\nname\n]?.\nstring\n \nelse\n \n{\n\n        \nthrow\n \nAbort\n(.\nbadRequest\n)\n\n    \n}\n\n\n    \nreturn\n \nHello, \n\\(\nname\n)\n!\n\n\n}\n\n\n\n\n\n\nThe above snippet will only work if the request is sent with JSON data.\n\n\nResponse\n\n\nTo respond with JSON, simply create a \nJSON\n object and add values to it.\n\n\ndrop\n.\nget\n(\nversion\n)\n \n{\n \nrequest\n \nin\n\n    \nvar\n \njson\n \n=\n \nJSON\n()\n\n    \ntry\n \njson\n.\nset\n(\nversion\n,\n \n1.0\n)\n\n    \nreturn\n \njson\n\n\n}\n\n\n\n\n\n\nConvertible\n\n\nMaking your classes and structs JSON convertible is a great way to interact with APIs in an organized and DRY way.\n\n\nRepresentable\n\n\nWhen something conforms to \nJSONRepresentable\n, it can be converted into JSON.\n\n\nextension\n \nUser\n:\n \nJSONRepresentable\n \n{\n\n    \nfunc\n \nmakeJSON\n()\n \nthrows\n \n-\n \nJSON\n \n{\n\n        \nvar\n \njson\n \n=\n \nJSON\n()\n\n        \ntry\n \njson\n.\nset\n(\nid\n,\n \nid\n)\n\n        \ntry\n \njson\n.\nset\n(\nname\n,\n \nname\n)\n\n        \ntry\n \njson\n.\nset\n(\nage\n,\n \nage\n)\n\n        \nreturn\n \njson\n\n    \n}\n\n\n}\n\n\n\n\n\n\nNow you can simply return \nuser.makeJSON()\n in your routes.\n\n\ndrop\n.\nget\n(\nusers\n,\n \nUser\n.\ninit\n)\n \n{\n \nreq\n,\n \nuser\n \nin\n\n    \nreturn\n \ntry\n \nuser\n.\nmakeJSON\n()\n\n\n}\n\n\n\n\n\n\nYou can even go a step further and conform your model to \nResponseRepresentable\n. Since it's already \nJSONRepresentable\n\nyou will get the conformance for free.\n\n\nextension\n \nUser\n:\n \nResponseRepresentable\n \n{\n \n}\n\n\n\n\n\n\nNow you can return the model by itself. It will automatically call \n.makeJSON()\n.\n\n\ndrop\n.\nget\n(\nusers\n,\n \nUser\n.\ninit\n)\n \n{\n \nreq\n,\n \nuser\n \nin\n\n    \nreturn\n \ntry\n \nuser\n\n\n}\n\n\n\n\n\n\nInitializable\n\n\nWhen something conforms to \nJSONInitializable\n, it can be created from JSON.\n\n\nextension\n \nUser\n:\n \nJSONInitializable\n \n{\n\n    \nconvenience\n \ninit\n(\njson\n:\n \nJSON\n)\n \nthrows\n \n{\n\n        \ntry\n \nself\n.\ninit\n(\n\n            \nname\n:\n \njson\n.\nget\n(\nname\n),\n\n            \nage\n:\n \njson\n.\nget\n(\nage\n)\n\n        \n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nNow you can simply call \nUser(json: ...)\n to create a user from JSON.\n\n\ndrop\n.\npost\n(\nusers\n)\n \n{\n \nreq\n \nin\n\n    \nguard\n \nlet\n \njson\n \n=\n \nreq\n.\njson\n \nelse\n \n{\n\n        \nthrow\n \nAbort\n(.\nbadRequest\n)\n\n    \n}\n\n\n    \nlet\n \nuser\n \n=\n \ntry\n \nUser\n(\njson\n:\n \njson\n)\n\n    \ntry\n \nuser\n.\nsave\n()\n\n    \nreturn\n \nuser\n\n\n}", 
            "title": "Overview"
        }, 
        {
            "location": "/json/overview/#json", 
            "text": "JSON is an integral part of Vapor. It powers Vapor's  Config  and is easy to use in both requests and responses.", 
            "title": "JSON"
        }, 
        {
            "location": "/json/overview/#request", 
            "text": "JSON is automatically available in  request.data  alongside Form URL Encoded, Form Data, and Query data. This allows you to focus on making a great API, not worrying about what content types data will be sent in.  drop . get ( hello )   {   request   in \n     guard   let   name   =   request . data [ name ]?. string   else   { \n         throw   Abort (. badRequest ) \n     } \n     return   Hello,  \\( name ) !  }   This will return a greeting for any HTTP method or content type that the  name  is sent as, including JSON.", 
            "title": "Request"
        }, 
        {
            "location": "/json/overview/#json-only", 
            "text": "To specifically target JSON, use the  request.json  property.  drop . post ( json )   {   request   in \n     guard   let   name   =   request . json ?[ name ]?. string   else   { \n         throw   Abort (. badRequest ) \n     } \n\n     return   Hello,  \\( name ) !  }   The above snippet will only work if the request is sent with JSON data.", 
            "title": "JSON Only"
        }, 
        {
            "location": "/json/overview/#response", 
            "text": "To respond with JSON, simply create a  JSON  object and add values to it.  drop . get ( version )   {   request   in \n     var   json   =   JSON () \n     try   json . set ( version ,   1.0 ) \n     return   json  }", 
            "title": "Response"
        }, 
        {
            "location": "/json/overview/#convertible", 
            "text": "Making your classes and structs JSON convertible is a great way to interact with APIs in an organized and DRY way.", 
            "title": "Convertible"
        }, 
        {
            "location": "/json/overview/#representable", 
            "text": "When something conforms to  JSONRepresentable , it can be converted into JSON.  extension   User :   JSONRepresentable   { \n     func   makeJSON ()   throws   -   JSON   { \n         var   json   =   JSON () \n         try   json . set ( id ,   id ) \n         try   json . set ( name ,   name ) \n         try   json . set ( age ,   age ) \n         return   json \n     }  }   Now you can simply return  user.makeJSON()  in your routes.  drop . get ( users ,   User . init )   {   req ,   user   in \n     return   try   user . makeJSON ()  }   You can even go a step further and conform your model to  ResponseRepresentable . Since it's already  JSONRepresentable \nyou will get the conformance for free.  extension   User :   ResponseRepresentable   {   }   Now you can return the model by itself. It will automatically call  .makeJSON() .  drop . get ( users ,   User . init )   {   req ,   user   in \n     return   try   user  }", 
            "title": "Representable"
        }, 
        {
            "location": "/json/overview/#initializable", 
            "text": "When something conforms to  JSONInitializable , it can be created from JSON.  extension   User :   JSONInitializable   { \n     convenience   init ( json :   JSON )   throws   { \n         try   self . init ( \n             name :   json . get ( name ), \n             age :   json . get ( age ) \n         ) \n     }  }   Now you can simply call  User(json: ...)  to create a user from JSON.  drop . post ( users )   {   req   in \n     guard   let   json   =   req . json   else   { \n         throw   Abort (. badRequest ) \n     } \n\n     let   user   =   try   User ( json :   json ) \n     try   user . save () \n     return   user  }", 
            "title": "Initializable"
        }, 
        {
            "location": "/routing/package/", 
            "text": "", 
            "title": "Package"
        }, 
        {
            "location": "/routing/overview/", 
            "text": "Basic Routing\n\n\nRouting is one of the most critical parts of a web framework. The router decides which requests get which responses.\n\n\nVapor has a plethora of functionality for routing including route builders, groups, and collections. In this section, we will look at the basics of routing.\n\n\nRegister\n\n\nThe most basic route includes a method, path, and closure.\n\n\ndrop\n.\nget\n(\nwelcome\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \nHello\n\n\n}\n\n\n\n\n\n\nThe standard HTTP methods are available including \nget\n, \npost\n, \nput\n, \npatch\n, \ndelete\n, and \noptions\n.\n\n\ndrop\n.\npost\n(\nform\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \nSubmitted with a POST request\n\n\n}\n\n\n\n\n\n\nNesting\n\n\nTo nest paths (adding \n/\ns in the URL), simply add commas.\n\n\ndrop\n.\nget\n(\nfoo\n,\n \nbar\n,\n \nbaz\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \nYou requested /foo/bar/baz\n\n\n}\n\n\n\n\n\n\nYou can also use \n/\n, but commas are often easier to type and work better with type safe route \nparameters\n.\n\n\nAlternate\n\n\nAn alternate syntax that accepts a \nMethod\n as the first parameter is also available.\n\n\ndrop\n.\nadd\n(.\ntrace\n,\n \nwelcome\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \nHello\n\n\n}\n\n\n\n\n\n\nThis may be useful if you want to register routes dynamically or use a less common method.\n\n\nRequest\n\n\nEach route closure is given a single \nRequest\n. This contains all of the data associated with the request that led to your route closure being called.\n\n\nResponse Representable\n\n\nA route closure can return in three ways:\n\n\n\n\nResponse\n\n\nResponseRepresentable\n\n\nthrow\n\n\n\n\nResponse\n\n\nA custom \nResponse\n can be returned.\n\n\ndrop\n.\nget\n(\nvapor\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \nResponse\n(\nredirect\n:\n \nhttp://vapor.codes\n)\n\n\n}\n\n\n\n\n\n\nThis is useful for creating special responses like redirects. It is also useful for cases where you want to add cookies or other items to the response.\n\n\nResponse Representable\n\n\nAs you have seen in the previous examples, \nString\ns can be returned in route closures. This is because they conform to \nResponseRepresentable\n\n\nA lot of types in Vapor conform to this protocol by default:\n- String\n- Int\n- \nJSON\n\n- \nModel\n\n\ndrop\n.\nget\n(\njson\n)\n \n{\n \nrequest\n \nin\n\n    \nvar\n \njson\n \n=\n \nJSON\n()\n\n    \ntry\n \njson\n.\nset\n(\nnumber\n,\n \n123\n)\n\n    \ntry\n \njson\n.\nset\n(\ntext\n,\n \nunicorns\n)\n\n    \ntry\n \njson\n.\nset\n(\nbool\n,\n \nfalse\n)\n\n    \nreturn\n \njson\n\n\n}\n\n\n\n\n\n\nThrowing\n\n\nIf you are unable to return a response, you may \nthrow\n any object that conforms to \nError\n. Vapor comes with a default error enum \nAbort\n.\n\n\ndrop\n.\nget\n(\n404\n)\n \n{\n \nrequest\n \nin\n\n    \nthrow\n \nAbort\n(.\nnotFound\n)\n\n\n}\n\n\n\n\n\n\nYou can customize the message of these errors by using \nAbort\n\n\ndrop\n.\nget\n(\nerror\n)\n \n{\n \nrequest\n \nin\n\n    \nthrow\n \nAbort\n(.\nbadRequest\n,\n \nreason\n:\n \nSorry \ud83d\ude31\n)\n\n\n}\n\n\n\n\n\n\nThese errors are caught by default in the \nErrorMiddleware\n where they are turned into a JSON response like the following.\n\n\n{\n\n    \nerror:\n \ntrue,\n\n    \nmessage:\n \nthe message\n\n\n}\n\n\n\n\n\n\nIf you want to override this behavior, remove the \nErrorMiddleware\n (key: \n\"error\"\n) from the \nDroplet\n's middleware and add your own.\n\n\nFallback\n\n\nFallback routes allow you to match multiple layers of nesting slashes.\n\n\napp\n.\nget\n(\nanything\n,\n \n*\n)\n \n{\n \nrequest\n \nin\n\n    \nreturn\n \nMatches anything after /anything\n\n\n}\n\n\n\n\n\n\nFor example, the above route matches all of the following and more:\n\n\n\n\n/anything\n\n\n/anything/foo\n\n\n/anything/foo/bar\n\n\n/anything/foo/bar/baz\n\n\n...", 
            "title": "Overview"
        }, 
        {
            "location": "/routing/overview/#basic-routing", 
            "text": "Routing is one of the most critical parts of a web framework. The router decides which requests get which responses.  Vapor has a plethora of functionality for routing including route builders, groups, and collections. In this section, we will look at the basics of routing.", 
            "title": "Basic Routing"
        }, 
        {
            "location": "/routing/overview/#register", 
            "text": "The most basic route includes a method, path, and closure.  drop . get ( welcome )   {   request   in \n     return   Hello  }   The standard HTTP methods are available including  get ,  post ,  put ,  patch ,  delete , and  options .  drop . post ( form )   {   request   in \n     return   Submitted with a POST request  }", 
            "title": "Register"
        }, 
        {
            "location": "/routing/overview/#nesting", 
            "text": "To nest paths (adding  / s in the URL), simply add commas.  drop . get ( foo ,   bar ,   baz )   {   request   in \n     return   You requested /foo/bar/baz  }   You can also use  / , but commas are often easier to type and work better with type safe route  parameters .", 
            "title": "Nesting"
        }, 
        {
            "location": "/routing/overview/#alternate", 
            "text": "An alternate syntax that accepts a  Method  as the first parameter is also available.  drop . add (. trace ,   welcome )   {   request   in \n     return   Hello  }   This may be useful if you want to register routes dynamically or use a less common method.", 
            "title": "Alternate"
        }, 
        {
            "location": "/routing/overview/#request", 
            "text": "Each route closure is given a single  Request . This contains all of the data associated with the request that led to your route closure being called.", 
            "title": "Request"
        }, 
        {
            "location": "/routing/overview/#response-representable", 
            "text": "A route closure can return in three ways:   Response  ResponseRepresentable  throw", 
            "title": "Response Representable"
        }, 
        {
            "location": "/routing/overview/#response", 
            "text": "A custom  Response  can be returned.  drop . get ( vapor )   {   request   in \n     return   Response ( redirect :   http://vapor.codes )  }   This is useful for creating special responses like redirects. It is also useful for cases where you want to add cookies or other items to the response.", 
            "title": "Response"
        }, 
        {
            "location": "/routing/overview/#response-representable_1", 
            "text": "As you have seen in the previous examples,  String s can be returned in route closures. This is because they conform to  ResponseRepresentable  A lot of types in Vapor conform to this protocol by default:\n- String\n- Int\n-  JSON \n-  Model  drop . get ( json )   {   request   in \n     var   json   =   JSON () \n     try   json . set ( number ,   123 ) \n     try   json . set ( text ,   unicorns ) \n     try   json . set ( bool ,   false ) \n     return   json  }", 
            "title": "Response Representable"
        }, 
        {
            "location": "/routing/overview/#throwing", 
            "text": "If you are unable to return a response, you may  throw  any object that conforms to  Error . Vapor comes with a default error enum  Abort .  drop . get ( 404 )   {   request   in \n     throw   Abort (. notFound )  }   You can customize the message of these errors by using  Abort  drop . get ( error )   {   request   in \n     throw   Abort (. badRequest ,   reason :   Sorry \ud83d\ude31 )  }   These errors are caught by default in the  ErrorMiddleware  where they are turned into a JSON response like the following.  { \n     error:   true, \n     message:   the message  }   If you want to override this behavior, remove the  ErrorMiddleware  (key:  \"error\" ) from the  Droplet 's middleware and add your own.", 
            "title": "Throwing"
        }, 
        {
            "location": "/routing/overview/#fallback", 
            "text": "Fallback routes allow you to match multiple layers of nesting slashes.  app . get ( anything ,   * )   {   request   in \n     return   Matches anything after /anything  }   For example, the above route matches all of the following and more:   /anything  /anything/foo  /anything/foo/bar  /anything/foo/bar/baz  ...", 
            "title": "Fallback"
        }, 
        {
            "location": "/routing/parameters/", 
            "text": "Routing Parameters\n\n\nTraditional web frameworks leave room for error in routing by using strings for route parameter names and types. Vapor takes advantage of Swift's closures to provide a safer and more intuitive method for accessing route parameters.\n\n\n\n\nSeealso\n\n\nRoute parameters refer to segments of the URL path (e.g., \n/users/:id\n). For query parameters (e.g., \n?foo=bar\n) see \nrequest query parameters\n.\n\n\n\n\nType Safe\n\n\nTo create a type safe route simply replace one of the parts of your path with a \nType\n.\n\n\ndrop\n.\nget\n(\nusers\n,\n \nInt\n.\nparameter\n)\n \n{\n \nreq\n \nin\n\n    \nlet\n \nuserId\n \n=\n \ntry\n \nreq\n.\nparameters\n.\nnext\n(\nInt\n.\nself\n)\n\n    \nreturn\n \nYou requested User #\n\\(\nuserId\n)\n\n\n}\n\n\n\n\n\n\nThis creates a route that matches \nusers/:id\n where the \n:id\n is an \nInt\n. Here's what it would look like using manual route parameters.\n\n\ndrop\n.\nget\n(\nusers\n,\n \n:id\n)\n \n{\n \nrequest\n \nin\n\n    \nguard\n \nlet\n \nuserId\n \n=\n \nrequest\n.\nparameters\n[\nid\n]?.\nint\n \nelse\n \n{\n\n        \nthrow\n \nAbort\n.\nbadRequest\n\n    \n}\n\n\n    \nreturn\n \nYou requested User #\n\\(\nuserId\n)\n\n\n}\n\n\n\n\n\n\nHere you can see that type safe routing saves ~3 lines of code and also prevents runtime errors like misspelling \n:id\n.\n\n\nParameterizable\n\n\nAny type conforming to \nParameterizable\n can be used as a parameter. By default, all Vapor \nModels\n conform.\n\n\nUsing this, our previous example with users can be further simplified.\n\n\ndrop\n.\nget\n(\nusers\n,\n \nUser\n.\nparameter\n)\n \n{\n \nreq\n \nin\n\n    \nlet\n \nuser\n \n=\n \ntry\n \nreq\n.\nparameters\n.\nnext\n(\nUser\n.\nself\n)\n\n\n    \nreturn\n \nYou requested \n\\(\nuser\n.\nname\n)\n\n\n}\n\n\n\n\n\n\nHere the identifier supplied is automatically used to lookup a user. For example, if \n/users/5\n is requested, the \nUser\n model will be asked for a user with identifier \n5\n. If one is found, the request succeeds and the closure is called. If not, a not found error is thrown.\n\n\nHere is what this would look like if we looked the model up manually.\n\n\ndrop\n.\nget\n(\nusers\n,\n \nInt\n.\nparameter\n)\n \n{\n \nreq\n \nin\n\n    \nlet\n \nuserId\n \n=\n \ntry\n \nreq\n.\nparameters\n.\nnext\n(\nInt\n.\nself\n)\n\n    \nguard\n \nlet\n \nuser\n \n=\n \ntry\n \nUser\n.\nfind\n(\nuserId\n)\n \nelse\n \n{\n\n        \nthrow\n \nAbort\n.\nnotFound\n\n    \n}\n\n\n    \nreturn\n \nYou requested \n\\(\nuser\n.\nname\n)\n\n\n}\n\n\n\n\n\n\nProtocol\n\n\nYou can conform your own types to \nParameterizable\n.\n\n\nimport\n \nRouting\n\n\n\nextension\n \nFoo\n:\n \nParameterizable\n \n{\n\n    \n/// This unique slug is used to identify\n\n    \n/// the parameter in the router\n\n    \nstatic\n \nvar\n \nuniqueSlug\n:\n \nString\n \n{\n\n        \nreturn\n \nfoo\n\n    \n}\n\n\n\n    \nstatic\n \nfunc\n \nmake\n(\nfor\n \nparameter\n:\n \nString\n)\n \nthrows\n \n-\n \nFoo\n \n{\n\n        \n/// custom lookup logic here\n\n        \n/// the parameter string contains the information\n\n        \n/// parsed from the URL.\n\n        \n...\n\n    \n}\n\n\n}\n\n\n\n\n\n\nNow you can use this type for type safe routing.\n\n\ndrop\n.\nget\n(\nusers\n,\n \nnickname\n,\n \nFoo\n.\nparameter\n)\n \n{\n \nreq\n \nin\n\n    \nlet\n \nfoo\n \n=\n \ntry\n \nreq\n.\nparameters\n.\nnext\n(\nFoo\n.\nself\n)\n\n    \n...\n\n\n}\n\n\n\n\n\n\nGroups\n\n\nType-safe parameters also work with \ngroups\n.\n\n\nlet\n \nuserGroup\n \n=\n \ndrop\n.\ngrouped\n(\nusers\n,\n \nUser\n.\nparameter\n)\n\n\nuserGroup\n.\nget\n(\nmessages\n)\n \n{\n \nreq\n \nin\n\n    \nlet\n \nuser\n \n=\n \ntry\n \nreq\n.\nparameters\n.\nnext\n(\nUser\n.\nself\n)\n\n\n    \n...\n\n\n}", 
            "title": "Parameters"
        }, 
        {
            "location": "/routing/parameters/#routing-parameters", 
            "text": "Traditional web frameworks leave room for error in routing by using strings for route parameter names and types. Vapor takes advantage of Swift's closures to provide a safer and more intuitive method for accessing route parameters.   Seealso  Route parameters refer to segments of the URL path (e.g.,  /users/:id ). For query parameters (e.g.,  ?foo=bar ) see  request query parameters .", 
            "title": "Routing Parameters"
        }, 
        {
            "location": "/routing/parameters/#type-safe", 
            "text": "To create a type safe route simply replace one of the parts of your path with a  Type .  drop . get ( users ,   Int . parameter )   {   req   in \n     let   userId   =   try   req . parameters . next ( Int . self ) \n     return   You requested User # \\( userId )  }   This creates a route that matches  users/:id  where the  :id  is an  Int . Here's what it would look like using manual route parameters.  drop . get ( users ,   :id )   {   request   in \n     guard   let   userId   =   request . parameters [ id ]?. int   else   { \n         throw   Abort . badRequest \n     } \n\n     return   You requested User # \\( userId )  }   Here you can see that type safe routing saves ~3 lines of code and also prevents runtime errors like misspelling  :id .", 
            "title": "Type Safe"
        }, 
        {
            "location": "/routing/parameters/#parameterizable", 
            "text": "Any type conforming to  Parameterizable  can be used as a parameter. By default, all Vapor  Models  conform.  Using this, our previous example with users can be further simplified.  drop . get ( users ,   User . parameter )   {   req   in \n     let   user   =   try   req . parameters . next ( User . self ) \n\n     return   You requested  \\( user . name )  }   Here the identifier supplied is automatically used to lookup a user. For example, if  /users/5  is requested, the  User  model will be asked for a user with identifier  5 . If one is found, the request succeeds and the closure is called. If not, a not found error is thrown.  Here is what this would look like if we looked the model up manually.  drop . get ( users ,   Int . parameter )   {   req   in \n     let   userId   =   try   req . parameters . next ( Int . self ) \n     guard   let   user   =   try   User . find ( userId )   else   { \n         throw   Abort . notFound \n     } \n\n     return   You requested  \\( user . name )  }", 
            "title": "Parameterizable"
        }, 
        {
            "location": "/routing/parameters/#protocol", 
            "text": "You can conform your own types to  Parameterizable .  import   Routing  extension   Foo :   Parameterizable   { \n     /// This unique slug is used to identify \n     /// the parameter in the router \n     static   var   uniqueSlug :   String   { \n         return   foo \n     } \n\n\n     static   func   make ( for   parameter :   String )   throws   -   Foo   { \n         /// custom lookup logic here \n         /// the parameter string contains the information \n         /// parsed from the URL. \n         ... \n     }  }   Now you can use this type for type safe routing.  drop . get ( users ,   nickname ,   Foo . parameter )   {   req   in \n     let   foo   =   try   req . parameters . next ( Foo . self ) \n     ...  }", 
            "title": "Protocol"
        }, 
        {
            "location": "/routing/parameters/#groups", 
            "text": "Type-safe parameters also work with  groups .  let   userGroup   =   drop . grouped ( users ,   User . parameter )  userGroup . get ( messages )   {   req   in \n     let   user   =   try   req . parameters . next ( User . self ) \n\n     ...  }", 
            "title": "Groups"
        }, 
        {
            "location": "/routing/group/", 
            "text": "Route Groups\n\n\nGrouping routes together makes it easy to add common prefixes, middleware, or hosts to multiple routes.\n\n\nRoute groups have two different forms: Group and Grouped.\n\n\nGroup\n\n\nGroup (without the \"ed\" at the end) takes a closure that is passed a \nRouteBuilder\n.\n\n\ndrop\n.\ngroup\n(\nv1\n)\n \n{\n \nv1\n \nin\n\n    \nv1\n.\nget\n(\nusers\n)\n \n{\n \nrequest\n \nin\n\n        \n// get the users\n\n    \n}\n\n\n}\n\n\n\n\n\n\nGrouped\n\n\nGrouped returns a \nRouteBuilder\n that you can pass around.\n\n\nlet\n \nv1\n \n=\n \ndrop\n.\ngrouped\n(\nv1\n)\n\n\nv1\n.\nget\n(\nusers\n)\n \n{\n \nrequest\n \nin\n\n    \n// get the users\n\n\n}\n\n\n\n\n\n\nMiddleware\n\n\nYou can add middleware to a group of routes. This is especially useful for authentication.\n\n\ndrop\n.\ngroup\n(\nAuthMiddleware\n())\n \n{\n \nauthorized\n \nin\n \n    \nauthorized\n.\nget\n(\ntoken\n)\n \n{\n \nrequest\n \nin\n\n        \n// has been authorized\n\n    \n}\n\n\n}\n\n\n\n\n\n\nHost\n\n\nYou can limit the host for a group of routes.\n\n\ndrop\n.\ngroup\n(\nhost\n:\n \nvapor.codes\n)\n \n{\n \nvapor\n \nin\n\n    \nvapor\n.\nget\n \n{\n \nrequest\n \nin\n\n        \n// only responds to requests to vapor.codes\n\n    \n}\n\n\n}\n\n\n\n\n\n\nChaining\n\n\nGroups can be chained together.\n\n\ndrop\n.\ngrouped\n(\nhost\n:\n \nvapor.codes\n).\ngrouped\n(\nAuthMiddleware\n()).\ngroup\n(\nv1\n)\n \n{\n \nauthedSecureV1\n \nin\n\n    \n// add routes here\n\n\n}", 
            "title": "Group"
        }, 
        {
            "location": "/routing/group/#route-groups", 
            "text": "Grouping routes together makes it easy to add common prefixes, middleware, or hosts to multiple routes.  Route groups have two different forms: Group and Grouped.", 
            "title": "Route Groups"
        }, 
        {
            "location": "/routing/group/#group", 
            "text": "Group (without the \"ed\" at the end) takes a closure that is passed a  RouteBuilder .  drop . group ( v1 )   {   v1   in \n     v1 . get ( users )   {   request   in \n         // get the users \n     }  }", 
            "title": "Group"
        }, 
        {
            "location": "/routing/group/#grouped", 
            "text": "Grouped returns a  RouteBuilder  that you can pass around.  let   v1   =   drop . grouped ( v1 )  v1 . get ( users )   {   request   in \n     // get the users  }", 
            "title": "Grouped"
        }, 
        {
            "location": "/routing/group/#middleware", 
            "text": "You can add middleware to a group of routes. This is especially useful for authentication.  drop . group ( AuthMiddleware ())   {   authorized   in  \n     authorized . get ( token )   {   request   in \n         // has been authorized \n     }  }", 
            "title": "Middleware"
        }, 
        {
            "location": "/routing/group/#host", 
            "text": "You can limit the host for a group of routes.  drop . group ( host :   vapor.codes )   {   vapor   in \n     vapor . get   {   request   in \n         // only responds to requests to vapor.codes \n     }  }", 
            "title": "Host"
        }, 
        {
            "location": "/routing/group/#chaining", 
            "text": "Groups can be chained together.  drop . grouped ( host :   vapor.codes ). grouped ( AuthMiddleware ()). group ( v1 )   {   authedSecureV1   in \n     // add routes here  }", 
            "title": "Chaining"
        }, 
        {
            "location": "/routing/collection/", 
            "text": "Route Collections\n\n\nRoute collections allow multiple routes and route groups to be organized in different files or modules.\n\n\nExample\n\n\nHere is an example of a route collection for the \nv1\n portion of an API.\n\n\nimport\n \nVapor\n\n\nimport\n \nHTTP\n\n\nimport\n \nRouting\n\n\n\nclass\n \nV1Collection\n:\n \nRouteCollection\n \n{\n\n    \nfunc\n \nbuild\n(\n_\n \nbuilder\n:\n \nRouteBuilder\n)\n \n{\n\n        \nlet\n \nv1\n \n=\n \nbuilder\n.\ngrouped\n(\nv1\n)\n\n        \nlet\n \nusers\n \n=\n \nv1\n.\ngrouped\n(\nusers\n)\n\n        \nlet\n \narticles\n \n=\n \nv1\n.\ngrouped\n(\narticles\n)\n\n\n        \nusers\n.\nget\n \n{\n \nrequest\n \nin\n\n            \nreturn\n \nRequested all users.\n\n        \n}\n\n\n        \narticles\n.\nget\n(\nArticle\n.\ninit\n)\n \n{\n \nrequest\n,\n \narticle\n \nin\n\n            \nreturn\n \nRequested \n\\(\narticle\n.\nname\n)\n\n        \n}\n\n    \n}\n\n\n}\n\n\n\n\n\n\nThis class could be place in any file, and we could add it to our droplet or even another route group.\n\n\nlet\n \nv1\n \n=\n \nV1Collection\n()\n\n\ndrop\n.\ncollection\n(\nv1\n)\n\n\n\n\n\n\nThe \nDroplet\n will then be passed to the \nbuild(_:)\n method of your route collection and have the various routes added to it.\n\n\nEmpty Initializable\n\n\nYou can add \nEmptyInitializable\n to your route collection if it has an empty \ninit\n method. This will allow you to add the route collection via its type name.\n\n\nclass\n \nV1Collection\n:\n \nRouteCollection\n,\n \nEmptyInitializable\n \n{\n\n    \ninit\n()\n \n{\n \n}\n\n    \n...\n\n\n\n\n\n\nNow we can add the collection without initializing it.\n\n\ndrop\n.\ncollection\n(\nV1Collection\n.\nself\n)", 
            "title": "Collection"
        }, 
        {
            "location": "/routing/collection/#route-collections", 
            "text": "Route collections allow multiple routes and route groups to be organized in different files or modules.", 
            "title": "Route Collections"
        }, 
        {
            "location": "/routing/collection/#example", 
            "text": "Here is an example of a route collection for the  v1  portion of an API.  import   Vapor  import   HTTP  import   Routing  class   V1Collection :   RouteCollection   { \n     func   build ( _   builder :   RouteBuilder )   { \n         let   v1   =   builder . grouped ( v1 ) \n         let   users   =   v1 . grouped ( users ) \n         let   articles   =   v1 . grouped ( articles ) \n\n         users . get   {   request   in \n             return   Requested all users. \n         } \n\n         articles . get ( Article . init )   {   request ,   article   in \n             return   Requested  \\( article . name ) \n         } \n     }  }   This class could be place in any file, and we could add it to our droplet or even another route group.  let   v1   =   V1Collection ()  drop . collection ( v1 )   The  Droplet  will then be passed to the  build(_:)  method of your route collection and have the various routes added to it.", 
            "title": "Example"
        }, 
        {
            "location": "/routing/collection/#empty-initializable", 
            "text": "You can add  EmptyInitializable  to your route collection if it has an empty  init  method. This will allow you to add the route collection via its type name.  class   V1Collection :   RouteCollection ,   EmptyInitializable   { \n     init ()   {   } \n     ...   Now we can add the collection without initializing it.  drop . collection ( V1Collection . self )", 
            "title": "Empty Initializable"
        }, 
        {
            "location": "/fluent/package/", 
            "text": "Using Fluent\n\n\nThis section outlines how to import the Fluent package both with or without a Vapor project.\n\n\nWith Vapor\n\n\nFluent comes included with most Vapor templates. However, if you have created a project from scratch you will need to add the provider to your \nPackage.swift\n file.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/vapor.git\n,\n \nmajorVersion\n:\n \n2\n),\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/fluent-provider.git\n,\n \nmajorVersion\n:\n \n1\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nThe Fluent provider package adds Fluent to your project and adds some additional, Vapor-specific conveniences like HTTP conformances. \n\n\nUsing \nimport FluentProvider\n will import both Fluent and Fluent's Vapor-specific APIs. \n\n\nWithout Vapor\n\n\nFluent is a powerful, pure-Swift ORM that can be used with any Server-Side Swift framework. To include it in your package, add it to your \nPackage.swift\n file.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/fluent.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Fluent\n to access Fluent's APIs.\n\n\n\n\nWarning\n\n\nModel\n is a Vapor + Fluent type, use \nEntity\n instead.\n\n\n\n\nDrivers\n\n\nFluent drivers allow Fluent models and queries to communicate with various database technologies like MySQL or Mongo. For a full list of drivers, check out the \nfluent-driver\n tag on GitHub.", 
            "title": "Package"
        }, 
        {
            "location": "/fluent/package/#using-fluent", 
            "text": "This section outlines how to import the Fluent package both with or without a Vapor project.", 
            "title": "Using Fluent"
        }, 
        {
            "location": "/fluent/package/#with-vapor", 
            "text": "Fluent comes included with most Vapor templates. However, if you have created a project from scratch you will need to add the provider to your  Package.swift  file.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         . Package ( url :   https://github.com/vapor/vapor.git ,   majorVersion :   2 ), \n         . Package ( url :   https://github.com/vapor/fluent-provider.git ,   majorVersion :   1 ) \n     ], \n     exclude :   [   ...   ]  )   The Fluent provider package adds Fluent to your project and adds some additional, Vapor-specific conveniences like HTTP conformances.   Using  import FluentProvider  will import both Fluent and Fluent's Vapor-specific APIs.", 
            "title": "With Vapor"
        }, 
        {
            "location": "/fluent/package/#without-vapor", 
            "text": "Fluent is a powerful, pure-Swift ORM that can be used with any Server-Side Swift framework. To include it in your package, add it to your  Package.swift  file.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/fluent.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Fluent  to access Fluent's APIs.   Warning  Model  is a Vapor + Fluent type, use  Entity  instead.", 
            "title": "Without Vapor"
        }, 
        {
            "location": "/fluent/package/#drivers", 
            "text": "Fluent drivers allow Fluent models and queries to communicate with various database technologies like MySQL or Mongo. For a full list of drivers, check out the  fluent-driver  tag on GitHub.", 
            "title": "Drivers"
        }, 
        {
            "location": "/fluent/getting-started/", 
            "text": "Getting Started with Fluent\n\n\nFluent provides an easy, simple, and safe API for working with your persisted data. Each database table/collection is represented by a \nModel\n that can be used to interact with the data. Fluent supports common operations like creating, reading, updating, and deleting models. It also supports more advanced operations like joining, relating, and soft deleting. \n\n\n\n\nNote\n\n\nDon't forget to add \nimport FluentProvider\n (or your other database provider) to the top of your Swift files.\n\n\n\n\nFluent ships with SQLite by default. You can use SQLite to quickly scaffolding your application with the in-memory database it provides. This is enabled by default in Vapor's default template. To learn more about configuring your database, check out the available \ndrivers\n.\n\n\nCreating a Model\n\n\nModels are the Swift representations of the data in your database. As such, they are central to most of Fluent's APIs.\n\n\nLet's take a look at what a simple model looks like.\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \nvar\n \nname\n:\n \nString\n\n    \nvar\n \nage\n:\n \nInt\n\n    \nlet\n \nstorage\n \n=\n \nStorage\n()\n\n\n    \ninit\n(\nname\n:\n \nString\n,\n \nage\n:\n \nInt\n)\n \n{\n\n        \nself\n.\nname\n \n=\n \nname\n\n        \nself\n.\nage\n \n=\n \nage\n\n    \n}\n\n\n    \n...\n\n\n}\n\n\n\n\n\n\nHere we are creating a simple class \nPet\n with a name and an age. We will add a simple init method for creating new pets.\n\n\nStorage\n\n\nThe \nstorage\n property is there to allow Fluent to store extra information on your model--things like the model's database id. \n\n\nRow\n\n\nThe \nRow\n struct represents a database row. Your models should be able to parse from and serialize to database rows.\n\n\nParse\n\n\nHere's the code for parsing the Pet from the database.\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \n...\n\n\n    \ninit\n(\nrow\n:\n \nRow\n)\n \nthrows\n \n{\n\n        \nname\n \n=\n \ntry\n \nrow\n.\nget\n(\nname\n)\n\n        \nage\n \n=\n \ntry\n \nrow\n.\nget\n(\nage\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nSerialize\n\n\nHere's the code for serializing the Pet to the database.\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \n...\n\n\n    \nfunc\n \nmakeRow\n()\n \nthrows\n \n-\n \nRow\n \n{\n\n        \nvar\n \nrow\n \n=\n \nRow\n()\n\n        \ntry\n \nrow\n.\nset\n(\nname\n,\n \nname\n)\n\n        \ntry\n \nrow\n.\nset\n(\nage\n,\n \nage\n)\n\n        \nreturn\n \nrow\n\n    \n}\n\n\n}\n\n\n\n\n\n\nPreparing the Database\n\n\nIn order to use your model, you may need to prepare your database with an appropriate schema.\n\n\nPreparation\n\n\nYou can do this by conforming your model to \nPreparation\n.\n\n\nextension\n \nPet\n:\n \nPreparation\n \n{\n\n    \nstatic\n \nfunc\n \nprepare\n(\n_\n \ndatabase\n:\n \nDatabase\n)\n \nthrows\n \n{\n\n        \ntry\n \ndatabase\n.\ncreate\n(\nself\n)\n \n{\n \npets\n \nin\n\n            \npets\n.\nid\n(\nfor\n:\n \nself\n)\n\n            \npets\n.\nstring\n(\nname\n)\n\n            \npets\n.\nint\n(\nage\n)\n\n        \n}\n\n    \n}\n \n\n    \nstatic\n \nfunc\n \nrevert\n(\n_\n \ndatabase\n:\n \nDatabase\n)\n \nthrows\n \n{\n\n        \ntry\n \ndatabase\n.\ndelete\n(\nself\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nHere we are creating a simple table that will look like this:\n\n\n\n\n\n\n\n\nid\n\n\nname\n\n\nage\n\n\n\n\n\n\n\n\n\n\ndatabase id type\n\n\nstring\n\n\nint\n\n\n\n\n\n\n\n\nAdd to Droplet\n\n\nNow you can add your model to the Droplet's prearations so the database is prepared when your application boots.\n\n\nimport\n \nVapor\n\n\nimport\n \nFluentProvider\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\ndrop\n.\npreparations\n.\nappend\n(\nPet\n.\nself\n)\n\n\n\n...\n\n\n\n\n\n\nUsing Models\n\n\nNow that we have created our model and prepared the database, we can use it to save and fetch data from the database.\n\n\nSave\n\n\nTo save a model, call \n.save()\n. A new identifier for the model will automatically be created.\n\n\nlet\n \ndog\n \n=\n \nPet\n(\nname\n:\n \nSpud\n,\n \nage\n:\n \n5\n)\n\n\ntry\n \ndog\n.\nsave\n()\n\n\nprint\n(\ndog\n.\nid\n)\n \n// the newly saved pet\ns id\n\n\n\n\n\n\nFind\n\n\nYou can fetch a model from the database using it's ID.\n\n\nguard\n \nlet\n \ndog\n \n=\n \ntry\n \nPet\n.\nfind\n(\n42\n)\n \nelse\n \n{\n\n    \nthrow\n \nAbort\n.\nnotFound\n\n\n}\n\n\n\nprint\n(\ndog\n.\nname\n)\n \n// the name of the dog with id 42\n\n\n\n\n\n\nFilter\n\n\nYou can also search for models using filters.\n\n\nlet\n \ndogs\n \n=\n \ntry\n \nPet\n.\nmakeQuery\n().\nfilter\n(\nage\n,\n \n.\ngreaterThan\n,\n \n2\n).\nall\n()\n\n\nprint\n(\ndogs\n)\n \n// all dogs older than 2\n\n\n\n\n\n\nDrivers\n\n\nCheck out the \ndatabase\n section for more information about different database drivers you can use with Fluent.", 
            "title": "Getting Started"
        }, 
        {
            "location": "/fluent/getting-started/#getting-started-with-fluent", 
            "text": "Fluent provides an easy, simple, and safe API for working with your persisted data. Each database table/collection is represented by a  Model  that can be used to interact with the data. Fluent supports common operations like creating, reading, updating, and deleting models. It also supports more advanced operations like joining, relating, and soft deleting.    Note  Don't forget to add  import FluentProvider  (or your other database provider) to the top of your Swift files.   Fluent ships with SQLite by default. You can use SQLite to quickly scaffolding your application with the in-memory database it provides. This is enabled by default in Vapor's default template. To learn more about configuring your database, check out the available  drivers .", 
            "title": "Getting Started with Fluent"
        }, 
        {
            "location": "/fluent/getting-started/#creating-a-model", 
            "text": "Models are the Swift representations of the data in your database. As such, they are central to most of Fluent's APIs.  Let's take a look at what a simple model looks like.  final   class   Pet :   Model   { \n     var   name :   String \n     var   age :   Int \n     let   storage   =   Storage () \n\n     init ( name :   String ,   age :   Int )   { \n         self . name   =   name \n         self . age   =   age \n     } \n\n     ...  }   Here we are creating a simple class  Pet  with a name and an age. We will add a simple init method for creating new pets.", 
            "title": "Creating a Model"
        }, 
        {
            "location": "/fluent/getting-started/#storage", 
            "text": "The  storage  property is there to allow Fluent to store extra information on your model--things like the model's database id.", 
            "title": "Storage"
        }, 
        {
            "location": "/fluent/getting-started/#row", 
            "text": "The  Row  struct represents a database row. Your models should be able to parse from and serialize to database rows.", 
            "title": "Row"
        }, 
        {
            "location": "/fluent/getting-started/#parse", 
            "text": "Here's the code for parsing the Pet from the database.  final   class   Pet :   Model   { \n     ... \n\n     init ( row :   Row )   throws   { \n         name   =   try   row . get ( name ) \n         age   =   try   row . get ( age ) \n     }  }", 
            "title": "Parse"
        }, 
        {
            "location": "/fluent/getting-started/#serialize", 
            "text": "Here's the code for serializing the Pet to the database.  final   class   Pet :   Model   { \n     ... \n\n     func   makeRow ()   throws   -   Row   { \n         var   row   =   Row () \n         try   row . set ( name ,   name ) \n         try   row . set ( age ,   age ) \n         return   row \n     }  }", 
            "title": "Serialize"
        }, 
        {
            "location": "/fluent/getting-started/#preparing-the-database", 
            "text": "In order to use your model, you may need to prepare your database with an appropriate schema.", 
            "title": "Preparing the Database"
        }, 
        {
            "location": "/fluent/getting-started/#preparation", 
            "text": "You can do this by conforming your model to  Preparation .  extension   Pet :   Preparation   { \n     static   func   prepare ( _   database :   Database )   throws   { \n         try   database . create ( self )   {   pets   in \n             pets . id ( for :   self ) \n             pets . string ( name ) \n             pets . int ( age ) \n         } \n     }  \n\n     static   func   revert ( _   database :   Database )   throws   { \n         try   database . delete ( self ) \n     }  }   Here we are creating a simple table that will look like this:     id  name  age      database id type  string  int", 
            "title": "Preparation"
        }, 
        {
            "location": "/fluent/getting-started/#add-to-droplet", 
            "text": "Now you can add your model to the Droplet's prearations so the database is prepared when your application boots.  import   Vapor  import   FluentProvider  let   drop   =   try   Droplet ()  drop . preparations . append ( Pet . self )  ...", 
            "title": "Add to Droplet"
        }, 
        {
            "location": "/fluent/getting-started/#using-models", 
            "text": "Now that we have created our model and prepared the database, we can use it to save and fetch data from the database.", 
            "title": "Using Models"
        }, 
        {
            "location": "/fluent/getting-started/#save", 
            "text": "To save a model, call  .save() . A new identifier for the model will automatically be created.  let   dog   =   Pet ( name :   Spud ,   age :   5 )  try   dog . save ()  print ( dog . id )   // the newly saved pet s id", 
            "title": "Save"
        }, 
        {
            "location": "/fluent/getting-started/#find", 
            "text": "You can fetch a model from the database using it's ID.  guard   let   dog   =   try   Pet . find ( 42 )   else   { \n     throw   Abort . notFound  }  print ( dog . name )   // the name of the dog with id 42", 
            "title": "Find"
        }, 
        {
            "location": "/fluent/getting-started/#filter", 
            "text": "You can also search for models using filters.  let   dogs   =   try   Pet . makeQuery (). filter ( age ,   . greaterThan ,   2 ). all ()  print ( dogs )   // all dogs older than 2", 
            "title": "Filter"
        }, 
        {
            "location": "/fluent/getting-started/#drivers", 
            "text": "Check out the  database  section for more information about different database drivers you can use with Fluent.", 
            "title": "Drivers"
        }, 
        {
            "location": "/fluent/model/", 
            "text": "Model\n\n\nModels are the Swift representations of the data in your database. As such, they are central to most of Fluent's APIs. \n\n\nThis guide is an overview of the protocol requirements and methods associated with models.\n\n\n\n\nSeealso\n\n\nCheck out the \ngetting started\n guide for an introductory overview on using models.\n\n\n\n\nCRUD\n\n\nModels have several basic methods for creating, reading, updating, and deleting.\n\n\nSave\n\n\nPersists the entity into the data store and sets the \nid\n property.\n\n\nlet\n \npet\n \n=\n \nPet\n(\nname\n:\n \nSpud\n,\n \nage\n:\n \n2\n)\n\n\ntry\n \npet\n.\nsave\n()\n\n\n\n\n\n\nFind\n\n\nFinds the model with the supplied identifier or returns \nnil\n.\n\n\nguard\n \nlet\n \npet\n \n=\n \ntry\n \nPets\n.\nfind\n(\n42\n)\n \nelse\n \n{\n\n    \nthrow\n \nAbort\n.\nnotFound\n\n\n}\n\n\nprint\n(\npet\n.\nname\n)\n\n\n\n\n\n\nDelete\n\n\nDeletes the entity from the data store if the entity has previously been fetched or saved.\n\n\ntry\n \npet\n.\ndelete\n()\n\n\n\n\n\n\nAll\n\n\nReturns all entities for this model.\n\n\nfor\n \npet\n \nin\n \ntry\n \nPets\n.\nall\n()\n \n{\n\n    \nprint\n(\npet\n.\nname\n)\n\n\n}\n\n\n\n\n\n\nCount\n\n\nReturns a count of all entities for this model.\n\n\nlet\n \ncount\n \n=\n \ntry\n \nPets\n.\ncount\n()\n\n\n\n\n\n\nChunk\n\n\nReturns chunked arrays of a supplied size for all of the entities for this model.\n\n\nThis is a great way to parse through all models of a large data set.\n\n\ntry\n \nPets\n.\nchunk\n(\n20\n)\n \n{\n \npets\n \nin\n\n    \n//\n\n\n}\n\n\n\n\n\n\nQuery\n\n\nCreates a \nQuery\n instance for this \nModel\n.\n\n\nlet\n \nquery\n \n=\n \ntry\n \nPet\n.\nmakeQuery\n()\n\n\n\n\n\n\nTo learn more about crafting complex queries, see the \nquery\n section.\n\n\nTimestamps\n\n\nTo add timestamps to your model, simply conform it to \nTimestampable\n.\n\n\nextension\n \nUser\n:\n \nTimestampable\n \n{\n \n}\n\n\n\n\n\n\nYou can access the updated at and created at times on any model instance.\n\n\nuser\n.\nupdatedAt\n \n// Date?\n\n\nuser\n.\ncreatedAt\n \n// Date?\n\n\n\n\n\n\nWhen filtering or sorting on the timestamp data, you can use the timestamp keys from the class.\n\n\nlet\n \nnewUsers\n \n=\n \ntry\n \nUser\n\n    \n.\nmakeQuery\n()\n\n    \n.\nfilter\n(\nUser\n.\ncreatedAtKey\n,\n \n.\ngreaterThan\n,\n \n...)\n\n    \n.\nall\n()\n\n\n\n\n\n\nYou can also override the timestamp keys if you have custom needs.\n\n\nextension\n \nUser\n:\n \nTimestampable\n \n{\n\n    \nstatic\n \nvar\n \nupdatedAtKey\n:\n \nString\n \n{\n \nreturn\n \ncustom_updated_at\n \n}\n\n    \nstatic\n \nvar\n \ncreatedAtKey\n:\n \nString\n \n{\n \nreturn\n \ncustom_created_at\n \n}\n\n\n}\n\n\n\n\n\n\nMigration\n\n\nTimestampable\n models will automatically have created at and updated at keys added during\n\ndatabase create\n calls.\n\n\nShould you need to manually add \nTimestampable\n to an existing model, you can use the \ndate()\n method\nin a \nmigration\n.\n\n\ndatabase\n.\nmodify\n(\nUser\n.\nself\n)\n \n{\n \nbuilder\n \nin\n\n    \nbuilder\n.\ndate\n(\nUser\n.\ncreatedAtKey\n)\n\n    \nbuilder\n.\ndate\n(\nUser\n.\nupdatedAtKey\n)\n\n\n}\n\n\n\n\n\n\nSoft Delete\n\n\nSoft delete is a way of \"deleting\" a model from all fetch and update queries to Fluent but not actually deleting the model from the database. Soft deleted models can also be restored. \n\n\nTo make your model soft deletable, simply conform it to \nSoftDeletable\n.\n\n\nextension\n \nUser\n:\n \nSoftDeletable\n \n{\n \n}\n\n\n\n\n\n\nOnce your model is soft deletable, all calls to \ndelete()\n will set the deleted at flag instead of actually\ndeleting the model.\n\n\nTo restore a model, call \n.restore()\n. To actually delete a model from the database, call \n.forceDelete()\n.\n\n\nYou can also override the soft delete key if you have custom needs.\n\n\nextension\n \nUser\n:\n \nSoftDeletable\n \n{\n\n    \nstatic\n \nvar\n \ndeletedAtKey\n:\n \nString\n \n{\n \nreturn\n \ncustom_deleted_at\n \n}\n\n\n}\n\n\n\n\n\n\nIncluding Deleted\n\n\nWhen a model is soft deleted, it will be affected by any queries made with the Fluent query builder.\n\n\nTo include soft deleted models, for instance if you want to restore them, use the \n.withSoftDeleted()\n method\non the query builder.\n\n\nlet\n \nallUsers\n \n=\n \ntry\n \nUser\n.\nmakeQuery\n().\nwithSoftDeleted\n().\nall\n()\n\n\n\n\n\n\nLifecycle\n\n\nYou can hook into the soft delete events of a model.\n\n\nextension\n \nUser\n:\n \nSoftDeletable\n \n{\n\n    \nfunc\n \nwillSoftDelete\n()\n \nthrows\n \n{\n \n...\n \n}\n\n    \nfunc\n \ndidSoftDelete\n()\n \n{\n \n...\n \n}\n\n    \nfunc\n \nwillForceDelete\n()\n \nthrows\n \n{\n \n...\n \n}\n\n    \nfunc\n \ndidForceDelete\n()\n \n{\n \n...\n \n}\n\n    \nfunc\n \nwillRestore\n()\n \nthrows\n \n{\n \n...\n \n}\n\n    \nfunc\n \ndidRestore\n()\n \n{\n \n...\n \n}\n\n\n}\n\n\n\n\n\n\n\n\nNote\n\n\nThrowing during a \nwill\n hook will prevent the action from happening.\n\n\n\n\nMigration\n\n\nSoftDeletable\n models will automatically have a deleted at key added during\n\ndatabase create\n calls.\n\n\nShould you need to manually add \nSoftDeletable\n to an existing model, you can use the \ndate()\n method\nin a \nmigration\n.\n\n\ndatabase\n.\nmodify\n(\nUser\n.\nself\n)\n \n{\n \nbuilder\n \nin\n\n    \nbuilder\n.\ndate\n(\nUser\n.\ndeletedAtKey\n,\n \noptional\n:\n \ntrue\n)\n\n\n}\n\n\n\n\n\n\nConvenience\n\n\nAssert Exists\n\n\nThe identifier property of a model is optional since models may not have been saved yet.\n\n\nYou can get the identifier or throw an error if the model has not been saved yet by calling \nassertExists()\n.\n\n\nlet\n \nid\n \n=\n \ntry\n \npet\n.\nassertExists\n()\n\n\nprint\n(\nid\n)\n \n// not optional\n\n\n\n\n\n\nLife Cycle\n\n\nThe following life-cycle methods can be implemented on your model to hook into internal operations.\n\n\n/// Called before the entity will be created.\n\n\n/// Throwing will cancel the creation.\n\n\nfunc\n \nwillCreate\n()\n \nthrows\n\n\n\n/// Called after the entity has been created.\n\n\nfunc\n \ndidCreate\n()\n\n\n\n/// Called before the entity will be updated.\n\n\n/// Throwing will cancel the update.\n\n\nfunc\n \nwillUpdate\n()\n \nthrows\n\n\n\n/// Called after the entity has been updated.\n\n\nfunc\n \ndidUpdate\n()\n\n\n\n/// Called before the entity will be deleted.\n\n\n/// Throwing will cancel the deletion.\n\n\nfunc\n \nwillDelete\n()\n \nthrows\n\n\n\n/// Called after the entity has been deleted.\n\n\nfunc\n \ndidDelete\n()\n\n\n\n\n\n\n\n\nNote\n\n\nThrowing in a \nwillFoo()\n method will cancel the operation.\n\n\n\n\nHere's an example of implementing the \ndidDelete\n method.\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \n...\n \n\n    \nfunc\n \ndidDelete\n()\n \n{\n\n        \nprint\n(\nDeleted \n\\(\nname\n)\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nEntity\n\n\nEntity is the base Fluent protocol that Model conforms to. It is responsible for providing all information the \ndatabase or query may need when saving, fetching, or deleting your models.\n\n\nName\n\n\nThe singular relational name of this model. Also used for internal storage. Example: Pet = \"pet\".\n\n\nThis value should usually not be overriden. \n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \nstatic\n \nlet\n \nname\n \n=\n \npet\n\n\n}\n\n\n\n\n\n\nEntity\n\n\nThe plural relational name of this model. Used as the collection or table name. \n\n\nExample: Pet = \"pets\".\n\n\nThis value should be overriden if the table name for your model is non-standard.\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \nstatic\n \nlet\n \nentity\n \n=\n \npets\n\n\n}\n\n\n\n\n\n\nID Type\n\n\nThe type of identifier used for both the local and foreign id keys. \n\n\nExample: uuid, integer, etc.\n\n\nThis value should be overriden if a particular model in your database uses a different ID type.\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \nstatic\n \nlet\n \nidType\n \n=\n \n.\nuuid\n\n\n}\n\n\n\n\n\n\nThis can also be overridden at the database level using config.\n\n\nConfig/fluent.json\n\n\n{\n\n    \nidType\n:\n \nuuid\n\n\n}\n\n\n\n\n\n\nOr programatically.\n\n\ndrop\n.\ndatabase\n?.\nidType\n \n=\n \n.\nuuid\n\n\n\n\n\n\nKey Naming Convention\n\n\nThe naming convetion to use for foreign id keys, table names, etc.\n\n\nExample: snake_case vs. camelCase.\n\n\nThis value should be overridden if a particular model in your database uses a different key naming convention.\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \nstatic\n \nlet\n \nkeyNamingConvention\n \n=\n \n.\nsnake_case\n\n\n}\n\n\n\n\n\n\nThis can also be overridden at the database level using config.\n\n\nConfig/fluent.json\n\n\n{\n\n    \nkeyNamingConvention\n:\n \nsnake_case\n\n\n}\n\n\n\n\n\n\nOr programatically.\n\n\ndrop\n.\ndatabase\n?.\nkeyNamingConvention\n \n=\n \n.\nsnake_case\n\n\n\n\n\n\nID Key\n\n\nThe name of the column that corresponds to this entity's identifying key.\n\n\nThe default is 'database.driver.idKey', and then \"id\"\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \nstatic\n \nlet\n \nidKey\n \n=\n \nid\n\n\n}\n\n\n\n\n\n\nForeign ID Key\n\n\nThe name of the column that points to this entity's id when referenced from other tables or collections.\n\n\nExample: \"foo_id\".\n\n\nfinal\n \nclass\n \nPet\n:\n \nModel\n \n{\n\n    \nstatic\n \nlet\n \nforeignIdKey\n \n=\n \npet_id\n\n\n}", 
            "title": "Model"
        }, 
        {
            "location": "/fluent/model/#model", 
            "text": "Models are the Swift representations of the data in your database. As such, they are central to most of Fluent's APIs.   This guide is an overview of the protocol requirements and methods associated with models.   Seealso  Check out the  getting started  guide for an introductory overview on using models.", 
            "title": "Model"
        }, 
        {
            "location": "/fluent/model/#crud", 
            "text": "Models have several basic methods for creating, reading, updating, and deleting.", 
            "title": "CRUD"
        }, 
        {
            "location": "/fluent/model/#save", 
            "text": "Persists the entity into the data store and sets the  id  property.  let   pet   =   Pet ( name :   Spud ,   age :   2 )  try   pet . save ()", 
            "title": "Save"
        }, 
        {
            "location": "/fluent/model/#find", 
            "text": "Finds the model with the supplied identifier or returns  nil .  guard   let   pet   =   try   Pets . find ( 42 )   else   { \n     throw   Abort . notFound  }  print ( pet . name )", 
            "title": "Find"
        }, 
        {
            "location": "/fluent/model/#delete", 
            "text": "Deletes the entity from the data store if the entity has previously been fetched or saved.  try   pet . delete ()", 
            "title": "Delete"
        }, 
        {
            "location": "/fluent/model/#all", 
            "text": "Returns all entities for this model.  for   pet   in   try   Pets . all ()   { \n     print ( pet . name )  }", 
            "title": "All"
        }, 
        {
            "location": "/fluent/model/#count", 
            "text": "Returns a count of all entities for this model.  let   count   =   try   Pets . count ()", 
            "title": "Count"
        }, 
        {
            "location": "/fluent/model/#chunk", 
            "text": "Returns chunked arrays of a supplied size for all of the entities for this model.  This is a great way to parse through all models of a large data set.  try   Pets . chunk ( 20 )   {   pets   in \n     //  }", 
            "title": "Chunk"
        }, 
        {
            "location": "/fluent/model/#query", 
            "text": "Creates a  Query  instance for this  Model .  let   query   =   try   Pet . makeQuery ()   To learn more about crafting complex queries, see the  query  section.", 
            "title": "Query"
        }, 
        {
            "location": "/fluent/model/#timestamps", 
            "text": "To add timestamps to your model, simply conform it to  Timestampable .  extension   User :   Timestampable   {   }   You can access the updated at and created at times on any model instance.  user . updatedAt   // Date?  user . createdAt   // Date?   When filtering or sorting on the timestamp data, you can use the timestamp keys from the class.  let   newUsers   =   try   User \n     . makeQuery () \n     . filter ( User . createdAtKey ,   . greaterThan ,   ...) \n     . all ()   You can also override the timestamp keys if you have custom needs.  extension   User :   Timestampable   { \n     static   var   updatedAtKey :   String   {   return   custom_updated_at   } \n     static   var   createdAtKey :   String   {   return   custom_created_at   }  }", 
            "title": "Timestamps"
        }, 
        {
            "location": "/fluent/model/#migration", 
            "text": "Timestampable  models will automatically have created at and updated at keys added during database create  calls.  Should you need to manually add  Timestampable  to an existing model, you can use the  date()  method\nin a  migration .  database . modify ( User . self )   {   builder   in \n     builder . date ( User . createdAtKey ) \n     builder . date ( User . updatedAtKey )  }", 
            "title": "Migration"
        }, 
        {
            "location": "/fluent/model/#soft-delete", 
            "text": "Soft delete is a way of \"deleting\" a model from all fetch and update queries to Fluent but not actually deleting the model from the database. Soft deleted models can also be restored.   To make your model soft deletable, simply conform it to  SoftDeletable .  extension   User :   SoftDeletable   {   }   Once your model is soft deletable, all calls to  delete()  will set the deleted at flag instead of actually\ndeleting the model.  To restore a model, call  .restore() . To actually delete a model from the database, call  .forceDelete() .  You can also override the soft delete key if you have custom needs.  extension   User :   SoftDeletable   { \n     static   var   deletedAtKey :   String   {   return   custom_deleted_at   }  }", 
            "title": "Soft Delete"
        }, 
        {
            "location": "/fluent/model/#including-deleted", 
            "text": "When a model is soft deleted, it will be affected by any queries made with the Fluent query builder.  To include soft deleted models, for instance if you want to restore them, use the  .withSoftDeleted()  method\non the query builder.  let   allUsers   =   try   User . makeQuery (). withSoftDeleted (). all ()", 
            "title": "Including Deleted"
        }, 
        {
            "location": "/fluent/model/#lifecycle", 
            "text": "You can hook into the soft delete events of a model.  extension   User :   SoftDeletable   { \n     func   willSoftDelete ()   throws   {   ...   } \n     func   didSoftDelete ()   {   ...   } \n     func   willForceDelete ()   throws   {   ...   } \n     func   didForceDelete ()   {   ...   } \n     func   willRestore ()   throws   {   ...   } \n     func   didRestore ()   {   ...   }  }    Note  Throwing during a  will  hook will prevent the action from happening.", 
            "title": "Lifecycle"
        }, 
        {
            "location": "/fluent/model/#migration_1", 
            "text": "SoftDeletable  models will automatically have a deleted at key added during database create  calls.  Should you need to manually add  SoftDeletable  to an existing model, you can use the  date()  method\nin a  migration .  database . modify ( User . self )   {   builder   in \n     builder . date ( User . deletedAtKey ,   optional :   true )  }", 
            "title": "Migration"
        }, 
        {
            "location": "/fluent/model/#convenience", 
            "text": "", 
            "title": "Convenience"
        }, 
        {
            "location": "/fluent/model/#assert-exists", 
            "text": "The identifier property of a model is optional since models may not have been saved yet.  You can get the identifier or throw an error if the model has not been saved yet by calling  assertExists() .  let   id   =   try   pet . assertExists ()  print ( id )   // not optional", 
            "title": "Assert Exists"
        }, 
        {
            "location": "/fluent/model/#life-cycle", 
            "text": "The following life-cycle methods can be implemented on your model to hook into internal operations.  /// Called before the entity will be created.  /// Throwing will cancel the creation.  func   willCreate ()   throws  /// Called after the entity has been created.  func   didCreate ()  /// Called before the entity will be updated.  /// Throwing will cancel the update.  func   willUpdate ()   throws  /// Called after the entity has been updated.  func   didUpdate ()  /// Called before the entity will be deleted.  /// Throwing will cancel the deletion.  func   willDelete ()   throws  /// Called after the entity has been deleted.  func   didDelete ()    Note  Throwing in a  willFoo()  method will cancel the operation.   Here's an example of implementing the  didDelete  method.  final   class   Pet :   Model   { \n     ...  \n\n     func   didDelete ()   { \n         print ( Deleted  \\( name ) ) \n     }  }", 
            "title": "Life Cycle"
        }, 
        {
            "location": "/fluent/model/#entity", 
            "text": "Entity is the base Fluent protocol that Model conforms to. It is responsible for providing all information the \ndatabase or query may need when saving, fetching, or deleting your models.", 
            "title": "Entity"
        }, 
        {
            "location": "/fluent/model/#name", 
            "text": "The singular relational name of this model. Also used for internal storage. Example: Pet = \"pet\".  This value should usually not be overriden.   final   class   Pet :   Model   { \n     static   let   name   =   pet  }", 
            "title": "Name"
        }, 
        {
            "location": "/fluent/model/#entity_1", 
            "text": "The plural relational name of this model. Used as the collection or table name.   Example: Pet = \"pets\".  This value should be overriden if the table name for your model is non-standard.  final   class   Pet :   Model   { \n     static   let   entity   =   pets  }", 
            "title": "Entity"
        }, 
        {
            "location": "/fluent/model/#id-type", 
            "text": "The type of identifier used for both the local and foreign id keys.   Example: uuid, integer, etc.  This value should be overriden if a particular model in your database uses a different ID type.  final   class   Pet :   Model   { \n     static   let   idType   =   . uuid  }   This can also be overridden at the database level using config.  Config/fluent.json  { \n     idType :   uuid  }   Or programatically.  drop . database ?. idType   =   . uuid", 
            "title": "ID Type"
        }, 
        {
            "location": "/fluent/model/#key-naming-convention", 
            "text": "The naming convetion to use for foreign id keys, table names, etc.  Example: snake_case vs. camelCase.  This value should be overridden if a particular model in your database uses a different key naming convention.  final   class   Pet :   Model   { \n     static   let   keyNamingConvention   =   . snake_case  }   This can also be overridden at the database level using config.  Config/fluent.json  { \n     keyNamingConvention :   snake_case  }   Or programatically.  drop . database ?. keyNamingConvention   =   . snake_case", 
            "title": "Key Naming Convention"
        }, 
        {
            "location": "/fluent/model/#id-key", 
            "text": "The name of the column that corresponds to this entity's identifying key.  The default is 'database.driver.idKey', and then \"id\"  final   class   Pet :   Model   { \n     static   let   idKey   =   id  }", 
            "title": "ID Key"
        }, 
        {
            "location": "/fluent/model/#foreign-id-key", 
            "text": "The name of the column that points to this entity's id when referenced from other tables or collections.  Example: \"foo_id\".  final   class   Pet :   Model   { \n     static   let   foreignIdKey   =   pet_id  }", 
            "title": "Foreign ID Key"
        }, 
        {
            "location": "/fluent/database/", 
            "text": "Database\n\n\nA Fluent database is responsible for managing connections to your underlying data store and sending queries to the implementation-specific driver you have chosen.\n\n\nDrivers\n\n\nBy default, Fluent includes in-memory and SQLite drivers. There are several drivers available to add to your Vapor application.\n\n\nAvailable\n\n\n\n\n\n\n\n\nType\n\n\nKey\n\n\nPackage\n\n\nClass\n\n\nOfficial\n\n\n\n\n\n\n\n\n\n\nMemory\n\n\nmemory\n\n\nFluent Provider\n\n\nFluent.MemoryDriver\n\n\nYes\n\n\n\n\n\n\nSQlite\n\n\nsqlite\n\n\nFluent Provider\n\n\nFluent.SQLiteDriver\n\n\nYes\n\n\n\n\n\n\nMySQL\n\n\nmysql\n\n\nMySQLProvider\n\n\nMySQLDriver.Driver\n\n\nYes\n\n\n\n\n\n\nPostgreSQL\n\n\nN/A\n\n\nPostgreSQLProvider\n\n\nN/A\n\n\nNo\n\n\n\n\n\n\nMongoDB\n\n\nN/A\n\n\nMongoProvider\n\n\nN/A\n\n\nNo\n\n\n\n\n\n\n\n\nClick on the provider package for more information about how to use it.\n\n\nYou can search for a list of available \nVapor database providers\n on GitHub.\n\n\nDroplet\n\n\nYou can access the database from the Droplet.\n\n\ndrop\n.\ndatabase\n \n// Database?\n\n\n\n\n\n\nPreparations\n\n\nMost databases, like SQL databases, require the schema for a model to be created before it is stored. \nAdding a preparation to your model will allow you to prepare the database while your app boots.\n\n\nextension\n \nUser\n:\n \nPreparation\n \n{\n\n    \n/// Prepares a table/collection in the database\n\n    \n/// for storing Users\n\n    \nstatic\n \nfunc\n \nprepare\n(\n_\n \ndatabase\n:\n \nDatabase\n)\n \nthrows\n \n{\n\n        \ntry\n \ndatabase\n.\ncreate\n(\nself\n)\n \n{\n \nbuilder\n \nin\n\n            \nbuilder\n.\nid\n()\n\n            \nbuilder\n.\nstring\n(\nname\n)\n\n            \nbuilder\n.\nint\n(\nage\n)\n\n        \n}\n\n    \n}\n\n\n    \n/// Undoes what was done in `prepare`\n\n    \nstatic\n \nfunc\n \nrevert\n(\n_\n \ndatabase\n:\n \nDatabase\n)\n \nthrows\n \n{\n\n        \ntry\n \ndatabase\n.\ndelete\n(\nself\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nThe above prepare statement results in SQL similar to the following:\n\n\nCREATE\n \nTABLE\n \n`\nusers\n`\n \n(\n`\nid\n`\n \nINTEGER\n \nPRIMARY\n \nKEY\n \nNOT\n \nNULL\n,\n \n`\nname\n`\n \nTEXT\n \nNOT\n \nNULL\n,\n \n`\nage\n`\n \nINTEGER\n \nNOT\n \nNULL\n)\n\n\n\n\n\n\nOnce you have created you preparation, add it to the Droplet's prepratations array.\n\n\ndrop\n.\npreparations\n.\nappend\n(\nUser\n.\nself\n)\n\n\n\n\n\n\nCreate\n\n\nThe following methods are available on while creating and modifing the database.\n\n\n\n\n\n\n\n\nMethod\n\n\nType\n\n\n\n\n\n\n\n\n\n\nid\n\n\nPrimary Identifier\n\n\n\n\n\n\nforeignId\n\n\nForeign Identifier\n\n\n\n\n\n\nint\n\n\nInteger\n\n\n\n\n\n\nstring\n\n\nString\n\n\n\n\n\n\ndouble\n\n\nDouble\n\n\n\n\n\n\nbool\n\n\nBoolean\n\n\n\n\n\n\nbytes\n\n\nData\n\n\n\n\n\n\ndate\n\n\nDate + Time\n\n\n\n\n\n\n\n\nYou can use any of these methods on a builder inside \n.create()\n.\n\n\ntry\n \ndatabase\n.\ncreate\n(\nself\n)\n \n{\n \nbuilder\n \nin\n\n    \nbuilder\n.\ndouble\n(\nlatitude\n)\n\n    \nbuilder\n.\ndouble\n(\nlongitude\n)\n\n\n}\n\n\n\n\n\n\nForeign Keys\n\n\nForeign keys are automatically added with \n.foreignId()\n. To add a foreign key manually, use the\n\n.foreignKey\n method.\n\n\ntry\n \ndatabase\n.\ncreate\n(\nself\n)\n \n{\n \nbuilder\n \nin\n\n    \nbuilder\n.\nforeignKey\n(\nuser_id\n,\n \nreferences\n:\n \nid\n,\n \non\n:\n \nUser\n.\nself\n)\n\n\n}\n\n\n\n\n\n\nTo disable automatic foreign keys, set \nautoForeignKeys\n to false in the \nConfig/fluent.json\n file.\n\n\n{\n\n    \nautoForeignKeys\n:\n \nfalse\n\n\n}\n\n\n\n\n\n\nModifier\n\n\nExisting schema can be modified using the \n.modify()\n property. All of the methods from \n.create()\n\nare available here as well.\n\n\ntry\n \ndatabase\n.\nmodify\n(\nself\n)\n \n{\n \nbuilder\n \nin\n\n    \nbuilder\n.\nstring\n(\nname\n)\n\n    \nbuilder\n.\ndelete\n(\nage\n)\n\n\n}\n\n\n\n\n\n\nMigrations\n\n\nOther times, you may want to make some modifications to your data set while migrating to a new version or\njust performing general cleanup. \n\n\nstruct\n \nDeleteOldEntries\n:\n \nPreparation\n \n{\n\n    \nstatic\n \nfunc\n \nprepare\n(\n_\n \ndatabase\n:\n \nDatabase\n)\n \nthrows\n \n{\n\n        \ntry\n \nLog\n.\nmakeQuery\n().\nfilter\n(...).\ndelete\n()\n\n    \n}\n\n\n    \n...\n\n\n}\n\n\n\n\n\n\nRun\n\n\nYour preparations will run every time you run your application. You can run your preparations without booting\nyour server by calling:\n\n\nvapor run prepare\n\n\n\n\n\nRevert\n\n\nUse the revert method to undo any work you did in the prepare method. \n\n\nextension\n \nUser\n:\n \nPreparation\n \n{\n\n    \n...\n\n\n\n    \nstatic\n \nfunc\n \nrevert\n(\n_\n \ndatabase\n:\n \nDatabase\n)\n \nthrows\n \n{\n\n        \ntry\n \ndatabase\n.\ndelete\n(\nself\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nYou can run the reversions by calling:\n\n\nvapor run prepare --revert\n\n\n\n\n\nThis will revert the latest batch of preparations. To revert the entire database, run the following:\n\n\nvapor run prepare --revert --all\n\n\n\n\n\nLog\n\n\nLogging queries is a great way to find optimizations for your application and track down bugs.\n\n\ndrop\n.\ndatabase\n?.\nlog\n \n=\n \n{\n \nquery\n \nin\n\n    \nprint\n(\nquery\n)\n\n\n}\n\n\n\n\n\n\nYou can assign a closure to the \nlog\n property on the database. Any time a query is run, the closure\nwill be called with a \nQueryLog\n object containing a string describing the statement and the time it ran.\n\n\nTransactions\n\n\nTransactions allow you to group multiple queries into one single unit of work. If any one of the \nqueries experiences a problem, the entire transaction will be rolled back.\n\n\ndrop\n.\ndatabase\n?.\ntransaction\n \n{\n \nconn\n \nin\n\n    \ntry\n \nuser\n.\npets\n.\nmakeQuery\n(\nconn\n).\ndelete\n()\n\n    \ntry\n \nuser\n.\nmakeQuery\n(\nconn\n).\ndelete\n()\n\n\n}\n\n\n\n\n\n\nDrivers that do not support transactions will throw an error if this method is called.\n\n\nYou can use the \n.makeQuery(_: Executor)\n method to create queries that will run on the \nconnection supplied to the closure.\n\n\n\n\nWarning\n\n\nYou must use the connection supplied to the closure for queries you want to include\nin the transaction.\n\n\n\n\nIndexes\n\n\nAn index is a copy of selected columns of data from a table that can be searched very efficiently.\n\n\nYou can add them to your database by calling \n.index()\n\n\ntry\n \ndatabase\n.\nindex\n(\nname\n,\n \nfor\n:\n \nUser\n.\nself\n)\n\n\n\n\n\n\nYou can delete them by calling \n.deleteIndex()\n\n\ntry\n \ndatabase\n.\ndeleteIndex\n(\nname\n,\n \nfor\n:\n \nUser\n.\nself\n)", 
            "title": "Database"
        }, 
        {
            "location": "/fluent/database/#database", 
            "text": "A Fluent database is responsible for managing connections to your underlying data store and sending queries to the implementation-specific driver you have chosen.", 
            "title": "Database"
        }, 
        {
            "location": "/fluent/database/#drivers", 
            "text": "By default, Fluent includes in-memory and SQLite drivers. There are several drivers available to add to your Vapor application.", 
            "title": "Drivers"
        }, 
        {
            "location": "/fluent/database/#available", 
            "text": "Type  Key  Package  Class  Official      Memory  memory  Fluent Provider  Fluent.MemoryDriver  Yes    SQlite  sqlite  Fluent Provider  Fluent.SQLiteDriver  Yes    MySQL  mysql  MySQLProvider  MySQLDriver.Driver  Yes    PostgreSQL  N/A  PostgreSQLProvider  N/A  No    MongoDB  N/A  MongoProvider  N/A  No     Click on the provider package for more information about how to use it.  You can search for a list of available  Vapor database providers  on GitHub.", 
            "title": "Available"
        }, 
        {
            "location": "/fluent/database/#droplet", 
            "text": "You can access the database from the Droplet.  drop . database   // Database?", 
            "title": "Droplet"
        }, 
        {
            "location": "/fluent/database/#preparations", 
            "text": "Most databases, like SQL databases, require the schema for a model to be created before it is stored. \nAdding a preparation to your model will allow you to prepare the database while your app boots.  extension   User :   Preparation   { \n     /// Prepares a table/collection in the database \n     /// for storing Users \n     static   func   prepare ( _   database :   Database )   throws   { \n         try   database . create ( self )   {   builder   in \n             builder . id () \n             builder . string ( name ) \n             builder . int ( age ) \n         } \n     } \n\n     /// Undoes what was done in `prepare` \n     static   func   revert ( _   database :   Database )   throws   { \n         try   database . delete ( self ) \n     }  }   The above prepare statement results in SQL similar to the following:  CREATE   TABLE   ` users `   ( ` id `   INTEGER   PRIMARY   KEY   NOT   NULL ,   ` name `   TEXT   NOT   NULL ,   ` age `   INTEGER   NOT   NULL )   Once you have created you preparation, add it to the Droplet's prepratations array.  drop . preparations . append ( User . self )", 
            "title": "Preparations"
        }, 
        {
            "location": "/fluent/database/#create", 
            "text": "The following methods are available on while creating and modifing the database.     Method  Type      id  Primary Identifier    foreignId  Foreign Identifier    int  Integer    string  String    double  Double    bool  Boolean    bytes  Data    date  Date + Time     You can use any of these methods on a builder inside  .create() .  try   database . create ( self )   {   builder   in \n     builder . double ( latitude ) \n     builder . double ( longitude )  }", 
            "title": "Create"
        }, 
        {
            "location": "/fluent/database/#foreign-keys", 
            "text": "Foreign keys are automatically added with  .foreignId() . To add a foreign key manually, use the .foreignKey  method.  try   database . create ( self )   {   builder   in \n     builder . foreignKey ( user_id ,   references :   id ,   on :   User . self )  }   To disable automatic foreign keys, set  autoForeignKeys  to false in the  Config/fluent.json  file.  { \n     autoForeignKeys :   false  }", 
            "title": "Foreign Keys"
        }, 
        {
            "location": "/fluent/database/#modifier", 
            "text": "Existing schema can be modified using the  .modify()  property. All of the methods from  .create() \nare available here as well.  try   database . modify ( self )   {   builder   in \n     builder . string ( name ) \n     builder . delete ( age )  }", 
            "title": "Modifier"
        }, 
        {
            "location": "/fluent/database/#migrations", 
            "text": "Other times, you may want to make some modifications to your data set while migrating to a new version or\njust performing general cleanup.   struct   DeleteOldEntries :   Preparation   { \n     static   func   prepare ( _   database :   Database )   throws   { \n         try   Log . makeQuery (). filter (...). delete () \n     } \n\n     ...  }", 
            "title": "Migrations"
        }, 
        {
            "location": "/fluent/database/#run", 
            "text": "Your preparations will run every time you run your application. You can run your preparations without booting\nyour server by calling:  vapor run prepare", 
            "title": "Run"
        }, 
        {
            "location": "/fluent/database/#revert", 
            "text": "Use the revert method to undo any work you did in the prepare method.   extension   User :   Preparation   { \n     ... \n\n\n     static   func   revert ( _   database :   Database )   throws   { \n         try   database . delete ( self ) \n     }  }   You can run the reversions by calling:  vapor run prepare --revert  This will revert the latest batch of preparations. To revert the entire database, run the following:  vapor run prepare --revert --all", 
            "title": "Revert"
        }, 
        {
            "location": "/fluent/database/#log", 
            "text": "Logging queries is a great way to find optimizations for your application and track down bugs.  drop . database ?. log   =   {   query   in \n     print ( query )  }   You can assign a closure to the  log  property on the database. Any time a query is run, the closure\nwill be called with a  QueryLog  object containing a string describing the statement and the time it ran.", 
            "title": "Log"
        }, 
        {
            "location": "/fluent/database/#transactions", 
            "text": "Transactions allow you to group multiple queries into one single unit of work. If any one of the \nqueries experiences a problem, the entire transaction will be rolled back.  drop . database ?. transaction   {   conn   in \n     try   user . pets . makeQuery ( conn ). delete () \n     try   user . makeQuery ( conn ). delete ()  }   Drivers that do not support transactions will throw an error if this method is called.  You can use the  .makeQuery(_: Executor)  method to create queries that will run on the \nconnection supplied to the closure.   Warning  You must use the connection supplied to the closure for queries you want to include\nin the transaction.", 
            "title": "Transactions"
        }, 
        {
            "location": "/fluent/database/#indexes", 
            "text": "An index is a copy of selected columns of data from a table that can be searched very efficiently.  You can add them to your database by calling  .index()  try   database . index ( name ,   for :   User . self )   You can delete them by calling  .deleteIndex()  try   database . deleteIndex ( name ,   for :   User . self )", 
            "title": "Indexes"
        }, 
        {
            "location": "/fluent/query/", 
            "text": "Query\n\n\nFluent's query builder provides a simple interface for creating complex database queries. The \nQuery\n class itself (raw queries excluded) is the sole method by which Fluent communicates with your database.\n\n\nMake\n\n\nYou can create a new query builder from any model class.\n\n\nlet\n \nquery\n \n=\n \ntry\n \nPost\n.\nmakeQuery\n()\n\n\n\n\n\n\nYou can also create queries from an instance. This is especially useful if you need to use a special\ndatabase connection (like for \ntransactions\n) to save or update a model.\n\n\nguard\n \nlet\n \npost\n \n=\n \ntry\n \nPost\n.\nfind\n(\n42\n)\n \nelse\n \n{\n \n...\n \n}\n\n\npost\n.\ncontent\n \n=\n \nUpdated\n\n\nlet\n \nquery\n \n=\n \ntry\n \npost\n.\nmakeQuery\n(\nconn\n).\nsave\n()\n\n\n\n\n\n\nFetch\n\n\nYou have multiple options for fetching the results of a query.\n\n\nAll\n\n\nThe simplest option, \n.all()\n returns all rows relevant to the query.\n\n\nlet users = try User.makeQuery().filter(...).all()\n\n\n\n\n\nFirst\n\n\nYou can take only the first row as well with \n.first()\n. \n\n\nlet user = try User.makeQuery().filter(...).first()\n\n\n\n\n\nFluent will automatically limit the results to \n1\n to increase\nthe performance of the query.\n\n\nChunk\n\n\nIf you want to fetch a large amount of models from the database, using \n.chunk()\n can help reduce the\namount of memory required for the query by fetching chunks of data at a time.\n\n\nUser.makeQuery().filter(...).chunk(32) { users in\n    print(users)\n}\n\n\n\n\n\nFilter\n\n\nFilters allow you to choose exactly what subset of data you want to modify or fetch. There are three different\ntypes of filters.\n\n\nCompare\n\n\nCompare filters perform a comparison between a field on your model in the database and a supplied value.\n\n\ntry\n \nquery\n.\nfilter\n(\nage\n,\n \n.\ngreaterThanOrEquals\n,\n \n21\n)\n\n\n\n\n\n\nYou can also use operators.\n\n\ntry\n \nquery\n.\nfilter\n(\nage\n \n=\n \n21\n)\n\n\n\n\n\n\n\n\n\n\n\n\nCase\n\n\nOperator\n\n\nType\n\n\n\n\n\n\n\n\n\n\n.equals\n\n\n==\n\n\nEquals\n\n\n\n\n\n\n.greaterThan\n\n\n\n\nGreater Than\n\n\n\n\n\n\n.lessThan\n\n\n\n\nLess Than\n\n\n\n\n\n\n.greaterThanOrEquals\n\n\n=\n\n\nGreater Than Or Equals\n\n\n\n\n\n\n.lessThanOrEquals\n\n\n=\n\n\nLess Than Or Equals\n\n\n\n\n\n\n.notEquals\n\n\n!=\n\n\nNot Equals\n\n\n\n\n\n\n.hasSuffix\n\n\n\n\nHas Suffix\n\n\n\n\n\n\n.hasPrefix\n\n\n\n\nHas Prefix\n\n\n\n\n\n\n.contains\n\n\n\n\nContains\n\n\n\n\n\n\n.custom(String)\n\n\n\n\nCustom\n\n\n\n\n\n\n\n\n\n\nTip\n\n\nYou can omit the comparison type for \n.equals\n, e.g., \nquery.filter(\"age\", 23)\n\n\n\n\nSubset\n\n\nYou can also filter by fields being in a set of data.\n\n\ntry\n \nquery\n.\nfilter\n(\nfavoriteColor\n,\n \nin\n:\n \n[\npink\n,\n \nblue\n])\n\n\n\n\n\n\nOr the opposite.\n\n\ntry\n \nquery\n.\nfilter\n(\nfavoriteColor\n,\n \nnotIn\n:\n \n[\nbrown\n,\n \nblack\n])\n\n\n\n\n\n\nGroup\n\n\nBy default, all query filters are joined by AND logic. You can create groups of filters within\nyour query that are joined with AND or OR logic.\n\n\ntry\n \nquery\n.\nor\n \n{\n \norGroup\n \nin\n\n    \ntry\n \norGroup\n.\nfilter\n(\nage\n,\n \n.\ngreaterThan\n,\n \n75\n)\n\n    \ntry\n \norGroup\n.\nfilter\n(\nage\n,\n \n.\nlessThan\n,\n \n18\n)\n\n\n}\n\n\n\n\n\n\nThis will result in SQL similar to the following:\n\n\nSELECT\n \n*\n \nFROM\n \n`\nusers\n`\n \nWHERE\n \n(\n`\nage\n`\n \n \n75\n \nOR\n \n`\nage\n`\n \n \n18\n);\n\n\n\n\n\n\n.and()\n is also available in case you need to switch back to joining filters with AND nested inside of an OR.\n\n\nComplex Example\n\n\nlet\n \nusers\n \n=\n \ntry\n \nUser\n\n    \n.\nmakeQuery\n()\n\n    \n.\nfilter\n(\nplanetOfOrigin\n,\n \n.\ngreaterThan\n,\n \nEarth\n)\n\n    \n.\nor\n \n{\n \norGroup\n \nin\n\n        \norGroup\n.\nand\n \n{\n \nandGroup\n \nin\n\n            \nandGroup\n.\nfilter\n(\nname\n,\n \nRick\n)\n\n            \nandGroup\n.\nfilter\n(\nfavoriteFood\n,\n \nBeer\n)\n\n        \n}\n\n        \norGroup\n.\nand\n \n{\n \nandGroup\n \nin\n\n            \nandGroup\n.\nfilter\n(\nname\n,\n \nMorty\n)\n\n            \nandGroup\n.\nfilter\n(\nfavoriteFood\n,\n \nEyeholes\n)\n\n        \n}\n\n    \n}\n\n    \n.\nall\n()\n\n\n\n\n\n\nThis will result in SQL similar to the following:\n\n\nSELECT\n \n*\n \nFROM\n \n`\nusers\n`\n \n    \nWHERE\n \n`\nplanetOfOrigin\n`\n \n=\n \nEarth\n \nAND\n \n(\n\n           \n(`\nname\n`\n \n=\n \nRick\n \nAND\n \n`\nfavoriteFood\n`\n \n=\n \nBeer\n)\n\n        \nOR\n \n(`\nname\n`\n \n=\n \nMorty\n \nAND\n \n`\nfavoriteFood\n`\n \n=\n \nEyeholes\n)\n\n    \n)\n\n\n\n\n\n\n\n\nNote\n\n\nKeep in mind that the AND/OR logic for a group applies only to the filters added \nwithin\n the group. All filters outside of a filter group will be joined by AND.\n\n\n\n\nRaw\n\n\nRaw filters can be used to filter by values that should not be parameterized.\n\n\ntry\n \nquery\n.\nfilter\n(\nraw\n:\n \ndate \n= CURRENT_TIMESTAMP\n)\n\n\n\n\n\n\nDistinct\n\n\nTo select only distinct models from the database, add \n.distinct()\n to your query.\n\n\ntry\n \nquery\n.\ndistinct\n()\n\n\n\n\n\n\nLimit / Offset\n\n\nTo limit or offset your query, use the \n.limit()\n method.\n\n\ntry\n \nquery\n.\nlimit\n(\n20\n,\n \noffset\n:\n \n5\n)\n\n\n\n\n\n\nSort\n\n\nTo sort the results of your query, use the \n.sort()\n method\n\n\ntry\n \nquery\n.\nsort\n(\nage\n,\n \n.\ndescending\n)\n\n\n\n\n\n\nRaw\n\n\nShould you need to perform a query that the query builder does not support, you can use the raw query.\n\n\ntry\n \ndrop\n.\ndatabase\n?.\nraw\n(\nSELECT @@version\n)\n\n\n\n\n\n\nYou can also use the database of a given model.\n\n\nUser.database?.raw(\nSELECT * FROM `users`\n)\n\n\n\n\n\nBesides providing a more expressive interface for querying your database, the query builder also takes measures to increase security by automatically sanitizing input. Because of this, try to use the query class wherever you can over performing raw queries.", 
            "title": "Query"
        }, 
        {
            "location": "/fluent/query/#query", 
            "text": "Fluent's query builder provides a simple interface for creating complex database queries. The  Query  class itself (raw queries excluded) is the sole method by which Fluent communicates with your database.", 
            "title": "Query"
        }, 
        {
            "location": "/fluent/query/#make", 
            "text": "You can create a new query builder from any model class.  let   query   =   try   Post . makeQuery ()   You can also create queries from an instance. This is especially useful if you need to use a special\ndatabase connection (like for  transactions ) to save or update a model.  guard   let   post   =   try   Post . find ( 42 )   else   {   ...   }  post . content   =   Updated  let   query   =   try   post . makeQuery ( conn ). save ()", 
            "title": "Make"
        }, 
        {
            "location": "/fluent/query/#fetch", 
            "text": "You have multiple options for fetching the results of a query.", 
            "title": "Fetch"
        }, 
        {
            "location": "/fluent/query/#all", 
            "text": "The simplest option,  .all()  returns all rows relevant to the query.  let users = try User.makeQuery().filter(...).all()", 
            "title": "All"
        }, 
        {
            "location": "/fluent/query/#first", 
            "text": "You can take only the first row as well with  .first() .   let user = try User.makeQuery().filter(...).first()  Fluent will automatically limit the results to  1  to increase\nthe performance of the query.", 
            "title": "First"
        }, 
        {
            "location": "/fluent/query/#chunk", 
            "text": "If you want to fetch a large amount of models from the database, using  .chunk()  can help reduce the\namount of memory required for the query by fetching chunks of data at a time.  User.makeQuery().filter(...).chunk(32) { users in\n    print(users)\n}", 
            "title": "Chunk"
        }, 
        {
            "location": "/fluent/query/#filter", 
            "text": "Filters allow you to choose exactly what subset of data you want to modify or fetch. There are three different\ntypes of filters.", 
            "title": "Filter"
        }, 
        {
            "location": "/fluent/query/#compare", 
            "text": "Compare filters perform a comparison between a field on your model in the database and a supplied value.  try   query . filter ( age ,   . greaterThanOrEquals ,   21 )   You can also use operators.  try   query . filter ( age   =   21 )      Case  Operator  Type      .equals  ==  Equals    .greaterThan   Greater Than    .lessThan   Less Than    .greaterThanOrEquals  =  Greater Than Or Equals    .lessThanOrEquals  =  Less Than Or Equals    .notEquals  !=  Not Equals    .hasSuffix   Has Suffix    .hasPrefix   Has Prefix    .contains   Contains    .custom(String)   Custom      Tip  You can omit the comparison type for  .equals , e.g.,  query.filter(\"age\", 23)", 
            "title": "Compare"
        }, 
        {
            "location": "/fluent/query/#subset", 
            "text": "You can also filter by fields being in a set of data.  try   query . filter ( favoriteColor ,   in :   [ pink ,   blue ])   Or the opposite.  try   query . filter ( favoriteColor ,   notIn :   [ brown ,   black ])", 
            "title": "Subset"
        }, 
        {
            "location": "/fluent/query/#group", 
            "text": "By default, all query filters are joined by AND logic. You can create groups of filters within\nyour query that are joined with AND or OR logic.  try   query . or   {   orGroup   in \n     try   orGroup . filter ( age ,   . greaterThan ,   75 ) \n     try   orGroup . filter ( age ,   . lessThan ,   18 )  }   This will result in SQL similar to the following:  SELECT   *   FROM   ` users `   WHERE   ( ` age `     75   OR   ` age `     18 );   .and()  is also available in case you need to switch back to joining filters with AND nested inside of an OR.", 
            "title": "Group"
        }, 
        {
            "location": "/fluent/query/#complex-example", 
            "text": "let   users   =   try   User \n     . makeQuery () \n     . filter ( planetOfOrigin ,   . greaterThan ,   Earth ) \n     . or   {   orGroup   in \n         orGroup . and   {   andGroup   in \n             andGroup . filter ( name ,   Rick ) \n             andGroup . filter ( favoriteFood ,   Beer ) \n         } \n         orGroup . and   {   andGroup   in \n             andGroup . filter ( name ,   Morty ) \n             andGroup . filter ( favoriteFood ,   Eyeholes ) \n         } \n     } \n     . all ()   This will result in SQL similar to the following:  SELECT   *   FROM   ` users `  \n     WHERE   ` planetOfOrigin `   =   Earth   AND   ( \n            (` name `   =   Rick   AND   ` favoriteFood `   =   Beer ) \n         OR   (` name `   =   Morty   AND   ` favoriteFood `   =   Eyeholes ) \n     )    Note  Keep in mind that the AND/OR logic for a group applies only to the filters added  within  the group. All filters outside of a filter group will be joined by AND.", 
            "title": "Complex Example"
        }, 
        {
            "location": "/fluent/query/#raw", 
            "text": "Raw filters can be used to filter by values that should not be parameterized.  try   query . filter ( raw :   date  = CURRENT_TIMESTAMP )", 
            "title": "Raw"
        }, 
        {
            "location": "/fluent/query/#distinct", 
            "text": "To select only distinct models from the database, add  .distinct()  to your query.  try   query . distinct ()", 
            "title": "Distinct"
        }, 
        {
            "location": "/fluent/query/#limit-offset", 
            "text": "To limit or offset your query, use the  .limit()  method.  try   query . limit ( 20 ,   offset :   5 )", 
            "title": "Limit / Offset"
        }, 
        {
            "location": "/fluent/query/#sort", 
            "text": "To sort the results of your query, use the  .sort()  method  try   query . sort ( age ,   . descending )", 
            "title": "Sort"
        }, 
        {
            "location": "/fluent/query/#raw_1", 
            "text": "Should you need to perform a query that the query builder does not support, you can use the raw query.  try   drop . database ?. raw ( SELECT @@version )   You can also use the database of a given model.  User.database?.raw( SELECT * FROM `users` )  Besides providing a more expressive interface for querying your database, the query builder also takes measures to increase security by automatically sanitizing input. Because of this, try to use the query class wherever you can over performing raw queries.", 
            "title": "Raw"
        }, 
        {
            "location": "/fluent/relations/", 
            "text": "Relations\n\n\nFluent relations allow you to relate your models in three different ways:\n\n\n\n\n\n\n\n\nType\n\n\nRelations\n\n\n\n\n\n\n\n\n\n\nOne to One\n\n\nParent / Child\n\n\n\n\n\n\nOne to Many\n\n\nParent / Children\n\n\n\n\n\n\nMany to Many\n\n\nSiblings\n\n\n\n\n\n\n\n\nOne to Many\n\n\nWe'll start with one-to-many since it's the easiest type of relation to understand.\n\n\nTake the following database schema:\n\n\nusers\n\n\n\n\n\n\n\n\nid\n\n\nname\n\n\n\n\n\n\n\n\n\n\nid type\n\n\nstring\n\n\n\n\n\n\n\n\npets\n\n\n\n\n\n\n\n\nid\n\n\nname\n\n\nuser_id\n\n\n\n\n\n\n\n\n\n\nid type\n\n\nstring\n\n\nid type\n\n\n\n\n\n\n\n\n\n\nSeealso\n\n\nVisit the \ndatabase preparations\n guide for more information\non how to create schema.\n\n\n\n\nHere each pet has exactly one owner (a user) and each owner can have multiple pets. \nThis is a one-to-many relationship. One owner has many pets.\n\n\n\n\nTip\n\n\nUse the \nbuilder.foreignId()\n to create foreign ids like \nuser_id\n. This will automatically\ncreate foreign key constraints and follow pre-set key naming conventions.\n\n\n\n\nChildren\n\n\nTo access the user's pets, we will use the \nChildren\n relation.\n\n\nextension\n \nUser\n \n{\n\n    \nvar\n \npets\n:\n \nChildren\nUser\n,\n \nPet\n \n{\n\n        \nreturn\n \nchildren\n()\n\n    \n}\n\n\n}\n\n\n\n\n\n\nImagine the children relation as \nChildren\nParent, Child\n or \nChildren\nFrom, To\n.\nHere we are relating \nfrom\n the user type \nto\n the pet type.\n\n\nWe can now use this relation to get all of the user's pets.\n\n\nlet\n \npets\n \n=\n \ntry\n \nuser\n.\npets\n.\nall\n()\n \n// [Pet]\n\n\n\n\n\n\nThis will create SQL similar to:\n\n\nSELECT\n \n*\n \nFROM\n \n`\npets\n`\n \nWHERE\n \n`\nuser_id\n`\n \n=\n \n...\n;\n\n\n\n\n\n\nRelations work similarly to \nqueries\n.\n\n\nlet\n \npet\n \n=\n \ntry\n \nuser\n.\npets\n.\nfilter\n(\nname\n,\n \nSpud\n).\nfirst\n()\n\n\n\n\n\n\nParent\n\n\nTo access a pet's owner from the pet, we will use the \nParent\n relation.\n\n\nextension\n \nPet\n \n{\n\n    \nlet\n \nuserId\n:\n \nIdentifier\n\n\n    \n...\n\n\n    \nvar\n \nowner\n:\n \nParent\nPet\n,\n \nUser\n \n{\n\n        \nreturn\n \nparent\n(\nid\n:\n \nuserId\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nImagine the parent relation as \nParent\nChild, Parent\n or \nParent\nFrom, To\n.\nHere we are relating \nfrom\n the pet type \nto\n the parent type.\n\n\n\n\nNote\n\n\nNotice the \nParent\n relation requires an identifier to be passed in. \nMake sure to load this identifier in your model's \ninit(row:)\n method.\n\n\n\n\nWe can now use this relation to get the pet's owner.\n\n\nlet\n \nowner\n \n=\n \ntry\n \npet\n.\nowner\n.\nget\n()\n \n// User?\n\n\n\n\n\n\nMigration\n\n\nAdding a parent identifier to the child table can be done using the \n.parent()\n method on\nthe schema builder.\n\n\ntry\n \ndatabase\n.\ncreate\n(\nPet\n.\nself\n)\n \n{\n \nbuilder\n \nin\n\n    \n...\n\n    \nbuilder\n.\nparent\n(\nUser\n.\nself\n)\n\n\n}\n\n\n\n\n\n\nOne to One\n\n\nOne-to-one relations work exactly the same as one-to-many relations. You can use the\ncode from the previous example and simply call \n.first()\n and all calls from the parent type.\n\n\nHowever, you can add a convenience for doing this. Let's assume we wanted to change the previous\nexample from one-to-many to one-to-one.\n\n\nextension\n \nUser\n \n{\n\n   \nfunc\n \npet\n()\n \nthrows\n \n-\n \nPet\n?\n \n{\n\n        \nreturn\n \ntry\n \nchildren\n().\nfirst\n()\n\n   \n}\n \n\n}\n\n\n\n\n\n\nMany to Many\n\n\nMany to many relations require a table in between to store which model is related to which. \nThis table is called a pivot table.\n\n\nYou can use any entity you want as a pivot, but Fluent provides a default one called \nPivot\n.\n\n\nTake the following schema.\n\n\npets\n\n\n\n\n\n\n\n\nid\n\n\nname\n\n\n\n\n\n\n\n\n\n\nid type\n\n\nstring\n\n\n\n\n\n\n\n\npet_toy\n\n\n\n\n\n\n\n\nid\n\n\npet_id\n\n\ntoy_id\n\n\n\n\n\n\n\n\n\n\nid type\n\n\nid type\n\n\nid type\n\n\n\n\n\n\n\n\ntoys\n\n\n\n\n\n\n\n\nid\n\n\nname\n\n\n\n\n\n\n\n\n\n\nid type\n\n\nstring\n\n\n\n\n\n\n\n\nHere each pet can own many toys and each toy can belong to many pets. This is a many-to-many relationship.\n\n\nSiblings\n\n\nTo represent this many-to-many relationship, we will use the \nSiblings\n relation.\n\n\nextension\n \nPet\n \n{\n\n    \nvar\n \ntoys\n:\n \nSiblings\nPet\n,\n \nToy\n,\n \nPivot\nPet\n,\n \nToy\n \n{\n\n        \nreturn\n \nsiblings\n()\n\n    \n}\n\n\n}\n\n\n\n\n\n\nImagine the siblings relations as \nSiblings\nFrom, To, Through\n.\nHere we are relating \nfrom\n the pet type \nto\n the toy type \nthrough\n the pet/toy pivot.\n\n\n\n\nNote\n\n\nThe generic syntax might look a little intimidating at first, but it allows for a very powerful API.\n\n\n\n\nWith this relation added on pets, we can fetch a pet's toys.\n\n\nlet\n \ntoys\n \n=\n \npet\n.\ntoys\n.\nall\n()\n \n// [Toy]\n\n\n\n\n\n\nThe siblings relation works similarly to \nqueries\n and parent/children relations. \n\n\nMigration\n\n\nIf you are using a \nPivot\n type, you can simply add it to your Droplet's preparation array.\n\n\ndrop\n.\npreparations\n.\nappend\n(\nPivot\nPet\n,\n \nToy\n)\n\n\n\n\n\n\nIf you are using a \nPivot\n for your \"through\" model, it will also have methods for adding and removing models from the relation.\n\n\nAdd\n\n\nTo add a new model to the relation, use the \n.add()\n method.\n\n\ntry\n \npet\n.\ntoys\n.\nadd\n(\ntoy\n)\n\n\n\n\n\n\n\n\nNote\n\n\nThe newly created pivot will be returned.\n\n\n\n\nRemove\n\n\nTo remove a model from being related, use the \n.remove()\n method.\n\n\ntry\n \npet\n.\ntoys\n.\nremove\n(\ntoy\n)\n\n\n\n\n\n\nIs Attached\n\n\nTo check if a model is related, use the \n.isAttached()\n method.\n\n\nif\n \ntry\n \npet\n.\ntoys\n.\nisAttached\n(\nto\n:\n \ntoy\n)\n \n{\n\n    \n// it is attached\n\n\n}\n\n\n\n\n\n\nCustom Through\n\n\nYou can use any entity type as the \"through\" entity in your siblings relation. \n\n\nextension\n \nUser\n \n{\n\n    \nvar\n \nposts\n:\n \nSiblings\nUser\n,\n \nPost\n,\n \nComment\n \n{\n\n        \nreturn\n \nsiblings\n()\n\n    \n}\n\n\n}\n\n\n\n\n\n\nIn the above example we are pivoting on the comments entity to retreive all posts the user\nhas commented on.\n\n\nAs long as the \"through\" entity has a \nuser_id\n and \npost_id\n, the siblings relation will work.\n\n\n\n\nNote\n\n\nIf the \nComment\nentity does not conform to \nPivotProtocol\n, the \n\nadd\n, \nremove\n, and \nisAttached\n methods will not be available.", 
            "title": "Relations"
        }, 
        {
            "location": "/fluent/relations/#relations", 
            "text": "Fluent relations allow you to relate your models in three different ways:     Type  Relations      One to One  Parent / Child    One to Many  Parent / Children    Many to Many  Siblings", 
            "title": "Relations"
        }, 
        {
            "location": "/fluent/relations/#one-to-many", 
            "text": "We'll start with one-to-many since it's the easiest type of relation to understand.  Take the following database schema:  users     id  name      id type  string     pets     id  name  user_id      id type  string  id type      Seealso  Visit the  database preparations  guide for more information\non how to create schema.   Here each pet has exactly one owner (a user) and each owner can have multiple pets. \nThis is a one-to-many relationship. One owner has many pets.   Tip  Use the  builder.foreignId()  to create foreign ids like  user_id . This will automatically\ncreate foreign key constraints and follow pre-set key naming conventions.", 
            "title": "One to Many"
        }, 
        {
            "location": "/fluent/relations/#children", 
            "text": "To access the user's pets, we will use the  Children  relation.  extension   User   { \n     var   pets :   Children User ,   Pet   { \n         return   children () \n     }  }   Imagine the children relation as  Children Parent, Child  or  Children From, To .\nHere we are relating  from  the user type  to  the pet type.  We can now use this relation to get all of the user's pets.  let   pets   =   try   user . pets . all ()   // [Pet]   This will create SQL similar to:  SELECT   *   FROM   ` pets `   WHERE   ` user_id `   =   ... ;   Relations work similarly to  queries .  let   pet   =   try   user . pets . filter ( name ,   Spud ). first ()", 
            "title": "Children"
        }, 
        {
            "location": "/fluent/relations/#parent", 
            "text": "To access a pet's owner from the pet, we will use the  Parent  relation.  extension   Pet   { \n     let   userId :   Identifier \n\n     ... \n\n     var   owner :   Parent Pet ,   User   { \n         return   parent ( id :   userId ) \n     }  }   Imagine the parent relation as  Parent Child, Parent  or  Parent From, To .\nHere we are relating  from  the pet type  to  the parent type.   Note  Notice the  Parent  relation requires an identifier to be passed in. \nMake sure to load this identifier in your model's  init(row:)  method.   We can now use this relation to get the pet's owner.  let   owner   =   try   pet . owner . get ()   // User?", 
            "title": "Parent"
        }, 
        {
            "location": "/fluent/relations/#migration", 
            "text": "Adding a parent identifier to the child table can be done using the  .parent()  method on\nthe schema builder.  try   database . create ( Pet . self )   {   builder   in \n     ... \n     builder . parent ( User . self )  }", 
            "title": "Migration"
        }, 
        {
            "location": "/fluent/relations/#one-to-one", 
            "text": "One-to-one relations work exactly the same as one-to-many relations. You can use the\ncode from the previous example and simply call  .first()  and all calls from the parent type.  However, you can add a convenience for doing this. Let's assume we wanted to change the previous\nexample from one-to-many to one-to-one.  extension   User   { \n    func   pet ()   throws   -   Pet ?   { \n         return   try   children (). first () \n    }   }", 
            "title": "One to One"
        }, 
        {
            "location": "/fluent/relations/#many-to-many", 
            "text": "Many to many relations require a table in between to store which model is related to which. \nThis table is called a pivot table.  You can use any entity you want as a pivot, but Fluent provides a default one called  Pivot .  Take the following schema.  pets     id  name      id type  string     pet_toy     id  pet_id  toy_id      id type  id type  id type     toys     id  name      id type  string     Here each pet can own many toys and each toy can belong to many pets. This is a many-to-many relationship.", 
            "title": "Many to Many"
        }, 
        {
            "location": "/fluent/relations/#siblings", 
            "text": "To represent this many-to-many relationship, we will use the  Siblings  relation.  extension   Pet   { \n     var   toys :   Siblings Pet ,   Toy ,   Pivot Pet ,   Toy   { \n         return   siblings () \n     }  }   Imagine the siblings relations as  Siblings From, To, Through .\nHere we are relating  from  the pet type  to  the toy type  through  the pet/toy pivot.   Note  The generic syntax might look a little intimidating at first, but it allows for a very powerful API.   With this relation added on pets, we can fetch a pet's toys.  let   toys   =   pet . toys . all ()   // [Toy]   The siblings relation works similarly to  queries  and parent/children relations.", 
            "title": "Siblings"
        }, 
        {
            "location": "/fluent/relations/#migration_1", 
            "text": "If you are using a  Pivot  type, you can simply add it to your Droplet's preparation array.  drop . preparations . append ( Pivot Pet ,   Toy )   If you are using a  Pivot  for your \"through\" model, it will also have methods for adding and removing models from the relation.", 
            "title": "Migration"
        }, 
        {
            "location": "/fluent/relations/#add", 
            "text": "To add a new model to the relation, use the  .add()  method.  try   pet . toys . add ( toy )    Note  The newly created pivot will be returned.", 
            "title": "Add"
        }, 
        {
            "location": "/fluent/relations/#remove", 
            "text": "To remove a model from being related, use the  .remove()  method.  try   pet . toys . remove ( toy )", 
            "title": "Remove"
        }, 
        {
            "location": "/fluent/relations/#is-attached", 
            "text": "To check if a model is related, use the  .isAttached()  method.  if   try   pet . toys . isAttached ( to :   toy )   { \n     // it is attached  }", 
            "title": "Is Attached"
        }, 
        {
            "location": "/fluent/relations/#custom-through", 
            "text": "You can use any entity type as the \"through\" entity in your siblings relation.   extension   User   { \n     var   posts :   Siblings User ,   Post ,   Comment   { \n         return   siblings () \n     }  }   In the above example we are pivoting on the comments entity to retreive all posts the user\nhas commented on.  As long as the \"through\" entity has a  user_id  and  post_id , the siblings relation will work.   Note  If the  Comment entity does not conform to  PivotProtocol , the  add ,  remove , and  isAttached  methods will not be available.", 
            "title": "Custom Through"
        }, 
        {
            "location": "/cache/package/", 
            "text": "Using Cache\n\n\nThis package is included with the Vapor dependency, use\n\n\nimport\n \nCache", 
            "title": "Package"
        }, 
        {
            "location": "/cache/package/#using-cache", 
            "text": "This package is included with the Vapor dependency, use  import   Cache", 
            "title": "Using Cache"
        }, 
        {
            "location": "/cache/overview/", 
            "text": "Cache\n\n\nVapor's \nCacheProtocol\n allows you to store and fetch items from a cache using optional expiration dates.\n\n\nBy default, the Droplet's cache is set to \nMemoryCache\n. See the various \nproviders\n below.\n\n\nStore\n\n\nStoring data into the cache is straightforward.\n\n\ntry\n \ndrop\n.\ncache\n.\nset\n(\nhello\n,\n \nworld\n)\n\n\n\n\n\n\nExpiration\n\n\nWhen storing data, you can also supply an expiration date.\n\n\ntry\n \ndrop\n.\ncache\n.\nset\n(\nephemeral\n,\n \n42\n,\n \nexpiration\n:\n \nDate\n(\ntimeIntervalSinceNow\n:\n \n30\n))\n\n\n\n\n\n\nIn the above example, the supplied key value pair will expire after 30 seconds.\n\n\nFetch\n\n\nYou can retreive data from the cache using the \n.get()\n method.\n\n\ntry\n \ndrop\n.\ncache\n.\nget\n(\nhello\n)\n \n// \nworld\n\n\n\n\n\n\nDelete\n\n\nKeys can be deleted from the cache using the \n.delete()\n method.\n\n\ntry\n \ndrop\n.\ncache\n.\ndelete\n(\nhello\n)\n\n\n\n\n\n\nProviders\n\n\nHere is a list of official cache providers. You can \nsearch GitHub\n for additional packages.\n\n\n\n\n\n\n\n\nType\n\n\nKey\n\n\nDescription\n\n\nPackage\n\n\nClass\n\n\n\n\n\n\n\n\n\n\nMemory\n\n\nmemory\n\n\nIn-memory cache. Not persisted.\n\n\nVapor\n\n\nMemoryCache\n\n\n\n\n\n\nFluent\n\n\nfluent\n\n\nUses Fluent database.\n\n\nFluent Provider\n\n\nFluentCache\n\n\n\n\n\n\nRedis\n\n\nredis\n\n\nUses Redis database.\n\n\nRedisProvider\n\n\nRedisCache\n\n\n\n\n\n\n\n\nHow to Use\n\n\nTo use a different cache provider besides the default \nMemoryCache\n, make sure you have added the provider to your Package.\n\n\nimport\n \nVapor\n\n\nimport\n \npackage\nProvider\n\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\ntry\n \nconfig\n.\naddProvider\n(\npackage\nProvider\n.\nProvider\n.\nself\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n\n...\n\n\n\n\n\n\nThen change the Droplet's configuration file.\n\n\nConfig/droplet.json\n\n\n{\n\n    \ncache\n:\n \nkey\n\n\n}", 
            "title": "Overview"
        }, 
        {
            "location": "/cache/overview/#cache", 
            "text": "Vapor's  CacheProtocol  allows you to store and fetch items from a cache using optional expiration dates.  By default, the Droplet's cache is set to  MemoryCache . See the various  providers  below.", 
            "title": "Cache"
        }, 
        {
            "location": "/cache/overview/#store", 
            "text": "Storing data into the cache is straightforward.  try   drop . cache . set ( hello ,   world )", 
            "title": "Store"
        }, 
        {
            "location": "/cache/overview/#expiration", 
            "text": "When storing data, you can also supply an expiration date.  try   drop . cache . set ( ephemeral ,   42 ,   expiration :   Date ( timeIntervalSinceNow :   30 ))   In the above example, the supplied key value pair will expire after 30 seconds.", 
            "title": "Expiration"
        }, 
        {
            "location": "/cache/overview/#fetch", 
            "text": "You can retreive data from the cache using the  .get()  method.  try   drop . cache . get ( hello )   //  world", 
            "title": "Fetch"
        }, 
        {
            "location": "/cache/overview/#delete", 
            "text": "Keys can be deleted from the cache using the  .delete()  method.  try   drop . cache . delete ( hello )", 
            "title": "Delete"
        }, 
        {
            "location": "/cache/overview/#providers", 
            "text": "Here is a list of official cache providers. You can  search GitHub  for additional packages.     Type  Key  Description  Package  Class      Memory  memory  In-memory cache. Not persisted.  Vapor  MemoryCache    Fluent  fluent  Uses Fluent database.  Fluent Provider  FluentCache    Redis  redis  Uses Redis database.  RedisProvider  RedisCache", 
            "title": "Providers"
        }, 
        {
            "location": "/cache/overview/#how-to-use", 
            "text": "To use a different cache provider besides the default  MemoryCache , make sure you have added the provider to your Package.  import   Vapor  import   package Provider  let   config   =   try   Config ()  try   config . addProvider ( package Provider . Provider . self )  let   drop   =   try   Droplet ( config )  ...   Then change the Droplet's configuration file.  Config/droplet.json  { \n     cache :   key  }", 
            "title": "How to Use"
        }, 
        {
            "location": "/mysql/package/", 
            "text": "Using MySQL\n\n\nThis section outlines how to import the MySQL package both with or without a Vapor project.\n\n\nWith Vapor\n\n\nThe easiest way to use MySQL with Vapor is to include the MySQL provider. \n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/vapor.git\n,\n \nmajorVersion\n:\n \n2\n),\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/mysql-provider.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nThe MySQL provider package adds MySQL to your project and adds some additional, Vapor-specific conveniences like \ndrop.mysql()\n. \n\n\nUsing \nimport MySQLProvider\n will import both Fluent and Fluent's Vapor-specific APIs. \n\n\nWith Fluent\n\n\nFluent is a powerful, pure-Swift ORM that can be used with any Server-Side Swift framework. The MySQL driver allows you to use a MySQL database to power your models and queries.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/fluent.git\n,\n \nmajorVersion\n:\n \n2\n),\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/mysql-driver.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport MySQLDriver\n to access the \nMySQLDriver\n class which you can use to initialize a Fluent \nDatabase\n. \n\n\nJust MySQL\n\n\nAt the core of the MySQL provider and MySQL driver is a Swift wrapper around the C MySQL client. This package can be used by itself to send raw, parameterized queries to your MySQL database.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/mysql.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport MySQL\n to access the \nMySQL.Database\n class.", 
            "title": "Package"
        }, 
        {
            "location": "/mysql/package/#using-mysql", 
            "text": "This section outlines how to import the MySQL package both with or without a Vapor project.", 
            "title": "Using MySQL"
        }, 
        {
            "location": "/mysql/package/#with-vapor", 
            "text": "The easiest way to use MySQL with Vapor is to include the MySQL provider.   import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         . Package ( url :   https://github.com/vapor/vapor.git ,   majorVersion :   2 ), \n         . Package ( url :   https://github.com/vapor/mysql-provider.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   The MySQL provider package adds MySQL to your project and adds some additional, Vapor-specific conveniences like  drop.mysql() .   Using  import MySQLProvider  will import both Fluent and Fluent's Vapor-specific APIs.", 
            "title": "With Vapor"
        }, 
        {
            "location": "/mysql/package/#with-fluent", 
            "text": "Fluent is a powerful, pure-Swift ORM that can be used with any Server-Side Swift framework. The MySQL driver allows you to use a MySQL database to power your models and queries.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/fluent.git ,   majorVersion :   2 ), \n         . Package ( url :   https://github.com/vapor/mysql-driver.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import MySQLDriver  to access the  MySQLDriver  class which you can use to initialize a Fluent  Database .", 
            "title": "With Fluent"
        }, 
        {
            "location": "/mysql/package/#just-mysql", 
            "text": "At the core of the MySQL provider and MySQL driver is a Swift wrapper around the C MySQL client. This package can be used by itself to send raw, parameterized queries to your MySQL database.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/mysql.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import MySQL  to access the  MySQL.Database  class.", 
            "title": "Just MySQL"
        }, 
        {
            "location": "/mysql/provider/", 
            "text": "MySQL Provider\n\n\nAfter you've \nadded the MySQL Provider package\n to your project, setting the provider up in code is easy.\n\n\nAdd to Droplet\n\n\nFirst, register the \nMySQLProvider.Provider\n with your Droplet.\n\n\nimport\n \nVapor\n\n\nimport\n \nMySQLProvider\n\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\ntry\n \nconfig\n.\naddProvider\n(\nMySQLProvider\n.\nProvider\n.\nself\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n...\n\n\n\n\n\n\nConfigure Fluent\n\n\nOnce the provider is added to your Droplet, you can configure Fluent to use the MySQL driver.\n\n\nConfig/fluent.json\n\n\n{\n\n    \ndriver\n:\n \nmysql\n\n\n}\n\n\n\n\n\n\n\n\nSeealso\n\n\nLearn more about configuration files in the \nSettings guide\n.\n\n\n\n\nConfigure MySQL\n\n\nIf you run your application now, you will likely see an error that the MySQL configuration file is missing. Let's add that now.\n\n\nBasic\n\n\nHere is an example of a simple MySQL configuration file.\n\n\nConfig/mysql.json\n\n\n{\n\n    \nhostname\n:\n \n127.0.0.1\n,\n\n    \nuser\n:\n \nroot\n,\n\n    \npassword\n:\n \npassword\n,\n\n    \ndatabase\n:\n \nhello\n\n\n}\n\n\n\n\n\n\n\n\nNote\n\n\nIt's a good idea to store the MySQL configuration file in the \nConfig/secrets\n folder since it contains sensitive information.\n\n\n\n\nURL\n\n\nYou can also pass the MySQL credentials as a URL.\n\n\nConfig/mysql.json\n\n\n{\n\n    \nurl\n:\n \nhttp://root:password@127.0.0.1/hello\n\n\n}\n\n\n\n\n\n\nRead Replicas\n\n\nRead replicas can be supplied by passing a single \nmaster\n hostname and an array of \nreadReplicas\n hostnames.\n\n\nConfig/mysql.json\n\n\n{\n\n    \nmaster\n:\n \nmaster.mysql.foo.com\n,\n\n    \nreadReplicas\n:\n \n[\nread01.mysql.foo.com\n,\n \nread02.mysql.foo.com\n],\n\n    \nuser\n:\n \nroot\n,\n\n    \npassword\n:\n \npassword\n,\n\n    \ndatabase\n:\n \nhello\n\n\n}\n\n\n\n\n\n\n\n\nTip\n\n\nYou can also provide the \nreadReplicas\n as a comma-separated string.\n\n\n\n\nDriver\n\n\nYou can get access to the \nMySQL Driver\n on the droplet.\n\n\nimport\n \nVapor\n\n\nimport\n \nMySQLProvider\n\n\n\nlet\n \nmysqlDriver\n \n=\n \ntry\n \ndrop\n.\nmysql\n()\n\n\n\n\n\n\nConfigure Cache\n\n\nYou can also choose to use your Fluent database (now set to MySQL) for caching. \n\n\nConfig/droplet.json\n\n\n{\n\n    \ndriver\n:\n \nfluent\n\n\n}\n\n\n\n\n\n\nLearn more about \ncaching here\n.\n\n\nDone\n\n\nNext time you boot your Droplet, you should see:\n\n\nDatabase prepared\n\n\n\n\n\nYou are now ready to \nstart using Fluent\n with your MySQL database.", 
            "title": "Provider"
        }, 
        {
            "location": "/mysql/provider/#mysql-provider", 
            "text": "After you've  added the MySQL Provider package  to your project, setting the provider up in code is easy.", 
            "title": "MySQL Provider"
        }, 
        {
            "location": "/mysql/provider/#add-to-droplet", 
            "text": "First, register the  MySQLProvider.Provider  with your Droplet.  import   Vapor  import   MySQLProvider  let   config   =   try   Config ()  try   config . addProvider ( MySQLProvider . Provider . self )  let   drop   =   try   Droplet ( config )  ...", 
            "title": "Add to Droplet"
        }, 
        {
            "location": "/mysql/provider/#configure-fluent", 
            "text": "Once the provider is added to your Droplet, you can configure Fluent to use the MySQL driver.  Config/fluent.json  { \n     driver :   mysql  }    Seealso  Learn more about configuration files in the  Settings guide .", 
            "title": "Configure Fluent"
        }, 
        {
            "location": "/mysql/provider/#configure-mysql", 
            "text": "If you run your application now, you will likely see an error that the MySQL configuration file is missing. Let's add that now.", 
            "title": "Configure MySQL"
        }, 
        {
            "location": "/mysql/provider/#basic", 
            "text": "Here is an example of a simple MySQL configuration file.  Config/mysql.json  { \n     hostname :   127.0.0.1 , \n     user :   root , \n     password :   password , \n     database :   hello  }    Note  It's a good idea to store the MySQL configuration file in the  Config/secrets  folder since it contains sensitive information.", 
            "title": "Basic"
        }, 
        {
            "location": "/mysql/provider/#url", 
            "text": "You can also pass the MySQL credentials as a URL.  Config/mysql.json  { \n     url :   http://root:password@127.0.0.1/hello  }", 
            "title": "URL"
        }, 
        {
            "location": "/mysql/provider/#read-replicas", 
            "text": "Read replicas can be supplied by passing a single  master  hostname and an array of  readReplicas  hostnames.  Config/mysql.json  { \n     master :   master.mysql.foo.com , \n     readReplicas :   [ read01.mysql.foo.com ,   read02.mysql.foo.com ], \n     user :   root , \n     password :   password , \n     database :   hello  }    Tip  You can also provide the  readReplicas  as a comma-separated string.", 
            "title": "Read Replicas"
        }, 
        {
            "location": "/mysql/provider/#driver", 
            "text": "You can get access to the  MySQL Driver  on the droplet.  import   Vapor  import   MySQLProvider  let   mysqlDriver   =   try   drop . mysql ()", 
            "title": "Driver"
        }, 
        {
            "location": "/mysql/provider/#configure-cache", 
            "text": "You can also choose to use your Fluent database (now set to MySQL) for caching.   Config/droplet.json  { \n     driver :   fluent  }   Learn more about  caching here .", 
            "title": "Configure Cache"
        }, 
        {
            "location": "/mysql/provider/#done", 
            "text": "Next time you boot your Droplet, you should see:  Database prepared  You are now ready to  start using Fluent  with your MySQL database.", 
            "title": "Done"
        }, 
        {
            "location": "/mysql/driver/", 
            "text": "MySQL Driver\n\n\nFluent uses the MySQL driver to talk to your MySQL database. Although you won't need to use it most of the time, it does \nhave some handy features.\n\n\nRaw\n\n\nSometimes you need to bypass Fluent and send raw queries to the database.\n\n\nlet\n \nresult\n \n=\n \ntry\n \nmysqlDriver\n.\nraw\n(\nSELECT @@version\n)\n\n\n\n\n\n\n\n\nNote\n\n\nIf you are using Vapor, you can get access to the MySQL Driver with \ndrop.mysql()\n\n\n\n\nTransaction\n\n\nIf you are performing multiple queries that depend on eachother, you can use transactions to make sure nothing gets saved\nto the database if one of the queries fails.\n\n\ntry\n \nmysqlDriver\n.\ntransaction\n \n{\n \nconn\n \nin\n\n    \n// delete user\ns pets, then delete user\n\n    \n// if one of these fails, the transaction will rollback\n\n    \ntry\n \nuser\n.\npets\n.\nmakeQuery\n(\nconn\n).\ndelete\n()\n\n    \ntry\n \nuser\n.\nmakeQuery\n(\nconn\n).\ndelete\n()\n\n\n}\n\n\n\n\n\n\n\n\nWarning\n\n\nMake sure to use the connection supplied to the closure for all queries you want included in the transaction.\n\n\n\n\nManual\n\n\nYou can also manually send a query to the driver without going through Fluent.\n\n\nlet\n \nquery\n \n=\n \ntry\n \nUser\n.\nmakeQuery\n()\n\n\n...\n\n\n\nlet\n \nresults\n \n=\n \ntry\n \nmysqlDriver\n.\nquery\n(\nquery\n)", 
            "title": "Driver"
        }, 
        {
            "location": "/mysql/driver/#mysql-driver", 
            "text": "Fluent uses the MySQL driver to talk to your MySQL database. Although you won't need to use it most of the time, it does \nhave some handy features.", 
            "title": "MySQL Driver"
        }, 
        {
            "location": "/mysql/driver/#raw", 
            "text": "Sometimes you need to bypass Fluent and send raw queries to the database.  let   result   =   try   mysqlDriver . raw ( SELECT @@version )    Note  If you are using Vapor, you can get access to the MySQL Driver with  drop.mysql()", 
            "title": "Raw"
        }, 
        {
            "location": "/mysql/driver/#transaction", 
            "text": "If you are performing multiple queries that depend on eachother, you can use transactions to make sure nothing gets saved\nto the database if one of the queries fails.  try   mysqlDriver . transaction   {   conn   in \n     // delete user s pets, then delete user \n     // if one of these fails, the transaction will rollback \n     try   user . pets . makeQuery ( conn ). delete () \n     try   user . makeQuery ( conn ). delete ()  }    Warning  Make sure to use the connection supplied to the closure for all queries you want included in the transaction.", 
            "title": "Transaction"
        }, 
        {
            "location": "/mysql/driver/#manual", 
            "text": "You can also manually send a query to the driver without going through Fluent.  let   query   =   try   User . makeQuery ()  ...  let   results   =   try   mysqlDriver . query ( query )", 
            "title": "Manual"
        }, 
        {
            "location": "/redis/package/", 
            "text": "Using Redis\n\n\nThis section outlines how to import the Redis package both with or without a Vapor project.\n\n\nWith Vapor\n\n\nThe easiest way to use Redis with Vapor is to include the Redis provider. \n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/vapor.git\n,\n \nmajorVersion\n:\n \n2\n),\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/redis-provider.git\n,\n \nmajorVersion\n:\n \n1\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nThe Redis provider package adds Redis to your project and conforms it to Vapor's \nCacheProtocol\n. \n\n\nUse \nimport RedisProvider\n.\n\n\nJust Redis\n\n\nAt the core of the Redis provider is a pure Swift Redis client. This package can be used by itself to send raw cache queries to your Redis database.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/redis.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Redis\n.", 
            "title": "Package"
        }, 
        {
            "location": "/redis/package/#using-redis", 
            "text": "This section outlines how to import the Redis package both with or without a Vapor project.", 
            "title": "Using Redis"
        }, 
        {
            "location": "/redis/package/#with-vapor", 
            "text": "The easiest way to use Redis with Vapor is to include the Redis provider.   import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         . Package ( url :   https://github.com/vapor/vapor.git ,   majorVersion :   2 ), \n         . Package ( url :   https://github.com/vapor/redis-provider.git ,   majorVersion :   1 ) \n     ], \n     exclude :   [   ...   ]  )   The Redis provider package adds Redis to your project and conforms it to Vapor's  CacheProtocol .   Use  import RedisProvider .", 
            "title": "With Vapor"
        }, 
        {
            "location": "/redis/package/#just-redis", 
            "text": "At the core of the Redis provider is a pure Swift Redis client. This package can be used by itself to send raw cache queries to your Redis database.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/redis.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Redis .", 
            "title": "Just Redis"
        }, 
        {
            "location": "/redis/provider/", 
            "text": "Redis Provider\n\n\nAfter you've \nadded the Redis Provider package\n to your project, setting the provider up in code is easy.\n\n\nAdd to Droplet\n\n\nFirst, register the \nRedisProvider.Provider\n with your Droplet.\n\n\nimport\n \nVapor\n\n\nimport\n \nRedisProvider\n\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\ntry\n \nconfig\n.\naddProvider\n(\nRedisProvider\n.\nProvider\n.\nself\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n...\n\n\n\n\n\n\nConfigure Vapor\n\n\nOnce the provider is added to your Droplet, you can configure Vapor to use Redis for caching.\n\n\nConfig/droplet.json\n\n\n{\n\n    \ncache\n:\n \nredis\n\n\n}\n\n\n\n\n\n\n\n\nSeealso\n\n\nLearn more about configuration files in the \nSettings guide\n.\n\n\n\n\nConfigure Redis\n\n\nIf you run your application now, you will likely see an error that the Redis configuration file is missing. Let's add that now.\n\n\nBasic\n\n\nHere is an example of a simple Redis configuration file.\n\n\nConfig/redis.json\n\n\n{\n\n    \nhostname\n:\n \n127.0.0.1\n,\n\n    \nport\n:\n \n6379\n,\n\n    \npassword\n:\n \nsecret\n,\n\n    \ndatabase\n:\n \n2\n\n\n}\n\n\n\n\n\n\nBoth password and database are optional.\n\n\n\n\nNote\n\n\nIt's a good idea to store the Redis configuration file in the \nConfig/secrets\n folder since it may contain sensitive information.\n\n\n\n\nURL\n\n\nYou can also pass the Redis credentials as a URL.\n\n\nConfig/Redis.json\n\n\n{\n\n    \nurl\n:\n \nredis://:secret@127.0.0.1:6379/2\n\n\n}\n\n\n\n\n\n\nBoth password and database are optional.\n\n\nDone\n\n\nYou are now ready to \nstart using Cache\n with your Redis database.", 
            "title": "Provider"
        }, 
        {
            "location": "/redis/provider/#redis-provider", 
            "text": "After you've  added the Redis Provider package  to your project, setting the provider up in code is easy.", 
            "title": "Redis Provider"
        }, 
        {
            "location": "/redis/provider/#add-to-droplet", 
            "text": "First, register the  RedisProvider.Provider  with your Droplet.  import   Vapor  import   RedisProvider  let   config   =   try   Config ()  try   config . addProvider ( RedisProvider . Provider . self )  let   drop   =   try   Droplet ( config )  ...", 
            "title": "Add to Droplet"
        }, 
        {
            "location": "/redis/provider/#configure-vapor", 
            "text": "Once the provider is added to your Droplet, you can configure Vapor to use Redis for caching.  Config/droplet.json  { \n     cache :   redis  }    Seealso  Learn more about configuration files in the  Settings guide .", 
            "title": "Configure Vapor"
        }, 
        {
            "location": "/redis/provider/#configure-redis", 
            "text": "If you run your application now, you will likely see an error that the Redis configuration file is missing. Let's add that now.", 
            "title": "Configure Redis"
        }, 
        {
            "location": "/redis/provider/#basic", 
            "text": "Here is an example of a simple Redis configuration file.  Config/redis.json  { \n     hostname :   127.0.0.1 , \n     port :   6379 , \n     password :   secret , \n     database :   2  }   Both password and database are optional.   Note  It's a good idea to store the Redis configuration file in the  Config/secrets  folder since it may contain sensitive information.", 
            "title": "Basic"
        }, 
        {
            "location": "/redis/provider/#url", 
            "text": "You can also pass the Redis credentials as a URL.  Config/Redis.json  { \n     url :   redis://:secret@127.0.0.1:6379/2  }   Both password and database are optional.", 
            "title": "URL"
        }, 
        {
            "location": "/redis/provider/#done", 
            "text": "You are now ready to  start using Cache  with your Redis database.", 
            "title": "Done"
        }, 
        {
            "location": "/auth/package/", 
            "text": "Using Auth\n\n\nThis section outlines how to import the Auth package both with or without a Vapor project.\n\n\nWith Vapor\n\n\nThe easiest way to use Auth with Vapor is to include the Auth provider. \n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/vapor.git\n,\n \nmajorVersion\n:\n \n2\n),\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/auth-provider.git\n,\n \nmajorVersion\n:\n \n1\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nThe Auth provider package adds Auth to your project and adds some additional, vapor-specific conveniences like auth middleware. \n\n\nUsing \nimport AuthProvider\n will import all of the auth middleware and the Authentication and Authorization modules. \n\n\nJust Auth\n\n\nAt the core of the Auth provider is an Authentication and Authorization module based on Fluent.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/auth.git\n,\n \nmajorVersion\n:\n \n1\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Auth\n to access the core auth classes.", 
            "title": "Package"
        }, 
        {
            "location": "/auth/package/#using-auth", 
            "text": "This section outlines how to import the Auth package both with or without a Vapor project.", 
            "title": "Using Auth"
        }, 
        {
            "location": "/auth/package/#with-vapor", 
            "text": "The easiest way to use Auth with Vapor is to include the Auth provider.   import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         . Package ( url :   https://github.com/vapor/vapor.git ,   majorVersion :   2 ), \n         . Package ( url :   https://github.com/vapor/auth-provider.git ,   majorVersion :   1 ) \n     ], \n     exclude :   [   ...   ]  )   The Auth provider package adds Auth to your project and adds some additional, vapor-specific conveniences like auth middleware.   Using  import AuthProvider  will import all of the auth middleware and the Authentication and Authorization modules.", 
            "title": "With Vapor"
        }, 
        {
            "location": "/auth/package/#just-auth", 
            "text": "At the core of the Auth provider is an Authentication and Authorization module based on Fluent.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/auth.git ,   majorVersion :   1 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Auth  to access the core auth classes.", 
            "title": "Just Auth"
        }, 
        {
            "location": "/auth/provider/", 
            "text": "Auth Provider\n\n\nAfter you've \nadded the Auth Provider package\n to your project, setting the provider up in code is easy.\n\n\nAdd to Droplet\n\n\nRegister the \nAuthProvider.Provider\n with your Droplet.\n\n\nimport\n \nVapor\n\n\nimport\n \nAuthProvider\n\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\ntry\n \nconfig\n.\naddProvider\n(\nAuthProvider\n.\nProvider\n.\nself\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n...\n\n\n\n\n\n\nDone\n\n\nYou are now ready to start using the Auth package.", 
            "title": "Provider"
        }, 
        {
            "location": "/auth/provider/#auth-provider", 
            "text": "After you've  added the Auth Provider package  to your project, setting the provider up in code is easy.", 
            "title": "Auth Provider"
        }, 
        {
            "location": "/auth/provider/#add-to-droplet", 
            "text": "Register the  AuthProvider.Provider  with your Droplet.  import   Vapor  import   AuthProvider  let   config   =   try   Config ()  try   config . addProvider ( AuthProvider . Provider . self )  let   drop   =   try   Droplet ( config )  ...", 
            "title": "Add to Droplet"
        }, 
        {
            "location": "/auth/provider/#done", 
            "text": "You are now ready to start using the Auth package.", 
            "title": "Done"
        }, 
        {
            "location": "/auth/getting-started/", 
            "text": "Getting Started\n\n\nVapor's \nAuth Provider\n package makes implementing authentication and \nauthorization easy and secure. It supports common auth patterns such as:\n\n\n\n\nToken (bearer) authentication\n\n\nUsername + password (basic) authentication\n\n\nPermission-based authorization\n\n\nSession-based persistance \n\n\n\n\nAuth's modular, protocol-based nature also makes it a great foundation for custom auth needs.\n\n\nPackage\n\n\nTo use Auth, you will need to have the \nAuth Provider\n added to your project.\nThis is as simple as adding the following line to your \nPackage.swift\n file.\n\n\n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/auth-provider.git\n,\n \n...)\n\n\n\n\n\n\nCheck out the \nPackage\n section for more information.\n\n\nExample\n\n\nLet's take a look at how we can implement a simple, token-based authentication system using Vapor and Auth.\n\n\nUser\n\n\nWe will start by creating a model to represent our user. If you already have a user class, you can skip this step.\n\n\nimport\n \nVapor\n\n\nimport\n \nFluentProvider\n\n\n\nfinal\n \nclass\n \nExampleUser\n:\n \nModel\n \n{\n\n    \nlet\n \nname\n:\n \nString\n\n\n    \n...\n\n\n}\n\n\n\nextension\n \nExampleUser\n:\n \nPreparation\n \n{\n \n...\n \n}\n\n\n\n\n\n\nHere we create a very simple user with just one property: a name.\n\n\n\n\nSeealso\n\n\nWe're omitting most of \nModel\n and \nPreparation\n protocol requirements. Check out Fluent's \n\nGetting Started\n for more information about these protocols.\n\n\n\n\nToken\n\n\nNext let's create a model to represent our authentication tokens. These will be stored in a separate database\ntable or collection called \"tokens\".\n\n\nWhen a user logs in, we will create a new token for them. They will then use this token on subsequent requests\ninstead of their username and password. \n\n\nFor now, here's what our simple token model will look like.\n\n\nimport\n \nVapor\n\n\nimport\n \nFluentProvider\n\n\n\nfinal\n \nclass\n \nExampleToken\n:\n \nModel\n \n{\n\n    \nlet\n \ntoken\n:\n \nString\n\n    \nlet\n \nuserId\n:\n \nIdentifier\n\n\n    \nvar\n \nuser\n:\n \nParent\nExampleToken\n,\n \nExampleUser\n \n{\n\n        \nreturn\n \nparent\n(\nid\n:\n \nuserId\n)\n\n    \n}\n\n\n    \n...\n\n\n}\n\n\n\nextension\n \nExampleToken\n:\n \nPreparation\n \n{\n \n...\n \n}\n\n\n\n\n\n\nThis token has two properties:\n\n\n\n\ntoken: a unique, random string that we will send in requests\n\n\nuserId: the identifier for the user to whom this token belongs\n\n\n\n\n\n\nSeealso\n\n\nWe're using Fluent relations here. Check out Fluent's  \nRelations\n\nsection for more information.\n\n\n\n\nToken Authenticatable\n\n\nNow that we have our example user and token, we can make our user authenticatable with the token.\n\n\nThis might sound complicated, but it's actually pretty easy:\n\n\nimport\n \nAuthProvider\n\n\n\nextension\n \nExampleUser\n:\n \nTokenAuthenticatable\n \n{\n\n    \n// the token model that should be queried\n\n    \n// to authenticate this user\n\n    \npublic\n \ntypealias\n \nTokenType\n \n=\n \nExampleToken\n\n\n}\n\n\n\n\n\n\nNow that our example user is \nTokenAuthenticatable\n, we can move on to the next step!\n\n\nUser Helper\n\n\nLet's add a simple convenience method on request for accessing the authenticated user.\n\n\nextension\n \nRequest\n \n{\n\n    \nfunc\n \nuser\n()\n \nthrows\n \n-\n \nExampleUser\n \n{\n\n        \nreturn\n \ntry\n \nauth\n.\nassertAuthenticated\n()\n\n    \n}\n\n\n}\n\n\n\n\n\n\nThis is a nice shortcut that will come in handy in a few steps.\n\n\nMiddleware\n\n\nTo require authentication we need to add the \nTokenAuthenticationMiddleware\n. You can apply this middleware\nto individual routes or to the entire Droplet. For simplicity, we'll apply it to the Droplet.\n\n\nimport\n \nVapor\n\n\nimport\n \nAuthProvider\n\n\nimport\n \nFluentProvider\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\ndrop\n.\npreparations\n \n+=\n \n[\nExampleUser\n.\nself\n,\n \nExampleToken\n.\nself\n]\n\n\n\nlet\n \ntokenMiddleware\n \n=\n \nTokenAuthenticationMiddleware\n(\nExampleUser\n.\nself\n)\n\n\ndrop\n.\nmiddleware\n.\nappend\n(\ntokenMiddleware\n)\n\n\n\n\n\n\nSince our \nExampleUser\n class is \nTokenAuthenticatable\n, we can pass it into the middleware's init method.\n\n\n\n\nSeealso\n\n\nIf you only want to require authentication for certain routes, look at our \n\nRoute Group\n section in the routing docs.\n\n\n\n\nRoute\n\n\nNow that our Droplet is setup to require token authentication globally, let's add a route to\nreturn the authenticated user's name.\n\n\ndrop\n.\nget\n(\nme\n)\n \n{\n \nreq\n \nin\n\n    \n// return the authenticated user\ns name\n\n    \nreturn\n \ntry\n \nreq\n.\nuser\n().\nname\n\n\n}\n\n\n\n\n\n\n\n\nTip\n\n\nWe're using the \n.user()\n convenience we added to \nRequest\n here. It is a shortcut\nfor \nlet user = try req.auth.assertAuthenticated(ExampleUser.self)\n \n\n\n\n\nDatabase\n\n\nThat's it! We now have a functioning authentication system. Let's add a couple of entries\nto our database and test it out.\n\n\nUsers\n\n\n\n\n\n\n\n\nid\n\n\nname\n\n\n\n\n\n\n\n\n\n\n1\n\n\nBob\n\n\n\n\n\n\n\n\nTokens\n\n\n\n\n\n\n\n\nid\n\n\ntoken\n\n\nuser_id\n\n\n\n\n\n\n\n\n\n\n1\n\n\nfoo\n\n\n1\n\n\n\n\n\n\n\n\nRequest\n\n\nNow we can make a request to our Vapor app.\n\n\nGET\n \n/me\n \nHTTP\n/\n1.1\n\n\nAuthorization\n:\n \nBearer foo\n\n\n\n\n\n\nAnd we should get a response like.\n\n\nHTTP\n/\n1.1\n \n200\n \nOK\n\n\nContent-Type\n:\n \ntext/plain\n\n\nBob\n\n\n\n\n\nBad Token\n\n\nTo make sure it's secure, let's test using a token that's not in our database.\n\n\nGET\n \n/me\n \nHTTP\n/\n1.1\n\n\nAuthorization\n:\n \nBearer not-a-token\n\n\n\n\n\n\nAnd we should get a response like.\n\n\nHTTP\n/\n1.1\n \n403\n \nForbidden\n\n\n\n\n\n\nNext Steps\n\n\nTo build this out into a production-ready authentication system, you will need to build some\nadditional routes for creating users and creating tokens. \n\n\nContinue on in the Auth section to learn more about different types of authentication.", 
            "title": "Getting Started"
        }, 
        {
            "location": "/auth/getting-started/#getting-started", 
            "text": "Vapor's  Auth Provider  package makes implementing authentication and \nauthorization easy and secure. It supports common auth patterns such as:   Token (bearer) authentication  Username + password (basic) authentication  Permission-based authorization  Session-based persistance    Auth's modular, protocol-based nature also makes it a great foundation for custom auth needs.", 
            "title": "Getting Started"
        }, 
        {
            "location": "/auth/getting-started/#package", 
            "text": "To use Auth, you will need to have the  Auth Provider  added to your project.\nThis is as simple as adding the following line to your  Package.swift  file.  . Package ( url :   https://github.com/vapor/auth-provider.git ,   ...)   Check out the  Package  section for more information.", 
            "title": "Package"
        }, 
        {
            "location": "/auth/getting-started/#example", 
            "text": "Let's take a look at how we can implement a simple, token-based authentication system using Vapor and Auth.", 
            "title": "Example"
        }, 
        {
            "location": "/auth/getting-started/#user", 
            "text": "We will start by creating a model to represent our user. If you already have a user class, you can skip this step.  import   Vapor  import   FluentProvider  final   class   ExampleUser :   Model   { \n     let   name :   String \n\n     ...  }  extension   ExampleUser :   Preparation   {   ...   }   Here we create a very simple user with just one property: a name.   Seealso  We're omitting most of  Model  and  Preparation  protocol requirements. Check out Fluent's  Getting Started  for more information about these protocols.", 
            "title": "User"
        }, 
        {
            "location": "/auth/getting-started/#token", 
            "text": "Next let's create a model to represent our authentication tokens. These will be stored in a separate database\ntable or collection called \"tokens\".  When a user logs in, we will create a new token for them. They will then use this token on subsequent requests\ninstead of their username and password.   For now, here's what our simple token model will look like.  import   Vapor  import   FluentProvider  final   class   ExampleToken :   Model   { \n     let   token :   String \n     let   userId :   Identifier \n\n     var   user :   Parent ExampleToken ,   ExampleUser   { \n         return   parent ( id :   userId ) \n     } \n\n     ...  }  extension   ExampleToken :   Preparation   {   ...   }   This token has two properties:   token: a unique, random string that we will send in requests  userId: the identifier for the user to whom this token belongs    Seealso  We're using Fluent relations here. Check out Fluent's   Relations \nsection for more information.", 
            "title": "Token"
        }, 
        {
            "location": "/auth/getting-started/#token-authenticatable", 
            "text": "Now that we have our example user and token, we can make our user authenticatable with the token.  This might sound complicated, but it's actually pretty easy:  import   AuthProvider  extension   ExampleUser :   TokenAuthenticatable   { \n     // the token model that should be queried \n     // to authenticate this user \n     public   typealias   TokenType   =   ExampleToken  }   Now that our example user is  TokenAuthenticatable , we can move on to the next step!", 
            "title": "Token Authenticatable"
        }, 
        {
            "location": "/auth/getting-started/#user-helper", 
            "text": "Let's add a simple convenience method on request for accessing the authenticated user.  extension   Request   { \n     func   user ()   throws   -   ExampleUser   { \n         return   try   auth . assertAuthenticated () \n     }  }   This is a nice shortcut that will come in handy in a few steps.", 
            "title": "User Helper"
        }, 
        {
            "location": "/auth/getting-started/#middleware", 
            "text": "To require authentication we need to add the  TokenAuthenticationMiddleware . You can apply this middleware\nto individual routes or to the entire Droplet. For simplicity, we'll apply it to the Droplet.  import   Vapor  import   AuthProvider  import   FluentProvider  let   drop   =   try   Droplet ()  drop . preparations   +=   [ ExampleUser . self ,   ExampleToken . self ]  let   tokenMiddleware   =   TokenAuthenticationMiddleware ( ExampleUser . self )  drop . middleware . append ( tokenMiddleware )   Since our  ExampleUser  class is  TokenAuthenticatable , we can pass it into the middleware's init method.   Seealso  If you only want to require authentication for certain routes, look at our  Route Group  section in the routing docs.", 
            "title": "Middleware"
        }, 
        {
            "location": "/auth/getting-started/#route", 
            "text": "Now that our Droplet is setup to require token authentication globally, let's add a route to\nreturn the authenticated user's name.  drop . get ( me )   {   req   in \n     // return the authenticated user s name \n     return   try   req . user (). name  }    Tip  We're using the  .user()  convenience we added to  Request  here. It is a shortcut\nfor  let user = try req.auth.assertAuthenticated(ExampleUser.self)", 
            "title": "Route"
        }, 
        {
            "location": "/auth/getting-started/#database", 
            "text": "That's it! We now have a functioning authentication system. Let's add a couple of entries\nto our database and test it out.", 
            "title": "Database"
        }, 
        {
            "location": "/auth/getting-started/#users", 
            "text": "id  name      1  Bob", 
            "title": "Users"
        }, 
        {
            "location": "/auth/getting-started/#tokens", 
            "text": "id  token  user_id      1  foo  1", 
            "title": "Tokens"
        }, 
        {
            "location": "/auth/getting-started/#request", 
            "text": "Now we can make a request to our Vapor app.  GET   /me   HTTP / 1.1  Authorization :   Bearer foo   And we should get a response like.  HTTP / 1.1   200   OK  Content-Type :   text/plain \n\nBob", 
            "title": "Request"
        }, 
        {
            "location": "/auth/getting-started/#bad-token", 
            "text": "To make sure it's secure, let's test using a token that's not in our database.  GET   /me   HTTP / 1.1  Authorization :   Bearer not-a-token   And we should get a response like.  HTTP / 1.1   403   Forbidden", 
            "title": "Bad Token"
        }, 
        {
            "location": "/auth/getting-started/#next-steps", 
            "text": "To build this out into a production-ready authentication system, you will need to build some\nadditional routes for creating users and creating tokens.   Continue on in the Auth section to learn more about different types of authentication.", 
            "title": "Next Steps"
        }, 
        {
            "location": "/auth/helper/", 
            "text": "Auth Helper\n\n\nThe Auth package adds a convenience property on every request that makes it \neasy to authenticate, persist, and unauthenticate users.\n\n\nAuthentication\n\n\nChecking\n\n\nYou can get the currently authenticated user.\n\n\nlet\n \nuser\n \n=\n \nreq\n.\nauth\n.\nauthenticated\n(\nUser\n.\nself\n)\n\n\n\n\n\n\nYou can check to see if the user is authenticated.\n\n\nif\n \nreq\n.\nauth\n.\nisAuthenticated\n(\nUser\n.\nself\n)\n \n{\n\n    \n...\n \n\n}\n\n\n\n\n\n\nYou can also assert that the user is authenticated.\n\n\nlet\n \nuser\n \n=\n \ntry\n \nreq\n.\nauth\n.\nassertAuthenticated\n(\nUser\n.\nself\n)\n\n\n\n\n\n\n\n\nNote\n\n\nA 403 Forbidden error will be thrown if the user is not authenticated.\n\n\n\n\nManual\n\n\nYou can manually authenticate a user.\n\n\nif\n \nlet\n \nuser\n \n=\n \ntry\n \nUser\n.\nfind\n(\n1\n)\n \n{\n\n    \nreq\n.\nauth\n.\nauthenticate\n(\nuser\n)\n \n\n}\n\n\n\n\n\n\nYou can also unauthenticate the currently authenticated user.\n\n\ntry\n \nreq\n.\nauth\n.\nunauthenticate\n()\n\n\n\n\n\n\n\n\nNote\n\n\nIf the user is \nPersistable\n, they will also be unpersisted.\n\n\n\n\nHeaders\n\n\nThe helper can be used to access common authorization headers.\n\n\nprint\n(\nreq\n.\nauth\n.\nheader\n)\n\n\n\n\n\n\nToken\n\n\nThe header has additional conveniences for parsing out bearer tokens.\n\n\nprint\n(\nreq\n.\nauth\n.\nheader\n?.\nbearer\n)\n\n\n\n\n\n\n\n\nTip\n\n\nYou can use \n_authorizationBasic\n and \n_authorizationBearer\n to send tokens in the URL string.\n\n\n\n\nPassword\n\n\nAnd basic auth username + password.\n\n\nprint\n(\nreq\n.\nauth\n.\nheader\n?.\nbasic\n)", 
            "title": "Helper"
        }, 
        {
            "location": "/auth/helper/#auth-helper", 
            "text": "The Auth package adds a convenience property on every request that makes it \neasy to authenticate, persist, and unauthenticate users.", 
            "title": "Auth Helper"
        }, 
        {
            "location": "/auth/helper/#authentication", 
            "text": "", 
            "title": "Authentication"
        }, 
        {
            "location": "/auth/helper/#checking", 
            "text": "You can get the currently authenticated user.  let   user   =   req . auth . authenticated ( User . self )   You can check to see if the user is authenticated.  if   req . auth . isAuthenticated ( User . self )   { \n     ...   }   You can also assert that the user is authenticated.  let   user   =   try   req . auth . assertAuthenticated ( User . self )    Note  A 403 Forbidden error will be thrown if the user is not authenticated.", 
            "title": "Checking"
        }, 
        {
            "location": "/auth/helper/#manual", 
            "text": "You can manually authenticate a user.  if   let   user   =   try   User . find ( 1 )   { \n     req . auth . authenticate ( user )   }   You can also unauthenticate the currently authenticated user.  try   req . auth . unauthenticate ()    Note  If the user is  Persistable , they will also be unpersisted.", 
            "title": "Manual"
        }, 
        {
            "location": "/auth/helper/#headers", 
            "text": "The helper can be used to access common authorization headers.  print ( req . auth . header )", 
            "title": "Headers"
        }, 
        {
            "location": "/auth/helper/#token", 
            "text": "The header has additional conveniences for parsing out bearer tokens.  print ( req . auth . header ?. bearer )    Tip  You can use  _authorizationBasic  and  _authorizationBearer  to send tokens in the URL string.", 
            "title": "Token"
        }, 
        {
            "location": "/auth/helper/#password", 
            "text": "And basic auth username + password.  print ( req . auth . header ?. basic )", 
            "title": "Password"
        }, 
        {
            "location": "/auth/password/", 
            "text": "Username + Password (Basic) Auth\n\n\nThe \nAuthorization: Basic ...\n header can be used to send username and password credentials\nfor authentication.\n\n\nThis page will show you how to use this type of authentication in your web app.\n\n\n\n\nNote\n\n\nSending and storing passwords should be avoided wherever possible. Use tokens or\nsessions persistance to prevent the need for sending the password in every request.\n\n\n\n\nPassword Authenticatable\n\n\nStart by conforming your user model to the \nPasswordAuthenticatable\n protocol.\n\n\nimport\n \nAuthProvider\n\n\n\nextension\n \nUser\n:\n \nPasswordAuthenticatable\n \n{\n \n}\n\n\n\n\n\n\nCustom\n\n\nIf your user conforms to \nModel\n, all of the required methods will be implemented automatically. However,\nyou can implement them if you want to do something custom.\n\n\nextension\n \nUser\n:\n \nPasswordAuthenticatable\n \n{\n\n    \n/// Return the user matching the supplied\n\n    \n/// username and password\n\n    \nstatic\n \nfunc\n \nauthenticate\n(\n_\n:\n \nPassword\n)\n \nthrows\n \n-\n \nSelf\n \n{\n\n        \n// something custom\n\n    \n}\n\n\n    \n/// The entity\ns hashed password used for\n\n    \n/// validating against Password credentials\n\n    \n/// with a PasswordVerifier\n\n    \nvar\n \nhashedPassword\n:\n \nString\n?\n \n{\n\n        \n// something custom\n\n    \n}\n\n\n    \n/// The key under which the user\ns username,\n\n    \n/// email, or other identifing value is stored.\n\n    \nstatic\n \nvar\n \nusernameKey\n:\n \nString\n \n{\n\n        \n// something custom\n\n    \n}\n\n\n    \n/// The key under which the user\ns password\n\n    \n/// is stored.\n\n    \nstatic\n \nvar\n \npasswordKey\n:\n \nString\n \n{\n\n        \n// something custom\n\n    \n}\n\n\n    \n/// Optional password verifier to use when\n\n    \n/// comparing plaintext passwords from the \n\n    \n/// Authorization header to hashed passwords\n\n    \n/// in the database.\n\n    \nstatic\n \nvar\n \npasswordVerifier\n:\n \nPasswordVerifier\n?\n \n{\n\n        \n// some hasher\n\n    \n}\n\n\n}\n\n\n\n\n\n\nMiddleware\n\n\nOnce your model conforms to the \nPasswordAuthenticatable\n protocol, you can create the middleware.\n\n\nimport\n \nVapor\n\n\nimport\n \nAuthProvider\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\ndrop\n.\nmiddleware\n \n+=\n \n[\n\n    \nPasswordAuthenticationMiddleware\n(\nUser\n.\nself\n)\n\n\n]\n\n\n\ntry\n \ndrop\n.\nrun\n()\n\n\n\n\n\n\nAll routes on this Droplet will now be protected by the password middleware.\n\n\n\n\nSeealso\n\n\nIf you only want to require authentication for certain routes, look at our \n\nRoute Group\n section in the routing docs.\n\n\n\n\nCall \nreq.user.authenticated(User.self)\n to get access to the authenticated user.\n\n\nRequest\n\n\nNow we can make a request to our Vapor app.\n\n\nGET\n \n/me\n \nHTTP\n/\n1.1\n\n\nAuthorization\n:\n \nBasic dmFwb3I6Zm9v \n\n\n\n\n\n\n\n\nNote\n\n\ndmFwb3I6Zm9v\n is \"vapor:foo\" base64 encoded where \"vapor\" is the username and \n\"foo\" is the password. This is the format of Basic authorization headers.\n\n\n\n\nAnd we should get a response like.\n\n\nHTTP\n/\n1.1\n \n200\n \nOK\n\n\nContent-Type\n:\n \ntext/plain\n\n\nVapor", 
            "title": "Password"
        }, 
        {
            "location": "/auth/password/#username-password-basic-auth", 
            "text": "The  Authorization: Basic ...  header can be used to send username and password credentials\nfor authentication.  This page will show you how to use this type of authentication in your web app.   Note  Sending and storing passwords should be avoided wherever possible. Use tokens or\nsessions persistance to prevent the need for sending the password in every request.", 
            "title": "Username + Password (Basic) Auth"
        }, 
        {
            "location": "/auth/password/#password-authenticatable", 
            "text": "Start by conforming your user model to the  PasswordAuthenticatable  protocol.  import   AuthProvider  extension   User :   PasswordAuthenticatable   {   }", 
            "title": "Password Authenticatable"
        }, 
        {
            "location": "/auth/password/#custom", 
            "text": "If your user conforms to  Model , all of the required methods will be implemented automatically. However,\nyou can implement them if you want to do something custom.  extension   User :   PasswordAuthenticatable   { \n     /// Return the user matching the supplied \n     /// username and password \n     static   func   authenticate ( _ :   Password )   throws   -   Self   { \n         // something custom \n     } \n\n     /// The entity s hashed password used for \n     /// validating against Password credentials \n     /// with a PasswordVerifier \n     var   hashedPassword :   String ?   { \n         // something custom \n     } \n\n     /// The key under which the user s username, \n     /// email, or other identifing value is stored. \n     static   var   usernameKey :   String   { \n         // something custom \n     } \n\n     /// The key under which the user s password \n     /// is stored. \n     static   var   passwordKey :   String   { \n         // something custom \n     } \n\n     /// Optional password verifier to use when \n     /// comparing plaintext passwords from the  \n     /// Authorization header to hashed passwords \n     /// in the database. \n     static   var   passwordVerifier :   PasswordVerifier ?   { \n         // some hasher \n     }  }", 
            "title": "Custom"
        }, 
        {
            "location": "/auth/password/#middleware", 
            "text": "Once your model conforms to the  PasswordAuthenticatable  protocol, you can create the middleware.  import   Vapor  import   AuthProvider  let   drop   =   try   Droplet ()  drop . middleware   +=   [ \n     PasswordAuthenticationMiddleware ( User . self )  ]  try   drop . run ()   All routes on this Droplet will now be protected by the password middleware.   Seealso  If you only want to require authentication for certain routes, look at our  Route Group  section in the routing docs.   Call  req.user.authenticated(User.self)  to get access to the authenticated user.", 
            "title": "Middleware"
        }, 
        {
            "location": "/auth/password/#request", 
            "text": "Now we can make a request to our Vapor app.  GET   /me   HTTP / 1.1  Authorization :   Basic dmFwb3I6Zm9v     Note  dmFwb3I6Zm9v  is \"vapor:foo\" base64 encoded where \"vapor\" is the username and \n\"foo\" is the password. This is the format of Basic authorization headers.   And we should get a response like.  HTTP / 1.1   200   OK  Content-Type :   text/plain \n\nVapor", 
            "title": "Request"
        }, 
        {
            "location": "/auth/persist/", 
            "text": "Persisting Auth\n\n\nPersisting authentication means that a user does not need to provide their credentials with every request.\nThis is useful for web apps where a user should only have to log in once. \n\n\n\n\nNote\n\n\nFor APIs, it's recommended that the user send a token with every request.\nSee \nGetting Started\n for an example about Token auth.\n\n\n\n\nSessions\n\n\nSessions are built into Vapor by default and are an easy way to persist users in your web app.\n\n\nSessionPersistable\n\n\nThe first step is to conform your user model to the \nSessionPersistable\n protocol.\n\n\nimport\n \nAuthProvider\n\n\n\nextension\n \nUser\n:\n \nSessionPersistable\n \n{}\n\n\n\n\n\n\nIf your user is a Model, the protocol methods will be implemented automatically. However,\nyou can implement them if you want to do something custom.\n\n\nimport\n \nAuthProvider\n\n\nimport\n \nHTTP\n\n\n\nextension\n \nUser\n:\n \nSessionPersistable\n \n{\n\n    \nfunc\n \npersist\n(\nfor\n:\n \nRequest\n)\n \nthrows\n \n{\n\n        \n// something custom\n\n    \n}\n\n\n    \nstatic\n \nfunc\n \nfetchPersisted\n(\nfor\n:\n \nRequest\n)\n \nthrows\n \n-\n \nSelf\n?\n \n{\n\n        \n// something custom\n\n    \n}\n\n\n}\n\n\n\n\n\n\nMiddleware\n\n\nNow that the user is \nSessionPersistable\n, we can create our middleware.\n\n\nSessions\n\n\nFirst let's start by creating \nSessionMiddleware\n. We'll use the \nMemorySessions()\n to get started.\n\n\nlet\n \nmemory\n \n=\n \nMemorySessions\n()\n\n\nlet\n \nsessionsMiddleware\n \n=\n \nSessionsMiddleware\n(\nmemory\n)\n\n\n\n\n\n\nPersist\n\n\nNow let's create the \nPersistMiddleware\n. This will take care of persisting our user once they've \nbeen authenticated. \n\n\nlet\n \npersistMiddleware\n \n=\n \nPersistMiddleware\n(\nUser\n.\nself\n)\n\n\n\n\n\n\nSince our user conforms to \nSessionPersistable\n (and thus \nPersistable\n), we can pass it \ninto this middleware's init.\n\n\nAuthentication\n\n\nNow to create the authentication middleware of your choice. We'll use \nPasswordAuthenticationMiddleware\n\nwhich requires an \nAuthorization: Basic ...\n header with the user's username and password.\n\n\nlet\n \npasswordMiddleware\n \n=\n \nPasswordAuthenticationMiddleware\n(\nUser\n.\nself\n)\n\n\n\n\n\n\n\n\nNote\n\n\nUser\n must conform to \nPasswordAuthenticatable\n to be used with this middleware.\nSee the \nPassword\n section to learn more.\n\n\n\n\nDroplet\n\n\nNow we can create a Droplet and add all of our middleware.\n\n\nimport\n \nVapor\n\n\nimport\n \nSessions\n\n\nimport\n \nAuthProvider\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\ndrop\n.\nmiddleware\n \n+=\n \n[\n\n    \nsessionsMiddleware\n,\n \npersistMiddleware\n,\n \npasswordMiddleware\n\n\n]\n\n\n\n\n\n\n\n\nSeealso\n\n\nIf you only want to require authentication for certain routes, look at our \n\nRoute Group\n section in the routing docs.\n\n\n\n\nRoute\n\n\nNow you can add a route to return the authenticated user.\n\n\ndrop\n.\nget\n(\nme\n)\n \n{\n \nreq\n \nin\n\n    \n// return the authenticated user\n\n    \nreturn\n \ntry\n \nreq\n.\nauth\n.\nassertAuthenticated\n(\nUser\n.\nself\n)\n\n\n}\n\n\n\n\n\n\nRequest\n\n\nNow we can make a request to our Vapor app.\n\n\nGET\n \n/me\n \nHTTP\n/\n1.1\n\n\nAuthorization\n:\n \nBasic dmFwb3I6Zm9v \n\n\n\n\n\n\n\n\nNote\n\n\ndmFwb3I6Zm9v\n is \"vapor:foo\" base64 encoded where \"vapor\" is the username and \n\"foo\" is the password. This is the format of Basic authorization headers.\n\n\n\n\nAnd we should get a response like.\n\n\nHTTP\n/\n1.1\n \n200\n \nOK\n\n\nContent-Type\n:\n \ntext/plain\n\n\nSet-Cookie\n:\n \nvapor-session=...\n\n\nVapor\n\n\n\n\n\nNotice the \nvapor-session\n in the response. This can be used in subsequent requests instead of \nthe username and password.", 
            "title": "Persist"
        }, 
        {
            "location": "/auth/persist/#persisting-auth", 
            "text": "Persisting authentication means that a user does not need to provide their credentials with every request.\nThis is useful for web apps where a user should only have to log in once.    Note  For APIs, it's recommended that the user send a token with every request.\nSee  Getting Started  for an example about Token auth.", 
            "title": "Persisting Auth"
        }, 
        {
            "location": "/auth/persist/#sessions", 
            "text": "Sessions are built into Vapor by default and are an easy way to persist users in your web app.", 
            "title": "Sessions"
        }, 
        {
            "location": "/auth/persist/#sessionpersistable", 
            "text": "The first step is to conform your user model to the  SessionPersistable  protocol.  import   AuthProvider  extension   User :   SessionPersistable   {}   If your user is a Model, the protocol methods will be implemented automatically. However,\nyou can implement them if you want to do something custom.  import   AuthProvider  import   HTTP  extension   User :   SessionPersistable   { \n     func   persist ( for :   Request )   throws   { \n         // something custom \n     } \n\n     static   func   fetchPersisted ( for :   Request )   throws   -   Self ?   { \n         // something custom \n     }  }", 
            "title": "SessionPersistable"
        }, 
        {
            "location": "/auth/persist/#middleware", 
            "text": "Now that the user is  SessionPersistable , we can create our middleware.", 
            "title": "Middleware"
        }, 
        {
            "location": "/auth/persist/#sessions_1", 
            "text": "First let's start by creating  SessionMiddleware . We'll use the  MemorySessions()  to get started.  let   memory   =   MemorySessions ()  let   sessionsMiddleware   =   SessionsMiddleware ( memory )", 
            "title": "Sessions"
        }, 
        {
            "location": "/auth/persist/#persist", 
            "text": "Now let's create the  PersistMiddleware . This will take care of persisting our user once they've \nbeen authenticated.   let   persistMiddleware   =   PersistMiddleware ( User . self )   Since our user conforms to  SessionPersistable  (and thus  Persistable ), we can pass it \ninto this middleware's init.", 
            "title": "Persist"
        }, 
        {
            "location": "/auth/persist/#authentication", 
            "text": "Now to create the authentication middleware of your choice. We'll use  PasswordAuthenticationMiddleware \nwhich requires an  Authorization: Basic ...  header with the user's username and password.  let   passwordMiddleware   =   PasswordAuthenticationMiddleware ( User . self )    Note  User  must conform to  PasswordAuthenticatable  to be used with this middleware.\nSee the  Password  section to learn more.", 
            "title": "Authentication"
        }, 
        {
            "location": "/auth/persist/#droplet", 
            "text": "Now we can create a Droplet and add all of our middleware.  import   Vapor  import   Sessions  import   AuthProvider  let   drop   =   try   Droplet ()  drop . middleware   +=   [ \n     sessionsMiddleware ,   persistMiddleware ,   passwordMiddleware  ]    Seealso  If you only want to require authentication for certain routes, look at our  Route Group  section in the routing docs.", 
            "title": "Droplet"
        }, 
        {
            "location": "/auth/persist/#route", 
            "text": "Now you can add a route to return the authenticated user.  drop . get ( me )   {   req   in \n     // return the authenticated user \n     return   try   req . auth . assertAuthenticated ( User . self )  }", 
            "title": "Route"
        }, 
        {
            "location": "/auth/persist/#request", 
            "text": "Now we can make a request to our Vapor app.  GET   /me   HTTP / 1.1  Authorization :   Basic dmFwb3I6Zm9v     Note  dmFwb3I6Zm9v  is \"vapor:foo\" base64 encoded where \"vapor\" is the username and \n\"foo\" is the password. This is the format of Basic authorization headers.   And we should get a response like.  HTTP / 1.1   200   OK  Content-Type :   text/plain  Set-Cookie :   vapor-session=... \n\nVapor  Notice the  vapor-session  in the response. This can be used in subsequent requests instead of \nthe username and password.", 
            "title": "Request"
        }, 
        {
            "location": "/sessions/sessions/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\nSessions\n\n\nSessions help you store information about a user between requests. As long as the client supports cookies, sessions are easy to create.\n\n\nMiddleware\n\n\nEnable sessions on your \nDroplet\n by adding an instance of \nSessionMiddleware\n.\n\n\nimport\n \nSessions\n\n\n\nlet\n \nmemory\n \n=\n \nMemorySessions\n()\n\n\nlet\n \nsessions\n \n=\n \nSessionsMiddleware\n(\nsessions\n:\n \nmemory\n)\n\n\n\n\n\n\nThen add to the \nDroplet\n.\n\n\nlet drop = Droplet()\ndrop.middleware.append(sessions)\n\n\n\n\n\n\n\nNote: If you'd like to enable or disable the middleware based on config files, check out \nmiddleware\n.\n\n\n\n\nRequest\n\n\nAfter \nSessionMiddleware\n has been enabled, you can access the \nreq.sessions()\n method to get access to session data.\n\n\nlet\n \ndata\n \n=\n \ntry\n \nreq\n.\nsession\n().\ndata\n\n\n\n\n\n\nExample\n\n\nLet's create an example that remembers the user's name.\n\n\nStore\n\n\ndrop\n.\npost\n(\nremember\n)\n \n{\n \nreq\n \nin\n\n    \nguard\n \nlet\n \nname\n \n=\n \nreq\n.\ndata\n[\nname\n]?.\nstring\n \nelse\n \n{\n\n        \nthrow\n \nAbort\n.\nbadRequest\n\n    \n}\n\n\n    \ntry\n \nreq\n.\nsession\n().\ndata\n[\nname\n]\n \n=\n \nNode\n.\nstring\n(\nname\n)\n\n\n    \nreturn\n \nRemebered name.\n\n\n}\n\n\n\n\n\n\nOn \nPOST /remember\n, fetch a \nname\n from the request input, then store this name into the session data.\n\n\nFetch\n\n\nOn \nGET /remember\n, fetch the \nname\n from the session data and return it.\n\n\ndrop\n.\nget\n(\nremember\n)\n \n{\n \nreq\n \nin\n\n    \nguard\n \nlet\n \nname\n \n=\n \ntry\n \nreq\n.\nsession\n().\ndata\n[\nname\n]?.\nstring\n \nelse\n \n{\n\n        \nreturn\n \nthrow\n \nAbort\n.\ncustom\n(\nstatus\n:\n \n.\nbadRequest\n,\n \nmessage\n:\n \nPlease POST the name first.\n)\n\n    \n}\n\n\n    \nreturn\n \nname\n\n\n}\n\n\n\n\n\n\nCookie\n\n\nThe session will be stored using the \nvapor-session\n cookie.", 
            "title": "Sessions"
        }, 
        {
            "location": "/sessions/sessions/#sessions", 
            "text": "Sessions help you store information about a user between requests. As long as the client supports cookies, sessions are easy to create.", 
            "title": "Sessions"
        }, 
        {
            "location": "/sessions/sessions/#middleware", 
            "text": "Enable sessions on your  Droplet  by adding an instance of  SessionMiddleware .  import   Sessions  let   memory   =   MemorySessions ()  let   sessions   =   SessionsMiddleware ( sessions :   memory )   Then add to the  Droplet .  let drop = Droplet()\ndrop.middleware.append(sessions)   Note: If you'd like to enable or disable the middleware based on config files, check out  middleware .", 
            "title": "Middleware"
        }, 
        {
            "location": "/sessions/sessions/#request", 
            "text": "After  SessionMiddleware  has been enabled, you can access the  req.sessions()  method to get access to session data.  let   data   =   try   req . session (). data", 
            "title": "Request"
        }, 
        {
            "location": "/sessions/sessions/#example", 
            "text": "Let's create an example that remembers the user's name.", 
            "title": "Example"
        }, 
        {
            "location": "/sessions/sessions/#store", 
            "text": "drop . post ( remember )   {   req   in \n     guard   let   name   =   req . data [ name ]?. string   else   { \n         throw   Abort . badRequest \n     } \n\n     try   req . session (). data [ name ]   =   Node . string ( name ) \n\n     return   Remebered name.  }   On  POST /remember , fetch a  name  from the request input, then store this name into the session data.", 
            "title": "Store"
        }, 
        {
            "location": "/sessions/sessions/#fetch", 
            "text": "On  GET /remember , fetch the  name  from the session data and return it.  drop . get ( remember )   {   req   in \n     guard   let   name   =   try   req . session (). data [ name ]?. string   else   { \n         return   throw   Abort . custom ( status :   . badRequest ,   message :   Please POST the name first. ) \n     } \n\n     return   name  }", 
            "title": "Fetch"
        }, 
        {
            "location": "/sessions/sessions/#cookie", 
            "text": "The session will be stored using the  vapor-session  cookie.", 
            "title": "Cookie"
        }, 
        {
            "location": "/http/request/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\n\n\nModule: \nimport HTTP\n\n\n\n\nRequest\n\n\nThe most common part of the \nHTTP\n library we'll be interacting with is the \nRequest\n type. Here's a look at some of the most commonly used attributes in this type.\n\n\npublic\n \nvar\n \nmethod\n:\n \nMethod\n\n\npublic\n \nvar\n \nuri\n:\n \nURI\n\n\npublic\n \nvar\n \nparameters\n:\n \nNode\n\n\npublic\n \nvar\n \nheaders\n:\n \n[\nHeaderKey\n:\n \nString\n]\n\n\npublic\n \nvar\n \nbody\n:\n \nBody\n\n\npublic\n \nvar\n \ndata\n:\n \nContent\n\n\n\n\n\n\nMethod\n\n\nThe HTTP \nMethod\n associated with the \nRequest\n, ie: \nGET\n, \nPOST\n, \nPUT\n, \nPATCH\n, \nDELETE\n.\n\n\nURI\n\n\nThe associated \nURI\n of the request. We will use this to access attributes about the \nuri\n the request was sent to.\n\n\nFor example, given the following uri: \nhttp://vapor.codes/example?query=hi#fragments-too\n\n\nlet\n \nscheme\n \n=\n \nrequest\n.\nuri\n.\nscheme\n \n// http\n\n\nlet\n \nhost\n \n=\n \nrequest\n.\nuri\n.\nhost\n \n// vapor.codes\n\n\n\nlet\n \npath\n \n=\n \nrequest\n.\nuri\n.\npath\n \n// /example\n\n\nlet\n \nquery\n \n=\n \nrequest\n.\nuri\n.\nquery\n \n// query=hi\n\n\nlet\n \nfragment\n \n=\n \nrequest\n.\nuri\n.\nfragment\n \n// fragments-too\n\n\n\n\n\n\nRoute Parameters\n\n\nThe url parameters associated with the request. For example, if we have a path registered as \nhello/:name/age/:age\n, we would be able to access those in our request, like so:\n\n\nlet\n \nname\n \n=\n \nrequest\n.\nparameters\n[\nname\n]\n \n// String?\n\n\nlet\n \nage\n \n=\n \nrequest\n.\nparameters\n[\nage\n]?.\nint\n \n// Int?\n\n\n\n\n\n\nOr, to automatically throw on \nnil\n or invalid variable, you can also \nextract\n\n\nlet\n \nname\n \n=\n \ntry\n \nrequest\n.\nparameters\n.\nextract\n(\nname\n)\n \nas\n \nString\n\n\nlet\n \nage\n \n=\n \ntry\n \nrequest\n.\nparameters\n.\nextract\n(\nage\n)\n \nas\n \nInt\n\n\n\n\n\n\nThese extract functions can cast to any \nNodeInitializable\n type, including your own custom types. Make sure to check out \nNode\n for more info.\n\n\n\n\nNote: Vapor also provides type safe routing in the routing section of our docs.\n\n\n\n\nHeaders\n\n\nThese are the headers associated with the request. If you are preparing an outgoing request, this can be used to add your own keys.\n\n\nlet\n \ncontentType\n \n=\n \nrequest\n.\nheaders\n[\nContent-Type\n]\n  \n\n\n\n\n\nOr for outgoing requests:\n\n\nlet\n \nrequest\n \n=\n \nRequest\n \n...\n\n\nrequest\n.\nheaders\n[\nContent-Type\n]\n \n=\n \napplication/json\n\n\nrequest\n.\nheaders\n[\nAuthorization\n]\n \n=\n \n...\n \nmy\n \nauth\n \ntoken\n\n\n\n\n\n\nExtending Headers\n\n\nWe generally seek to improve code bases by removing stringly typed code where possible. We can add variables to the headers using generic extensions.\n\n\nextension\n \nHTTP\n.\nKeyAccessible\n \nwhere\n \nKey\n \n==\n \nHeaderKey\n,\n \nValue\n \n==\n \nString\n \n{\n\n    \nvar\n \ncustomKey\n:\n \nString\n?\n \n{\n\n      \nget\n \n{\n\n        \nreturn\n \nself\n[\nCustom-Key\n]\n\n      \n}\n\n      \nset\n \n{\n\n        \nself\n[\nCustom-Key\n]\n \n=\n \nnewValue\n\n      \n}\n\n    \n}\n\n\n}\n\n\n\n\n\n\nWith this pattern implemented, our string \n\"Custom-Key\"\n is contained in one section of our code. We can now access like this:\n\n\nlet\n \ncustomKey\n \n=\n \nrequest\n.\nheaders\n.\ncustomKey\n\n\n\n// or\n\n\n\nlet\n \nrequest\n \n=\n \n...\n\n\nrequest\n.\nheaders\n.\ncustomKey\n \n=\n \nmy custom value\n\n\n\n\n\n\nBody\n\n\nThis is the body associated with the request and represents the general data payload. You can view more about body in the associated \ndocs\n\n\nFor incoming requests, we'll often pull out the associated bytes like so:\n\n\nlet\n \nrawBytes\n \n=\n \nrequest\n.\nbody\n.\nbytes\n\n\n\n\n\n\nContent\n\n\nGenerally when we're sending or receiving requests, we're using them as a way to transport content. For this, Vapor provides a convenient \ndata\n variable associated with the request that prioritizes content in a consistent way.\n\n\nFor example, say I receive a request to \nhttp://vapor.codes?hello=world\n.\n\n\nlet\n \nworld\n \n=\n \nrequest\n.\ndata\n[\nhello\n]?.\nstring\n\n\n\n\n\n\nThis same code will work if I receive a JSON request, for example:\n\n\n{\n\n  \nhello\n:\n \nworld\n\n\n}\n\n\n\n\n\n\nWill still be accessible through data.\n\n\nlet\n \nworld\n \n=\n \nrequest\n.\ndata\n[\nhello\n]?.\nstring\n\n\n\n\n\n\n\n\nNote: Force unwrap should never be used.\n\n\n\n\nThis also applies to multi-part requests and can even be extended to new types such as XML or YAML through middleware.\n\n\nIf you'd prefer to access given types more explicitly, that's totally fine. The \ndata\n variable is purely opt-in convenience for those who want it.\n\n\nJSON\n\n\nTo access JSON directly on a given request, use the following:\n\n\nlet\n \njson\n \n=\n \nrequest\n.\njson\n[\nhello\n]\n\n\n\n\n\n\nQuery Parameters\n\n\nThe same applies to query convenience:\n\n\nlet\n \nquery\n \n=\n \nrequest\n.\nquery\n?[\nhello\n]\n  \n// String?\n\n\nlet\n \nname\n \n=\n \nrequest\n.\nquery\n?[\nname\n]?.\nstring\n \n// String?\n\n\nlet\n \nage\n \n=\n \nrequest\n.\nquery\n?[\nage\n]?.\nint\n \n// Int?\n\n\nlet\n \nrating\n \n=\n \nrequest\n.\nquery\n?[\nrating\n]?.\ndouble\n \n// Double?\n\n\n\n\n\n\nKey Paths\n\n\nKey paths work on most Vapor types that can have nested key value objects. Here's a couple examples of how to access given the following json:\n\n\n{\n\n  \nmetadata\n:\n \nsome metadata\n,\n\n  \nartists\n \n:\n \n{\n\n    \nhref\n:\n \nhttp://someurl.com\n,\n\n    \nitems\n:\n \n[\n\n      \n{\n\n        \nname\n:\n \nVan Gogh\n,\n\n      \n},\n\n      \n{\n\n        \nname\n:\n \nMozart\n\n      \n}\n\n    \n]\n\n  \n}\n\n\n}\n\n\n\n\n\n\nWe could access the data in the following ways:\n\n\nMetadata\n\n\nAccess top level values\n\n\nlet\n \ntype\n \n=\n \nrequest\n.\ndata\n[\nmetadata\n].\nstring\n \n// \nsome metadata\n\n\n\n\n\n\nItems\n\n\nAccess nested values\n\n\nlet\n \nitems\n \n=\n \nrequest\n.\ndata\n[\nartists\n,\n \nitems\n]\n \n// [[\nname\n: \nVan Gogh\n], [\nname\n: \nMozart\n]]\n\n\n\n\n\n\nMixing Arrays and Objects\n\n\nGet first artists\n\n\nlet\n \nfirst\n \n=\n \nrequest\n.\ndata\n[\nartists\n,\n \nitems\n,\n \n0\n]\n \n// [\nname\n: \nVan Gogh\n]\n\n\n\n\n\n\nArray Item\n\n\nGet key from array item\n\n\nlet\n \nfirstName\n \n=\n \nrequest\n.\ndata\n[\nartists\n,\n \nitems\n,\n \n0\n,\n \nname\n]\n \n// \nVan Gogh\n\n\n\n\n\n\nArray Comprehension\n\n\nWe can also smartly map an array of keys, for example, to just get the names of all of the artists, we could use the following\n\n\nlet\n \nnames\n \n=\n \nrequest\n.\ndata\n[\nartists\n,\n \nitems\n,\n \nname\n]\n \n// [\nVan Gogh\n, \nMozart\n]", 
            "title": "Request"
        }, 
        {
            "location": "/http/request/#request", 
            "text": "The most common part of the  HTTP  library we'll be interacting with is the  Request  type. Here's a look at some of the most commonly used attributes in this type.  public   var   method :   Method  public   var   uri :   URI  public   var   parameters :   Node  public   var   headers :   [ HeaderKey :   String ]  public   var   body :   Body  public   var   data :   Content", 
            "title": "Request"
        }, 
        {
            "location": "/http/request/#method", 
            "text": "The HTTP  Method  associated with the  Request , ie:  GET ,  POST ,  PUT ,  PATCH ,  DELETE .", 
            "title": "Method"
        }, 
        {
            "location": "/http/request/#uri", 
            "text": "The associated  URI  of the request. We will use this to access attributes about the  uri  the request was sent to.  For example, given the following uri:  http://vapor.codes/example?query=hi#fragments-too  let   scheme   =   request . uri . scheme   // http  let   host   =   request . uri . host   // vapor.codes  let   path   =   request . uri . path   // /example  let   query   =   request . uri . query   // query=hi  let   fragment   =   request . uri . fragment   // fragments-too", 
            "title": "URI"
        }, 
        {
            "location": "/http/request/#route-parameters", 
            "text": "The url parameters associated with the request. For example, if we have a path registered as  hello/:name/age/:age , we would be able to access those in our request, like so:  let   name   =   request . parameters [ name ]   // String?  let   age   =   request . parameters [ age ]?. int   // Int?   Or, to automatically throw on  nil  or invalid variable, you can also  extract  let   name   =   try   request . parameters . extract ( name )   as   String  let   age   =   try   request . parameters . extract ( age )   as   Int   These extract functions can cast to any  NodeInitializable  type, including your own custom types. Make sure to check out  Node  for more info.   Note: Vapor also provides type safe routing in the routing section of our docs.", 
            "title": "Route Parameters"
        }, 
        {
            "location": "/http/request/#headers", 
            "text": "These are the headers associated with the request. If you are preparing an outgoing request, this can be used to add your own keys.  let   contentType   =   request . headers [ Content-Type ]     Or for outgoing requests:  let   request   =   Request   ...  request . headers [ Content-Type ]   =   application/json  request . headers [ Authorization ]   =   ...   my   auth   token", 
            "title": "Headers"
        }, 
        {
            "location": "/http/request/#extending-headers", 
            "text": "We generally seek to improve code bases by removing stringly typed code where possible. We can add variables to the headers using generic extensions.  extension   HTTP . KeyAccessible   where   Key   ==   HeaderKey ,   Value   ==   String   { \n     var   customKey :   String ?   { \n       get   { \n         return   self [ Custom-Key ] \n       } \n       set   { \n         self [ Custom-Key ]   =   newValue \n       } \n     }  }   With this pattern implemented, our string  \"Custom-Key\"  is contained in one section of our code. We can now access like this:  let   customKey   =   request . headers . customKey  // or  let   request   =   ...  request . headers . customKey   =   my custom value", 
            "title": "Extending Headers"
        }, 
        {
            "location": "/http/request/#body", 
            "text": "This is the body associated with the request and represents the general data payload. You can view more about body in the associated  docs  For incoming requests, we'll often pull out the associated bytes like so:  let   rawBytes   =   request . body . bytes", 
            "title": "Body"
        }, 
        {
            "location": "/http/request/#content", 
            "text": "Generally when we're sending or receiving requests, we're using them as a way to transport content. For this, Vapor provides a convenient  data  variable associated with the request that prioritizes content in a consistent way.  For example, say I receive a request to  http://vapor.codes?hello=world .  let   world   =   request . data [ hello ]?. string   This same code will work if I receive a JSON request, for example:  { \n   hello :   world  }   Will still be accessible through data.  let   world   =   request . data [ hello ]?. string    Note: Force unwrap should never be used.   This also applies to multi-part requests and can even be extended to new types such as XML or YAML through middleware.  If you'd prefer to access given types more explicitly, that's totally fine. The  data  variable is purely opt-in convenience for those who want it.", 
            "title": "Content"
        }, 
        {
            "location": "/http/request/#json", 
            "text": "To access JSON directly on a given request, use the following:  let   json   =   request . json [ hello ]", 
            "title": "JSON"
        }, 
        {
            "location": "/http/request/#query-parameters", 
            "text": "The same applies to query convenience:  let   query   =   request . query ?[ hello ]    // String?  let   name   =   request . query ?[ name ]?. string   // String?  let   age   =   request . query ?[ age ]?. int   // Int?  let   rating   =   request . query ?[ rating ]?. double   // Double?", 
            "title": "Query Parameters"
        }, 
        {
            "location": "/http/request/#key-paths", 
            "text": "Key paths work on most Vapor types that can have nested key value objects. Here's a couple examples of how to access given the following json:  { \n   metadata :   some metadata , \n   artists   :   { \n     href :   http://someurl.com , \n     items :   [ \n       { \n         name :   Van Gogh , \n       }, \n       { \n         name :   Mozart \n       } \n     ] \n   }  }   We could access the data in the following ways:", 
            "title": "Key Paths"
        }, 
        {
            "location": "/http/request/#metadata", 
            "text": "Access top level values  let   type   =   request . data [ metadata ]. string   //  some metadata", 
            "title": "Metadata"
        }, 
        {
            "location": "/http/request/#items", 
            "text": "Access nested values  let   items   =   request . data [ artists ,   items ]   // [[ name :  Van Gogh ], [ name :  Mozart ]]", 
            "title": "Items"
        }, 
        {
            "location": "/http/request/#mixing-arrays-and-objects", 
            "text": "Get first artists  let   first   =   request . data [ artists ,   items ,   0 ]   // [ name :  Van Gogh ]", 
            "title": "Mixing Arrays and Objects"
        }, 
        {
            "location": "/http/request/#array-item", 
            "text": "Get key from array item  let   firstName   =   request . data [ artists ,   items ,   0 ,   name ]   //  Van Gogh", 
            "title": "Array Item"
        }, 
        {
            "location": "/http/request/#array-comprehension", 
            "text": "We can also smartly map an array of keys, for example, to just get the names of all of the artists, we could use the following  let   names   =   request . data [ artists ,   items ,   name ]   // [ Van Gogh ,  Mozart ]", 
            "title": "Array Comprehension"
        }, 
        {
            "location": "/http/response/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\n\n\nModule: \nimport HTTP\n\n\n\n\nResponse\n\n\nWhen building endpoints, we'll often be returning responses for requests. If we're making outgoing requests, we'll be receiving them.\n\n\npublic\n \nlet\n \nstatus\n:\n \nStatus\n\n\npublic\n \nvar\n \nheaders\n:\n \n[\nHeaderKey\n:\n \nString\n]\n\n\npublic\n \nvar\n \nbody\n:\n \nBody\n\n\npublic\n \nvar\n \ndata\n:\n \nContent\n\n\n\n\n\n\nStatus\n\n\nThe http status associated with the event, for example \n.ok\n == 200 ok.\n\n\nHeaders\n\n\nThese are the headers associated with the request. If you are preparing an outgoing response, this can be used to add your own keys.\n\n\nlet\n \ncontentType\n \n=\n \nresponse\n.\nheaders\n[\nContent-Type\n]\n  \n\n\n\n\n\nOr for outgoing response:\n\n\nlet\n \nresponse\n \n=\n \nresponse\n \n...\n\n\nresponse\n.\nheaders\n[\nContent-Type\n]\n \n=\n \napplication/json\n\n\nresponse\n.\nheaders\n[\nAuthorization\n]\n \n=\n \n...\n \nmy\n \nauth\n \ntoken\n\n\n\n\n\n\nExtending Headers\n\n\nWe generally seek to improve code bases by removing stringly typed code where possible. We can add variables to the headers using generic extensions.\n\n\nextension\n \nHTTP\n.\nKeyAccessible\n \nwhere\n \nKey\n \n==\n \nHeaderKey\n,\n \nValue\n \n==\n \nString\n \n{\n\n    \nvar\n \ncustomKey\n:\n \nString\n?\n \n{\n\n      \nget\n \n{\n\n        \nreturn\n \nself\n[\nCustom-Key\n]\n\n      \n}\n\n      \nset\n \n{\n\n        \nself\n[\nCustom-Key\n]\n \n=\n \nnewValue\n\n      \n}\n\n    \n}\n\n\n}\n\n\n\n\n\n\nWith this pattern implemented, our string \n\"Custom-Key\"\n is contained in one section of our code. We can now access like this:\n\n\nlet\n \ncustomKey\n \n=\n \nresponse\n.\nheaders\n.\ncustomKey\n\n\n\n// or\n\n\n\nlet\n \nrequest\n \n=\n \n...\n\n\nresponse\n.\nheaders\n.\ncustomKey\n \n=\n \nmy custom value\n\n\n\n\n\n\nBody\n\n\nThis is the body associated with the response and represents the general data payload. You can view more about body in the associated \ndocs\n\n\nFor responses, the body is most commonly set at initialization. With two main types.\n\n\nBodyRepresentable\n\n\nThings that can be converted to bytes, ie:\n\n\nlet\n \nresponse\n \n=\n \nResponse\n(\nstatus\n:\n \n.\nok\n,\n \nbody\n:\n \nsome string\n)\n\n\n\n\n\n\nIn the above example, the \nString\n will be automatically converted to a body. Your own types can do this as well.\n\n\nBytes Directly\n\n\nIf we already have our bytes array, we can pass it into the body like so:\n\n\nlet\n \nresponse\n \n=\n \nResponse\n(\nstatus\n:\n \n.\nok\n,\n \nbody\n:\n \n.\ndata\n(\nmyArrayOfBytes\n))\n\n\n\n\n\n\nChunked\n\n\nTo send an \nHTTP.Response\n in chunks, we can pass a closure that we'll use to send our response body in parts.\n\n\nlet\n \nresponse\n \n=\n \nResponse\n(\nstatus\n:\n \n.\nok\n)\n \n{\n \nchunker\n \nin\n\n  \nfor\n \nname\n \nin\n \n[\njoe\n,\n \npam\n,\n \ncheryl\n]\n \n{\n\n      \nsleep\n(\n1\n)\n\n      \ntry\n \nchunker\n.\nsend\n(\nname\n)\n\n  \n}\n\n\n  \ntry\n \nchunker\n.\nclose\n()\n\n\n}\n\n\n\n\n\n\n\n\nMake sure to call \nclose()\n before the chunker leaves scope.\n\n\n\n\nContent\n\n\nWe can access content the same we do in a \nrequest\n. This most commonly applies to outgoing requests.\n\n\nlet\n \npokemonResponse\n \n=\n \ntry\n \ndrop\n.\nclient\n.\nget\n(\nhttp://pokeapi.co/api/v2/pokemon/\n)\n\n\nlet\n \nnames\n \n=\n \npokemonResponse\n.\ndata\n[\nresults\n,\n \nname\n]?.\narray\n\n\n\n\n\n\nJSON\n\n\nTo access JSON directly on a given response, use the following:\n\n\nlet\n \njson\n \n=\n \nrequest\n.\nresponse\n[\nhello\n]\n\n\n\n\n\n\nKey Paths\n\n\nFor more on KeyPaths, visit \nhere", 
            "title": "Response"
        }, 
        {
            "location": "/http/response/#response", 
            "text": "When building endpoints, we'll often be returning responses for requests. If we're making outgoing requests, we'll be receiving them.  public   let   status :   Status  public   var   headers :   [ HeaderKey :   String ]  public   var   body :   Body  public   var   data :   Content", 
            "title": "Response"
        }, 
        {
            "location": "/http/response/#status", 
            "text": "The http status associated with the event, for example  .ok  == 200 ok.", 
            "title": "Status"
        }, 
        {
            "location": "/http/response/#headers", 
            "text": "These are the headers associated with the request. If you are preparing an outgoing response, this can be used to add your own keys.  let   contentType   =   response . headers [ Content-Type ]     Or for outgoing response:  let   response   =   response   ...  response . headers [ Content-Type ]   =   application/json  response . headers [ Authorization ]   =   ...   my   auth   token", 
            "title": "Headers"
        }, 
        {
            "location": "/http/response/#extending-headers", 
            "text": "We generally seek to improve code bases by removing stringly typed code where possible. We can add variables to the headers using generic extensions.  extension   HTTP . KeyAccessible   where   Key   ==   HeaderKey ,   Value   ==   String   { \n     var   customKey :   String ?   { \n       get   { \n         return   self [ Custom-Key ] \n       } \n       set   { \n         self [ Custom-Key ]   =   newValue \n       } \n     }  }   With this pattern implemented, our string  \"Custom-Key\"  is contained in one section of our code. We can now access like this:  let   customKey   =   response . headers . customKey  // or  let   request   =   ...  response . headers . customKey   =   my custom value", 
            "title": "Extending Headers"
        }, 
        {
            "location": "/http/response/#body", 
            "text": "This is the body associated with the response and represents the general data payload. You can view more about body in the associated  docs  For responses, the body is most commonly set at initialization. With two main types.", 
            "title": "Body"
        }, 
        {
            "location": "/http/response/#bodyrepresentable", 
            "text": "Things that can be converted to bytes, ie:  let   response   =   Response ( status :   . ok ,   body :   some string )   In the above example, the  String  will be automatically converted to a body. Your own types can do this as well.", 
            "title": "BodyRepresentable"
        }, 
        {
            "location": "/http/response/#bytes-directly", 
            "text": "If we already have our bytes array, we can pass it into the body like so:  let   response   =   Response ( status :   . ok ,   body :   . data ( myArrayOfBytes ))", 
            "title": "Bytes Directly"
        }, 
        {
            "location": "/http/response/#chunked", 
            "text": "To send an  HTTP.Response  in chunks, we can pass a closure that we'll use to send our response body in parts.  let   response   =   Response ( status :   . ok )   {   chunker   in \n   for   name   in   [ joe ,   pam ,   cheryl ]   { \n       sleep ( 1 ) \n       try   chunker . send ( name ) \n   } \n\n   try   chunker . close ()  }    Make sure to call  close()  before the chunker leaves scope.", 
            "title": "Chunked"
        }, 
        {
            "location": "/http/response/#content", 
            "text": "We can access content the same we do in a  request . This most commonly applies to outgoing requests.  let   pokemonResponse   =   try   drop . client . get ( http://pokeapi.co/api/v2/pokemon/ )  let   names   =   pokemonResponse . data [ results ,   name ]?. array", 
            "title": "Content"
        }, 
        {
            "location": "/http/response/#json", 
            "text": "To access JSON directly on a given response, use the following:  let   json   =   request . response [ hello ]", 
            "title": "JSON"
        }, 
        {
            "location": "/http/response/#key-paths", 
            "text": "For more on KeyPaths, visit  here", 
            "title": "Key Paths"
        }, 
        {
            "location": "/http/middleware/", 
            "text": "Middleware\n\n\nMiddleware is an essential part of any modern web framework. It allows you to modify requests and responses as they pass between the client and your server.\n\n\nYou can imagine middleware as a chain of logic connecting your server to the client requesting your web app.\n\n\nVersion Middleware\n\n\nAs an example, let's create a middleware that will add the version of our API to each response. The middleware would look something like this:\n\n\nimport\n \nHTTP\n\n\n\nfinal\n \nclass\n \nVersionMiddleware\n:\n \nMiddleware\n \n{\n\n    \nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n,\n \nchainingTo\n \nnext\n:\n \nResponder\n)\n \nthrows\n \n-\n \nResponse\n \n{\n\n        \nlet\n \nresponse\n \n=\n \ntry\n \nnext\n.\nrespond\n(\nto\n:\n \nrequest\n)\n\n\n        \nresponse\n.\nheaders\n[\nVersion\n]\n \n=\n \nAPI v1.0\n\n\n        \nreturn\n \nresponse\n\n    \n}\n\n\n}\n\n\n\n\n\n\nWe then supply this middleware to our \nDroplet\n.\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\ndrop\n.\nmiddleware\n.\nappend\n(\nVersionMiddleware\n())\n\n\n\n\n\n\nYou can imagine our \nVersionMiddleware\n sitting in the middle of a chain that connects the client and our server. Every request and response that hits our server must go through this chain of middleware.\n\n\n\n\nBreakdown\n\n\nLet's break down the middleware line by line.\n\n\nlet\n \nresponse\n \n=\n \ntry\n \nnext\n.\nrespond\n(\nto\n:\n \nrequest\n)\n\n\n\n\n\n\nSince the \nVersionMiddleware\n in this example is not interested in modifying the request, we immediately ask the next middleware in the chain to respond to the request. This goes all the way down the chain to the \nDroplet\n and comes back with the response that should be sent to the client.\n\n\nresponse\n.\nheaders\n[\nVersion\n]\n \n=\n \nAPI v1.0\n\n\n\n\n\n\nWe then \nmodify\n the response to contain a Version header.\n\n\nreturn\n \nresponse\n\n\n\n\n\n\nThe response is returned and will chain back up any remaining middleware and back to the client.\n\n\nRequest\n\n\nThe middleware can also modify or interact with the request.\n\n\nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n,\n \nchainingTo\n \nnext\n:\n \nResponder\n)\n \nthrows\n \n-\n \nResponse\n \n{\n\n    \nguard\n \nrequest\n.\ncookies\n[\ntoken\n]\n \n==\n \nsecret\n \nelse\n \n{\n\n        \nthrow\n \nAbort\n(.\nbadRequest\n)\n\n    \n}\n\n\n    \nreturn\n \ntry\n \nnext\n.\nrespond\n(\nto\n:\n \nrequest\n)\n\n\n}\n\n\n\n\n\n\nThis middleware will require that the request has a cookie named \ntoken\n that equals \nsecret\n or else the request will be aborted.\n\n\nErrors\n\n\nMiddleware is the perfect place to catch errors thrown from anywhere in your application. When you let the middleware catch errors, you can remove a lot of duplicated logic from your route closures. Take a look at the following example:\n\n\nenum\n \nFooError\n:\n \nError\n \n{\n\n    \ncase\n \nfooServiceUnavailable\n\n\n}\n\n\n\n\n\n\nSay there is a custom error that either you defined or one of the APIs you are using \nthrows\n. This error must be caught when thrown, or else it will end up as an internal server error (500) which may be unexpected to a user. The most obvious solution is to catch the error in the route closure.\n\n\napp\n.\nget\n(\nfoo\n)\n \n{\n \nrequest\n \nin\n\n    \nlet\n \nfoo\n:\n \nFoo\n\n    \ndo\n \n{\n\n        \nfoo\n \n=\n \ntry\n \ngetFooFromService\n()\n\n    \n}\n \ncatch\n \n{\n\n        \nthrow\n \nAbort\n(.\nbadRequest\n)\n\n    \n}\n\n\n    \n// continue with Foo object\n\n\n}\n\n\n\n\n\n\nThis solution works, but it would get repetitive if multiple routes need to handle the error. Luckily, this error could be caught in a middleware instead.\n\n\nfinal\n \nclass\n \nFooErrorMiddleware\n:\n \nMiddleware\n \n{\n\n    \nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n,\n \nchainingTo\n \nnext\n:\n \nResponder\n)\n \nthrows\n \n-\n \nResponse\n \n{\n\n        \ndo\n \n{\n\n            \nreturn\n \ntry\n \nnext\n.\nrespond\n(\nto\n:\n \nrequest\n)\n\n        \n}\n \ncatch\n \nFooError\n.\nfooServiceUnavailable\n \n{\n\n            \nthrow\n \nAbort\n(\n\n                \n.\nbadRequest\n,\n\n                \nreason\n:\n \nSorry, we were unable to query the Foo service.\n\n            \n)\n\n        \n}\n\n    \n}\n\n\n}\n\n\n\n\n\n\nWe just need to append this middleware to the \nDroplet\n.\n\n\ndrop\n.\nmiddleware\n.\nappend\n(\nFooErrorMiddleware\n())\n\n\n\n\n\n\nNow our route closures look a lot better and we don't have to worry about code duplication.\n\n\napp\n.\nget\n(\nfoo\n)\n \n{\n \nrequest\n \nin\n\n    \nlet\n \nfoo\n \n=\n \ntry\n \ngetFooFromService\n()\n\n\n    \n// continue with Foo object\n\n\n}\n\n\n\n\n\n\nRoute Groups\n\n\nFor more granularity, Middleware can be applied to specific route groups.\n\n\nlet\n \nauthed\n \n=\n \ndrop\n.\ngrouped\n(\nAuthMiddleware\n())\n\n\nauthed\n.\nget\n(\nsecure\n)\n \n{\n \nreq\n \nin\n\n    \nreturn\n \nSecrets\n.\nall\n().\nmakeJSON\n()\n\n\n}\n\n\n\n\n\n\nAnything added to the \nauthed\n group must pass through \nAuthMiddleware\n. Because of this, we can assume all traffic to \n/secure\n has been authorized. Learn more in \nRouting\n.\n\n\nConfiguration\n\n\nAppending middleware to the \ndrop.middleware\n array is the simplest way to add middleware--it will be used every time the application starts.\n\n\nYou can also use the \nconfiguration\n files to enabled or disable middleware for more control. This is especially useful if you have middleware that should, for example, run only in production.\n\n\nAppending configurable middleware looks like the following:\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\ndrop\n.\naddConfigurable\n(\nmiddleware\n:\n \nmyMiddleware\n,\n \nname\n:\n \nmy-middleware\n)\n\n\n\n\n\n\nThen, in the \nConfig/droplet.json\n file, add \nmy-middleware\n to the appropriate \nmiddleware\n array.\n\n\n{\n\n    \n...\n\n    \nmiddleware\n:\n \n{\n\n        \nserver\n:\n \n[\n\n            \n...\n\n            \nmy-middleware\n,\n\n            \n...\n\n        \n],\n\n        \nclient\n:\n \n[\n\n            \n...\n\n        \n]\n\n    \n},\n\n    \n...\n\n\n}\n\n\n\n\n\n\nIf the name of the added middleware appears in the \nserver\n array for the loaded configuration, it will be added to the server's middleware when the application boots.\n\n\nLikewise, if the middleware appears in the \nclient\n array for the loaded configuration, it will be added to the client's middleware.\n\n\nOne middleware can be appended to both the Client and the Server, and can be added multiple times. The ordering of middleware is respected.\n\n\nAdvanced\n\n\nExtensions\n\n\nMiddleware pairs great with request/response extensions and storage. This example shows you how to dynamically return either HTML or JSON responses for a Model depending on the type of client.\n\n\nMiddleware\n\n\nfinal\n \nclass\n \nPokemonMiddleware\n:\n \nMiddleware\n \n{\n\n    \nlet\n \nview\n:\n \nViewProtocol\n\n    \ninit\n(\n_\n \nview\n:\n \nViewProtocol\n)\n \n{\n\n        \nself\n.\nview\n \n=\n \nview\n\n    \n}\n\n\n    \nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n,\n \nchainingTo\n \nnext\n:\n \nResponder\n)\n \nthrows\n \n-\n \nResponse\n \n{\n\n        \nlet\n \nresponse\n \n=\n \ntry\n \nnext\n.\nrespond\n(\nto\n:\n \nrequest\n)\n\n\n        \nif\n \nlet\n \npokemon\n \n=\n \nresponse\n.\npokemon\n \n{\n\n            \nif\n \nrequest\n.\naccept\n.\nprefers\n(\nhtml\n)\n \n{\n\n                \nresponse\n.\nview\n \n=\n \ntry\n \nview\n.\nmake\n(\npokemon.mustache\n,\n \npokemon\n)\n\n            \n}\n \nelse\n \n{\n\n                \nresponse\n.\njson\n \n=\n \ntry\n \npokemon\n.\nmakeJSON\n()\n\n            \n}\n\n        \n}\n\n\n        \nreturn\n \nresponse\n\n    \n}\n\n\n}\n\n\n\n\n\n\nResponse\n\n\nAnd the extension to \nResponse\n.\n\n\nextension\n \nResponse\n \n{\n\n    \nvar\n \npokemon\n:\n \nPokemon\n?\n \n{\n\n        \nget\n \n{\n \nreturn\n \nstorage\n[\npokemon\n]\n \nas\n?\n \nPokemon\n \n}\n\n        \nset\n \n{\n \nstorage\n[\npokemon\n]\n \n=\n \nnewValue\n \n}\n\n    \n}\n\n\n}\n\n\n\n\n\n\nIn this example, we added a new property to response capable of holding a Pok\u00e9mon object. If the middleware finds a response with one of these Pok\u00e9mon objects, it will dynamically check whether the client prefers HTML. If the client is a browser like Safari and prefers HTML, it will return a Mustache view. If the client does not prefer HTML, it will return JSON.\n\n\nUsage\n\n\nYour closures can now look something like this:\n\n\nimport\n \nVapor\n\n\nimport\n \nHTTP\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\nlet\n \npokemonMiddleware\n \n=\n \nPokemonMiddleware\n(\ndrop\n.\nview\n)\n\n\ndrop\n.\nmiddleware\n.\nappend\n(\npokemonMiddleware\n)\n\n\n\ndrop\n.\nget\n(\npokemon\n,\n \nPokemon\n.\nself\n)\n \n{\n \nrequest\n,\n \npokemon\n \nin\n\n    \nlet\n \nresponse\n \n=\n \nResponse\n()\n\n    \nresponse\n.\npokemon\n \n=\n \npokemon\n\n    \nreturn\n \nresponse\n\n\n}\n\n\n\n\n\n\nResponse Representable\n\n\nIf you want to go a step further, you can make \nPokemon\n conform to \nResponseRepresentable\n.\n\n\nimport\n \nHTTP\n\n\n\nextension\n \nPokemon\n:\n \nResponseRepresentable\n \n{\n\n    \nfunc\n \nmakeResponse\n()\n \nthrows\n \n-\n \nResponse\n \n{\n\n        \nlet\n \nresponse\n \n=\n \nResponse\n()\n\n        \nresponse\n.\npokemon\n \n=\n \nself\n\n        \nreturn\n \nresponse\n\n    \n}\n\n\n}\n\n\n\n\n\n\nNow your route closures are greatly simplified.\n\n\ndrop\n.\nget\n(\npokemon\n,\n \nPokemon\n.\nself\n)\n \n{\n \nrequest\n,\n \npokemon\n \nin\n\n    \nreturn\n \npokemon\n\n\n}\n\n\n\n\n\n\nMiddleware is incredibly powerful. Combined with extensions, it allows you to add functionality that feels native to the framework.", 
            "title": "Middleware"
        }, 
        {
            "location": "/http/middleware/#middleware", 
            "text": "Middleware is an essential part of any modern web framework. It allows you to modify requests and responses as they pass between the client and your server.  You can imagine middleware as a chain of logic connecting your server to the client requesting your web app.", 
            "title": "Middleware"
        }, 
        {
            "location": "/http/middleware/#version-middleware", 
            "text": "As an example, let's create a middleware that will add the version of our API to each response. The middleware would look something like this:  import   HTTP  final   class   VersionMiddleware :   Middleware   { \n     func   respond ( to   request :   Request ,   chainingTo   next :   Responder )   throws   -   Response   { \n         let   response   =   try   next . respond ( to :   request ) \n\n         response . headers [ Version ]   =   API v1.0 \n\n         return   response \n     }  }   We then supply this middleware to our  Droplet .  let   drop   =   Droplet ()  drop . middleware . append ( VersionMiddleware ())   You can imagine our  VersionMiddleware  sitting in the middle of a chain that connects the client and our server. Every request and response that hits our server must go through this chain of middleware.", 
            "title": "Version Middleware"
        }, 
        {
            "location": "/http/middleware/#breakdown", 
            "text": "Let's break down the middleware line by line.  let   response   =   try   next . respond ( to :   request )   Since the  VersionMiddleware  in this example is not interested in modifying the request, we immediately ask the next middleware in the chain to respond to the request. This goes all the way down the chain to the  Droplet  and comes back with the response that should be sent to the client.  response . headers [ Version ]   =   API v1.0   We then  modify  the response to contain a Version header.  return   response   The response is returned and will chain back up any remaining middleware and back to the client.", 
            "title": "Breakdown"
        }, 
        {
            "location": "/http/middleware/#request", 
            "text": "The middleware can also modify or interact with the request.  func   respond ( to   request :   Request ,   chainingTo   next :   Responder )   throws   -   Response   { \n     guard   request . cookies [ token ]   ==   secret   else   { \n         throw   Abort (. badRequest ) \n     } \n\n     return   try   next . respond ( to :   request )  }   This middleware will require that the request has a cookie named  token  that equals  secret  or else the request will be aborted.", 
            "title": "Request"
        }, 
        {
            "location": "/http/middleware/#errors", 
            "text": "Middleware is the perfect place to catch errors thrown from anywhere in your application. When you let the middleware catch errors, you can remove a lot of duplicated logic from your route closures. Take a look at the following example:  enum   FooError :   Error   { \n     case   fooServiceUnavailable  }   Say there is a custom error that either you defined or one of the APIs you are using  throws . This error must be caught when thrown, or else it will end up as an internal server error (500) which may be unexpected to a user. The most obvious solution is to catch the error in the route closure.  app . get ( foo )   {   request   in \n     let   foo :   Foo \n     do   { \n         foo   =   try   getFooFromService () \n     }   catch   { \n         throw   Abort (. badRequest ) \n     } \n\n     // continue with Foo object  }   This solution works, but it would get repetitive if multiple routes need to handle the error. Luckily, this error could be caught in a middleware instead.  final   class   FooErrorMiddleware :   Middleware   { \n     func   respond ( to   request :   Request ,   chainingTo   next :   Responder )   throws   -   Response   { \n         do   { \n             return   try   next . respond ( to :   request ) \n         }   catch   FooError . fooServiceUnavailable   { \n             throw   Abort ( \n                 . badRequest , \n                 reason :   Sorry, we were unable to query the Foo service. \n             ) \n         } \n     }  }   We just need to append this middleware to the  Droplet .  drop . middleware . append ( FooErrorMiddleware ())   Now our route closures look a lot better and we don't have to worry about code duplication.  app . get ( foo )   {   request   in \n     let   foo   =   try   getFooFromService () \n\n     // continue with Foo object  }", 
            "title": "Errors"
        }, 
        {
            "location": "/http/middleware/#route-groups", 
            "text": "For more granularity, Middleware can be applied to specific route groups.  let   authed   =   drop . grouped ( AuthMiddleware ())  authed . get ( secure )   {   req   in \n     return   Secrets . all (). makeJSON ()  }   Anything added to the  authed  group must pass through  AuthMiddleware . Because of this, we can assume all traffic to  /secure  has been authorized. Learn more in  Routing .", 
            "title": "Route Groups"
        }, 
        {
            "location": "/http/middleware/#configuration", 
            "text": "Appending middleware to the  drop.middleware  array is the simplest way to add middleware--it will be used every time the application starts.  You can also use the  configuration  files to enabled or disable middleware for more control. This is especially useful if you have middleware that should, for example, run only in production.  Appending configurable middleware looks like the following:  let   drop   =   Droplet ()  drop . addConfigurable ( middleware :   myMiddleware ,   name :   my-middleware )   Then, in the  Config/droplet.json  file, add  my-middleware  to the appropriate  middleware  array.  { \n     ... \n     middleware :   { \n         server :   [ \n             ... \n             my-middleware , \n             ... \n         ], \n         client :   [ \n             ... \n         ] \n     }, \n     ...  }   If the name of the added middleware appears in the  server  array for the loaded configuration, it will be added to the server's middleware when the application boots.  Likewise, if the middleware appears in the  client  array for the loaded configuration, it will be added to the client's middleware.  One middleware can be appended to both the Client and the Server, and can be added multiple times. The ordering of middleware is respected.", 
            "title": "Configuration"
        }, 
        {
            "location": "/http/middleware/#advanced", 
            "text": "", 
            "title": "Advanced"
        }, 
        {
            "location": "/http/middleware/#extensions", 
            "text": "Middleware pairs great with request/response extensions and storage. This example shows you how to dynamically return either HTML or JSON responses for a Model depending on the type of client.", 
            "title": "Extensions"
        }, 
        {
            "location": "/http/middleware/#middleware_1", 
            "text": "final   class   PokemonMiddleware :   Middleware   { \n     let   view :   ViewProtocol \n     init ( _   view :   ViewProtocol )   { \n         self . view   =   view \n     } \n\n     func   respond ( to   request :   Request ,   chainingTo   next :   Responder )   throws   -   Response   { \n         let   response   =   try   next . respond ( to :   request ) \n\n         if   let   pokemon   =   response . pokemon   { \n             if   request . accept . prefers ( html )   { \n                 response . view   =   try   view . make ( pokemon.mustache ,   pokemon ) \n             }   else   { \n                 response . json   =   try   pokemon . makeJSON () \n             } \n         } \n\n         return   response \n     }  }", 
            "title": "Middleware"
        }, 
        {
            "location": "/http/middleware/#response", 
            "text": "And the extension to  Response .  extension   Response   { \n     var   pokemon :   Pokemon ?   { \n         get   {   return   storage [ pokemon ]   as ?   Pokemon   } \n         set   {   storage [ pokemon ]   =   newValue   } \n     }  }   In this example, we added a new property to response capable of holding a Pok\u00e9mon object. If the middleware finds a response with one of these Pok\u00e9mon objects, it will dynamically check whether the client prefers HTML. If the client is a browser like Safari and prefers HTML, it will return a Mustache view. If the client does not prefer HTML, it will return JSON.", 
            "title": "Response"
        }, 
        {
            "location": "/http/middleware/#usage", 
            "text": "Your closures can now look something like this:  import   Vapor  import   HTTP  let   drop   =   try   Droplet ()  let   pokemonMiddleware   =   PokemonMiddleware ( drop . view )  drop . middleware . append ( pokemonMiddleware )  drop . get ( pokemon ,   Pokemon . self )   {   request ,   pokemon   in \n     let   response   =   Response () \n     response . pokemon   =   pokemon \n     return   response  }", 
            "title": "Usage"
        }, 
        {
            "location": "/http/middleware/#response-representable", 
            "text": "If you want to go a step further, you can make  Pokemon  conform to  ResponseRepresentable .  import   HTTP  extension   Pokemon :   ResponseRepresentable   { \n     func   makeResponse ()   throws   -   Response   { \n         let   response   =   Response () \n         response . pokemon   =   self \n         return   response \n     }  }   Now your route closures are greatly simplified.  drop . get ( pokemon ,   Pokemon . self )   {   request ,   pokemon   in \n     return   pokemon  }   Middleware is incredibly powerful. Combined with extensions, it allows you to add functionality that feels native to the framework.", 
            "title": "Response Representable"
        }, 
        {
            "location": "/http/body/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\n\n\nModule: \nimport HTTP\n\n\n\n\nBody\n\n\nThe \nHTTP.Body\n represents the payload of an \nHTTP.Message\n, and is used to pass the underlying data. Some examples of this in practice would be \nJSON\n, \nHTML\n text, or the bytes of an image. Let's look at the implementation:\n\n\npublic\n \nenum\n \nBody\n \n{\n\n    \ncase\n \ndata\n(\nBytes\n)\n\n    \ncase\n \nchunked\n((\nChunkStream\n)\n \nthrows\n \n-\n \nVoid\n)\n\n\n}\n\n\n\n\n\n\nData Case\n\n\nThe \ndata\n case is by far the most common use for a \nBody\n in an \nHTTP.Message\n. It is simply an array of bytes. The serialization protocol or type associated with these bytes is usually defined by the \nContent-Type\n header. Let's look at some examples.\n\n\nApplication/JSON\n\n\nIf our \nContent-Type\n header contains \napplication/json\n, then the underlying bytes represent serialized JSON.\n\n\nif\n \nlet\n \ncontentType\n \n=\n \nreq\n.\nheaders\n[\nContent-Type\n],\n \ncontentType\n.\ncontains\n(\napplication/json\n),\n \nlet\n \nbytes\n \n=\n \nreq\n.\nbody\n.\nbytes\n \n{\n\n  \nlet\n \njson\n \n=\n \ntry\n \nJSON\n(\nbytes\n:\n \nbytes\n)\n\n  \nprint\n(\nGot JSON: \n\\(\njson\n)\n)\n\n\n}\n\n\n\n\n\n\nImage/PNG\n\n\nIf our \nContent-Type\n contains \nimage/png\n, then the underlying bytes represent an encoded png.\n\n\nif\n \nlet\n \ncontentType\n \n=\n \nreq\n.\nheaders\n[\nContent-Type\n],\n \ncontentType\n.\ncontains\n(\nimage/png\n),\n \nlet\n \nbytes\n \n=\n \nreq\n.\nbody\n.\nbytes\n \n{\n\n  \ntry\n \ndatabase\n.\nsave\n(\nimage\n:\n \nbytes\n)\n\n\n}\n\n\n\n\n\n\nChunked Case\n\n\nThe \nchunked\n case only applies to outgoing \nHTTP.Message\ns in Vapor. It is traditionally a responder's role to collect an entire chunked encoding before passing it on. We can use this to send a body asynchronously.\n\n\nlet\n \nbody\n:\n \nBody\n \n=\n \nBody\n.\nchunked\n(\nsender\n)\n\n\nreturn\n \nResponse\n(\nstatus\n:\n \n.\nok\n,\n \nbody\n:\n \nbody\n)\n\n\n\n\n\n\nWe can implement this manually, or use Vapor's built in convenience initializer for chunked bodies:\n\n\nreturn\n \nResponse\n(\nstatus\n:\n \n.\nok\n)\n \n{\n \nchunker\n \nin\n\n  \nfor\n \nname\n \nin\n \n[\njoe\n,\n \npam\n,\n \ncheryl\n]\n \n{\n\n      \nsleep\n(\n1\n)\n\n      \ntry\n \nchunker\n.\nsend\n(\nname\n)\n\n  \n}\n\n\n  \ntry\n \nchunker\n.\nclose\n()\n\n\n}\n\n\n\n\n\n\n\n\nMake sure to call \nclose()\n before the chunker leaves scope.\n\n\n\n\nBodyRepresentable\n\n\nIn addition to the concrete \nBody\n type, as is common in Vapor, we also have wide support for \nBodyRepresentable\n. This means objects that we're commonly converting to \nBody\n type can be used interchangeably. For example:\n\n\nreturn\n \nResponse\n(\nbody\n:\n \nHello, World!\n)\n\n\n\n\n\n\nIn the above example, string is converted to bytes and added to the body.\n\n\n\n\nIn practice, it is better to use \nreturn \"Hello, World!\"\n. Vapor will automatically be able to set the \nContent-Type\n to appropriate values.\n\n\n\n\nLet's look at how it's implemented:\n\n\npublic\n \nprotocol\n \nBodyRepresentable\n \n{\n\n    \nfunc\n \nmakeBody\n()\n \n-\n \nBody\n\n\n}\n\n\n\n\n\n\nCustom\n\n\nWe can conform our own types to this as well where applicable. Let's pretend we have a custom data type, \n.vpr\n. Let's conform our \nVPR\n file type model:\n\n\nextension\n \nVPRFile\n:\n \nHTTP\n.\nBodyRepresentable\n \n{\n\n  \nfunc\n \nmakeBody\n()\n \n-\n \nBody\n \n{\n\n    \n// collect bytes\n\n    \nreturn\n \n.\ndata\n(\nbytes\n)\n\n  \n}\n\n\n}\n\n\n\n\n\n\n\n\nYou may have noticed above, that the protocol throws, but our implementation does not. This is completely valid in Swift and will allow you to not throw if you're ever calling the function manually.\n\n\n\n\nNow we're able to include our \nVPR\n file directly in our \nResponses\n.\n\n\ndrop\n.\nget\n(\nfiles\n,\n \n:file-name\n)\n \n{\n \nrequest\n \nin\n\n  \nlet\n \nfilename\n \n=\n \ntry\n \nrequest\n.\nparameters\n.\nextract\n(\nfile-name\n)\n \nas\n \nString\n\n  \nlet\n \nfile\n \n=\n \nVPRFileManager\n.\nfetch\n(\nfilename\n)\n\n  \nreturn\n \nResponse\n(\nstatus\n:\n \n.\nok\n,\n \nheaders\n:\n \n[\nContent-Type\n:\n \nfile/vpr\n],\n \nbody\n:\n \nfile\n)\n\n\n}\n\n\n\n\n\n\nIn practice, if we're repeating this often, we'll probably conform \nVPRFile\n directly to \nResponseRepresentable\n\n\nextension\n \nVPRFile\n:\n \nHTTP\n.\nResponseRepresentable\n \n{\n\n  \nfunc\n \nmakeResponse\n()\n \n-\n \nResponse\n \n{\n\n    \nreturn\n \nResponse\n(\n\n      \nstatus\n:\n \n.\nok\n,\n\n      \nheaders\n:\n \n[\nContent-Type\n:\n \nfile/vpr\n],\n\n      \nbody\n:\n \nfile\n\n    \n)\n\n  \n}\n\n\n}\n\n\n\n\n\n\nHere's our above example now:\n\n\ndrop\n.\nget\n(\nfiles\n,\n \n:file-name\n)\n \n{\n \nrequest\n \nin\n\n  \nlet\n \nfilename\n \n=\n \ntry\n \nrequest\n.\nparameters\n.\nextract\n(\nfile-name\n)\n \nas\n \nString\n\n  \nreturn\n \nVPRFileManager\n.\nfetch\n(\nfilename\n)\n\n\n}\n\n\n\n\n\n\nWe could also use type-safe routing to make this even more concise:\n\n\ndrop\n.\nget\n(\nfiles\n,\n \nString\n.\nself\n)\n \n{\n \nrequest\n,\n \nfilename\n \nin\n\n  \nreturn\n \nVPRFileManager\n.\nfetch\n(\nfilename\n)\n\n\n}", 
            "title": "Body"
        }, 
        {
            "location": "/http/body/#body", 
            "text": "The  HTTP.Body  represents the payload of an  HTTP.Message , and is used to pass the underlying data. Some examples of this in practice would be  JSON ,  HTML  text, or the bytes of an image. Let's look at the implementation:  public   enum   Body   { \n     case   data ( Bytes ) \n     case   chunked (( ChunkStream )   throws   -   Void )  }", 
            "title": "Body"
        }, 
        {
            "location": "/http/body/#data-case", 
            "text": "The  data  case is by far the most common use for a  Body  in an  HTTP.Message . It is simply an array of bytes. The serialization protocol or type associated with these bytes is usually defined by the  Content-Type  header. Let's look at some examples.", 
            "title": "Data Case"
        }, 
        {
            "location": "/http/body/#applicationjson", 
            "text": "If our  Content-Type  header contains  application/json , then the underlying bytes represent serialized JSON.  if   let   contentType   =   req . headers [ Content-Type ],   contentType . contains ( application/json ),   let   bytes   =   req . body . bytes   { \n   let   json   =   try   JSON ( bytes :   bytes ) \n   print ( Got JSON:  \\( json ) )  }", 
            "title": "Application/JSON"
        }, 
        {
            "location": "/http/body/#imagepng", 
            "text": "If our  Content-Type  contains  image/png , then the underlying bytes represent an encoded png.  if   let   contentType   =   req . headers [ Content-Type ],   contentType . contains ( image/png ),   let   bytes   =   req . body . bytes   { \n   try   database . save ( image :   bytes )  }", 
            "title": "Image/PNG"
        }, 
        {
            "location": "/http/body/#chunked-case", 
            "text": "The  chunked  case only applies to outgoing  HTTP.Message s in Vapor. It is traditionally a responder's role to collect an entire chunked encoding before passing it on. We can use this to send a body asynchronously.  let   body :   Body   =   Body . chunked ( sender )  return   Response ( status :   . ok ,   body :   body )   We can implement this manually, or use Vapor's built in convenience initializer for chunked bodies:  return   Response ( status :   . ok )   {   chunker   in \n   for   name   in   [ joe ,   pam ,   cheryl ]   { \n       sleep ( 1 ) \n       try   chunker . send ( name ) \n   } \n\n   try   chunker . close ()  }    Make sure to call  close()  before the chunker leaves scope.", 
            "title": "Chunked Case"
        }, 
        {
            "location": "/http/body/#bodyrepresentable", 
            "text": "In addition to the concrete  Body  type, as is common in Vapor, we also have wide support for  BodyRepresentable . This means objects that we're commonly converting to  Body  type can be used interchangeably. For example:  return   Response ( body :   Hello, World! )   In the above example, string is converted to bytes and added to the body.   In practice, it is better to use  return \"Hello, World!\" . Vapor will automatically be able to set the  Content-Type  to appropriate values.   Let's look at how it's implemented:  public   protocol   BodyRepresentable   { \n     func   makeBody ()   -   Body  }", 
            "title": "BodyRepresentable"
        }, 
        {
            "location": "/http/body/#custom", 
            "text": "We can conform our own types to this as well where applicable. Let's pretend we have a custom data type,  .vpr . Let's conform our  VPR  file type model:  extension   VPRFile :   HTTP . BodyRepresentable   { \n   func   makeBody ()   -   Body   { \n     // collect bytes \n     return   . data ( bytes ) \n   }  }    You may have noticed above, that the protocol throws, but our implementation does not. This is completely valid in Swift and will allow you to not throw if you're ever calling the function manually.   Now we're able to include our  VPR  file directly in our  Responses .  drop . get ( files ,   :file-name )   {   request   in \n   let   filename   =   try   request . parameters . extract ( file-name )   as   String \n   let   file   =   VPRFileManager . fetch ( filename ) \n   return   Response ( status :   . ok ,   headers :   [ Content-Type :   file/vpr ],   body :   file )  }   In practice, if we're repeating this often, we'll probably conform  VPRFile  directly to  ResponseRepresentable  extension   VPRFile :   HTTP . ResponseRepresentable   { \n   func   makeResponse ()   -   Response   { \n     return   Response ( \n       status :   . ok , \n       headers :   [ Content-Type :   file/vpr ], \n       body :   file \n     ) \n   }  }   Here's our above example now:  drop . get ( files ,   :file-name )   {   request   in \n   let   filename   =   try   request . parameters . extract ( file-name )   as   String \n   return   VPRFileManager . fetch ( filename )  }   We could also use type-safe routing to make this even more concise:  drop . get ( files ,   String . self )   {   request ,   filename   in \n   return   VPRFileManager . fetch ( filename )  }", 
            "title": "Custom"
        }, 
        {
            "location": "/http/response-representable/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\n\n\nModule: \nimport HTTP\n\n\n\n\nResponseRepresentable\n\n\nTraditionally HTTP servers take a \nRequest\n and return a \nResponse\n. Vapor is no different, but we can take advantage of Swift's powerful protocols to be a bit more flexible to the user facing API.\n\n\nLet's start with the definition of \nResponseRepresentable\n\n\npublic\n \nprotocol\n \nResponseRepresentable\n \n{\n\n    \nfunc\n \nmakeResponse\n()\n \nthrows\n \n-\n \nResponse\n\n\n}\n\n\n\n\n\n\nBy conforming to this protocol, we can more flexibly return things that conform instead of creating the response manually each time. Vapor provides some of these by default. Including (but not limited to):\n\n\nString\n\n\nBecause string conforms to \nResponseRepresentable\n, we can return it directly in a Vapor route handler.\n\n\ndrop\n.\nget\n(\nhello\n)\n \n{\n \nrequest\n \nin\n\n  \nreturn\n \nHello, World!\n\n\n}\n\n\n\n\n\n\nJSON\n\n\nJSON\n can be returned directly instead of recreating a response each time.\n\n\ndrop\n.\nget\n(\nhello\n)\n \n{\n \nrequest\n \nin\n\n  \nreturn\n \ntry\n \nJSON\n(\nnode\n:\n \n[\n\n      \nhello\n:\n \nworld\n,\n\n      \nsome-numbers\n:\n \n[\n\n        \n1\n,\n\n        \n2\n,\n\n        \n3\n\n      \n]\n\n    \n]\n\n  \n)\n\n\n}\n\n\n\n\n\n\nResponse\n\n\nOf course, we can also return Responses for anything not covered:\n\n\ndrop\n.\nget\n(\nhello\n)\n \n{\n \nrequest\n \nin\n\n  \nreturn\n \nResponse\n(\nstatus\n:\n \n.\nok\n,\n \nheaders\n:\n \n[\nContent-Type\n:\n \ntext/plain\n],\n \nbody\n:\n \nHello, World!\n)\n\n\n}\n\n\n\n\n\n\nConforming\n\n\nAll we need to do to return our own objects is conform them to \nResponseRepresentable\n. Let's look at an example type, a simple blog post model:\n\n\nimport\n \nFoundation\n\n\n\nstruct\n \nBlogPost\n \n{\n\n  \nlet\n \nid\n:\n \nString\n\n  \nlet\n \ncontent\n:\n \nString\n\n  \nlet\n \ncreatedAt\n:\n \nNSDate\n\n\n}\n\n\n\n\n\n\nAnd now, let's conform it to response representable.\n\n\nimport\n \nHTTP\n\n\nimport\n \nFoundation\n\n\n\nextension\n \nBlogPost\n:\n \nResponseRepresentable\n \n{\n\n  \nfunc\n \nmakeResponse\n()\n \nthrows\n \n-\n \nResponse\n \n{\n\n    \nlet\n \njson\n \n=\n \ntry\n \nJSON\n(\nnode\n:\n\n      \n[\n\n        \nid\n:\n \nid\n,\n\n        \ncontent\n:\n \ncontent\n,\n\n        \ncreated-at\n:\n \ncreatedAt\n.\ntimeIntervalSince1970\n\n      \n]\n\n    \n)\n\n    \nreturn\n \ntry\n \njson\n.\nmakeResponse\n()\n\n  \n}\n\n\n}\n\n\n\n\n\n\n\n\nDon't forget to import HTTP.\n\n\n\n\nNow that we've modeled our BlogPost, we can return it directly in route handlers.\n\n\ndrop\n.\npost\n(\npost\n)\n \n{\n \nreq\n \nin\n\n  \nguard\n \nlet\n \ncontent\n \n=\n \nrequest\n.\ndata\n[\ncontent\n]\n \nelse\n \n{\n \nthrow\n \nError\n.\nmissingContent\n \n}\n\n  \nlet\n \npost\n \n=\n \nPost\n(\ncontent\n:\n \ncontent\n)\n\n  \ntry\n \npost\n.\nsave\n(\nto\n:\n \ndatabase\n)\n\n  \nreturn\n \npost\n\n\n}", 
            "title": "ResponseRepresentable"
        }, 
        {
            "location": "/http/response-representable/#responserepresentable", 
            "text": "Traditionally HTTP servers take a  Request  and return a  Response . Vapor is no different, but we can take advantage of Swift's powerful protocols to be a bit more flexible to the user facing API.  Let's start with the definition of  ResponseRepresentable  public   protocol   ResponseRepresentable   { \n     func   makeResponse ()   throws   -   Response  }   By conforming to this protocol, we can more flexibly return things that conform instead of creating the response manually each time. Vapor provides some of these by default. Including (but not limited to):", 
            "title": "ResponseRepresentable"
        }, 
        {
            "location": "/http/response-representable/#string", 
            "text": "Because string conforms to  ResponseRepresentable , we can return it directly in a Vapor route handler.  drop . get ( hello )   {   request   in \n   return   Hello, World!  }", 
            "title": "String"
        }, 
        {
            "location": "/http/response-representable/#json", 
            "text": "JSON  can be returned directly instead of recreating a response each time.  drop . get ( hello )   {   request   in \n   return   try   JSON ( node :   [ \n       hello :   world , \n       some-numbers :   [ \n         1 , \n         2 , \n         3 \n       ] \n     ] \n   )  }", 
            "title": "JSON"
        }, 
        {
            "location": "/http/response-representable/#response", 
            "text": "Of course, we can also return Responses for anything not covered:  drop . get ( hello )   {   request   in \n   return   Response ( status :   . ok ,   headers :   [ Content-Type :   text/plain ],   body :   Hello, World! )  }", 
            "title": "Response"
        }, 
        {
            "location": "/http/response-representable/#conforming", 
            "text": "All we need to do to return our own objects is conform them to  ResponseRepresentable . Let's look at an example type, a simple blog post model:  import   Foundation  struct   BlogPost   { \n   let   id :   String \n   let   content :   String \n   let   createdAt :   NSDate  }   And now, let's conform it to response representable.  import   HTTP  import   Foundation  extension   BlogPost :   ResponseRepresentable   { \n   func   makeResponse ()   throws   -   Response   { \n     let   json   =   try   JSON ( node : \n       [ \n         id :   id , \n         content :   content , \n         created-at :   createdAt . timeIntervalSince1970 \n       ] \n     ) \n     return   try   json . makeResponse () \n   }  }    Don't forget to import HTTP.   Now that we've modeled our BlogPost, we can return it directly in route handlers.  drop . post ( post )   {   req   in \n   guard   let   content   =   request . data [ content ]   else   {   throw   Error . missingContent   } \n   let   post   =   Post ( content :   content ) \n   try   post . save ( to :   database ) \n   return   post  }", 
            "title": "Conforming"
        }, 
        {
            "location": "/http/responder/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\n\n\nModule: \nimport HTTP\n\n\n\n\nResponder\n\n\nThe \nResponder\n is a simple protocol defining the behavior of objects that can accept a \nRequest\n and return a \nResponse\n. Most notably in Vapor, it is the core API endpoint that connects the \nDroplet\n to the \nServer\n. Let's look at the definition:\n\n\npublic\n \nprotocol\n \nResponder\n \n{\n\n    \nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n)\n \nthrows\n \n-\n \nResponse\n\n\n}\n\n\n\n\n\n\n\n\nThe responder protocol is most notably related to Droplet and it's relationship with a server. Average users will not likely interact with it much.\n\n\n\n\nSimple\n\n\nOf course, Vapor provides some conveniences for this, and in practice, we will often call:\n\n\ntry\n \ndrop\n.\nrun\n()\n\n\n\n\n\n\nManual\n\n\nAs we just mentioned, the Vapor \nDroplet\n itself conforms to \nResponder\n, connecting it to the \nServer\n. This means if we wanted to serve our droplet manually, we could do:\n\n\nlet\n \nserver\n \n=\n \ntry\n \nServer\nTCPServerStream\n,\n \nParser\nRequest\n,\n \nSerializer\nResponse\n(\nport\n:\n \nport\n)\n\n\ntry\n \nserver\n.\nstart\n(\nresponder\n:\n \ndroplet\n)\n  \n{\n \nerror\n \nin\n\n    \nprint\n(\nGot error: \n\\(\nerror\n)\n)\n\n\n}\n\n\n\n\n\n\nAdvanced\n\n\nWe can conform our own objects to \nResponder\n and pass them to \nServers\n. Let's look at an example:\n\n\nfinal\n \nclass\n \nResponder\n:\n \nHTTP\n.\nResponder\n \n{\n\n    \nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n)\n \nthrows\n \n-\n \nResponse\n \n{\n\n        \nlet\n \nbody\n \n=\n \nHello World\n.\nmakeBody\n()\n\n        \nreturn\n \nResponse\n(\nbody\n:\n \nbody\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nThis only returns \n\"Hello World\"\n for every request, it's most commonly going to be linked with a router of some type.\n\n\nfinal\n \nclass\n \nResponder\n:\n \nHTTP\n.\nResponder\n \n{\n\n    \nlet\n \nrouter\n:\n \nRouter\n \n=\n \n...\n\n\n    \nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n)\n \nthrows\n \n-\n \nResponse\n \n{\n\n        \nreturn\n \ntry\n \nrouter\n.\nroute\n(\nrequest\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nWe'll then pass this responder to a server and let it go.\n\n\nlet\n \nserver\n \n=\n \ntry\n \nServer\nTCPServerStream\n,\n \nParser\nRequest\n,\n \nSerializer\nResponse\n(\nport\n:\n \nport\n)\n\n\n\nprint\n(\nvisit http://localhost:\n\\(\nport\n)\n/\n)\n\n\ntry\n \nserver\n.\nstart\n(\nresponder\n:\n \nResponder\n())\n \n{\n \nerror\n \nin\n\n    \nprint\n(\nGot error: \n\\(\nerror\n)\n)\n\n\n}\n\n\n\n\n\n\nThis can be used as a jumping off point for applications looking to implement features manually.\n\n\nClient\n\n\nThe \nHTTP.Client\n is itself a \nResponder\n although, instead of handling the \nRequest\n itself, it passes it on to the underlying URI.", 
            "title": "Responder"
        }, 
        {
            "location": "/http/responder/#responder", 
            "text": "The  Responder  is a simple protocol defining the behavior of objects that can accept a  Request  and return a  Response . Most notably in Vapor, it is the core API endpoint that connects the  Droplet  to the  Server . Let's look at the definition:  public   protocol   Responder   { \n     func   respond ( to   request :   Request )   throws   -   Response  }    The responder protocol is most notably related to Droplet and it's relationship with a server. Average users will not likely interact with it much.", 
            "title": "Responder"
        }, 
        {
            "location": "/http/responder/#simple", 
            "text": "Of course, Vapor provides some conveniences for this, and in practice, we will often call:  try   drop . run ()", 
            "title": "Simple"
        }, 
        {
            "location": "/http/responder/#manual", 
            "text": "As we just mentioned, the Vapor  Droplet  itself conforms to  Responder , connecting it to the  Server . This means if we wanted to serve our droplet manually, we could do:  let   server   =   try   Server TCPServerStream ,   Parser Request ,   Serializer Response ( port :   port )  try   server . start ( responder :   droplet )    {   error   in \n     print ( Got error:  \\( error ) )  }", 
            "title": "Manual"
        }, 
        {
            "location": "/http/responder/#advanced", 
            "text": "We can conform our own objects to  Responder  and pass them to  Servers . Let's look at an example:  final   class   Responder :   HTTP . Responder   { \n     func   respond ( to   request :   Request )   throws   -   Response   { \n         let   body   =   Hello World . makeBody () \n         return   Response ( body :   body ) \n     }  }   This only returns  \"Hello World\"  for every request, it's most commonly going to be linked with a router of some type.  final   class   Responder :   HTTP . Responder   { \n     let   router :   Router   =   ... \n\n     func   respond ( to   request :   Request )   throws   -   Response   { \n         return   try   router . route ( request ) \n     }  }   We'll then pass this responder to a server and let it go.  let   server   =   try   Server TCPServerStream ,   Parser Request ,   Serializer Response ( port :   port )  print ( visit http://localhost: \\( port ) / )  try   server . start ( responder :   Responder ())   {   error   in \n     print ( Got error:  \\( error ) )  }   This can be used as a jumping off point for applications looking to implement features manually.", 
            "title": "Advanced"
        }, 
        {
            "location": "/http/responder/#client", 
            "text": "The  HTTP.Client  is itself a  Responder  although, instead of handling the  Request  itself, it passes it on to the underlying URI.", 
            "title": "Client"
        }, 
        {
            "location": "/http/client/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\n\n\nModule: \nimport HTTP\n\n\n\n\nClient\n\n\nThe client provided by \nHTTP\n is used to make outgoing requests to remote servers. Let's look at a simple outgoing request.\n\n\nQuickStart\n\n\nLet's jump right in to make a simple HTTP Request. Here's a basic \nGET\n request using your Vapor \nDroplet\n.\n\n\nlet\n \nquery\n \n=\n \n...\n\n\nlet\n \nspotifyResponse\n \n=\n \ntry\n \ndrop\n.\nclient\n.\nget\n(\nhttps://api.spotify.com/v1/search?type=artist\nq=\n\\(\nquery\n)\n)\n\n\nprint\n(\nspotifyR\n)\n\n\n\n\n\n\nClean Up\n\n\nThe url above can be a little tricky to read, so let's use the query parameter to clean it up a little bit:\n\n\ntry\n \ndrop\n.\nclient\n.\nget\n(\nhttps://api.spotify.com/v1/search\n,\n \nquery\n:\n \n[\ntype\n:\n \nartist\n,\n \nq\n:\n \nquery\n])\n\n\n\n\n\n\nContinued\n\n\nIn addition to \nGET\n requests, Vapor's client provides support for most common HTTP functions. \nGET\n, \nPOST\n, \nPUT\n, \nPATCH\n, \nDELETE\n\n\nPOST as json\n\n\ntry\n \ndrop\n.\nclient\n.\npost\n(\nhttp://some-endpoint/json\n,\n \nheaders\n:\n \n[\nContent-Type\n:\n \napplication/json\n],\n \nbody\n:\n \nmyJSON\n.\nmakeBody\n())\n\n\n\n\n\n\nPOST as x-www-form-urlencoded\n\n\ntry\n \ndrop\n.\nclient\n.\npost\n(\nhttp://some-endpoint\n,\n \nheaders\n:\n \n[\n\n  \nContent-Type\n:\n \napplication/x-www-form-urlencoded\n\n\n],\n \nbody\n:\n \nBody\n.\ndata\n(\n \nNode\n(\nnode\n:\n \n[\n\n  \nemail\n:\n \nmymail@vapor.codes\n\n\n]).\nformURLEncoded\n()))\n               \n\n\n\n\n\nFull Request\n\n\nTo access additional functionality or custom methods, use the underlying \nrequest\n function directly.\n\n\npublic\n \nstatic\n \nfunc\n \nget\n(\n_\n \nmethod\n:\n \nMethod\n,\n\n                       \n_\n \nuri\n:\n \nString\n,\n\n                       \nheaders\n:\n \n[\nHeaderKey\n:\n \nString\n]\n \n=\n \n[:],\n\n                       \nquery\n:\n \n[\nString\n:\n \nCustomStringConvertible\n]\n \n=\n \n[:],\n\n                       \nbody\n:\n \nBody\n \n=\n \n[])\n \nthrows\n \n-\n \nResponse\n\n\n\n\n\n\nFor example:\n\n\ntry\n \ndrop\n.\nclient\n.\nrequest\n(.\nother\n(\nmethod\n:\n \nCUSTOM\n),\n \nhttp://some-domain\n,\n \nheaders\n:\n \n[\nMy\n:\n \nHeader\n],\n \nquery\n:\n \n[\nkey\n:\n \nvalue\n],\n \nbody\n:\n \n[])\n\n\n\n\n\n\nConfig\n\n\nThe \nConfig/clients.json\n file can be used to modify the client's settings.\n\n\nTLS\n\n\nHost and certificate verification can be disabled.\n\n\n\n\nNote: Use extreme caution when modifying these settings.\n\n\n\n\n{\n\n    \ntls\n:\n \n{\n\n        \nverifyHost\n:\n \nfalse\n,\n\n        \nverifyCertificates\n:\n \nfalse\n\n    \n}\n\n\n}\n\n\n\n\n\n\nMozilla\n\n\nThe Mozilla certificates are included by default to make fetching content from secure sites easy.\n\n\n{\n\n    \ntls\n:\n \n{\n\n        \ncertificates\n:\n \nmozilla\n\n    \n}\n\n\n}\n\n\n\n\n\n\nAdvanced\n\n\nIn addition to our Droplet, we can also use and interact with the \nClient\n manually. Here's how our default implementation in Vapor looks:\n\n\nlet\n \nresponse\n \n=\n \ntry\n \nClient\nTCPClientStream\n.\nget\n(\nhttp://some-endpoint/mine\n)\n\n\n\n\n\n\nThe first thing we likely noticed is \nTCPClientStream\n being used as a Generic value. This will be the underlying connection that the \nHTTP.Client\n can use when performing the request. By conforming to the underlying \nClientStream\n, an \nHTTP.Client\n can accept custom stream implementations seamlessly.\n\n\nSave Connection\n\n\nUp to this point, we've been interacting with the Client via \nclass\n or \nstatic\n level functions. This allows us to end the connection upon a completed request and is the recommended interaction for most use cases. For some advanced situations, we may want to reuse a connection. For these, we can initialize our client and perform multiple requests like this.\n\n\nlet\n \npokemonClient\n \n=\n \ntry\n \ndrop\n?.\nclient\n.\nmake\n(\nscheme\n:\n \nhttp\n,\n \nhost\n:\n \npokeapi.co\n)\n\n\nfor\n \ni\n \nin\n \n0.\n..\n1\n \n{\n\n    \nlet\n \nresponse\n \n=\n \ntry\n \npokemonClient\n?.\nget\n(\npath\n:\n \n/api/v2/pokemon/\n,\n \nquery\n:\n \n[\nlimit\n:\n \n20\n,\n \noffset\n:\n \ni\n])\n\n    \nprint\n(\nresponse: \n\\(\nresponse\n)\n)\n\n\n}\n\n\n\n\n\n\nClientProtocol\n\n\nUp to this point, we've focused on the built in \nHTTP.Client\n, but users can also include their own customized clients by conforming to \nHTTP.ClientProtocol\n. Let's look at the implementation:\n\n\npublic\n \nprotocol\n \nResponder\n \n{\n\n    \nfunc\n \nrespond\n(\nto\n \nrequest\n:\n \nRequest\n)\n \nthrows\n \n-\n \nResponse\n\n\n}\n\n\n\npublic\n \nprotocol\n \nProgram\n \n{\n\n    \nvar\n \nhost\n:\n \nString\n \n{\n \nget\n \n}\n\n    \nvar\n \nport\n:\n \nInt\n \n{\n \nget\n \n}\n\n    \nvar\n \nsecurityLayer\n:\n \nSecurityLayer\n \n{\n \nget\n \n}\n\n    \n// default implemented\n\n    \ninit\n(\nhost\n:\n \nString\n,\n \nport\n:\n \nInt\n,\n \nsecurityLayer\n:\n \nSecurityLayer\n)\n \nthrows\n\n\n}\n\n\n\npublic\n \nprotocol\n \nClientProtocol\n:\n \nProgram\n,\n \nResponder\n \n{\n\n    \nvar\n \nscheme\n:\n \nString\n \n{\n \nget\n \n}\n\n    \nvar\n \nstream\n:\n \nStream\n \n{\n \nget\n \n}\n\n    \ninit\n(\nscheme\n:\n \nString\n,\n \nhost\n:\n \nString\n,\n \nport\n:\n \nInt\n,\n \nsecurityLayer\n:\n \nSecurityLayer\n)\n \nthrows\n\n\n}\n\n\n\n\n\n\nBy conforming to these underlying functions, we immediately gain access to the public \nClientProtocol\n apis we viewed above.\n\n\nCustomize Droplet\n\n\nIf we've introduced a custom conformance to \nHTTP.ClientProtocol\n, we can pass this into our droplet without changing the underlying behavior in our application.\n\n\nFor example:\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\n\ndrop\n.\nclient\n \n=\n \nMyCustomClient\n.\nself\n\n\n\n\n\n\nGoing forward, all of your calls to \ndrop.client\n will use \nMyCustomClient.self\n:\n\n\ndrop\n.\nclient\n.\nget\n(...\n \n// uses `MyCustomClient`", 
            "title": "Client"
        }, 
        {
            "location": "/http/client/#client", 
            "text": "The client provided by  HTTP  is used to make outgoing requests to remote servers. Let's look at a simple outgoing request.", 
            "title": "Client"
        }, 
        {
            "location": "/http/client/#quickstart", 
            "text": "Let's jump right in to make a simple HTTP Request. Here's a basic  GET  request using your Vapor  Droplet .  let   query   =   ...  let   spotifyResponse   =   try   drop . client . get ( https://api.spotify.com/v1/search?type=artist q= \\( query ) )  print ( spotifyR )", 
            "title": "QuickStart"
        }, 
        {
            "location": "/http/client/#clean-up", 
            "text": "The url above can be a little tricky to read, so let's use the query parameter to clean it up a little bit:  try   drop . client . get ( https://api.spotify.com/v1/search ,   query :   [ type :   artist ,   q :   query ])", 
            "title": "Clean Up"
        }, 
        {
            "location": "/http/client/#continued", 
            "text": "In addition to  GET  requests, Vapor's client provides support for most common HTTP functions.  GET ,  POST ,  PUT ,  PATCH ,  DELETE", 
            "title": "Continued"
        }, 
        {
            "location": "/http/client/#post-as-json", 
            "text": "try   drop . client . post ( http://some-endpoint/json ,   headers :   [ Content-Type :   application/json ],   body :   myJSON . makeBody ())", 
            "title": "POST as json"
        }, 
        {
            "location": "/http/client/#post-as-x-www-form-urlencoded", 
            "text": "try   drop . client . post ( http://some-endpoint ,   headers :   [ \n   Content-Type :   application/x-www-form-urlencoded  ],   body :   Body . data (   Node ( node :   [ \n   email :   mymail@vapor.codes  ]). formURLEncoded ()))", 
            "title": "POST as x-www-form-urlencoded"
        }, 
        {
            "location": "/http/client/#full-request", 
            "text": "To access additional functionality or custom methods, use the underlying  request  function directly.  public   static   func   get ( _   method :   Method , \n                        _   uri :   String , \n                        headers :   [ HeaderKey :   String ]   =   [:], \n                        query :   [ String :   CustomStringConvertible ]   =   [:], \n                        body :   Body   =   [])   throws   -   Response   For example:  try   drop . client . request (. other ( method :   CUSTOM ),   http://some-domain ,   headers :   [ My :   Header ],   query :   [ key :   value ],   body :   [])", 
            "title": "Full Request"
        }, 
        {
            "location": "/http/client/#config", 
            "text": "The  Config/clients.json  file can be used to modify the client's settings.", 
            "title": "Config"
        }, 
        {
            "location": "/http/client/#tls", 
            "text": "Host and certificate verification can be disabled.   Note: Use extreme caution when modifying these settings.   { \n     tls :   { \n         verifyHost :   false , \n         verifyCertificates :   false \n     }  }", 
            "title": "TLS"
        }, 
        {
            "location": "/http/client/#mozilla", 
            "text": "The Mozilla certificates are included by default to make fetching content from secure sites easy.  { \n     tls :   { \n         certificates :   mozilla \n     }  }", 
            "title": "Mozilla"
        }, 
        {
            "location": "/http/client/#advanced", 
            "text": "In addition to our Droplet, we can also use and interact with the  Client  manually. Here's how our default implementation in Vapor looks:  let   response   =   try   Client TCPClientStream . get ( http://some-endpoint/mine )   The first thing we likely noticed is  TCPClientStream  being used as a Generic value. This will be the underlying connection that the  HTTP.Client  can use when performing the request. By conforming to the underlying  ClientStream , an  HTTP.Client  can accept custom stream implementations seamlessly.", 
            "title": "Advanced"
        }, 
        {
            "location": "/http/client/#save-connection", 
            "text": "Up to this point, we've been interacting with the Client via  class  or  static  level functions. This allows us to end the connection upon a completed request and is the recommended interaction for most use cases. For some advanced situations, we may want to reuse a connection. For these, we can initialize our client and perform multiple requests like this.  let   pokemonClient   =   try   drop ?. client . make ( scheme :   http ,   host :   pokeapi.co )  for   i   in   0. .. 1   { \n     let   response   =   try   pokemonClient ?. get ( path :   /api/v2/pokemon/ ,   query :   [ limit :   20 ,   offset :   i ]) \n     print ( response:  \\( response ) )  }", 
            "title": "Save Connection"
        }, 
        {
            "location": "/http/client/#clientprotocol", 
            "text": "Up to this point, we've focused on the built in  HTTP.Client , but users can also include their own customized clients by conforming to  HTTP.ClientProtocol . Let's look at the implementation:  public   protocol   Responder   { \n     func   respond ( to   request :   Request )   throws   -   Response  }  public   protocol   Program   { \n     var   host :   String   {   get   } \n     var   port :   Int   {   get   } \n     var   securityLayer :   SecurityLayer   {   get   } \n     // default implemented \n     init ( host :   String ,   port :   Int ,   securityLayer :   SecurityLayer )   throws  }  public   protocol   ClientProtocol :   Program ,   Responder   { \n     var   scheme :   String   {   get   } \n     var   stream :   Stream   {   get   } \n     init ( scheme :   String ,   host :   String ,   port :   Int ,   securityLayer :   SecurityLayer )   throws  }   By conforming to these underlying functions, we immediately gain access to the public  ClientProtocol  apis we viewed above.", 
            "title": "ClientProtocol"
        }, 
        {
            "location": "/http/client/#customize-droplet", 
            "text": "If we've introduced a custom conformance to  HTTP.ClientProtocol , we can pass this into our droplet without changing the underlying behavior in our application.  For example:  let   drop   =   Droplet ()  drop . client   =   MyCustomClient . self   Going forward, all of your calls to  drop.client  will use  MyCustomClient.self :  drop . client . get (...   // uses `MyCustomClient`", 
            "title": "Customize Droplet"
        }, 
        {
            "location": "/http/server/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\nServer\n\n\nThe server is responsible for accepting connections from clients, parsing their requests, and delivering them a response. \n\n\nDefault\n\n\nStarting your Droplet with a default server is simple.\n\n\nimport\n \nVapor\n\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\n\ndrop\n.\nrun\n()\n\n\n\n\n\n\nThe default server will bind to host \n0.0.0.0\n at port \n8080\n.\n\n\nConfig\n\n\nIf you are using a \nConfig/servers.json\n file, this is where you can easily change your host and port or even boot multiple servers.\n\n\n{\n\n    \ndefault\n:\n \n{\n\n        \nport\n:\n \n$PORT:8080\n,\n\n        \nhost\n:\n \n0.0.0.0\n,\n\n        \nsecurityLayer\n:\n \nnone\n\n    \n}\n\n\n}\n\n\n\n\n\n\nThe default \nservers.json\n is above. The port with try to resolve the environment variable \n$PORT\n or fallback to \n8080\n.\n\n\nMultiple\n\n\nYou can start multiple servers in the same application. This is especially useful if you want to boot an \nHTTP\n and \nHTTPS\n server side by side.\n\n\n{\n\n    \nplaintext\n:\n \n{\n\n        \nport\n:\n \n80\n,\n\n        \nhost\n:\n \nvapor.codes\n,\n\n        \nsecurityLayer\n:\n \nnone\n\n    \n},\n\n    \nsecure\n:\n \n{\n\n        \nport\n:\n \n443\n,\n\n        \nhost\n:\n \nvapor.codes\n,\n\n        \nsecurityLayer\n:\n \ntls\n,\n\n        \ntls\n:\n \n{\n\n            \ncertificates\n:\n \nnone\n,\n\n            \nsignature\n:\n \nselfSigned\n\n        \n}\n\n    \n},\n\n\n}\n\n\n\n\n\n\nTLS\n\n\nTLS (formerly SSL) can be configured with a variety of different certificate and signature types.\n\n\nVerify\n\n\nVerificiation of hosts and certificates can be disabled. They are enabled by default.\n\n\n\n\nNote: Be extremely careful when disabling these options.\n\n\n\n\ntls\n:\n \n{\n\n   \nverifyHost\n:\n \nfalse\n,\n\n   \nverifyCertificates\n:\n \nfalse\n\n\n}\n\n\n\n\n\n\nCertificates\n\n\nNone\n\n\ntls\n:\n \n{\n\n    \ncertificates\n:\n \nnone\n\n\n}\n\n\n\n\n\n\nChain\n\n\ntls\n:\n \n{\n\n    \ncertificates\n:\n \nchain\n,\n\n    \nchainFile\n:\n \n/path/to/chainfile\n\n\n}\n\n\n\n\n\n\nFiles\n\n\ntls\n:\n \n{\n\n    \ncertificates\n:\n \nfiles\n,\n\n    \ncertificateFile\n:\n \n/path/to/cert.pem\n,\n\n    \nprivateKeyFile\n:\n \n/path/to/key.pem\n\n\n}\n\n\n\n\n\n\nCertificate Authority\n\n\ntls\n:\n \n{\n\n    \ncertificates\n:\n \nca\n\n\n}\n\n\n\n\n\n\nSignature\n\n\nSelf Signed\n\n\ntls\n:\n \n{\n\n    \nsignature\n:\n \nselfSigned\n\n\n}\n\n\n\n\n\n\nSigned File\n\n\ntls\n:\n \n{\n\n    \nsignature\n:\n \nsignedFile\n,\n\n    \ncaCertificateFile\n:\n \n/path/to/file\n\n\n}\n\n\n\n\n\n\nSigned Directory\n\n\ntls\n:\n \n{\n\n    \nsignature\n:\n \nsignedDirectory\n,\n\n    \ncaCertificateDirectory\n:\n \n/path/to/dir\n\n\n}\n\n\n\n\n\n\nExample\n\n\nHere is an example \nservers.json\n file using certificate files with a self signed signature and host verification redundantly set to \ntrue\n.\n\n\n{\n\n    \nsecure\n:\n \n{\n\n        \nport\n:\n \n8443\n,\n\n        \nhost\n:\n \n0.0.0.0\n,\n\n        \nsecurityLayer\n:\n \ntls\n,\n\n        \ntls\n:\n \n{\n\n            \nverifyHost\n:\n \ntrue\n,\n\n            \ncertificates\n:\n \nfiles\n,\n\n            \ncertificateFile\n:\n \n/vapor/certs/cert.pem\n,\n\n            \nprivateKeyFile\n:\n \n/vapor/certs/key.pem\n,\n\n            \nsignature\n:\n \nselfSigned\n\n        \n}\n\n    \n}\n\n\n}\n\n\n\n\n\n\nManual\n\n\nServers can also be configured manually, without configuration files. \n\n\n\n\nNote: If servers are configured programatically, they override any config settings.\n\n\n\n\nSimple\n\n\nThe \nrun\n method on the Droplet takes a dictionary of server configuration objects. The key is the name of the server.\n\n\nimport\n \nVapor\n\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\n\ndrop\n.\nrun\n(\nservers\n:\n \n[\n\n    \ndefault\n:\n \n(\nhost\n:\n \nvapor.codes\n,\n \nport\n:\n \n8080\n,\n \nsecurityLayer\n:\n \n.\nnone\n)\n\n\n]\n\n\n\n\n\n\nTLS\n\n\nTLS can also be configured manually, and works similarly to the \nservers.json\n config files described above.\n\n\n```swift\nimport Vapor\nimport TLS\n\n\nlet drop = Droplet()\n\n\nlet config = try TLS.Config(\n    mode: .server,\n    certificates: .files(\n        certificateFile: \"/Users/tanner/Desktop/certs/cert.pem\", \n        privateKeyFile: \"/Users/tanner/Desktop/certs/key.pem\", \n        signature: .selfSigned\n    ),\n    verifyHost: true,\n    verifyCertificates: true\n)\n\n\ndrop.run(servers: [\n    \"plaintext\": (\"vapor.codes\", 8080, .none),\n    \"secure\": (\"vapor.codes\", 8443, .tls(config)),\n])\n````", 
            "title": "Server"
        }, 
        {
            "location": "/http/server/#server", 
            "text": "The server is responsible for accepting connections from clients, parsing their requests, and delivering them a response.", 
            "title": "Server"
        }, 
        {
            "location": "/http/server/#default", 
            "text": "Starting your Droplet with a default server is simple.  import   Vapor  let   drop   =   Droplet ()  drop . run ()   The default server will bind to host  0.0.0.0  at port  8080 .", 
            "title": "Default"
        }, 
        {
            "location": "/http/server/#config", 
            "text": "If you are using a  Config/servers.json  file, this is where you can easily change your host and port or even boot multiple servers.  { \n     default :   { \n         port :   $PORT:8080 , \n         host :   0.0.0.0 , \n         securityLayer :   none \n     }  }   The default  servers.json  is above. The port with try to resolve the environment variable  $PORT  or fallback to  8080 .", 
            "title": "Config"
        }, 
        {
            "location": "/http/server/#multiple", 
            "text": "You can start multiple servers in the same application. This is especially useful if you want to boot an  HTTP  and  HTTPS  server side by side.  { \n     plaintext :   { \n         port :   80 , \n         host :   vapor.codes , \n         securityLayer :   none \n     }, \n     secure :   { \n         port :   443 , \n         host :   vapor.codes , \n         securityLayer :   tls , \n         tls :   { \n             certificates :   none , \n             signature :   selfSigned \n         } \n     },  }", 
            "title": "Multiple"
        }, 
        {
            "location": "/http/server/#tls", 
            "text": "TLS (formerly SSL) can be configured with a variety of different certificate and signature types.", 
            "title": "TLS"
        }, 
        {
            "location": "/http/server/#verify", 
            "text": "Verificiation of hosts and certificates can be disabled. They are enabled by default.   Note: Be extremely careful when disabling these options.   tls :   { \n    verifyHost :   false , \n    verifyCertificates :   false  }", 
            "title": "Verify"
        }, 
        {
            "location": "/http/server/#certificates", 
            "text": "", 
            "title": "Certificates"
        }, 
        {
            "location": "/http/server/#none", 
            "text": "tls :   { \n     certificates :   none  }", 
            "title": "None"
        }, 
        {
            "location": "/http/server/#chain", 
            "text": "tls :   { \n     certificates :   chain , \n     chainFile :   /path/to/chainfile  }", 
            "title": "Chain"
        }, 
        {
            "location": "/http/server/#files", 
            "text": "tls :   { \n     certificates :   files , \n     certificateFile :   /path/to/cert.pem , \n     privateKeyFile :   /path/to/key.pem  }", 
            "title": "Files"
        }, 
        {
            "location": "/http/server/#certificate-authority", 
            "text": "tls :   { \n     certificates :   ca  }", 
            "title": "Certificate Authority"
        }, 
        {
            "location": "/http/server/#signature", 
            "text": "", 
            "title": "Signature"
        }, 
        {
            "location": "/http/server/#self-signed", 
            "text": "tls :   { \n     signature :   selfSigned  }", 
            "title": "Self Signed"
        }, 
        {
            "location": "/http/server/#signed-file", 
            "text": "tls :   { \n     signature :   signedFile , \n     caCertificateFile :   /path/to/file  }", 
            "title": "Signed File"
        }, 
        {
            "location": "/http/server/#signed-directory", 
            "text": "tls :   { \n     signature :   signedDirectory , \n     caCertificateDirectory :   /path/to/dir  }", 
            "title": "Signed Directory"
        }, 
        {
            "location": "/http/server/#example", 
            "text": "Here is an example  servers.json  file using certificate files with a self signed signature and host verification redundantly set to  true .  { \n     secure :   { \n         port :   8443 , \n         host :   0.0.0.0 , \n         securityLayer :   tls , \n         tls :   { \n             verifyHost :   true , \n             certificates :   files , \n             certificateFile :   /vapor/certs/cert.pem , \n             privateKeyFile :   /vapor/certs/key.pem , \n             signature :   selfSigned \n         } \n     }  }", 
            "title": "Example"
        }, 
        {
            "location": "/http/server/#manual", 
            "text": "Servers can also be configured manually, without configuration files.    Note: If servers are configured programatically, they override any config settings.", 
            "title": "Manual"
        }, 
        {
            "location": "/http/server/#simple", 
            "text": "The  run  method on the Droplet takes a dictionary of server configuration objects. The key is the name of the server.  import   Vapor  let   drop   =   Droplet ()  drop . run ( servers :   [ \n     default :   ( host :   vapor.codes ,   port :   8080 ,   securityLayer :   . none )  ]", 
            "title": "Simple"
        }, 
        {
            "location": "/http/server/#tls_1", 
            "text": "TLS can also be configured manually, and works similarly to the  servers.json  config files described above.  ```swift\nimport Vapor\nimport TLS  let drop = Droplet()  let config = try TLS.Config(\n    mode: .server,\n    certificates: .files(\n        certificateFile: \"/Users/tanner/Desktop/certs/cert.pem\", \n        privateKeyFile: \"/Users/tanner/Desktop/certs/key.pem\", \n        signature: .selfSigned\n    ),\n    verifyHost: true,\n    verifyCertificates: true\n)  drop.run(servers: [\n    \"plaintext\": (\"vapor.codes\", 8080, .none),\n    \"secure\": (\"vapor.codes\", 8443, .tls(config)),\n])\n````", 
            "title": "TLS"
        }, 
        {
            "location": "/http/cors/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\nCORS\n\n\nVapor by default provides a middleware for implementing proper support for Cross-Origin Resource Sharing (CORS) named \nCORSMiddleware\n.\n\n\n\"Cross-Origin Resource Sharing (CORS) is a specification that enables truly open access across domain-boundaries. If you serve public content, please consider using CORS to open it up for universal JavaScript / browser access.\" - \nhttp://enable-cors.org/\n\n\nTo learn more about middlewares, please visit the Middleware section of the documentation \nhere\n.\n\n\n\n\nImage Author: \nWikipedia\n\n\nBasic\n\n\nFirst of all, add the CORS middleware into your droplet middlewares array.\n\n\n#\n \nInsert\n \nCORS\n \nbefore\n \nany\n \nother\n \nmiddlewares\n\n\ndrop\n.\nmiddleware\n.\ninsert\n(\nCORSMiddleware\n(),\n \nat\n:\n \n0\n)\n\n\n\n\n\n\n\n\nNote: Make sure you insert CORS middleware before any other throwing middlewares, like the AbortMiddleware or similar. Otherwise the proper headers might not be added to the response.\n\n\n\n\nCORSMiddleware\n has a default configuration which should suit most users, with values as follows:\n\n\n\n\nAllowed Origin\n \n\n\nValue of origin header in the request.\n\n\n\n\n\n\nAllowed Methods\n \n\n\nGET\n, \nPOST\n, \nPUT\n, \nOPTIONS\n, \nDELETE\n, \nPATCH\n\n\n\n\n\n\nAllowed Headers\n\n\nAccept\n, \nAuthorization\n, \nContent-Type\n, \nOrigin\n, \nX-Requested-With\n\n\n\n\n\n\n\n\nAdvanced\n\n\nAll settings and presets can be customized by advanced users. There's two ways of doing this, either you programatically create and configure a \nCORSConfiguration\n object or you can put your configuration into a Vapor's JSON config file.\n\n\nSee below for how to set up both and what are the options.\n\n\nConfiguration\n\n\nThe \nCORSConfiguration\n struct is used to configure the \nCORSMiddleware\n. You can instanitate one like this:\n\n\nlet\n \nconfiguration\n \n=\n \nCORSConfiguration\n(\nallowedOrigin\n:\n \n.\ncustom\n(\nhttps://vapor.codes\n),\n\n                                          \nallowedMethods\n:\n \n[.\nget\n,\n \n.\npost\n,\n \n.\noptions\n],\n\n                                          \nallowedHeaders\n:\n \n[\nAccept\n,\n \nAuthorization\n],\n\n                                          \nallowCredentials\n:\n \nfalse\n,\n\n                                          \ncacheExpiration\n:\n \n600\n,\n\n                                          \nexposedHeaders\n:\n \n[\nCache-Control\n,\n \nContent-Language\n])\n\n\n\n\n\n\nAfter creating a configuration you can add the CORS middleware.\n\n\ndrop\n.\nmiddleware\n.\ninsert\n(\nCORSMiddleware\n(\nconfiguration\n:\n \nconfiguration\n),\n \nat\n:\n \n0\n)\n\n\n\n\n\n\n\n\nNote: Please consult the documentation in the source code of the \nCORSConfiguration\n for more information about available values for the settings.\n\n\n\n\nJSON Config\n\n\nOptionally, \nCORSMiddleware\n can be configured using the Vapor's \nConfig\n which is created out of the json files contained in your Config folder. You will need to create a file called \ncors.json\n or \nCORS.json\n in your Config folder in your project and add the required keys.\n\n\nExample of how such a file could look as follows:\n\n\n{\n\n    \nallowedOrigin\n:\n \norigin\n,\n\n    \nallowedMethods\n:\n \nGET,POST,PUT,OPTIONS,DELETE,PATCH\n,\n\n    \nallowedHeaders\n:\n \n[\nAccept\n,\n \nAuthorization\n,\n \nContent-Type\n,\n \nOrigin\n,\n \nX-Requested-With\n]\n\n\n}\n\n\n\n\n\n\n\n\nNote: Following keys are required: \nallowedOrigin\n, \nallowedMethods\n, \nallowedHeaders\n. If they are not present an error will be thrown while instantiating the middleware.\n\n\nOptionally you can also specify the keys \nallowCredentials\n (Bool), \ncacheExpiration\n (Int) and \nexposedHeaders\n ([String]).\n\n\n\n\nAfterwards you can add the middleware using the a throwing overload of the initialiser that accepts Vapor's \nConfig\n.\n\n\nlet\n \ndrop\n \n=\n \nDroplet\n()\n\n\n\ndo\n \n{\n\n    \ndrop\n.\nmiddleware\n.\ninsert\n(\ntry\n \nCORSMiddleware\n(\nconfiguration\n:\n \ndrop\n.\nconfig\n),\n \nat\n:\n \n0\n)\n\n\n}\n \ncatch\n \n{\n\n    \nfatalError\n(\nError creating CORSMiddleware, please check that you\nve setup cors.json correctly.\n)\n\n\n}", 
            "title": "CORS"
        }, 
        {
            "location": "/http/cors/#cors", 
            "text": "Vapor by default provides a middleware for implementing proper support for Cross-Origin Resource Sharing (CORS) named  CORSMiddleware .  \"Cross-Origin Resource Sharing (CORS) is a specification that enables truly open access across domain-boundaries. If you serve public content, please consider using CORS to open it up for universal JavaScript / browser access.\" -  http://enable-cors.org/  To learn more about middlewares, please visit the Middleware section of the documentation  here .   Image Author:  Wikipedia", 
            "title": "CORS"
        }, 
        {
            "location": "/http/cors/#basic", 
            "text": "First of all, add the CORS middleware into your droplet middlewares array.  #   Insert   CORS   before   any   other   middlewares  drop . middleware . insert ( CORSMiddleware (),   at :   0 )    Note: Make sure you insert CORS middleware before any other throwing middlewares, like the AbortMiddleware or similar. Otherwise the proper headers might not be added to the response.   CORSMiddleware  has a default configuration which should suit most users, with values as follows:   Allowed Origin    Value of origin header in the request.    Allowed Methods    GET ,  POST ,  PUT ,  OPTIONS ,  DELETE ,  PATCH    Allowed Headers  Accept ,  Authorization ,  Content-Type ,  Origin ,  X-Requested-With", 
            "title": "Basic"
        }, 
        {
            "location": "/http/cors/#advanced", 
            "text": "All settings and presets can be customized by advanced users. There's two ways of doing this, either you programatically create and configure a  CORSConfiguration  object or you can put your configuration into a Vapor's JSON config file.  See below for how to set up both and what are the options.", 
            "title": "Advanced"
        }, 
        {
            "location": "/http/cors/#configuration", 
            "text": "The  CORSConfiguration  struct is used to configure the  CORSMiddleware . You can instanitate one like this:  let   configuration   =   CORSConfiguration ( allowedOrigin :   . custom ( https://vapor.codes ), \n                                           allowedMethods :   [. get ,   . post ,   . options ], \n                                           allowedHeaders :   [ Accept ,   Authorization ], \n                                           allowCredentials :   false , \n                                           cacheExpiration :   600 , \n                                           exposedHeaders :   [ Cache-Control ,   Content-Language ])   After creating a configuration you can add the CORS middleware.  drop . middleware . insert ( CORSMiddleware ( configuration :   configuration ),   at :   0 )    Note: Please consult the documentation in the source code of the  CORSConfiguration  for more information about available values for the settings.", 
            "title": "Configuration"
        }, 
        {
            "location": "/http/cors/#json-config", 
            "text": "Optionally,  CORSMiddleware  can be configured using the Vapor's  Config  which is created out of the json files contained in your Config folder. You will need to create a file called  cors.json  or  CORS.json  in your Config folder in your project and add the required keys.  Example of how such a file could look as follows:  { \n     allowedOrigin :   origin , \n     allowedMethods :   GET,POST,PUT,OPTIONS,DELETE,PATCH , \n     allowedHeaders :   [ Accept ,   Authorization ,   Content-Type ,   Origin ,   X-Requested-With ]  }    Note: Following keys are required:  allowedOrigin ,  allowedMethods ,  allowedHeaders . If they are not present an error will be thrown while instantiating the middleware.  Optionally you can also specify the keys  allowCredentials  (Bool),  cacheExpiration  (Int) and  exposedHeaders  ([String]).   Afterwards you can add the middleware using the a throwing overload of the initialiser that accepts Vapor's  Config .  let   drop   =   Droplet ()  do   { \n     drop . middleware . insert ( try   CORSMiddleware ( configuration :   drop . config ),   at :   0 )  }   catch   { \n     fatalError ( Error creating CORSMiddleware, please check that you ve setup cors.json correctly. )  }", 
            "title": "JSON Config"
        }, 
        {
            "location": "/leaf/package/", 
            "text": "Using Leaf\n\n\nThis section outlines how to import the Leaf package both with or without a Vapor project.\n\n\nWith Vapor\n\n\nThe easiest way to use Leaf with Vapor is to include the Leaf provider. \n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/vapor.git\n,\n \nmajorVersion\n:\n \n2\n),\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/leaf-provider.git\n,\n \nmajorVersion\n:\n \n1\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nThe Leaf provider package adds Leaf to your project and adds some additional, Vapor-specific conveniences like \ndrop.stem()\n. \n\n\nUse \nimport LeafProvider\n.\n\n\nJust Leaf\n\n\nAt the core of the Leaf provider is a fast, pure Swift templating engine. You can use it with any of your Swift packages or server-side Swift applications.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/leaf.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Leaf\n to access the \nLeaf.Stem\n class.", 
            "title": "Package"
        }, 
        {
            "location": "/leaf/package/#using-leaf", 
            "text": "This section outlines how to import the Leaf package both with or without a Vapor project.", 
            "title": "Using Leaf"
        }, 
        {
            "location": "/leaf/package/#with-vapor", 
            "text": "The easiest way to use Leaf with Vapor is to include the Leaf provider.   import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         . Package ( url :   https://github.com/vapor/vapor.git ,   majorVersion :   2 ), \n         . Package ( url :   https://github.com/vapor/leaf-provider.git ,   majorVersion :   1 ) \n     ], \n     exclude :   [   ...   ]  )   The Leaf provider package adds Leaf to your project and adds some additional, Vapor-specific conveniences like  drop.stem() .   Use  import LeafProvider .", 
            "title": "With Vapor"
        }, 
        {
            "location": "/leaf/package/#just-leaf", 
            "text": "At the core of the Leaf provider is a fast, pure Swift templating engine. You can use it with any of your Swift packages or server-side Swift applications.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/leaf.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Leaf  to access the  Leaf.Stem  class.", 
            "title": "Just Leaf"
        }, 
        {
            "location": "/leaf/provider/", 
            "text": "Leaf Provider\n\n\nAfter you've \nadded the Leaf Provider package\n to your project, setting the provider up in code is easy.\n\n\nAdd to Droplet\n\n\nFirst, register the \nLeafProvider.Provider\n with your Droplet.\n\n\nimport\n \nVapor\n\n\nimport\n \nLeafProvider\n\n\n\nlet\n \nconfig\n \n=\n \ntry\n \nConfig\n()\n\n\ntry\n \nconfig\n.\naddProvider\n(\nLeafProvider\n.\nProvider\n.\nself\n)\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n(\nconfig\n)\n\n\n\n...\n\n\n\n\n\n\nConfigure Droplet\n\n\nOnce the provider is added to your Droplet, you can configure your Droplet to use the Leaf view renderer.\n\n\nConfig/droplet.json\n\n\n{\n\n    \n...,\n\n    \nview\n:\n \nleaf\n,\n\n    \n...\n\n\n}\n\n\n\n\n\n\n\n\nSeealso\n\n\nLearn more about configuration files in the \nSettings guide\n.\n\n\n\n\nManual\n\n\nYou can also set the \ndrop.view\n property manually if you want to hardcode your view renderer.\n\n\nimport\n \nVapor\n\n\nimport\n \nLeafProvider\n\n\n\nlet\n \ndrop\n \n=\n \ntry\n \nDroplet\n()\n\n\n\ndrop\n.\nview\n \n=\n \nLeafRenderer\n(\nviewsDir\n:\n \ndrop\n.\nviewsDir\n)\n\n\n\n\n\n\nDone\n\n\nNext time you boot your application, your views will be rendered using Leaf.", 
            "title": "Provider"
        }, 
        {
            "location": "/leaf/provider/#leaf-provider", 
            "text": "After you've  added the Leaf Provider package  to your project, setting the provider up in code is easy.", 
            "title": "Leaf Provider"
        }, 
        {
            "location": "/leaf/provider/#add-to-droplet", 
            "text": "First, register the  LeafProvider.Provider  with your Droplet.  import   Vapor  import   LeafProvider  let   config   =   try   Config ()  try   config . addProvider ( LeafProvider . Provider . self )  let   drop   =   try   Droplet ( config )  ...", 
            "title": "Add to Droplet"
        }, 
        {
            "location": "/leaf/provider/#configure-droplet", 
            "text": "Once the provider is added to your Droplet, you can configure your Droplet to use the Leaf view renderer.  Config/droplet.json  { \n     ..., \n     view :   leaf , \n     ...  }    Seealso  Learn more about configuration files in the  Settings guide .", 
            "title": "Configure Droplet"
        }, 
        {
            "location": "/leaf/provider/#manual", 
            "text": "You can also set the  drop.view  property manually if you want to hardcode your view renderer.  import   Vapor  import   LeafProvider  let   drop   =   try   Droplet ()  drop . view   =   LeafRenderer ( viewsDir :   drop . viewsDir )", 
            "title": "Manual"
        }, 
        {
            "location": "/leaf/provider/#done", 
            "text": "Next time you boot your application, your views will be rendered using Leaf.", 
            "title": "Done"
        }, 
        {
            "location": "/leaf/leaf/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\nLeaf\n\n\nWelcome to Leaf. Leaf's goal is to be a simple templating language that can make generating views easier. There's a lot of great templating languages, use what's best for you, maybe that's Leaf! The goals of Leaf are as follows:\n\n\n\n\nSmall set of strictly enforced rules\n\n\nConsistency\n\n\nParser first mentality\n\n\nExtensibility\n\n\n\n\nSyntax\n\n\nStructure\n\n\nLeaf Tags are made up of 4 Elements:\n  - Token: \n#\n is the Token\n  - Name: A \nstring\n that identifies the tag\n  - Parameter List: \n()\n May accept 0 or more arguments\n  - Body (optional): \n{}\n Must be separated from the Parameter List by a space\n\n\nThere can be many different usages of these 4 elements depending on the Tag's implementation. Let's look at a few examples of how Leaf's built-in Tags might be used:\n\n\n\n\n#()\n\n\n#(variable)\n\n\n#import(\"template\")\n\n\n#export(\"link\") { \na href=\"#()\"\n/a\n }\n\n\n#index(friends, \"0\")\n\n\n#loop(friends, \"friend\") { \nli\n#(friend.name)\n/li\n }\n\n\n#raw() { \na href=\"#raw\"\nAnything goes!@#$%^\n*\n/a\n }\n\n\n\n\nUsing the \n#\n token in HTML\n\n\nThe \n#\n token cannot be escaped. Use the \n#()\n or \n#raw() {}\n Tag to output a \n#\n in a Leaf Template. \n#()\n =\n \n#\n\n\nRaw HTML\n\n\nAll Leaf output is escaped by default. Use the \n#raw() {}\n Tag for unescaped output.\n\n#raw() { \na href=\"#link\"\nLink\n/a\n }\n =\n \na href=\"#link\"\nLink\n/a\n\n\n\n\nIMPORTANT!  Make sure you are not using the \n#raw() {}\n Tag with user input.\n\n\n\n\nChaining\n\n\nThe double token: \n##\n indicates a chain. It can be applied to any standard Tag. If the previous Tag fails, the chained Tag will be given an opportunity to run.\n\n\n#if(hasFriends) ##embed(\ngetFriends\n)\n\n\n\n\n\nLeaf's built-in Tags\n\n\nToken: \n#()\n\n\n#() #()hashtags #()FTW =\n # #Hashtags #FTW\n\n\n\n\n\nRaw: \n#raw() {}\n\n\n#raw() {\n    Do whatever w/ #\ns here, this code won\nt be rendered as leaf document and is not escaped.\n    It\ns a great place for things like Javascript or large HTML sections.\n}\n\n\n\n\n\nEqual: \n#equal(lhs, rhs) {}\n\n\n#equal(leaf, leaf) { Leaf == Leaf } =\n Leaf == Leaf\n#equal(leaf, mustache) { Leaf == Mustache } =\n\n\n\n\n\n\nVariable: \n#(variable)\n\n\nHello, #(name)!\n\n\n\n\n\nLoop: \n#loop(object, \"index\")\n\n\n#loop(friends, \nfriend\n) {\n  Hello, #(friend.name)!\n}\n\n\n\n\n\nIndex: \n#index(object, _ index: Int|String)\n\n\nHello, #index(friends, 0)!\nHello, #index(friends, \nbest\n)!\n\n\n\n\n\nIf - Else: \n#if(bool) ##else() { this }\n\n\n#if(entering) {\n  Hello, there!\n} ##if(leaving) {\n  Goodbye!\n} ##else() {\n  I\nve been here the whole time.\n}\n\n\n\n\n\nImport: \n#import(\"template\")\n\n\nExport: \n#export(\"template\") { Leaf/HTML }\n\n\nExtend: \n#extend(\"template\")\n\n\nEmbed: \n#embed(\"template\")\n\n\n\n\nWhen using these Layout Tags, omit the template file's .leaf extension.\n\n\n\n\n/// base.leaf\n\n!DOCTYPE html\n\n#import(\nhtml\n)\n\n/// html.leaf\n#extend(\nbase\n)\n\n#export(\nhtml\n) { \nhtml\n#embed(\nbody\n)\n/html\n }\n\n/// body.leaf\n\nbody\n/body\n\n\n\n\n\n\nLeaf renders \nhtml.leaf\n as:\n\n\n!DOCTYPE html\n\n\nhtml\nbody\n/body\n/html\n\n\n\n\n\n\nCustom Tags\n\n\nLook at the existing tags for advanced scenarios, let's look at a basic example by creating \nIndex\n together. This tag will take two arguments, an array, and an index to access.\n\n\nclass\n \nIndex\n:\n \nBasicTag\n \n{\n\n  \nlet\n \nname\n \n=\n \nindex\n\n\n  \nfunc\n \nrun\n(\narguments\n:\n \n[\nArgument\n])\n \nthrows\n \n-\n \nNode\n?\n \n{\n\n    \nguard\n\n      \narguments\n.\ncount\n \n==\n \n2\n,\n\n      \nlet\n \narray\n \n=\n \narguments\n[\n0\n].\nvalue\n?.\nnodeArray\n,\n\n      \nlet\n \nindex\n \n=\n \narguments\n[\n1\n].\nvalue\n?.\nint\n,\n\n      \nindex\n \n \narray\n.\ncount\n\n    \nelse\n \n{\n \nreturn\n \nnil\n \n}\n\n        \nreturn\n \narray\n[\nindex\n]\n\n    \n}\n\n\n}\n\n\n\n\n\n\nWe can now register this Tag in our \nmain.swift\n file with:\n\n\nif\n \nlet\n \nleaf\n \n=\n \ndrop\n.\nview\n \nas\n?\n \nLeafRenderer\n \n{\n\n    \nleaf\n.\nstem\n.\nregister\n(\nIndex\n())\n\n\n}\n\n\n\n\n\n\nAnd use it just like we did \nabove\n.\n\n\n\n\nNote: Use of non-alphanumeric characters in Tag Names is \nstrongly discouraged\n and may be disallowed in future versions of Leaf.\n\n\n\n\nSyntax Highlighting\n\n\nAtom\n\n\nlanguage-leaf\n by ButkiewiczP\n\n\nXcode\n\n\nIt is not currently possible to implement Leaf Syntax Highlighting in Xcode, however, using Xcode's HTML Syntax Coloring can help a bit. Select one or more Leaf files and then choose Editor \n Syntax Coloring \n HTML.  Your selected Leaf files will now use Xcode's HTML Syntax Coloring.  Unfortunately the usefulness of this is limited because this association will be removed when \nvapor xcode\n is run.\n\n\nThere appears to be a way to \nmake Xcode file associations persist\n but that requires a bit more kung-fu.\n\n\nVS Code\n\n\nhtml-leaf\n by FranciscoAmado\n\n\nCLion \n AppCode\n\n\nSome preliminary work has been done to implement a Leaf Plugin for CLion \n AppCode but lack of skill and interest in Java has slowed progress! If you have IntelliJ SDK experience and want to help with this, message Tom Holland on \nVapor Slack", 
            "title": "Overview"
        }, 
        {
            "location": "/leaf/leaf/#leaf", 
            "text": "Welcome to Leaf. Leaf's goal is to be a simple templating language that can make generating views easier. There's a lot of great templating languages, use what's best for you, maybe that's Leaf! The goals of Leaf are as follows:   Small set of strictly enforced rules  Consistency  Parser first mentality  Extensibility", 
            "title": "Leaf"
        }, 
        {
            "location": "/leaf/leaf/#syntax", 
            "text": "", 
            "title": "Syntax"
        }, 
        {
            "location": "/leaf/leaf/#structure", 
            "text": "Leaf Tags are made up of 4 Elements:\n  - Token:  #  is the Token\n  - Name: A  string  that identifies the tag\n  - Parameter List:  ()  May accept 0 or more arguments\n  - Body (optional):  {}  Must be separated from the Parameter List by a space  There can be many different usages of these 4 elements depending on the Tag's implementation. Let's look at a few examples of how Leaf's built-in Tags might be used:   #()  #(variable)  #import(\"template\")  #export(\"link\") {  a href=\"#()\" /a  }  #index(friends, \"0\")  #loop(friends, \"friend\") {  li #(friend.name) /li  }  #raw() {  a href=\"#raw\" Anything goes!@#$%^ * /a  }", 
            "title": "Structure"
        }, 
        {
            "location": "/leaf/leaf/#using-the-token-in-html", 
            "text": "The  #  token cannot be escaped. Use the  #()  or  #raw() {}  Tag to output a  #  in a Leaf Template.  #()  =   #", 
            "title": "Using the # token in HTML"
        }, 
        {
            "location": "/leaf/leaf/#raw-html", 
            "text": "All Leaf output is escaped by default. Use the  #raw() {}  Tag for unescaped output. #raw() {  a href=\"#link\" Link /a  }  =   a href=\"#link\" Link /a   IMPORTANT!  Make sure you are not using the  #raw() {}  Tag with user input.", 
            "title": "Raw HTML"
        }, 
        {
            "location": "/leaf/leaf/#chaining", 
            "text": "The double token:  ##  indicates a chain. It can be applied to any standard Tag. If the previous Tag fails, the chained Tag will be given an opportunity to run.  #if(hasFriends) ##embed( getFriends )", 
            "title": "Chaining"
        }, 
        {
            "location": "/leaf/leaf/#leafs-built-in-tags", 
            "text": "", 
            "title": "Leaf's built-in Tags"
        }, 
        {
            "location": "/leaf/leaf/#token", 
            "text": "#() #()hashtags #()FTW =  # #Hashtags #FTW", 
            "title": "Token: #()"
        }, 
        {
            "location": "/leaf/leaf/#raw-raw", 
            "text": "#raw() {\n    Do whatever w/ # s here, this code won t be rendered as leaf document and is not escaped.\n    It s a great place for things like Javascript or large HTML sections.\n}", 
            "title": "Raw: #raw() {}"
        }, 
        {
            "location": "/leaf/leaf/#equal-equallhs-rhs", 
            "text": "#equal(leaf, leaf) { Leaf == Leaf } =  Leaf == Leaf\n#equal(leaf, mustache) { Leaf == Mustache } =", 
            "title": "Equal: #equal(lhs, rhs) {}"
        }, 
        {
            "location": "/leaf/leaf/#variable-variable", 
            "text": "Hello, #(name)!", 
            "title": "Variable: #(variable)"
        }, 
        {
            "location": "/leaf/leaf/#loop-loopobject-index", 
            "text": "#loop(friends,  friend ) {\n  Hello, #(friend.name)!\n}", 
            "title": "Loop: #loop(object, \"index\")"
        }, 
        {
            "location": "/leaf/leaf/#index-indexobject-_-index-intstring", 
            "text": "Hello, #index(friends, 0)!\nHello, #index(friends,  best )!", 
            "title": "Index: #index(object, _ index: Int|String)"
        }, 
        {
            "location": "/leaf/leaf/#if-else-ifbool-else-this", 
            "text": "#if(entering) {\n  Hello, there!\n} ##if(leaving) {\n  Goodbye!\n} ##else() {\n  I ve been here the whole time.\n}", 
            "title": "If - Else: #if(bool) ##else() { this }"
        }, 
        {
            "location": "/leaf/leaf/#import-importtemplate", 
            "text": "", 
            "title": "Import: #import(\"template\")"
        }, 
        {
            "location": "/leaf/leaf/#export-exporttemplate-leafhtml", 
            "text": "", 
            "title": "Export: #export(\"template\") { Leaf/HTML }"
        }, 
        {
            "location": "/leaf/leaf/#extend-extendtemplate", 
            "text": "", 
            "title": "Extend: #extend(\"template\")"
        }, 
        {
            "location": "/leaf/leaf/#embed-embedtemplate", 
            "text": "When using these Layout Tags, omit the template file's .leaf extension.   /// base.leaf !DOCTYPE html \n#import( html )\n\n/// html.leaf\n#extend( base )\n\n#export( html ) {  html #embed( body ) /html  }\n\n/// body.leaf body /body   Leaf renders  html.leaf  as:  !DOCTYPE html  html body /body /html", 
            "title": "Embed: #embed(\"template\")"
        }, 
        {
            "location": "/leaf/leaf/#custom-tags", 
            "text": "Look at the existing tags for advanced scenarios, let's look at a basic example by creating  Index  together. This tag will take two arguments, an array, and an index to access.  class   Index :   BasicTag   { \n   let   name   =   index \n\n   func   run ( arguments :   [ Argument ])   throws   -   Node ?   { \n     guard \n       arguments . count   ==   2 , \n       let   array   =   arguments [ 0 ]. value ?. nodeArray , \n       let   index   =   arguments [ 1 ]. value ?. int , \n       index     array . count \n     else   {   return   nil   } \n         return   array [ index ] \n     }  }   We can now register this Tag in our  main.swift  file with:  if   let   leaf   =   drop . view   as ?   LeafRenderer   { \n     leaf . stem . register ( Index ())  }   And use it just like we did  above .   Note: Use of non-alphanumeric characters in Tag Names is  strongly discouraged  and may be disallowed in future versions of Leaf.", 
            "title": "Custom Tags"
        }, 
        {
            "location": "/leaf/leaf/#syntax-highlighting", 
            "text": "", 
            "title": "Syntax Highlighting"
        }, 
        {
            "location": "/leaf/leaf/#atom", 
            "text": "language-leaf  by ButkiewiczP", 
            "title": "Atom"
        }, 
        {
            "location": "/leaf/leaf/#xcode", 
            "text": "It is not currently possible to implement Leaf Syntax Highlighting in Xcode, however, using Xcode's HTML Syntax Coloring can help a bit. Select one or more Leaf files and then choose Editor   Syntax Coloring   HTML.  Your selected Leaf files will now use Xcode's HTML Syntax Coloring.  Unfortunately the usefulness of this is limited because this association will be removed when  vapor xcode  is run.  There appears to be a way to  make Xcode file associations persist  but that requires a bit more kung-fu.", 
            "title": "Xcode"
        }, 
        {
            "location": "/leaf/leaf/#vs-code", 
            "text": "html-leaf  by FranciscoAmado", 
            "title": "VS Code"
        }, 
        {
            "location": "/leaf/leaf/#clion-appcode", 
            "text": "Some preliminary work has been done to implement a Leaf Plugin for CLion   AppCode but lack of skill and interest in Java has slowed progress! If you have IntelliJ SDK experience and want to help with this, message Tom Holland on  Vapor Slack", 
            "title": "CLion & AppCode"
        }, 
        {
            "location": "/validation/overview/", 
            "text": "Warning\n\n\nThis section may contain outdated information.\n\n\n\n\nValidation\n\n\nVapor provides a few different ways to validate data coming into your application. Let's start by looking at the most common.\n\n\nCommon Usage\n\n\nSeveral useful convenience validators are included by default. You can use these to validate data coming into your application, or combine them and create your own.\n\n\nLet's look at the most common way to validate data.\n\n\nclass\n \nEmployee\n \n{\n\n    \nvar\n \nemail\n:\n \nValid\nEmail\n\n    \nvar\n \nname\n:\n \nValid\nName\n\n\n    \ninit\n(\nrequest\n:\n \nRequest\n)\n \nthrows\n \n{\n\n        \nname\n \n=\n \ntry\n \nrequest\n.\ndata\n[\nname\n].\nvalidated\n()\n\n        \nemail\n \n=\n \ntry\n \nrequest\n.\ndata\n[\nemail\n].\nvalidated\n()\n\n    \n}\n\n\n}\n\n\n\n\n\n\nHere we have a typical Employee model with an \nemail\n and \nname\n property. By declaring both of these properties as \nValid\n, you are ensuring that these properties can only ever contain valid data. The Swift type checking system will prevent anything that does not pass validation from being stored.\n\n\nTo store something in a \nValid\n property, you must use the \n.validated()\n method. This is available for any data returned by \nrequest.data\n.\n\n\nEmail\n is a real \nvalidator\n included with Vapor, but \nName\n is not. Let's take a look at how you can create a Validator.\n\n\nValid\nOnlyAlphanumeric\n\n\nValid\nEmail\n\n\nValid\nUnique\nT\n\n\nValid\nMatches\nT\n\n\nValid\nIn\nT\n\n\nValid\nContains\nT\n\n\nValid\nCount\nT\n\n\n\n\n\n\nValidators vs. ValidationSuites\n\n\nValidators, like \nCount\n or \nContains\n can have multiple configurations. For example:\n\n\nlet\n \nname\n:\n \nValid\nCount\nString\n \n=\n \ntry\n \nVapor\n.\nvalidated\n(\nby\n:\n \nCount\n.\nmax\n(\n5\n))\n\n\n\n\n\n\nHere we are validating that the \nString\n is at most 5 characters long. The type of \nValid\nCount\n tells us that the string has been validated to be a certain count, but it does not tell us exactly what that count was. The string could have been validated to be less than three characters or more than one million.\n\n\nBecause of this, \nValidators\n themselves are not as type safe as some applications might desire. \nValidationSuites\n fix this. They combine multiple \nValidators\n and/or \nValidationSuites\n together to represent exactly what type of data should be considered valid.\n\n\nCustom Validator\n\n\nHere is how to create a custom \nValidationSuite\n.\n\n\nclass\n \nName\n:\n \nValidationSuite\n \n{\n\n    \nstatic\n \nfunc\n \nvalidate\n(\ninput\n \nvalue\n:\n \nString\n)\n \nthrows\n \n{\n\n        \nlet\n \nevaluation\n \n=\n \nOnlyAlphanumeric\n.\nself\n\n            \n \nCount\n.\nmin\n(\n5\n)\n\n            \n \nCount\n.\nmax\n(\n20\n)\n\n\n        \ntry\n \nevaluation\n.\nvalidate\n(\ninput\n:\n \nvalue\n)\n\n    \n}\n\n\n}\n\n\n\n\n\n\nYou only have to implement one method. In this method, use any other validators or logic to create your custom validator. Here we are defining a \nName\n as only accepting alphanumeric Strings that are between 5 and 20 characters.\n\n\nNow we can be sure that anything of type \nValid\nName\n follows these rules.\n\n\nCombining Validators\n\n\nIn the \nName\n validator, you can see that \n is being used to combine validators. You can use \n as well as \n||\n to combine any validator as you would boolean values with an \nif\n statement.\n\n\nYou can also use \n!\n to invert the validator.\n\n\nlet\n \nsymbols\n \n=\n \ninput\n.\nvalidated\n(\nby\n:\n \n!\nOnlyAlphanumeric\n.\nself\n)\n\n\n\n\n\n\nTesting Validity\n\n\nWhile \nvalidated() throw\n is the most common method for validating, there are two others.\n\n\nlet\n \npassed\n \n=\n \ninput\n.\npasses\n(\nCount\n.\nmin\n(\n5\n))\n\n\nlet\n \nvalid\n \n=\n \ntry\n \ninput\n.\ntested\n(\nCount\n.\nmin\n(\n5\n))\n\n\n\n\n\n\npasses()\n returns a boolean indicating whether or not the test passed. \ntested()\n will throw if the validation does not pass. But unlike \nvalidated()\n which returns a \nValid\n type, \ntested()\n returns the original type of the item it was called on.\n\n\nValidation Failures\n\n\nVapor will automatically catch validation failures in the \nValidationMiddleware\n. But you can catch them on your own, or customize responses for certain types of failures.\n\n\ndo\n \n{\n\n    \n//validation here\n\n\n}\n \ncatch\n \nlet\n \nerror\n \nas\n \nValidationErrorProtocol\n \n{\n\n    \nprint\n(\nerror\n.\nmessage\n)\n\n\n}", 
            "title": "Validation"
        }, 
        {
            "location": "/validation/overview/#validation", 
            "text": "Vapor provides a few different ways to validate data coming into your application. Let's start by looking at the most common.", 
            "title": "Validation"
        }, 
        {
            "location": "/validation/overview/#common-usage", 
            "text": "Several useful convenience validators are included by default. You can use these to validate data coming into your application, or combine them and create your own.  Let's look at the most common way to validate data.  class   Employee   { \n     var   email :   Valid Email \n     var   name :   Valid Name \n\n     init ( request :   Request )   throws   { \n         name   =   try   request . data [ name ]. validated () \n         email   =   try   request . data [ email ]. validated () \n     }  }   Here we have a typical Employee model with an  email  and  name  property. By declaring both of these properties as  Valid , you are ensuring that these properties can only ever contain valid data. The Swift type checking system will prevent anything that does not pass validation from being stored.  To store something in a  Valid  property, you must use the  .validated()  method. This is available for any data returned by  request.data .  Email  is a real  validator  included with Vapor, but  Name  is not. Let's take a look at how you can create a Validator.  Valid OnlyAlphanumeric  Valid Email  Valid Unique T  Valid Matches T  Valid In T  Valid Contains T  Valid Count T", 
            "title": "Common Usage"
        }, 
        {
            "location": "/validation/overview/#validators-vs-validationsuites", 
            "text": "Validators, like  Count  or  Contains  can have multiple configurations. For example:  let   name :   Valid Count String   =   try   Vapor . validated ( by :   Count . max ( 5 ))   Here we are validating that the  String  is at most 5 characters long. The type of  Valid Count  tells us that the string has been validated to be a certain count, but it does not tell us exactly what that count was. The string could have been validated to be less than three characters or more than one million.  Because of this,  Validators  themselves are not as type safe as some applications might desire.  ValidationSuites  fix this. They combine multiple  Validators  and/or  ValidationSuites  together to represent exactly what type of data should be considered valid.", 
            "title": "Validators vs. ValidationSuites"
        }, 
        {
            "location": "/validation/overview/#custom-validator", 
            "text": "Here is how to create a custom  ValidationSuite .  class   Name :   ValidationSuite   { \n     static   func   validate ( input   value :   String )   throws   { \n         let   evaluation   =   OnlyAlphanumeric . self \n               Count . min ( 5 ) \n               Count . max ( 20 ) \n\n         try   evaluation . validate ( input :   value ) \n     }  }   You only have to implement one method. In this method, use any other validators or logic to create your custom validator. Here we are defining a  Name  as only accepting alphanumeric Strings that are between 5 and 20 characters.  Now we can be sure that anything of type  Valid Name  follows these rules.", 
            "title": "Custom Validator"
        }, 
        {
            "location": "/validation/overview/#combining-validators", 
            "text": "In the  Name  validator, you can see that   is being used to combine validators. You can use   as well as  ||  to combine any validator as you would boolean values with an  if  statement.  You can also use  !  to invert the validator.  let   symbols   =   input . validated ( by :   ! OnlyAlphanumeric . self )", 
            "title": "Combining Validators"
        }, 
        {
            "location": "/validation/overview/#testing-validity", 
            "text": "While  validated() throw  is the most common method for validating, there are two others.  let   passed   =   input . passes ( Count . min ( 5 ))  let   valid   =   try   input . tested ( Count . min ( 5 ))   passes()  returns a boolean indicating whether or not the test passed.  tested()  will throw if the validation does not pass. But unlike  validated()  which returns a  Valid  type,  tested()  returns the original type of the item it was called on.", 
            "title": "Testing Validity"
        }, 
        {
            "location": "/validation/overview/#validation-failures", 
            "text": "Vapor will automatically catch validation failures in the  ValidationMiddleware . But you can catch them on your own, or customize responses for certain types of failures.  do   { \n     //validation here  }   catch   let   error   as   ValidationErrorProtocol   { \n     print ( error . message )  }", 
            "title": "Validation Failures"
        }, 
        {
            "location": "/node/package/", 
            "text": "Using Node\n\n\nWith Vapor\n\n\nThis package is included with Vapor by default, just add:\n\n\nimport\n \nNode\n\n\n\n\n\n\nWithout Vapor\n\n\nNode provides a lot of conveniences for any server-side, or client side Swift project. To include it in your package, add the following to your \nPackage.swift\n file.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/node.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Node\n to access Node's APIs\n.", 
            "title": "Package"
        }, 
        {
            "location": "/node/package/#using-node", 
            "text": "", 
            "title": "Using Node"
        }, 
        {
            "location": "/node/package/#with-vapor", 
            "text": "This package is included with Vapor by default, just add:  import   Node", 
            "title": "With Vapor"
        }, 
        {
            "location": "/node/package/#without-vapor", 
            "text": "Node provides a lot of conveniences for any server-side, or client side Swift project. To include it in your package, add the following to your  Package.swift  file.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/node.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Node  to access Node's APIs\n.", 
            "title": "Without Vapor"
        }, 
        {
            "location": "/core/package/", 
            "text": "Using Core\n\n\nWith Vapor\n\n\nThis package is included with Vapor by default, just add:\n\n\nimport\n \nCore\n\n\n\n\n\n\nWithout Vapor\n\n\nCore provides a lot of conveniences for any server-side Swift project. To include it in your package, add the following to your \nPackage.swift\n file.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/core.git\n,\n \nmajorVersion\n:\n \n2\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Core\n to access Core's APIs.", 
            "title": "Package"
        }, 
        {
            "location": "/core/package/#using-core", 
            "text": "", 
            "title": "Using Core"
        }, 
        {
            "location": "/core/package/#with-vapor", 
            "text": "This package is included with Vapor by default, just add:  import   Core", 
            "title": "With Vapor"
        }, 
        {
            "location": "/core/package/#without-vapor", 
            "text": "Core provides a lot of conveniences for any server-side Swift project. To include it in your package, add the following to your  Package.swift  file.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/core.git ,   majorVersion :   2 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Core  to access Core's APIs.", 
            "title": "Without Vapor"
        }, 
        {
            "location": "/core/overview/", 
            "text": "Core\n\n\nCore provides some conveniences for common tasks.\n\n\n\n\nNote\n\n\nUse \nimport Core\n\n\n\n\nBackground\n\n\nEasily create a background thread using \nbackground()\n\n\nprint\n(\nhello\n)\n\n\n\ntry\n \nbackground\n \n{\n\n    \nprint\n(\nworld\n)\n  \n\n}\n\n\n\n\n\n\nPortal\n\n\nPortals allow you to make async tasks blocking.\n\n\nlet\n \nresult\n \n=\n \ntry\n \nPortal\n.\nopen\n \n{\n \nportal\n \nin\n\n    \nsomeAsyncTask\n \n{\n \nresult\n \nin\n\n        \nportal\n.\nclose\n(\nwith\n:\n \nresult\n)\n\n    \n}\n\n\n}\n\n\n\nprint\n(\nresult\n)\n \n// the result from the async task\n\n\n\n\n\n\nRFC1123\n\n\nCreate RFC1123 type dates.\n\n\nlet\n \nnow\n \n=\n \nDate\n().\nrfc1123\n \n// string \n\n\n\n\n\n\nYou can also parse RFC1123 strings.\n\n\nlet parsed = Date(rfc1123: \nMon, 10 Apr 2017 11:26:13 GMT\n)", 
            "title": "Overview"
        }, 
        {
            "location": "/core/overview/#core", 
            "text": "Core provides some conveniences for common tasks.   Note  Use  import Core", 
            "title": "Core"
        }, 
        {
            "location": "/core/overview/#background", 
            "text": "Easily create a background thread using  background()  print ( hello )  try   background   { \n     print ( world )    }", 
            "title": "Background"
        }, 
        {
            "location": "/core/overview/#portal", 
            "text": "Portals allow you to make async tasks blocking.  let   result   =   try   Portal . open   {   portal   in \n     someAsyncTask   {   result   in \n         portal . close ( with :   result ) \n     }  }  print ( result )   // the result from the async task", 
            "title": "Portal"
        }, 
        {
            "location": "/core/overview/#rfc1123", 
            "text": "Create RFC1123 type dates.  let   now   =   Date (). rfc1123   // string    You can also parse RFC1123 strings.  let parsed = Date(rfc1123:  Mon, 10 Apr 2017 11:26:13 GMT )", 
            "title": "RFC1123"
        }, 
        {
            "location": "/bits/package/", 
            "text": "Using Bits\n\n\nWith Vapor\n\n\nThis package is included with Vapor by default, just add:\n\n\nimport\n \nBits\n\n\n\n\n\n\nWithout Vapor\n\n\nBits provides a lot of byte-manipulation conveniences for any server-side Swift project. To include it in your package, add the following to your \nPackage.swift\n file.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/bits.git\n,\n \nmajorVersion\n:\n \n1\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Bits\n to access Bits' APIs.", 
            "title": "Package"
        }, 
        {
            "location": "/bits/package/#using-bits", 
            "text": "", 
            "title": "Using Bits"
        }, 
        {
            "location": "/bits/package/#with-vapor", 
            "text": "This package is included with Vapor by default, just add:  import   Bits", 
            "title": "With Vapor"
        }, 
        {
            "location": "/bits/package/#without-vapor", 
            "text": "Bits provides a lot of byte-manipulation conveniences for any server-side Swift project. To include it in your package, add the following to your  Package.swift  file.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/bits.git ,   majorVersion :   1 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Bits  to access Bits' APIs.", 
            "title": "Without Vapor"
        }, 
        {
            "location": "/bits/overview/", 
            "text": "Bits\n\n\nThe bits package is included in Vapor by default and provides a convenient API for working with bytes.\n\n\n\n\nNote\n\n\nUse \nimport Bits\n to use this package.\n\n\n\n\nTypealias\n\n\nThe bits package provides two type-aliases for bytes.\n\n\ntypealias\n \nByte\n \n=\n \nUInt8\n\n\ntypealias\n \nBytes\n \n=\n \n[\nByte\n]\n\n\n\n\n\n\nBytesConvertible\n\n\nIt's quite often that we want to convert objects to and from byte arrays when we're working. The BytesConvertible helps define objects that have these capabilities. This is implemented already on most objects in Vapor that can/should be converted to and from byte arrays.\n\n\nlet\n \nhello\n \n=\n \nString\n(\nbytes\n:\n \n[\n72\n,\n \n101\n,\n \n108\n,\n \n108\n,\n \n111\n])\n\n\nlet\n \nbytes\n \n=\n \nhello\n.\nmakeBytes\n()\n \n\n\n\n\n\nString\n\n\nConverting from bytes to string using the UTF-8 encoding is easy.\n\n\nlet\n \nbytes\n \n=\n \nhello\n.\nmakeBytes\n()\n\n\nlet\n \nstring\n \n=\n \nbytes\n.\nmakeString\n()\n\n\nprint\n(\nstring\n)\n \n// \nhello\n\n\n\n\n\n\nByte\n\n\nThe upper and lowercase latin alphabet and some additional control characters are statically typed on the \nByte\n.\n\n\nlet\n \nbytes\n:\n \nBytes\n \n=\n \n[.\nh\n,\n \n.\ne\n,\n \n.\nl\n,\n \n.\nl\n,\n \n.\no\n]\n\n\nprint\n(\nbytes\n.\nmakeString\n())\n \n// \nhello\n\n\n\n\n\n\nThis makes byte manipulation and comparison easy and is useful for building things like parsers and serializers.\n\n\nlet\n \nbyte\n:\n \nByte\n \n=\n \n65\n\n\nif\n \nbyte\n \n==\n \n.\nA\n \n{\n\n    \nprint\n(\nfound A!\n)\n\n\n}", 
            "title": "Overview"
        }, 
        {
            "location": "/bits/overview/#bits", 
            "text": "The bits package is included in Vapor by default and provides a convenient API for working with bytes.   Note  Use  import Bits  to use this package.", 
            "title": "Bits"
        }, 
        {
            "location": "/bits/overview/#typealias", 
            "text": "The bits package provides two type-aliases for bytes.  typealias   Byte   =   UInt8  typealias   Bytes   =   [ Byte ]", 
            "title": "Typealias"
        }, 
        {
            "location": "/bits/overview/#bytesconvertible", 
            "text": "It's quite often that we want to convert objects to and from byte arrays when we're working. The BytesConvertible helps define objects that have these capabilities. This is implemented already on most objects in Vapor that can/should be converted to and from byte arrays.  let   hello   =   String ( bytes :   [ 72 ,   101 ,   108 ,   108 ,   111 ])  let   bytes   =   hello . makeBytes ()", 
            "title": "BytesConvertible"
        }, 
        {
            "location": "/bits/overview/#string", 
            "text": "Converting from bytes to string using the UTF-8 encoding is easy.  let   bytes   =   hello . makeBytes ()  let   string   =   bytes . makeString ()  print ( string )   //  hello", 
            "title": "String"
        }, 
        {
            "location": "/bits/overview/#byte", 
            "text": "The upper and lowercase latin alphabet and some additional control characters are statically typed on the  Byte .  let   bytes :   Bytes   =   [. h ,   . e ,   . l ,   . l ,   . o ]  print ( bytes . makeString ())   //  hello   This makes byte manipulation and comparison easy and is useful for building things like parsers and serializers.  let   byte :   Byte   =   65  if   byte   ==   . A   { \n     print ( found A! )  }", 
            "title": "Byte"
        }, 
        {
            "location": "/debugging/package/", 
            "text": "Using Debugging\n\n\nWith Vapor\n\n\nThis package is included with Vapor by default, just add:\n\n\nimport\n \nDebugging\n\n\n\n\n\n\nWithout Vapor\n\n\nDebugging is a convenient protocol for providing more information about error messages. You can use it in any of your Swift projects.\n\n\nimport\n \nPackageDescription\n\n\n\nlet\n \npackage\n \n=\n \nPackage\n(\n\n    \nname\n:\n \nProject\n,\n\n    \ndependencies\n:\n \n[\n\n        \n...\n\n        \n.\nPackage\n(\nurl\n:\n \nhttps://github.com/vapor/debugging.git\n,\n \nmajorVersion\n:\n \n1\n)\n\n    \n],\n\n    \nexclude\n:\n \n[\n \n...\n \n]\n\n\n)\n\n\n\n\n\n\nUse \nimport Debugging\n to access Debugging' APIs.", 
            "title": "Package"
        }, 
        {
            "location": "/debugging/package/#using-debugging", 
            "text": "", 
            "title": "Using Debugging"
        }, 
        {
            "location": "/debugging/package/#with-vapor", 
            "text": "This package is included with Vapor by default, just add:  import   Debugging", 
            "title": "With Vapor"
        }, 
        {
            "location": "/debugging/package/#without-vapor", 
            "text": "Debugging is a convenient protocol for providing more information about error messages. You can use it in any of your Swift projects.  import   PackageDescription  let   package   =   Package ( \n     name :   Project , \n     dependencies :   [ \n         ... \n         . Package ( url :   https://github.com/vapor/debugging.git ,   majorVersion :   1 ) \n     ], \n     exclude :   [   ...   ]  )   Use  import Debugging  to access Debugging' APIs.", 
            "title": "Without Vapor"
        }, 
        {
            "location": "/debugging/overview/", 
            "text": "Debugging\n\n\nConforming your error types to \nDebuggable\n allows Vapor to create richer error messages and makes debugging easier.\n\n\nimport\n \nDebugging\n\n\n\nextension\n \nFooError\n:\n \nDebuggable\n \n{\n\n    \n// conform here\n\n\n}\n\n\n\n\n\n\nNow when a \nFooError\n is thrown, you will get a nice message in your console.\n\n\nFoo Error: You \ndo\n not have a \n`\nfoo\n`\n.\n\nIdentifier: DebuggingTests.FooError.noFoo\n\nHere are some possible causes: \n- You did not \nset\n the flongwaffle.\n- The session ended before a \n`\nFoo\n`\n could be made.\n- The universe conspires against us all.\n- Computers are hard.\n\nThese suggestions could address the issue: \n- You really want to use a \n`\nBar\n`\n here.\n- Take up the guitar and move to the beach.\n\nVapor\ns documentation talks about this: \n- http://documentation.com/Foo\n- http://documentation.com/foo/noFoo", 
            "title": "Overview"
        }, 
        {
            "location": "/debugging/overview/#debugging", 
            "text": "Conforming your error types to  Debuggable  allows Vapor to create richer error messages and makes debugging easier.  import   Debugging  extension   FooError :   Debuggable   { \n     // conform here  }   Now when a  FooError  is thrown, you will get a nice message in your console.  Foo Error: You  do  not have a  ` foo ` .\n\nIdentifier: DebuggingTests.FooError.noFoo\n\nHere are some possible causes: \n- You did not  set  the flongwaffle.\n- The session ended before a  ` Foo `  could be made.\n- The universe conspires against us all.\n- Computers are hard.\n\nThese suggestions could address the issue: \n- You really want to use a  ` Bar `  here.\n- Take up the guitar and move to the beach.\n\nVapor s documentation talks about this: \n- http://documentation.com/Foo\n- http://documentation.com/foo/noFoo", 
            "title": "Debugging"
        }, 
        {
            "location": "/advanced/modules/", 
            "text": "", 
            "title": "Modules"
        }, 
        {
            "location": "/switch/1_5/", 
            "text": "Redirecting...", 
            "title": "1.5"
        }, 
        {
            "location": "/switch/1_5/#redirecting", 
            "text": "", 
            "title": "Redirecting..."
        }, 
        {
            "location": "/switch/2_0/", 
            "text": "Redirecting...", 
            "title": "2.0"
        }, 
        {
            "location": "/switch/2_0/#redirecting", 
            "text": "", 
            "title": "Redirecting..."
        }
    ]
}