Commit Graph

3 Commits

Author SHA1 Message Date
Keen Yee Liau 323be39297 fix(language-service): hybrid visitor returns parent node of BoundAttribute (#38995)
For the following example, the cursor is between `keySpan` and `valueSpan`
of the `BoundAttribute`.
```html
<test-cmp [foo]¦="bar"></test-cmp>
```
Our hybrid visitor will return `Element`in this case, which is the parent
node of the `BoundAttribute`.
This is because we only look at the `keySpan` and `valueSpan`, and not
the source span. The last element in the AST path is `Element`, so it gets
returned.

In this PR, I propose fixing this by adding a sentinel value `undefined`
to the AST path to signal that we've found a source span but the cursor is
neither in the key span nor the value span.

PR Close #38995
2020-09-25 14:35:08 -04:00
Andrew Scott 4c8766573d refactor(language-service): Update hybrid visitor to use keySpan for bound attributes (#38955)
The keySpan in bound attributes provides more fine-grained location information and can be used
to disambiguate multiple bound attributes in a single microsyntax binding. Previously,
this case could not distinguish between the two different attributes because
the sourceSpans were identical and valueSpans would not match if the cursor
was located in a key.

PR Close #38955
2020-09-24 11:35:12 -04:00
Keen Yee Liau b48cc6ead5 feat(language-service): introduce hybrid visitor to locate AST node (#38540)
This commit introduces two visitors, one for Template AST and the other
for Expression AST to allow us to easily find the node that most closely
corresponds to a given cursor position.

This is crucial because many language service APIs take in a `position`
parameter, and the information returned depends on how well we can find
a good candidate node.

In View Engine implementation of language service, the search for the node
and the processing of information to return the result are strongly coupled.
This makes the code hard to understand and hard to debug because the stack
trace is often littered with layers of visitor calls.

With this new feature, we could test the "searching" part separately and
colocate all the logic (aka hacks) that's required to retrieve an accurate
span for a given node.

Right now, only the most "narrow" node is returned by the main exported
function `findNodeAtPosition`. If needed, we could expose the entire AST
path, or expose other methods to provide more context for a node.

Note that due to limitations in the template AST interface, there are
a few known cases where microsyntax spans are not recorded properly.
This will be dealt with in a follow-up PR.

PR Close #38540
2020-08-24 09:24:18 -07:00