@lokascript/semantic 1.2.1 → 2.0.0

package/dist/languages/th.js

@@ -0,0 +1,1123 @@
+// src/registry.ts
+var tokenizers = /* @__PURE__ */ new Map();
+var profiles = /* @__PURE__ */ new Map();
+var patternCache = /* @__PURE__ */ new Map();
+function registerLanguage(code, tokenizer, profile) {
+  tokenizers.set(code, tokenizer);
+  profiles.set(code, profile);
+  patternCache.delete(code);
+}
+
+// src/tokenizers/token-utils.ts
+var TokenStreamImpl = class {
+  constructor(tokens, language) {
+    this.pos = 0;
+    this.tokens = tokens;
+    this.language = language;
+  }
+  peek(offset = 0) {
+    const index = this.pos + offset;
+    if (index < 0 || index >= this.tokens.length) {
+      return null;
+    }
+    return this.tokens[index];
+  }
+  advance() {
+    if (this.isAtEnd()) {
+      throw new Error("Unexpected end of token stream");
+    }
+    return this.tokens[this.pos++];
+  }
+  isAtEnd() {
+    return this.pos >= this.tokens.length;
+  }
+  mark() {
+    return { position: this.pos };
+  }
+  reset(mark) {
+    this.pos = mark.position;
+  }
+  position() {
+    return this.pos;
+  }
+  /**
+   * Get remaining tokens as an array.
+   */
+  remaining() {
+    return this.tokens.slice(this.pos);
+  }
+  /**
+   * Consume tokens while predicate is true.
+   */
+  takeWhile(predicate) {
+    const result = [];
+    while (!this.isAtEnd() && predicate(this.peek())) {
+      result.push(this.advance());
+    }
+    return result;
+  }
+  /**
+   * Skip tokens while predicate is true.
+   */
+  skipWhile(predicate) {
+    while (!this.isAtEnd() && predicate(this.peek())) {
+      this.advance();
+    }
+  }
+};
+function createPosition(start, end) {
+  return { start, end };
+}
+function createToken(value, kind, position, normalizedOrOptions) {
+  if (typeof normalizedOrOptions === "string") {
+    return { value, kind, position, normalized: normalizedOrOptions };
+  }
+  if (normalizedOrOptions) {
+    const { normalized, stem, stemConfidence } = normalizedOrOptions;
+    return {
+      value,
+      kind,
+      position,
+      ...normalized !== void 0 && { normalized },
+      ...stem !== void 0 && { stem },
+      ...stemConfidence !== void 0 && { stemConfidence }
+    };
+  }
+  return { value, kind, position };
+}
+function isWhitespace(char) {
+  return /\s/.test(char);
+}
+function isSelectorStart(char) {
+  return char === "#" || char === "." || char === "[" || char === "@" || char === "*" || char === "<";
+}
+function isQuote(char) {
+  return char === '"' || char === "'" || char === "`" || char === "\u300C" || char === "\u300D";
+}
+function isDigit(char) {
+  return /\d/.test(char);
+}
+function isAsciiLetter(char) {
+  return /[a-zA-Z]/.test(char);
+}
+function isAsciiIdentifierChar(char) {
+  return /[a-zA-Z0-9_-]/.test(char);
+}
+
+// src/tokenizers/char-classifiers.ts
+function createUnicodeRangeClassifier(ranges) {
+  return (char) => {
+    const code = char.charCodeAt(0);
+    return ranges.some(([start, end]) => code >= start && code <= end);
+  };
+}
+
+// src/tokenizers/extractors.ts
+function extractCssSelector(input, startPos) {
+  if (startPos >= input.length) return null;
+  const char = input[startPos];
+  if (!isSelectorStart(char)) return null;
+  let pos = startPos;
+  let selector = "";
+  if (char === "#" || char === ".") {
+    selector += input[pos++];
+    while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+      selector += input[pos++];
+    }
+    if (selector.length <= 1) return null;
+    if (pos < input.length && input[pos] === "." && char === "#") {
+      const methodStart = pos + 1;
+      let methodEnd = methodStart;
+      while (methodEnd < input.length && isAsciiIdentifierChar(input[methodEnd])) {
+        methodEnd++;
+      }
+      if (methodEnd < input.length && input[methodEnd] === "(") {
+        return selector;
+      }
+    }
+  } else if (char === "[") {
+    let depth = 1;
+    let inQuote = false;
+    let quoteChar = null;
+    let escaped = false;
+    selector += input[pos++];
+    while (pos < input.length && depth > 0) {
+      const c = input[pos];
+      selector += c;
+      if (escaped) {
+        escaped = false;
+      } else if (c === "\\") {
+        escaped = true;
+      } else if (inQuote) {
+        if (c === quoteChar) {
+          inQuote = false;
+          quoteChar = null;
+        }
+      } else {
+        if (c === '"' || c === "'" || c === "`") {
+          inQuote = true;
+          quoteChar = c;
+        } else if (c === "[") {
+          depth++;
+        } else if (c === "]") {
+          depth--;
+        }
+      }
+      pos++;
+    }
+    if (depth !== 0) return null;
+  } else if (char === "@") {
+    selector += input[pos++];
+    while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+      selector += input[pos++];
+    }
+    if (selector.length <= 1) return null;
+  } else if (char === "*") {
+    selector += input[pos++];
+    while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+      selector += input[pos++];
+    }
+    if (selector.length <= 1) return null;
+  } else if (char === "<") {
+    selector += input[pos++];
+    if (pos >= input.length || !isAsciiLetter(input[pos])) return null;
+    while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+      selector += input[pos++];
+    }
+    while (pos < input.length) {
+      const modChar = input[pos];
+      if (modChar === ".") {
+        selector += input[pos++];
+        if (pos >= input.length || !isAsciiIdentifierChar(input[pos])) {
+          return null;
+        }
+        while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+          selector += input[pos++];
+        }
+      } else if (modChar === "#") {
+        selector += input[pos++];
+        if (pos >= input.length || !isAsciiIdentifierChar(input[pos])) {
+          return null;
+        }
+        while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+          selector += input[pos++];
+        }
+      } else if (modChar === "[") {
+        let depth = 1;
+        let inQuote = false;
+        let quoteChar = null;
+        let escaped = false;
+        selector += input[pos++];
+        while (pos < input.length && depth > 0) {
+          const c = input[pos];
+          selector += c;
+          if (escaped) {
+            escaped = false;
+          } else if (c === "\\") {
+            escaped = true;
+          } else if (inQuote) {
+            if (c === quoteChar) {
+              inQuote = false;
+              quoteChar = null;
+            }
+          } else {
+            if (c === '"' || c === "'" || c === "`") {
+              inQuote = true;
+              quoteChar = c;
+            } else if (c === "[") {
+              depth++;
+            } else if (c === "]") {
+              depth--;
+            }
+          }
+          pos++;
+        }
+        if (depth !== 0) return null;
+      } else {
+        break;
+      }
+    }
+    while (pos < input.length && isWhitespace(input[pos])) {
+      selector += input[pos++];
+    }
+    if (pos < input.length && input[pos] === "/") {
+      selector += input[pos++];
+      while (pos < input.length && isWhitespace(input[pos])) {
+        selector += input[pos++];
+      }
+    }
+    if (pos >= input.length || input[pos] !== ">") return null;
+    selector += input[pos++];
+  }
+  return selector || null;
+}
+function isPossessiveMarker(input, pos) {
+  if (pos >= input.length || input[pos] !== "'") return false;
+  if (pos + 1 >= input.length) return false;
+  const nextChar = input[pos + 1].toLowerCase();
+  if (nextChar !== "s") return false;
+  if (pos + 2 >= input.length) return true;
+  const afterS = input[pos + 2];
+  return isWhitespace(afterS) || afterS === "*" || !isAsciiIdentifierChar(afterS);
+}
+function extractStringLiteral(input, startPos) {
+  if (startPos >= input.length) return null;
+  const openQuote = input[startPos];
+  if (!isQuote(openQuote)) return null;
+  if (openQuote === "'" && isPossessiveMarker(input, startPos)) {
+    return null;
+  }
+  const closeQuoteMap = {
+    '"': '"',
+    "'": "'",
+    "`": "`",
+    "\u300C": "\u300D"
+  };
+  const closeQuote = closeQuoteMap[openQuote];
+  if (!closeQuote) return null;
+  let pos = startPos + 1;
+  let literal = openQuote;
+  let escaped = false;
+  while (pos < input.length) {
+    const char = input[pos];
+    literal += char;
+    if (escaped) {
+      escaped = false;
+    } else if (char === "\\") {
+      escaped = true;
+    } else if (char === closeQuote) {
+      return literal;
+    }
+    pos++;
+  }
+  return literal;
+}
+function isUrlStart(input, pos) {
+  if (pos >= input.length) return false;
+  const char = input[pos];
+  const next = input[pos + 1] || "";
+  const third = input[pos + 2] || "";
+  if (char === "/" && next !== "/" && /[a-zA-Z0-9._-]/.test(next)) {
+    return true;
+  }
+  if (char === "/" && next === "/" && /[a-zA-Z]/.test(third)) {
+    return true;
+  }
+  if (char === "." && (next === "/" || next === "." && third === "/")) {
+    return true;
+  }
+  const slice = input.slice(pos, pos + 8).toLowerCase();
+  if (slice.startsWith("http://") || slice.startsWith("https://")) {
+    return true;
+  }
+  return false;
+}
+function extractUrl(input, startPos) {
+  if (!isUrlStart(input, startPos)) return null;
+  let pos = startPos;
+  let url = "";
+  const urlChars = /[a-zA-Z0-9/:._\-?&=%@+~!$'()*,;[\]]/;
+  while (pos < input.length) {
+    const char = input[pos];
+    if (char === "#") {
+      if (url.length > 0 && /[a-zA-Z0-9/.]$/.test(url)) {
+        url += char;
+        pos++;
+        while (pos < input.length && /[a-zA-Z0-9_-]/.test(input[pos])) {
+          url += input[pos++];
+        }
+      }
+      break;
+    }
+    if (urlChars.test(char)) {
+      url += char;
+      pos++;
+    } else {
+      break;
+    }
+  }
+  if (url.length < 2) return null;
+  return url;
+}
+function extractNumber(input, startPos) {
+  if (startPos >= input.length) return null;
+  const char = input[startPos];
+  if (!isDigit(char) && char !== "-" && char !== "+") return null;
+  let pos = startPos;
+  let number = "";
+  if (input[pos] === "-" || input[pos] === "+") {
+    number += input[pos++];
+  }
+  if (pos >= input.length || !isDigit(input[pos])) {
+    return null;
+  }
+  while (pos < input.length && isDigit(input[pos])) {
+    number += input[pos++];
+  }
+  if (pos < input.length && input[pos] === ".") {
+    number += input[pos++];
+    while (pos < input.length && isDigit(input[pos])) {
+      number += input[pos++];
+    }
+  }
+  if (pos < input.length) {
+    const suffix = input.slice(pos, pos + 2);
+    if (suffix === "ms") {
+      number += "ms";
+    } else if (input[pos] === "s" || input[pos] === "m" || input[pos] === "h") {
+      number += input[pos];
+    }
+  }
+  return number;
+}
+
+// src/tokenizers/base-tokenizer.ts
+var _BaseTokenizer = class _BaseTokenizer {
+  constructor() {
+    /** Keywords derived from profile, sorted longest-first for greedy matching */
+    this.profileKeywords = [];
+    /** 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:
+   * - profile.keywords (primary + alternatives)
+   * - profile.references (me, it, you, etc.)
+   * - profile.roleMarkers (into, from, with, etc.)
+   *
+   * Results are sorted longest-first for greedy matching (important for non-space languages).
+   * Extras take precedence over profile entries when there are duplicates.
+   *
+   * @param profile - Language profile containing keyword translations
+   * @param extras - Additional keyword entries to include (literals, positional, events)
+   */
+  initializeKeywordsFromProfile(profile, extras = []) {
+    const keywordMap = /* @__PURE__ */ new Map();
+    if (profile.keywords) {
+      for (const [normalized, translation] of Object.entries(profile.keywords)) {
+        keywordMap.set(translation.primary, {
+          native: translation.primary,
+          normalized: translation.normalized || normalized
+        });
+        if (translation.alternatives) {
+          for (const alt of translation.alternatives) {
+            keywordMap.set(alt, {
+              native: alt,
+              normalized: translation.normalized || normalized
+            });
+          }
+        }
+      }
+    }
+    if (profile.references) {
+      for (const [normalized, native] of Object.entries(profile.references)) {
+        keywordMap.set(native, { native, normalized });
+      }
+      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)) {
+        if (marker.primary) {
+          keywordMap.set(marker.primary, { native: marker.primary, normalized: role });
+        }
+        if (marker.alternatives) {
+          for (const alt of marker.alternatives) {
+            keywordMap.set(alt, { native: alt, normalized: role });
+          }
+        }
+      }
+    }
+    for (const extra of extras) {
+      keywordMap.set(extra.native, extra);
+    }
+    this.profileKeywords = Array.from(keywordMap.values()).sort(
+      (a, b) => b.native.length - a.native.length
+    );
+    this.profileKeywordMap = /* @__PURE__ */ new Map();
+    for (const keyword of this.profileKeywords) {
+      this.profileKeywordMap.set(keyword.native.toLowerCase(), keyword);
+      const normalized = this.removeDiacritics(keyword.native);
+      if (normalized !== keyword.native && !this.profileKeywordMap.has(normalized.toLowerCase())) {
+        this.profileKeywordMap.set(normalized.toLowerCase(), keyword);
+      }
+    }
+  }
+  /**
+   * Remove diacritical marks from a word for normalization.
+   * Primarily for Arabic (shadda, fatha, kasra, damma, sukun, etc.)
+   * but could be extended for other languages.
+   *
+   * @param word - Word to normalize
+   * @returns Word without diacritics
+   */
+  removeDiacritics(word) {
+    return word.replace(/[\u064B-\u0652\u0670]/g, "");
+  }
+  /**
+   * Try to match a keyword from profile at the current position.
+   * Uses longest-first greedy matching (important for non-space languages).
+   *
+   * @param input - Input string
+   * @param pos - Current position
+   * @returns Token if matched, null otherwise
+   */
+  tryProfileKeyword(input, pos) {
+    for (const entry of this.profileKeywords) {
+      if (input.slice(pos).startsWith(entry.native)) {
+        return createToken(
+          entry.native,
+          "keyword",
+          createPosition(pos, pos + entry.native.length),
+          entry.normalized
+        );
+      }
+    }
+    return null;
+  }
+  /**
+   * Check if the remaining input starts with any known keyword.
+   * Useful for non-space languages to detect word boundaries.
+   *
+   * @param input - Input string
+   * @param pos - Current position
+   * @returns true if a keyword starts at this position
+   */
+  isKeywordStart(input, pos) {
+    const remaining = input.slice(pos);
+    return this.profileKeywords.some((entry) => remaining.startsWith(entry.native));
+  }
+  /**
+   * Look up a keyword by native word (case-insensitive).
+   * O(1) lookup using the keyword map.
+   *
+   * @param native - Native word to look up
+   * @returns KeywordEntry if found, undefined otherwise
+   */
+  lookupKeyword(native) {
+    return this.profileKeywordMap.get(native.toLowerCase());
+  }
+  /**
+   * Check if a word is a known keyword (case-insensitive).
+   * O(1) lookup using the keyword map.
+   *
+   * @param native - Native word to check
+   * @returns true if the word is a keyword
+   */
+  isKeyword(native) {
+    return this.profileKeywordMap.has(native.toLowerCase());
+  }
+  /**
+   * Set the morphological normalizer for this tokenizer.
+   */
+  setNormalizer(normalizer) {
+    this.normalizer = normalizer;
+  }
+  /**
+   * Try to normalize a word using the morphological normalizer.
+   * Returns null if no normalizer is set or normalization fails.
+   *
+   * Note: We don't check isNormalizable() here because the individual tokenizers
+   * historically called normalize() directly without that check. The normalize()
+   * method itself handles returning noChange() for words that can't be normalized.
+   */
+  tryNormalize(word) {
+    if (!this.normalizer) return null;
+    const result = this.normalizer.normalize(word);
+    if (result.stem !== word && result.confidence >= 0.7) {
+      return result;
+    }
+    return null;
+  }
+  /**
+   * Try morphological normalization and keyword lookup.
+   *
+   * If the word can be normalized to a stem that matches a known keyword,
+   * returns a keyword token with morphological metadata (stem, stemConfidence).
+   *
+   * This is the common pattern for handling conjugated verbs across languages:
+   * 1. Normalize the word (e.g., "toggled" → "toggle")
+   * 2. Look up the stem in the keyword map
+   * 3. Create a token with both the original form and stem metadata
+   *
+   * @param word - The word to normalize and look up
+   * @param startPos - Start position for the token
+   * @param endPos - End position for the token
+   * @returns Token if stem matches a keyword, null otherwise
+   */
+  tryMorphKeywordMatch(word, startPos, endPos) {
+    const result = this.tryNormalize(word);
+    if (!result) return null;
+    const stemEntry = this.lookupKeyword(result.stem);
+    if (!stemEntry) return null;
+    const tokenOptions = {
+      normalized: stemEntry.normalized,
+      stem: result.stem,
+      stemConfidence: result.confidence
+    };
+    return createToken(word, "keyword", createPosition(startPos, endPos), tokenOptions);
+  }
+  /**
+   * Try to extract a CSS selector at the current position.
+   */
+  trySelector(input, pos) {
+    const selector = extractCssSelector(input, pos);
+    if (selector) {
+      return createToken(selector, "selector", createPosition(pos, pos + selector.length));
+    }
+    return null;
+  }
+  /**
+   * Try to extract an event modifier at the current position.
+   * Event modifiers are .once, .debounce(N), .throttle(N), .queue(strategy)
+   */
+  tryEventModifier(input, pos) {
+    if (input[pos] !== ".") {
+      return null;
+    }
+    const match = input.slice(pos).match(/^\.(?:once|debounce|throttle|queue)(?:\(([^)]+)\))?(?:\s|$|\.)/);
+    if (!match) {
+      return null;
+    }
+    const fullMatch = match[0].replace(/(\s|\.)$/, "");
+    const modifierName = fullMatch.slice(1).split("(")[0];
+    const value = match[1];
+    const token = createToken(
+      fullMatch,
+      "event-modifier",
+      createPosition(pos, pos + fullMatch.length)
+    );
+    return {
+      ...token,
+      metadata: {
+        modifierName,
+        value: value ? modifierName === "queue" ? value : parseInt(value, 10) : void 0
+      }
+    };
+  }
+  /**
+   * Try to extract a string literal at the current position.
+   */
+  tryString(input, pos) {
+    const literal = extractStringLiteral(input, pos);
+    if (literal) {
+      return createToken(literal, "literal", createPosition(pos, pos + literal.length));
+    }
+    return null;
+  }
+  /**
+   * Try to extract a number at the current position.
+   */
+  tryNumber(input, pos) {
+    const number = extractNumber(input, pos);
+    if (number) {
+      return createToken(number, "literal", createPosition(pos, pos + number.length));
+    }
+    return null;
+  }
+  /**
+   * Try to match a time unit from a list of patterns.
+   *
+   * @param input - Input string
+   * @param pos - Position after the number
+   * @param timeUnits - Array of time unit mappings (native pattern → standard suffix)
+   * @param skipWhitespace - Whether to skip whitespace before time unit (default: false)
+   * @returns Object with matched suffix and new position, or null if no match
+   */
+  tryMatchTimeUnit(input, pos, timeUnits, skipWhitespace = false) {
+    let unitPos = pos;
+    if (skipWhitespace) {
+      while (unitPos < input.length && isWhitespace(input[unitPos])) {
+        unitPos++;
+      }
+    }
+    const remaining = input.slice(unitPos);
+    for (const unit of timeUnits) {
+      const candidate = remaining.slice(0, unit.length);
+      const matches = unit.caseInsensitive ? candidate.toLowerCase() === unit.pattern.toLowerCase() : candidate === unit.pattern;
+      if (matches) {
+        if (unit.notFollowedBy) {
+          const nextChar = remaining[unit.length] || "";
+          if (nextChar === unit.notFollowedBy) continue;
+        }
+        if (unit.checkBoundary) {
+          const nextChar = remaining[unit.length] || "";
+          if (isAsciiIdentifierChar(nextChar)) continue;
+        }
+        return { suffix: unit.suffix, endPos: unitPos + unit.length };
+      }
+    }
+    return null;
+  }
+  /**
+   * Parse a base number (sign, integer, decimal) without time units.
+   * Returns the number string and end position.
+   *
+   * @param input - Input string
+   * @param startPos - Start position
+   * @param allowSign - Whether to allow +/- sign (default: true)
+   * @returns Object with number string and end position, or null
+   */
+  parseBaseNumber(input, startPos, allowSign = true) {
+    let pos = startPos;
+    let number = "";
+    if (allowSign && (input[pos] === "-" || input[pos] === "+")) {
+      number += input[pos++];
+    }
+    if (pos >= input.length || !isDigit(input[pos])) {
+      return null;
+    }
+    while (pos < input.length && isDigit(input[pos])) {
+      number += input[pos++];
+    }
+    if (pos < input.length && input[pos] === ".") {
+      number += input[pos++];
+      while (pos < input.length && isDigit(input[pos])) {
+        number += input[pos++];
+      }
+    }
+    if (!number || number === "-" || number === "+") return null;
+    return { number, endPos: pos };
+  }
+  /**
+   * Try to extract a number with native language time units.
+   *
+   * This is a template method that handles the common pattern:
+   * 1. Parse the base number (sign, integer, decimal)
+   * 2. Try to match native language time units
+   * 3. Fall back to standard time units (ms, s, m, h)
+   *
+   * @param input - Input string
+   * @param pos - Start position
+   * @param nativeTimeUnits - Language-specific time unit mappings
+   * @param options - Configuration options
+   * @returns Token if number found, null otherwise
+   */
+  tryNumberWithTimeUnits(input, pos, nativeTimeUnits, options = {}) {
+    const { allowSign = true, skipWhitespace = false } = options;
+    const baseResult = this.parseBaseNumber(input, pos, allowSign);
+    if (!baseResult) return null;
+    let { number, endPos } = baseResult;
+    const allUnits = [...nativeTimeUnits, ..._BaseTokenizer.STANDARD_TIME_UNITS];
+    const timeMatch = this.tryMatchTimeUnit(input, endPos, allUnits, skipWhitespace);
+    if (timeMatch) {
+      number += timeMatch.suffix;
+      endPos = timeMatch.endPos;
+    }
+    return createToken(number, "literal", createPosition(pos, endPos));
+  }
+  /**
+   * Try to extract a URL at the current position.
+   * Handles /path, ./path, ../path, //domain.com, http://, https://
+   */
+  tryUrl(input, pos) {
+    const url = extractUrl(input, pos);
+    if (url) {
+      return createToken(url, "url", createPosition(pos, pos + url.length));
+    }
+    return null;
+  }
+  /**
+   * Try to extract a variable reference (:varname) at the current position.
+   * In hyperscript, :x refers to a local variable named x.
+   */
+  tryVariableRef(input, pos) {
+    if (input[pos] !== ":") return null;
+    if (pos + 1 >= input.length) return null;
+    if (!isAsciiIdentifierChar(input[pos + 1])) return null;
+    let endPos = pos + 1;
+    while (endPos < input.length && isAsciiIdentifierChar(input[endPos])) {
+      endPos++;
+    }
+    const varRef = input.slice(pos, endPos);
+    return createToken(varRef, "identifier", createPosition(pos, endPos));
+  }
+  /**
+   * Try to extract an operator or punctuation token at the current position.
+   * Handles two-character operators (==, !=, etc.) and single-character operators.
+   */
+  tryOperator(input, pos) {
+    const twoChar = input.slice(pos, pos + 2);
+    if (["==", "!=", "<=", ">=", "&&", "||", "->"].includes(twoChar)) {
+      return createToken(twoChar, "operator", createPosition(pos, pos + 2));
+    }
+    const oneChar = input[pos];
+    if (["<", ">", "!", "+", "-", "*", "/", "="].includes(oneChar)) {
+      return createToken(oneChar, "operator", createPosition(pos, pos + 1));
+    }
+    if (["(", ")", "{", "}", ",", ";", ":"].includes(oneChar)) {
+      return createToken(oneChar, "punctuation", createPosition(pos, pos + 1));
+    }
+    return null;
+  }
+  /**
+   * Try to match a multi-character particle from a list.
+   *
+   * Used by languages like Japanese, Korean, and Chinese that have
+   * multi-character particles (e.g., Japanese から, まで, より).
+   *
+   * @param input - Input string
+   * @param pos - Current position
+   * @param particles - Array of multi-character particles to match
+   * @returns Token if matched, null otherwise
+   */
+  tryMultiCharParticle(input, pos, particles) {
+    for (const particle of particles) {
+      if (input.slice(pos, pos + particle.length) === particle) {
+        return createToken(particle, "particle", createPosition(pos, pos + particle.length));
+      }
+    }
+    return null;
+  }
+};
+/**
+ * Configuration for native language time units.
+ * Maps patterns to their standard suffix (ms, s, m, h).
+ */
+_BaseTokenizer.STANDARD_TIME_UNITS = [
+  { pattern: "ms", suffix: "ms", length: 2 },
+  { pattern: "s", suffix: "s", length: 1, checkBoundary: true },
+  { pattern: "m", suffix: "m", length: 1, checkBoundary: true, notFollowedBy: "s" },
+  { pattern: "h", suffix: "h", length: 1, checkBoundary: true }
+];
+var BaseTokenizer = _BaseTokenizer;
+
+// src/generators/profiles/thai.ts
+var thaiProfile = {
+  code: "th",
+  name: "Thai",
+  nativeName: "\u0E44\u0E17\u0E22",
+  direction: "ltr",
+  wordOrder: "SVO",
+  markingStrategy: "preposition",
+  usesSpaces: false,
+  // Thai doesn't use spaces between words
+  defaultVerbForm: "base",
+  verb: {
+    position: "second",
+    subjectDrop: true
+  },
+  references: {
+    me: "\u0E09\u0E31\u0E19",
+    it: "\u0E21\u0E31\u0E19",
+    you: "\u0E04\u0E38\u0E13",
+    result: "\u0E1C\u0E25\u0E25\u0E31\u0E1E\u0E18\u0E4C",
+    event: "\u0E40\u0E2B\u0E15\u0E38\u0E01\u0E32\u0E23\u0E13\u0E4C",
+    target: "\u0E40\u0E1B\u0E49\u0E32\u0E2B\u0E21\u0E32\u0E22",
+    body: "\u0E1A\u0E2D\u0E14\u0E35\u0E49"
+  },
+  possessive: {
+    marker: "\u0E02\u0E2D\u0E07",
+    markerPosition: "between",
+    keywords: {
+      // Thai uses ของ (khong) + pronoun, but also full possessive phrases
+      // "my" - ของฉัน (khong chan), ของผม (khong phom, male speaker)
+      \u0E02\u0E2D\u0E07\u0E09\u0E31\u0E19: "me",
+      \u0E02\u0E2D\u0E07\u0E1C\u0E21: "me",
+      // "your" - ของคุณ (khong khun)
+      \u0E02\u0E2D\u0E07\u0E04\u0E38\u0E13: "you",
+      // "its/his/her" - ของมัน (khong man)
+      \u0E02\u0E2D\u0E07\u0E21\u0E31\u0E19: "it",
+      \u0E02\u0E2D\u0E07\u0E40\u0E02\u0E32: "it",
+      // his
+      \u0E02\u0E2D\u0E07\u0E40\u0E18\u0E2D: "it"
+      // her
+    }
+  },
+  roleMarkers: {
+    patient: { primary: "", position: "before" },
+    // Thai is isolating - no case markers
+    destination: { primary: "\u0E43\u0E19", alternatives: ["\u0E44\u0E1B\u0E22\u0E31\u0E07"], position: "before" },
+    source: { primary: "\u0E08\u0E32\u0E01", position: "before" },
+    style: { primary: "\u0E14\u0E49\u0E27\u0E22", position: "before" },
+    event: { primary: "\u0E40\u0E21\u0E37\u0E48\u0E2D", position: "before" }
+  },
+  keywords: {
+    // Class/Attribute operations
+    toggle: { primary: "\u0E2A\u0E25\u0E31\u0E1A", alternatives: [], normalized: "toggle" },
+    add: { primary: "\u0E40\u0E1E\u0E34\u0E48\u0E21", alternatives: [], normalized: "add" },
+    remove: { primary: "\u0E25\u0E1A", alternatives: ["\u0E25\u0E1A\u0E2D\u0E2D\u0E01"], normalized: "remove" },
+    // Content operations
+    put: { primary: "\u0E43\u0E2A\u0E48", alternatives: ["\u0E27\u0E32\u0E07"], normalized: "put" },
+    append: { primary: "\u0E40\u0E1E\u0E34\u0E48\u0E21\u0E17\u0E49\u0E32\u0E22", alternatives: [], normalized: "append" },
+    prepend: { primary: "\u0E40\u0E1E\u0E34\u0E48\u0E21\u0E2B\u0E19\u0E49\u0E32", alternatives: [], normalized: "prepend" },
+    take: { primary: "\u0E23\u0E31\u0E1A", alternatives: [], normalized: "take" },
+    make: { primary: "\u0E2A\u0E23\u0E49\u0E32\u0E07", alternatives: [], normalized: "make" },
+    clone: { primary: "\u0E04\u0E31\u0E14\u0E25\u0E2D\u0E01", alternatives: ["\u0E2A\u0E33\u0E40\u0E19\u0E32"], normalized: "clone" },
+    swap: { primary: "\u0E2A\u0E25\u0E31\u0E1A\u0E17\u0E35\u0E48", alternatives: [], normalized: "swap" },
+    morph: { primary: "\u0E41\u0E1B\u0E25\u0E07\u0E23\u0E48\u0E32\u0E07", alternatives: [], normalized: "morph" },
+    // Variable operations
+    set: { primary: "\u0E15\u0E31\u0E49\u0E07", alternatives: ["\u0E01\u0E33\u0E2B\u0E19\u0E14"], normalized: "set" },
+    get: { primary: "\u0E23\u0E31\u0E1A\u0E04\u0E48\u0E32", alternatives: [], normalized: "get" },
+    increment: { primary: "\u0E40\u0E1E\u0E34\u0E48\u0E21\u0E04\u0E48\u0E32", alternatives: [], normalized: "increment" },
+    decrement: { primary: "\u0E25\u0E14\u0E04\u0E48\u0E32", alternatives: [], normalized: "decrement" },
+    log: { primary: "\u0E1A\u0E31\u0E19\u0E17\u0E36\u0E01", alternatives: [], normalized: "log" },
+    // Visibility
+    show: { primary: "\u0E41\u0E2A\u0E14\u0E07", alternatives: [], normalized: "show" },
+    hide: { primary: "\u0E0B\u0E48\u0E2D\u0E19", alternatives: [], normalized: "hide" },
+    transition: { primary: "\u0E40\u0E1B\u0E25\u0E35\u0E48\u0E22\u0E19", alternatives: [], normalized: "transition" },
+    // Events
+    on: { primary: "\u0E40\u0E21\u0E37\u0E48\u0E2D", alternatives: ["\u0E15\u0E2D\u0E19"], normalized: "on" },
+    trigger: { primary: "\u0E17\u0E23\u0E34\u0E01\u0E40\u0E01\u0E2D\u0E23\u0E4C", alternatives: [], normalized: "trigger" },
+    send: { primary: "\u0E2A\u0E48\u0E07", alternatives: [], normalized: "send" },
+    // DOM focus
+    focus: { primary: "\u0E42\u0E1F\u0E01\u0E31\u0E2A", alternatives: [], normalized: "focus" },
+    blur: { primary: "\u0E40\u0E1A\u0E25\u0E2D", alternatives: [], normalized: "blur" },
+    // Common event names (for event handler patterns)
+    click: { primary: "\u0E04\u0E25\u0E34\u0E01", normalized: "click" },
+    hover: { primary: "\u0E42\u0E2E\u0E40\u0E27\u0E2D\u0E23\u0E4C", alternatives: ["\u0E27\u0E32\u0E07\u0E40\u0E21\u0E32\u0E2A\u0E4C"], normalized: "hover" },
+    submit: { primary: "\u0E2A\u0E48\u0E07", alternatives: ["\u0E2A\u0E48\u0E07\u0E02\u0E49\u0E2D\u0E21\u0E39\u0E25"], normalized: "submit" },
+    input: { primary: "\u0E1B\u0E49\u0E2D\u0E19", alternatives: ["\u0E01\u0E23\u0E2D\u0E01"], normalized: "input" },
+    change: { primary: "\u0E40\u0E1B\u0E25\u0E35\u0E48\u0E22\u0E19", alternatives: ["\u0E40\u0E1B\u0E25\u0E35\u0E48\u0E22\u0E19\u0E41\u0E1B\u0E25\u0E07"], normalized: "change" },
+    // Navigation
+    go: { primary: "\u0E44\u0E1B", alternatives: ["\u0E44\u0E1B\u0E17\u0E35\u0E48"], normalized: "go" },
+    // Async
+    wait: { primary: "\u0E23\u0E2D", alternatives: [], normalized: "wait" },
+    fetch: { primary: "\u0E14\u0E36\u0E07\u0E02\u0E49\u0E2D\u0E21\u0E39\u0E25", alternatives: [], normalized: "fetch" },
+    settle: { primary: "\u0E04\u0E07\u0E17\u0E35\u0E48", alternatives: [], normalized: "settle" },
+    // Control flow
+    if: { primary: "\u0E16\u0E49\u0E32", alternatives: ["\u0E2B\u0E32\u0E01"], normalized: "if" },
+    when: { primary: "\u0E40\u0E21\u0E37\u0E48\u0E2D", normalized: "when" },
+    where: { primary: "\u0E17\u0E35\u0E48\u0E44\u0E2B\u0E19", normalized: "where" },
+    else: { primary: "\u0E44\u0E21\u0E48\u0E07\u0E31\u0E49\u0E19", alternatives: ["\u0E44\u0E21\u0E48\u0E40\u0E0A\u0E48\u0E19\u0E19\u0E31\u0E49\u0E19"], normalized: "else" },
+    repeat: { primary: "\u0E17\u0E33\u0E0B\u0E49\u0E33", alternatives: [], normalized: "repeat" },
+    for: { primary: "\u0E2A\u0E33\u0E2B\u0E23\u0E31\u0E1A", alternatives: [], normalized: "for" },
+    while: { primary: "\u0E43\u0E19\u0E02\u0E13\u0E30\u0E17\u0E35\u0E48", alternatives: [], normalized: "while" },
+    continue: { primary: "\u0E15\u0E48\u0E2D\u0E44\u0E1B", alternatives: [], normalized: "continue" },
+    halt: { primary: "\u0E2B\u0E22\u0E38\u0E14", alternatives: [], normalized: "halt" },
+    throw: { primary: "\u0E42\u0E22\u0E19", alternatives: [], normalized: "throw" },
+    call: { primary: "\u0E40\u0E23\u0E35\u0E22\u0E01", alternatives: [], normalized: "call" },
+    return: { primary: "\u0E04\u0E37\u0E19\u0E04\u0E48\u0E32", alternatives: ["\u0E01\u0E25\u0E31\u0E1A"], normalized: "return" },
+    then: { primary: "\u0E41\u0E25\u0E49\u0E27", alternatives: [], normalized: "then" },
+    and: { primary: "\u0E41\u0E25\u0E30", alternatives: [], normalized: "and" },
+    end: { primary: "\u0E08\u0E1A", alternatives: [], normalized: "end" },
+    // Advanced
+    js: { primary: "\u0E40\u0E08\u0E40\u0E2D\u0E2A", alternatives: ["js"], normalized: "js" },
+    async: { primary: "\u0E2D\u0E30\u0E0B\u0E34\u0E07\u0E04\u0E4C", alternatives: [], normalized: "async" },
+    tell: { primary: "\u0E1A\u0E2D\u0E01", alternatives: [], normalized: "tell" },
+    default: { primary: "\u0E04\u0E48\u0E32\u0E40\u0E23\u0E34\u0E48\u0E21\u0E15\u0E49\u0E19", alternatives: [], normalized: "default" },
+    init: { primary: "\u0E40\u0E23\u0E34\u0E48\u0E21\u0E15\u0E49\u0E19", alternatives: [], normalized: "init" },
+    behavior: { primary: "\u0E1E\u0E24\u0E15\u0E34\u0E01\u0E23\u0E23\u0E21", alternatives: [], normalized: "behavior" },
+    install: { primary: "\u0E15\u0E34\u0E14\u0E15\u0E31\u0E49\u0E07", alternatives: [], normalized: "install" },
+    measure: { primary: "\u0E27\u0E31\u0E14", alternatives: [], normalized: "measure" },
+    beep: { primary: "\u0E1A\u0E35\u0E4A\u0E1A", alternatives: [], normalized: "beep" },
+    break: { primary: "\u0E2B\u0E22\u0E38\u0E14", alternatives: [], normalized: "break" },
+    copy: { primary: "\u0E04\u0E31\u0E14\u0E25\u0E2D\u0E01", alternatives: [], normalized: "copy" },
+    exit: { primary: "\u0E2D\u0E2D\u0E01", alternatives: [], normalized: "exit" },
+    pick: { primary: "\u0E40\u0E25\u0E37\u0E2D\u0E01", alternatives: [], normalized: "pick" },
+    render: { primary: "\u0E41\u0E2A\u0E14\u0E07\u0E1C\u0E25", alternatives: [], normalized: "render" },
+    // Modifiers
+    into: { primary: "\u0E43\u0E19", alternatives: [], normalized: "into" },
+    before: { primary: "\u0E01\u0E48\u0E2D\u0E19", alternatives: [], normalized: "before" },
+    after: { primary: "\u0E2B\u0E25\u0E31\u0E07", alternatives: [], normalized: "after" },
+    until: { primary: "\u0E08\u0E19\u0E16\u0E36\u0E07", alternatives: [], normalized: "until" },
+    event: { primary: "\u0E40\u0E2B\u0E15\u0E38\u0E01\u0E32\u0E23\u0E13\u0E4C", alternatives: [], normalized: "event" },
+    from: { primary: "\u0E08\u0E32\u0E01", normalized: "from" }
+  },
+  tokenization: {
+    boundaryStrategy: "character"
+    // Character-based like Chinese/Japanese
+  },
+  eventHandler: {
+    keyword: { primary: "\u0E40\u0E21\u0E37\u0E48\u0E2D", alternatives: ["\u0E15\u0E2D\u0E19"], normalized: "on" },
+    sourceMarker: { primary: "\u0E08\u0E32\u0E01", position: "before" },
+    // Event marker: เมื่อ (when), used in SVO pattern
+    // Pattern: เมื่อ [event] [verb] [patient] ใน [destination?]
+    // Example: เมื่อ คลิก สลับ .active ใน #button
+    eventMarker: { primary: "\u0E40\u0E21\u0E37\u0E48\u0E2D", alternatives: ["\u0E15\u0E2D\u0E19"], position: "before" },
+    temporalMarkers: ["\u0E40\u0E21\u0E37\u0E48\u0E2D", "\u0E15\u0E2D\u0E19"]
+    // temporal conjunctions (when)
+  }
+};
+
+// src/tokenizers/thai.ts
+var isThai = createUnicodeRangeClassifier([[3584, 3711]]);
+var THAI_EXTRAS = [
+  // Values/Literals
+  { native: "\u0E08\u0E23\u0E34\u0E07", normalized: "true" },
+  { native: "\u0E40\u0E17\u0E47\u0E08", normalized: "false" },
+  { native: "\u0E27\u0E48\u0E32\u0E07", normalized: "null" },
+  { native: "\u0E44\u0E21\u0E48\u0E01\u0E33\u0E2B\u0E19\u0E14", normalized: "undefined" },
+  // Positional
+  { native: "\u0E41\u0E23\u0E01", normalized: "first" },
+  { native: "\u0E2A\u0E38\u0E14\u0E17\u0E49\u0E32\u0E22", normalized: "last" },
+  { native: "\u0E16\u0E31\u0E14\u0E44\u0E1B", normalized: "next" },
+  { native: "\u0E01\u0E48\u0E2D\u0E19\u0E2B\u0E19\u0E49\u0E32", normalized: "previous" },
+  { native: "\u0E43\u0E01\u0E25\u0E49\u0E2A\u0E38\u0E14", normalized: "closest" },
+  { native: "\u0E15\u0E49\u0E19\u0E17\u0E32\u0E07", normalized: "parent" },
+  // Events
+  { native: "\u0E04\u0E25\u0E34\u0E01", normalized: "click" },
+  { native: "\u0E40\u0E1B\u0E25\u0E35\u0E48\u0E22\u0E19\u0E41\u0E1B\u0E25\u0E07", normalized: "change" },
+  { native: "\u0E2A\u0E48\u0E07", normalized: "submit" },
+  { native: "\u0E2D\u0E34\u0E19\u0E1E\u0E38\u0E15", normalized: "input" },
+  { native: "\u0E42\u0E2B\u0E25\u0E14", normalized: "load" },
+  { native: "\u0E40\u0E25\u0E37\u0E48\u0E2D\u0E19", normalized: "scroll" },
+  // Additional modifiers
+  { native: "\u0E40\u0E27\u0E25\u0E32", normalized: "when" },
+  { native: "\u0E44\u0E1B\u0E22\u0E31\u0E07", normalized: "to" },
+  { native: "\u0E14\u0E49\u0E27\u0E22", normalized: "with" },
+  { native: "\u0E41\u0E25\u0E30", normalized: "and" }
+];
+var ThaiTokenizer = class extends BaseTokenizer {
+  constructor() {
+    super();
+    this.language = "th";
+    this.direction = "ltr";
+    this.initializeKeywordsFromProfile(thaiProfile, THAI_EXTRAS);
+  }
+  tokenize(input) {
+    const tokens = [];
+    let pos = 0;
+    while (pos < input.length) {
+      if (isWhitespace(input[pos])) {
+        pos++;
+        continue;
+      }
+      if (isSelectorStart(input[pos])) {
+        const modifierToken = this.tryEventModifier(input, pos);
+        if (modifierToken) {
+          tokens.push(modifierToken);
+          pos = modifierToken.position.end;
+          continue;
+        }
+        if (this.tryPropertyAccess(input, pos, tokens)) {
+          pos++;
+          continue;
+        }
+        const selectorToken = this.trySelector(input, pos);
+        if (selectorToken) {
+          tokens.push(selectorToken);
+          pos = selectorToken.position.end;
+          continue;
+        }
+      }
+      if (isQuote(input[pos])) {
+        const stringToken = this.tryString(input, pos);
+        if (stringToken) {
+          tokens.push(stringToken);
+          pos = stringToken.position.end;
+          continue;
+        }
+      }
+      if (isUrlStart(input, pos)) {
+        const urlToken = this.tryUrl(input, pos);
+        if (urlToken) {
+          tokens.push(urlToken);
+          pos = urlToken.position.end;
+          continue;
+        }
+      }
+      if (isDigit(input[pos]) || input[pos] === "-" && pos + 1 < input.length && isDigit(input[pos + 1])) {
+        const numberToken = this.tryNumber(input, pos);
+        if (numberToken) {
+          tokens.push(numberToken);
+          pos = numberToken.position.end;
+          continue;
+        }
+      }
+      if (input[pos] === ":") {
+        const startPos2 = pos;
+        pos++;
+        let varName = "";
+        while (pos < input.length && (isAsciiIdentifierChar(input[pos]) || isThai(input[pos]))) {
+          varName += input[pos];
+          pos++;
+        }
+        if (varName) {
+          tokens.push(
+            createToken(":" + varName, "identifier", createPosition(startPos2, pos), ":" + varName)
+          );
+          continue;
+        }
+        pos = startPos2;
+      }
+      if (isThai(input[pos])) {
+        const startPos2 = pos;
+        const keywordToken = this.tryProfileKeyword(input, pos);
+        if (keywordToken) {
+          tokens.push(keywordToken);
+          pos = keywordToken.position.end;
+          continue;
+        }
+        let word = "";
+        while (pos < input.length && isThai(input[pos])) {
+          if (word.length > 0 && this.isKeywordStart(input, pos)) {
+            break;
+          }
+          word += input[pos];
+          pos++;
+        }
+        if (word) {
+          tokens.push(createToken(word, "identifier", createPosition(startPos2, pos)));
+        }
+        continue;
+      }
+      if (isAsciiIdentifierChar(input[pos])) {
+        const startPos2 = pos;
+        let word = "";
+        while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+          word += input[pos];
+          pos++;
+        }
+        tokens.push(
+          createToken(word, "identifier", createPosition(startPos2, pos), word.toLowerCase())
+        );
+        continue;
+      }
+      const startPos = pos;
+      tokens.push(createToken(input[pos], "operator", createPosition(startPos, pos + 1)));
+      pos++;
+    }
+    return new TokenStreamImpl(tokens, this.language);
+  }
+  classifyToken(value) {
+    if (this.isKeyword(value)) return "keyword";
+    if (value.startsWith(".") || value.startsWith("#") || value.startsWith("[") || value.startsWith("*") || value.startsWith("<"))
+      return "selector";
+    if (value.startsWith(":")) return "identifier";
+    if (value.startsWith('"') || value.startsWith("'")) return "literal";
+    if (/^-?\d/.test(value)) return "literal";
+    return "identifier";
+  }
+};
+var thaiTokenizer = new ThaiTokenizer();
+
+// src/languages/th.ts
+registerLanguage("th", thaiTokenizer, thaiProfile);
+export {
+  thaiProfile,
+  thaiTokenizer
+};
+//# sourceMappingURL=th.js.map