This does not remove the computation of available declarations. This CL
seemed big enough without that, so I (or someone else) can get that in
the next CL.
Change-Id: I67ab49b75c8a415ccfaef16c4e49a00026a6091f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/341160
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Sorry for the long CL; once again I couldn't find a reasonable way to
split it up into smaller pieces.
This CL causes all of the elements from imports to be generated by the
new approach.
It does contain a new flag that is used to preserve the current
semantics when computing completions for Cider. I hope to be able to
make the new approach efficient enough that we can remove this flag,
but I can't realistically do that until more of the approach is
implemented.
I'm also not fond of the way it fails to handle conflicts in the import
scope. There's a TODO marking the places that would need to be changed,
but the current behavior is that if the same name N is imported from
two or more imports and they refer to different elements, the first one
found will be suggested and others won't. I don't think we want to
suggest any of them unless we also fix the imports so that the name
won't be conflicting. I'm not sure how we want to do that, but it needs
to work with whatever optimization approach we choose.
I think the changes to the tests are all positive, but please double
check that I haven't missed a good reason for the previous behavior.
Change-Id: I70d4514a67988f654984ac20f33b2d7a3cfc78e2
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/340282
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Reviewed-by: Keerti Parthasarathy <keertip@google.com>
Sorry for the length of the CL. I couldn't really think of a good way
to split it up further.
The first pass was previously suggesting everything declared in scopes
below the class level. This CL expands that to include everything
declared in the current library. It does not include imported
declarations. It also doesn't yet handle members suggested after a
period. Those are both being handled by constributors.
Please look over the tests carefully. There are more changes than I
would have liked, although I think they're mostly for the better (as
in, the suggestions are more accurate), but I could easily have
missed an invalid change.
Change-Id: I98b3cfc8440ee5da09a8e6828c946d65d8d77759
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/338482
Reviewed-by: Samuel Rawlins <srawlins@google.com>
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Reviewed-by: Keerti Parthasarathy <keertip@google.com>
I'm fairly sure the failing tests are recent breakages. I was only able
to break it without noticing because we didn't have adequate testing.
The next CL should fix the tests.
Change-Id: I69f5fde6e7cd2e872b2ffab38905467d366caa9f
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/336625
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Reviewed-by: Keerti Parthasarathy <keertip@google.com>
Sharing the visibility tracker this way prevents shadowed elements from
being suggested and improves performance.
Most of these changes can be reverted once the local reference
contributor has been removed.
Change-Id: Ib156ae56455e8d85a31163f9e59ce4dd93f5920b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/333124
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
I found the old style of tests difficult to read and even harder to
debug, so I decided to convert them. Fortunately, a small conversion
program did most of the work.
I didn't change the actual test code except in a couple of places where
the tests were failing because the test code produced diagnostics beyond
those expected for incomplete code (such as unreferenced variables).
As much as possible I kept the original pass/fail sense of the test.
There were a few tests that were failing because the expectations were
just wrong, so in those cases I allowed the test to pass with the right
expectations. I tried to verify the expectations of the tests, but I'm
sure I missed some.
Change-Id: I80b7f914e75012562d3423cc6eb242bb8be11cc1
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/331215
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Fixes some cases where the same keyword was being suggested twice
because we were completing at the beginning of a statement which can
also be the beginning of an expression.
Change-Id: I7c12da8caed14438f9f1a16fa7121d1cbd96b891
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/328769
Reviewed-by: Samuel Rawlins <srawlins@google.com>
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Sorry about the size of the change.
This started as a CL to move the computation of keywords with a record
pattern from the contributor to the pass. In the process of doing so I
realized that the previous computation of the completion node (the node
that's visited) was too aggressive about moving upward in the AST. I
believe the issue happens when we reach a node in a list of similar
nodes (such as a statement in a block).
I changed the computation of the completion node, which necessitated a
lot of changes to existing code. It also resulted in changing several
tests, most for the better I think, but please look carefully for
places where you think the changes aren't reasonable. I'm also
interested in feedback about the change to the computation of the
completion node.
Change-Id: I2c6ac76fe88f672a206c3692d9ff51d2e487ec0e
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/324620
Reviewed-by: Samuel Rawlins <srawlins@google.com>
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Currently the checks that we have are ironically both repetitive and
inconsistent. There are also tests that are not precise / demonstrative.
Change-Id: I1ad0de5b9fb982d8091e4f80a6e5c5ebbad0f418
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/308580
Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Samuel Rawlins <srawlins@google.com>
I'm not sure how to convert the doc tests. They're currently depending
on being able to set up some state in the server that I can't reach
from the new test framework. I'm guessing that I might be able to define
dartdoc comments that would populate the data structures in the same way,
but if you can think of an easier approach I'd love to hear about it.
Change-Id: I20afcf0865c1a45b682444f8bce8c4e3c7fdc45c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/305782
Commit-Queue: Brian Wilkerson <brianwilkerson@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Parker Lougheed