@cavelang/tree-sitter-cave 0.27.3 → 0.29.0

package/Authors.md

@@ -0,0 +1,5 @@
+# Authors
+
+By adding their name to this file, all listed authors confirm they have copyright over their contributions to this project and are waiving these rights, placing their work in the public domain in accordance with the CC0 1.0 Universal Public Domain Dedication.
+
+- Mirek Rusin (Switzerland)

package/README.md

@@ -8,12 +8,17 @@ needs no external scanner: indentation is skipped and qualifier/continuation
 lines are recognized by their leading verb. Parent attachment (spec §8) is
 semantic and left to consumers such as `@cavelang/parser`.
 
+Entity and attribute names accept Unicode letters, combining marks and
+numbers, with `/`, `-`, `_`, and `.` as structural characters. Numeric
+values include negative scalars and negative trajectory endpoints.
+
 ## Contents
 
 - `grammar.js` — the grammar (spec §16, §3–§8)
 - `queries/highlights.scm` — highlight captures (nvim/helix vocabulary);
   the single source used by `@cavelang/highlight` (terminal ANSI) and the
-  CAVE VSCode extension (semantic tokens)
+  CAVE VSCode extension (semantic tokens), including trajectory arrows as
+  operators
 - `src/` (generated parser) and `tree-sitter-cave.wasm` are **not
   committed** — generated artifacts are unauditable, so `pnpm build`
   produces both on demand (tree-sitter-cli fetches wasi-sdk itself; no
@@ -22,6 +27,10 @@ semantic and left to consumers such as `@cavelang/parser`.
 
 ## Consuming
 
+The published entry points are `@cavelang/tree-sitter-cave/wasm` for the
+generated grammar, `@cavelang/tree-sitter-cave/highlights` for the shared
+query, and `@cavelang/tree-sitter-cave/package.json` for package metadata.
+
 ```js
 import { Language, Parser } from 'web-tree-sitter'
 

package/grammar.js

@@ -137,11 +137,14 @@ module.exports = grammar({
     // keywords (`NOT`, `WHEN`, …) win as strings over this regex.
     verb: () => /[A-Z][A-Z-]*/,
 
-    entity: () => /[A-Za-z_][A-Za-z0-9_./-]*/,
+    // Entity names may use Unicode letters, combining marks and numbers.
+    // The first code point remains a letter or `_`; scope and
+    // kebab-case separators keep their existing ASCII spellings (§4.1).
+    entity: () => /[\p{L}_][\p{L}\p{M}\p{N}_./-]*/u,
 
-    attribute: () => token(/[A-Za-z_][A-Za-z0-9_./-]*:/),
+    attribute: () => token(/[\p{L}_][\p{L}\p{M}\p{N}_./-]*:/u),
 
-    number: () => /~?[0-9][0-9A-Za-z.,%_-]*/,
+    number: () => /~?-?[0-9][0-9A-Za-z.,%_-]*/,
 
     string: () => token(seq('"', /[^"\r\n]*/, '"')),
 

package/package.json

@@ -1,7 +1,17 @@
 {
   "name": "@cavelang/tree-sitter-cave",
-  "version": "0.27.3",
+  "version": "0.29.0",
   "description": "Tree-sitter grammar for CAVE — the canonical grammar artifact behind editor and terminal highlighting.",
+  "keywords": [
+    "cave",
+    "knowledge-graph",
+    "knowledge-representation",
+    "typescript"
+  ],
+  "homepage": "https://github.com/mirek/cave#readme",
+  "bugs": {
+    "url": "https://github.com/mirek/cave/issues"
+  },
   "license": "CC0-1.0",
   "author": "Mirek Rusin (https://github.com/mirek)",
   "repository": {
@@ -24,7 +34,8 @@
     "queries",
     "tree-sitter-cave.wasm",
     "README.md",
-    "License.md"
+    "License.md",
+    "Authors.md"
   ],
   "publishConfig": {
     "access": "public"

package/queries/highlights.scm

@@ -24,6 +24,7 @@
 (comparison_op) @operator
 (importance) @operator
 (uncertainty "+/-" @operator)
+(value "->" @operator)
 
 (tag "#" @punctuation.special)
 (tag ":" @punctuation.delimiter)

package/src/grammar.json

@@ -688,18 +688,20 @@
     },
     "entity": {
       "type": "PATTERN",
-      "value": "[A-Za-z_][A-Za-z0-9_./-]*"
+      "value": "[\\p{L}_][\\p{L}\\p{M}\\p{N}_./-]*",
+      "flags": "u"
     },
     "attribute": {
       "type": "TOKEN",
       "content": {
         "type": "PATTERN",
-        "value": "[A-Za-z_][A-Za-z0-9_./-]*:"
+        "value": "[\\p{L}_][\\p{L}\\p{M}\\p{N}_./-]*:",
+        "flags": "u"
       }
     },
     "number": {
       "type": "PATTERN",
-      "value": "~?[0-9][0-9A-Za-z.,%_-]*"
+      "value": "~?-?[0-9][0-9A-Za-z.,%_-]*"
     },
     "string": {
       "type": "TOKEN",