Glob Spec Regression

[tris203]

• Update 3.18.0 meta model. #2212

Changed some glob spec details, around "zero or more" vs "one or more" for *

Is this a regression of

• fix typo in glob pattern spec #2132

Or an intentional change in behaviour?

[justinmk]

cc @dbaeumer @ulugbekna

[dbaeumer]

A * should match zero or more to be consistent with standard matching theme. This was a mistake in the original text.

[tris203]

A * should match zero or more to be consistent with standard matching theme. This was a mistake in the original text.

Ok. The #2212 PR changed it back from zero to one
So this is a regression

[dbaeumer]

The meta model is actually generated from the implementation. I fixed this on the implementation side.

[tris203]

The meta model is actually generated from the implementation. I fixed this on the implementation side.

Thank you. It still seems to be outdated in the meta model.

language-server-protocol/_specifications/lsp/3.18/metaModel/metaModel.json

Line 10333 in 75332dc

"documentation": "The glob pattern to match. Glob patterns can have the following syntax:\n- `*` to match one or more characters in a path segment\n- `?` to match on one character in a path segment\n- `**` to match any number of path segments, including none\n- `{}` to group sub patterns into an OR expression. (e.g. `**​/*.{ts,js}` matches all TypeScript and JavaScript files)\n- `[]` to declare a range of characters to match in a path segment (e.g., `example.[0-9]` to match on `example.0`, `example.1`, …)\n- `[!...]` to negate a range of characters to match in a path segment (e.g., `example.[!0-9]` to match on `example.a`, `example.b`, but not `example.0`)"

Do we need a regeneration?

Thanks