@bearcove/codemirror-lang-styx 0.1.0

package/dist/index.js

@@ -0,0 +1,263 @@
+// src/syntax.grammar.ts
+import { LRParser } from "@lezer/lr";
+
+// src/heredoc.ts
+import { ExternalTokenizer } from "@lezer/lr";
+
+// src/syntax.grammar.terms.ts
+var heredoc = 39;
+
+// src/heredoc.ts
+function isDelimiterStart(ch) {
+  return ch >= 65 && ch <= 90;
+}
+function isDelimiterChar(ch) {
+  return ch >= 65 && ch <= 90 || ch >= 48 && ch <= 57 || ch === 95;
+}
+function isLangHintChar(ch) {
+  return ch >= 97 && ch <= 122 || ch >= 48 && ch <= 57 || ch === 95 || ch === 46 || ch === 45;
+}
+var heredocTokenizer = new ExternalTokenizer(
+  (input, stack) => {
+    if (input.next !== 60) return;
+    input.advance();
+    if (input.next !== 60) return;
+    input.advance();
+    if (!isDelimiterStart(input.next)) return;
+    let delimiter = "";
+    while (isDelimiterChar(input.next)) {
+      delimiter += String.fromCharCode(input.next);
+      input.advance();
+    }
+    if (input.next === 44) {
+      input.advance();
+      if (input.next >= 97 && input.next <= 122) {
+        while (isLangHintChar(input.next)) {
+          input.advance();
+        }
+      }
+    }
+    if (input.next !== 10 && input.next !== 13) return;
+    if (input.next === 13) input.advance();
+    if (input.next === 10) input.advance();
+    while (input.next !== -1) {
+      while (input.next === 32 || input.next === 9) {
+        input.advance();
+      }
+      let matchPos = 0;
+      let isMatch = true;
+      const lineStart = input.pos;
+      for (let i = 0; i < delimiter.length && input.next !== -1; i++) {
+        if (input.next !== delimiter.charCodeAt(i)) {
+          isMatch = false;
+          break;
+        }
+        input.advance();
+        matchPos++;
+      }
+      if (isMatch && matchPos === delimiter.length) {
+        if (input.next === 10 || input.next === 13 || input.next === -1) {
+          input.acceptToken(heredoc);
+          return;
+        }
+      }
+      while (input.next !== 10 && input.next !== 13 && input.next !== -1) {
+        input.advance();
+      }
+      if (input.next === 13) input.advance();
+      if (input.next === 10) input.advance();
+    }
+    input.acceptToken(heredoc);
+  },
+  { contextual: false }
+);
+
+// src/highlight.ts
+import { styleTags, tags as t } from "@lezer/highlight";
+var styxHighlight = styleTags({
+  // Keys (property names) - first atom in an Entry
+  "KeyAtom/BareScalar": t.propertyName,
+  "KeyAtom/QuotedScalar": t.propertyName,
+  "KeyPayload/BareScalar": t.propertyName,
+  "KeyPayload/QuotedScalar": t.propertyName,
+  // Values - second atom in an Entry
+  "ValueAtom/BareScalar": t.string,
+  "ValueAtom/QuotedScalar": t.string,
+  "ValuePayload/BareScalar": t.string,
+  "ValuePayload/QuotedScalar": t.string,
+  // Sequence items
+  "SeqAtom/BareScalar": t.string,
+  "SeqAtom/QuotedScalar": t.string,
+  "SeqPayload/BareScalar": t.string,
+  "SeqPayload/QuotedScalar": t.string,
+  // Tags (@foo)
+  Tag: t.tagName,
+  // Raw strings and heredocs
+  RawScalar: t.special(t.string),
+  Heredoc: t.special(t.string),
+  // Inline attributes (key>value pairs)
+  Attributes: t.special(t.variableName),
+  Unit: t.null,
+  Comment: t.lineComment,
+  DocComment: t.docComment,
+  "( )": t.paren,
+  "{ }": t.brace,
+  ",": t.separator
+});
+
+// src/syntax.grammar.ts
+var parser = LRParser.deserialize({
+  version: 14,
+  states: "*tQVQPOOOOQO'#C^'#C^OzQPO'#C_OOQQ'#Cb'#CbOOQQ'#Cd'#CdOOQQ'#Ce'#CeO#aQPO'#CnOOQQ'#Cs'#CsOOQQ'#Ct'#CtOOQQ'#Cv'#CvO$VQPO'#ChOOQQ'#Cw'#CwO$rQRO'#CaOOQQ'#Ca'#CaO%lQRO'#C`O&SQPO'#DWOOQO'#DW'#DWOOQO'#C|'#C|QVQPOOOOQO'#C}'#C}OOQO-E6{-E6{OOQO'#Cp'#CpOOQO'#DP'#DPO!lQPO'#CoOOQO'#Cr'#CrO&XQPO'#CrO&cQPO'#CoOOQQ,59Y,59YO&nQPO,59YOOQO'#DO'#DOO#hQPO'#CiOOQO'#Cu'#CuO&sQPO'#CjOOQO'#Cj'#CjO'gQPO'#CiOOQQ,59S,59SO'nQPO,59SOOQQ'#Cc'#CcOOQQ,58{,58{OOQO'#C{'#C{OOQO'#Cz'#CzO'sQPO'#CxOOQO'#Cx'#CxOOQO,58z,58zOOQO,59r,59rOOQO-E6z-E6zOOQO-E6}-E6}O(QQPO,59ZOOQO,59^,59^O(]QPO'#DPO(QQPO,59ZO(QQPO,59ZOOQQ1G.t1G.tOOQO-E6|-E6|O(sQPO,59TOOQO'#Ck'#CkOOQO,59U,59UO(sQPO,59TOOQO'#DR'#DRO(sQPO,59TOOQQ1G.n1G.nOOQO'#Cy'#CyOOQO,59d,59dO(zQPO1G.uO(zQPO1G.uOOQO,59l,59lOOQO-E7O-E7OO)VQPO1G.oO)VQPO1G.oOOQO,59m,59mOOQO-E7P-E7PO)^QPO7+$aP)iQPO'#DQO)vQPO7+$ZP#hQPO'#DO",
+  stateData: "*W~OyOS~OZYOaUO{PO|`O}cO!ORO!PSO!QTO!RVO!SWO!TXO~O}cO|RXZRXaRX!ORX!PRX!QRX!RRX!SRX!TRX~OZYOaUOeeO|eO}cO!ORO!PSO!QTO!RVO!SWO!TXO~O`kO~P!lOZYOaUO|mO!ORO!PSO!QTO!RVO!SWO!TXO~OYsO~P#hOZYOaUO!PSO!QTO!RVO!SWO~OZTXaTXwTX|TX!OTX!PTX!QTX!RTX!STX!TTX`TXeTX~P$^OwwO!ORO!TXO|SX`SXeSX~P$^O||O~O!ORO!TXO~P$^OeeO|eO`cX~O`!UO~OY^XZ^Xa^X|^X!O^X!P^X!Q^X!R^X!S^X!T^X~P$^OY]X~P#hOY!^O~O|lX`lXelX~P$^OeeO|eO`ca~O}cO!ORO!TXO`sXesX|sX~P$^OY]a~P#hOeeO|eO`ci~OY]i~P#hOeeO|eO`cq~O}cO!ORO!TXO~P$^OY]q~P#hO!O!R}{!Q!S!T{~",
+  goto: "(Q{PP|!Q![!i!r#[#_#_PP#_$Q$T$gPP#_$j$mP$z#_#_%U%b%z&T&W&Z%w&^&d&o'T'h'rPPPP'|T_ObS_ObXiUg!R!jS_ObWhUg!R!jR!Qi_^OUbgi!R!j^[OUbgi!R!jdpYnr!W!Z!]!e!f!k!lRy^Rv[^ZOUbgi!R!jdoYnr!W!Z!]!e!f!k!lQu[Qx^Q!XpR!_yRtYQrYQ!WnW![r!W!]!fX!g!Z!e!k!lR!YpRlUYfUg!S!a!iX!Rj!P!T!bQjUQ!PgT!c!R!jeqYnr!W!Z!]!e!f!k!l^ZOUbgi!R!jdoYnr!W!Z!]!e!f!k!lRx^_]OUbgi!R!jR{^R!`yRz^QbOR}b[QOUbg!R!jRdQQnYY!Vn!Z!e!k!lQ!ZrS!e!W!]R!k!fQgUW!Og!S!a!iQ!SjS!a!P!TR!i!bQ!TjQ!b!PT!d!T!bQ!]rQ!f!WT!h!]!fTaOb",
+  nodeNames: "\u26A0 Document Comment DocComment Entry KeyExpr Tag KeyPayload QuotedScalar RawScalar ) ( Sequence SeqContent SeqItem SeqPayload } { Object ObjContent ObjSep , ObjItem Unit Attributes SeqAtom BareScalar KeyAtom ValueExpr ValuePayload ValueAtom Heredoc",
+  maxTerm: 51,
+  nodeProps: [
+    ["openedBy", 10, "(", 16, "{"],
+    ["closedBy", 11, ")", 17, "}"]
+  ],
+  propSources: [styxHighlight],
+  skippedNodes: [0],
+  repeatNodeCount: 6,
+  tokenData: "B[~RmOX!|XY(UYZ(aZ]!|]^(f^p!|pq(Uqr!|rs(lsx!|xy)syz)xz|!||})}}!P!|!P!Q*S!Q!^!|!^!_$Q!a!b!|!b!c?b!c#f!|#f#g@U#g#o!|#o#pBQ#p#q!|#q#rBV#r;'S!|;'S;=`(O<%lO!|~#R_!T~OX!|Z]!|^p!|qr!|sx!|z|!|}!^!|!^!_$Q!_!`!|!`!a$|!a#o!|#p#q!|#r;'S!|;'S;=`(O<%lO!|~$T]OX$QZ]$Q^p$Qqr$Qsx$Qz|$Q}!`$Q!`!a$|!a#o$Q#p#q$Q#r;'S$Q;'S;=`'r<%lO$Q~%PZOX%rZ]%r^p%rqr%rsx%rz|%r}#o%r#p#q%r#r;'S%r;'S;=`'x<%lO%r~%w]!S~OX%rXY&pZ]%r^p%rpq&pqr%rsx%rz|%r}#o%r#p#q%r#r;'S%r;'S;=`'x<%lO%r~&s_OX$QXY&pZ]$Q^p$Qpq&pqr$Qsx$Qz|$Q}!_$Q!a!b$Q!c#o$Q#p#q$Q#r;'S$Q;'S;=`'r<%lO$Q~'uP;=`<%l$Q~'{P;=`<%l%r~(RP;=`<%l!|~(ZQy~XY(Upq(U~(fO|~~(iPYZ(a~(oXOY(lZ](l^r(lrs)[s#O(l#O#P)a#P;'S(l;'S;=`)m<%lO(l~)aO!P~~)dRO;'S(l;'S;=`)m<%lO(l~)pP;=`<%l(l~)xOZ~~)}OY~~*SOe~~*Xa!T~OX!|Z]!|^p!|qr!|sx!|z|!|}!P!|!P!Q+^!Q!^!|!^!_$Q!_!`!|!`!a$|!a#o!|#p#q!|#r;'S!|;'S;=`(O<%lO!|~+ch!T~OX,}XY.pZ],}^p,}pq.pqr,}rs.psx,}xz.pz|,}|}.p}!P,}!P!Q5u!Q!^,}!^!_/_!_!`,}!`!a0x!a#o,}#o#p.p#p#q,}#q#r.p#r;'S,};'S;=`5o<%lO,}~-Uh{~!T~OX,}XY.pZ],}^p,}pq.pqr,}rs.psx,}xz.pz|,}|}.p}!^,}!^!_/_!_!`,}!`!a0x!a#Q,}#Q#R!|#R#o,}#o#p.p#p#q,}#q#r.p#r;'S,};'S;=`5o<%lO,}~.uU{~OY.pZ].p^#Q.p#R;'S.p;'S;=`/X<%lO.p~/[P;=`<%l.p~/df{~OX/_XY.pZ]/_^p/_pq.pqr/_rs.psx/_xz.pz|/_|}.p}!`/_!`!a0x!a#Q/_#Q#R$Q#R#o/_#o#p.p#p#q/_#q#r.p#r;'S/_;'S;=`5c<%lO/_~0}d{~OX2]XY.pZ]2]^p2]pq.pqr2]rs.psx2]xz.pz|2]|}.p}#Q2]#Q#R%r#R#o2]#o#p.p#p#q2]#q#r.p#r;'S2];'S;=`5i<%lO2]~2dd{~!S~OX2]XY3rZ]2]^p2]pq3rqr2]rs.psx2]xz.pz|2]|}.p}#Q2]#Q#R%r#R#o2]#o#p.p#p#q2]#q#r.p#r;'S2];'S;=`5i<%lO2]~3wh{~OX/_XY3rZ]/_^p/_pq3rqr/_rs.psx/_xz.pz|/_|}.p}!_/_!_!a.p!a!b/_!b!c.p!c#Q/_#Q#R$Q#R#o/_#o#p.p#p#q/_#q#r.p#r;'S/_;'S;=`5c<%lO/_~5fP;=`<%l/_~5lP;=`<%l2]~5rP;=`<%l,}~5zj!T~OX5uXY7lYZ8XZ]5u]^8^^p5upq7lqr5urs7lsx5uxz7lz|5u|}7l}!^5u!^!_8j!_!`5u!`!a:X!a#Q5u#Q#R!|#R#o5u#o#p7l#p#q5u#q#r7l#r;'S5u;'S;=`?[<%lO5u~7oWOY7lYZ8XZ]7l]^8^^#Q7l#R;'S7l;'S;=`8d<%lO7l~8^O}~~8aPYZ8X~8gP;=`<%l7l~8mhOX8jXY7lYZ8XZ]8j]^8^^p8jpq7lqr8jrs7lsx8jxz7lz|8j|}7l}!`8j!`!a:X!a#Q8j#Q#R$Q#R#o8j#o#p7l#p#q8j#q#r7l#r;'S8j;'S;=`?O<%lO8j~:[fOX;pXY7lYZ8XZ];p]^8^^p;ppq7lqr;prs7lsx;pxz7lz|;p|}7l}#Q;p#Q#R%r#R#o;p#o#p7l#p#q;p#q#r7l#r;'S;p;'S;=`?U<%lO;p~;uf!S~OX;pXY=ZYZ8XZ];p]^8^^p;ppq=Zqr;prs7lsx;pxz7lz|;p|}7l}#Q;p#Q#R%r#R#o;p#o#p7l#p#q;p#q#r7l#r;'S;p;'S;=`?U<%lO;p~=^jOX8jXY=ZYZ8XZ]8j]^8^^p8jpq=Zqr8jrs7lsx8jxz7lz|8j|}7l}!_8j!_!a7l!a!b8j!b!c7l!c#Q8j#Q#R$Q#R#o8j#o#p7l#p#q8j#q#r7l#r;'S8j;'S;=`?O<%lO8j~?RP;=`<%l8j~?XP;=`<%l;p~?_P;=`<%l5u~?gR!R~!c!}?p#R#S?p#T#o?p~?uT!O~}!O?p!Q![?p!c!}?p#R#S?p#T#o?p~@Za!T~OX!|Z]!|^p!|qr!|rsA`st@Utx!|z|!|}!^!|!^!_$Q!_!`!|!`!a$|!a#o!|#p#q!|#r;'S!|;'S;=`(O<%lO!|~AcTOrA`rsArs;'SA`;'S;=`Az<%lOA`~AwP!Q~stAr~A}P;=`<%lA`~BVOa~~B[O`~",
+  tokenizers: [0, heredocTokenizer],
+  topRules: { "Document": [0, 1] },
+  tokenPrec: 413
+});
+
+// src/index.ts
+import {
+  LRLanguage,
+  LanguageSupport,
+  foldNodeProp,
+  foldInside,
+  indentNodeProp,
+  continuedIndent,
+  syntaxTree,
+  foldService
+} from "@codemirror/language";
+import { completeFromList } from "@codemirror/autocomplete";
+import { parseMixed } from "@lezer/common";
+function parseHeredocText(text) {
+  const match = text.match(/^<<([A-Z][A-Z0-9_]*)(?:,([a-z][a-z0-9_.-]*))?\r?\n/);
+  if (!match) return null;
+  const delimiter = match[1];
+  const langHint = match[2] || null;
+  const headerLen = match[0].length;
+  const delimPattern = new RegExp(`^[ \\t]*${delimiter}$`, "m");
+  const contentMatch = text.slice(headerLen).match(delimPattern);
+  if (!contentMatch || contentMatch.index === void 0) {
+    return {
+      delimiter,
+      langHint,
+      contentStart: headerLen,
+      contentEnd: text.length
+    };
+  }
+  return {
+    delimiter,
+    langHint,
+    contentStart: headerLen,
+    contentEnd: headerLen + contentMatch.index
+  };
+}
+function createMixedParser(nestedLanguages) {
+  const langMap = /* @__PURE__ */ new Map();
+  for (const { tag, language } of nestedLanguages) {
+    langMap.set(tag, language.language.parser);
+  }
+  return parseMixed((node, input) => {
+    if (node.type.name !== "Heredoc") return null;
+    const text = input.read(node.from, node.to);
+    const parsed = parseHeredocText(text);
+    if (!parsed || !parsed.langHint) return null;
+    const nestedParser = langMap.get(parsed.langHint);
+    if (!nestedParser) return null;
+    return {
+      parser: nestedParser,
+      overlay: [{ from: node.from + parsed.contentStart, to: node.from + parsed.contentEnd }]
+    };
+  });
+}
+var styxFoldService = foldService.of((state, lineStart, lineEnd) => {
+  const tree = syntaxTree(state);
+  let node = tree.resolveInner(lineEnd, -1);
+  for (let cur = node; cur; cur = cur.parent) {
+    if (cur.type.name === "Object" || cur.type.name === "Sequence") {
+      const first = cur.firstChild;
+      const last = cur.lastChild;
+      if (first && last && first.to < last.from && first.from >= lineStart) {
+        return { from: first.to, to: last.from };
+      }
+    }
+  }
+  return null;
+});
+var baseProps = [
+  indentNodeProp.add({
+    Object: continuedIndent({ except: /^\s*\}/ }),
+    Sequence: continuedIndent({ except: /^\s*\)/ })
+  }),
+  foldNodeProp.add({
+    Object: foldInside,
+    Sequence: foldInside
+  })
+];
+var styxLanguage = LRLanguage.define({
+  name: "styx",
+  parser: parser.configure({ props: baseProps }),
+  languageData: {
+    commentTokens: { line: "//" },
+    closeBrackets: { brackets: ["(", "{", '"'] }
+  }
+});
+function createStyxLanguage(nestedLanguages) {
+  if (nestedLanguages.length === 0) {
+    return styxLanguage;
+  }
+  const mixedParser = parser.configure({
+    props: baseProps,
+    wrap: createMixedParser(nestedLanguages)
+  });
+  return LRLanguage.define({
+    name: "styx",
+    parser: mixedParser,
+    languageData: {
+      commentTokens: { line: "//" },
+      closeBrackets: { brackets: ["(", "{", '"'] }
+    }
+  });
+}
+var builtinTags = [
+  "@string",
+  "@int",
+  "@float",
+  "@bool",
+  "@null",
+  "@object",
+  "@array",
+  "@optional",
+  "@required",
+  "@default",
+  "@enum",
+  "@pattern",
+  "@min",
+  "@max",
+  "@minLength",
+  "@maxLength"
+].map((label) => ({ label, type: "keyword" }));
+var styxCompletion = styxLanguage.data.of({
+  autocomplete: completeFromList(builtinTags)
+});
+function styx(config = {}) {
+  const nestedLanguages = config.nestedLanguages || [];
+  const lang = createStyxLanguage(nestedLanguages);
+  const nestedSupports = nestedLanguages.flatMap((n) => n.language.support);
+  return new LanguageSupport(lang, [styxCompletion, styxFoldService, ...nestedSupports]);
+}
+export {
+  parser,
+  styx,
+  styxLanguage
+};

package/package.json

@@ -0,0 +1,63 @@
+{
+  "name": "@bearcove/codemirror-lang-styx",
+  "version": "0.1.0",
+  "description": "Styx language support for CodeMirror 6",
+  "type": "module",
+  "main": "dist/index.cjs",
+  "module": "dist/index.js",
+  "types": "dist/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./dist/index.d.ts",
+      "import": "./dist/index.js",
+      "require": "./dist/index.cjs"
+    }
+  },
+  "files": [
+    "dist",
+    "src"
+  ],
+  "keywords": [
+    "codemirror",
+    "lezer",
+    "styx",
+    "syntax-highlighting",
+    "editor"
+  ],
+  "author": "Amos Wenger <amos@bearcove.net>",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/bearcove/styx.git",
+    "directory": "editors/codemirror-styx"
+  },
+  "bugs": {
+    "url": "https://github.com/bearcove/styx/issues"
+  },
+  "homepage": "https://styx.bearcove.net",
+  "devDependencies": {
+    "@lezer/generator": "^1.7.0",
+    "tsup": "^8.0.0",
+    "typescript": "^5.3.0",
+    "vitest": "^1.0.0"
+  },
+  "peerDependencies": {
+    "@codemirror/autocomplete": "^6.0.0",
+    "@codemirror/language": "^6.0.0",
+    "@codemirror/state": "^6.0.0",
+    "@codemirror/view": "^6.0.0",
+    "@lezer/common": "^1.0.0",
+    "@lezer/highlight": "^1.0.0",
+    "@lezer/lr": "^1.0.0"
+  },
+  "publishConfig": {
+    "access": "public"
+  },
+  "scripts": {
+    "build": "npm run build:grammar && npm run build:bundle",
+    "build:grammar": "lezer-generator src/styx.grammar -o src/syntax.grammar.ts",
+    "build:bundle": "tsup src/index.ts --format esm,cjs --dts --clean",
+    "dev": "npm run build:grammar && tsup src/index.ts --format esm,cjs --dts --watch",
+    "test": "vitest run"
+  }
+}

package/src/heredoc.ts

@@ -0,0 +1,116 @@
+import { ExternalTokenizer } from "@lezer/lr";
+import { heredoc as Heredoc } from "./syntax.grammar.terms";
+
+// Helper: check if char is valid delimiter start [A-Z]
+function isDelimiterStart(ch: number): boolean {
+  return ch >= 65 && ch <= 90; // A-Z
+}
+
+// Helper: check if char is valid delimiter char [A-Z0-9_]
+function isDelimiterChar(ch: number): boolean {
+  return (ch >= 65 && ch <= 90) || (ch >= 48 && ch <= 57) || ch === 95;
+}
+
+// Helper: check if char is valid lang hint char [a-z0-9_.-]
+function isLangHintChar(ch: number): boolean {
+  return (ch >= 97 && ch <= 122) || (ch >= 48 && ch <= 57) ||
+         ch === 95 || ch === 46 || ch === 45;
+}
+
+/**
+ * External tokenizer that matches an entire heredoc as a single token.
+ *
+ * Format: <<DELIM[,lang]\n...content...\nDELIM
+ *
+ * The token includes:
+ * - The opening marker (<<DELIM or <<DELIM,lang)
+ * - The newline after the marker
+ * - All content lines
+ * - The closing delimiter
+ */
+export const heredocTokenizer = new ExternalTokenizer(
+  (input, stack) => {
+    // Must start with <<
+    if (input.next !== 60 /* < */) return;
+    input.advance();
+    if (input.next !== 60 /* < */) return;
+    input.advance();
+
+    // Must have delimiter starting with [A-Z]
+    if (!isDelimiterStart(input.next)) return;
+
+    // Read delimiter name
+    let delimiter = "";
+    while (isDelimiterChar(input.next)) {
+      delimiter += String.fromCharCode(input.next);
+      input.advance();
+    }
+
+    // Optional lang hint after comma
+    if (input.next === 44 /* , */) {
+      input.advance();
+      // Consume lang hint (a-z start, then a-z0-9_.- continuation)
+      if (input.next >= 97 && input.next <= 122) {
+        while (isLangHintChar(input.next)) {
+          input.advance();
+        }
+      }
+    }
+
+    // Must be followed by newline
+    if (input.next !== 10 && input.next !== 13) return;
+
+    // Consume newline
+    if (input.next === 13) input.advance(); // \r
+    if (input.next === 10) input.advance(); // \n
+
+    // Now scan content lines until we find the delimiter at start of line
+    while (input.next !== -1) {
+      // At start of line - check for delimiter
+      // Skip optional leading whitespace (for indented heredocs)
+      while (input.next === 32 || input.next === 9) {
+        input.advance();
+      }
+
+      // Check if this line starts with the delimiter
+      let matchPos = 0;
+      let isMatch = true;
+
+      // We need to peek ahead without consuming if it's not a match
+      // Unfortunately ExternalTokenizer doesn't have peek, so we'll
+      // consume and track position
+      const lineStart = input.pos;
+
+      for (let i = 0; i < delimiter.length && input.next !== -1; i++) {
+        if (input.next !== delimiter.charCodeAt(i)) {
+          isMatch = false;
+          break;
+        }
+        input.advance();
+        matchPos++;
+      }
+
+      if (isMatch && matchPos === delimiter.length) {
+        // Check that delimiter is followed by newline or EOF
+        if (input.next === 10 || input.next === 13 || input.next === -1) {
+          // Found the end! Accept the token
+          input.acceptToken(Heredoc);
+          return;
+        }
+      }
+
+      // Not a match - consume rest of line
+      while (input.next !== 10 && input.next !== 13 && input.next !== -1) {
+        input.advance();
+      }
+
+      // Consume newline
+      if (input.next === 13) input.advance();
+      if (input.next === 10) input.advance();
+    }
+
+    // EOF without finding end delimiter - still accept as (unterminated) heredoc
+    input.acceptToken(Heredoc);
+  },
+  { contextual: false }
+);

package/src/highlight.ts

@@ -0,0 +1,37 @@
+import { styleTags, tags as t } from "@lezer/highlight";
+
+export const styxHighlight = styleTags({
+  // Keys (property names) - first atom in an Entry
+  "KeyAtom/BareScalar": t.propertyName,
+  "KeyAtom/QuotedScalar": t.propertyName,
+  "KeyPayload/BareScalar": t.propertyName,
+  "KeyPayload/QuotedScalar": t.propertyName,
+
+  // Values - second atom in an Entry
+  "ValueAtom/BareScalar": t.string,
+  "ValueAtom/QuotedScalar": t.string,
+  "ValuePayload/BareScalar": t.string,
+  "ValuePayload/QuotedScalar": t.string,
+
+  // Sequence items
+  "SeqAtom/BareScalar": t.string,
+  "SeqAtom/QuotedScalar": t.string,
+  "SeqPayload/BareScalar": t.string,
+  "SeqPayload/QuotedScalar": t.string,
+
+  // Tags (@foo)
+  Tag: t.tagName,
+
+  // Raw strings and heredocs
+  RawScalar: t.special(t.string),
+  Heredoc: t.special(t.string),
+
+  // Inline attributes (key>value pairs)
+  Attributes: t.special(t.variableName),
+  Unit: t.null,
+  Comment: t.lineComment,
+  DocComment: t.docComment,
+  "( )": t.paren,
+  "{ }": t.brace,
+  ",": t.separator,
+});