@lokascript/semantic 1.3.0 → 2.0.0

package/dist/languages/tr.js

@@ -8,7 +8,7 @@ function registerLanguage(code, tokenizer, profile) {
   patternCache.delete(code);
 }
 
-// src/tokenizers/base.ts
+// src/tokenizers/token-utils.ts
 var TokenStreamImpl = class {
   constructor(tokens, language) {
     this.pos = 0;
@@ -74,17 +74,14 @@ function createToken(value, kind, position, normalizedOrOptions) {
   }
   if (normalizedOrOptions) {
     const { normalized: normalized2, stem, stemConfidence } = normalizedOrOptions;
-    const token = { value, kind, position };
-    if (normalized2 !== void 0) {
-      token.normalized = normalized2;
-    }
-    if (stem !== void 0) {
-      token.stem = stem;
-      if (stemConfidence !== void 0) {
-        token.stemConfidence = stemConfidence;
-      }
-    }
-    return token;
+    return {
+      value,
+      kind,
+      position,
+      ...normalized2 !== void 0 && { normalized: normalized2 },
+      ...stem !== void 0 && { stem },
+      ...stemConfidence !== void 0 && { stemConfidence }
+    };
   }
   return { value, kind, position };
 }
@@ -106,11 +103,15 @@ function isAsciiLetter(char) {
 function isAsciiIdentifierChar(char) {
   return /[a-zA-Z0-9_-]/.test(char);
 }
+
+// src/tokenizers/char-classifiers.ts
 function createLatinCharClassifiers(letterPattern) {
   const isLetter = (char) => letterPattern.test(char);
   const isIdentifierChar = (char) => isLetter(char) || /[0-9_-]/.test(char);
   return { isLetter, isIdentifierChar };
 }
+
+// src/tokenizers/extractors.ts
 function extractCssSelector(input, startPos) {
   if (startPos >= input.length) return null;
   const char = input[startPos];
@@ -368,6 +369,8 @@ function extractNumber(input, startPos) {
   }
   return number;
 }
+
+// src/tokenizers/base-tokenizer.ts
 var _BaseTokenizer = class _BaseTokenizer {
   constructor() {
     /** Keywords derived from profile, sorted longest-first for greedy matching */
@@ -375,6 +378,34 @@ var _BaseTokenizer = class _BaseTokenizer {
     /** Map for O(1) keyword lookups by lowercase native word */
     this.profileKeywordMap = /* @__PURE__ */ new Map();
   }
+  /**
+   * Check if current position is a property access (obj.prop) vs CSS selector (.active).
+   * Property access: no whitespace before '.', previous token is identifier/keyword/selector.
+   * Also detects standalone method calls: .identifier( pattern.
+   *
+   * Returns true if '.' was emitted as an operator token and pos should advance by 1.
+   * Returns false if this is a CSS selector and should be handled by trySelector().
+   */
+  tryPropertyAccess(input, pos, tokens) {
+    if (input[pos] !== ".") return false;
+    const lastToken = tokens[tokens.length - 1];
+    const hasWhitespaceBefore = lastToken && lastToken.position.end < pos;
+    const isPropertyAccess = lastToken && !hasWhitespaceBefore && (lastToken.kind === "identifier" || lastToken.kind === "keyword" || lastToken.kind === "selector");
+    if (isPropertyAccess) {
+      tokens.push(createToken(".", "operator", createPosition(pos, pos + 1)));
+      return true;
+    }
+    const methodStart = pos + 1;
+    let methodEnd = methodStart;
+    while (methodEnd < input.length && isAsciiIdentifierChar(input[methodEnd])) {
+      methodEnd++;
+    }
+    if (methodEnd < input.length && input[methodEnd] === "(") {
+      tokens.push(createToken(".", "operator", createPosition(pos, pos + 1)));
+      return true;
+    }
+    return false;
+  }
   /**
    * Initialize keyword mappings from a language profile.
    * Builds a list of native→english mappings from:
@@ -410,6 +441,11 @@ var _BaseTokenizer = class _BaseTokenizer {
       for (const [normalized2, native] of Object.entries(profile.references)) {
         keywordMap.set(native, { native, normalized: normalized2 });
       }
+      for (const canonical of Object.keys(profile.references)) {
+        if (!keywordMap.has(canonical)) {
+          keywordMap.set(canonical, { native: canonical, normalized: canonical });
+        }
+      }
     }
     if (profile.roleMarkers) {
       for (const [role, marker] of Object.entries(profile.roleMarkers)) {
@@ -1234,8 +1270,14 @@ var turkishProfile = {
     default: { primary: "varsay\u0131lan", normalized: "default" },
     init: { primary: "ba\u015Flat", normalized: "init" },
     behavior: { primary: "davran\u0131\u015F", normalized: "behavior" },
-    install: { primary: "y\xFCkle", normalized: "install" },
+    install: { primary: "y\xFCkle", alternatives: ["kur", "y\xFCklemek"], normalized: "install" },
     measure: { primary: "\xF6l\xE7", normalized: "measure" },
+    beep: { primary: "bip", normalized: "beep" },
+    break: { primary: "dur", normalized: "break" },
+    copy: { primary: "kopyala", normalized: "copy" },
+    exit: { primary: "\xE7\u0131k", normalized: "exit" },
+    pick: { primary: "se\xE7", normalized: "pick" },
+    render: { primary: "render", normalized: "render" },
     // Modifiers
     into: { primary: "i\xE7ine", normalized: "into" },
     before: { primary: "\xF6nce", normalized: "before" },
@@ -1267,8 +1309,7 @@ var POSTPOSITIONS = /* @__PURE__ */ new Set([
   // until, as far as
   "gibi",
   // like
-  "sonra",
-  // after
+  // 'sonra' removed — classified as keyword('then') for hyperscript clause chaining
   "\xF6nce",
   // before
   "\xFCzerinde",
@@ -1374,6 +1415,11 @@ var TURKISH_EXTRAS = [
   { native: "milisaniye", normalized: "ms" },
   { native: "dakika", normalized: "m" },
   { native: "saat", normalized: "h" },
+  // Then/conjunction (clause chaining)
+  { native: "sonra", normalized: "then" },
+  { native: "ard\u0131ndan", normalized: "then" },
+  { native: "ardindan", normalized: "then" },
+  { native: "daha sonra", normalized: "then" },
   // Logical
   { native: "ve", normalized: "and" },
   { native: "veya", normalized: "or" },
@@ -1415,6 +1461,10 @@ var TurkishTokenizer = class extends BaseTokenizer {
           pos = modifierToken.position.end;
           continue;
         }
+        if (this.tryPropertyAccess(input, pos, tokens)) {
+          pos++;
+          continue;
+        }
         const selectorToken = this.trySelector(input, pos);
         if (selectorToken) {
           tokens.push(selectorToken);
@@ -1475,7 +1525,8 @@ var TurkishTokenizer = class extends BaseTokenizer {
     if (POSTPOSITIONS.has(lower)) return "particle";
     if (CASE_SUFFIXES.has(lower)) return "particle";
     if (this.isKeyword(lower)) return "keyword";
-    if (token.startsWith("#") || token.startsWith(".") || token.startsWith("[")) return "selector";
+    if (token.startsWith("#") || token.startsWith(".") || token.startsWith("[") || token.startsWith("*") || token.startsWith("<"))
+      return "selector";
     if (token.startsWith('"') || token.startsWith("'")) return "literal";
     if (/^\d/.test(token)) return "literal";
     return "identifier";