@mailwoman/corpus 9.3.0 → 10.0.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (991) hide show
  1. package/README.md +16 -0
  2. package/lib/adapters/geonames/adapter.ts +3 -3
  3. package/lib/adapters/geonames/postal/adapter.ts +122 -0
  4. package/lib/adapters/index.ts +39 -32
  5. package/lib/adapters/openaddresses/adapter.ts +4 -7
  6. package/lib/adapters/osm/adapter.ts +278 -0
  7. package/lib/adapters/overture/adapter.ts +22 -7
  8. package/lib/adapters/utils/index.ts +1 -1
  9. package/lib/adapters/wof/admin/json/adapter.ts +239 -0
  10. package/lib/adapters/wof/json-rows.ts +181 -0
  11. package/lib/adapters/wof/postalcode-json/adapter.ts +196 -0
  12. package/lib/au/adapters/gnaf/adapter.ts +152 -0
  13. package/lib/au/adapters/gnaf/assemble.ts +225 -0
  14. package/lib/cz/recipes/pcfirst-preposition.ts +112 -0
  15. package/lib/de/recipes/locale.ts +360 -0
  16. package/lib/fr/adapters/ban/adapter.ts +159 -0
  17. package/lib/fr/recipes/admin-split.ts +243 -0
  18. package/lib/fr/recipes/bare-street.ts +171 -0
  19. package/lib/fr/recipes/fragment.ts +334 -0
  20. package/lib/fr/recipes/lieudit.ts +274 -0
  21. package/lib/fr/recipes/order.ts +246 -0
  22. package/lib/fr/tools/fetch/ban.ts +289 -0
  23. package/lib/index.ts +1 -1
  24. package/lib/international/recipes/locale.ts +598 -0
  25. package/lib/jp/tools/fetch/houjin.ts +141 -0
  26. package/lib/kr/adapters/juso/label-rows.ts +259 -0
  27. package/lib/kr/adapters/localdata/align.ts +299 -0
  28. package/lib/kr/adapters/localdata/key-index.ts +185 -0
  29. package/lib/kr/adapters/localdata/permit-rows.ts +186 -0
  30. package/lib/kr/tools/fetch/juso.ts +172 -0
  31. package/lib/kr/tools/fetch/localdata.ts +206 -0
  32. package/lib/nl/recipes/postcode.ts +117 -0
  33. package/lib/no/recipes/fragment.ts +213 -0
  34. package/lib/no/recipes/street/index.ts +91 -0
  35. package/lib/no/recipes/street/led.ts +135 -0
  36. package/lib/recipes/anchor-absorption.ts +2 -2
  37. package/lib/recipes/bare/country.ts +95 -0
  38. package/lib/recipes/bare/postcode/eval.ts +99 -0
  39. package/lib/recipes/bare/postcode/index.ts +327 -0
  40. package/lib/recipes/boundary-stress.ts +3 -1
  41. package/lib/recipes/country-balanced.ts +3 -2
  42. package/lib/recipes/house-venue.ts +2 -1
  43. package/lib/recipes/index.ts +28 -21
  44. package/lib/recipes/intersection.ts +3 -2
  45. package/lib/recipes/po/box/cedex/au-nz.ts +82 -0
  46. package/lib/recipes/po/box/cedex/ca.ts +58 -0
  47. package/lib/recipes/po/box/cedex/fr.ts +76 -0
  48. package/lib/recipes/po/box/cedex/phrases.ts +108 -0
  49. package/lib/recipes/po/box/cedex/recipe.ts +203 -0
  50. package/lib/recipes/po/box/cedex/sources.ts +189 -0
  51. package/lib/recipes/po/box/cedex/types.ts +35 -0
  52. package/lib/recipes/po/box/cedex/us.ts +69 -0
  53. package/lib/recipes/po/box/cedex/vocabulary.ts +106 -0
  54. package/lib/recipes/po/box/cedex.ts +9 -0
  55. package/lib/recipes/po/box/index.ts +128 -0
  56. package/lib/recipes/scaffold.ts +18 -42
  57. package/lib/recipes/street/affix.ts +807 -0
  58. package/lib/recipes/street/bare.ts +102 -0
  59. package/lib/recipes/street.ts +3 -1
  60. package/lib/recipes/sub/venue/context.ts +149 -0
  61. package/lib/recipes/sub/venue/index.ts +997 -0
  62. package/lib/recipes/sub/venue/sources.ts +739 -0
  63. package/lib/recipes/trailing-region.ts +9 -9
  64. package/lib/recipes/unit.ts +3 -3
  65. package/lib/sg/recipes/register.ts +270 -0
  66. package/lib/sg/tools/fetch/acra.ts +179 -0
  67. package/lib/si/recipes/bare-village.ts +99 -0
  68. package/lib/south-asia/recipes/register.ts +214 -0
  69. package/lib/synthesizers/anchor-absorption.ts +1 -1
  70. package/lib/synthesizers/locale.ts +261 -0
  71. package/lib/synthesizers/street.ts +1 -1
  72. package/lib/synthesizers/utils.ts +2 -2
  73. package/lib/tools/audit.ts +3 -2
  74. package/lib/tools/corpus-stats.ts +3 -3
  75. package/lib/tools/fetch/download/index.ts +10 -0
  76. package/lib/tools/fetch/download/manifest.ts +102 -0
  77. package/lib/tools/fetch/download/network.ts +226 -0
  78. package/lib/tools/fetch/download/retry.ts +173 -0
  79. package/lib/tools/fetch/geonames/dump.ts +257 -0
  80. package/lib/tools/fetch/geonames/postal.ts +180 -0
  81. package/lib/tools/fetch/index.ts +47 -20
  82. package/lib/tools/fetch/openaddresses.ts +2 -2
  83. package/lib/tools/fetch/ourairports.ts +2 -2
  84. package/lib/tools/fetch/ppd.ts +1 -3
  85. package/lib/tools/fetch/wikidata-subvenue.ts +2 -2
  86. package/lib/tools/golden/promote.ts +340 -0
  87. package/lib/tools/golden/relabel-street/directory.ts +340 -0
  88. package/lib/tools/golden/relabel-street/index.ts +9 -0
  89. package/lib/tools/golden/relabel-street/row.ts +361 -0
  90. package/lib/tools/index.ts +10 -10
  91. package/lib/tools/ingest-csv.ts +5 -7
  92. package/lib/tools/lint/slice/vocab.ts +376 -0
  93. package/lib/tools/overlay-manifest.ts +92 -25
  94. package/lib/tools/overture-subvenue.ts +1 -1
  95. package/lib/tools/postcode-triples.ts +138 -14
  96. package/lib/tools/sub/venue/harvest.ts +234 -0
  97. package/lib/tools/sub/venue/head-nouns.ts +245 -0
  98. package/lib/tools/sub/venue/lexicon.ts +429 -0
  99. package/lib/tools/sub/venue/surfaces.ts +95 -0
  100. package/lib/tools/sub/venue/table.ts +281 -0
  101. package/lib/tools/sub/venue/wikidata.ts +85 -0
  102. package/lib/tw/tools/fetch/gcis.ts +180 -0
  103. package/lib/types.ts +1 -1
  104. package/lib/us/adapters/fcc-bdc/adapter.ts +173 -0
  105. package/lib/us/adapters/state/hi-schools/adapter.ts +160 -0
  106. package/lib/us/adapters/state/ia-contractors/adapter.ts +136 -0
  107. package/lib/us/adapters/state/ny-notaries/adapter.ts +133 -0
  108. package/lib/us/adapters/state/tx-notaries/adapter.ts +149 -0
  109. package/lib/us/adapters/tiger/adapter.ts +250 -0
  110. package/lib/us/adapters/usgov/hrsa-fqhc/adapter.ts +156 -0
  111. package/lib/us/adapters/usgov/imls-pls/adapter.ts +134 -0
  112. package/lib/us/adapters/usgov/irs-bmf/adapter.ts +145 -0
  113. package/lib/us/adapters/usgov/nad/adapter.ts +324 -0
  114. package/lib/us/adapters/usgov/nppes/adapter.ts +143 -0
  115. package/lib/us/adapters/usgov/samhsa-treatment-locator/adapter.ts +182 -0
  116. package/lib/us/tools/fetch/hrsa.ts +59 -0
  117. package/lib/us/tools/fetch/imls-pls.ts +149 -0
  118. package/lib/us/tools/fetch/nad.ts +341 -0
  119. package/lib/us/tools/fetch/nppes.ts +165 -0
  120. package/lib/us/tools/fetch/state/hi-schools.ts +161 -0
  121. package/lib/us/tools/fetch/state/sources.ts +146 -0
  122. package/lib/us/tools/fetch/tiger-full.ts +308 -0
  123. package/lib/utils/align.ts +1 -1
  124. package/lib/utils/golden.ts +12 -8
  125. package/lib/utils/wof-json.ts +7 -13
  126. package/out/adapters/geonames/adapter.d.ts +1 -1
  127. package/out/adapters/geonames/adapter.js +3 -3
  128. package/out/adapters/geonames/adapter.js.map +1 -1
  129. package/out/adapters/geonames/postal/adapter.d.ts.map +1 -0
  130. package/out/adapters/geonames/postal/adapter.js +107 -0
  131. package/out/adapters/geonames/postal/adapter.js.map +1 -0
  132. package/out/adapters/index.d.ts +16 -16
  133. package/out/adapters/index.d.ts.map +1 -1
  134. package/out/adapters/index.js +34 -32
  135. package/out/adapters/index.js.map +1 -1
  136. package/out/adapters/openaddresses/adapter.d.ts.map +1 -1
  137. package/out/adapters/openaddresses/adapter.js +4 -6
  138. package/out/adapters/openaddresses/adapter.js.map +1 -1
  139. package/out/adapters/osm/adapter.d.ts +95 -0
  140. package/out/adapters/osm/adapter.d.ts.map +1 -0
  141. package/out/adapters/osm/adapter.js +211 -0
  142. package/out/adapters/osm/adapter.js.map +1 -0
  143. package/out/adapters/overture/adapter.d.ts +10 -0
  144. package/out/adapters/overture/adapter.d.ts.map +1 -1
  145. package/out/adapters/overture/adapter.js +20 -7
  146. package/out/adapters/overture/adapter.js.map +1 -1
  147. package/out/adapters/synth-po-box/adapter.js.map +1 -1
  148. package/out/adapters/utils/index.d.ts +1 -1
  149. package/out/adapters/utils/index.d.ts.map +1 -1
  150. package/out/adapters/utils/index.js.map +1 -1
  151. package/out/adapters/wof/admin/json/adapter.d.ts +79 -0
  152. package/out/adapters/wof/admin/json/adapter.d.ts.map +1 -0
  153. package/out/adapters/wof/admin/json/adapter.js +204 -0
  154. package/out/adapters/wof/admin/json/adapter.js.map +1 -0
  155. package/out/adapters/wof/json-rows.d.ts +73 -0
  156. package/out/adapters/wof/json-rows.d.ts.map +1 -0
  157. package/out/adapters/wof/json-rows.js +115 -0
  158. package/out/adapters/wof/json-rows.js.map +1 -0
  159. package/out/adapters/wof/postalcode-json/adapter.d.ts +65 -0
  160. package/out/adapters/wof/postalcode-json/adapter.d.ts.map +1 -0
  161. package/out/adapters/wof/postalcode-json/adapter.js +167 -0
  162. package/out/adapters/wof/postalcode-json/adapter.js.map +1 -0
  163. package/out/au/adapters/gnaf/adapter.d.ts.map +1 -0
  164. package/out/au/adapters/gnaf/adapter.js +127 -0
  165. package/out/au/adapters/gnaf/adapter.js.map +1 -0
  166. package/out/au/adapters/gnaf/assemble.d.ts +65 -0
  167. package/out/au/adapters/gnaf/assemble.d.ts.map +1 -0
  168. package/out/au/adapters/gnaf/assemble.js +154 -0
  169. package/out/au/adapters/gnaf/assemble.js.map +1 -0
  170. package/out/cn/recipes/organizational-units.d.ts +36 -0
  171. package/out/cn/recipes/organizational-units.d.ts.map +1 -0
  172. package/out/cn/recipes/organizational-units.js +159 -0
  173. package/out/cn/recipes/organizational-units.js.map +1 -0
  174. package/out/cz/recipes/pcfirst-preposition.d.ts +22 -0
  175. package/out/cz/recipes/pcfirst-preposition.d.ts.map +1 -0
  176. package/out/cz/recipes/pcfirst-preposition.js +101 -0
  177. package/out/cz/recipes/pcfirst-preposition.js.map +1 -0
  178. package/out/de/recipes/locale.d.ts +44 -0
  179. package/out/de/recipes/locale.d.ts.map +1 -0
  180. package/out/de/recipes/locale.js +289 -0
  181. package/out/de/recipes/locale.js.map +1 -0
  182. package/out/env.d.ts +1 -1
  183. package/out/fr/adapters/ban/adapter.d.ts.map +1 -0
  184. package/out/fr/adapters/ban/adapter.js +123 -0
  185. package/out/fr/adapters/ban/adapter.js.map +1 -0
  186. package/out/fr/adapters/ban/street-decompose.d.ts.map +1 -0
  187. package/out/fr/adapters/ban/street-decompose.js.map +1 -0
  188. package/out/fr/recipes/admin-split.d.ts +38 -0
  189. package/out/fr/recipes/admin-split.d.ts.map +1 -0
  190. package/out/fr/recipes/admin-split.js +192 -0
  191. package/out/fr/recipes/admin-split.js.map +1 -0
  192. package/out/fr/recipes/bare-street.d.ts +28 -0
  193. package/out/fr/recipes/bare-street.d.ts.map +1 -0
  194. package/out/fr/recipes/bare-street.js +144 -0
  195. package/out/fr/recipes/bare-street.js.map +1 -0
  196. package/out/fr/recipes/fragment.d.ts +77 -0
  197. package/out/fr/recipes/fragment.d.ts.map +1 -0
  198. package/out/fr/recipes/fragment.js +277 -0
  199. package/out/fr/recipes/fragment.js.map +1 -0
  200. package/out/fr/recipes/lieudit.d.ts +32 -0
  201. package/out/fr/recipes/lieudit.d.ts.map +1 -0
  202. package/out/fr/recipes/lieudit.js +211 -0
  203. package/out/fr/recipes/lieudit.js.map +1 -0
  204. package/out/fr/recipes/order.d.ts +38 -0
  205. package/out/fr/recipes/order.d.ts.map +1 -0
  206. package/out/fr/recipes/order.js +198 -0
  207. package/out/fr/recipes/order.js.map +1 -0
  208. package/out/fr/tools/fetch/ban.d.ts +31 -0
  209. package/out/fr/tools/fetch/ban.d.ts.map +1 -0
  210. package/out/fr/tools/fetch/ban.js +239 -0
  211. package/out/fr/tools/fetch/ban.js.map +1 -0
  212. package/out/index.d.ts +1 -1
  213. package/out/index.d.ts.map +1 -1
  214. package/out/index.js +1 -1
  215. package/out/index.js.map +1 -1
  216. package/out/international/recipes/locale.d.ts +136 -0
  217. package/out/international/recipes/locale.d.ts.map +1 -0
  218. package/out/international/recipes/locale.js +458 -0
  219. package/out/international/recipes/locale.js.map +1 -0
  220. package/out/jp/adapters/wof-admin/adapter.d.ts.map +1 -0
  221. package/out/jp/adapters/wof-admin/adapter.js.map +1 -0
  222. package/out/jp/tools/fetch/houjin.d.ts +24 -0
  223. package/out/jp/tools/fetch/houjin.d.ts.map +1 -0
  224. package/out/jp/tools/fetch/houjin.js +102 -0
  225. package/out/jp/tools/fetch/houjin.js.map +1 -0
  226. package/out/kr/adapters/juso/label-rows.d.ts +76 -0
  227. package/out/kr/adapters/juso/label-rows.d.ts.map +1 -0
  228. package/out/kr/adapters/juso/label-rows.js +194 -0
  229. package/out/kr/adapters/juso/label-rows.js.map +1 -0
  230. package/out/kr/adapters/localdata/align.d.ts +43 -0
  231. package/out/kr/adapters/localdata/align.d.ts.map +1 -0
  232. package/out/kr/adapters/localdata/align.js +223 -0
  233. package/out/kr/adapters/localdata/align.js.map +1 -0
  234. package/out/kr/adapters/localdata/key-index.d.ts +71 -0
  235. package/out/kr/adapters/localdata/key-index.d.ts.map +1 -0
  236. package/out/kr/adapters/localdata/key-index.js +133 -0
  237. package/out/kr/adapters/localdata/key-index.js.map +1 -0
  238. package/out/kr/adapters/localdata/permit-rows.d.ts +66 -0
  239. package/out/kr/adapters/localdata/permit-rows.d.ts.map +1 -0
  240. package/out/kr/adapters/localdata/permit-rows.js +131 -0
  241. package/out/kr/adapters/localdata/permit-rows.js.map +1 -0
  242. package/out/kr/tools/fetch/juso.d.ts +30 -0
  243. package/out/kr/tools/fetch/juso.d.ts.map +1 -0
  244. package/out/kr/tools/fetch/juso.js +120 -0
  245. package/out/kr/tools/fetch/juso.js.map +1 -0
  246. package/out/kr/tools/fetch/localdata.d.ts +30 -0
  247. package/out/kr/tools/fetch/localdata.d.ts.map +1 -0
  248. package/out/kr/tools/fetch/localdata.js +139 -0
  249. package/out/kr/tools/fetch/localdata.js.map +1 -0
  250. package/out/nl/recipes/postcode.d.ts +26 -0
  251. package/out/nl/recipes/postcode.d.ts.map +1 -0
  252. package/out/nl/recipes/postcode.js +103 -0
  253. package/out/nl/recipes/postcode.js.map +1 -0
  254. package/out/no/recipes/fragment.d.ts +43 -0
  255. package/out/no/recipes/fragment.d.ts.map +1 -0
  256. package/out/no/recipes/fragment.js +183 -0
  257. package/out/no/recipes/fragment.js.map +1 -0
  258. package/out/no/recipes/street/index.d.ts +18 -0
  259. package/out/no/recipes/street/index.d.ts.map +1 -0
  260. package/out/no/recipes/street/index.js +72 -0
  261. package/out/no/recipes/street/index.js.map +1 -0
  262. package/out/no/recipes/street/led.d.ts +24 -0
  263. package/out/no/recipes/street/led.d.ts.map +1 -0
  264. package/out/no/recipes/street/led.js +114 -0
  265. package/out/no/recipes/street/led.js.map +1 -0
  266. package/out/recipes/anchor-absorption.d.ts.map +1 -1
  267. package/out/recipes/anchor-absorption.js +2 -1
  268. package/out/recipes/anchor-absorption.js.map +1 -1
  269. package/out/recipes/bare/country.d.ts +25 -0
  270. package/out/recipes/bare/country.d.ts.map +1 -0
  271. package/out/recipes/bare/country.js +88 -0
  272. package/out/recipes/bare/country.js.map +1 -0
  273. package/out/recipes/bare/postcode/eval.d.ts +29 -0
  274. package/out/recipes/bare/postcode/eval.d.ts.map +1 -0
  275. package/out/recipes/bare/postcode/eval.js +85 -0
  276. package/out/recipes/bare/postcode/eval.js.map +1 -0
  277. package/out/recipes/bare/postcode/index.d.ts +62 -0
  278. package/out/recipes/bare/postcode/index.d.ts.map +1 -0
  279. package/out/recipes/bare/postcode/index.js +267 -0
  280. package/out/recipes/bare/postcode/index.js.map +1 -0
  281. package/out/recipes/boundary-stress.d.ts.map +1 -1
  282. package/out/recipes/boundary-stress.js +2 -1
  283. package/out/recipes/boundary-stress.js.map +1 -1
  284. package/out/recipes/country-balanced.d.ts.map +1 -1
  285. package/out/recipes/country-balanced.js +2 -1
  286. package/out/recipes/country-balanced.js.map +1 -1
  287. package/out/recipes/house-venue.d.ts.map +1 -1
  288. package/out/recipes/house-venue.js +2 -1
  289. package/out/recipes/house-venue.js.map +1 -1
  290. package/out/recipes/index.d.ts.map +1 -1
  291. package/out/recipes/index.js +28 -21
  292. package/out/recipes/index.js.map +1 -1
  293. package/out/recipes/intersection.d.ts.map +1 -1
  294. package/out/recipes/intersection.js +2 -1
  295. package/out/recipes/intersection.js.map +1 -1
  296. package/out/recipes/po/box/cedex/au-nz.d.ts +7 -0
  297. package/out/recipes/po/box/cedex/au-nz.d.ts.map +1 -0
  298. package/out/recipes/po/box/cedex/au-nz.js +56 -0
  299. package/out/recipes/po/box/cedex/au-nz.js.map +1 -0
  300. package/out/recipes/po/box/cedex/ca.d.ts +7 -0
  301. package/out/recipes/po/box/cedex/ca.d.ts.map +1 -0
  302. package/out/recipes/po/box/cedex/ca.js +38 -0
  303. package/out/recipes/po/box/cedex/ca.js.map +1 -0
  304. package/out/recipes/po/box/cedex/fr.d.ts +7 -0
  305. package/out/recipes/po/box/cedex/fr.d.ts.map +1 -0
  306. package/out/recipes/po/box/cedex/fr.js +57 -0
  307. package/out/recipes/po/box/cedex/fr.js.map +1 -0
  308. package/out/recipes/po/box/cedex/phrases.d.ts +8 -0
  309. package/out/recipes/po/box/cedex/phrases.d.ts.map +1 -0
  310. package/out/recipes/po/box/cedex/phrases.js +76 -0
  311. package/out/recipes/po/box/cedex/phrases.js.map +1 -0
  312. package/out/recipes/po/box/cedex/recipe.d.ts +9 -0
  313. package/out/recipes/po/box/cedex/recipe.d.ts.map +1 -0
  314. package/out/recipes/po/box/cedex/recipe.js +153 -0
  315. package/out/recipes/po/box/cedex/recipe.js.map +1 -0
  316. package/out/recipes/po/box/cedex/sources.d.ts +86 -0
  317. package/out/recipes/po/box/cedex/sources.d.ts.map +1 -0
  318. package/out/recipes/po/box/cedex/sources.js +143 -0
  319. package/out/recipes/po/box/cedex/sources.js.map +1 -0
  320. package/out/recipes/po/box/cedex/types.d.ts +31 -0
  321. package/out/recipes/po/box/cedex/types.d.ts.map +1 -0
  322. package/out/recipes/po/box/cedex/types.js +5 -0
  323. package/out/recipes/po/box/cedex/types.js.map +1 -0
  324. package/out/recipes/po/box/cedex/us.d.ts +7 -0
  325. package/out/recipes/po/box/cedex/us.d.ts.map +1 -0
  326. package/out/recipes/po/box/cedex/us.js +51 -0
  327. package/out/recipes/po/box/cedex/us.js.map +1 -0
  328. package/out/recipes/po/box/cedex/vocabulary.d.ts +68 -0
  329. package/out/recipes/po/box/cedex/vocabulary.d.ts.map +1 -0
  330. package/out/recipes/po/box/cedex/vocabulary.js +90 -0
  331. package/out/recipes/po/box/cedex/vocabulary.js.map +1 -0
  332. package/out/recipes/po/box/cedex.d.ts +9 -0
  333. package/out/recipes/po/box/cedex.d.ts.map +1 -0
  334. package/out/recipes/po/box/cedex.js +9 -0
  335. package/out/recipes/po/box/cedex.js.map +1 -0
  336. package/out/recipes/po/box/index.d.ts +17 -0
  337. package/out/recipes/po/box/index.d.ts.map +1 -0
  338. package/out/recipes/po/box/index.js +105 -0
  339. package/out/recipes/po/box/index.js.map +1 -0
  340. package/out/recipes/scaffold.d.ts +20 -1
  341. package/out/recipes/scaffold.d.ts.map +1 -1
  342. package/out/recipes/scaffold.js +18 -49
  343. package/out/recipes/scaffold.js.map +1 -1
  344. package/out/recipes/street/affix.d.ts +89 -0
  345. package/out/recipes/street/affix.d.ts.map +1 -0
  346. package/out/recipes/street/affix.js +565 -0
  347. package/out/recipes/street/affix.js.map +1 -0
  348. package/out/recipes/street/bare.d.ts +25 -0
  349. package/out/recipes/street/bare.d.ts.map +1 -0
  350. package/out/recipes/street/bare.js +87 -0
  351. package/out/recipes/street/bare.js.map +1 -0
  352. package/out/recipes/street.d.ts.map +1 -1
  353. package/out/recipes/street.js +2 -1
  354. package/out/recipes/street.js.map +1 -1
  355. package/out/recipes/sub/venue/context.d.ts +29 -0
  356. package/out/recipes/sub/venue/context.d.ts.map +1 -0
  357. package/out/recipes/sub/venue/context.js +106 -0
  358. package/out/recipes/sub/venue/context.js.map +1 -0
  359. package/out/recipes/sub/venue/index.d.ts +270 -0
  360. package/out/recipes/sub/venue/index.d.ts.map +1 -0
  361. package/out/recipes/sub/venue/index.js +728 -0
  362. package/out/recipes/sub/venue/index.js.map +1 -0
  363. package/out/recipes/sub/venue/sources.d.ts +209 -0
  364. package/out/recipes/sub/venue/sources.d.ts.map +1 -0
  365. package/out/recipes/sub/venue/sources.js +555 -0
  366. package/out/recipes/sub/venue/sources.js.map +1 -0
  367. package/out/recipes/trailing-region.d.ts.map +1 -1
  368. package/out/recipes/trailing-region.js +8 -2
  369. package/out/recipes/trailing-region.js.map +1 -1
  370. package/out/recipes/unit.d.ts.map +1 -1
  371. package/out/recipes/unit.js +2 -1
  372. package/out/recipes/unit.js.map +1 -1
  373. package/out/sg/recipes/register.d.ts +69 -0
  374. package/out/sg/recipes/register.d.ts.map +1 -0
  375. package/out/sg/recipes/register.js +211 -0
  376. package/out/sg/recipes/register.js.map +1 -0
  377. package/out/sg/tools/fetch/acra.d.ts +23 -0
  378. package/out/sg/tools/fetch/acra.d.ts.map +1 -0
  379. package/out/sg/tools/fetch/acra.js +126 -0
  380. package/out/sg/tools/fetch/acra.js.map +1 -0
  381. package/out/si/recipes/bare-village.d.ts +32 -0
  382. package/out/si/recipes/bare-village.d.ts.map +1 -0
  383. package/out/si/recipes/bare-village.js +89 -0
  384. package/out/si/recipes/bare-village.js.map +1 -0
  385. package/out/south-asia/recipes/register.d.ts +51 -0
  386. package/out/south-asia/recipes/register.d.ts.map +1 -0
  387. package/out/south-asia/recipes/register.js +154 -0
  388. package/out/south-asia/recipes/register.js.map +1 -0
  389. package/out/synthesizers/anchor-absorption.d.ts +1 -1
  390. package/out/synthesizers/anchor-absorption.d.ts.map +1 -1
  391. package/out/synthesizers/anchor-absorption.js.map +1 -1
  392. package/out/synthesizers/boundary-stress.js.map +1 -1
  393. package/out/synthesizers/house-venue.js.map +1 -1
  394. package/out/synthesizers/intersection.js.map +1 -1
  395. package/out/synthesizers/locale.d.ts +116 -0
  396. package/out/synthesizers/locale.d.ts.map +1 -0
  397. package/out/synthesizers/locale.js +150 -0
  398. package/out/synthesizers/locale.js.map +1 -0
  399. package/out/synthesizers/no-street.d.ts.map +1 -1
  400. package/out/synthesizers/no-street.js.map +1 -1
  401. package/out/synthesizers/po-box.js.map +1 -1
  402. package/out/synthesizers/street.d.ts.map +1 -1
  403. package/out/synthesizers/street.js +1 -1
  404. package/out/synthesizers/street.js.map +1 -1
  405. package/out/synthesizers/utils.js +1 -1
  406. package/out/synthesizers/utils.js.map +1 -1
  407. package/out/test-kit/corpus-recipe.d.ts.map +1 -1
  408. package/out/test-kit/corpus-recipe.js.map +1 -1
  409. package/out/tools/audit.d.ts.map +1 -1
  410. package/out/tools/audit.js +3 -2
  411. package/out/tools/audit.js.map +1 -1
  412. package/out/tools/corpus-stats.js +3 -3
  413. package/out/tools/corpus-stats.js.map +1 -1
  414. package/out/tools/fetch/download/index.d.ts +10 -0
  415. package/out/tools/fetch/download/index.d.ts.map +1 -0
  416. package/out/tools/fetch/download/index.js +10 -0
  417. package/out/tools/fetch/download/index.js.map +1 -0
  418. package/out/tools/fetch/download/manifest.d.ts +47 -0
  419. package/out/tools/fetch/download/manifest.d.ts.map +1 -0
  420. package/out/tools/fetch/download/manifest.js +74 -0
  421. package/out/tools/fetch/download/manifest.js.map +1 -0
  422. package/out/tools/fetch/download/network.d.ts +105 -0
  423. package/out/tools/fetch/download/network.d.ts.map +1 -0
  424. package/out/tools/fetch/download/network.js +130 -0
  425. package/out/tools/fetch/download/network.js.map +1 -0
  426. package/out/tools/fetch/download/retry.d.ts +45 -0
  427. package/out/tools/fetch/download/retry.d.ts.map +1 -0
  428. package/out/tools/fetch/download/retry.js +136 -0
  429. package/out/tools/fetch/download/retry.js.map +1 -0
  430. package/out/tools/fetch/geonames/dump.d.ts +91 -0
  431. package/out/tools/fetch/geonames/dump.d.ts.map +1 -0
  432. package/out/tools/fetch/geonames/dump.js +171 -0
  433. package/out/tools/fetch/geonames/dump.js.map +1 -0
  434. package/out/tools/fetch/geonames/postal.d.ts +76 -0
  435. package/out/tools/fetch/geonames/postal.d.ts.map +1 -0
  436. package/out/tools/fetch/geonames/postal.js +126 -0
  437. package/out/tools/fetch/geonames/postal.js.map +1 -0
  438. package/out/tools/fetch/index.d.ts +47 -20
  439. package/out/tools/fetch/index.d.ts.map +1 -1
  440. package/out/tools/fetch/index.js +47 -20
  441. package/out/tools/fetch/index.js.map +1 -1
  442. package/out/tools/fetch/openaddresses.d.ts +1 -1
  443. package/out/tools/fetch/openaddresses.d.ts.map +1 -1
  444. package/out/tools/fetch/openaddresses.js +1 -1
  445. package/out/tools/fetch/openaddresses.js.map +1 -1
  446. package/out/tools/fetch/ourairports.d.ts +1 -1
  447. package/out/tools/fetch/ourairports.d.ts.map +1 -1
  448. package/out/tools/fetch/ourairports.js +1 -1
  449. package/out/tools/fetch/ourairports.js.map +1 -1
  450. package/out/tools/fetch/ppd.js +1 -3
  451. package/out/tools/fetch/ppd.js.map +1 -1
  452. package/out/tools/fetch/wikidata-subvenue.d.ts +1 -1
  453. package/out/tools/fetch/wikidata-subvenue.d.ts.map +1 -1
  454. package/out/tools/fetch/wikidata-subvenue.js +1 -1
  455. package/out/tools/fetch/wikidata-subvenue.js.map +1 -1
  456. package/out/tools/golden/expand.d.ts +90 -0
  457. package/out/tools/golden/expand.d.ts.map +1 -0
  458. package/out/tools/golden/expand.js +406 -0
  459. package/out/tools/golden/expand.js.map +1 -0
  460. package/out/tools/golden/promote.d.ts +72 -0
  461. package/out/tools/golden/promote.d.ts.map +1 -0
  462. package/out/tools/golden/promote.js +243 -0
  463. package/out/tools/golden/promote.js.map +1 -0
  464. package/out/tools/golden/relabel-street/directory.d.ts +83 -0
  465. package/out/tools/golden/relabel-street/directory.d.ts.map +1 -0
  466. package/out/tools/golden/relabel-street/directory.js +216 -0
  467. package/out/tools/golden/relabel-street/directory.js.map +1 -0
  468. package/out/tools/golden/relabel-street/index.d.ts +9 -0
  469. package/out/tools/golden/relabel-street/index.d.ts.map +1 -0
  470. package/out/tools/golden/relabel-street/index.js +9 -0
  471. package/out/tools/golden/relabel-street/index.js.map +1 -0
  472. package/out/tools/golden/relabel-street/row.d.ts +121 -0
  473. package/out/tools/golden/relabel-street/row.d.ts.map +1 -0
  474. package/out/tools/golden/relabel-street/row.js +232 -0
  475. package/out/tools/golden/relabel-street/row.js.map +1 -0
  476. package/out/tools/index.d.ts +10 -10
  477. package/out/tools/index.d.ts.map +1 -1
  478. package/out/tools/index.js +10 -10
  479. package/out/tools/index.js.map +1 -1
  480. package/out/tools/ingest-csv.d.ts.map +1 -1
  481. package/out/tools/ingest-csv.js +5 -7
  482. package/out/tools/ingest-csv.js.map +1 -1
  483. package/out/tools/lint/slice/index.d.ts +93 -0
  484. package/out/tools/lint/slice/index.d.ts.map +1 -0
  485. package/out/tools/lint/slice/index.js +322 -0
  486. package/out/tools/lint/slice/index.js.map +1 -0
  487. package/out/tools/lint/slice/vocab.d.ts +90 -0
  488. package/out/tools/lint/slice/vocab.d.ts.map +1 -0
  489. package/out/tools/lint/slice/vocab.js +254 -0
  490. package/out/tools/lint/slice/vocab.js.map +1 -0
  491. package/out/tools/overlay-manifest.d.ts +56 -6
  492. package/out/tools/overlay-manifest.d.ts.map +1 -1
  493. package/out/tools/overlay-manifest.js +56 -15
  494. package/out/tools/overlay-manifest.js.map +1 -1
  495. package/out/tools/overture-subvenue.d.ts +1 -1
  496. package/out/tools/postcode-triples.d.ts +36 -0
  497. package/out/tools/postcode-triples.d.ts.map +1 -1
  498. package/out/tools/postcode-triples.js +91 -12
  499. package/out/tools/postcode-triples.js.map +1 -1
  500. package/out/tools/slice/kryptonite.d.ts +32 -0
  501. package/out/tools/slice/kryptonite.d.ts.map +1 -0
  502. package/out/tools/slice/kryptonite.js +99 -0
  503. package/out/tools/slice/kryptonite.js.map +1 -0
  504. package/out/tools/slice/translit.d.ts +51 -0
  505. package/out/tools/slice/translit.d.ts.map +1 -0
  506. package/out/tools/slice/translit.js +166 -0
  507. package/out/tools/slice/translit.js.map +1 -0
  508. package/out/tools/sub/venue/harvest.d.ts +100 -0
  509. package/out/tools/sub/venue/harvest.d.ts.map +1 -0
  510. package/out/tools/sub/venue/harvest.js +168 -0
  511. package/out/tools/sub/venue/harvest.js.map +1 -0
  512. package/out/tools/sub/venue/head-nouns.d.ts +50 -0
  513. package/out/tools/sub/venue/head-nouns.d.ts.map +1 -0
  514. package/out/tools/sub/venue/head-nouns.js.map +1 -0
  515. package/out/tools/sub/venue/lexicon.d.ts +156 -0
  516. package/out/tools/sub/venue/lexicon.d.ts.map +1 -0
  517. package/out/tools/sub/venue/lexicon.js +283 -0
  518. package/out/tools/sub/venue/lexicon.js.map +1 -0
  519. package/out/tools/sub/venue/promotions.d.ts +94 -0
  520. package/out/tools/sub/venue/promotions.d.ts.map +1 -0
  521. package/out/tools/sub/venue/promotions.js +266 -0
  522. package/out/tools/sub/venue/promotions.js.map +1 -0
  523. package/out/tools/sub/venue/surfaces.d.ts +45 -0
  524. package/out/tools/sub/venue/surfaces.d.ts.map +1 -0
  525. package/out/tools/sub/venue/surfaces.js.map +1 -0
  526. package/out/tools/sub/venue/table.d.ts +229 -0
  527. package/out/tools/sub/venue/table.d.ts.map +1 -0
  528. package/out/tools/sub/venue/table.js.map +1 -0
  529. package/out/tools/sub/venue/wikidata.d.ts +26 -0
  530. package/out/tools/sub/venue/wikidata.d.ts.map +1 -0
  531. package/out/tools/sub/venue/wikidata.js +62 -0
  532. package/out/tools/sub/venue/wikidata.js.map +1 -0
  533. package/out/tw/tools/fetch/gcis.d.ts +23 -0
  534. package/out/tw/tools/fetch/gcis.d.ts.map +1 -0
  535. package/out/tw/tools/fetch/gcis.js +135 -0
  536. package/out/tw/tools/fetch/gcis.js.map +1 -0
  537. package/out/types.d.ts +1 -1
  538. package/out/types.d.ts.map +1 -1
  539. package/out/us/adapters/fcc-bdc/adapter.d.ts.map +1 -0
  540. package/out/us/adapters/fcc-bdc/adapter.js +142 -0
  541. package/out/us/adapters/fcc-bdc/adapter.js.map +1 -0
  542. package/out/us/adapters/state/hi-schools/adapter.d.ts.map +1 -0
  543. package/out/us/adapters/state/hi-schools/adapter.js +136 -0
  544. package/out/us/adapters/state/hi-schools/adapter.js.map +1 -0
  545. package/out/us/adapters/state/hi-schools/workbook.d.ts.map +1 -0
  546. package/out/us/adapters/state/hi-schools/workbook.js.map +1 -0
  547. package/out/us/adapters/state/ia-contractors/adapter.d.ts.map +1 -0
  548. package/out/us/adapters/state/ia-contractors/adapter.js +104 -0
  549. package/out/us/adapters/state/ia-contractors/adapter.js.map +1 -0
  550. package/out/us/adapters/state/ny-notaries/adapter.d.ts.map +1 -0
  551. package/out/us/adapters/state/ny-notaries/adapter.js +112 -0
  552. package/out/us/adapters/state/ny-notaries/adapter.js.map +1 -0
  553. package/out/us/adapters/state/tx-notaries/adapter.d.ts.map +1 -0
  554. package/out/us/adapters/state/tx-notaries/adapter.js +118 -0
  555. package/out/us/adapters/state/tx-notaries/adapter.js.map +1 -0
  556. package/out/us/adapters/tiger/adapter.d.ts +55 -0
  557. package/out/us/adapters/tiger/adapter.d.ts.map +1 -0
  558. package/out/us/adapters/tiger/adapter.js +197 -0
  559. package/out/us/adapters/tiger/adapter.js.map +1 -0
  560. package/out/us/adapters/tiger/street-decompose.d.ts.map +1 -0
  561. package/out/us/adapters/tiger/street-decompose.js.map +1 -0
  562. package/out/us/adapters/usgov/hrsa-fqhc/adapter.d.ts.map +1 -0
  563. package/out/us/adapters/usgov/hrsa-fqhc/adapter.js +121 -0
  564. package/out/us/adapters/usgov/hrsa-fqhc/adapter.js.map +1 -0
  565. package/out/us/adapters/usgov/imls-pls/adapter.d.ts.map +1 -0
  566. package/out/us/adapters/usgov/imls-pls/adapter.js +106 -0
  567. package/out/us/adapters/usgov/imls-pls/adapter.js.map +1 -0
  568. package/out/us/adapters/usgov/irs-bmf/adapter.d.ts.map +1 -0
  569. package/out/us/adapters/usgov/irs-bmf/adapter.js +116 -0
  570. package/out/us/adapters/usgov/irs-bmf/adapter.js.map +1 -0
  571. package/out/us/adapters/usgov/nad/adapter.d.ts.map +1 -0
  572. package/out/us/adapters/usgov/nad/adapter.js +239 -0
  573. package/out/us/adapters/usgov/nad/adapter.js.map +1 -0
  574. package/out/us/adapters/usgov/nppes/adapter.d.ts.map +1 -0
  575. package/out/us/adapters/usgov/nppes/adapter.js +110 -0
  576. package/out/us/adapters/usgov/nppes/adapter.js.map +1 -0
  577. package/out/us/adapters/usgov/samhsa-treatment-locator/adapter.d.ts.map +1 -0
  578. package/out/us/adapters/usgov/samhsa-treatment-locator/adapter.js +143 -0
  579. package/out/us/adapters/usgov/samhsa-treatment-locator/adapter.js.map +1 -0
  580. package/out/us/fips-state.d.ts +41 -0
  581. package/out/us/fips-state.d.ts.map +1 -0
  582. package/out/us/fips-state.js +100 -0
  583. package/out/us/fips-state.js.map +1 -0
  584. package/out/us/tools/fetch/hrsa.d.ts +18 -0
  585. package/out/us/tools/fetch/hrsa.d.ts.map +1 -0
  586. package/out/us/tools/fetch/hrsa.js +46 -0
  587. package/out/us/tools/fetch/hrsa.js.map +1 -0
  588. package/out/us/tools/fetch/imls-pls.d.ts +25 -0
  589. package/out/us/tools/fetch/imls-pls.d.ts.map +1 -0
  590. package/out/us/tools/fetch/imls-pls.js +108 -0
  591. package/out/us/tools/fetch/imls-pls.js.map +1 -0
  592. package/out/us/tools/fetch/nad.d.ts +70 -0
  593. package/out/us/tools/fetch/nad.d.ts.map +1 -0
  594. package/out/us/tools/fetch/nad.js +234 -0
  595. package/out/us/tools/fetch/nad.js.map +1 -0
  596. package/out/us/tools/fetch/nppes.d.ts +26 -0
  597. package/out/us/tools/fetch/nppes.d.ts.map +1 -0
  598. package/out/us/tools/fetch/nppes.js +126 -0
  599. package/out/us/tools/fetch/nppes.js.map +1 -0
  600. package/out/us/tools/fetch/state/hi-schools.d.ts +26 -0
  601. package/out/us/tools/fetch/state/hi-schools.d.ts.map +1 -0
  602. package/out/us/tools/fetch/state/hi-schools.js +112 -0
  603. package/out/us/tools/fetch/state/hi-schools.js.map +1 -0
  604. package/out/us/tools/fetch/state/sources.d.ts +26 -0
  605. package/out/us/tools/fetch/state/sources.d.ts.map +1 -0
  606. package/out/us/tools/fetch/state/sources.js +111 -0
  607. package/out/us/tools/fetch/state/sources.js.map +1 -0
  608. package/out/us/tools/fetch/tiger-full.d.ts +22 -0
  609. package/out/us/tools/fetch/tiger-full.d.ts.map +1 -0
  610. package/out/us/tools/fetch/tiger-full.js +217 -0
  611. package/out/us/tools/fetch/tiger-full.js.map +1 -0
  612. package/out/utils/align.d.ts +1 -1
  613. package/out/utils/align.d.ts.map +1 -1
  614. package/out/utils/align.js.map +1 -1
  615. package/out/utils/golden.d.ts +7 -3
  616. package/out/utils/golden.d.ts.map +1 -1
  617. package/out/utils/golden.js +12 -8
  618. package/out/utils/golden.js.map +1 -1
  619. package/out/utils/parquet.d.ts.map +1 -1
  620. package/out/utils/split.js.map +1 -1
  621. package/out/utils/wof-json.d.ts.map +1 -1
  622. package/out/utils/wof-json.js +7 -12
  623. package/out/utils/wof-json.js.map +1 -1
  624. package/out/ve/recipes/reviewed-postcode-tail.d.ts.map +1 -0
  625. package/out/ve/recipes/reviewed-postcode-tail.js.map +1 -0
  626. package/package.json +196 -46
  627. package/lib/adapters/ban/adapter.ts +0 -190
  628. package/lib/adapters/fcc-bdc/adapter.ts +0 -174
  629. package/lib/adapters/geonames-postal/adapter.ts +0 -122
  630. package/lib/adapters/gnaf/adapter.ts +0 -149
  631. package/lib/adapters/gnaf/assemble.ts +0 -225
  632. package/lib/adapters/state-hi-schools/adapter.ts +0 -170
  633. package/lib/adapters/state-ia-contractors/adapter.ts +0 -144
  634. package/lib/adapters/state-ny-notaries/adapter.ts +0 -141
  635. package/lib/adapters/state-tx-notaries/adapter.ts +0 -157
  636. package/lib/adapters/tiger/adapter.ts +0 -252
  637. package/lib/adapters/usgov-hrsa-fqhc/adapter.ts +0 -167
  638. package/lib/adapters/usgov-imls-pls/adapter.ts +0 -143
  639. package/lib/adapters/usgov-irs-bmf/adapter.ts +0 -151
  640. package/lib/adapters/usgov-nad/adapter.ts +0 -334
  641. package/lib/adapters/usgov-nppes/adapter.ts +0 -151
  642. package/lib/adapters/usgov-samhsa-treatment-locator/adapter.ts +0 -193
  643. package/lib/adapters/wof-admin-json/adapter.ts +0 -231
  644. package/lib/adapters/wof-json-rows.ts +0 -129
  645. package/lib/adapters/wof-postalcode-json/adapter.ts +0 -176
  646. package/lib/name-prone-us-suffixes.ts +0 -12
  647. package/lib/recipes/bare-country.ts +0 -95
  648. package/lib/recipes/cz-pcfirst-preposition.ts +0 -111
  649. package/lib/recipes/fr-admin-split.ts +0 -242
  650. package/lib/recipes/fr-bare-street.ts +0 -170
  651. package/lib/recipes/fr-fragment.ts +0 -333
  652. package/lib/recipes/fr-lieudit.ts +0 -301
  653. package/lib/recipes/fr-order.ts +0 -245
  654. package/lib/recipes/german.ts +0 -359
  655. package/lib/recipes/locale.ts +0 -599
  656. package/lib/recipes/nl-postcode.ts +0 -115
  657. package/lib/recipes/no-fragment.ts +0 -219
  658. package/lib/recipes/no-street-led.ts +0 -141
  659. package/lib/recipes/no-street.ts +0 -90
  660. package/lib/recipes/po-box-cedex.ts +0 -976
  661. package/lib/recipes/po-box.ts +0 -123
  662. package/lib/recipes/si-bare-village.ts +0 -97
  663. package/lib/recipes/street-affix.ts +0 -804
  664. package/lib/recipes/street-bare.ts +0 -100
  665. package/lib/recipes/sub-venue-sources.ts +0 -877
  666. package/lib/recipes/sub-venue.ts +0 -997
  667. package/lib/synthesizers/german.ts +0 -260
  668. package/lib/tools/fetch/ban.ts +0 -289
  669. package/lib/tools/fetch/download.ts +0 -262
  670. package/lib/tools/fetch/geonames-dump.ts +0 -257
  671. package/lib/tools/fetch/geonames-postal.ts +0 -180
  672. package/lib/tools/fetch/hrsa.ts +0 -59
  673. package/lib/tools/fetch/imls-pls.ts +0 -149
  674. package/lib/tools/fetch/nad.ts +0 -341
  675. package/lib/tools/fetch/nppes.ts +0 -165
  676. package/lib/tools/fetch/state-hi-schools.ts +0 -161
  677. package/lib/tools/fetch/state-sources.ts +0 -146
  678. package/lib/tools/fetch/tiger-full.ts +0 -308
  679. package/lib/tools/golden-promote.ts +0 -337
  680. package/lib/tools/golden-relabel-street.ts +0 -683
  681. package/lib/tools/lint-slice-vocab.ts +0 -388
  682. package/lib/tools/sub-venue/harvest.ts +0 -234
  683. package/lib/tools/sub-venue/head-nouns.ts +0 -245
  684. package/lib/tools/sub-venue/surfaces.ts +0 -95
  685. package/lib/tools/sub-venue/table.ts +0 -281
  686. package/lib/tools/sub-venue/wikidata.ts +0 -85
  687. package/lib/tools/sub-venue-lexicon.ts +0 -429
  688. package/out/adapters/ban/adapter.d.ts.map +0 -1
  689. package/out/adapters/ban/adapter.js +0 -149
  690. package/out/adapters/ban/adapter.js.map +0 -1
  691. package/out/adapters/ban/street-decompose.d.ts.map +0 -1
  692. package/out/adapters/ban/street-decompose.js.map +0 -1
  693. package/out/adapters/fcc-bdc/adapter.d.ts.map +0 -1
  694. package/out/adapters/fcc-bdc/adapter.js +0 -144
  695. package/out/adapters/fcc-bdc/adapter.js.map +0 -1
  696. package/out/adapters/geonames-postal/adapter.d.ts.map +0 -1
  697. package/out/adapters/geonames-postal/adapter.js +0 -107
  698. package/out/adapters/geonames-postal/adapter.js.map +0 -1
  699. package/out/adapters/gnaf/adapter.d.ts.map +0 -1
  700. package/out/adapters/gnaf/adapter.js +0 -124
  701. package/out/adapters/gnaf/adapter.js.map +0 -1
  702. package/out/adapters/gnaf/assemble.d.ts +0 -65
  703. package/out/adapters/gnaf/assemble.d.ts.map +0 -1
  704. package/out/adapters/gnaf/assemble.js +0 -154
  705. package/out/adapters/gnaf/assemble.js.map +0 -1
  706. package/out/adapters/state-hi-schools/adapter.d.ts.map +0 -1
  707. package/out/adapters/state-hi-schools/adapter.js +0 -145
  708. package/out/adapters/state-hi-schools/adapter.js.map +0 -1
  709. package/out/adapters/state-hi-schools/workbook.d.ts.map +0 -1
  710. package/out/adapters/state-hi-schools/workbook.js.map +0 -1
  711. package/out/adapters/state-ia-contractors/adapter.d.ts.map +0 -1
  712. package/out/adapters/state-ia-contractors/adapter.js +0 -111
  713. package/out/adapters/state-ia-contractors/adapter.js.map +0 -1
  714. package/out/adapters/state-ny-notaries/adapter.d.ts.map +0 -1
  715. package/out/adapters/state-ny-notaries/adapter.js +0 -119
  716. package/out/adapters/state-ny-notaries/adapter.js.map +0 -1
  717. package/out/adapters/state-tx-notaries/adapter.d.ts.map +0 -1
  718. package/out/adapters/state-tx-notaries/adapter.js +0 -125
  719. package/out/adapters/state-tx-notaries/adapter.js.map +0 -1
  720. package/out/adapters/tiger/adapter.d.ts +0 -55
  721. package/out/adapters/tiger/adapter.d.ts.map +0 -1
  722. package/out/adapters/tiger/adapter.js +0 -201
  723. package/out/adapters/tiger/adapter.js.map +0 -1
  724. package/out/adapters/tiger/street-decompose.d.ts.map +0 -1
  725. package/out/adapters/tiger/street-decompose.js.map +0 -1
  726. package/out/adapters/usgov-hrsa-fqhc/adapter.d.ts.map +0 -1
  727. package/out/adapters/usgov-hrsa-fqhc/adapter.js +0 -132
  728. package/out/adapters/usgov-hrsa-fqhc/adapter.js.map +0 -1
  729. package/out/adapters/usgov-imls-pls/adapter.d.ts.map +0 -1
  730. package/out/adapters/usgov-imls-pls/adapter.js +0 -114
  731. package/out/adapters/usgov-imls-pls/adapter.js.map +0 -1
  732. package/out/adapters/usgov-irs-bmf/adapter.d.ts.map +0 -1
  733. package/out/adapters/usgov-irs-bmf/adapter.js +0 -120
  734. package/out/adapters/usgov-irs-bmf/adapter.js.map +0 -1
  735. package/out/adapters/usgov-nad/adapter.d.ts.map +0 -1
  736. package/out/adapters/usgov-nad/adapter.js +0 -249
  737. package/out/adapters/usgov-nad/adapter.js.map +0 -1
  738. package/out/adapters/usgov-nppes/adapter.d.ts.map +0 -1
  739. package/out/adapters/usgov-nppes/adapter.js +0 -118
  740. package/out/adapters/usgov-nppes/adapter.js.map +0 -1
  741. package/out/adapters/usgov-samhsa-treatment-locator/adapter.d.ts.map +0 -1
  742. package/out/adapters/usgov-samhsa-treatment-locator/adapter.js +0 -154
  743. package/out/adapters/usgov-samhsa-treatment-locator/adapter.js.map +0 -1
  744. package/out/adapters/wof-admin-jp/adapter.d.ts.map +0 -1
  745. package/out/adapters/wof-admin-jp/adapter.js.map +0 -1
  746. package/out/adapters/wof-admin-json/adapter.d.ts +0 -79
  747. package/out/adapters/wof-admin-json/adapter.d.ts.map +0 -1
  748. package/out/adapters/wof-admin-json/adapter.js +0 -197
  749. package/out/adapters/wof-admin-json/adapter.js.map +0 -1
  750. package/out/adapters/wof-json-rows.d.ts +0 -62
  751. package/out/adapters/wof-json-rows.d.ts.map +0 -1
  752. package/out/adapters/wof-json-rows.js +0 -83
  753. package/out/adapters/wof-json-rows.js.map +0 -1
  754. package/out/adapters/wof-postalcode-json/adapter.d.ts +0 -65
  755. package/out/adapters/wof-postalcode-json/adapter.d.ts.map +0 -1
  756. package/out/adapters/wof-postalcode-json/adapter.js +0 -151
  757. package/out/adapters/wof-postalcode-json/adapter.js.map +0 -1
  758. package/out/codex/us-fips-state.d.ts +0 -41
  759. package/out/codex/us-fips-state.d.ts.map +0 -1
  760. package/out/codex/us-fips-state.js +0 -100
  761. package/out/codex/us-fips-state.js.map +0 -1
  762. package/out/name-prone-us-suffixes.d.ts +0 -12
  763. package/out/name-prone-us-suffixes.d.ts.map +0 -1
  764. package/out/name-prone-us-suffixes.js +0 -12
  765. package/out/name-prone-us-suffixes.js.map +0 -1
  766. package/out/recipes/bare-country.d.ts +0 -25
  767. package/out/recipes/bare-country.d.ts.map +0 -1
  768. package/out/recipes/bare-country.js +0 -87
  769. package/out/recipes/bare-country.js.map +0 -1
  770. package/out/recipes/cn-organizational-units.d.ts +0 -36
  771. package/out/recipes/cn-organizational-units.d.ts.map +0 -1
  772. package/out/recipes/cn-organizational-units.js +0 -159
  773. package/out/recipes/cn-organizational-units.js.map +0 -1
  774. package/out/recipes/cz-pcfirst-preposition.d.ts +0 -22
  775. package/out/recipes/cz-pcfirst-preposition.d.ts.map +0 -1
  776. package/out/recipes/cz-pcfirst-preposition.js +0 -100
  777. package/out/recipes/cz-pcfirst-preposition.js.map +0 -1
  778. package/out/recipes/fr-admin-split.d.ts +0 -38
  779. package/out/recipes/fr-admin-split.d.ts.map +0 -1
  780. package/out/recipes/fr-admin-split.js +0 -192
  781. package/out/recipes/fr-admin-split.js.map +0 -1
  782. package/out/recipes/fr-bare-street.d.ts +0 -28
  783. package/out/recipes/fr-bare-street.d.ts.map +0 -1
  784. package/out/recipes/fr-bare-street.js +0 -143
  785. package/out/recipes/fr-bare-street.js.map +0 -1
  786. package/out/recipes/fr-fragment.d.ts +0 -77
  787. package/out/recipes/fr-fragment.d.ts.map +0 -1
  788. package/out/recipes/fr-fragment.js +0 -276
  789. package/out/recipes/fr-fragment.js.map +0 -1
  790. package/out/recipes/fr-lieudit.d.ts +0 -32
  791. package/out/recipes/fr-lieudit.d.ts.map +0 -1
  792. package/out/recipes/fr-lieudit.js +0 -227
  793. package/out/recipes/fr-lieudit.js.map +0 -1
  794. package/out/recipes/fr-order.d.ts +0 -38
  795. package/out/recipes/fr-order.d.ts.map +0 -1
  796. package/out/recipes/fr-order.js +0 -197
  797. package/out/recipes/fr-order.js.map +0 -1
  798. package/out/recipes/german.d.ts +0 -44
  799. package/out/recipes/german.d.ts.map +0 -1
  800. package/out/recipes/german.js +0 -288
  801. package/out/recipes/german.js.map +0 -1
  802. package/out/recipes/locale.d.ts +0 -136
  803. package/out/recipes/locale.d.ts.map +0 -1
  804. package/out/recipes/locale.js +0 -460
  805. package/out/recipes/locale.js.map +0 -1
  806. package/out/recipes/nl-postcode.d.ts +0 -26
  807. package/out/recipes/nl-postcode.d.ts.map +0 -1
  808. package/out/recipes/nl-postcode.js +0 -102
  809. package/out/recipes/nl-postcode.js.map +0 -1
  810. package/out/recipes/no-fragment.d.ts +0 -43
  811. package/out/recipes/no-fragment.d.ts.map +0 -1
  812. package/out/recipes/no-fragment.js +0 -182
  813. package/out/recipes/no-fragment.js.map +0 -1
  814. package/out/recipes/no-street-led.d.ts +0 -24
  815. package/out/recipes/no-street-led.d.ts.map +0 -1
  816. package/out/recipes/no-street-led.js +0 -113
  817. package/out/recipes/no-street-led.js.map +0 -1
  818. package/out/recipes/no-street.d.ts +0 -18
  819. package/out/recipes/no-street.d.ts.map +0 -1
  820. package/out/recipes/no-street.js +0 -71
  821. package/out/recipes/no-street.js.map +0 -1
  822. package/out/recipes/po-box-cedex.d.ts +0 -38
  823. package/out/recipes/po-box-cedex.d.ts.map +0 -1
  824. package/out/recipes/po-box-cedex.js +0 -767
  825. package/out/recipes/po-box-cedex.js.map +0 -1
  826. package/out/recipes/po-box.d.ts +0 -17
  827. package/out/recipes/po-box.d.ts.map +0 -1
  828. package/out/recipes/po-box.js +0 -100
  829. package/out/recipes/po-box.js.map +0 -1
  830. package/out/recipes/reviewed-postcode-tail.d.ts.map +0 -1
  831. package/out/recipes/reviewed-postcode-tail.js.map +0 -1
  832. package/out/recipes/si-bare-village.d.ts +0 -32
  833. package/out/recipes/si-bare-village.d.ts.map +0 -1
  834. package/out/recipes/si-bare-village.js +0 -88
  835. package/out/recipes/si-bare-village.js.map +0 -1
  836. package/out/recipes/street-affix.d.ts +0 -89
  837. package/out/recipes/street-affix.d.ts.map +0 -1
  838. package/out/recipes/street-affix.js +0 -563
  839. package/out/recipes/street-affix.js.map +0 -1
  840. package/out/recipes/street-bare.d.ts +0 -25
  841. package/out/recipes/street-bare.d.ts.map +0 -1
  842. package/out/recipes/street-bare.js +0 -86
  843. package/out/recipes/street-bare.js.map +0 -1
  844. package/out/recipes/sub-venue-sources.d.ts +0 -229
  845. package/out/recipes/sub-venue-sources.d.ts.map +0 -1
  846. package/out/recipes/sub-venue-sources.js +0 -653
  847. package/out/recipes/sub-venue-sources.js.map +0 -1
  848. package/out/recipes/sub-venue.d.ts +0 -269
  849. package/out/recipes/sub-venue.d.ts.map +0 -1
  850. package/out/recipes/sub-venue.js +0 -725
  851. package/out/recipes/sub-venue.js.map +0 -1
  852. package/out/synthesizers/german.d.ts +0 -115
  853. package/out/synthesizers/german.d.ts.map +0 -1
  854. package/out/synthesizers/german.js +0 -149
  855. package/out/synthesizers/german.js.map +0 -1
  856. package/out/tools/fetch/ban.d.ts +0 -31
  857. package/out/tools/fetch/ban.d.ts.map +0 -1
  858. package/out/tools/fetch/ban.js +0 -239
  859. package/out/tools/fetch/ban.js.map +0 -1
  860. package/out/tools/fetch/download.d.ts +0 -117
  861. package/out/tools/fetch/download.d.ts.map +0 -1
  862. package/out/tools/fetch/download.js +0 -160
  863. package/out/tools/fetch/download.js.map +0 -1
  864. package/out/tools/fetch/geonames-dump.d.ts +0 -91
  865. package/out/tools/fetch/geonames-dump.d.ts.map +0 -1
  866. package/out/tools/fetch/geonames-dump.js +0 -171
  867. package/out/tools/fetch/geonames-dump.js.map +0 -1
  868. package/out/tools/fetch/geonames-postal.d.ts +0 -76
  869. package/out/tools/fetch/geonames-postal.d.ts.map +0 -1
  870. package/out/tools/fetch/geonames-postal.js +0 -126
  871. package/out/tools/fetch/geonames-postal.js.map +0 -1
  872. package/out/tools/fetch/hrsa.d.ts +0 -18
  873. package/out/tools/fetch/hrsa.d.ts.map +0 -1
  874. package/out/tools/fetch/hrsa.js +0 -46
  875. package/out/tools/fetch/hrsa.js.map +0 -1
  876. package/out/tools/fetch/imls-pls.d.ts +0 -25
  877. package/out/tools/fetch/imls-pls.d.ts.map +0 -1
  878. package/out/tools/fetch/imls-pls.js +0 -108
  879. package/out/tools/fetch/imls-pls.js.map +0 -1
  880. package/out/tools/fetch/nad.d.ts +0 -70
  881. package/out/tools/fetch/nad.d.ts.map +0 -1
  882. package/out/tools/fetch/nad.js +0 -234
  883. package/out/tools/fetch/nad.js.map +0 -1
  884. package/out/tools/fetch/nppes.d.ts +0 -26
  885. package/out/tools/fetch/nppes.d.ts.map +0 -1
  886. package/out/tools/fetch/nppes.js +0 -126
  887. package/out/tools/fetch/nppes.js.map +0 -1
  888. package/out/tools/fetch/state-hi-schools.d.ts +0 -26
  889. package/out/tools/fetch/state-hi-schools.d.ts.map +0 -1
  890. package/out/tools/fetch/state-hi-schools.js +0 -112
  891. package/out/tools/fetch/state-hi-schools.js.map +0 -1
  892. package/out/tools/fetch/state-sources.d.ts +0 -26
  893. package/out/tools/fetch/state-sources.d.ts.map +0 -1
  894. package/out/tools/fetch/state-sources.js +0 -111
  895. package/out/tools/fetch/state-sources.js.map +0 -1
  896. package/out/tools/fetch/tiger-full.d.ts +0 -22
  897. package/out/tools/fetch/tiger-full.d.ts.map +0 -1
  898. package/out/tools/fetch/tiger-full.js +0 -217
  899. package/out/tools/fetch/tiger-full.js.map +0 -1
  900. package/out/tools/golden-expand.d.ts +0 -90
  901. package/out/tools/golden-expand.d.ts.map +0 -1
  902. package/out/tools/golden-expand.js +0 -406
  903. package/out/tools/golden-expand.js.map +0 -1
  904. package/out/tools/golden-promote.d.ts +0 -72
  905. package/out/tools/golden-promote.d.ts.map +0 -1
  906. package/out/tools/golden-promote.js +0 -240
  907. package/out/tools/golden-promote.js.map +0 -1
  908. package/out/tools/golden-relabel-street.d.ts +0 -196
  909. package/out/tools/golden-relabel-street.d.ts.map +0 -1
  910. package/out/tools/golden-relabel-street.js +0 -440
  911. package/out/tools/golden-relabel-street.js.map +0 -1
  912. package/out/tools/lint-slice-vocab.d.ts +0 -90
  913. package/out/tools/lint-slice-vocab.d.ts.map +0 -1
  914. package/out/tools/lint-slice-vocab.js +0 -267
  915. package/out/tools/lint-slice-vocab.js.map +0 -1
  916. package/out/tools/lint-slice.d.ts +0 -93
  917. package/out/tools/lint-slice.d.ts.map +0 -1
  918. package/out/tools/lint-slice.js +0 -322
  919. package/out/tools/lint-slice.js.map +0 -1
  920. package/out/tools/slice-kryptonite.d.ts +0 -32
  921. package/out/tools/slice-kryptonite.d.ts.map +0 -1
  922. package/out/tools/slice-kryptonite.js +0 -99
  923. package/out/tools/slice-kryptonite.js.map +0 -1
  924. package/out/tools/slice-translit.d.ts +0 -51
  925. package/out/tools/slice-translit.d.ts.map +0 -1
  926. package/out/tools/slice-translit.js +0 -166
  927. package/out/tools/slice-translit.js.map +0 -1
  928. package/out/tools/sub-venue/harvest.d.ts +0 -100
  929. package/out/tools/sub-venue/harvest.d.ts.map +0 -1
  930. package/out/tools/sub-venue/harvest.js +0 -168
  931. package/out/tools/sub-venue/harvest.js.map +0 -1
  932. package/out/tools/sub-venue/head-nouns.d.ts +0 -50
  933. package/out/tools/sub-venue/head-nouns.d.ts.map +0 -1
  934. package/out/tools/sub-venue/head-nouns.js.map +0 -1
  935. package/out/tools/sub-venue/surfaces.d.ts +0 -45
  936. package/out/tools/sub-venue/surfaces.d.ts.map +0 -1
  937. package/out/tools/sub-venue/surfaces.js.map +0 -1
  938. package/out/tools/sub-venue/table.d.ts +0 -229
  939. package/out/tools/sub-venue/table.d.ts.map +0 -1
  940. package/out/tools/sub-venue/table.js.map +0 -1
  941. package/out/tools/sub-venue/wikidata.d.ts +0 -26
  942. package/out/tools/sub-venue/wikidata.d.ts.map +0 -1
  943. package/out/tools/sub-venue/wikidata.js +0 -62
  944. package/out/tools/sub-venue/wikidata.js.map +0 -1
  945. package/out/tools/sub-venue-lexicon.d.ts +0 -156
  946. package/out/tools/sub-venue-lexicon.d.ts.map +0 -1
  947. package/out/tools/sub-venue-lexicon.js +0 -283
  948. package/out/tools/sub-venue-lexicon.js.map +0 -1
  949. package/out/tools/sub-venue-promotions.d.ts +0 -94
  950. package/out/tools/sub-venue-promotions.d.ts.map +0 -1
  951. package/out/tools/sub-venue-promotions.js +0 -266
  952. package/out/tools/sub-venue-promotions.js.map +0 -1
  953. /package/lib/{recipes/cn-organizational-units.ts → cn/recipes/organizational-units.ts} +0 -0
  954. /package/lib/{adapters → fr/adapters}/ban/street-decompose.ts +0 -0
  955. /package/lib/{adapters/wof-admin-jp → jp/adapters/wof-admin}/adapter.ts +0 -0
  956. /package/lib/tools/{golden-expand.ts → golden/expand.ts} +0 -0
  957. /package/lib/tools/{lint-slice.ts → lint/slice/index.ts} +0 -0
  958. /package/lib/tools/{slice-kryptonite.ts → slice/kryptonite.ts} +0 -0
  959. /package/lib/tools/{slice-translit.ts → slice/translit.ts} +0 -0
  960. /package/lib/tools/{sub-venue-promotions.ts → sub/venue/promotions.ts} +0 -0
  961. /package/lib/{adapters/state-hi-schools → us/adapters/state/hi-schools}/workbook.ts +0 -0
  962. /package/lib/{adapters → us/adapters}/tiger/street-decompose.ts +0 -0
  963. /package/lib/{codex/us-fips-state.ts → us/fips-state.ts} +0 -0
  964. /package/lib/{recipes → ve/recipes}/reviewed-postcode-tail.ts +0 -0
  965. /package/out/adapters/{geonames-postal → geonames/postal}/adapter.d.ts +0 -0
  966. /package/out/{adapters → au/adapters}/gnaf/adapter.d.ts +0 -0
  967. /package/out/{adapters → fr/adapters}/ban/adapter.d.ts +0 -0
  968. /package/out/{adapters → fr/adapters}/ban/street-decompose.d.ts +0 -0
  969. /package/out/{adapters → fr/adapters}/ban/street-decompose.js +0 -0
  970. /package/out/{adapters/wof-admin-jp → jp/adapters/wof-admin}/adapter.d.ts +0 -0
  971. /package/out/{adapters/wof-admin-jp → jp/adapters/wof-admin}/adapter.js +0 -0
  972. /package/out/tools/{sub-venue → sub/venue}/head-nouns.js +0 -0
  973. /package/out/tools/{sub-venue → sub/venue}/surfaces.js +0 -0
  974. /package/out/tools/{sub-venue → sub/venue}/table.js +0 -0
  975. /package/out/{adapters → us/adapters}/fcc-bdc/adapter.d.ts +0 -0
  976. /package/out/{adapters/state-hi-schools → us/adapters/state/hi-schools}/adapter.d.ts +0 -0
  977. /package/out/{adapters/state-hi-schools → us/adapters/state/hi-schools}/workbook.d.ts +0 -0
  978. /package/out/{adapters/state-hi-schools → us/adapters/state/hi-schools}/workbook.js +0 -0
  979. /package/out/{adapters/state-ia-contractors → us/adapters/state/ia-contractors}/adapter.d.ts +0 -0
  980. /package/out/{adapters/state-ny-notaries → us/adapters/state/ny-notaries}/adapter.d.ts +0 -0
  981. /package/out/{adapters/state-tx-notaries → us/adapters/state/tx-notaries}/adapter.d.ts +0 -0
  982. /package/out/{adapters → us/adapters}/tiger/street-decompose.d.ts +0 -0
  983. /package/out/{adapters → us/adapters}/tiger/street-decompose.js +0 -0
  984. /package/out/{adapters/usgov-hrsa-fqhc → us/adapters/usgov/hrsa-fqhc}/adapter.d.ts +0 -0
  985. /package/out/{adapters/usgov-imls-pls → us/adapters/usgov/imls-pls}/adapter.d.ts +0 -0
  986. /package/out/{adapters/usgov-irs-bmf → us/adapters/usgov/irs-bmf}/adapter.d.ts +0 -0
  987. /package/out/{adapters/usgov-nad → us/adapters/usgov/nad}/adapter.d.ts +0 -0
  988. /package/out/{adapters/usgov-nppes → us/adapters/usgov/nppes}/adapter.d.ts +0 -0
  989. /package/out/{adapters/usgov-samhsa-treatment-locator → us/adapters/usgov/samhsa-treatment-locator}/adapter.d.ts +0 -0
  990. /package/out/{recipes → ve/recipes}/reviewed-postcode-tail.d.ts +0 -0
  991. /package/out/{recipes → ve/recipes}/reviewed-postcode-tail.js +0 -0
@@ -1,406 +0,0 @@
1
- /**
2
- * @copyright Sister Software
3
- * @license AGPL-3.0
4
- * @author Teffen Ellis, et al.
5
- *
6
- * Generate golden-set candidate entries by LLM-driven surface-form synthesis from a verified-label
7
- * seed pulled out of the corpus's labeled test split.
8
- *
9
- * ## Why this approach
10
- *
11
- * The Phase 2 golden set has 74 entries; session-notes.md called for ≥500/locale. Manual curation
12
- * doesn't scale. Pure-LLM generation (invent raw + labels from scratch) is too noisy — labels
13
- * would be unverified.
14
- *
15
- * This module takes the middle path:
16
- *
17
- * 1. **Seeds come from corpus-v0.2.0 test slice** — already through the alignment pipeline, so labels
18
- * are pipeline-verified.
19
- * 2. **LLM only varies the surface form** — case, abbreviations, reordering, dropped components. The
20
- * component VALUES (locality string, postcode digits, etc.) are preserved verbatim.
21
- * 3. **Programmatic validator drops hallucinations** — every component value must appear as a
22
- * substring (case-insensitive, whitespace-normalized) of the variant's raw. Failures dropped
23
- * silently; cost is wasted tokens, never bad-labeled golden entries.
24
- *
25
- * ## Usage
26
- *
27
- * ```sh
28
- * DEEPSEEK_API_KEY=sk-... \
29
- * mailwoman corpus golden expand \
30
- * --count 1000 \
31
- * --variants 5 \
32
- * --output data/eval/golden/candidates/expand-$(date +%Y%m%d-%H%M%S).jsonl
33
- * ```
34
- *
35
- * ## Env
36
- *
37
- * - `DEEPSEEK_API_KEY` — required for provider `deepseek`
38
- * - `ANTHROPIC_API_KEY` — required for provider `anthropic`
39
- *
40
- * ## What this module does NOT do
41
- *
42
- * - Does not commit anything or modify the versioned golden dir. Candidates land in
43
- * `data/eval/golden/candidates/` for operator review (skim, prune, then run
44
- * `mailwoman corpus golden promote`).
45
- * - Does not score the LLM's quality — that's an eyeball job after pilot lands.
46
- * - Does not retry hallucinated candidates. Cost of wasted tokens is trivial (~$0.0006/each).
47
- */
48
- import { APIClient, pluckResponseData } from "@mailwoman/core/api";
49
- import { dataRootPath } from "@mailwoman/core/data-root";
50
- import { makeDirectories } from "@mailwoman/core/fs/writers";
51
- import { tryParsingJSON } from "@mailwoman/core/json";
52
- import { isPresent } from "@mailwoman/core/objects";
53
- import { foldCaseWhitespace } from "@mailwoman/normalize/fold";
54
- import { dirname } from "path-ts";
55
- import { createNewlineWriter } from "spliterator";
56
- import { $private } from "#env";
57
- import { streamParquetRows } from "#utils/parquet";
58
- // ── Types ─────────────────────────────────────────────────────────────────
59
- /**
60
- * Longest raw line still plausibly a single address rather than a concatenated record.
61
- */
62
- const MAX_CANDIDATE_LENGTH = 500;
63
- // ── Seed loading ──────────────────────────────────────────────────────────
64
- /**
65
- * Decode BIO labels + tokens into a verified components map. Mirrors the Python `decode_components` in
66
- * mailwoman_train/eval.py — first-occurrence-wins per tag, contiguous B-X/I-X runs concatenated with a single space
67
- * (the canonical separator used by corpus alignment).
68
- */
69
- function decodeComponents(tokens, labels) {
70
- const out = {};
71
- let currentTag = null;
72
- let currentTokens = [];
73
- const flush = () => {
74
- if (currentTag && currentTokens.length && !(currentTag in out)) {
75
- out[currentTag] = currentTokens.join(" ").trim();
76
- }
77
- currentTag = null;
78
- currentTokens = [];
79
- };
80
- for (let i = 0; i < labels.length; i++) {
81
- const label = labels[i];
82
- const tok = tokens[i] ?? "";
83
- if (label === "O") {
84
- flush();
85
- continue;
86
- }
87
- const [prefix, tag] = label.split("-", 2);
88
- if (prefix === "B" || currentTag !== tag) {
89
- flush();
90
- currentTag = tag;
91
- currentTokens = [tok];
92
- }
93
- else {
94
- currentTokens.push(tok);
95
- }
96
- }
97
- flush();
98
- return out;
99
- }
100
- async function loadSeeds(corpusPath, count, includeSources, report) {
101
- const paths = corpusPath
102
- .split(",")
103
- .map((p) => p.trim())
104
- .filter(isPresent);
105
- report?.(`reading seeds from ${paths.length} slice(s) (target: ${count}, stratified)`);
106
- if (includeSources) {
107
- report?.(` include-sources filter: ${Array.from(includeSources).join(", ")}`);
108
- }
109
- // Stratified sampling: read all rows from all slices, group by source. Bounded by per-source
110
- // reservoir: keep at most max(2*count, 5000) rows per source so we don't blow memory on train
111
- // slices (1M rows × many slices). Sampling later is uniform within each pool.
112
- const bySource = new Map();
113
- const PER_SOURCE_CAP = Math.max(2 * count, 5000);
114
- let scanned = 0;
115
- let skippedThinComponents = 0;
116
- for (const path of paths) {
117
- for await (const row of streamParquetRows(path)) {
118
- scanned++;
119
- // Source allow-list (--include-sources) — applied early to skip parsing rows we won't use
120
- if (includeSources && !includeSources.has(row.source))
121
- continue;
122
- const components = decodeComponents(row.tokens ?? [], row.labels ?? []);
123
- // Skip rows with too few components — single-name wof-admin entries don't make useful seeds
124
- if (Object.keys(components).length < 2) {
125
- skippedThinComponents++;
126
- continue;
127
- }
128
- const seed = {
129
- raw: row.raw,
130
- components,
131
- country: row.country,
132
- source: row.source,
133
- source_id: row.source_id,
134
- };
135
- let bucket = bySource.get(row.source);
136
- if (!bucket) {
137
- bucket = [];
138
- bySource.set(row.source, bucket);
139
- }
140
- if (bucket.length < PER_SOURCE_CAP) {
141
- bucket.push(seed);
142
- }
143
- }
144
- }
145
- report?.(` scanned ${scanned} rows across ${paths.length} slice(s); thin-components dropped: ${skippedThinComponents}`);
146
- report?.(` per-source pool sizes:`);
147
- for (const [src, pool] of bySource) {
148
- report?.(` ${src}: ${pool.length}`);
149
- }
150
- // Round-robin sample. Each source gives floor(count / nSources) seeds; rounding goes
151
- // to sources in alphabetical order. If a pool is smaller than its target, take all of it.
152
- const sources = Array.from(bySource.keys()).toSorted();
153
- const perSource = Math.floor(count / sources.length);
154
- const remainder = count - perSource * sources.length;
155
- const picked = [];
156
- for (let i = 0; i < sources.length; i++) {
157
- const src = sources[i];
158
- const pool = bySource.get(src);
159
- const target = perSource + (i < remainder ? 1 : 0);
160
- // Random subsample without replacement — deterministic via shuffle then slice
161
- for (let j = pool.length - 1; j > 0; j--) {
162
- const k = Math.floor(Math.random() * (j + 1));
163
- [pool[j], pool[k]] = [pool[k], pool[j]];
164
- }
165
- const take = Math.min(target, pool.length);
166
- picked.push(...pool.slice(0, take));
167
- if (take < target) {
168
- report?.(` ⚠ ${src}: requested ${target}, pool had ${pool.length}`);
169
- }
170
- }
171
- report?.(` → loaded ${picked.length} seeds across ${sources.length} sources`);
172
- return picked;
173
- }
174
- const SYSTEM_PROMPT = `You are a postal-address surface-form generator. Given a structured address, produce realistic variants a human might type into a geocoder.
175
-
176
- CONSTRAINT — you MUST preserve every kept component value verbatim in the output. You may:
177
- - vary case (UPPER, lower, Title Case)
178
- - abbreviate (Saint → St, Avenue → Ave, Boulevard → Blvd, North → N, etc.)
179
- - vary punctuation (commas, dashes, spaces, line breaks)
180
- - reorder components (postcode-first, country-first, address-only)
181
- - drop OPTIONAL components — list them in "dropped"
182
-
183
- OPTIONAL components (allowed to drop): country, postcode, dependent_locality, subregion, cedex.
184
- REQUIRED components (must keep, even if input has them): locality, region (when present),
185
- street, house_number, venue (when present).
186
-
187
- ALWAYS keep AT LEAST 2 components in the final raw text. Single-component variants like
188
- "VT" or "Paris" are USELESS as eval entries — do not produce them.
189
-
190
- You MUST NOT:
191
- - introduce typos (validator drops these silently — wasted tokens)
192
- - invent new component values
193
- - produce text longer than 500 characters
194
- - output a degenerate single-token answer
195
-
196
- OUTPUT a JSON array of N objects, each shaped {"raw": "...", "dropped": ["..."]}.`;
197
- function buildUserPrompt(seed, n) {
198
- return `INPUT:
199
- ${JSON.stringify({ raw: seed.raw, components: seed.components, country: seed.country }, null, 2)}
200
-
201
- N: ${n}`;
202
- }
203
- function makeDeepseekProvider(model) {
204
- const apiKey = $private.DEEPSEEK_API_KEY;
205
- if (!apiKey)
206
- throw new Error("DEEPSEEK_API_KEY env var is required for provider deepseek");
207
- return {
208
- name: "deepseek",
209
- model,
210
- async generateVariants(seed, n) {
211
- const data = await new APIClient({
212
- displayName: "deepseek",
213
- retry: true,
214
- axios: { headers: { Authorization: `Bearer ${apiKey}`, "Content-Type": "application/json" } },
215
- })
216
- .fetch({
217
- url: "https://api.deepseek.com/chat/completions",
218
- method: "post",
219
- timeout: 60_000,
220
- data: {
221
- model,
222
- messages: [
223
- { role: "system", content: SYSTEM_PROMPT },
224
- { role: "user", content: buildUserPrompt(seed, n) },
225
- ],
226
- response_format: { type: "json_object" },
227
- max_tokens: 800,
228
- temperature: 0.7,
229
- },
230
- })
231
- .then(pluckResponseData);
232
- const content = data.choices?.[0]?.message.content ?? "{}";
233
- return parseCandidates(content);
234
- },
235
- };
236
- }
237
- function makeAnthropicProvider(model) {
238
- const apiKey = $private.ANTHROPIC_API_KEY;
239
- if (!apiKey)
240
- throw new Error("ANTHROPIC_API_KEY env var is required for provider anthropic");
241
- return {
242
- name: "anthropic",
243
- model,
244
- async generateVariants(seed, n) {
245
- const data = await new APIClient({
246
- displayName: "anthropic",
247
- retry: true,
248
- axios: {
249
- headers: {
250
- "x-api-key": apiKey,
251
- "anthropic-version": "2023-06-01",
252
- "Content-Type": "application/json",
253
- },
254
- },
255
- })
256
- .fetch({
257
- url: "https://api.anthropic.com/v1/messages",
258
- method: "post",
259
- timeout: 60_000,
260
- data: {
261
- model,
262
- max_tokens: 800,
263
- system: SYSTEM_PROMPT,
264
- messages: [{ role: "user", content: buildUserPrompt(seed, n) }],
265
- },
266
- })
267
- .then(pluckResponseData);
268
- const text = data.content?.find((c) => c.type === "text")?.text ?? "[]";
269
- return parseCandidates(text);
270
- },
271
- };
272
- }
273
- function parseCandidates(text) {
274
- // Strip markdown fences the model sometimes wraps around JSON
275
- const cleaned = text.replaceAll(/^```(?:json)?\n?|\n?```$/g, "").trim();
276
- const parsed = tryParsingJSON(cleaned);
277
- if (Array.isArray(parsed))
278
- return parsed;
279
- // Some providers wrap in {"variants": [...]} or {"candidates": [...]}
280
- if (typeof parsed === "object" && parsed !== null) {
281
- for (const key of ["variants", "candidates", "results"]) {
282
- const v = parsed[key];
283
- if (Array.isArray(v))
284
- return v;
285
- }
286
- }
287
- return [];
288
- }
289
- // ── Validator ─────────────────────────────────────────────────────────────
290
- /**
291
- * Components that are NEVER allowed to be dropped — keeps degenerate single-token candidates out.
292
- */
293
- const REQUIRED_COMPONENT_TAGS = new Set(["locality", "region", "street", "house_number", "venue"]);
294
- function validate(seed, candidate) {
295
- if (!candidate.raw || typeof candidate.raw !== "string")
296
- return false;
297
- if (candidate.raw.length > MAX_CANDIDATE_LENGTH)
298
- return false;
299
- if (/```|<\/?\w+>|^\s*\{/.test(candidate.raw))
300
- return false;
301
- const normRaw = foldCaseWhitespace(candidate.raw);
302
- const dropped = new Set(candidate.dropped);
303
- // LLM cannot drop required components present in the seed.
304
- for (const tag of dropped) {
305
- if (REQUIRED_COMPONENT_TAGS.has(tag) && seed.components[tag])
306
- return false;
307
- }
308
- // Every kept component value must appear verbatim (post-normalization) in the candidate raw.
309
- let keptCount = 0;
310
- for (const [tag, value] of Object.entries(seed.components)) {
311
- if (dropped.has(tag))
312
- continue;
313
- if (!value)
314
- continue;
315
- if (!normRaw.includes(foldCaseWhitespace(value)))
316
- return false;
317
- keptCount++;
318
- }
319
- // Reject degenerate single-component candidates ("VT", "Paris" alone).
320
- if (keptCount < 2)
321
- return false;
322
- return true;
323
- }
324
- // ── Main pipeline ─────────────────────────────────────────────────────────
325
- export async function expandGolden(options = {}, report) {
326
- const corpusPath = (options.corpus ?? dataRootPath("corpus", "versioned", "v0.2.0", "corpus-v0.2.0", "test", "part-0000.parquet")).toString();
327
- const count = options.count ?? 100;
328
- const variants = options.variants ?? 5;
329
- const providerName = options.provider ?? "deepseek";
330
- const model = options.model ?? (providerName === "anthropic" ? "claude-haiku-4-5-20251001" : "deepseek-chat");
331
- const concurrencyLimit = options.concurrency ?? 4;
332
- const includeSources = options.includeSources ? new Set(options.includeSources.split(",").map((s) => s.trim())) : null;
333
- const ts = new Date().toISOString().replaceAll(/[:.]/g, "-").slice(0, 19);
334
- const outputPath = options.output ?? `data/eval/golden/candidates/expand-${ts}.jsonl`;
335
- const provider = providerName === "anthropic" ? makeAnthropicProvider(model) : makeDeepseekProvider(model);
336
- report?.(`provider: ${provider.name} model: ${provider.model}`);
337
- const seeds = await loadSeeds(corpusPath, count, includeSources, report);
338
- if (!seeds.length) {
339
- throw new Error("no seeds loaded — corpus path or filter is wrong");
340
- }
341
- await makeDirectories(dirname(outputPath));
342
- const outRows = [];
343
- let kept = 0;
344
- let dropped = 0;
345
- let errored = 0;
346
- // Bounded-concurrency worker pool
347
- let cursor = 0;
348
- const workers = Array.from({ length: Math.min(concurrencyLimit, seeds.length) }, async () => {
349
- while (true) {
350
- const i = cursor++;
351
- if (i >= seeds.length)
352
- return;
353
- const seed = seeds[i];
354
- try {
355
- const candidates = await provider.generateVariants(seed, variants);
356
- for (const cand of candidates) {
357
- if (validate(seed, cand)) {
358
- const goldenCandidate = {
359
- raw: cand.raw,
360
- components: { ...seed.components },
361
- country: seed.country,
362
- source: `expand-golden:${provider.name}`,
363
- seed_source_id: seed.source_id,
364
- seed_source_adapter: seed.source,
365
- dropped_components: cand.dropped ?? [],
366
- provenance: { provider: provider.name, model: provider.model },
367
- };
368
- // Remove dropped components from the components map
369
- for (const tag of goldenCandidate.dropped_components) {
370
- // oxlint-disable-next-line typescript/no-dynamic-delete -- removing one key from a plain record; the object is not on a hot path
371
- delete goldenCandidate.components[tag];
372
- }
373
- outRows.push(goldenCandidate);
374
- kept++;
375
- }
376
- else {
377
- dropped++;
378
- }
379
- }
380
- }
381
- catch (error) {
382
- errored++;
383
- report?.(` ✗ seed ${seed.source_id}: ${error.message}`);
384
- }
385
- if ((i + 1) % 10 === 0) {
386
- report?.(` progress: ${i + 1}/${seeds.length} kept=${kept} dropped=${dropped} errored=${errored}`);
387
- }
388
- }
389
- });
390
- await Promise.all(workers);
391
- {
392
- await using out = createNewlineWriter(outputPath);
393
- for (const row of outRows) {
394
- await out.write(JSON.stringify(row));
395
- }
396
- }
397
- report?.(`=== summary ===`);
398
- report?.(`seeds processed: ${seeds.length}`);
399
- report?.(`candidates kept: ${kept}`);
400
- report?.(`candidates dropped (validator): ${dropped}`);
401
- report?.(`seeds with errors: ${errored}`);
402
- report?.(`yield: ${seeds.length ? ((kept / (seeds.length * variants)) * 100).toFixed(1) : "0"}%`);
403
- report?.(`output: ${outputPath}`);
404
- return { seedsProcessed: seeds.length, kept, dropped, errored, outputPath };
405
- }
406
- //# sourceMappingURL=golden-expand.js.map
@@ -1 +0,0 @@
1
- {"version":3,"file":"golden-expand.js","sourceRoot":"","sources":["../../lib/tools/golden-expand.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA8CG;AAEH,OAAO,EAAE,SAAS,EAAE,iBAAiB,EAAE,MAAM,qBAAqB,CAAA;AAClE,OAAO,EAAE,YAAY,EAAE,MAAM,2BAA2B,CAAA;AACxD,OAAO,EAAE,eAAe,EAAE,MAAM,4BAA4B,CAAA;AAC5D,OAAO,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAA;AACrD,OAAO,EAAE,SAAS,EAAE,MAAM,yBAAyB,CAAA;AACnD,OAAO,EAAE,kBAAkB,EAAE,MAAM,2BAA2B,CAAA;AAC9D,OAAO,EAAE,OAAO,EAAE,MAAM,SAAS,CAAA;AACjC,OAAO,EAAE,mBAAmB,EAAE,MAAM,aAAa,CAAA;AAEjD,OAAO,EAAE,QAAQ,EAAE,MAAM,MAAM,CAAA;AAC/B,OAAO,EAAE,iBAAiB,EAAE,MAAM,gBAAgB,CAAA;AAElD,6EAA6E;AAE7E;;GAEG;AACH,MAAM,oBAAoB,GAAG,GAAG,CAAA;AA+EhC,6EAA6E;AAE7E;;;;GAIG;AACH,SAAS,gBAAgB,CAAC,MAAgB,EAAE,MAAgB;IAC3D,MAAM,GAAG,GAA2B,EAAE,CAAA;IACtC,IAAI,UAAU,GAAkB,IAAI,CAAA;IACpC,IAAI,aAAa,GAAa,EAAE,CAAA;IAEhC,MAAM,KAAK,GAAG,GAAG,EAAE;QAClB,IAAI,UAAU,IAAI,aAAa,CAAC,MAAM,IAAI,CAAC,CAAC,UAAU,IAAI,GAAG,CAAC,EAAE,CAAC;YAChE,GAAG,CAAC,UAAU,CAAC,GAAG,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;QACjD,CAAC;QAED,UAAU,GAAG,IAAI,CAAA;QACjB,aAAa,GAAG,EAAE,CAAA;IACnB,CAAC,CAAA;IAED,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACxC,MAAM,KAAK,GAAG,MAAM,CAAC,CAAC,CAAE,CAAA;QACxB,MAAM,GAAG,GAAG,MAAM,CAAC,CAAC,CAAC,IAAI,EAAE,CAAA;QAE3B,IAAI,KAAK,KAAK,GAAG,EAAE,CAAC;YACnB,KAAK,EAAE,CAAA;YAEP,SAAQ;QACT,CAAC;QAED,MAAM,CAAC,MAAM,EAAE,GAAG,CAAC,GAAG,KAAK,CAAC,KAAK,CAAC,GAAG,EAAE,CAAC,CAAC,CAAA;QAEzC,IAAI,MAAM,KAAK,GAAG,IAAI,UAAU,KAAK,GAAG,EAAE,CAAC;YAC1C,KAAK,EAAE,CAAA;YACP,UAAU,GAAG,GAAI,CAAA;YACjB,aAAa,GAAG,CAAC,GAAG,CAAC,CAAA;QACtB,CAAC;aAAM,CAAC;YACP,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;QACxB,CAAC;IACF,CAAC;IAED,KAAK,EAAE,CAAA;IAEP,OAAO,GAAG,CAAA;AACX,CAAC;AAED,KAAK,UAAU,SAAS,CACvB,UAAkB,EAClB,KAAa,EACb,cAAkC,EAClC,MAA+B;IAE/B,MAAM,KAAK,GAAG,UAAU;SACtB,KAAK,CAAC,GAAG,CAAC;SACV,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;SACpB,MAAM,CAAC,SAAS,CAAC,CAAA;IAEnB,MAAM,EAAE,CAAC,sBAAsB,KAAK,CAAC,MAAM,sBAAsB,KAAK,eAAe,CAAC,CAAA;IAEtF,IAAI,cAAc,EAAE,CAAC;QACpB,MAAM,EAAE,CAAC,6BAA6B,KAAK,CAAC,IAAI,CAAC,cAAc,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAA;IAC/E,CAAC;IAED,6FAA6F;IAC7F,8FAA8F;IAC9F,8EAA8E;IAC9E,MAAM,QAAQ,GAAG,IAAI,GAAG,EAAkB,CAAA;IAC1C,MAAM,cAAc,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,KAAK,EAAE,IAAI,CAAC,CAAA;IAChD,IAAI,OAAO,GAAG,CAAC,CAAA;IACf,IAAI,qBAAqB,GAAG,CAAC,CAAA;IAE7B,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QAC1B,IAAI,KAAK,EAAE,MAAM,GAAG,IAAI,iBAAiB,CAAY,IAAI,CAAC,EAAE,CAAC;YAC5D,OAAO,EAAE,CAAA;YAET,0FAA0F;YAC1F,IAAI,cAAc,IAAI,CAAC,cAAc,CAAC,GAAG,CAAC,GAAG,CAAC,MAAM,CAAC;gBAAE,SAAQ;YAC/D,MAAM,UAAU,GAAG,gBAAgB,CAAC,GAAG,CAAC,MAAM,IAAI,EAAE,EAAE,GAAG,CAAC,MAAM,IAAI,EAAE,CAAC,CAAA;YAEvE,4FAA4F;YAC5F,IAAI,MAAM,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBACxC,qBAAqB,EAAE,CAAA;gBAEvB,SAAQ;YACT,CAAC;YAED,MAAM,IAAI,GAAS;gBAClB,GAAG,EAAE,GAAG,CAAC,GAAG;gBACZ,UAAU;gBACV,OAAO,EAAE,GAAG,CAAC,OAAO;gBACpB,MAAM,EAAE,GAAG,CAAC,MAAM;gBAClB,SAAS,EAAE,GAAG,CAAC,SAAS;aACxB,CAAA;YAED,IAAI,MAAM,GAAG,QAAQ,CAAC,GAAG,CAAC,GAAG,CAAC,MAAM,CAAC,CAAA;YAErC,IAAI,CAAC,MAAM,EAAE,CAAC;gBACb,MAAM,GAAG,EAAE,CAAA;gBACX,QAAQ,CAAC,GAAG,CAAC,GAAG,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;YACjC,CAAC;YAED,IAAI,MAAM,CAAC,MAAM,GAAG,cAAc,EAAE,CAAC;gBACpC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;YAClB,CAAC;QACF,CAAC;IACF,CAAC;IAED,MAAM,EAAE,CACP,aAAa,OAAO,gBAAgB,KAAK,CAAC,MAAM,uCAAuC,qBAAqB,EAAE,CAC9G,CAAA;IAED,MAAM,EAAE,CAAC,0BAA0B,CAAC,CAAA;IAEpC,KAAK,MAAM,CAAC,GAAG,EAAE,IAAI,CAAC,IAAI,QAAQ,EAAE,CAAC;QACpC,MAAM,EAAE,CAAC,OAAO,GAAG,KAAK,IAAI,CAAC,MAAM,EAAE,CAAC,CAAA;IACvC,CAAC;IAED,qFAAqF;IACrF,0FAA0F;IAC1F,MAAM,OAAO,GAAG,KAAK,CAAC,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAA;IACtD,MAAM,SAAS,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,GAAG,OAAO,CAAC,MAAM,CAAC,CAAA;IACpD,MAAM,SAAS,GAAG,KAAK,GAAG,SAAS,GAAG,OAAO,CAAC,MAAM,CAAA;IACpD,MAAM,MAAM,GAAW,EAAE,CAAA;IAEzB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACzC,MAAM,GAAG,GAAG,OAAO,CAAC,CAAC,CAAE,CAAA;QACvB,MAAM,IAAI,GAAG,QAAQ,CAAC,GAAG,CAAC,GAAG,CAAE,CAAA;QAC/B,MAAM,MAAM,GAAG,SAAS,GAAG,CAAC,CAAC,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAA;QAElD,8EAA8E;QAC9E,KAAK,IAAI,CAAC,GAAG,IAAI,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YAC1C,MAAM,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAC5C;YAAA,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAE,EAAE,IAAI,CAAC,CAAC,CAAE,CAAC,CAAA;QAC3C,CAAC;QAED,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,MAAM,EAAE,IAAI,CAAC,MAAM,CAAC,CAAA;QAC1C,MAAM,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAA;QAEnC,IAAI,IAAI,GAAG,MAAM,EAAE,CAAC;YACnB,MAAM,EAAE,CAAC,SAAS,GAAG,eAAe,MAAM,cAAc,IAAI,CAAC,MAAM,EAAE,CAAC,CAAA;QACvE,CAAC;IACF,CAAC;IAED,MAAM,EAAE,CAAC,cAAc,MAAM,CAAC,MAAM,iBAAiB,OAAO,CAAC,MAAM,UAAU,CAAC,CAAA;IAE9E,OAAO,MAAM,CAAA;AACd,CAAC;AAUD,MAAM,aAAa,GAAG;;;;;;;;;;;;;;;;;;;;;;kFAsB4D,CAAA;AAElF,SAAS,eAAe,CAAC,IAAU,EAAE,CAAS;IAC7C,OAAO;EACN,IAAI,CAAC,SAAS,CAAC,EAAE,GAAG,EAAE,IAAI,CAAC,GAAG,EAAE,UAAU,EAAE,IAAI,CAAC,UAAU,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC;;KAE3F,CAAC,EAAE,CAAA;AACR,CAAC;AAED,SAAS,oBAAoB,CAAC,KAAa;IAC1C,MAAM,MAAM,GAAG,QAAQ,CAAC,gBAAgB,CAAA;IAExC,IAAI,CAAC,MAAM;QAAE,MAAM,IAAI,KAAK,CAAC,4DAA4D,CAAC,CAAA;IAE1F,OAAO;QACN,IAAI,EAAE,UAAU;QAChB,KAAK;QACL,KAAK,CAAC,gBAAgB,CAAC,IAAI,EAAE,CAAC;YAC7B,MAAM,IAAI,GAAG,MAAM,IAAI,SAAS,CAAC;gBAChC,WAAW,EAAE,UAAU;gBACvB,KAAK,EAAE,IAAI;gBACX,KAAK,EAAE,EAAE,OAAO,EAAE,EAAE,aAAa,EAAE,UAAU,MAAM,EAAE,EAAE,cAAc,EAAE,kBAAkB,EAAE,EAAE;aAC7F,CAAC;iBACA,KAAK,CAAwD;gBAC7D,GAAG,EAAE,2CAA2C;gBAChD,MAAM,EAAE,MAAM;gBACd,OAAO,EAAE,MAAM;gBACf,IAAI,EAAE;oBACL,KAAK;oBACL,QAAQ,EAAE;wBACT,EAAE,IAAI,EAAE,QAAQ,EAAE,OAAO,EAAE,aAAa,EAAE;wBAC1C,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,eAAe,CAAC,IAAI,EAAE,CAAC,CAAC,EAAE;qBACnD;oBACD,eAAe,EAAE,EAAE,IAAI,EAAE,aAAa,EAAE;oBACxC,UAAU,EAAE,GAAG;oBACf,WAAW,EAAE,GAAG;iBAChB;aACD,CAAC;iBACD,IAAI,CAAC,iBAAiB,CAAC,CAAA;YAEzB,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,EAAE,OAAO,CAAC,OAAO,IAAI,IAAI,CAAA;YAE1D,OAAO,eAAe,CAAC,OAAO,CAAC,CAAA;QAChC,CAAC;KACD,CAAA;AACF,CAAC;AAED,SAAS,qBAAqB,CAAC,KAAa;IAC3C,MAAM,MAAM,GAAG,QAAQ,CAAC,iBAAiB,CAAA;IAEzC,IAAI,CAAC,MAAM;QAAE,MAAM,IAAI,KAAK,CAAC,8DAA8D,CAAC,CAAA;IAE5F,OAAO;QACN,IAAI,EAAE,WAAW;QACjB,KAAK;QACL,KAAK,CAAC,gBAAgB,CAAC,IAAI,EAAE,CAAC;YAC7B,MAAM,IAAI,GAAG,MAAM,IAAI,SAAS,CAAC;gBAChC,WAAW,EAAE,WAAW;gBACxB,KAAK,EAAE,IAAI;gBACX,KAAK,EAAE;oBACN,OAAO,EAAE;wBACR,WAAW,EAAE,MAAM;wBACnB,mBAAmB,EAAE,YAAY;wBACjC,cAAc,EAAE,kBAAkB;qBAClC;iBACD;aACD,CAAC;iBACA,KAAK,CAAsD;gBAC3D,GAAG,EAAE,uCAAuC;gBAC5C,MAAM,EAAE,MAAM;gBACd,OAAO,EAAE,MAAM;gBACf,IAAI,EAAE;oBACL,KAAK;oBACL,UAAU,EAAE,GAAG;oBACf,MAAM,EAAE,aAAa;oBACrB,QAAQ,EAAE,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,eAAe,CAAC,IAAI,EAAE,CAAC,CAAC,EAAE,CAAC;iBAC/D;aACD,CAAC;iBACD,IAAI,CAAC,iBAAiB,CAAC,CAAA;YAEzB,MAAM,IAAI,GAAG,IAAI,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,MAAM,CAAC,EAAE,IAAI,IAAI,IAAI,CAAA;YAEvE,OAAO,eAAe,CAAC,IAAI,CAAC,CAAA;QAC7B,CAAC;KACD,CAAA;AACF,CAAC;AAED,SAAS,eAAe,CAAC,IAAY;IACpC,8DAA8D;IAC9D,MAAM,OAAO,GAAG,IAAI,CAAC,UAAU,CAAC,2BAA2B,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;IAEvE,MAAM,MAAM,GAAG,cAAc,CAAC,OAAO,CAAC,CAAA;IAEtC,IAAI,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC;QAAE,OAAO,MAAqB,CAAA;IAEvD,sEAAsE;IACtE,IAAI,OAAO,MAAM,KAAK,QAAQ,IAAI,MAAM,KAAK,IAAI,EAAE,CAAC;QACnD,KAAK,MAAM,GAAG,IAAI,CAAC,UAAU,EAAE,YAAY,EAAE,SAAS,CAAC,EAAE,CAAC;YACzD,MAAM,CAAC,GAAI,MAAkC,CAAC,GAAG,CAAC,CAAA;YAElD,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC;gBAAE,OAAO,CAAgB,CAAA;QAC9C,CAAC;IACF,CAAC;IAED,OAAO,EAAE,CAAA;AACV,CAAC;AAED,6EAA6E;AAE7E;;GAEG;AACH,MAAM,uBAAuB,GAAG,IAAI,GAAG,CAAC,CAAC,UAAU,EAAE,QAAQ,EAAE,QAAQ,EAAE,cAAc,EAAE,OAAO,CAAC,CAAC,CAAA;AAElG,SAAS,QAAQ,CAAC,IAAU,EAAE,SAAoB;IACjD,IAAI,CAAC,SAAS,CAAC,GAAG,IAAI,OAAO,SAAS,CAAC,GAAG,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAA;IAErE,IAAI,SAAS,CAAC,GAAG,CAAC,MAAM,GAAG,oBAAoB;QAAE,OAAO,KAAK,CAAA;IAE7D,IAAI,qBAAqB,CAAC,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC;QAAE,OAAO,KAAK,CAAA;IAC3D,MAAM,OAAO,GAAG,kBAAkB,CAAC,SAAS,CAAC,GAAG,CAAC,CAAA;IACjD,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC,SAAS,CAAC,OAAO,CAAC,CAAA;IAE1C,2DAA2D;IAC3D,KAAK,MAAM,GAAG,IAAI,OAAO,EAAE,CAAC;QAC3B,IAAI,uBAAuB,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC;YAAE,OAAO,KAAK,CAAA;IAC3E,CAAC;IAED,6FAA6F;IAC7F,IAAI,SAAS,GAAG,CAAC,CAAA;IAEjB,KAAK,MAAM,CAAC,GAAG,EAAE,KAAK,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,IAAI,CAAC,UAAU,CAAC,EAAE,CAAC;QAC5D,IAAI,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC;YAAE,SAAQ;QAE9B,IAAI,CAAC,KAAK;YAAE,SAAQ;QAEpB,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,kBAAkB,CAAC,KAAK,CAAC,CAAC;YAAE,OAAO,KAAK,CAAA;QAE9D,SAAS,EAAE,CAAA;IACZ,CAAC;IAED,uEAAuE;IACvE,IAAI,SAAS,GAAG,CAAC;QAAE,OAAO,KAAK,CAAA;IAE/B,OAAO,IAAI,CAAA;AACZ,CAAC;AAED,6EAA6E;AAE7E,MAAM,CAAC,KAAK,UAAU,YAAY,CACjC,UAA+B,EAAE,EACjC,MAA+B;IAE/B,MAAM,UAAU,GAAG,CAClB,OAAO,CAAC,MAAM,IAAI,YAAY,CAAC,QAAQ,EAAE,WAAW,EAAE,QAAQ,EAAE,eAAe,EAAE,MAAM,EAAE,mBAAmB,CAAC,CAC7G,CAAC,QAAQ,EAAE,CAAA;IAEZ,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,IAAI,GAAG,CAAA;IAClC,MAAM,QAAQ,GAAG,OAAO,CAAC,QAAQ,IAAI,CAAC,CAAA;IACtC,MAAM,YAAY,GAAG,OAAO,CAAC,QAAQ,IAAI,UAAU,CAAA;IACnD,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,IAAI,CAAC,YAAY,KAAK,WAAW,CAAC,CAAC,CAAC,2BAA2B,CAAC,CAAC,CAAC,eAAe,CAAC,CAAA;IAC7G,MAAM,gBAAgB,GAAG,OAAO,CAAC,WAAW,IAAI,CAAC,CAAA;IACjD,MAAM,cAAc,GAAG,OAAO,CAAC,cAAc,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,OAAO,CAAC,cAAc,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAA;IACtH,MAAM,EAAE,GAAG,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC,UAAU,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAA;IACzE,MAAM,UAAU,GAAG,OAAO,CAAC,MAAM,IAAI,sCAAsC,EAAE,QAAQ,CAAA;IAErF,MAAM,QAAQ,GAAG,YAAY,KAAK,WAAW,CAAC,CAAC,CAAC,qBAAqB,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,oBAAoB,CAAC,KAAK,CAAC,CAAA;IAC1G,MAAM,EAAE,CAAC,aAAa,QAAQ,CAAC,IAAI,YAAY,QAAQ,CAAC,KAAK,EAAE,CAAC,CAAA;IAEhE,MAAM,KAAK,GAAG,MAAM,SAAS,CAAC,UAAU,EAAE,KAAK,EAAE,cAAc,EAAE,MAAM,CAAC,CAAA;IAExE,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC;QACnB,MAAM,IAAI,KAAK,CAAC,kDAAkD,CAAC,CAAA;IACpE,CAAC;IAED,MAAM,eAAe,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC,CAAA;IAC1C,MAAM,OAAO,GAAsB,EAAE,CAAA;IACrC,IAAI,IAAI,GAAG,CAAC,CAAA;IACZ,IAAI,OAAO,GAAG,CAAC,CAAA;IACf,IAAI,OAAO,GAAG,CAAC,CAAA;IAEf,kCAAkC;IAClC,IAAI,MAAM,GAAG,CAAC,CAAA;IAEd,MAAM,OAAO,GAAG,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,IAAI,CAAC,GAAG,CAAC,gBAAgB,EAAE,KAAK,CAAC,MAAM,CAAC,EAAE,EAAE,KAAK,IAAI,EAAE;QAC3F,OAAO,IAAI,EAAE,CAAC;YACb,MAAM,CAAC,GAAG,MAAM,EAAE,CAAA;YAElB,IAAI,CAAC,IAAI,KAAK,CAAC,MAAM;gBAAE,OAAM;YAC7B,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAE,CAAA;YAEtB,IAAI,CAAC;gBACJ,MAAM,UAAU,GAAG,MAAM,QAAQ,CAAC,gBAAgB,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAA;gBAElE,KAAK,MAAM,IAAI,IAAI,UAAU,EAAE,CAAC;oBAC/B,IAAI,QAAQ,CAAC,IAAI,EAAE,IAAI,CAAC,EAAE,CAAC;wBAC1B,MAAM,eAAe,GAAoB;4BACxC,GAAG,EAAE,IAAI,CAAC,GAAG;4BACb,UAAU,EAAE,EAAE,GAAG,IAAI,CAAC,UAAU,EAAE;4BAClC,OAAO,EAAE,IAAI,CAAC,OAAO;4BACrB,MAAM,EAAE,iBAAiB,QAAQ,CAAC,IAAI,EAAE;4BACxC,cAAc,EAAE,IAAI,CAAC,SAAS;4BAC9B,mBAAmB,EAAE,IAAI,CAAC,MAAM;4BAChC,kBAAkB,EAAE,IAAI,CAAC,OAAO,IAAI,EAAE;4BACtC,UAAU,EAAE,EAAE,QAAQ,EAAE,QAAQ,CAAC,IAAI,EAAE,KAAK,EAAE,QAAQ,CAAC,KAAK,EAAE;yBAC9D,CAAA;wBAED,oDAAoD;wBACpD,KAAK,MAAM,GAAG,IAAI,eAAe,CAAC,kBAAkB,EAAE,CAAC;4BACtD,iIAAiI;4BACjI,OAAO,eAAe,CAAC,UAAU,CAAC,GAAG,CAAC,CAAA;wBACvC,CAAC;wBAED,OAAO,CAAC,IAAI,CAAC,eAAe,CAAC,CAAA;wBAE7B,IAAI,EAAE,CAAA;oBACP,CAAC;yBAAM,CAAC;wBACP,OAAO,EAAE,CAAA;oBACV,CAAC;gBACF,CAAC;YACF,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBAChB,OAAO,EAAE,CAAA;gBACT,MAAM,EAAE,CAAC,YAAY,IAAI,CAAC,SAAS,KAAM,KAAe,CAAC,OAAO,EAAE,CAAC,CAAA;YACpE,CAAC;YAED,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,KAAK,CAAC,EAAE,CAAC;gBACxB,MAAM,EAAE,CAAC,eAAe,CAAC,GAAG,CAAC,IAAI,KAAK,CAAC,MAAM,UAAU,IAAI,aAAa,OAAO,aAAa,OAAO,EAAE,CAAC,CAAA;YACvG,CAAC;QACF,CAAC;IACF,CAAC,CAAC,CAAA;IAEF,MAAM,OAAO,CAAC,GAAG,CAAC,OAAO,CAAC,CAAA;IAE1B,CAAC;QACA,YAAY,GAAG,GAAG,mBAAmB,CAAC,UAAU,CAAC,CAAA;QAEjD,KAAK,MAAM,GAAG,IAAI,OAAO,EAAE,CAAC;YAC3B,MAAM,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC,CAAA;QACrC,CAAC;IACF,CAAC;IAED,MAAM,EAAE,CAAC,iBAAiB,CAAC,CAAA;IAC3B,MAAM,EAAE,CAAC,qBAAqB,KAAK,CAAC,MAAM,EAAE,CAAC,CAAA;IAC7C,MAAM,EAAE,CAAC,qBAAqB,IAAI,EAAE,CAAC,CAAA;IACrC,MAAM,EAAE,CAAC,mCAAmC,OAAO,EAAE,CAAC,CAAA;IACtD,MAAM,EAAE,CAAC,sBAAsB,OAAO,EAAE,CAAC,CAAA;IACzC,MAAM,EAAE,CAAC,UAAU,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,KAAK,CAAC,MAAM,GAAG,QAAQ,CAAC,CAAC,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,GAAG,CAAC,CAAA;IACjG,MAAM,EAAE,CAAC,qBAAqB,UAAU,EAAE,CAAC,CAAA;IAE3C,OAAO,EAAE,cAAc,EAAE,KAAK,CAAC,MAAM,EAAE,IAAI,EAAE,OAAO,EAAE,OAAO,EAAE,UAAU,EAAE,CAAA;AAC5E,CAAC"}
@@ -1,72 +0,0 @@
1
- /**
2
- * @copyright Sister Software
3
- * @license AGPL-3.0
4
- * @author Teffen Ellis, et al.
5
- *
6
- * Promote LLM-synthesized golden-set candidates into a versioned golden dir, with
7
- * human-typed-likelihood filters + dedup. Companion to `golden-expand.ts`.
8
- *
9
- * ## What it does
10
- *
11
- * 1. Reads a candidates JSONL from `data/eval/golden/candidates/`
12
- * 2. Reads the previous-version golden dir for forward-copy + dedup base
13
- * 3. Applies filters that drop candidates unlikely to be human-typed:
14
- *
15
- * - Components-glued-without-commas (5+ components but <2 separators → freeform jumble)
16
- * - Postcode-at-start with many other components (US/UK conventions put postcode last; FR puts it
17
- * before locality only)
18
- * - Suspicious-token signals (unmatched brackets, control chars, etc.)
19
- * 4. Dedupes by normalized raw (case-insensitive, whitespace-collapsed)
20
- * 5. Splits by country (US/FR/other) and writes `data/eval/golden/v<X.Y.Z>/{us,fr,other}.jsonl`
21
- * 6. Forward-copies the prior version's adversarial.jsonl + README as-is
22
- * 7. Writes MANIFEST.json with sha256 of each output file
23
- *
24
- * ## Usage
25
- *
26
- * ```sh
27
- * mailwoman corpus golden promote \
28
- * --input data/eval/golden/candidates/expand-20260518-162627.jsonl \
29
- * --bump-to v0.1.1 \
30
- * --prior v0.1.0
31
- * ```
32
- */
33
- export interface PromoteStats {
34
- candidatesIn: number;
35
- filteredOut: {
36
- glued: number;
37
- postcodeLeading: number;
38
- suspicious: number;
39
- duplicate: number;
40
- forwardDup: number;
41
- };
42
- kept: number;
43
- perCountry: Record<string, number>;
44
- }
45
- export interface PromoteGoldenOptions {
46
- /**
47
- * Candidates JSONL (required).
48
- */
49
- input: string;
50
- /**
51
- * Target golden version dir (required, e.g. `v0.1.1`).
52
- */
53
- bumpTo: string;
54
- /**
55
- * Previous version to forward-copy + dedup against. Default `v0.1.0`.
56
- */
57
- prior?: string;
58
- /**
59
- * Golden dir root. Default `data/eval/golden`.
60
- */
61
- goldenRoot?: string;
62
- /**
63
- * Skip the human-typed-likelihood filters (keep everything that passed expand-golden's validator).
64
- */
65
- noFilters?: boolean;
66
- /**
67
- * Report what would be written but don't touch disk.
68
- */
69
- dryRun?: boolean;
70
- }
71
- export declare function promoteGolden(options: PromoteGoldenOptions, report?: (line: string) => void): Promise<PromoteStats>;
72
- //# sourceMappingURL=golden-promote.d.ts.map
@@ -1 +0,0 @@
1
- {"version":3,"file":"golden-promote.d.ts","sourceRoot":"","sources":["../../lib/tools/golden-promote.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA+BG;AA4BH,MAAM,WAAW,YAAY;IAC5B,YAAY,EAAE,MAAM,CAAA;IACpB,WAAW,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,eAAe,EAAE,MAAM,CAAC;QAAC,UAAU,EAAE,MAAM,CAAC;QAAC,SAAS,EAAE,MAAM,CAAC;QAAC,UAAU,EAAE,MAAM,CAAA;KAAE,CAAA;IAClH,IAAI,EAAE,MAAM,CAAA;IACZ,UAAU,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;CAClC;AAED,MAAM,WAAW,oBAAoB;IACpC;;OAEG;IACH,KAAK,EAAE,MAAM,CAAA;IACb;;OAEG;IACH,MAAM,EAAE,MAAM,CAAA;IACd;;OAEG;IACH,KAAK,CAAC,EAAE,MAAM,CAAA;IACd;;OAEG;IACH,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB;;OAEG;IACH,SAAS,CAAC,EAAE,OAAO,CAAA;IACnB;;OAEG;IACH,MAAM,CAAC,EAAE,OAAO,CAAA;CAChB;AA0DD,wBAAsB,aAAa,CAClC,OAAO,EAAE,oBAAoB,EAC7B,MAAM,CAAC,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,IAAI,GAC7B,OAAO,CAAC,YAAY,CAAC,CAwLvB"}