Some languages have the notion of modules, and to represent those
we've started to use a `@module` tag, as discussed in
https://github.com/elixir-lang/tree-sitter-elixir/issues/15.
Because historically we've used the constructor highlight color for
modules in JS/Ruby, it's defined to map to the same color.
* Fix bugs related to named wildcard patterns vs regular wildcard patterns.
* Fix handling of extra nodes during query analysis. Previously, the
expected child_index was updated incorrectly after an extra node,
leading to false "impossible pattern" errors.
* Refine logic for avoiding unnecessary state-splitting due to fallible steps.
Compute *two* different analysis results related to step fallibility:
* `root_pattern_guaranteed` which, like before, summarizes whether the
entire pattern is guaranteed to match once this step is reached.
* `parent_pattern_guaranteed` - which just indicates whether the
immediate parent pattern is guaranteed. This is now used when
deciding whether it's necessary to split a match state.
When descending into a hidden child rule, the current stack entry
can be reused if it is currently at the end of its rule.
This fixes a test failure when analyzing a Ruby query. The
failure was introduced due to some changes to the Ruby grammar.
This optimization allows us to impose a _smaller_ limit on
the stack size, which should make query analysis faster and
more memory-efficient.
It seems that (some implementations of?) `snprintf` returns -1 and sets `errno` to `EINVAL` if a `maxlen` of zero is passed. This causes the count to underflow and `ts_subtree__write_to_string` returns a gigantic size which the succeeding malloc will refuse to allocate.
These tests are easier to write and maintain if the grammars are just JS,
like grammars normally are. It doesn't slow the tests down significantly
to shell out to `node` for each of these grammars.
