@lokascript/semantic 1.0.0 → 1.1.0

package/dist/languages/ko.d.ts

@@ -1,4 +1,4 @@
-import { B as BaseTokenizer, T as TokenStream, a as TokenKind, L as LanguageProfile } from '../types-C4dcj53L.js';
+import { B as BaseTokenizer, T as TokenStream, a as TokenKind, L as LanguageProfile } from '../types-BY3Id07j.js';
 
 /**
  * Korean Tokenizer
@@ -33,6 +33,13 @@ declare class KoreanTokenizer extends BaseTokenizer {
      * Korean time units attach directly without whitespace.
      */
     private extractKoreanNumber;
+    /**
+     * Try to split a temporal event suffix from a word token.
+     * This handles compact forms like 클릭할때 → 클릭 + 할때
+     *
+     * @returns Split tokens if a suffix is found, null otherwise
+     */
+    private trySplitTemporalSuffix;
 }
 /**
  * Singleton instance.

package/dist/languages/ko.js

@@ -1162,7 +1162,7 @@ var koreanProfile = {
   },
   roleMarkers: {
     patient: { primary: "\uC744", alternatives: ["\uB97C"], position: "after" },
-    destination: { primary: "\uC5D0", alternatives: ["\uC73C\uB85C", "\uB85C", "\uC5D0\uC11C"], position: "after" },
+    destination: { primary: "\uC5D0", alternatives: ["\uC73C\uB85C", "\uB85C", "\uC5D0\uC11C", "\uC758"], position: "after" },
     source: { primary: "\uC5D0\uC11C", alternatives: ["\uBD80\uD130"], position: "after" },
     style: { primary: "\uB85C", alternatives: ["\uC73C\uB85C"], position: "after" },
     event: { primary: "\uC744", alternatives: ["\uB97C"], position: "after" }
@@ -1198,6 +1198,12 @@ var koreanProfile = {
     // DOM focus
     focus: { primary: "\uD3EC\uCEE4\uC2A4", normalized: "focus" },
     blur: { primary: "\uBE14\uB7EC", normalized: "blur" },
+    // Common event names (for event handler patterns)
+    click: { primary: "\uD074\uB9AD", normalized: "click" },
+    hover: { primary: "\uD638\uBC84", normalized: "hover" },
+    submit: { primary: "\uC81C\uCD9C", normalized: "submit" },
+    input: { primary: "\uC785\uB825", normalized: "input" },
+    change: { primary: "\uBCC0\uACBD", normalized: "change" },
     // Navigation
     go: { primary: "\uC774\uB3D9", normalized: "go" },
     // Async
@@ -1241,6 +1247,15 @@ var koreanProfile = {
   tokenization: {
     particles: ["\uC744", "\uB97C", "\uC774", "\uAC00", "\uC740", "\uB294", "\uC5D0", "\uC5D0\uC11C", "\uC73C\uB85C", "\uB85C", "\uC640", "\uACFC", "\uB3C4"],
     boundaryStrategy: "space"
+  },
+  eventHandler: {
+    // Event marker: 할 때 (when/at the time of), used in SOV pattern
+    // Pattern: [event] 할 때 [destination] 의 [patient] 를 [action]
+    // Example: 클릭 할 때 #button 의 .active 를 토글
+    // Compact forms (no space): 클릭할때 .active를토글
+    eventMarker: { primary: "\uD560 \uB54C", alternatives: ["\uD560\uB54C", "\uB54C", "\uC5D0"], position: "after" },
+    temporalMarkers: ["\uD560 \uB54C", "\uD560\uB54C", "\uB54C"]
+    // temporal markers (with and without space)
   }
 };
 
@@ -1314,6 +1329,7 @@ var SINGLE_CHAR_PARTICLES = /* @__PURE__ */ new Set([
   "\uB9CC"
 ]);
 var MULTI_CHAR_PARTICLES = ["\uC5D0\uC11C", "\uC73C\uB85C", "\uBD80\uD130", "\uAE4C\uC9C0", "\uCC98\uB7FC", "\uBCF4\uB2E4"];
+var TEMPORAL_EVENT_SUFFIXES = ["\uD560\uB54C", "\uD558\uBA74", "\uD558\uB2C8\uAE4C", "\uD560 \uB54C"];
 var PARTICLE_ROLES = /* @__PURE__ */ new Map([
   // Subject markers (vowel harmony pair)
   [
@@ -1442,59 +1458,23 @@ var KOREAN_EXTRAS = [
   { native: "\uD0A4\uC5C5", normalized: "keyup" },
   { native: "\uB9C8\uC6B0\uC2A4\uC624\uBC84", normalized: "mouseover" },
   { native: "\uB9C8\uC6B0\uC2A4\uC544\uC6C3", normalized: "mouseout" },
-  // References (additional forms)
+  // References (additional forms not in profile)
   { native: "\uB0B4", normalized: "my" },
   { native: "\uADF8\uAC83\uC758", normalized: "its" },
-  // Conditional event forms (native idioms)
-  { native: "\uD558\uBA74", normalized: "on" },
-  { native: "\uC73C\uBA74", normalized: "on" },
-  { native: "\uBA74", normalized: "on" },
-  { native: "\uD560\uB54C", normalized: "on" },
-  { native: "\uD560 \uB54C", normalized: "on" },
-  { native: "\uC744\uB54C", normalized: "on" },
-  { native: "\uC744 \uB54C", normalized: "on" },
-  { native: "\uD558\uB2C8\uAE4C", normalized: "on" },
-  { native: "\uB2C8\uAE4C", normalized: "on" },
-  // Control flow helpers
-  { native: "\uADF8\uB7EC\uBA74", normalized: "then" },
-  { native: "\uADF8\uB807\uC9C0\uC54A\uC73C\uBA74", normalized: "otherwise" },
-  { native: "\uC911\uB2E8", normalized: "break" },
   // Logical
   { native: "\uADF8\uB9AC\uACE0", normalized: "and" },
   { native: "\uB610\uB294", normalized: "or" },
   { native: "\uC544\uB2C8", normalized: "not" },
   { native: "\uC774\uB2E4", normalized: "is" },
-  // Command overrides (ensure correct mapping when profile has multiple meanings)
-  { native: "\uCD94\uAC00", normalized: "add" },
-  // Profile may have this as 'append'
-  // Attached particle forms (native idioms - particle + verb without space)
-  // Object particle 를 (after vowel)
-  { native: "\uB97C\uD1A0\uAE00", normalized: "toggle" },
-  { native: "\uB97C\uC804\uD658", normalized: "toggle" },
-  { native: "\uB97C\uCD94\uAC00", normalized: "add" },
-  { native: "\uB97C\uC81C\uAC70", normalized: "remove" },
-  { native: "\uB97C\uC0AD\uC81C", normalized: "remove" },
-  { native: "\uB97C\uC99D\uAC00", normalized: "increment" },
-  { native: "\uB97C\uAC10\uC18C", normalized: "decrement" },
-  { native: "\uB97C\uD45C\uC2DC", normalized: "show" },
-  { native: "\uB97C\uC228\uAE30\uB2E4", normalized: "hide" },
-  { native: "\uB97C\uC124\uC815", normalized: "set" },
-  // Object particle 을 (after consonant)
-  { native: "\uC744\uD1A0\uAE00", normalized: "toggle" },
-  { native: "\uC744\uC804\uD658", normalized: "toggle" },
-  { native: "\uC744\uCD94\uAC00", normalized: "add" },
-  { native: "\uC744\uC81C\uAC70", normalized: "remove" },
-  { native: "\uC744\uC0AD\uC81C", normalized: "remove" },
-  { native: "\uC744\uC99D\uAC00", normalized: "increment" },
-  { native: "\uC744\uAC10\uC18C", normalized: "decrement" },
-  { native: "\uC744\uD45C\uC2DC", normalized: "show" },
-  { native: "\uC744\uC228\uAE30\uB2E4", normalized: "hide" },
-  { native: "\uC744\uC124\uC815", normalized: "set" },
   // Time units
   { native: "\uCD08", normalized: "s" },
   { native: "\uBC00\uB9AC\uCD08", normalized: "ms" },
   { native: "\uBD84", normalized: "m" },
   { native: "\uC2DC\uAC04", normalized: "h" }
+  // Note: Attached particle+verb forms (를토글, 을토글, etc.) are intentionally
+  // NOT included because they cause ambiguous parsing. The separate particle + verb
+  // pattern (를 + 토글) is preferred for consistent semantic analysis.
+  // This follows the same approach as the Japanese tokenizer.
 ];
 var KOREAN_TIME_UNITS = [
   { pattern: "\uBC00\uB9AC\uCD08", suffix: "ms", length: 3 },
@@ -1565,7 +1545,13 @@ var KoreanTokenizer = class extends BaseTokenizer {
       if (isKorean(input[pos])) {
         const wordToken = this.extractKoreanWord(input, pos);
         if (wordToken) {
-          tokens.push(wordToken);
+          const splitResult = this.trySplitTemporalSuffix(wordToken);
+          if (splitResult) {
+            tokens.push(splitResult.stemToken);
+            tokens.push(splitResult.suffixToken);
+          } else {
+            tokens.push(wordToken);
+          }
           pos = wordToken.position.end;
           continue;
         }
@@ -1644,6 +1630,13 @@ var KoreanTokenizer = class extends BaseTokenizer {
         }
       }
       if (!allKorean) continue;
+      if (PARTICLES.has(candidate) && startPos === startPos) {
+        const afterCandidate = startPos + len;
+        const nextChar = afterCandidate < input.length ? input[afterCandidate] : "";
+        if (nextChar === "" || isWhitespace(nextChar) || !isKorean(nextChar)) {
+          return null;
+        }
+      }
       const keywordEntry2 = this.lookupKeyword(candidate);
       if (keywordEntry2) {
         return createToken(
@@ -1661,6 +1654,9 @@ var KoreanTokenizer = class extends BaseTokenizer {
     while (pos < input.length) {
       const char = input[pos];
       const nextChar = pos + 1 < input.length ? input[pos + 1] : "";
+      if (word.length === 0 && SINGLE_CHAR_PARTICLES.has(char)) {
+        return null;
+      }
       if (SINGLE_CHAR_PARTICLES.has(char) && word.length > 0) {
         const isWordBoundary = nextChar === "" || isWhitespace(nextChar) || !isKorean(nextChar) || SINGLE_CHAR_PARTICLES.has(nextChar);
         if (isWordBoundary) {
@@ -1687,6 +1683,9 @@ var KoreanTokenizer = class extends BaseTokenizer {
       }
     }
     if (!word) return null;
+    if (PARTICLES.has(word)) {
+      return null;
+    }
     const keywordEntry = this.lookupKeyword(word);
     if (keywordEntry) {
       return createToken(word, "keyword", createPosition(startPos, pos), keywordEntry.normalized);
@@ -1717,6 +1716,39 @@ var KoreanTokenizer = class extends BaseTokenizer {
       skipWhitespace: false
     });
   }
+  /**
+   * Try to split a temporal event suffix from a word token.
+   * This handles compact forms like 클릭할때 → 클릭 + 할때
+   *
+   * @returns Split tokens if a suffix is found, null otherwise
+   */
+  trySplitTemporalSuffix(wordToken) {
+    const word = wordToken.value;
+    for (const suffix of TEMPORAL_EVENT_SUFFIXES) {
+      if (word.endsWith(suffix) && word.length > suffix.length) {
+        const stem = word.slice(0, -suffix.length);
+        const stemLower = stem.toLowerCase();
+        const keywordEntry = this.lookupKeyword(stemLower);
+        if (!keywordEntry) continue;
+        const stemEnd = wordToken.position.start + stem.length;
+        const stemToken = createToken(
+          stem,
+          "keyword",
+          createPosition(wordToken.position.start, stemEnd),
+          keywordEntry.normalized
+        );
+        const suffixToken = createToken(
+          suffix,
+          "keyword",
+          createPosition(stemEnd, wordToken.position.end),
+          "when"
+          // Normalize temporal suffixes to 'when'
+        );
+        return { stemToken, suffixToken };
+      }
+    }
+    return null;
+  }
 };
 var koreanTokenizer = new KoreanTokenizer();