@lokascript/semantic 1.0.0

package/dist/languages/ar.js

@@ -0,0 +1,1622 @@
+// 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/base.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: 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 };
+}
+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);
+}
+function createUnicodeRangeClassifier(ranges) {
+  return (char) => {
+    const code = char.charCodeAt(0);
+    return ranges.some(([start, end]) => code >= start && code <= end);
+  };
+}
+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;
+}
+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();
+  }
+  /**
+   * 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 [normalized2, translation] of Object.entries(profile.keywords)) {
+        keywordMap.set(translation.primary, {
+          native: translation.primary,
+          normalized: translation.normalized || normalized2
+        });
+        if (translation.alternatives) {
+          for (const alt of translation.alternatives) {
+            keywordMap.set(alt, {
+              native: alt,
+              normalized: translation.normalized || normalized2
+            });
+          }
+        }
+      }
+    }
+    if (profile.references) {
+      for (const [normalized2, native] of Object.entries(profile.references)) {
+        keywordMap.set(native, { native, normalized: normalized2 });
+      }
+    }
+    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 normalized2 = this.removeDiacritics(keyword.native);
+      if (normalized2 !== keyword.native && !this.profileKeywordMap.has(normalized2.toLowerCase())) {
+        this.profileKeywordMap.set(normalized2.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/tokenizers/morphology/types.ts
+function noChange(word) {
+  return { stem: word, confidence: 1 };
+}
+function normalized(stem, confidence, metadata) {
+  if (metadata) {
+    return { stem, confidence, metadata };
+  }
+  return { stem, confidence };
+}
+
+// src/tokenizers/morphology/arabic-normalizer.ts
+function isArabic(char) {
+  const code = char.charCodeAt(0);
+  return code >= 1536 && code <= 1791 || // Arabic
+  code >= 1872 && code <= 1919 || // Arabic Supplement
+  code >= 2208 && code <= 2303 || // Arabic Extended-A
+  code >= 64336 && code <= 65023 || // Arabic Presentation Forms-A
+  code >= 65136 && code <= 65279;
+}
+function containsArabic(word) {
+  for (const char of word) {
+    if (isArabic(char)) return true;
+  }
+  return false;
+}
+function removeDiacritics(word) {
+  return word.replace(/[\u064B-\u0652\u0670]/g, "");
+}
+var COMBINED_PREFIXES = [
+  // Conjunction + article combinations (4 chars)
+  { pattern: "\u0648\u0627\u0644", confidencePenalty: 0.15, prefixType: "conjunction" },
+  // wa + al
+  { pattern: "\u0641\u0627\u0644", confidencePenalty: 0.15, prefixType: "conjunction" },
+  // fa + al
+  { pattern: "\u0628\u0627\u0644", confidencePenalty: 0.15, prefixType: "preposition" },
+  // bi + al
+  { pattern: "\u0643\u0627\u0644", confidencePenalty: 0.15, prefixType: "preposition" },
+  // ka + al
+  { pattern: "\u0644\u0644", confidencePenalty: 0.12, prefixType: "preposition" }
+  // li + al (assimilation)
+];
+var SINGLE_PREFIXES = [
+  // Definite article (2 chars) - can leave 2-char stem
+  { pattern: "\u0627\u0644", confidencePenalty: 0.08, prefixType: "article", minRemaining: 2 },
+  // Conjunctions and prepositions (1 char) - need longer stem to be safe
+  { pattern: "\u0648", confidencePenalty: 0.08, prefixType: "conjunction", minRemaining: 3 },
+  // wa- (and)
+  { pattern: "\u0641", confidencePenalty: 0.08, prefixType: "conjunction", minRemaining: 3 },
+  // fa- (then/so)
+  { pattern: "\u0628", confidencePenalty: 0.1, prefixType: "preposition", minRemaining: 3 },
+  // bi- (with/by)
+  { pattern: "\u0644", confidencePenalty: 0.1, prefixType: "preposition", minRemaining: 3 },
+  // li- (to/for)
+  { pattern: "\u0643", confidencePenalty: 0.1, prefixType: "preposition", minRemaining: 3 }
+  // ka- (like/as)
+];
+var VERB_PREFIXES = [
+  { pattern: "\u064A", confidencePenalty: 0.12, prefixType: "verb-marker", minRemaining: 3 },
+  // ya- (he/it)
+  { pattern: "\u062A", confidencePenalty: 0.12, prefixType: "verb-marker", minRemaining: 3 },
+  // ta- (she/you)
+  { pattern: "\u0646", confidencePenalty: 0.12, prefixType: "verb-marker", minRemaining: 3 },
+  // na- (we)
+  { pattern: "\u0623", confidencePenalty: 0.12, prefixType: "verb-marker", minRemaining: 3 },
+  // a- (I)
+  { pattern: "\u0627", confidencePenalty: 0.12, prefixType: "verb-marker", minRemaining: 3 }
+  // a- without hamza
+];
+var SUFFIXES = [
+  // Plural forms
+  { pattern: "\u0648\u0646", confidencePenalty: 0.1, type: "masculine-plural" },
+  { pattern: "\u064A\u0646", confidencePenalty: 0.1, type: "masculine-plural-accusative" },
+  { pattern: "\u0627\u062A", confidencePenalty: 0.1, type: "feminine-plural" },
+  // Dual forms
+  { pattern: "\u0627\u0646", confidencePenalty: 0.1, type: "dual-nominative" },
+  { pattern: "\u064A\u0646", confidencePenalty: 0.1, type: "dual-accusative" },
+  // Pronoun suffixes
+  { pattern: "\u0647\u0627", confidencePenalty: 0.1, type: "pronoun-her" },
+  { pattern: "\u0647\u0645", confidencePenalty: 0.1, type: "pronoun-them" },
+  { pattern: "\u0647\u0646", confidencePenalty: 0.1, type: "pronoun-them-f" },
+  { pattern: "\u0646\u0627", confidencePenalty: 0.1, type: "pronoun-us" },
+  { pattern: "\u0643\u0645", confidencePenalty: 0.1, type: "pronoun-you-pl" },
+  { pattern: "\u0643", confidencePenalty: 0.08, type: "pronoun-you" },
+  { pattern: "\u0647", confidencePenalty: 0.08, type: "pronoun-him" },
+  { pattern: "\u064A", confidencePenalty: 0.08, type: "pronoun-me" },
+  // Feminine marker
+  { pattern: "\u0629", confidencePenalty: 0.08, type: "feminine" }
+];
+var ArabicMorphologicalNormalizer = class {
+  constructor() {
+    this.language = "ar";
+  }
+  /**
+   * Check if a word might be an Arabic word that can be normalized.
+   */
+  isNormalizable(word) {
+    if (!containsArabic(word)) return false;
+    if (word.length < 2) return false;
+    return true;
+  }
+  /**
+   * Normalize an Arabic word by stripping prefixes and suffixes.
+   */
+  normalize(word) {
+    let stem = removeDiacritics(word);
+    let confidence = 1;
+    const removedPrefixes = [];
+    const removedSuffixes = [];
+    for (const rule of COMBINED_PREFIXES) {
+      if (stem.startsWith(rule.pattern)) {
+        const remaining = stem.slice(rule.pattern.length);
+        if (remaining.length >= 2) {
+          stem = remaining;
+          confidence -= rule.confidencePenalty;
+          removedPrefixes.push(rule.pattern);
+          break;
+        }
+      }
+    }
+    if (removedPrefixes.length === 0) {
+      for (const rule of SINGLE_PREFIXES) {
+        if (stem.startsWith(rule.pattern)) {
+          const remaining = stem.slice(rule.pattern.length);
+          const minLen = rule.minRemaining ?? 2;
+          if (remaining.length >= minLen) {
+            stem = remaining;
+            confidence -= rule.confidencePenalty;
+            removedPrefixes.push(rule.pattern);
+            break;
+          }
+        }
+      }
+    }
+    const looksLikeNoun = stem.endsWith("\u0627\u062A") || stem.endsWith("\u0629") || stem.endsWith("\u0648\u0646") || stem.endsWith("\u064A\u0646") || stem.endsWith("\u0647\u0627") || stem.endsWith("\u0647\u0645") || stem.endsWith("\u0647\u0646") || stem.endsWith("\u0646\u0627") || stem.endsWith("\u0643\u0645");
+    if (!looksLikeNoun && (removedPrefixes.length === 0 || removedPrefixes[0] === "\u0648" || removedPrefixes[0] === "\u0641")) {
+      for (const rule of VERB_PREFIXES) {
+        if (stem.startsWith(rule.pattern)) {
+          const remaining = stem.slice(rule.pattern.length);
+          const minLen = rule.minRemaining ?? 3;
+          if (remaining.length >= minLen) {
+            stem = remaining;
+            confidence -= rule.confidencePenalty;
+            removedPrefixes.push(rule.pattern);
+            break;
+          }
+        }
+      }
+    }
+    for (const rule of SUFFIXES) {
+      if (stem.endsWith(rule.pattern)) {
+        const remaining = stem.slice(0, -rule.pattern.length);
+        if (remaining.length >= 2) {
+          stem = remaining;
+          confidence -= rule.confidencePenalty;
+          removedSuffixes.push(rule.pattern);
+        }
+      }
+    }
+    confidence = Math.max(0.5, confidence);
+    if (removedPrefixes.length === 0 && removedSuffixes.length === 0) {
+      return noChange(word);
+    }
+    return normalized(stem, confidence, {
+      removedPrefixes,
+      removedSuffixes
+    });
+  }
+};
+var arabicMorphologicalNormalizer = new ArabicMorphologicalNormalizer();
+
+// src/generators/profiles/arabic.ts
+var arabicProfile = {
+  code: "ar",
+  name: "Arabic",
+  nativeName: "\u0627\u0644\u0639\u0631\u0628\u064A\u0629",
+  direction: "rtl",
+  wordOrder: "VSO",
+  markingStrategy: "preposition",
+  usesSpaces: true,
+  verb: {
+    position: "start",
+    subjectDrop: true
+  },
+  references: {
+    me: "\u0623\u0646\u0627",
+    // "I/me" - first person
+    it: "\u0647\u0648",
+    // "it" (masculine)
+    you: "\u0623\u0646\u062A",
+    // "you"
+    result: "\u0627\u0644\u0646\u062A\u064A\u062C\u0629",
+    event: "\u0627\u0644\u062D\u062F\u062B",
+    target: "\u0627\u0644\u0647\u062F\u0641",
+    body: "\u0627\u0644\u062C\u0633\u0645"
+  },
+  possessive: {
+    marker: "",
+    // No explicit marker - uses possessive pronouns
+    markerPosition: "after-object",
+    usePossessiveAdjectives: true,
+    specialForms: {
+      // Arabic: "value لي" (value for-me) - possessive pronoun follows property
+      me: "\u0644\u064A",
+      // "for me" / "mine"
+      it: "\u0644\u0647",
+      // "for it" / "its"
+      you: "\u0644\u0643"
+      // "for you" / "yours"
+    },
+    keywords: {
+      // "my" / "mine"
+      \u0644\u064A: "me",
+      // "your" / "yours"
+      \u0644\u0643: "you",
+      // "its/his/her"
+      \u0644\u0647: "it",
+      // his/its (masculine)
+      \u0644\u0647\u0627: "it"
+      // her/its (feminine)
+    }
+  },
+  roleMarkers: {
+    destination: { primary: "\u0639\u0644\u0649", alternatives: ["\u0641\u064A", "\u0625\u0644\u0649", "\u0628"], position: "before" },
+    source: { primary: "\u0645\u0646", position: "before" },
+    patient: { primary: "", position: "before" },
+    style: { primary: "\u0628\u0640", alternatives: ["\u0628\u0627\u0633\u062A\u062E\u062F\u0627\u0645"], position: "before" }
+  },
+  keywords: {
+    // Class/Attribute operations
+    toggle: { primary: "\u0628\u062F\u0651\u0644", alternatives: ["\u0628\u062F\u0644", "\u063A\u064A\u0651\u0631", "\u063A\u064A\u0631"], normalized: "toggle" },
+    add: { primary: "\u0623\u0636\u0641", alternatives: ["\u0627\u0636\u0641", "\u0632\u0650\u062F"], normalized: "add" },
+    remove: { primary: "\u0627\u062D\u0630\u0641", alternatives: ["\u0623\u0632\u0644", "\u0627\u0645\u0633\u062D"], normalized: "remove" },
+    // Content operations
+    put: { primary: "\u0636\u0639", alternatives: ["\u0627\u062C\u0639\u0644"], normalized: "put" },
+    append: { primary: "\u0623\u0644\u062D\u0642", normalized: "append" },
+    prepend: { primary: "\u0633\u0628\u0642", normalized: "prepend" },
+    take: { primary: "\u062E\u0630", alternatives: ["\u0627\u062D\u0635\u0644"], normalized: "take" },
+    make: { primary: "\u0627\u0635\u0646\u0639", alternatives: ["\u0623\u0646\u0634\u0626"], normalized: "make" },
+    clone: { primary: "\u0627\u0633\u062A\u0646\u0633\u062E", alternatives: ["\u0627\u0646\u0633\u062E"], normalized: "clone" },
+    swap: { primary: "\u0627\u0633\u062A\u0628\u062F\u0644", alternatives: ["\u062A\u0628\u0627\u062F\u0644"], normalized: "swap" },
+    morph: { primary: "\u062D\u0648\u0651\u0644", alternatives: ["\u063A\u064A\u0651\u0631"], normalized: "morph" },
+    // Variable operations
+    set: { primary: "\u0627\u0636\u0628\u0637", alternatives: ["\u0639\u064A\u0651\u0646", "\u062D\u062F\u062F"], normalized: "set" },
+    get: { primary: "\u0627\u062D\u0635\u0644", alternatives: ["\u062E\u0630"], normalized: "get" },
+    increment: { primary: "\u0632\u0650\u062F", alternatives: ["\u0632\u062F", "\u0627\u0631\u0641\u0639"], normalized: "increment" },
+    decrement: { primary: "\u0623\u0646\u0642\u0635", alternatives: ["\u0627\u0646\u0642\u0635", "\u0642\u0644\u0644"], normalized: "decrement" },
+    log: { primary: "\u0633\u062C\u0644", normalized: "log" },
+    // Visibility
+    show: { primary: "\u0627\u0638\u0647\u0631", alternatives: ["\u0623\u0638\u0647\u0631", "\u0627\u0639\u0631\u0636"], normalized: "show" },
+    hide: { primary: "\u0627\u062E\u0641", alternatives: ["\u0623\u062E\u0641\u0650", "\u0627\u062E\u0641\u064A"], normalized: "hide" },
+    transition: { primary: "\u0627\u0646\u062A\u0642\u0627\u0644", alternatives: ["\u0627\u0646\u062A\u0642\u0644"], normalized: "transition" },
+    // Events
+    on: { primary: "\u0639\u0644\u0649", alternatives: ["\u0639\u0646\u062F", "\u0644\u062F\u0649", "\u062D\u064A\u0646"], normalized: "on" },
+    trigger: { primary: "\u062A\u0634\u063A\u064A\u0644", alternatives: ["\u0623\u0637\u0644\u0642", "\u0641\u0639\u0651\u0644"], normalized: "trigger" },
+    send: { primary: "\u0623\u0631\u0633\u0644", normalized: "send" },
+    // DOM focus
+    focus: { primary: "\u062A\u0631\u0643\u064A\u0632", alternatives: ["\u0631\u0643\u0632"], normalized: "focus" },
+    blur: { primary: "\u0636\u0628\u0627\u0628\u064A\u0629", alternatives: ["\u0634\u0648\u0634"], normalized: "blur" },
+    // Navigation
+    go: { primary: "\u0627\u0630\u0647\u0628", alternatives: ["\u0627\u0646\u062A\u0642\u0644"], normalized: "go" },
+    // Async
+    wait: { primary: "\u0627\u0646\u062A\u0638\u0631", normalized: "wait" },
+    fetch: { primary: "\u0627\u062D\u0636\u0631", alternatives: ["\u062C\u0644\u0628"], normalized: "fetch" },
+    settle: { primary: "\u0627\u0633\u062A\u0642\u0631", normalized: "settle" },
+    // Control flow
+    if: { primary: "\u0625\u0630\u0627", normalized: "if" },
+    when: { primary: "\u0639\u0646\u062F\u0645\u0627", normalized: "when" },
+    where: { primary: "\u0623\u064A\u0646", normalized: "where" },
+    else: { primary: "\u0648\u0625\u0644\u0627", alternatives: ["\u062E\u0644\u0627\u0641 \u0630\u0644\u0643"], normalized: "else" },
+    repeat: { primary: "\u0643\u0631\u0631", normalized: "repeat" },
+    for: { primary: "\u0644\u0643\u0644", normalized: "for" },
+    while: { primary: "\u0628\u064A\u0646\u0645\u0627", normalized: "while" },
+    continue: { primary: "\u0648\u0627\u0635\u0644", normalized: "continue" },
+    halt: { primary: "\u0623\u0648\u0642\u0641", alternatives: ["\u062A\u0648\u0642\u0641"], normalized: "halt" },
+    throw: { primary: "\u0627\u0631\u0645", alternatives: ["\u0627\u0631\u0645\u0650"], normalized: "throw" },
+    call: { primary: "\u0627\u0633\u062A\u062F\u0639", alternatives: ["\u0646\u0627\u062F\u0650"], normalized: "call" },
+    return: { primary: "\u0627\u0631\u062C\u0639", alternatives: ["\u0639\u064F\u062F"], normalized: "return" },
+    then: { primary: "\u062B\u0645", alternatives: ["\u0628\u0639\u062F\u0647\u0627", "\u062B\u0645\u0651"], normalized: "then" },
+    and: { primary: "\u0648\u0623\u064A\u0636\u0627\u064B", alternatives: ["\u0623\u064A\u0636\u0627\u064B"], normalized: "and" },
+    end: { primary: "\u0646\u0647\u0627\u064A\u0629", alternatives: ["\u0627\u0646\u062A\u0647\u0649", "\u0622\u062E\u0631"], normalized: "end" },
+    // Advanced
+    js: { primary: "\u062C\u0627\u0641\u0627\u0633\u0643\u0631\u0628\u062A", alternatives: ["js"], normalized: "js" },
+    async: { primary: "\u0645\u062A\u0632\u0627\u0645\u0646", normalized: "async" },
+    tell: { primary: "\u0623\u062E\u0628\u0631", normalized: "tell" },
+    default: { primary: "\u0627\u0641\u062A\u0631\u0627\u0636\u064A", normalized: "default" },
+    init: { primary: "\u062A\u0647\u064A\u0626\u0629", alternatives: ["\u0628\u062F\u0621"], normalized: "init" },
+    behavior: { primary: "\u0633\u0644\u0648\u0643", normalized: "behavior" },
+    install: { primary: "\u062A\u062B\u0628\u064A\u062A", alternatives: ["\u062B\u0628\u0651\u062A"], normalized: "install" },
+    measure: { primary: "\u0642\u064A\u0627\u0633", alternatives: ["\u0642\u0650\u0633"], normalized: "measure" },
+    // Modifiers
+    into: { primary: "\u0641\u064A", alternatives: ["\u0625\u0644\u0649"], normalized: "into" },
+    before: { primary: "\u0642\u0628\u0644", normalized: "before" },
+    after: { primary: "\u0628\u0639\u062F", normalized: "after" },
+    // Event modifiers (for repeat until event)
+    until: { primary: "\u062D\u062A\u0649", normalized: "until" },
+    event: { primary: "\u062D\u062F\u062B", normalized: "event" },
+    from: { primary: "\u0645\u0646", normalized: "from" }
+  },
+  tokenization: {
+    prefixes: ["\u0627\u0644", "\u0648", "\u0641", "\u0628", "\u0643", "\u0644"]
+  }
+};
+
+// src/tokenizers/arabic.ts
+var isArabic2 = createUnicodeRangeClassifier([
+  [1536, 1791],
+  // Arabic
+  [1872, 1919],
+  // Arabic Supplement
+  [2208, 2303],
+  // Arabic Extended-A
+  [64336, 65023],
+  // Arabic Presentation Forms-A
+  [65136, 65279]
+  // Arabic Presentation Forms-B
+]);
+var ATTACHED_PREFIXES = /* @__PURE__ */ new Set([
+  "\u0628\u0640",
+  // bi- (with, by)
+  "\u0644\u0640",
+  // li- (to, for)
+  "\u0643\u0640",
+  // ka- (like, as)
+  "\u0648\u0640"
+  // wa- (and)
+]);
+var PROCLITICS = /* @__PURE__ */ new Map([
+  // Conjunctions (single character)
+  ["\u0648", { normalized: "and", type: "conjunction" }],
+  // wa - conjunction "and"
+  ["\u0641", { normalized: "then", type: "conjunction" }],
+  // fa - conjunction "then/so"
+  // Attached prefix prepositions
+  ["\u0628", { normalized: "with", type: "preposition" }],
+  // bi- (with, by)
+  ["\u0644", { normalized: "to", type: "preposition" }],
+  // li- (to, for)
+  ["\u0643", { normalized: "like", type: "preposition" }],
+  // ka- (like, as)
+  // Multi-proclitic sequences (conjunction + preposition)
+  ["\u0648\u0644", { normalized: "and-to", type: "conjunction" }],
+  // wa + li-
+  ["\u0648\u0628", { normalized: "and-with", type: "conjunction" }],
+  // wa + bi-
+  ["\u0648\u0643", { normalized: "and-like", type: "conjunction" }],
+  // wa + ka-
+  ["\u0641\u0644", { normalized: "then-to", type: "conjunction" }],
+  // fa + li-
+  ["\u0641\u0628", { normalized: "then-with", type: "conjunction" }],
+  // fa + bi-
+  ["\u0641\u0643", { normalized: "then-like", type: "conjunction" }]
+  // fa + ka-
+]);
+var TEMPORAL_MARKERS = /* @__PURE__ */ new Map([
+  [
+    "\u0639\u0646\u062F\u0645\u0627",
+    {
+      normalized: "on",
+      formality: "formal",
+      confidence: 0.95,
+      description: "when (formal MSA)"
+    }
+  ],
+  [
+    "\u062D\u064A\u0646\u0645\u0627",
+    {
+      normalized: "on",
+      formality: "formal",
+      confidence: 0.93,
+      description: "when/whenever (formal)"
+    }
+  ],
+  [
+    "\u0639\u0646\u062F",
+    {
+      normalized: "on",
+      formality: "neutral",
+      confidence: 0.88,
+      description: "at/when (neutral)"
+    }
+  ],
+  [
+    "\u062D\u064A\u0646",
+    {
+      normalized: "on",
+      formality: "neutral",
+      confidence: 0.85,
+      description: "when/time (neutral)"
+    }
+  ],
+  [
+    "\u0644\u0645\u0651\u0627",
+    {
+      normalized: "on",
+      formality: "dialectal",
+      confidence: 0.7,
+      description: "when (dialectal, with shadda)"
+    }
+  ],
+  [
+    "\u0644\u0645\u0627",
+    {
+      normalized: "on",
+      formality: "dialectal",
+      confidence: 0.68,
+      description: "when (dialectal, no diacritic)"
+    }
+  ],
+  [
+    "\u0644\u062F\u0649",
+    {
+      normalized: "on",
+      formality: "neutral",
+      confidence: 0.82,
+      description: "at/with (temporal)"
+    }
+  ]
+]);
+var PREPOSITIONS = /* @__PURE__ */ new Set([
+  "\u0641\u064A",
+  // fī (in)
+  "\u0639\u0644\u0649",
+  // ʿalā (on)
+  "\u0645\u0646",
+  // min (from)
+  "\u0625\u0644\u0649",
+  // ilā (to)
+  "\u0627\u0644\u0649",
+  // ilā (alternative spelling)
+  // 'عند' removed - it's a temporal marker with metadata
+  "\u0645\u0639",
+  // maʿa (with)
+  "\u0639\u0646",
+  // ʿan (about, from)
+  "\u0642\u0628\u0644",
+  // qabl (before)
+  "\u0628\u0639\u062F",
+  // baʿd (after)
+  "\u0628\u064A\u0646"
+  // bayn (between)
+]);
+var ARABIC_EXTRAS = [
+  // Values/Literals
+  { native: "\u0635\u062D\u064A\u062D", normalized: "true" },
+  { native: "\u062E\u0637\u0623", normalized: "false" },
+  { native: "null", normalized: "null" },
+  { native: "\u0641\u0627\u0631\u063A", normalized: "null" },
+  { native: "\u063A\u064A\u0631 \u0645\u0639\u0631\u0641", normalized: "undefined" },
+  // Positional
+  { native: "\u0627\u0644\u0623\u0648\u0644", normalized: "first" },
+  { native: "\u0623\u0648\u0644", normalized: "first" },
+  { native: "\u0627\u0644\u0623\u062E\u064A\u0631", normalized: "last" },
+  { native: "\u0622\u062E\u0631", normalized: "last" },
+  { native: "\u0627\u0644\u062A\u0627\u0644\u064A", normalized: "next" },
+  { native: "\u0627\u0644\u0633\u0627\u0628\u0642", normalized: "previous" },
+  { native: "\u0627\u0644\u0623\u0642\u0631\u0628", normalized: "closest" },
+  { native: "\u0627\u0644\u0623\u0628", normalized: "parent" },
+  // Events
+  { native: "\u0627\u0644\u0646\u0642\u0631", normalized: "click" },
+  { native: "\u0646\u0642\u0631", normalized: "click" },
+  { native: "\u0627\u0644\u0625\u062F\u062E\u0627\u0644", normalized: "input" },
+  { native: "\u0625\u062F\u062E\u0627\u0644", normalized: "input" },
+  { native: "\u0627\u0644\u062A\u063A\u064A\u064A\u0631", normalized: "change" },
+  { native: "\u062A\u063A\u064A\u064A\u0631", normalized: "change" },
+  { native: "\u0627\u0644\u0625\u0631\u0633\u0627\u0644", normalized: "submit" },
+  { native: "\u0625\u0631\u0633\u0627\u0644", normalized: "submit" },
+  { native: "\u0627\u0644\u062A\u0631\u0643\u064A\u0632", normalized: "focus" },
+  { native: "\u0641\u0642\u062F\u0627\u0646 \u0627\u0644\u062A\u0631\u0643\u064A\u0632", normalized: "blur" },
+  { native: "\u0636\u063A\u0637", normalized: "keydown" },
+  { native: "\u0631\u0641\u0639", normalized: "keyup" },
+  { native: "\u062A\u0645\u0631\u064A\u0631 \u0627\u0644\u0641\u0623\u0631\u0629", normalized: "mouseover" },
+  { native: "\u0645\u063A\u0627\u062F\u0631\u0629 \u0627\u0644\u0641\u0623\u0631\u0629", normalized: "mouseout" },
+  { native: "\u062A\u062D\u0645\u064A\u0644", normalized: "load" },
+  { native: "\u062A\u0645\u0631\u064A\u0631", normalized: "scroll" },
+  // References
+  { native: "\u0623\u0646\u0627", normalized: "me" },
+  { native: "\u0647\u0648", normalized: "it" },
+  { native: "\u0647\u064A", normalized: "it" },
+  { native: "\u0627\u0644\u0646\u062A\u064A\u062C\u0629", normalized: "result" },
+  { native: "\u0627\u0644\u062D\u062F\u062B", normalized: "event" },
+  { native: "\u0627\u0644\u0647\u062F\u0641", normalized: "target" },
+  // Time units
+  { native: "\u062B\u0627\u0646\u064A\u0629", normalized: "s" },
+  { native: "\u062B\u0648\u0627\u0646\u064A", normalized: "s" },
+  { native: "\u0645\u0644\u064A \u062B\u0627\u0646\u064A\u0629", normalized: "ms" },
+  { native: "\u062F\u0642\u064A\u0642\u0629", normalized: "m" },
+  { native: "\u062F\u0642\u0627\u0626\u0642", normalized: "m" },
+  { native: "\u0633\u0627\u0639\u0629", normalized: "h" },
+  { native: "\u0633\u0627\u0639\u0627\u062A", normalized: "h" },
+  // Note: Temporal markers (عندما, حينما, etc.) are in TEMPORAL_MARKERS map
+  // with formality metadata, not in ARABIC_EXTRAS
+  // Additional spelling variants (without diacritics)
+  { native: "\u0628\u062F\u0644", normalized: "toggle" },
+  { native: "\u063A\u064A\u0631", normalized: "toggle" },
+  { native: "\u0627\u0636\u0641", normalized: "add" },
+  { native: "\u0627\u0632\u0644", normalized: "remove" },
+  { native: "\u0627\u0636\u0639", normalized: "put" },
+  { native: "\u064A\u0636\u0639", normalized: "put" },
+  { native: "\u0627\u062C\u0639\u0644", normalized: "put" },
+  { native: "\u0639\u064A\u0646", normalized: "set" },
+  { native: "\u0632\u062F", normalized: "increment" },
+  { native: "\u0627\u0631\u0641\u0639", normalized: "increment" },
+  { native: "\u0627\u0646\u0642\u0635", normalized: "decrement" },
+  { native: "\u0642\u0644\u0644", normalized: "decrement" },
+  { native: "\u0633\u062C\u0644", normalized: "log" },
+  { native: "\u0627\u0638\u0647\u0631", normalized: "show" },
+  { native: "\u0627\u0639\u0631\u0636", normalized: "show" },
+  { native: "\u0627\u062E\u0641", normalized: "hide" },
+  { native: "\u0627\u062E\u0641\u064A", normalized: "hide" },
+  { native: "\u0634\u063A\u0644", normalized: "trigger" },
+  { native: "\u0627\u0631\u0633\u0644", normalized: "send" },
+  { native: "\u0631\u0643\u0632", normalized: "focus" },
+  { native: "\u0634\u0648\u0634", normalized: "blur" },
+  { native: "\u0627\u0630\u0627", normalized: "if" },
+  { native: "\u0644\u0648", normalized: "if" },
+  { native: "\u0648\u0627\u0644\u0627", normalized: "else" },
+  { native: "\u062A\u0648\u0642\u0641", normalized: "halt" },
+  { native: "\u0627\u0646\u0633\u062E", normalized: "clone" },
+  // Control flow helpers
+  { native: "\u0625\u0630\u0646", normalized: "then" },
+  { native: "\u0641\u0625\u0646", normalized: "then" },
+  { native: "\u0646\u0647\u0627\u064A\u0629", normalized: "end" },
+  // Modifiers
+  { native: "\u0642\u0628\u0644", normalized: "before" },
+  { native: "\u0628\u0639\u062F", normalized: "after" }
+];
+var ARABIC_TIME_UNITS = [
+  { pattern: "\u0645\u0644\u064A \u062B\u0627\u0646\u064A\u0629", suffix: "ms", length: 9, caseInsensitive: false },
+  { pattern: "\u0645\u0644\u064A_\u062B\u0627\u0646\u064A\u0629", suffix: "ms", length: 8, caseInsensitive: false },
+  { pattern: "\u062F\u0642\u0627\u0626\u0642", suffix: "m", length: 5, caseInsensitive: false },
+  { pattern: "\u062F\u0642\u064A\u0642\u0629", suffix: "m", length: 5, caseInsensitive: false },
+  { pattern: "\u062B\u0648\u0627\u0646\u064A", suffix: "s", length: 5, caseInsensitive: false },
+  { pattern: "\u062B\u0627\u0646\u064A\u0629", suffix: "s", length: 5, caseInsensitive: false },
+  { pattern: "\u0633\u0627\u0639\u0627\u062A", suffix: "h", length: 5, caseInsensitive: false },
+  { pattern: "\u0633\u0627\u0639\u0629", suffix: "h", length: 4, caseInsensitive: false }
+];
+var ArabicTokenizer = class extends BaseTokenizer {
+  constructor() {
+    super();
+    this.language = "ar";
+    this.direction = "rtl";
+    this.initializeKeywordsFromProfile(arabicProfile, ARABIC_EXTRAS);
+    this.normalizer = new ArabicMorphologicalNormalizer();
+  }
+  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;
+        }
+        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])) {
+        const numberToken = this.extractArabicNumber(input, pos);
+        if (numberToken) {
+          tokens.push(numberToken);
+          pos = numberToken.position.end;
+          continue;
+        }
+      }
+      const varToken = this.tryVariableRef(input, pos);
+      if (varToken) {
+        tokens.push(varToken);
+        pos = varToken.position.end;
+        continue;
+      }
+      const prepToken = this.tryPreposition(input, pos);
+      if (prepToken) {
+        tokens.push(prepToken);
+        pos = prepToken.position.end;
+        continue;
+      }
+      if (isArabic2(input[pos])) {
+        const procliticResult = this.tryProclitic(input, pos);
+        if (procliticResult) {
+          tokens.push(procliticResult.conjunction);
+          pos = procliticResult.conjunction.position.end;
+          continue;
+        }
+        const wordToken = this.extractArabicWord(input, pos);
+        if (wordToken) {
+          tokens.push(wordToken);
+          pos = wordToken.position.end;
+          continue;
+        }
+      }
+      if (isAsciiIdentifierChar(input[pos])) {
+        const asciiToken = this.extractAsciiWord(input, pos);
+        if (asciiToken) {
+          tokens.push(asciiToken);
+          pos = asciiToken.position.end;
+          continue;
+        }
+      }
+      pos++;
+    }
+    return new TokenStreamImpl(tokens, "ar");
+  }
+  classifyToken(token) {
+    if (PREPOSITIONS.has(token)) return "particle";
+    if (this.isKeyword(token)) return "keyword";
+    if (token.startsWith("#") || token.startsWith(".") || token.startsWith("[")) return "selector";
+    if (token.startsWith('"') || token.startsWith("'")) return "literal";
+    if (/^\d/.test(token)) return "literal";
+    return "identifier";
+  }
+  /**
+   * Try to match an Arabic preposition.
+   * Attaches prepositionValue metadata for disambiguation in pattern matching.
+   */
+  tryPreposition(input, pos) {
+    const sortedPreps = Array.from(PREPOSITIONS).sort((a, b) => b.length - a.length);
+    for (const prep of sortedPreps) {
+      if (input.slice(pos, pos + prep.length) === prep) {
+        const nextPos = pos + prep.length;
+        if (nextPos >= input.length || isWhitespace(input[nextPos]) || !isArabic2(input[nextPos])) {
+          const token = createToken(prep, "particle", createPosition(pos, nextPos));
+          return {
+            ...token,
+            metadata: {
+              prepositionValue: prep
+            }
+          };
+        }
+      }
+    }
+    return null;
+  }
+  /**
+   * Try to extract a proclitic (conjunction or preposition) that's attached to the following word.
+   *
+   * Arabic proclitics attach directly to words without space:
+   * - والنقر → و + النقر (and + the-click)
+   * - فالتبديل → ف + التبديل (then + the-toggle)
+   * - بالنقر → ب + النقر (with + the-click)
+   * - ولالنقر → و + ل + النقر (and + to + the-click)
+   *
+   * This enables:
+   * - Polysyndetic coordination: A وB وC
+   * - Attached prepositions: بالنقر (with-the-click)
+   * - Multi-proclitic sequences: ولالنقر (and-to-the-click)
+   *
+   * Returns null if:
+   * - Not a proclitic character/sequence
+   * - Proclitic is standalone (followed by space)
+   * - Remaining word is too short (< 2 chars, to avoid false positives)
+   * - Full word is a recognized keyword (e.g., بدل should NOT be split to ب + دل)
+   *
+   * @see NATIVE_REVIEW_NEEDED.md for implementation rationale
+   */
+  tryProclitic(input, pos) {
+    let wordEnd = pos;
+    while (wordEnd < input.length && (isArabic2(input[wordEnd]) || input[wordEnd] === "\u0640")) {
+      wordEnd++;
+    }
+    const fullWord = input.slice(pos, wordEnd);
+    if (this.lookupKeyword(fullWord)) {
+      return null;
+    }
+    if (TEMPORAL_MARKERS.has(fullWord)) {
+      return null;
+    }
+    if (PREPOSITIONS.has(fullWord)) {
+      return null;
+    }
+    if (pos + 2 <= input.length) {
+      const twoChar = input.slice(pos, pos + 2);
+      const twoCharEntry = PROCLITICS.get(twoChar);
+      if (twoCharEntry) {
+        const nextPos2 = pos + 2;
+        if (nextPos2 < input.length && isArabic2(input[nextPos2])) {
+          let remainingLength2 = 0;
+          let checkPos2 = nextPos2;
+          while (checkPos2 < input.length && isArabic2(input[checkPos2])) {
+            remainingLength2++;
+            checkPos2++;
+          }
+          if (remainingLength2 >= 2) {
+            const singleCharProclitic = PROCLITICS.get(input[pos]);
+            if (singleCharProclitic) {
+              const afterSingleChar = input.slice(pos + 1, wordEnd);
+              if (this.lookupKeyword(afterSingleChar)) {
+              } else {
+                const tokenKind2 = twoCharEntry.type === "conjunction" ? "conjunction" : "particle";
+                return {
+                  conjunction: createToken(
+                    twoChar,
+                    tokenKind2,
+                    createPosition(pos, nextPos2),
+                    twoCharEntry.normalized
+                  )
+                };
+              }
+            } else {
+              const tokenKind2 = twoCharEntry.type === "conjunction" ? "conjunction" : "particle";
+              return {
+                conjunction: createToken(
+                  twoChar,
+                  tokenKind2,
+                  createPosition(pos, nextPos2),
+                  twoCharEntry.normalized
+                )
+              };
+            }
+          }
+        }
+      }
+    }
+    const char = input[pos];
+    const entry = PROCLITICS.get(char);
+    if (!entry) return null;
+    const nextPos = pos + 1;
+    if (nextPos >= input.length || !isArabic2(input[nextPos])) {
+      return null;
+    }
+    let remainingLength = 0;
+    let checkPos = nextPos;
+    while (checkPos < input.length && isArabic2(input[checkPos])) {
+      remainingLength++;
+      checkPos++;
+    }
+    if (remainingLength < 2) {
+      return null;
+    }
+    const tokenKind = entry.type === "conjunction" ? "conjunction" : "particle";
+    return {
+      conjunction: createToken(char, tokenKind, createPosition(pos, nextPos), entry.normalized)
+    };
+  }
+  /**
+   * Extract an Arabic word.
+   * Uses morphological normalization to handle prefix/suffix variations.
+   * Attaches metadata for temporal markers (formality, confidence).
+   */
+  extractArabicWord(input, startPos) {
+    let pos = startPos;
+    let word = "";
+    for (const prefix of ATTACHED_PREFIXES) {
+      const basePrefix = prefix.replace("\u0640", "");
+      if (input.slice(pos, pos + basePrefix.length) === basePrefix) {
+      }
+    }
+    while (pos < input.length && (isArabic2(input[pos]) || input[pos] === "\u0640")) {
+      word += input[pos++];
+    }
+    if (!word) return null;
+    const temporalMarker = TEMPORAL_MARKERS.get(word);
+    if (temporalMarker) {
+      const token = createToken(
+        word,
+        "keyword",
+        createPosition(startPos, pos),
+        temporalMarker.normalized
+      );
+      return {
+        ...token,
+        metadata: {
+          temporalFormality: temporalMarker.formality,
+          temporalConfidence: temporalMarker.confidence
+        }
+      };
+    }
+    const keywordEntry = this.lookupKeyword(word);
+    if (keywordEntry) {
+      return createToken(word, "keyword", createPosition(startPos, pos), keywordEntry.normalized);
+    }
+    if (PREPOSITIONS.has(word)) {
+      const token = createToken(word, "particle", createPosition(startPos, pos));
+      return {
+        ...token,
+        metadata: {
+          prepositionValue: word
+        }
+      };
+    }
+    const morphToken = this.tryMorphKeywordMatch(word, startPos, pos);
+    if (morphToken) return morphToken;
+    return createToken(word, "identifier", createPosition(startPos, pos));
+  }
+  /**
+   * Extract an ASCII word.
+   */
+  extractAsciiWord(input, startPos) {
+    let pos = startPos;
+    let word = "";
+    while (pos < input.length && isAsciiIdentifierChar(input[pos])) {
+      word += input[pos++];
+    }
+    if (!word) return null;
+    return createToken(word, "identifier", createPosition(startPos, pos));
+  }
+  /**
+   * Extract a number, including Arabic time unit suffixes.
+   * Arabic allows space between number and unit.
+   */
+  extractArabicNumber(input, startPos) {
+    return this.tryNumberWithTimeUnits(input, startPos, ARABIC_TIME_UNITS, {
+      allowSign: false,
+      skipWhitespace: true
+    });
+  }
+};
+var arabicTokenizer = new ArabicTokenizer();
+
+// src/languages/ar.ts
+registerLanguage("ar", arabicTokenizer, arabicProfile);
+export {
+  arabicProfile,
+  arabicTokenizer
+};
+//# sourceMappingURL=ar.js.map