thaizip 0.6.0 → 0.7.3

package/dist/index.cjs

@@ -4,8 +4,13 @@
 var TONE_MARKS = /[่-๋็์]/g;
 var ADDRESS_PREFIXES = /^(จังหวัด|อำเภอ|ตำบล|แขวง|เขต|จ\.|อ\.|ต\.|ข\.)/;
 function normalizeThaiAddressText(input) {
-  if (!input) return "";
-  let text = input.trim();
+  if (typeof input !== "string") {
+    throw new TypeError(
+      `[thaizip] normalizeThaiAddressText expected string, got ${input === null ? "null" : typeof input}`
+    );
+  }
+  if (input.length === 0) return "";
+  let text = input.normalize("NFC").trim();
   text = text.replace(ADDRESS_PREFIXES, "").trim();
   text = text.replace(TONE_MARKS, "");
   text = text.toLowerCase();
@@ -30,6 +35,69 @@ function extractTrigramsNormalized(text) {
 }
 
 // src/core/indexer.ts
+function typeErr(table, id, field, expected, value) {
+  return new TypeError(`[thaizip] ${table} ${id}: expected ${expected} for ${field}, got ${typeof value}`);
+}
+function typeName(value) {
+  return value === null ? "null" : typeof value;
+}
+function validateRawData(data) {
+  if (data === null || typeof data !== "object") {
+    throw new TypeError(`[thaizip] validateRawData expected a RawData object, got ${typeName(data)}`);
+  }
+  const { provinces, amphures, tambons } = data;
+  if (!Array.isArray(provinces)) {
+    throw new TypeError(`[thaizip] RawData.provinces must be an array, got ${typeName(provinces)}`);
+  }
+  if (!Array.isArray(amphures)) {
+    throw new TypeError(`[thaizip] RawData.amphures must be an array, got ${typeName(amphures)}`);
+  }
+  if (!Array.isArray(tambons)) {
+    throw new TypeError(`[thaizip] RawData.tambons must be an array, got ${typeName(tambons)}`);
+  }
+  const seenProvinceIds = /* @__PURE__ */ new Set();
+  for (let i = 0; i < provinces.length; i++) {
+    const p = provinces[i];
+    if (p === null || typeof p !== "object") {
+      throw new TypeError(`[thaizip] province[${i}]: expected object, got ${typeName(p)}`);
+    }
+    if (typeof p.id !== "number") throw typeErr("province", p.id, "id", "number", p.id);
+    if (seenProvinceIds.has(p.id)) throw new TypeError(`[thaizip] duplicate province id: ${p.id}`);
+    seenProvinceIds.add(p.id);
+    if (typeof p.name_th !== "string") throw typeErr("province", p.id, "name_th", "string", p.name_th);
+    if (typeof p.name_en !== "string") throw typeErr("province", p.id, "name_en", "string", p.name_en);
+  }
+  const seenAmphureIds = /* @__PURE__ */ new Set();
+  for (let i = 0; i < amphures.length; i++) {
+    const a = amphures[i];
+    if (a === null || typeof a !== "object") {
+      throw new TypeError(`[thaizip] amphure[${i}]: expected object, got ${typeName(a)}`);
+    }
+    if (typeof a.id !== "number") throw typeErr("amphure", a.id, "id", "number", a.id);
+    if (seenAmphureIds.has(a.id)) throw new TypeError(`[thaizip] duplicate amphure id: ${a.id}`);
+    seenAmphureIds.add(a.id);
+    if (typeof a.name_th !== "string") throw typeErr("amphure", a.id, "name_th", "string", a.name_th);
+    if (typeof a.name_en !== "string") throw typeErr("amphure", a.id, "name_en", "string", a.name_en);
+    if (typeof a.province_id !== "number") throw typeErr("amphure", a.id, "province_id", "number", a.province_id);
+  }
+  const seenTambonIds = /* @__PURE__ */ new Set();
+  for (let i = 0; i < tambons.length; i++) {
+    const t = tambons[i];
+    if (t === null || typeof t !== "object") {
+      throw new TypeError(`[thaizip] tambon[${i}]: expected object, got ${typeName(t)}`);
+    }
+    if (typeof t.id !== "number") throw typeErr("tambon", t.id, "id", "number", t.id);
+    if (seenTambonIds.has(t.id)) throw new TypeError(`[thaizip] duplicate tambon id: ${t.id}`);
+    seenTambonIds.add(t.id);
+    if (typeof t.name_th !== "string") throw typeErr("tambon", t.id, "name_th", "string", t.name_th);
+    if (typeof t.name_en !== "string") throw typeErr("tambon", t.id, "name_en", "string", t.name_en);
+    const zipType = typeof t.zip_code;
+    if (zipType !== "string" && zipType !== "number") {
+      throw typeErr("tambon", t.id, "zip_code", "string or number", t.zip_code);
+    }
+    if (typeof t.amphure_id !== "number") throw typeErr("tambon", t.id, "amphure_id", "number", t.amphure_id);
+  }
+}
 function addTrigrams(map, trigrams, idx) {
   for (const trigram of trigrams) {
     let set = map.get(trigram);
@@ -47,6 +115,9 @@ function combinedTrigrams(nameTh, nameEn) {
   return s;
 }
 function buildThaiAddressIndex(data, options) {
+  if (options?.validate !== false) {
+    validateRawData(data);
+  }
   const { provinces, amphures, tambons } = data;
   const provMap = new Map(
     provinces.filter((p) => !p.deleted_at).map((p) => [p.id, p])
@@ -65,6 +136,10 @@ function buildThaiAddressIndex(data, options) {
   const records = [];
   const map = /* @__PURE__ */ new Map();
   const zipIndex = /* @__PURE__ */ new Map();
+  const normTambon = [];
+  const normTambonEn = [];
+  const byProvince = /* @__PURE__ */ new Map();
+  const byAmphure = /* @__PURE__ */ new Map();
   for (const tambon of tambons) {
     if (tambon.deleted_at) continue;
     const amphure = ampMap.get(tambon.amphure_id);
@@ -94,70 +169,340 @@ function buildThaiAddressIndex(data, options) {
     const existing = zipIndex.get(record.zipCode);
     if (existing) existing.push(idx);
     else zipIndex.set(record.zipCode, [idx]);
-    addTrigrams(map, extractTrigrams(record.tambonNameTh), idx);
-    addTrigrams(map, extractTrigrams(record.tambonNameEn), idx);
+    const provList = byProvince.get(province.id);
+    if (provList) provList.push(idx);
+    else byProvince.set(province.id, [idx]);
+    const ampList = byAmphure.get(amphure.id);
+    if (ampList) ampList.push(idx);
+    else byAmphure.set(amphure.id, [idx]);
+    const normTh = normalizeThaiAddressText(record.tambonNameTh);
+    normTambon.push(normTh);
+    addTrigrams(map, extractTrigramsNormalized(normTh), idx);
+    const normEn = normalizeThaiAddressText(record.tambonNameEn);
+    normTambonEn.push(normEn);
+    addTrigrams(map, extractTrigramsNormalized(normEn), idx);
     addTrigrams(map, extractTrigrams(record.zipCode), idx);
     addTrigrams(map, provTrigrams.get(province.id), idx);
     addTrigrams(map, ampTrigrams.get(amphure.id), idx);
   }
-  return { map, records, zipIndex };
+  const sortedZipKeys = [...zipIndex.keys()].sort();
+  const sortedZipPostings = sortedZipKeys.map((z) => zipIndex.get(z));
+  return { map, records, zipIndex, sortedZipKeys, sortedZipPostings, normTambon, normTambonEn, byProvince, byAmphure };
+}
+
+// src/core/romanize.ts
+var ALIASES = /* @__PURE__ */ new Map([
+  // --- Bangkok khet (districts) — concatenated ("no space") forms ---------
+  ["phranakhon", "phra nakhon"],
+  ["nongchok", "nong chok"],
+  ["bangrak", "bang rak"],
+  ["bangkhen", "bang khen"],
+  ["bangkapi", "bang kapi"],
+  ["pathumwan", "pathum wan"],
+  ["patumwan", "pathum wan"],
+  ["pomprap", "pom prap sattru phai"],
+  ["phrakhanong", "phra khanong"],
+  ["minburi", "min buri"],
+  ["latkrabang", "lat krabang"],
+  ["ladkrabang", "lat krabang"],
+  ["yannawa", "yan nawa"],
+  ["phayathai", "phaya thai"],
+  ["thonburi", "thon buri"],
+  ["bangkokyai", "bangkok yai"],
+  ["huaikhwang", "huai khwang"],
+  ["huaykhwang", "huai khwang"],
+  ["huaykwang", "huai khwang"],
+  ["khlongsan", "khlong san"],
+  ["klongsan", "khlong san"],
+  ["talingchan", "taling chan"],
+  ["bangkoknoi", "bangkok noi"],
+  ["bangkhunthian", "bang khun thian"],
+  ["phasicharoen", "phasi charoen"],
+  ["nongkhaem", "nong khaem"],
+  ["ratburana", "rat burana"],
+  ["bangphlat", "bang phlat"],
+  ["dindaeng", "din daeng"],
+  ["buengkum", "bueng kum"],
+  ["bangsue", "bang sue"],
+  ["bangkholaem", "bang kho laem"],
+  ["khlongtoei", "khlong toei"],
+  ["klongtoei", "khlong toei"],
+  ["klongtoey", "khlong toei"],
+  ["khlongtoey", "khlong toei"],
+  ["suanluang", "suan luang"],
+  ["chomthong", "chom thong"],
+  ["donmueang", "don mueang"],
+  ["donmuang", "don mueang"],
+  ["latphrao", "lat phrao"],
+  ["ladprao", "lat phrao"],
+  ["lardprao", "lat phrao"],
+  ["lat prao", "lat phrao"],
+  ["bangkhae", "bang khae"],
+  ["laksi", "lak si"],
+  ["saimai", "sai mai"],
+  ["khannayao", "khan na yao"],
+  ["saphansung", "saphan sung"],
+  ["wangthonglang", "wang thonglang"],
+  ["khlongsamwa", "khlong sam wa"],
+  ["klongsamwa", "khlong sam wa"],
+  ["bangna", "bang na"],
+  ["thawiwatthana", "thawi watthana"],
+  ["thungkhru", "thung khru"],
+  ["bangbon", "bang bon"],
+  // --- Bangkok area / street nicknames with a common misspelling ----------
+  ["sathorn", "sathon"],
+  ["wattana", "watthana"],
+  ["silom", "si lom"],
+  // --- Provinces and other high-traffic areas ------------------------------
+  ["krungthep", "bangkok"],
+  ["korat", "nakhon ratchasima"],
+  ["nakornratchasima", "nakhon ratchasima"],
+  ["chiangmai", "chiang mai"],
+  ["chiangrai", "chiang rai"],
+  ["pathumthani", "pathum thani"],
+  ["patumthani", "pathum thani"],
+  ["samutprakan", "samut prakan"],
+  ["samutsakhon", "samut sakhon"],
+  ["samutsongkhram", "samut songkhram"],
+  ["nakhonpathom", "nakhon pathom"],
+  ["nakhonsawan", "nakhon sawan"],
+  ["nakhonsithammarat", "nakhon si thammarat"],
+  ["prachuapkhirikhan", "prachuap khiri khan"],
+  ["huahin", "hua hin"],
+  ["hatyai", "hat yai"],
+  ["hadyai", "hat yai"],
+  ["had yai", "hat yai"],
+  ["lopburi", "loburi"],
+  ["lop buri", "loburi"],
+  ["buengkan", "buogkan"],
+  ["bungkan", "buogkan"],
+  ["bung kan", "buogkan"],
+  ["udonthani", "udon thani"],
+  ["khonkaen", "khon kaen"],
+  ["suphanburi", "suphan buri"],
+  ["nongkhai", "nong khai"],
+  ["buriram", "buri ram"]
+]);
+var NOISE_WORDS_RE = /\b(?:sub[- ]?district|district|province|changwat|amphoe|amphur)\b/g;
+var WHITESPACE_RE = /\s+/g;
+function cleanup(text) {
+  const stripped = text.replace(NOISE_WORDS_RE, " ").replace(WHITESPACE_RE, " ").trim();
+  return stripped.length > 0 ? stripped : text;
+}
+function applyRomanizationAliases(normalized) {
+  if (!normalized) return normalized;
+  const direct = ALIASES.get(normalized);
+  if (direct !== void 0) return direct;
+  const cleaned = cleanup(normalized);
+  if (cleaned === normalized) return normalized;
+  const aliasedCleaned = ALIASES.get(cleaned);
+  return aliasedCleaned !== void 0 ? aliasedCleaned : cleaned;
 }
 
 // src/core/search.ts
 var ZIP_CODE_RE = /^\d+$/;
+var HAS_LATIN_RE = /[a-z]/i;
+var TH_COLLATOR = new Intl.Collator("th");
+function compareZip(a, b) {
+  if (a < b) return -1;
+  if (a > b) return 1;
+  return 0;
+}
+function sortZipMatches(matches, exactZip) {
+  matches.sort((a, b) => {
+    if (a.zipCode === exactZip && b.zipCode !== exactZip) return -1;
+    if (b.zipCode === exactZip && a.zipCode !== exactZip) return 1;
+    return compareZip(a.zipCode, b.zipCode);
+  });
+  return matches;
+}
+function normalizeLimit(value) {
+  if (Number.isNaN(value) || value === -Infinity) return 0;
+  if (value === Infinity) return value;
+  return Math.max(0, Math.floor(value));
+}
+function lookupByZipCode(index, zip, options) {
+  if (!index || !index.zipIndex || !index.records || typeof zip !== "string") return [];
+  const normalized = zip.trim();
+  if (!ZIP_CODE_RE.test(normalized) || normalized.length < 2) return [];
+  const zipLimit = normalizeLimit(options?.zipLimit ?? Infinity);
+  const matches = [];
+  const keys = index.sortedZipKeys;
+  const postings = index.sortedZipPostings;
+  if (keys !== void 0 && postings !== void 0) {
+    let lo = 0;
+    let hi = keys.length;
+    while (lo < hi) {
+      const mid = lo + hi >> 1;
+      if (keys[mid] < normalized) lo = mid + 1;
+      else hi = mid;
+    }
+    for (let i = lo; i < keys.length && keys[i].startsWith(normalized); i++) {
+      const indices = postings[i];
+      for (let j = 0; j < indices.length; j++) matches.push(index.records[indices[j]]);
+    }
+  } else {
+    for (const [z, indices] of index.zipIndex) {
+      if (z.startsWith(normalized)) {
+        for (const idx of indices) matches.push(index.records[idx]);
+      }
+    }
+    sortZipMatches(matches, normalized);
+  }
+  return matches.slice(0, zipLimit).map((record) => ({ ...record }));
+}
+function rankAgainstName(name, query) {
+  if (name === void 0 || name.length === 0) return 0;
+  if (name === query) return 3;
+  if (name.startsWith(query)) return 2;
+  if (name.includes(query)) return 1;
+  return 0;
+}
+function computeMatchRank(index, idx, query) {
+  const th = rankAgainstName(index.normTambon?.[idx], query);
+  if (th === 3) return 3;
+  const en = rankAgainstName(index.normTambonEn?.[idx], query);
+  return en > th ? en : th;
+}
+var scratchByIndex = /* @__PURE__ */ new WeakMap();
+function getScratch(index) {
+  const n = index.records.length;
+  let s = scratchByIndex.get(index);
+  if (s === void 0 || s.counts.length < n) {
+    s = { counts: new Uint32Array(n), touched: new Int32Array(n) };
+    scratchByIndex.set(index, s);
+  }
+  return s;
+}
 function searchThaiAddress(index, query, options) {
-  const limit = options?.limit ?? 10;
+  const limit = normalizeLimit(options?.limit ?? 10);
   const threshold = options?.threshold ?? 0.4;
-  if (!index || !query) return [];
+  if (!index || !index.map || !index.records || typeof query !== "string") return [];
+  if (query.length > 1e3) return [];
   const normalized = normalizeThaiAddressText(query);
   if (normalized.length === 0) return [];
   if (normalized.length > 300) return [];
   if (ZIP_CODE_RE.test(normalized)) {
     if (normalized.length < 2) return [];
-    const matches = [];
-    for (const [zip, indices] of index.zipIndex) {
-      if (zip.startsWith(normalized)) {
-        for (const idx of indices) matches.push(index.records[idx]);
-      }
-    }
-    matches.sort((a, b) => {
-      if (a.zipCode === normalized && b.zipCode !== normalized) return -1;
-      if (b.zipCode === normalized && a.zipCode !== normalized) return 1;
-      return a.zipCode.localeCompare(b.zipCode, void 0, { numeric: true });
+    return lookupByZipCode(index, normalized, {
+      zipLimit: options?.zipLimit ?? Infinity
     });
-    return matches.slice(0, limit);
   }
-  if (normalized.length < 3 && query.trim().length < 3) return [];
-  const queryTrigrams = extractTrigramsNormalized(normalized);
-  const hits = /* @__PURE__ */ new Map();
+  if (normalized.length < 3) return [];
+  const searchText = options?.romanizationAliases !== false && HAS_LATIN_RE.test(normalized) ? applyRomanizationAliases(normalized) : normalized;
+  const queryTrigrams = extractTrigramsNormalized(searchText);
+  const { counts, touched } = getScratch(index);
+  let touchedLen = 0;
   for (const trigram of queryTrigrams) {
     const candidates = index.map.get(trigram);
     if (!candidates) continue;
     for (const idx of candidates) {
-      hits.set(idx, (hits.get(idx) ?? 0) + 1);
+      if (counts[idx] === 0) touched[touchedLen++] = idx;
+      counts[idx]++;
     }
   }
   const scored = [];
-  for (const [idx, count] of hits) {
-    const score = count / queryTrigrams.size;
+  const totalTrigrams = queryTrigrams.size;
+  for (let i = 0; i < touchedLen; i++) {
+    const idx = touched[i];
+    const score = counts[idx] / totalTrigrams;
+    counts[idx] = 0;
     if (score >= threshold) {
-      scored.push({ idx, score });
+      scored.push({ idx, score, matchRank: computeMatchRank(index, idx, searchText) });
     }
   }
   scored.sort((a, b) => {
     if (b.score !== a.score) return b.score - a.score;
+    return b.matchRank - a.matchRank;
+  });
+  const windowSize = Math.max(limit * 4, 50);
+  const window = scored.slice(0, windowSize);
+  window.sort((a, b) => {
+    if (b.score !== a.score) return b.score - a.score;
+    if (b.matchRank !== a.matchRank) return b.matchRank - a.matchRank;
     const ra = index.records[a.idx];
     const rb = index.records[b.idx];
-    return ra.provinceNameTh.localeCompare(rb.provinceNameTh, "th") || ra.amphureNameTh.localeCompare(rb.amphureNameTh, "th") || ra.tambonNameTh.localeCompare(rb.tambonNameTh, "th");
+    return TH_COLLATOR.compare(ra.provinceNameTh, rb.provinceNameTh) || TH_COLLATOR.compare(ra.amphureNameTh, rb.amphureNameTh) || TH_COLLATOR.compare(ra.tambonNameTh, rb.tambonNameTh);
+  });
+  return window.slice(0, limit).map(({ idx }) => ({ ...index.records[idx] }));
+}
+
+// src/core/enumerate.ts
+var thCollator = new Intl.Collator("th");
+function byNameTh(a, b) {
+  return thCollator.compare(a.nameTh, b.nameTh);
+}
+function listProvinces(index) {
+  if (!index || !index.records) return [];
+  const seen = /* @__PURE__ */ new Map();
+  if (index.byProvince) {
+    for (const [provinceId, recordIndices] of index.byProvince) {
+      if (seen.has(provinceId) || recordIndices.length === 0) continue;
+      const r = index.records[recordIndices[0]];
+      seen.set(provinceId, { id: provinceId, nameTh: r.provinceNameTh, nameEn: r.provinceNameEn });
+    }
+  } else {
+    for (const r of index.records) {
+      if (!seen.has(r.provinceId)) {
+        seen.set(r.provinceId, { id: r.provinceId, nameTh: r.provinceNameTh, nameEn: r.provinceNameEn });
+      }
+    }
+  }
+  return Array.from(seen.values()).sort(byNameTh);
+}
+function listAmphures(index, provinceId) {
+  if (!index || !index.records) return [];
+  const seen = /* @__PURE__ */ new Map();
+  const addFrom = (r) => {
+    if (!seen.has(r.amphureId)) {
+      seen.set(r.amphureId, {
+        id: r.amphureId,
+        nameTh: r.amphureNameTh,
+        nameEn: r.amphureNameEn,
+        provinceId: r.provinceId
+      });
+    }
+  };
+  if (index.byProvince) {
+    const recordIndices = index.byProvince.get(provinceId);
+    if (!recordIndices) return [];
+    for (const idx of recordIndices) addFrom(index.records[idx]);
+  } else {
+    for (const r of index.records) {
+      if (r.provinceId === provinceId) addFrom(r);
+    }
+  }
+  return Array.from(seen.values()).sort(byNameTh);
+}
+function listTambons(index, amphureId) {
+  if (!index || !index.records) return [];
+  const toSummary = (r) => ({
+    id: r.tambonId,
+    nameTh: r.tambonNameTh,
+    nameEn: r.tambonNameEn,
+    amphureId: r.amphureId,
+    zipCode: r.zipCode
   });
-  return scored.slice(0, limit).map(({ idx }) => index.records[idx]);
+  let result;
+  if (index.byAmphure) {
+    const recordIndices = index.byAmphure.get(amphureId);
+    if (!recordIndices) return [];
+    result = recordIndices.map((idx) => toSummary(index.records[idx]));
+  } else {
+    result = index.records.filter((r) => r.amphureId === amphureId).map(toSummary);
+  }
+  return result.sort(byNameTh);
 }
 
 // src/core/formatter.ts
-function formatThaiAddressSuggestion(record) {
+function formatThaiAddressSuggestion(record, options) {
+  const labelTh = `${record.tambonNameTh} > ${record.amphureNameTh} > ${record.provinceNameTh} ${record.zipCode}`;
+  const labelEn = `${record.tambonNameEn} > ${record.amphureNameEn} > ${record.provinceNameEn} ${record.zipCode}`;
   return {
     id: String(record.tambonId),
-    label: `${record.tambonNameTh} > ${record.amphureNameTh} > ${record.provinceNameTh} ${record.zipCode}`,
+    label: options?.locale === "en" ? labelEn : labelTh,
+    labelTh,
+    labelEn,
     tambon: record.tambonNameTh,
     tambonEn: record.tambonNameEn,
     amphure: record.amphureNameTh,
@@ -186,10 +531,16 @@ function resolveThaiAddress(record) {
   };
 }
 
+exports.applyRomanizationAliases = applyRomanizationAliases;
 exports.buildThaiAddressIndex = buildThaiAddressIndex;
 exports.formatThaiAddressSuggestion = formatThaiAddressSuggestion;
+exports.listAmphures = listAmphures;
+exports.listProvinces = listProvinces;
+exports.listTambons = listTambons;
+exports.lookupByZipCode = lookupByZipCode;
 exports.normalizeThaiAddressText = normalizeThaiAddressText;
 exports.resolveThaiAddress = resolveThaiAddress;
 exports.searchThaiAddress = searchThaiAddress;
+exports.validateRawData = validateRawData;
 //# sourceMappingURL=index.cjs.map
 //# sourceMappingURL=index.cjs.map