Fix inconsistent conflict resolution for try and ternary
#75
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
alex-pinkus
force-pushed
the
try-ternary-ambiguity
branch
from
19bc69b to
1288b4e
Compare
In most places in the grammar, we rely on our conflicts resolving themselves with only one parse interpretation being valid. With `try` and ternary, that isn't the case: both of `try (foo() ? bar() : baz)` and `(try foo()) ? bar() : baz` are valid. This matters because `bar()` could also be a `throws` function, and our `try` is supposed to cover it. The rules for dynamic precedence are described here: tree-sitter/tree-sitter#678 Based on those, it seems like we were getting the correct parse result in our test cases solely because of some rule definition order, which "isn't really meant to be relied on." This exposes itself in some unrelated changes for a custom string interpolation (not included in this PR) where the rule order changes and flips our interpretation to the wrong one. Instead of relying on the ordering behavior, this change tells the parser to choose `try (foo() ? ...)` over `(try foo()) ? ...` by giving it a dynamic precedence value. This makes our ordering consistent even in the face of later rule changes.
alex-pinkus
force-pushed
the
try-ternary-ambiguity
branch
from
1288b4e to
edd6bdf
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In most places in the grammar, we rely on our conflicts resolving
themselves with only one parse interpretation being valid. With
tryandternary, that isn't the case: both of
try (foo() ? bar() : baz)and(try foo()) ? bar() : bazare valid. This matters becausebar()could also be a
throwsfunction, and ourtryis supposed to coverit.
The rules for dynamic precedence are described here:
tree-sitter/tree-sitter#678
Based on those, it seems like we were getting the correct parse result
in our test cases solely because of some rule definition order, which
"isn't really meant to be relied on." This exposes itself in some
unrelated changes for a custom string interpolation (not included in
this PR) where the rule order changes and flips our interpretation to
the wrong one.
Instead of relying on the ordering behavior, this change tells the
parser to choose
try (foo() ? ...)over(try foo()) ? ...by givingit a dynamic precedence value. This makes our ordering consistent even
in the face of later rule changes.