The only explanation I can think of for the failure in
https://go.dev/issue/49387#issuecomment-1303979877 is that maybe the
idle timeout started as soon as conn1 was closed, without waiting for
conn2 to be closed. That might be possible if the connection returned
by Dial was still in the server's accept queue, but never actually
accepted.
To eliminate that possibility, we can send an RPC on that connection
and wait for a response, as we already do with conn1. Since the conn1
RPC succeeded, we know that the connection is non-idle, conn2 should
be accepted, and the request on conn2 should succeed unconditionally.
Fixesgolang/go#49387 (hopefully for real this time).
Change-Id: Ie3e74f91d322223d82c000fdf1f3a0ed08afd20d
Reviewed-on: https://go-review.googlesource.com/c/tools/+/448096
gopls-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Fix the test failure demonstrated in the following failed builder:
https://build.golang.org/log/d0511c583201e8701e72066985ebf950d9f5511d
It should be OK to set multiple keys in the validated map. Support this
by keeping track of seen and deletion clock time. There are still
potential problems with this analysis (specifically, if a map is
constructed via SetAll), but we ignore those problems for now.
Change-Id: I5940d25f18afe31e13bc71f74d4eea7d737d593d
Reviewed-on: https://go-review.googlesource.com/c/tools/+/448696
Reviewed-by: Alan Donovan <adonovan@google.com>
Auto-Submit: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
I meant to do this in the first CL, but was prevented by
a bug which I have since reported and linked to from the code.
Change-Id: I651e728c535cdeb0885eae4d510fda3c24518dcf
Reviewed-on: https://go-review.googlesource.com/c/tools/+/448376
Auto-Submit: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The old code based on packages.Visit traversed in a
deterministic order, and didn't stop when it found its
target (the 'return false' only prunes that subtree).
This CL replaces it with a precomputation of the
PkgPath-to-*Package mapping.
The performance difference is small for this test but
it nearly dominates on a larger input (e.g. k8s).
Example code shouldn't steer users into asymptotic traps.
Change-Id: I19f4fc2c25da3d2ae00090704df30a54d8516bf5
Reviewed-on: https://go-review.googlesource.com/c/tools/+/447958
gopls-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
This change adds an internal API for marshalling and
unmarshalling a types.Package to "shallow" export data,
which does not index packages other than the main one.
The import function accepts a function that loads symbols
on demand (e.g. by recursively reading export data for
indirect dependencies).
The CL includes a test that the entire standard
library can be type-checked using shallow data.
Also:
- break dependency on go/ast.
- narrow the name and type of qualifiedObject.
- add (test) dependency on errgroup, and tidy go.mod.
Change-Id: I92d31efd343cf5dd6fca6d7b918a23749e2d1e83
Reviewed-on: https://go-review.googlesource.com/c/tools/+/447737
Run-TryBot: Alan Donovan <adonovan@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
We plan to add experimental features to this package for
use by gopls, but the directory structure makes this
tricky using the "internal directory" mechanism.
Change-Id: Ib842c0b100b167f6978c6ff783ea0e5d0704b4a7
Reviewed-on: https://go-review.googlesource.com/c/tools/+/447955
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Alan Donovan <adonovan@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Remove the internal guard preventing the loopclosure analyzer from
checking parallel subtests.
Also, improve the accuracy of the parallel subtest check:
- only consider statements after the final labeled statement in the
subtest body
- verify that the *testing.T value for which T.Parallel() is invoked
matches the argument to the subtest literal
Fixesgolang/go#55972
Change-Id: Ia2d9e08dfa88b5e31a9151872025272560d4b5e8
Reviewed-on: https://go-review.googlesource.com/c/tools/+/447256
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Tim King <taking@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
Prior to this CL we already shut down a jsonrpc2_v2.Conn when its
Reader breaks, which we expect to be the common shutdown path.
However, with certain kinds of connections (notably those over
stdin+stdout), it is possible for the Writer side to fail while
the Reader remains working.
If the Writer has failed, we have no way to return the required
Response messages for incoming calls, nor to write new Request
messages of our own. Since we have no way to return a response,
we will now mark those incoming calls as canceled.
However, even if the Writer has failed we may still be able to read
the responses for any outgoing calls that are already in flight. When
our in-flight calls complete, we could in theory even continue to
process Notification messages from the Reader; however, those are
unlikely to be useful with half the connection broken. It seems more
helpful — and less surprising — to go ahead and shut down the
connection completely when it becomes idle.
This is a redo of CL 446315, with additional fixes for bugs exposed on
the -race builders and some extra code cleanup from the process of
diagnosing those bugs.
Updates golang/go#46520.
Updates golang/go#49387.
Change-Id: I746409a7aa2c22d5651448ed0135b5ac21a9808e
Reviewed-on: https://go-review.googlesource.com/c/tools/+/447035
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Prior to this CL we already shut down a jsonrpc2_v2.Conn when its
Reader breaks, which we expect to be the common shutdown path.
However, with certain kinds of connections (notably those over
stdin+stdout), it is possible for the Writer side to fail while
the Reader remains working.
If the Writer has failed, we have no way to return the required
Response messages for incoming calls, nor to write new Request
messages of our own. Since we have no way to return a response,
we will now mark those incoming calls as canceled.
However, even if the Writer has failed we may still be able to read
the responses for any outgoing calls that are already in flight. When
our in-flight calls complete, we could in theory even continue to
process Notification messages from the Reader; however, those are
unlikely to be useful with half the connection broken. It seems more
helpful — and less surprising — to go ahead and shut down the
connection completely when it becomes idle.
Updates golang/go#46520.
Updates golang/go#49387.
Change-Id: I713f172ca7031f4211da321560fe7eae57960a48
Reviewed-on: https://go-review.googlesource.com/c/tools/+/446315
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Serve had a misleading name and signature: it did not actually block
on serving the connection, and never returned a non-nil error.
Updates golang/go#56281.
Change-Id: Ia6df0ba20066811b0551df3b3267dff2fffd7881
Reviewed-on: https://go-review.googlesource.com/c/tools/+/443678
Reviewed-by: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Also make (*Connection).Close safe to call from Bind.
A jsonrpc2_v2.Server has no good way to report an error from Bind.
If the Server saves the error to return from its own Wait method, that
might not ever actually happen: Wait waits for in-flight connections
to complete, but if some existing connection stays up then Wait will
not return.
If the Server goes ahead with establishing the connection and installs
its own Handler, that Handler needs to decide whether to serve the
error from Bind or something more opaque, and at that point Bind may
as well return a handler that makes that choice more precisely.
If the Server merely logs the error and closes the Connection, then
the Bind method itself may as well do that directly too.
It seems to me that the only winning move is not to play. Only Bind is
in a position to decide how to handle its errors appropriately, so it
should not return them to the Server.
Updates golang/go#56281.
Change-Id: I07dc43ddf31253ce23da21a92d2b6c0f8d4b3afe
Reviewed-on: https://go-review.googlesource.com/c/tools/+/443677
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Also implement and use the Shutdown method, which was mentioned in a
doc comment in CL 292169 but not actually present at that time.
With proper synchronization, we don't need heuristics to determine
whether an error is due to a connection or listener being closed.
We know whether we have called Close (and why), and we can assume
that if we have called Close then that is probably the reason for
any returned error.
Fixesgolang/go#56281.
Change-Id: I5e0ed7db0f736ca8df8cd8cf556b674e7a863a69
Reviewed-on: https://go-review.googlesource.com/c/tools/+/443675
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Prior to this CL, (*Server).run only filters out inactive connections
when it accepts a new connection. If existing connections complete,
their associated resources can't be garbage-collected until either the
next connection is accepted or the Listener is closed.
This change moves the open-connection accounting to an explicit hook
passed to newConnection, eliminating the need to call Wait entirely.
For golang/go#46047
Change-Id: I3732cb463fcea0c142f17f2b1510fdfd2dbc81da
Reviewed-on: https://go-review.googlesource.com/c/tools/+/388774
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Cottrell <iancottrell@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change fixes the semantics of Close to actually wait for
in-flight requests before closing the ReadWriteCloser. (Previously,
the Close method closed the ReadWriteCloser immediately, which I
suspect is what led to many of the failures observed in
golang/go#49387 and golang/go#46520.)
It achieves this by explicitly tracking the number of in-flight
requests, including requests with pending async responses, and
explicitly rejecting new Call requests (while keeping the read loop
open!) once Close has begun.
To make it easier for me to reason about the request lifetimes, I
reduced the number of long-lived goroutines from three to just one
(the Read loop), with an additional Handler goroutine that runs only
while the Handler queue is non-empty. Now, it is clearer (I hope!)
that the number of in-flight async requests strictly decreases
after Close has begun, even though the Read goroutine continues
to read requests (and, importantly, responses) and to forward
Notifications to the preempter.
For golang/go#49387
For golang/go#46520
Change-Id: Idf5960f848108a7ced78c5382099c8692e9b181e
Reviewed-on: https://go-review.googlesource.com/c/tools/+/388134
gopls-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
This cl updates go/internal/gcimporter the to anticiapte missing
targets and .a files the same way CL 442303 updates the two other
versions of gcimporter to do the same. It also adds a couple of
helpers to create importcfg files for the compiler and list the
locations of cached stdlib .a files in internal/goroot and
internal/testenv, the analogues of their import paths in the go
distribution.
Change-Id: Ie207882c13df0e886a51d31e7957a1e508331f10
Reviewed-on: https://go-review.googlesource.com/c/tools/+/445455
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Michael Matloob <matloob@golang.org>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
This change moves the facts package so that it can be reused by gopls.
It remains a tools-internal API.
A forthcoming reimplementation of gopls's analysis driver will
make use of the new packages and the features mentioned below.
Also:
- change parameter of read() callback from 'path string' to *Package.
- use NewDecoder().Decode() to amortize import-map computation across calls.
Change-Id: Id10cd02c0c241353524d568d5299d81457f571f8
Reviewed-on: https://go-review.googlesource.com/c/tools/+/444796
Reviewed-by: Tim King <taking@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
Some of the analyzers that are on by default in gopls produce many distracting
warnings. These are 'composites' and 'simplifycompositelit'.
This CL changes code to remove the 160 or so warnings.
An alternative would be to remove these from the default set of gopls analyzers.
Change-Id: Id5724a5aef260939dedd21a37e28f3d05bfa023d
Reviewed-on: https://go-review.googlesource.com/c/tools/+/443635
Reviewed-by: Heschi Kreinick <heschi@google.com>
Run-TryBot: Peter Weinberger <pjw@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This eliminates a race between a successful Accept call and a
concurrent Close call, which previously could have shut down the
'run' goroutine before Accept sent to the newConns channel, causing
Accept to deadlock.
In fact, it eliminates the long-running 'run' goroutine entirely
(replacing it with a time.Timer), and in the process avoids leaking
O(N) closed connections when only the very first one is long-lived.
It also eliminates a potential double-Close bug: if the run method had
called l.wrapped.Close due to an idle timeout, a subsequent call to
Close would invoke l.wrapped.Close again. The io.Closer method
explicitly documents doubled Close calls as undefined behavior, and
many Closer implementations (especially test fakes) panic or deadlock
in that case.
It also eliminates a timer leak if the Listener rapidly oscillates
between active and idle: previously the implementation used
time.After, but it now uses an explicit time.Timer which can be
stopped (and garbage-collected) when the listener becomes active.
Idleness is now tracked based on the connection's Close method rather
than Read: we have no guarantee in general that a caller will ever
actually invoke Read (if, for example, they Close the connection as
soon as it is dialed), but we can reasonably expect a caller to at
least try to ensure that Close is always called.
We now also verify, using a finalizer on a best-effort basis, that the
Close method on each connection is called. We use the finalizer to
verify the Close call — rather than to close the connection implicitly
— because closing the connection in a finalizer would delay the start
of the idle timer by an arbitrary and unbounded duration after the
last connection is actually no longer in use.
Fixesgolang/go#46047.
Fixesgolang/go#51435.
For golang/go#46520.
For golang/go#49387.
Change-Id: If173a3ed7a44aff14734b72c8340122e8d020f26
Reviewed-on: https://go-review.googlesource.com/c/tools/+/388597
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Cottrell <iancottrell@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
With the suffixes I end up a little lost in the “box” noise while I'm
reading — the channel ops alone suffice to make the storage mechanism
clear.
(To me, the mechanism of storing a value in a 1-buffered channel is
conceptually similar to storing it in a pointer, atomic.Pointer, or
similar — and we don't generally name those with a suffix either.)
For golang/go#46047.
For golang/go#46520.
For golang/go#49387.
Change-Id: I7f58a9ac532f597fe49ed70606d89bd8cbe33b55
Reviewed-on: https://go-review.googlesource.com/c/tools/+/443355
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Otherwise, the Await method on the corresponding AsyncCall will never
unblock, leading to a deadlock (detected by the test changes in
CL 388597).
For golang/go#46047
For golang/go#46520
For golang/go#49387
Change-Id: I7954f80786059772ddd7f8c98d8752d56d074919
Reviewed-on: https://go-review.googlesource.com/c/tools/+/388775
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Cottrell <iancottrell@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
(*Server).run calls Wait on all of its connections before returning,
but does not call Close explicitly. If Close is necessary to
release resources (as is often the case for a net.Conn),
the server ends up leaking resources.
For golang/go#46047
Change-Id: I4ad048c4f5e5d3f14f992eee6388acd42398c26b
Reviewed-on: https://go-review.googlesource.com/c/tools/+/388599
TryBot-Result: Gopher Robot <gobot@golang.org>
gopls-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
On darwin/cgo use readdir_r instead of syscall.ReadDirent() since the
later is simulated and slow (see: golang/go#30933).
Unlike CL 392094, this uses CGO instead of "//go:linkname" and assembly.
goos: darwin
goarch: arm64
FastWalk-10 68.9ms ±11% 20.7ms ± 2% -69.89% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
FastWalk-10 1.49MB ± 0% 1.51MB ± 0% +1.06% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
FastWalk-10 32.2k ± 0% 30.7k ± 0% -4.61% (p=0.016 n=5+4)
Fixesgolang/go#51356
Change-Id: Ia3afd06c8f14bd2036b2a1ea6e3cafbef81d3530
Reviewed-on: https://go-review.googlesource.com/c/tools/+/436780
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
Auto-Submit: Heschi Kreinick <heschi@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
The LSP protocol doesn't require edits to be sorted,
but some clients such as govim were relying on it,
as we learned when I recently removed the sort.
This change restores the sort behavior.
See also govim/govim#1171
Change-Id: I589b6cfde933ea6907859bf40acd33c1a7fcde02
Reviewed-on: https://go-review.googlesource.com/c/tools/+/442256
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
Read the entire API directory, so that we don't need to edit mkstdlib.go
when regenerating zstdlib.go for new go versions.
Joint with Dylan Le.
Co-Authored-By: dle8@u.rochester.edu.
Change-Id: If5c71bfcfd2104f923df10cf21d16b5280fed025
Reviewed-on: https://go-review.googlesource.com/c/tools/+/438504
Reviewed-by: Dylan Le <dungtuanle@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Findley <rfindley@google.com>
LineEdits has similar consistency preconditions to ApplyEdits.
Previously they were assumed, and bad input would create bad
output or crashes; now it uses the same validation logic
as ApplyEdits. Since it reports an error, computation of a
unified diff can also fail if the edits are inconsistent.
The ToUnified([]Edit) function now returns an error. For
convenience we also provide a wrapper (Unified) that cannot
fail since it calls Strings and ToUnified consistently.
LineEdits itself is now private.
Change-Id: I3780827f501d7d5c9665ec8be5656331c0dcda8e
Reviewed-on: https://go-review.googlesource.com/c/tools/+/440175
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
Spans are logically part of gopls, but could not be moved
into the gopls module because of a number of depenencies
from packagestest, analysis, and internal/diff.
Those edges are now broken.
Change-Id: Icba5ebec6b27974f832a1186120a4b87d5f87103
Reviewed-on: https://go-review.googlesource.com/c/tools/+/440176
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Previously, a diff of non-ASCII strings would incur
three complete copies of the buffers; this changes
reduces it to one.
Also:
- add diff.Bytes function, to avoid unnecessary conversions.
- remove unused diff.Lines and LineEdits functions from API.
- remove TODO to use []bytes everywhere.
We tried it in CL 439277 and didn't like it.
- Document that the diff is textual, even when given []byte.
Change-Id: I2da3257cc3d12c569218a2d7ce182452e8647a96
Reviewed-on: https://go-review.googlesource.com/c/tools/+/439835
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
gopls-CI: kokoro <noreply+kokoro@google.com>
Apply now checks that its edits are valid
(not out of bounds or overlapping),
and reports an error if not.
It also sorts them, if necessary, using (start, end)
as the key, to ensure that insertions (end=start)
are ordered before deletions at the same point
(but without changing the relative order of insertions).
Two other implementations of the diff.Apply algorithm
have been eliminated. (One of them failed to sort edits,
requiring the protocol sender to do so; that burden
is now gone.)
Change-Id: Ia76e485e6869db4a165835c3312fd14bc7d43db2
Reviewed-on: https://go-review.googlesource.com/c/tools/+/439278
Auto-Submit: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
The span package is properly part of gopls, and we'd like to move
it into that module. However, the packagestest package unfortunately
depends on span.Span as an alternative notation for Ranges.
This change decouples span.Span from packagestest.Range using a
new (unexported for now) rangeSetter interface, which Span implements.
Neither package depends on the other.
Technically this is a breaking API change:
all the Range methods have gone away, as have the Span, URI,
and Point types and their methods, which were accessible via
Range.Span(). However, clients would not be able to access
these internal types, and I think it is highly unlikely that
anyone depends on it. Nonethless this is a cautionary tale about
the risks from one innocuous-looking type alias declaration.
Change-Id: I8acb03f4acb1f798f304b03648445e37a44f9c45
Reviewed-on: https://go-review.googlesource.com/c/tools/+/439715
gopls-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
diff.TextEdit (now called simply Edit) no longer has a Span,
and no longer depends on the span package, which is really
part of gopls. Instead, it records only start/end byte
offsets within an (implied) file.
The diff algorithms have been simplified to avoid the need to
map offsets to line/column numbers (e.g. using a token.File).
All the conditions actually needed by the logic can be derived
by local string operations on the source text.
This change will allow us to move the span package into the
gopls module.
I was expecting that gopls would want to define its own
Span-augmented TextEdit type but, surprisingly, diff.Edit
is quite convenient to use throughout the entire repo:
in all places in gopls that manipulate Edits, the implied
file is obvious. In most cases, less conversion boilerplate
is required than before.
API Notes:
- diff.TextEdit -> Edit (it needn't be text)
- diff.ApplyEdits -> Apply
- source.protocolEditsFromSource is now private
Change-Id: I4d7cef078dfbd189b4aef477f845db320af6c5f6
Reviewed-on: https://go-review.googlesource.com/c/tools/+/436781
Run-TryBot: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
Computing the difference between two strings is logically an
infallible operation. This change makes the code reflect that. The
actual failures were unreachable given consistent inputs, but that was
hard to see from the complexity of the logic surrounding span.Span.
(The problem only occurs when converting offsets beyond the end of the
file to Spans, but the code preserves the integrity of offsets.)
gopls' "old" hooks.ComputeEdits impl (based on sergi/go-diff) now
reports a bug and returns a single diff for the entire file if it
panics.
Also, first steps towards simpler API and a
reusable diff package in x/tools:
- add TODO for new API. In particular, the diff package shouldn't care
about filenames, spans, and URIs. These are gopls concerns.
- diff.String is the main diff function.
- diff.Unified prints edits in unified form;
all its internals are now hidden.
- the ComputeEdits func type is moved to gopls (source.DiffFunction)
- remove all non-critical uses of myers.ComputeEdits. The only
remaining one is in gopls' defaults, but perhaps that gets
overridden by the default GoDiff setting in hooks, to BothDiffs
(sergi + pjw impls), so maybe it's now actually unused in practice?
Change-Id: I6ceb5c670897abbf285b243530a7372dfa41edf6
Reviewed-on: https://go-review.googlesource.com/c/tools/+/436778
Run-TryBot: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
gopls-CI: kokoro <noreply+kokoro@google.com>
GNU diff -u prints -0,0 for the special case of inserting
into an empty file. (I don't understand why this case is
special, but it's fair to assume it has been debugged;
perhaps patch(1) has a bug.)
This change matches that special behavior, and adds a test.
It also adds a (disabled) test for another bug I encountered.
Also, remove some overzealous uses of t.Helper() that served
to obscure the failure.
Change-Id: I3e0c389c478cde45163353130e36f2f8741a8659
Reviewed-on: https://go-review.googlesource.com/c/tools/+/436782
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>
The code that generates a name for a struct type,
and a default value for a struct field, now
supports the case where either contains a type
parameter.
Also:
- use type-checker's name for the struct type unless
it is a bare struct.
- populateValue: return *new(T) for a type parameter.
- various minor cleanups.
- various TODO comments for clarifications.
- fuzzy.FindBestMatch: use strings not identifiers.
Remove Find prefix (also FindMatchingIdentifiers).
Fixesgolang/go#54836
Change-Id: I4f6132598b4ac7e72ea1405e4a14d6a23c1eeeaa
Reviewed-on: https://go-review.googlesource.com/c/tools/+/436777
Auto-Submit: Alan Donovan <adonovan@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
An application that links in version 1.18 of (say)
go/types cannot process Go source files reported
by version 1.19 of 'go list'. There is no way to
tell go list to behave as if it was go1.18, so
for now we proceed with parsing/typechecking but,
in case of errors, we issue an additional informative
diagnostic.
Fixesgolang/go#55883Fixesgolang/go#55045
Updates golang/go#50825
Updates golang/go#52078
Change-Id: I5fd99b09742c136f4db7f71d2a75e4cdb730460d
Reviewed-on: https://go-review.googlesource.com/c/tools/+/435356
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
Reviewed-by: Zvonimir Pavlinovic <zpavlinovic@google.com>
We're almost certain that the go process shown by ps
is not the one that we're waiting for in runCmdContext,
but only by printing the pid can we be sure.
Updates golang/go#54461
Change-Id: I1ce9580de6ee6bc4557d76c2a6b05f5a3e2eb6c2
Reviewed-on: https://go-review.googlesource.com/c/tools/+/434637
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Add experimental new logic to the loopclosure analyzer that checks for
access to loop variables from parallel subtests. For now, this is gated
behind an internal variable so that we may experiment with it without
yet affecting cmd/vet.
Add an internal/loopclosure command that enables the experimental new
check.
Change-Id: Ibf7e388a58a2b3e712e66e095d06612cf3ea918c
Reviewed-on: https://go-review.googlesource.com/c/tools/+/430916
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Findley <rfindley@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Add a TODO and wait for a shorter period of time following Kill, per
post-submit advice from bcmills on CL 424075.
For golang/go#54461
Change-Id: Ia0e388c0119660844dad32629ebca4f122fded12
Reviewed-on: https://go-review.googlesource.com/c/tools/+/431075
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Findley <rfindley@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
When a go command hangs during gopls regression tests, print out
additional information about processes and file descriptors.
For golang/go#54461
Change-Id: I92aa4665e9056d15a274c154fce2783bed79718e
Reviewed-on: https://go-review.googlesource.com/c/tools/+/424075
gopls-CI: kokoro <noreply+kokoro@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL was created using the following commands:
./gopls/internal/migrate.sh
git add .
git codereview gofmt
For golang/go#54509
Change-Id: Iceeec602748a5e6f609c3ceda8d19157e5c94009
Reviewed-on: https://go-review.googlesource.com/c/tools/+/426796
gopls-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
Reviewed-by: Peter Weinberger <pjw@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Change DocumentChanges field type from []TextDocumentEdit to []DocumentChanges. DocumentChanges is a union type of TextDocumentEdit and RenameFile used for renaming package feature. It distinguishes a renaming directory ops from a text document edit ops.
For golang/go#41567
Change-Id: I25d106b34821effc53b712800f7175248e59ee25
Reviewed-on: https://go-review.googlesource.com/c/tools/+/427538
gopls-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Dylan Le <dungtuanle@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 410955 changes printing of unions which breaks two typeparams tests.
Use regexp matching and adjust the wanted results accordingly.
For golang/go#53279.
For golang/go#54822.
Change-Id: I7060df47d36ce3069570237dafff024aaad637a5
Reviewed-on: https://go-review.googlesource.com/c/tools/+/428055
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
gopls-CI: kokoro <noreply+kokoro@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Bryan C. Mills
Robert Findley
Alan Donovan
Michael Matloob
Peter Weinberger
Charlie Vieth
cui fliter
Dylan Le
Robert Griesemer