@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.
- package/README.md +16 -0
- package/lib/adapters/geonames/adapter.ts +3 -3
- package/lib/adapters/geonames/postal/adapter.ts +122 -0
- package/lib/adapters/index.ts +39 -32
- package/lib/adapters/openaddresses/adapter.ts +4 -7
- package/lib/adapters/osm/adapter.ts +278 -0
- package/lib/adapters/overture/adapter.ts +22 -7
- package/lib/adapters/utils/index.ts +1 -1
- package/lib/adapters/wof/admin/json/adapter.ts +239 -0
- package/lib/adapters/wof/json-rows.ts +181 -0
- package/lib/adapters/wof/postalcode-json/adapter.ts +196 -0
- package/lib/au/adapters/gnaf/adapter.ts +152 -0
- package/lib/au/adapters/gnaf/assemble.ts +225 -0
- package/lib/cz/recipes/pcfirst-preposition.ts +112 -0
- package/lib/de/recipes/locale.ts +360 -0
- package/lib/fr/adapters/ban/adapter.ts +159 -0
- package/lib/fr/recipes/admin-split.ts +243 -0
- package/lib/fr/recipes/bare-street.ts +171 -0
- package/lib/fr/recipes/fragment.ts +334 -0
- package/lib/fr/recipes/lieudit.ts +274 -0
- package/lib/fr/recipes/order.ts +246 -0
- package/lib/fr/tools/fetch/ban.ts +289 -0
- package/lib/index.ts +1 -1
- package/lib/international/recipes/locale.ts +598 -0
- package/lib/jp/tools/fetch/houjin.ts +141 -0
- package/lib/kr/adapters/juso/label-rows.ts +259 -0
- package/lib/kr/adapters/localdata/align.ts +299 -0
- package/lib/kr/adapters/localdata/key-index.ts +185 -0
- package/lib/kr/adapters/localdata/permit-rows.ts +186 -0
- package/lib/kr/tools/fetch/juso.ts +172 -0
- package/lib/kr/tools/fetch/localdata.ts +206 -0
- package/lib/nl/recipes/postcode.ts +117 -0
- package/lib/no/recipes/fragment.ts +213 -0
- package/lib/no/recipes/street/index.ts +91 -0
- package/lib/no/recipes/street/led.ts +135 -0
- package/lib/recipes/anchor-absorption.ts +2 -2
- package/lib/recipes/bare/country.ts +95 -0
- package/lib/recipes/bare/postcode/eval.ts +99 -0
- package/lib/recipes/bare/postcode/index.ts +327 -0
- package/lib/recipes/boundary-stress.ts +3 -1
- package/lib/recipes/country-balanced.ts +3 -2
- package/lib/recipes/house-venue.ts +2 -1
- package/lib/recipes/index.ts +28 -21
- package/lib/recipes/intersection.ts +3 -2
- package/lib/recipes/po/box/cedex/au-nz.ts +82 -0
- package/lib/recipes/po/box/cedex/ca.ts +58 -0
- package/lib/recipes/po/box/cedex/fr.ts +76 -0
- package/lib/recipes/po/box/cedex/phrases.ts +108 -0
- package/lib/recipes/po/box/cedex/recipe.ts +203 -0
- package/lib/recipes/po/box/cedex/sources.ts +189 -0
- package/lib/recipes/po/box/cedex/types.ts +35 -0
- package/lib/recipes/po/box/cedex/us.ts +69 -0
- package/lib/recipes/po/box/cedex/vocabulary.ts +106 -0
- package/lib/recipes/po/box/cedex.ts +9 -0
- package/lib/recipes/po/box/index.ts +128 -0
- package/lib/recipes/scaffold.ts +18 -42
- package/lib/recipes/street/affix.ts +807 -0
- package/lib/recipes/street/bare.ts +102 -0
- package/lib/recipes/street.ts +3 -1
- package/lib/recipes/sub/venue/context.ts +149 -0
- package/lib/recipes/sub/venue/index.ts +997 -0
- package/lib/recipes/sub/venue/sources.ts +739 -0
- package/lib/recipes/trailing-region.ts +9 -9
- package/lib/recipes/unit.ts +3 -3
- package/lib/sg/recipes/register.ts +270 -0
- package/lib/sg/tools/fetch/acra.ts +179 -0
- package/lib/si/recipes/bare-village.ts +99 -0
- package/lib/south-asia/recipes/register.ts +214 -0
- package/lib/synthesizers/anchor-absorption.ts +1 -1
- package/lib/synthesizers/locale.ts +261 -0
- package/lib/synthesizers/street.ts +1 -1
- package/lib/synthesizers/utils.ts +2 -2
- package/lib/tools/audit.ts +3 -2
- package/lib/tools/corpus-stats.ts +3 -3
- package/lib/tools/fetch/download/index.ts +10 -0
- package/lib/tools/fetch/download/manifest.ts +102 -0
- package/lib/tools/fetch/download/network.ts +226 -0
- package/lib/tools/fetch/download/retry.ts +173 -0
- package/lib/tools/fetch/geonames/dump.ts +257 -0
- package/lib/tools/fetch/geonames/postal.ts +180 -0
- package/lib/tools/fetch/index.ts +47 -20
- package/lib/tools/fetch/openaddresses.ts +2 -2
- package/lib/tools/fetch/ourairports.ts +2 -2
- package/lib/tools/fetch/ppd.ts +1 -3
- package/lib/tools/fetch/wikidata-subvenue.ts +2 -2
- package/lib/tools/golden/promote.ts +340 -0
- package/lib/tools/golden/relabel-street/directory.ts +340 -0
- package/lib/tools/golden/relabel-street/index.ts +9 -0
- package/lib/tools/golden/relabel-street/row.ts +361 -0
- package/lib/tools/index.ts +10 -10
- package/lib/tools/ingest-csv.ts +5 -7
- package/lib/tools/lint/slice/vocab.ts +376 -0
- package/lib/tools/overlay-manifest.ts +92 -25
- package/lib/tools/overture-subvenue.ts +1 -1
- package/lib/tools/postcode-triples.ts +138 -14
- package/lib/tools/sub/venue/harvest.ts +234 -0
- package/lib/tools/sub/venue/head-nouns.ts +245 -0
- package/lib/tools/sub/venue/lexicon.ts +429 -0
- package/lib/tools/sub/venue/surfaces.ts +95 -0
- package/lib/tools/sub/venue/table.ts +281 -0
- package/lib/tools/sub/venue/wikidata.ts +85 -0
- package/lib/tw/tools/fetch/gcis.ts +180 -0
- package/lib/types.ts +1 -1
- package/lib/us/adapters/fcc-bdc/adapter.ts +173 -0
- package/lib/us/adapters/state/hi-schools/adapter.ts +160 -0
- package/lib/us/adapters/state/ia-contractors/adapter.ts +136 -0
- package/lib/us/adapters/state/ny-notaries/adapter.ts +133 -0
- package/lib/us/adapters/state/tx-notaries/adapter.ts +149 -0
- package/lib/us/adapters/tiger/adapter.ts +250 -0
- package/lib/us/adapters/usgov/hrsa-fqhc/adapter.ts +156 -0
- package/lib/us/adapters/usgov/imls-pls/adapter.ts +134 -0
- package/lib/us/adapters/usgov/irs-bmf/adapter.ts +145 -0
- package/lib/us/adapters/usgov/nad/adapter.ts +324 -0
- package/lib/us/adapters/usgov/nppes/adapter.ts +143 -0
- package/lib/us/adapters/usgov/samhsa-treatment-locator/adapter.ts +182 -0
- package/lib/us/tools/fetch/hrsa.ts +59 -0
- package/lib/us/tools/fetch/imls-pls.ts +149 -0
- package/lib/us/tools/fetch/nad.ts +341 -0
- package/lib/us/tools/fetch/nppes.ts +165 -0
- package/lib/us/tools/fetch/state/hi-schools.ts +161 -0
- package/lib/us/tools/fetch/state/sources.ts +146 -0
- package/lib/us/tools/fetch/tiger-full.ts +308 -0
- package/lib/utils/align.ts +1 -1
- package/lib/utils/golden.ts +12 -8
- package/lib/utils/wof-json.ts +7 -13
- package/out/adapters/geonames/adapter.d.ts +1 -1
- package/out/adapters/geonames/adapter.js +3 -3
- package/out/adapters/geonames/adapter.js.map +1 -1
- package/out/adapters/geonames/postal/adapter.d.ts.map +1 -0
- package/out/adapters/geonames/postal/adapter.js +107 -0
- package/out/adapters/geonames/postal/adapter.js.map +1 -0
- package/out/adapters/index.d.ts +16 -16
- package/out/adapters/index.d.ts.map +1 -1
- package/out/adapters/index.js +34 -32
- package/out/adapters/index.js.map +1 -1
- package/out/adapters/openaddresses/adapter.d.ts.map +1 -1
- package/out/adapters/openaddresses/adapter.js +4 -6
- package/out/adapters/openaddresses/adapter.js.map +1 -1
- package/out/adapters/osm/adapter.d.ts +95 -0
- package/out/adapters/osm/adapter.d.ts.map +1 -0
- package/out/adapters/osm/adapter.js +211 -0
- package/out/adapters/osm/adapter.js.map +1 -0
- package/out/adapters/overture/adapter.d.ts +10 -0
- package/out/adapters/overture/adapter.d.ts.map +1 -1
- package/out/adapters/overture/adapter.js +20 -7
- package/out/adapters/overture/adapter.js.map +1 -1
- package/out/adapters/synth-po-box/adapter.js.map +1 -1
- package/out/adapters/utils/index.d.ts +1 -1
- package/out/adapters/utils/index.d.ts.map +1 -1
- package/out/adapters/utils/index.js.map +1 -1
- package/out/adapters/wof/admin/json/adapter.d.ts +79 -0
- package/out/adapters/wof/admin/json/adapter.d.ts.map +1 -0
- package/out/adapters/wof/admin/json/adapter.js +204 -0
- package/out/adapters/wof/admin/json/adapter.js.map +1 -0
- package/out/adapters/wof/json-rows.d.ts +73 -0
- package/out/adapters/wof/json-rows.d.ts.map +1 -0
- package/out/adapters/wof/json-rows.js +115 -0
- package/out/adapters/wof/json-rows.js.map +1 -0
- package/out/adapters/wof/postalcode-json/adapter.d.ts +65 -0
- package/out/adapters/wof/postalcode-json/adapter.d.ts.map +1 -0
- package/out/adapters/wof/postalcode-json/adapter.js +167 -0
- package/out/adapters/wof/postalcode-json/adapter.js.map +1 -0
- package/out/au/adapters/gnaf/adapter.d.ts.map +1 -0
- package/out/au/adapters/gnaf/adapter.js +127 -0
- package/out/au/adapters/gnaf/adapter.js.map +1 -0
- package/out/au/adapters/gnaf/assemble.d.ts +65 -0
- package/out/au/adapters/gnaf/assemble.d.ts.map +1 -0
- package/out/au/adapters/gnaf/assemble.js +154 -0
- package/out/au/adapters/gnaf/assemble.js.map +1 -0
- package/out/cn/recipes/organizational-units.d.ts +36 -0
- package/out/cn/recipes/organizational-units.d.ts.map +1 -0
- package/out/cn/recipes/organizational-units.js +159 -0
- package/out/cn/recipes/organizational-units.js.map +1 -0
- package/out/cz/recipes/pcfirst-preposition.d.ts +22 -0
- package/out/cz/recipes/pcfirst-preposition.d.ts.map +1 -0
- package/out/cz/recipes/pcfirst-preposition.js +101 -0
- package/out/cz/recipes/pcfirst-preposition.js.map +1 -0
- package/out/de/recipes/locale.d.ts +44 -0
- package/out/de/recipes/locale.d.ts.map +1 -0
- package/out/de/recipes/locale.js +289 -0
- package/out/de/recipes/locale.js.map +1 -0
- package/out/env.d.ts +1 -1
- package/out/fr/adapters/ban/adapter.d.ts.map +1 -0
- package/out/fr/adapters/ban/adapter.js +123 -0
- package/out/fr/adapters/ban/adapter.js.map +1 -0
- package/out/fr/adapters/ban/street-decompose.d.ts.map +1 -0
- package/out/fr/adapters/ban/street-decompose.js.map +1 -0
- package/out/fr/recipes/admin-split.d.ts +38 -0
- package/out/fr/recipes/admin-split.d.ts.map +1 -0
- package/out/fr/recipes/admin-split.js +192 -0
- package/out/fr/recipes/admin-split.js.map +1 -0
- package/out/fr/recipes/bare-street.d.ts +28 -0
- package/out/fr/recipes/bare-street.d.ts.map +1 -0
- package/out/fr/recipes/bare-street.js +144 -0
- package/out/fr/recipes/bare-street.js.map +1 -0
- package/out/fr/recipes/fragment.d.ts +77 -0
- package/out/fr/recipes/fragment.d.ts.map +1 -0
- package/out/fr/recipes/fragment.js +277 -0
- package/out/fr/recipes/fragment.js.map +1 -0
- package/out/fr/recipes/lieudit.d.ts +32 -0
- package/out/fr/recipes/lieudit.d.ts.map +1 -0
- package/out/fr/recipes/lieudit.js +211 -0
- package/out/fr/recipes/lieudit.js.map +1 -0
- package/out/fr/recipes/order.d.ts +38 -0
- package/out/fr/recipes/order.d.ts.map +1 -0
- package/out/fr/recipes/order.js +198 -0
- package/out/fr/recipes/order.js.map +1 -0
- package/out/fr/tools/fetch/ban.d.ts +31 -0
- package/out/fr/tools/fetch/ban.d.ts.map +1 -0
- package/out/fr/tools/fetch/ban.js +239 -0
- package/out/fr/tools/fetch/ban.js.map +1 -0
- package/out/index.d.ts +1 -1
- package/out/index.d.ts.map +1 -1
- package/out/index.js +1 -1
- package/out/index.js.map +1 -1
- package/out/international/recipes/locale.d.ts +136 -0
- package/out/international/recipes/locale.d.ts.map +1 -0
- package/out/international/recipes/locale.js +458 -0
- package/out/international/recipes/locale.js.map +1 -0
- package/out/jp/adapters/wof-admin/adapter.d.ts.map +1 -0
- package/out/jp/adapters/wof-admin/adapter.js.map +1 -0
- package/out/jp/tools/fetch/houjin.d.ts +24 -0
- package/out/jp/tools/fetch/houjin.d.ts.map +1 -0
- package/out/jp/tools/fetch/houjin.js +102 -0
- package/out/jp/tools/fetch/houjin.js.map +1 -0
- package/out/kr/adapters/juso/label-rows.d.ts +76 -0
- package/out/kr/adapters/juso/label-rows.d.ts.map +1 -0
- package/out/kr/adapters/juso/label-rows.js +194 -0
- package/out/kr/adapters/juso/label-rows.js.map +1 -0
- package/out/kr/adapters/localdata/align.d.ts +43 -0
- package/out/kr/adapters/localdata/align.d.ts.map +1 -0
- package/out/kr/adapters/localdata/align.js +223 -0
- package/out/kr/adapters/localdata/align.js.map +1 -0
- package/out/kr/adapters/localdata/key-index.d.ts +71 -0
- package/out/kr/adapters/localdata/key-index.d.ts.map +1 -0
- package/out/kr/adapters/localdata/key-index.js +133 -0
- package/out/kr/adapters/localdata/key-index.js.map +1 -0
- package/out/kr/adapters/localdata/permit-rows.d.ts +66 -0
- package/out/kr/adapters/localdata/permit-rows.d.ts.map +1 -0
- package/out/kr/adapters/localdata/permit-rows.js +131 -0
- package/out/kr/adapters/localdata/permit-rows.js.map +1 -0
- package/out/kr/tools/fetch/juso.d.ts +30 -0
- package/out/kr/tools/fetch/juso.d.ts.map +1 -0
- package/out/kr/tools/fetch/juso.js +120 -0
- package/out/kr/tools/fetch/juso.js.map +1 -0
- package/out/kr/tools/fetch/localdata.d.ts +30 -0
- package/out/kr/tools/fetch/localdata.d.ts.map +1 -0
- package/out/kr/tools/fetch/localdata.js +139 -0
- package/out/kr/tools/fetch/localdata.js.map +1 -0
- package/out/nl/recipes/postcode.d.ts +26 -0
- package/out/nl/recipes/postcode.d.ts.map +1 -0
- package/out/nl/recipes/postcode.js +103 -0
- package/out/nl/recipes/postcode.js.map +1 -0
- package/out/no/recipes/fragment.d.ts +43 -0
- package/out/no/recipes/fragment.d.ts.map +1 -0
- package/out/no/recipes/fragment.js +183 -0
- package/out/no/recipes/fragment.js.map +1 -0
- package/out/no/recipes/street/index.d.ts +18 -0
- package/out/no/recipes/street/index.d.ts.map +1 -0
- package/out/no/recipes/street/index.js +72 -0
- package/out/no/recipes/street/index.js.map +1 -0
- package/out/no/recipes/street/led.d.ts +24 -0
- package/out/no/recipes/street/led.d.ts.map +1 -0
- package/out/no/recipes/street/led.js +114 -0
- package/out/no/recipes/street/led.js.map +1 -0
- package/out/recipes/anchor-absorption.d.ts.map +1 -1
- package/out/recipes/anchor-absorption.js +2 -1
- package/out/recipes/anchor-absorption.js.map +1 -1
- package/out/recipes/bare/country.d.ts +25 -0
- package/out/recipes/bare/country.d.ts.map +1 -0
- package/out/recipes/bare/country.js +88 -0
- package/out/recipes/bare/country.js.map +1 -0
- package/out/recipes/bare/postcode/eval.d.ts +29 -0
- package/out/recipes/bare/postcode/eval.d.ts.map +1 -0
- package/out/recipes/bare/postcode/eval.js +85 -0
- package/out/recipes/bare/postcode/eval.js.map +1 -0
- package/out/recipes/bare/postcode/index.d.ts +62 -0
- package/out/recipes/bare/postcode/index.d.ts.map +1 -0
- package/out/recipes/bare/postcode/index.js +267 -0
- package/out/recipes/bare/postcode/index.js.map +1 -0
- package/out/recipes/boundary-stress.d.ts.map +1 -1
- package/out/recipes/boundary-stress.js +2 -1
- package/out/recipes/boundary-stress.js.map +1 -1
- package/out/recipes/country-balanced.d.ts.map +1 -1
- package/out/recipes/country-balanced.js +2 -1
- package/out/recipes/country-balanced.js.map +1 -1
- package/out/recipes/house-venue.d.ts.map +1 -1
- package/out/recipes/house-venue.js +2 -1
- package/out/recipes/house-venue.js.map +1 -1
- package/out/recipes/index.d.ts.map +1 -1
- package/out/recipes/index.js +28 -21
- package/out/recipes/index.js.map +1 -1
- package/out/recipes/intersection.d.ts.map +1 -1
- package/out/recipes/intersection.js +2 -1
- package/out/recipes/intersection.js.map +1 -1
- package/out/recipes/po/box/cedex/au-nz.d.ts +7 -0
- package/out/recipes/po/box/cedex/au-nz.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/au-nz.js +56 -0
- package/out/recipes/po/box/cedex/au-nz.js.map +1 -0
- package/out/recipes/po/box/cedex/ca.d.ts +7 -0
- package/out/recipes/po/box/cedex/ca.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/ca.js +38 -0
- package/out/recipes/po/box/cedex/ca.js.map +1 -0
- package/out/recipes/po/box/cedex/fr.d.ts +7 -0
- package/out/recipes/po/box/cedex/fr.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/fr.js +57 -0
- package/out/recipes/po/box/cedex/fr.js.map +1 -0
- package/out/recipes/po/box/cedex/phrases.d.ts +8 -0
- package/out/recipes/po/box/cedex/phrases.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/phrases.js +76 -0
- package/out/recipes/po/box/cedex/phrases.js.map +1 -0
- package/out/recipes/po/box/cedex/recipe.d.ts +9 -0
- package/out/recipes/po/box/cedex/recipe.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/recipe.js +153 -0
- package/out/recipes/po/box/cedex/recipe.js.map +1 -0
- package/out/recipes/po/box/cedex/sources.d.ts +86 -0
- package/out/recipes/po/box/cedex/sources.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/sources.js +143 -0
- package/out/recipes/po/box/cedex/sources.js.map +1 -0
- package/out/recipes/po/box/cedex/types.d.ts +31 -0
- package/out/recipes/po/box/cedex/types.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/types.js +5 -0
- package/out/recipes/po/box/cedex/types.js.map +1 -0
- package/out/recipes/po/box/cedex/us.d.ts +7 -0
- package/out/recipes/po/box/cedex/us.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/us.js +51 -0
- package/out/recipes/po/box/cedex/us.js.map +1 -0
- package/out/recipes/po/box/cedex/vocabulary.d.ts +68 -0
- package/out/recipes/po/box/cedex/vocabulary.d.ts.map +1 -0
- package/out/recipes/po/box/cedex/vocabulary.js +90 -0
- package/out/recipes/po/box/cedex/vocabulary.js.map +1 -0
- package/out/recipes/po/box/cedex.d.ts +9 -0
- package/out/recipes/po/box/cedex.d.ts.map +1 -0
- package/out/recipes/po/box/cedex.js +9 -0
- package/out/recipes/po/box/cedex.js.map +1 -0
- package/out/recipes/po/box/index.d.ts +17 -0
- package/out/recipes/po/box/index.d.ts.map +1 -0
- package/out/recipes/po/box/index.js +105 -0
- package/out/recipes/po/box/index.js.map +1 -0
- package/out/recipes/scaffold.d.ts +20 -1
- package/out/recipes/scaffold.d.ts.map +1 -1
- package/out/recipes/scaffold.js +18 -49
- package/out/recipes/scaffold.js.map +1 -1
- package/out/recipes/street/affix.d.ts +89 -0
- package/out/recipes/street/affix.d.ts.map +1 -0
- package/out/recipes/street/affix.js +565 -0
- package/out/recipes/street/affix.js.map +1 -0
- package/out/recipes/street/bare.d.ts +25 -0
- package/out/recipes/street/bare.d.ts.map +1 -0
- package/out/recipes/street/bare.js +87 -0
- package/out/recipes/street/bare.js.map +1 -0
- package/out/recipes/street.d.ts.map +1 -1
- package/out/recipes/street.js +2 -1
- package/out/recipes/street.js.map +1 -1
- package/out/recipes/sub/venue/context.d.ts +29 -0
- package/out/recipes/sub/venue/context.d.ts.map +1 -0
- package/out/recipes/sub/venue/context.js +106 -0
- package/out/recipes/sub/venue/context.js.map +1 -0
- package/out/recipes/sub/venue/index.d.ts +270 -0
- package/out/recipes/sub/venue/index.d.ts.map +1 -0
- package/out/recipes/sub/venue/index.js +728 -0
- package/out/recipes/sub/venue/index.js.map +1 -0
- package/out/recipes/sub/venue/sources.d.ts +209 -0
- package/out/recipes/sub/venue/sources.d.ts.map +1 -0
- package/out/recipes/sub/venue/sources.js +555 -0
- package/out/recipes/sub/venue/sources.js.map +1 -0
- package/out/recipes/trailing-region.d.ts.map +1 -1
- package/out/recipes/trailing-region.js +8 -2
- package/out/recipes/trailing-region.js.map +1 -1
- package/out/recipes/unit.d.ts.map +1 -1
- package/out/recipes/unit.js +2 -1
- package/out/recipes/unit.js.map +1 -1
- package/out/sg/recipes/register.d.ts +69 -0
- package/out/sg/recipes/register.d.ts.map +1 -0
- package/out/sg/recipes/register.js +211 -0
- package/out/sg/recipes/register.js.map +1 -0
- package/out/sg/tools/fetch/acra.d.ts +23 -0
- package/out/sg/tools/fetch/acra.d.ts.map +1 -0
- package/out/sg/tools/fetch/acra.js +126 -0
- package/out/sg/tools/fetch/acra.js.map +1 -0
- package/out/si/recipes/bare-village.d.ts +32 -0
- package/out/si/recipes/bare-village.d.ts.map +1 -0
- package/out/si/recipes/bare-village.js +89 -0
- package/out/si/recipes/bare-village.js.map +1 -0
- package/out/south-asia/recipes/register.d.ts +51 -0
- package/out/south-asia/recipes/register.d.ts.map +1 -0
- package/out/south-asia/recipes/register.js +154 -0
- package/out/south-asia/recipes/register.js.map +1 -0
- package/out/synthesizers/anchor-absorption.d.ts +1 -1
- package/out/synthesizers/anchor-absorption.d.ts.map +1 -1
- package/out/synthesizers/anchor-absorption.js.map +1 -1
- package/out/synthesizers/boundary-stress.js.map +1 -1
- package/out/synthesizers/house-venue.js.map +1 -1
- package/out/synthesizers/intersection.js.map +1 -1
- package/out/synthesizers/locale.d.ts +116 -0
- package/out/synthesizers/locale.d.ts.map +1 -0
- package/out/synthesizers/locale.js +150 -0
- package/out/synthesizers/locale.js.map +1 -0
- package/out/synthesizers/no-street.d.ts.map +1 -1
- package/out/synthesizers/no-street.js.map +1 -1
- package/out/synthesizers/po-box.js.map +1 -1
- package/out/synthesizers/street.d.ts.map +1 -1
- package/out/synthesizers/street.js +1 -1
- package/out/synthesizers/street.js.map +1 -1
- package/out/synthesizers/utils.js +1 -1
- package/out/synthesizers/utils.js.map +1 -1
- package/out/test-kit/corpus-recipe.d.ts.map +1 -1
- package/out/test-kit/corpus-recipe.js.map +1 -1
- package/out/tools/audit.d.ts.map +1 -1
- package/out/tools/audit.js +3 -2
- package/out/tools/audit.js.map +1 -1
- package/out/tools/corpus-stats.js +3 -3
- package/out/tools/corpus-stats.js.map +1 -1
- package/out/tools/fetch/download/index.d.ts +10 -0
- package/out/tools/fetch/download/index.d.ts.map +1 -0
- package/out/tools/fetch/download/index.js +10 -0
- package/out/tools/fetch/download/index.js.map +1 -0
- package/out/tools/fetch/download/manifest.d.ts +47 -0
- package/out/tools/fetch/download/manifest.d.ts.map +1 -0
- package/out/tools/fetch/download/manifest.js +74 -0
- package/out/tools/fetch/download/manifest.js.map +1 -0
- package/out/tools/fetch/download/network.d.ts +105 -0
- package/out/tools/fetch/download/network.d.ts.map +1 -0
- package/out/tools/fetch/download/network.js +130 -0
- package/out/tools/fetch/download/network.js.map +1 -0
- package/out/tools/fetch/download/retry.d.ts +45 -0
- package/out/tools/fetch/download/retry.d.ts.map +1 -0
- package/out/tools/fetch/download/retry.js +136 -0
- package/out/tools/fetch/download/retry.js.map +1 -0
- package/out/tools/fetch/geonames/dump.d.ts +91 -0
- package/out/tools/fetch/geonames/dump.d.ts.map +1 -0
- package/out/tools/fetch/geonames/dump.js +171 -0
- package/out/tools/fetch/geonames/dump.js.map +1 -0
- package/out/tools/fetch/geonames/postal.d.ts +76 -0
- package/out/tools/fetch/geonames/postal.d.ts.map +1 -0
- package/out/tools/fetch/geonames/postal.js +126 -0
- package/out/tools/fetch/geonames/postal.js.map +1 -0
- package/out/tools/fetch/index.d.ts +47 -20
- package/out/tools/fetch/index.d.ts.map +1 -1
- package/out/tools/fetch/index.js +47 -20
- package/out/tools/fetch/index.js.map +1 -1
- package/out/tools/fetch/openaddresses.d.ts +1 -1
- package/out/tools/fetch/openaddresses.d.ts.map +1 -1
- package/out/tools/fetch/openaddresses.js +1 -1
- package/out/tools/fetch/openaddresses.js.map +1 -1
- package/out/tools/fetch/ourairports.d.ts +1 -1
- package/out/tools/fetch/ourairports.d.ts.map +1 -1
- package/out/tools/fetch/ourairports.js +1 -1
- package/out/tools/fetch/ourairports.js.map +1 -1
- package/out/tools/fetch/ppd.js +1 -3
- package/out/tools/fetch/ppd.js.map +1 -1
- package/out/tools/fetch/wikidata-subvenue.d.ts +1 -1
- package/out/tools/fetch/wikidata-subvenue.d.ts.map +1 -1
- package/out/tools/fetch/wikidata-subvenue.js +1 -1
- package/out/tools/fetch/wikidata-subvenue.js.map +1 -1
- package/out/tools/golden/expand.d.ts +90 -0
- package/out/tools/golden/expand.d.ts.map +1 -0
- package/out/tools/golden/expand.js +406 -0
- package/out/tools/golden/expand.js.map +1 -0
- package/out/tools/golden/promote.d.ts +72 -0
- package/out/tools/golden/promote.d.ts.map +1 -0
- package/out/tools/golden/promote.js +243 -0
- package/out/tools/golden/promote.js.map +1 -0
- package/out/tools/golden/relabel-street/directory.d.ts +83 -0
- package/out/tools/golden/relabel-street/directory.d.ts.map +1 -0
- package/out/tools/golden/relabel-street/directory.js +216 -0
- package/out/tools/golden/relabel-street/directory.js.map +1 -0
- package/out/tools/golden/relabel-street/index.d.ts +9 -0
- package/out/tools/golden/relabel-street/index.d.ts.map +1 -0
- package/out/tools/golden/relabel-street/index.js +9 -0
- package/out/tools/golden/relabel-street/index.js.map +1 -0
- package/out/tools/golden/relabel-street/row.d.ts +121 -0
- package/out/tools/golden/relabel-street/row.d.ts.map +1 -0
- package/out/tools/golden/relabel-street/row.js +232 -0
- package/out/tools/golden/relabel-street/row.js.map +1 -0
- package/out/tools/index.d.ts +10 -10
- package/out/tools/index.d.ts.map +1 -1
- package/out/tools/index.js +10 -10
- package/out/tools/index.js.map +1 -1
- package/out/tools/ingest-csv.d.ts.map +1 -1
- package/out/tools/ingest-csv.js +5 -7
- package/out/tools/ingest-csv.js.map +1 -1
- package/out/tools/lint/slice/index.d.ts +93 -0
- package/out/tools/lint/slice/index.d.ts.map +1 -0
- package/out/tools/lint/slice/index.js +322 -0
- package/out/tools/lint/slice/index.js.map +1 -0
- package/out/tools/lint/slice/vocab.d.ts +90 -0
- package/out/tools/lint/slice/vocab.d.ts.map +1 -0
- package/out/tools/lint/slice/vocab.js +254 -0
- package/out/tools/lint/slice/vocab.js.map +1 -0
- package/out/tools/overlay-manifest.d.ts +56 -6
- package/out/tools/overlay-manifest.d.ts.map +1 -1
- package/out/tools/overlay-manifest.js +56 -15
- package/out/tools/overlay-manifest.js.map +1 -1
- package/out/tools/overture-subvenue.d.ts +1 -1
- package/out/tools/postcode-triples.d.ts +36 -0
- package/out/tools/postcode-triples.d.ts.map +1 -1
- package/out/tools/postcode-triples.js +91 -12
- package/out/tools/postcode-triples.js.map +1 -1
- package/out/tools/slice/kryptonite.d.ts +32 -0
- package/out/tools/slice/kryptonite.d.ts.map +1 -0
- package/out/tools/slice/kryptonite.js +99 -0
- package/out/tools/slice/kryptonite.js.map +1 -0
- package/out/tools/slice/translit.d.ts +51 -0
- package/out/tools/slice/translit.d.ts.map +1 -0
- package/out/tools/slice/translit.js +166 -0
- package/out/tools/slice/translit.js.map +1 -0
- package/out/tools/sub/venue/harvest.d.ts +100 -0
- package/out/tools/sub/venue/harvest.d.ts.map +1 -0
- package/out/tools/sub/venue/harvest.js +168 -0
- package/out/tools/sub/venue/harvest.js.map +1 -0
- package/out/tools/sub/venue/head-nouns.d.ts +50 -0
- package/out/tools/sub/venue/head-nouns.d.ts.map +1 -0
- package/out/tools/sub/venue/head-nouns.js.map +1 -0
- package/out/tools/sub/venue/lexicon.d.ts +156 -0
- package/out/tools/sub/venue/lexicon.d.ts.map +1 -0
- package/out/tools/sub/venue/lexicon.js +283 -0
- package/out/tools/sub/venue/lexicon.js.map +1 -0
- package/out/tools/sub/venue/promotions.d.ts +94 -0
- package/out/tools/sub/venue/promotions.d.ts.map +1 -0
- package/out/tools/sub/venue/promotions.js +266 -0
- package/out/tools/sub/venue/promotions.js.map +1 -0
- package/out/tools/sub/venue/surfaces.d.ts +45 -0
- package/out/tools/sub/venue/surfaces.d.ts.map +1 -0
- package/out/tools/sub/venue/surfaces.js.map +1 -0
- package/out/tools/sub/venue/table.d.ts +229 -0
- package/out/tools/sub/venue/table.d.ts.map +1 -0
- package/out/tools/sub/venue/table.js.map +1 -0
- package/out/tools/sub/venue/wikidata.d.ts +26 -0
- package/out/tools/sub/venue/wikidata.d.ts.map +1 -0
- package/out/tools/sub/venue/wikidata.js +62 -0
- package/out/tools/sub/venue/wikidata.js.map +1 -0
- package/out/tw/tools/fetch/gcis.d.ts +23 -0
- package/out/tw/tools/fetch/gcis.d.ts.map +1 -0
- package/out/tw/tools/fetch/gcis.js +135 -0
- package/out/tw/tools/fetch/gcis.js.map +1 -0
- package/out/types.d.ts +1 -1
- package/out/types.d.ts.map +1 -1
- package/out/us/adapters/fcc-bdc/adapter.d.ts.map +1 -0
- package/out/us/adapters/fcc-bdc/adapter.js +142 -0
- package/out/us/adapters/fcc-bdc/adapter.js.map +1 -0
- package/out/us/adapters/state/hi-schools/adapter.d.ts.map +1 -0
- package/out/us/adapters/state/hi-schools/adapter.js +136 -0
- package/out/us/adapters/state/hi-schools/adapter.js.map +1 -0
- package/out/us/adapters/state/hi-schools/workbook.d.ts.map +1 -0
- package/out/us/adapters/state/hi-schools/workbook.js.map +1 -0
- package/out/us/adapters/state/ia-contractors/adapter.d.ts.map +1 -0
- package/out/us/adapters/state/ia-contractors/adapter.js +104 -0
- package/out/us/adapters/state/ia-contractors/adapter.js.map +1 -0
- package/out/us/adapters/state/ny-notaries/adapter.d.ts.map +1 -0
- package/out/us/adapters/state/ny-notaries/adapter.js +112 -0
- package/out/us/adapters/state/ny-notaries/adapter.js.map +1 -0
- package/out/us/adapters/state/tx-notaries/adapter.d.ts.map +1 -0
- package/out/us/adapters/state/tx-notaries/adapter.js +118 -0
- package/out/us/adapters/state/tx-notaries/adapter.js.map +1 -0
- package/out/us/adapters/tiger/adapter.d.ts +55 -0
- package/out/us/adapters/tiger/adapter.d.ts.map +1 -0
- package/out/us/adapters/tiger/adapter.js +197 -0
- package/out/us/adapters/tiger/adapter.js.map +1 -0
- package/out/us/adapters/tiger/street-decompose.d.ts.map +1 -0
- package/out/us/adapters/tiger/street-decompose.js.map +1 -0
- package/out/us/adapters/usgov/hrsa-fqhc/adapter.d.ts.map +1 -0
- package/out/us/adapters/usgov/hrsa-fqhc/adapter.js +121 -0
- package/out/us/adapters/usgov/hrsa-fqhc/adapter.js.map +1 -0
- package/out/us/adapters/usgov/imls-pls/adapter.d.ts.map +1 -0
- package/out/us/adapters/usgov/imls-pls/adapter.js +106 -0
- package/out/us/adapters/usgov/imls-pls/adapter.js.map +1 -0
- package/out/us/adapters/usgov/irs-bmf/adapter.d.ts.map +1 -0
- package/out/us/adapters/usgov/irs-bmf/adapter.js +116 -0
- package/out/us/adapters/usgov/irs-bmf/adapter.js.map +1 -0
- package/out/us/adapters/usgov/nad/adapter.d.ts.map +1 -0
- package/out/us/adapters/usgov/nad/adapter.js +239 -0
- package/out/us/adapters/usgov/nad/adapter.js.map +1 -0
- package/out/us/adapters/usgov/nppes/adapter.d.ts.map +1 -0
- package/out/us/adapters/usgov/nppes/adapter.js +110 -0
- package/out/us/adapters/usgov/nppes/adapter.js.map +1 -0
- package/out/us/adapters/usgov/samhsa-treatment-locator/adapter.d.ts.map +1 -0
- package/out/us/adapters/usgov/samhsa-treatment-locator/adapter.js +143 -0
- package/out/us/adapters/usgov/samhsa-treatment-locator/adapter.js.map +1 -0
- package/out/us/fips-state.d.ts +41 -0
- package/out/us/fips-state.d.ts.map +1 -0
- package/out/us/fips-state.js +100 -0
- package/out/us/fips-state.js.map +1 -0
- package/out/us/tools/fetch/hrsa.d.ts +18 -0
- package/out/us/tools/fetch/hrsa.d.ts.map +1 -0
- package/out/us/tools/fetch/hrsa.js +46 -0
- package/out/us/tools/fetch/hrsa.js.map +1 -0
- package/out/us/tools/fetch/imls-pls.d.ts +25 -0
- package/out/us/tools/fetch/imls-pls.d.ts.map +1 -0
- package/out/us/tools/fetch/imls-pls.js +108 -0
- package/out/us/tools/fetch/imls-pls.js.map +1 -0
- package/out/us/tools/fetch/nad.d.ts +70 -0
- package/out/us/tools/fetch/nad.d.ts.map +1 -0
- package/out/us/tools/fetch/nad.js +234 -0
- package/out/us/tools/fetch/nad.js.map +1 -0
- package/out/us/tools/fetch/nppes.d.ts +26 -0
- package/out/us/tools/fetch/nppes.d.ts.map +1 -0
- package/out/us/tools/fetch/nppes.js +126 -0
- package/out/us/tools/fetch/nppes.js.map +1 -0
- package/out/us/tools/fetch/state/hi-schools.d.ts +26 -0
- package/out/us/tools/fetch/state/hi-schools.d.ts.map +1 -0
- package/out/us/tools/fetch/state/hi-schools.js +112 -0
- package/out/us/tools/fetch/state/hi-schools.js.map +1 -0
- package/out/us/tools/fetch/state/sources.d.ts +26 -0
- package/out/us/tools/fetch/state/sources.d.ts.map +1 -0
- package/out/us/tools/fetch/state/sources.js +111 -0
- package/out/us/tools/fetch/state/sources.js.map +1 -0
- package/out/us/tools/fetch/tiger-full.d.ts +22 -0
- package/out/us/tools/fetch/tiger-full.d.ts.map +1 -0
- package/out/us/tools/fetch/tiger-full.js +217 -0
- package/out/us/tools/fetch/tiger-full.js.map +1 -0
- package/out/utils/align.d.ts +1 -1
- package/out/utils/align.d.ts.map +1 -1
- package/out/utils/align.js.map +1 -1
- package/out/utils/golden.d.ts +7 -3
- package/out/utils/golden.d.ts.map +1 -1
- package/out/utils/golden.js +12 -8
- package/out/utils/golden.js.map +1 -1
- package/out/utils/parquet.d.ts.map +1 -1
- package/out/utils/split.js.map +1 -1
- package/out/utils/wof-json.d.ts.map +1 -1
- package/out/utils/wof-json.js +7 -12
- package/out/utils/wof-json.js.map +1 -1
- package/out/ve/recipes/reviewed-postcode-tail.d.ts.map +1 -0
- package/out/ve/recipes/reviewed-postcode-tail.js.map +1 -0
- package/package.json +196 -46
- package/lib/adapters/ban/adapter.ts +0 -190
- package/lib/adapters/fcc-bdc/adapter.ts +0 -174
- package/lib/adapters/geonames-postal/adapter.ts +0 -122
- package/lib/adapters/gnaf/adapter.ts +0 -149
- package/lib/adapters/gnaf/assemble.ts +0 -225
- package/lib/adapters/state-hi-schools/adapter.ts +0 -170
- package/lib/adapters/state-ia-contractors/adapter.ts +0 -144
- package/lib/adapters/state-ny-notaries/adapter.ts +0 -141
- package/lib/adapters/state-tx-notaries/adapter.ts +0 -157
- package/lib/adapters/tiger/adapter.ts +0 -252
- package/lib/adapters/usgov-hrsa-fqhc/adapter.ts +0 -167
- package/lib/adapters/usgov-imls-pls/adapter.ts +0 -143
- package/lib/adapters/usgov-irs-bmf/adapter.ts +0 -151
- package/lib/adapters/usgov-nad/adapter.ts +0 -334
- package/lib/adapters/usgov-nppes/adapter.ts +0 -151
- package/lib/adapters/usgov-samhsa-treatment-locator/adapter.ts +0 -193
- package/lib/adapters/wof-admin-json/adapter.ts +0 -231
- package/lib/adapters/wof-json-rows.ts +0 -129
- package/lib/adapters/wof-postalcode-json/adapter.ts +0 -176
- package/lib/name-prone-us-suffixes.ts +0 -12
- package/lib/recipes/bare-country.ts +0 -95
- package/lib/recipes/cz-pcfirst-preposition.ts +0 -111
- package/lib/recipes/fr-admin-split.ts +0 -242
- package/lib/recipes/fr-bare-street.ts +0 -170
- package/lib/recipes/fr-fragment.ts +0 -333
- package/lib/recipes/fr-lieudit.ts +0 -301
- package/lib/recipes/fr-order.ts +0 -245
- package/lib/recipes/german.ts +0 -359
- package/lib/recipes/locale.ts +0 -599
- package/lib/recipes/nl-postcode.ts +0 -115
- package/lib/recipes/no-fragment.ts +0 -219
- package/lib/recipes/no-street-led.ts +0 -141
- package/lib/recipes/no-street.ts +0 -90
- package/lib/recipes/po-box-cedex.ts +0 -976
- package/lib/recipes/po-box.ts +0 -123
- package/lib/recipes/si-bare-village.ts +0 -97
- package/lib/recipes/street-affix.ts +0 -804
- package/lib/recipes/street-bare.ts +0 -100
- package/lib/recipes/sub-venue-sources.ts +0 -877
- package/lib/recipes/sub-venue.ts +0 -997
- package/lib/synthesizers/german.ts +0 -260
- package/lib/tools/fetch/ban.ts +0 -289
- package/lib/tools/fetch/download.ts +0 -262
- package/lib/tools/fetch/geonames-dump.ts +0 -257
- package/lib/tools/fetch/geonames-postal.ts +0 -180
- package/lib/tools/fetch/hrsa.ts +0 -59
- package/lib/tools/fetch/imls-pls.ts +0 -149
- package/lib/tools/fetch/nad.ts +0 -341
- package/lib/tools/fetch/nppes.ts +0 -165
- package/lib/tools/fetch/state-hi-schools.ts +0 -161
- package/lib/tools/fetch/state-sources.ts +0 -146
- package/lib/tools/fetch/tiger-full.ts +0 -308
- package/lib/tools/golden-promote.ts +0 -337
- package/lib/tools/golden-relabel-street.ts +0 -683
- package/lib/tools/lint-slice-vocab.ts +0 -388
- package/lib/tools/sub-venue/harvest.ts +0 -234
- package/lib/tools/sub-venue/head-nouns.ts +0 -245
- package/lib/tools/sub-venue/surfaces.ts +0 -95
- package/lib/tools/sub-venue/table.ts +0 -281
- package/lib/tools/sub-venue/wikidata.ts +0 -85
- package/lib/tools/sub-venue-lexicon.ts +0 -429
- package/out/adapters/ban/adapter.d.ts.map +0 -1
- package/out/adapters/ban/adapter.js +0 -149
- package/out/adapters/ban/adapter.js.map +0 -1
- package/out/adapters/ban/street-decompose.d.ts.map +0 -1
- package/out/adapters/ban/street-decompose.js.map +0 -1
- package/out/adapters/fcc-bdc/adapter.d.ts.map +0 -1
- package/out/adapters/fcc-bdc/adapter.js +0 -144
- package/out/adapters/fcc-bdc/adapter.js.map +0 -1
- package/out/adapters/geonames-postal/adapter.d.ts.map +0 -1
- package/out/adapters/geonames-postal/adapter.js +0 -107
- package/out/adapters/geonames-postal/adapter.js.map +0 -1
- package/out/adapters/gnaf/adapter.d.ts.map +0 -1
- package/out/adapters/gnaf/adapter.js +0 -124
- package/out/adapters/gnaf/adapter.js.map +0 -1
- package/out/adapters/gnaf/assemble.d.ts +0 -65
- package/out/adapters/gnaf/assemble.d.ts.map +0 -1
- package/out/adapters/gnaf/assemble.js +0 -154
- package/out/adapters/gnaf/assemble.js.map +0 -1
- package/out/adapters/state-hi-schools/adapter.d.ts.map +0 -1
- package/out/adapters/state-hi-schools/adapter.js +0 -145
- package/out/adapters/state-hi-schools/adapter.js.map +0 -1
- package/out/adapters/state-hi-schools/workbook.d.ts.map +0 -1
- package/out/adapters/state-hi-schools/workbook.js.map +0 -1
- package/out/adapters/state-ia-contractors/adapter.d.ts.map +0 -1
- package/out/adapters/state-ia-contractors/adapter.js +0 -111
- package/out/adapters/state-ia-contractors/adapter.js.map +0 -1
- package/out/adapters/state-ny-notaries/adapter.d.ts.map +0 -1
- package/out/adapters/state-ny-notaries/adapter.js +0 -119
- package/out/adapters/state-ny-notaries/adapter.js.map +0 -1
- package/out/adapters/state-tx-notaries/adapter.d.ts.map +0 -1
- package/out/adapters/state-tx-notaries/adapter.js +0 -125
- package/out/adapters/state-tx-notaries/adapter.js.map +0 -1
- package/out/adapters/tiger/adapter.d.ts +0 -55
- package/out/adapters/tiger/adapter.d.ts.map +0 -1
- package/out/adapters/tiger/adapter.js +0 -201
- package/out/adapters/tiger/adapter.js.map +0 -1
- package/out/adapters/tiger/street-decompose.d.ts.map +0 -1
- package/out/adapters/tiger/street-decompose.js.map +0 -1
- package/out/adapters/usgov-hrsa-fqhc/adapter.d.ts.map +0 -1
- package/out/adapters/usgov-hrsa-fqhc/adapter.js +0 -132
- package/out/adapters/usgov-hrsa-fqhc/adapter.js.map +0 -1
- package/out/adapters/usgov-imls-pls/adapter.d.ts.map +0 -1
- package/out/adapters/usgov-imls-pls/adapter.js +0 -114
- package/out/adapters/usgov-imls-pls/adapter.js.map +0 -1
- package/out/adapters/usgov-irs-bmf/adapter.d.ts.map +0 -1
- package/out/adapters/usgov-irs-bmf/adapter.js +0 -120
- package/out/adapters/usgov-irs-bmf/adapter.js.map +0 -1
- package/out/adapters/usgov-nad/adapter.d.ts.map +0 -1
- package/out/adapters/usgov-nad/adapter.js +0 -249
- package/out/adapters/usgov-nad/adapter.js.map +0 -1
- package/out/adapters/usgov-nppes/adapter.d.ts.map +0 -1
- package/out/adapters/usgov-nppes/adapter.js +0 -118
- package/out/adapters/usgov-nppes/adapter.js.map +0 -1
- package/out/adapters/usgov-samhsa-treatment-locator/adapter.d.ts.map +0 -1
- package/out/adapters/usgov-samhsa-treatment-locator/adapter.js +0 -154
- package/out/adapters/usgov-samhsa-treatment-locator/adapter.js.map +0 -1
- package/out/adapters/wof-admin-jp/adapter.d.ts.map +0 -1
- package/out/adapters/wof-admin-jp/adapter.js.map +0 -1
- package/out/adapters/wof-admin-json/adapter.d.ts +0 -79
- package/out/adapters/wof-admin-json/adapter.d.ts.map +0 -1
- package/out/adapters/wof-admin-json/adapter.js +0 -197
- package/out/adapters/wof-admin-json/adapter.js.map +0 -1
- package/out/adapters/wof-json-rows.d.ts +0 -62
- package/out/adapters/wof-json-rows.d.ts.map +0 -1
- package/out/adapters/wof-json-rows.js +0 -83
- package/out/adapters/wof-json-rows.js.map +0 -1
- package/out/adapters/wof-postalcode-json/adapter.d.ts +0 -65
- package/out/adapters/wof-postalcode-json/adapter.d.ts.map +0 -1
- package/out/adapters/wof-postalcode-json/adapter.js +0 -151
- package/out/adapters/wof-postalcode-json/adapter.js.map +0 -1
- package/out/codex/us-fips-state.d.ts +0 -41
- package/out/codex/us-fips-state.d.ts.map +0 -1
- package/out/codex/us-fips-state.js +0 -100
- package/out/codex/us-fips-state.js.map +0 -1
- package/out/name-prone-us-suffixes.d.ts +0 -12
- package/out/name-prone-us-suffixes.d.ts.map +0 -1
- package/out/name-prone-us-suffixes.js +0 -12
- package/out/name-prone-us-suffixes.js.map +0 -1
- package/out/recipes/bare-country.d.ts +0 -25
- package/out/recipes/bare-country.d.ts.map +0 -1
- package/out/recipes/bare-country.js +0 -87
- package/out/recipes/bare-country.js.map +0 -1
- package/out/recipes/cn-organizational-units.d.ts +0 -36
- package/out/recipes/cn-organizational-units.d.ts.map +0 -1
- package/out/recipes/cn-organizational-units.js +0 -159
- package/out/recipes/cn-organizational-units.js.map +0 -1
- package/out/recipes/cz-pcfirst-preposition.d.ts +0 -22
- package/out/recipes/cz-pcfirst-preposition.d.ts.map +0 -1
- package/out/recipes/cz-pcfirst-preposition.js +0 -100
- package/out/recipes/cz-pcfirst-preposition.js.map +0 -1
- package/out/recipes/fr-admin-split.d.ts +0 -38
- package/out/recipes/fr-admin-split.d.ts.map +0 -1
- package/out/recipes/fr-admin-split.js +0 -192
- package/out/recipes/fr-admin-split.js.map +0 -1
- package/out/recipes/fr-bare-street.d.ts +0 -28
- package/out/recipes/fr-bare-street.d.ts.map +0 -1
- package/out/recipes/fr-bare-street.js +0 -143
- package/out/recipes/fr-bare-street.js.map +0 -1
- package/out/recipes/fr-fragment.d.ts +0 -77
- package/out/recipes/fr-fragment.d.ts.map +0 -1
- package/out/recipes/fr-fragment.js +0 -276
- package/out/recipes/fr-fragment.js.map +0 -1
- package/out/recipes/fr-lieudit.d.ts +0 -32
- package/out/recipes/fr-lieudit.d.ts.map +0 -1
- package/out/recipes/fr-lieudit.js +0 -227
- package/out/recipes/fr-lieudit.js.map +0 -1
- package/out/recipes/fr-order.d.ts +0 -38
- package/out/recipes/fr-order.d.ts.map +0 -1
- package/out/recipes/fr-order.js +0 -197
- package/out/recipes/fr-order.js.map +0 -1
- package/out/recipes/german.d.ts +0 -44
- package/out/recipes/german.d.ts.map +0 -1
- package/out/recipes/german.js +0 -288
- package/out/recipes/german.js.map +0 -1
- package/out/recipes/locale.d.ts +0 -136
- package/out/recipes/locale.d.ts.map +0 -1
- package/out/recipes/locale.js +0 -460
- package/out/recipes/locale.js.map +0 -1
- package/out/recipes/nl-postcode.d.ts +0 -26
- package/out/recipes/nl-postcode.d.ts.map +0 -1
- package/out/recipes/nl-postcode.js +0 -102
- package/out/recipes/nl-postcode.js.map +0 -1
- package/out/recipes/no-fragment.d.ts +0 -43
- package/out/recipes/no-fragment.d.ts.map +0 -1
- package/out/recipes/no-fragment.js +0 -182
- package/out/recipes/no-fragment.js.map +0 -1
- package/out/recipes/no-street-led.d.ts +0 -24
- package/out/recipes/no-street-led.d.ts.map +0 -1
- package/out/recipes/no-street-led.js +0 -113
- package/out/recipes/no-street-led.js.map +0 -1
- package/out/recipes/no-street.d.ts +0 -18
- package/out/recipes/no-street.d.ts.map +0 -1
- package/out/recipes/no-street.js +0 -71
- package/out/recipes/no-street.js.map +0 -1
- package/out/recipes/po-box-cedex.d.ts +0 -38
- package/out/recipes/po-box-cedex.d.ts.map +0 -1
- package/out/recipes/po-box-cedex.js +0 -767
- package/out/recipes/po-box-cedex.js.map +0 -1
- package/out/recipes/po-box.d.ts +0 -17
- package/out/recipes/po-box.d.ts.map +0 -1
- package/out/recipes/po-box.js +0 -100
- package/out/recipes/po-box.js.map +0 -1
- package/out/recipes/reviewed-postcode-tail.d.ts.map +0 -1
- package/out/recipes/reviewed-postcode-tail.js.map +0 -1
- package/out/recipes/si-bare-village.d.ts +0 -32
- package/out/recipes/si-bare-village.d.ts.map +0 -1
- package/out/recipes/si-bare-village.js +0 -88
- package/out/recipes/si-bare-village.js.map +0 -1
- package/out/recipes/street-affix.d.ts +0 -89
- package/out/recipes/street-affix.d.ts.map +0 -1
- package/out/recipes/street-affix.js +0 -563
- package/out/recipes/street-affix.js.map +0 -1
- package/out/recipes/street-bare.d.ts +0 -25
- package/out/recipes/street-bare.d.ts.map +0 -1
- package/out/recipes/street-bare.js +0 -86
- package/out/recipes/street-bare.js.map +0 -1
- package/out/recipes/sub-venue-sources.d.ts +0 -229
- package/out/recipes/sub-venue-sources.d.ts.map +0 -1
- package/out/recipes/sub-venue-sources.js +0 -653
- package/out/recipes/sub-venue-sources.js.map +0 -1
- package/out/recipes/sub-venue.d.ts +0 -269
- package/out/recipes/sub-venue.d.ts.map +0 -1
- package/out/recipes/sub-venue.js +0 -725
- package/out/recipes/sub-venue.js.map +0 -1
- package/out/synthesizers/german.d.ts +0 -115
- package/out/synthesizers/german.d.ts.map +0 -1
- package/out/synthesizers/german.js +0 -149
- package/out/synthesizers/german.js.map +0 -1
- package/out/tools/fetch/ban.d.ts +0 -31
- package/out/tools/fetch/ban.d.ts.map +0 -1
- package/out/tools/fetch/ban.js +0 -239
- package/out/tools/fetch/ban.js.map +0 -1
- package/out/tools/fetch/download.d.ts +0 -117
- package/out/tools/fetch/download.d.ts.map +0 -1
- package/out/tools/fetch/download.js +0 -160
- package/out/tools/fetch/download.js.map +0 -1
- package/out/tools/fetch/geonames-dump.d.ts +0 -91
- package/out/tools/fetch/geonames-dump.d.ts.map +0 -1
- package/out/tools/fetch/geonames-dump.js +0 -171
- package/out/tools/fetch/geonames-dump.js.map +0 -1
- package/out/tools/fetch/geonames-postal.d.ts +0 -76
- package/out/tools/fetch/geonames-postal.d.ts.map +0 -1
- package/out/tools/fetch/geonames-postal.js +0 -126
- package/out/tools/fetch/geonames-postal.js.map +0 -1
- package/out/tools/fetch/hrsa.d.ts +0 -18
- package/out/tools/fetch/hrsa.d.ts.map +0 -1
- package/out/tools/fetch/hrsa.js +0 -46
- package/out/tools/fetch/hrsa.js.map +0 -1
- package/out/tools/fetch/imls-pls.d.ts +0 -25
- package/out/tools/fetch/imls-pls.d.ts.map +0 -1
- package/out/tools/fetch/imls-pls.js +0 -108
- package/out/tools/fetch/imls-pls.js.map +0 -1
- package/out/tools/fetch/nad.d.ts +0 -70
- package/out/tools/fetch/nad.d.ts.map +0 -1
- package/out/tools/fetch/nad.js +0 -234
- package/out/tools/fetch/nad.js.map +0 -1
- package/out/tools/fetch/nppes.d.ts +0 -26
- package/out/tools/fetch/nppes.d.ts.map +0 -1
- package/out/tools/fetch/nppes.js +0 -126
- package/out/tools/fetch/nppes.js.map +0 -1
- package/out/tools/fetch/state-hi-schools.d.ts +0 -26
- package/out/tools/fetch/state-hi-schools.d.ts.map +0 -1
- package/out/tools/fetch/state-hi-schools.js +0 -112
- package/out/tools/fetch/state-hi-schools.js.map +0 -1
- package/out/tools/fetch/state-sources.d.ts +0 -26
- package/out/tools/fetch/state-sources.d.ts.map +0 -1
- package/out/tools/fetch/state-sources.js +0 -111
- package/out/tools/fetch/state-sources.js.map +0 -1
- package/out/tools/fetch/tiger-full.d.ts +0 -22
- package/out/tools/fetch/tiger-full.d.ts.map +0 -1
- package/out/tools/fetch/tiger-full.js +0 -217
- package/out/tools/fetch/tiger-full.js.map +0 -1
- package/out/tools/golden-expand.d.ts +0 -90
- package/out/tools/golden-expand.d.ts.map +0 -1
- package/out/tools/golden-expand.js +0 -406
- package/out/tools/golden-expand.js.map +0 -1
- package/out/tools/golden-promote.d.ts +0 -72
- package/out/tools/golden-promote.d.ts.map +0 -1
- package/out/tools/golden-promote.js +0 -240
- package/out/tools/golden-promote.js.map +0 -1
- package/out/tools/golden-relabel-street.d.ts +0 -196
- package/out/tools/golden-relabel-street.d.ts.map +0 -1
- package/out/tools/golden-relabel-street.js +0 -440
- package/out/tools/golden-relabel-street.js.map +0 -1
- package/out/tools/lint-slice-vocab.d.ts +0 -90
- package/out/tools/lint-slice-vocab.d.ts.map +0 -1
- package/out/tools/lint-slice-vocab.js +0 -267
- package/out/tools/lint-slice-vocab.js.map +0 -1
- package/out/tools/lint-slice.d.ts +0 -93
- package/out/tools/lint-slice.d.ts.map +0 -1
- package/out/tools/lint-slice.js +0 -322
- package/out/tools/lint-slice.js.map +0 -1
- package/out/tools/slice-kryptonite.d.ts +0 -32
- package/out/tools/slice-kryptonite.d.ts.map +0 -1
- package/out/tools/slice-kryptonite.js +0 -99
- package/out/tools/slice-kryptonite.js.map +0 -1
- package/out/tools/slice-translit.d.ts +0 -51
- package/out/tools/slice-translit.d.ts.map +0 -1
- package/out/tools/slice-translit.js +0 -166
- package/out/tools/slice-translit.js.map +0 -1
- package/out/tools/sub-venue/harvest.d.ts +0 -100
- package/out/tools/sub-venue/harvest.d.ts.map +0 -1
- package/out/tools/sub-venue/harvest.js +0 -168
- package/out/tools/sub-venue/harvest.js.map +0 -1
- package/out/tools/sub-venue/head-nouns.d.ts +0 -50
- package/out/tools/sub-venue/head-nouns.d.ts.map +0 -1
- package/out/tools/sub-venue/head-nouns.js.map +0 -1
- package/out/tools/sub-venue/surfaces.d.ts +0 -45
- package/out/tools/sub-venue/surfaces.d.ts.map +0 -1
- package/out/tools/sub-venue/surfaces.js.map +0 -1
- package/out/tools/sub-venue/table.d.ts +0 -229
- package/out/tools/sub-venue/table.d.ts.map +0 -1
- package/out/tools/sub-venue/table.js.map +0 -1
- package/out/tools/sub-venue/wikidata.d.ts +0 -26
- package/out/tools/sub-venue/wikidata.d.ts.map +0 -1
- package/out/tools/sub-venue/wikidata.js +0 -62
- package/out/tools/sub-venue/wikidata.js.map +0 -1
- package/out/tools/sub-venue-lexicon.d.ts +0 -156
- package/out/tools/sub-venue-lexicon.d.ts.map +0 -1
- package/out/tools/sub-venue-lexicon.js +0 -283
- package/out/tools/sub-venue-lexicon.js.map +0 -1
- package/out/tools/sub-venue-promotions.d.ts +0 -94
- package/out/tools/sub-venue-promotions.d.ts.map +0 -1
- package/out/tools/sub-venue-promotions.js +0 -266
- package/out/tools/sub-venue-promotions.js.map +0 -1
- /package/lib/{recipes/cn-organizational-units.ts → cn/recipes/organizational-units.ts} +0 -0
- /package/lib/{adapters → fr/adapters}/ban/street-decompose.ts +0 -0
- /package/lib/{adapters/wof-admin-jp → jp/adapters/wof-admin}/adapter.ts +0 -0
- /package/lib/tools/{golden-expand.ts → golden/expand.ts} +0 -0
- /package/lib/tools/{lint-slice.ts → lint/slice/index.ts} +0 -0
- /package/lib/tools/{slice-kryptonite.ts → slice/kryptonite.ts} +0 -0
- /package/lib/tools/{slice-translit.ts → slice/translit.ts} +0 -0
- /package/lib/tools/{sub-venue-promotions.ts → sub/venue/promotions.ts} +0 -0
- /package/lib/{adapters/state-hi-schools → us/adapters/state/hi-schools}/workbook.ts +0 -0
- /package/lib/{adapters → us/adapters}/tiger/street-decompose.ts +0 -0
- /package/lib/{codex/us-fips-state.ts → us/fips-state.ts} +0 -0
- /package/lib/{recipes → ve/recipes}/reviewed-postcode-tail.ts +0 -0
- /package/out/adapters/{geonames-postal → geonames/postal}/adapter.d.ts +0 -0
- /package/out/{adapters → au/adapters}/gnaf/adapter.d.ts +0 -0
- /package/out/{adapters → fr/adapters}/ban/adapter.d.ts +0 -0
- /package/out/{adapters → fr/adapters}/ban/street-decompose.d.ts +0 -0
- /package/out/{adapters → fr/adapters}/ban/street-decompose.js +0 -0
- /package/out/{adapters/wof-admin-jp → jp/adapters/wof-admin}/adapter.d.ts +0 -0
- /package/out/{adapters/wof-admin-jp → jp/adapters/wof-admin}/adapter.js +0 -0
- /package/out/tools/{sub-venue → sub/venue}/head-nouns.js +0 -0
- /package/out/tools/{sub-venue → sub/venue}/surfaces.js +0 -0
- /package/out/tools/{sub-venue → sub/venue}/table.js +0 -0
- /package/out/{adapters → us/adapters}/fcc-bdc/adapter.d.ts +0 -0
- /package/out/{adapters/state-hi-schools → us/adapters/state/hi-schools}/adapter.d.ts +0 -0
- /package/out/{adapters/state-hi-schools → us/adapters/state/hi-schools}/workbook.d.ts +0 -0
- /package/out/{adapters/state-hi-schools → us/adapters/state/hi-schools}/workbook.js +0 -0
- /package/out/{adapters/state-ia-contractors → us/adapters/state/ia-contractors}/adapter.d.ts +0 -0
- /package/out/{adapters/state-ny-notaries → us/adapters/state/ny-notaries}/adapter.d.ts +0 -0
- /package/out/{adapters/state-tx-notaries → us/adapters/state/tx-notaries}/adapter.d.ts +0 -0
- /package/out/{adapters → us/adapters}/tiger/street-decompose.d.ts +0 -0
- /package/out/{adapters → us/adapters}/tiger/street-decompose.js +0 -0
- /package/out/{adapters/usgov-hrsa-fqhc → us/adapters/usgov/hrsa-fqhc}/adapter.d.ts +0 -0
- /package/out/{adapters/usgov-imls-pls → us/adapters/usgov/imls-pls}/adapter.d.ts +0 -0
- /package/out/{adapters/usgov-irs-bmf → us/adapters/usgov/irs-bmf}/adapter.d.ts +0 -0
- /package/out/{adapters/usgov-nad → us/adapters/usgov/nad}/adapter.d.ts +0 -0
- /package/out/{adapters/usgov-nppes → us/adapters/usgov/nppes}/adapter.d.ts +0 -0
- /package/out/{adapters/usgov-samhsa-treatment-locator → us/adapters/usgov/samhsa-treatment-locator}/adapter.d.ts +0 -0
- /package/out/{recipes → ve/recipes}/reviewed-postcode-tail.d.ts +0 -0
- /package/out/{recipes → ve/recipes}/reviewed-postcode-tail.js +0 -0
package/out/tools/lint-slice.js
DELETED
|
@@ -1,322 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* @copyright Sister Software
|
|
3
|
-
* @license AGPL-3.0
|
|
4
|
-
* @author Teffen Ellis, et al.
|
|
5
|
-
*
|
|
6
|
-
* Corpus linter. Compares a new slice against pre-computed corpus statistics and flags patterns
|
|
7
|
-
* that would cause the class of failure we hit with v0.6.2's "5th Avenue Theatre" adversarial
|
|
8
|
-
* venue templates.
|
|
9
|
-
*
|
|
10
|
-
* Per DeepSeek turn 9 design (2026-05-29). v1 checks:
|
|
11
|
-
*
|
|
12
|
-
* 1. **Token-label distribution outliers.** For each token in the new slice, compare the slice's
|
|
13
|
-
* majority label to the corpus's majority label. Flag when the corpus has a
|
|
14
|
-
* confidently-established majority (>66%) AND the slice's majority differs AND both have
|
|
15
|
-
* non-trivial counts (slice ≥ 50, corpus ≥ 200).
|
|
16
|
-
* 2. **Label-vacuum tokens.** Token labeled with a tag that has ZERO instances in the corpus for that
|
|
17
|
-
* token, despite the token being well-represented in the corpus. Stronger signal than #1 —
|
|
18
|
-
* we're introducing a novel association, not shifting a distribution.
|
|
19
|
-
* 3. **Bigram-label collisions.** Identical (token_bigram, label_bigram) appears in slice while the
|
|
20
|
-
* same token_bigram has a DIFFERENT majority label_bigram in the corpus. The "5th Avenue"
|
|
21
|
-
* with [B-venue, I-venue] vs corpus's [B-house_number, I-street] case.
|
|
22
|
-
* 4. **Common-form anti-pattern rules.** Applies `lint-rules.json` — token-regex → forbidden-labels
|
|
23
|
-
* mappings — flagging matches.
|
|
24
|
-
* 5. **Basic sanity.** Truncated rows (tokens.length !== labels.length), all-O rows >90% of slice.
|
|
25
|
-
*
|
|
26
|
-
* Output: markdown report on stdout, optional JSON sidecar via `outJSON`. The command exits 0 if
|
|
27
|
-
* no errors, 1 if any errors (warnings don't check). Per the design, the MANIFEST entry for a
|
|
28
|
-
* flagged slice should require `lint_acknowledged: true` before training consumes it.
|
|
29
|
-
*
|
|
30
|
-
* Usage: mailwoman dev lint corpus-slice\
|
|
31
|
-
* --slice <new-slice.parquet>\
|
|
32
|
-
* --stats <corpus-stats.json>\
|
|
33
|
-
* [--rules <rules.json>]\
|
|
34
|
-
* [--out-md /tmp/lint-report.md]\
|
|
35
|
-
* [--out-json /tmp/lint-report.json]
|
|
36
|
-
*/
|
|
37
|
-
import { readLocalJSONFile } from "@mailwoman/core/fs/readers";
|
|
38
|
-
import { writeLocalFile, writeLocalJSONFile } from "@mailwoman/core/fs/writers";
|
|
39
|
-
import { resolvePackagePath } from "@mailwoman/core/module/resolvers";
|
|
40
|
-
import { accumulateCooccurrences, createCooccurrenceStats, SLICE_STATS_SEP as SEP, streamTokenLabelRows, } from "#utils/slice-stats";
|
|
41
|
-
/**
|
|
42
|
-
* Occurrences of a forbidden label before it is reported — one or two are noise, five is a pattern.
|
|
43
|
-
*/
|
|
44
|
-
const FORBIDDEN_LABEL_REPORT_THRESHOLD = 5;
|
|
45
|
-
/**
|
|
46
|
-
* Examples printed per finding before the list is truncated.
|
|
47
|
-
*/
|
|
48
|
-
const MAX_LISTED_EXAMPLES = 20;
|
|
49
|
-
/**
|
|
50
|
-
* Calibrated thresholds (DeepSeek turn 9). These can be tuned over time if new failure modes surface that the current
|
|
51
|
-
* numbers miss.
|
|
52
|
-
*/
|
|
53
|
-
const CORPUS_CONFIDENCE_FLOOR = 0.66;
|
|
54
|
-
const SLICE_MIN_COUNT = 50;
|
|
55
|
-
const CORPUS_MIN_COUNT = 200;
|
|
56
|
-
const VACUUM_SLICE_MIN_COUNT = 20;
|
|
57
|
-
const VACUUM_CORPUS_MIN_COUNT = 100;
|
|
58
|
-
const BIGRAM_MIN_COUNT = 10;
|
|
59
|
-
const ALL_O_RATIO_CEILING = 0.9;
|
|
60
|
-
/**
|
|
61
|
-
* Default `lint-rules.json` path — the rules ship beside this module in the source tree. tsc does not emit
|
|
62
|
-
* readFileSync'd JSON into `out/`, so the compiled tree falls back to the source-tree copy (corpus/out/src/tools/ →
|
|
63
|
-
* corpus/src/tools/). In-repo the `node` exports condition loads this module from source anyway, so the sibling URL is
|
|
64
|
-
* the common path.
|
|
65
|
-
*/
|
|
66
|
-
function defaultRulesPath() {
|
|
67
|
-
return resolvePackagePath("@mailwoman/corpus", "lib", "tools", "lint-rules.json");
|
|
68
|
-
}
|
|
69
|
-
async function statsFromSlice(rows) {
|
|
70
|
-
const co = createCooccurrenceStats();
|
|
71
|
-
const out = {
|
|
72
|
-
rowCount: 0,
|
|
73
|
-
tokens: co.tokens,
|
|
74
|
-
bigrams: co.bigrams,
|
|
75
|
-
truncatedRows: 0,
|
|
76
|
-
allORows: 0,
|
|
77
|
-
};
|
|
78
|
-
for await (const row of rows) {
|
|
79
|
-
out.rowCount++;
|
|
80
|
-
if (row.tokens.length !== row.labels.length) {
|
|
81
|
-
out.truncatedRows++;
|
|
82
|
-
continue;
|
|
83
|
-
}
|
|
84
|
-
if (row.labels.every((l) => l === "O")) {
|
|
85
|
-
out.allORows++;
|
|
86
|
-
}
|
|
87
|
-
accumulateCooccurrences(co, row.tokens, row.labels);
|
|
88
|
-
}
|
|
89
|
-
return out;
|
|
90
|
-
}
|
|
91
|
-
function majorityLabel(distribution) {
|
|
92
|
-
const entries = distribution instanceof Map ? [...distribution.entries()] : Object.entries(distribution);
|
|
93
|
-
let bestLabel = "";
|
|
94
|
-
let bestCount = 0;
|
|
95
|
-
let total = 0;
|
|
96
|
-
for (const [label, count] of entries) {
|
|
97
|
-
total += count;
|
|
98
|
-
if (count > bestCount) {
|
|
99
|
-
bestCount = count;
|
|
100
|
-
bestLabel = label;
|
|
101
|
-
}
|
|
102
|
-
}
|
|
103
|
-
return { label: bestLabel, count: bestCount, total, confidence: total === 0 ? 0 : bestCount / total };
|
|
104
|
-
}
|
|
105
|
-
function checkDistributionOutliers(slice, corpus) {
|
|
106
|
-
const flags = [];
|
|
107
|
-
for (const [token, sliceLabelMap] of slice.tokens) {
|
|
108
|
-
const corpusLabelMap = corpus.tokens[token];
|
|
109
|
-
if (!corpusLabelMap)
|
|
110
|
-
continue;
|
|
111
|
-
const sliceMaj = majorityLabel(sliceLabelMap);
|
|
112
|
-
const corpusMaj = majorityLabel(corpusLabelMap);
|
|
113
|
-
if (corpusMaj.confidence >= CORPUS_CONFIDENCE_FLOOR &&
|
|
114
|
-
sliceMaj.label !== corpusMaj.label &&
|
|
115
|
-
sliceMaj.count >= SLICE_MIN_COUNT &&
|
|
116
|
-
corpusMaj.total >= CORPUS_MIN_COUNT) {
|
|
117
|
-
flags.push({
|
|
118
|
-
check: "distribution-outlier",
|
|
119
|
-
severity: "error",
|
|
120
|
-
token,
|
|
121
|
-
sliceLabel: sliceMaj.label,
|
|
122
|
-
corpusLabel: corpusMaj.label,
|
|
123
|
-
sliceCount: sliceMaj.count,
|
|
124
|
-
corpusCount: corpusMaj.count,
|
|
125
|
-
detail: `Token "${token}": slice majority is ${sliceMaj.label} (${sliceMaj.count}/${sliceMaj.total}, ${(sliceMaj.confidence * 100).toFixed(0)}%), corpus majority is ${corpusMaj.label} (${corpusMaj.count}/${corpusMaj.total}, ${(corpusMaj.confidence * 100).toFixed(0)}%).`,
|
|
126
|
-
});
|
|
127
|
-
}
|
|
128
|
-
}
|
|
129
|
-
return flags;
|
|
130
|
-
}
|
|
131
|
-
function checkLabelVacuum(slice, corpus) {
|
|
132
|
-
const flags = [];
|
|
133
|
-
for (const [token, sliceLabelMap] of slice.tokens) {
|
|
134
|
-
const corpusLabelMap = corpus.tokens[token];
|
|
135
|
-
if (!corpusLabelMap)
|
|
136
|
-
continue;
|
|
137
|
-
const corpusTotal = Object.values(corpusLabelMap).reduce((a, b) => a + b, 0);
|
|
138
|
-
if (corpusTotal < VACUUM_CORPUS_MIN_COUNT)
|
|
139
|
-
continue;
|
|
140
|
-
for (const [label, sliceCount] of sliceLabelMap) {
|
|
141
|
-
if (sliceCount < VACUUM_SLICE_MIN_COUNT)
|
|
142
|
-
continue;
|
|
143
|
-
if (corpusLabelMap[label] === undefined || corpusLabelMap[label] === 0) {
|
|
144
|
-
flags.push({
|
|
145
|
-
check: "label-vacuum",
|
|
146
|
-
severity: "error",
|
|
147
|
-
token,
|
|
148
|
-
sliceLabel: label,
|
|
149
|
-
sliceCount,
|
|
150
|
-
corpusCount: corpusTotal,
|
|
151
|
-
detail: `Token "${token}": slice labels it ${label} ${sliceCount} times, but the corpus (${corpusTotal} instances of this token) has ZERO instances of this label.`,
|
|
152
|
-
});
|
|
153
|
-
}
|
|
154
|
-
}
|
|
155
|
-
}
|
|
156
|
-
return flags;
|
|
157
|
-
}
|
|
158
|
-
function checkBigramCollisions(slice, corpus) {
|
|
159
|
-
const flags = [];
|
|
160
|
-
for (const [bigram, sliceLabelMap] of slice.bigrams) {
|
|
161
|
-
const corpusLabelMap = corpus.bigrams[bigram];
|
|
162
|
-
if (!corpusLabelMap)
|
|
163
|
-
continue;
|
|
164
|
-
const sliceMaj = majorityLabel(sliceLabelMap);
|
|
165
|
-
const corpusMaj = majorityLabel(corpusLabelMap);
|
|
166
|
-
if (sliceMaj.label !== corpusMaj.label &&
|
|
167
|
-
sliceMaj.count >= BIGRAM_MIN_COUNT &&
|
|
168
|
-
corpusMaj.count >= BIGRAM_MIN_COUNT) {
|
|
169
|
-
const renderBigram = bigram.split(SEP).join(" ");
|
|
170
|
-
const renderSliceLabel = sliceMaj.label.split(SEP).join(" → ");
|
|
171
|
-
const renderCorpusLabel = corpusMaj.label.split(SEP).join(" → ");
|
|
172
|
-
flags.push({
|
|
173
|
-
check: "bigram-collision",
|
|
174
|
-
severity: "error",
|
|
175
|
-
bigram: renderBigram,
|
|
176
|
-
sliceLabel: renderSliceLabel,
|
|
177
|
-
corpusLabel: renderCorpusLabel,
|
|
178
|
-
sliceCount: sliceMaj.count,
|
|
179
|
-
corpusCount: corpusMaj.count,
|
|
180
|
-
detail: `Bigram "${renderBigram}": slice label-bigram is [${renderSliceLabel}] (${sliceMaj.count}×), corpus label-bigram is [${renderCorpusLabel}] (${corpusMaj.count}×). Same surface text, different structural reading.`,
|
|
181
|
-
});
|
|
182
|
-
}
|
|
183
|
-
}
|
|
184
|
-
return flags;
|
|
185
|
-
}
|
|
186
|
-
function checkRules(slice, rulesFile) {
|
|
187
|
-
const flags = [];
|
|
188
|
-
const compiled = rulesFile.rules.map((r) => ({
|
|
189
|
-
rule: r,
|
|
190
|
-
regex: new RegExp(r.pattern, r.pattern_case_sensitive ? "" : "i"),
|
|
191
|
-
}));
|
|
192
|
-
for (const [token, labelMap] of slice.tokens) {
|
|
193
|
-
for (const { rule, regex } of compiled) {
|
|
194
|
-
if (!regex.test(token))
|
|
195
|
-
continue;
|
|
196
|
-
for (const [label, count] of labelMap) {
|
|
197
|
-
if (rule.forbidden_labels.includes(label) && count >= FORBIDDEN_LABEL_REPORT_THRESHOLD) {
|
|
198
|
-
flags.push({
|
|
199
|
-
check: "anti-pattern-rule",
|
|
200
|
-
severity: rule.severity,
|
|
201
|
-
ruleID: rule.id,
|
|
202
|
-
token,
|
|
203
|
-
sliceLabel: label,
|
|
204
|
-
sliceCount: count,
|
|
205
|
-
detail: `Token "${token}" matched rule ${rule.id} and is labeled ${label} ${count} time(s). Rule message: ${rule.message}`,
|
|
206
|
-
});
|
|
207
|
-
}
|
|
208
|
-
}
|
|
209
|
-
}
|
|
210
|
-
}
|
|
211
|
-
return flags;
|
|
212
|
-
}
|
|
213
|
-
function checkSanity(slice) {
|
|
214
|
-
const flags = [];
|
|
215
|
-
if (slice.truncatedRows > 0) {
|
|
216
|
-
flags.push({
|
|
217
|
-
check: "truncated-rows",
|
|
218
|
-
severity: "error",
|
|
219
|
-
detail: `${slice.truncatedRows} row(s) have tokens.length !== labels.length. Pipeline alignment bug.`,
|
|
220
|
-
});
|
|
221
|
-
}
|
|
222
|
-
const allORatio = slice.allORows / Math.max(1, slice.rowCount);
|
|
223
|
-
if (allORatio >= ALL_O_RATIO_CEILING) {
|
|
224
|
-
flags.push({
|
|
225
|
-
check: "all-O-slice",
|
|
226
|
-
severity: "warn",
|
|
227
|
-
detail: `${slice.allORows}/${slice.rowCount} rows (${(allORatio * 100).toFixed(0)}%) are entirely O-labeled. Slice contributes no signal.`,
|
|
228
|
-
});
|
|
229
|
-
}
|
|
230
|
-
return flags;
|
|
231
|
-
}
|
|
232
|
-
function renderReport(opts, slice, flags) {
|
|
233
|
-
const errors = flags.filter((f) => f.severity === "error");
|
|
234
|
-
const warns = flags.filter((f) => f.severity === "warn");
|
|
235
|
-
const verdict = !errors.length ? "**PASS** ✓" : "**FLAGGED** ⚠";
|
|
236
|
-
const lines = [
|
|
237
|
-
`# Corpus Lint: ${verdict}`,
|
|
238
|
-
"",
|
|
239
|
-
`- **Slice:** \`${opts.slicePath}\``,
|
|
240
|
-
`- **Corpus stats:** \`${opts.statsPath}\``,
|
|
241
|
-
`- **Rules:** \`${opts.rulesPath}\``,
|
|
242
|
-
`- **Slice rows:** ${slice.rowCount}`,
|
|
243
|
-
`- **Unique tokens:** ${slice.tokens.size}`,
|
|
244
|
-
`- **Unique bigrams:** ${slice.bigrams.size}`,
|
|
245
|
-
"",
|
|
246
|
-
`**Errors:** ${errors.length} (checks the slice's inclusion unless MANIFEST sets \`lint_acknowledged: true\`)`,
|
|
247
|
-
`**Warnings:** ${warns.length} (advisory)`,
|
|
248
|
-
"",
|
|
249
|
-
];
|
|
250
|
-
if (!flags.length) {
|
|
251
|
-
lines.push("No anomalies detected.");
|
|
252
|
-
return lines.join("\n");
|
|
253
|
-
}
|
|
254
|
-
const byCheck = new Map();
|
|
255
|
-
for (const f of flags) {
|
|
256
|
-
const arr = byCheck.get(f.check) ?? [];
|
|
257
|
-
arr.push(f);
|
|
258
|
-
byCheck.set(f.check, arr);
|
|
259
|
-
}
|
|
260
|
-
for (const [check, list] of byCheck) {
|
|
261
|
-
lines.push(`## ${check} (${list.length})`);
|
|
262
|
-
lines.push("");
|
|
263
|
-
// Sort by sliceCount desc — highest-volume issues first
|
|
264
|
-
list.sort((a, b) => (b.sliceCount ?? 0) - (a.sliceCount ?? 0));
|
|
265
|
-
for (const f of list.slice(0, 20)) {
|
|
266
|
-
lines.push(`- **[${f.severity.toUpperCase()}]** ${f.detail}`);
|
|
267
|
-
}
|
|
268
|
-
if (list.length > MAX_LISTED_EXAMPLES) {
|
|
269
|
-
lines.push(`- ... and ${list.length - 20} more`);
|
|
270
|
-
}
|
|
271
|
-
lines.push("");
|
|
272
|
-
}
|
|
273
|
-
return lines.join("\n");
|
|
274
|
-
}
|
|
275
|
-
/**
|
|
276
|
-
* Lint a slice against corpus stats + the anti-pattern rules; print the markdown report to stdout.
|
|
277
|
-
*/
|
|
278
|
-
export async function lintCorpusSlice(options, report) {
|
|
279
|
-
const rulesPath = options.rulesPath ?? defaultRulesPath();
|
|
280
|
-
report?.(`Reading corpus stats from ${options.statsPath}...`);
|
|
281
|
-
const corpus = await readLocalJSONFile(options.statsPath);
|
|
282
|
-
report?.(` ${corpus.row_count} rows from ${corpus.slice_paths.length} slice(s); ${Object.keys(corpus.tokens).length} tokens, ${Object.keys(corpus.bigrams).length} bigrams`);
|
|
283
|
-
report?.(`Reading slice from ${options.slicePath}...`);
|
|
284
|
-
const slice = await statsFromSlice(streamTokenLabelRows(options.slicePath));
|
|
285
|
-
report?.(` ${slice.rowCount} rows`);
|
|
286
|
-
report?.(`Loading rules from ${rulesPath}...`);
|
|
287
|
-
const rulesFile = await readLocalJSONFile(rulesPath);
|
|
288
|
-
report?.(`Running checks...`);
|
|
289
|
-
const flags = [
|
|
290
|
-
...checkDistributionOutliers(slice, corpus),
|
|
291
|
-
...checkLabelVacuum(slice, corpus),
|
|
292
|
-
...checkBigramCollisions(slice, corpus),
|
|
293
|
-
...checkRules(slice, rulesFile),
|
|
294
|
-
...checkSanity(slice),
|
|
295
|
-
];
|
|
296
|
-
const rendered = renderReport({ slicePath: options.slicePath, statsPath: options.statsPath, rulesPath }, slice, flags);
|
|
297
|
-
console.log(rendered);
|
|
298
|
-
if (options.outMd) {
|
|
299
|
-
await writeLocalFile(rendered, options.outMd);
|
|
300
|
-
}
|
|
301
|
-
if (options.outJSON) {
|
|
302
|
-
await writeLocalJSONFile({
|
|
303
|
-
slice: options.slicePath,
|
|
304
|
-
stats: options.statsPath,
|
|
305
|
-
flags,
|
|
306
|
-
summary: {
|
|
307
|
-
errors: flags.filter((f) => f.severity === "error").length,
|
|
308
|
-
warnings: flags.filter((f) => f.severity === "warn").length,
|
|
309
|
-
},
|
|
310
|
-
}, options.outJSON);
|
|
311
|
-
}
|
|
312
|
-
const errorCount = flags.filter((f) => f.severity === "error").length;
|
|
313
|
-
const warningCount = flags.filter((f) => f.severity === "warn").length;
|
|
314
|
-
if (errorCount > 0) {
|
|
315
|
-
report?.(`LINT FAILED: ${errorCount} error(s).`);
|
|
316
|
-
}
|
|
317
|
-
else {
|
|
318
|
-
report?.("LINT PASSED.");
|
|
319
|
-
}
|
|
320
|
-
return { errors: errorCount, warnings: warningCount, findings: flags, report: rendered };
|
|
321
|
-
}
|
|
322
|
-
//# sourceMappingURL=lint-slice.js.map
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
{"version":3,"file":"lint-slice.js","sourceRoot":"","sources":["../../lib/tools/lint-slice.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAmCG;AAEH,OAAO,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAA;AAC9D,OAAO,EAAE,cAAc,EAAE,kBAAkB,EAAE,MAAM,4BAA4B,CAAA;AAC/E,OAAO,EAAE,kBAAkB,EAAE,MAAM,kCAAkC,CAAA;AAErE,OAAO,EACN,uBAAuB,EACvB,uBAAuB,EACvB,eAAe,IAAI,GAAG,EACtB,oBAAoB,GAEpB,MAAM,oBAAoB,CAAA;AAE3B;;GAEG;AACH,MAAM,gCAAgC,GAAG,CAAC,CAAA;AAE1C;;GAEG;AACH,MAAM,mBAAmB,GAAG,EAAE,CAAA;AAE9B;;;GAGG;AACH,MAAM,uBAAuB,GAAG,IAAI,CAAA;AACpC,MAAM,eAAe,GAAG,EAAE,CAAA;AAC1B,MAAM,gBAAgB,GAAG,GAAG,CAAA;AAC5B,MAAM,sBAAsB,GAAG,EAAE,CAAA;AACjC,MAAM,uBAAuB,GAAG,GAAG,CAAA;AACnC,MAAM,gBAAgB,GAAG,EAAE,CAAA;AAC3B,MAAM,mBAAmB,GAAG,GAAG,CAAA;AAE/B;;;;;GAKG;AACH,SAAS,gBAAgB;IACxB,OAAO,kBAAkB,CAAC,mBAAmB,EAAE,KAAK,EAAE,OAAO,EAAE,iBAAiB,CAAC,CAAA;AAClF,CAAC;AAwDD,KAAK,UAAU,cAAc,CAAC,IAA6B;IAC1D,MAAM,EAAE,GAAG,uBAAuB,EAAE,CAAA;IAEpC,MAAM,GAAG,GAAe;QACvB,QAAQ,EAAE,CAAC;QACX,MAAM,EAAE,EAAE,CAAC,MAAM;QACjB,OAAO,EAAE,EAAE,CAAC,OAAO;QACnB,aAAa,EAAE,CAAC;QAChB,QAAQ,EAAE,CAAC;KACX,CAAA;IAED,IAAI,KAAK,EAAE,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;QAC9B,GAAG,CAAC,QAAQ,EAAE,CAAA;QAEd,IAAI,GAAG,CAAC,MAAM,CAAC,MAAM,KAAK,GAAG,CAAC,MAAM,CAAC,MAAM,EAAE,CAAC;YAC7C,GAAG,CAAC,aAAa,EAAE,CAAA;YAEnB,SAAQ;QACT,CAAC;QAED,IAAI,GAAG,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,KAAK,GAAG,CAAC,EAAE,CAAC;YACxC,GAAG,CAAC,QAAQ,EAAE,CAAA;QACf,CAAC;QAED,uBAAuB,CAAC,EAAE,EAAE,GAAG,CAAC,MAAM,EAAE,GAAG,CAAC,MAAM,CAAC,CAAA;IACpD,CAAC;IAED,OAAO,GAAG,CAAA;AACX,CAAC;AAED,SAAS,aAAa,CAAC,YAA0D;IAMhF,MAAM,OAAO,GAAG,YAAY,YAAY,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,YAAY,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,YAAY,CAAC,CAAA;IACxG,IAAI,SAAS,GAAG,EAAE,CAAA;IAClB,IAAI,SAAS,GAAG,CAAC,CAAA;IACjB,IAAI,KAAK,GAAG,CAAC,CAAA;IAEb,KAAK,MAAM,CAAC,KAAK,EAAE,KAAK,CAAC,IAAI,OAAO,EAAE,CAAC;QACtC,KAAK,IAAI,KAAK,CAAA;QAEd,IAAI,KAAK,GAAG,SAAS,EAAE,CAAC;YACvB,SAAS,GAAG,KAAK,CAAA;YACjB,SAAS,GAAG,KAAK,CAAA;QAClB,CAAC;IACF,CAAC;IAED,OAAO,EAAE,KAAK,EAAE,SAAS,EAAE,KAAK,EAAE,SAAS,EAAE,KAAK,EAAE,UAAU,EAAE,KAAK,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,GAAG,KAAK,EAAE,CAAA;AACtG,CAAC;AA+BD,SAAS,yBAAyB,CAAC,KAAiB,EAAE,MAAmB;IACxE,MAAM,KAAK,GAAoB,EAAE,CAAA;IAEjC,KAAK,MAAM,CAAC,KAAK,EAAE,aAAa,CAAC,IAAI,KAAK,CAAC,MAAM,EAAE,CAAC;QACnD,MAAM,cAAc,GAAG,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAA;QAE3C,IAAI,CAAC,cAAc;YAAE,SAAQ;QAC7B,MAAM,QAAQ,GAAG,aAAa,CAAC,aAAa,CAAC,CAAA;QAC7C,MAAM,SAAS,GAAG,aAAa,CAAC,cAAc,CAAC,CAAA;QAE/C,IACC,SAAS,CAAC,UAAU,IAAI,uBAAuB;YAC/C,QAAQ,CAAC,KAAK,KAAK,SAAS,CAAC,KAAK;YAClC,QAAQ,CAAC,KAAK,IAAI,eAAe;YACjC,SAAS,CAAC,KAAK,IAAI,gBAAgB,EAClC,CAAC;YACF,KAAK,CAAC,IAAI,CAAC;gBACV,KAAK,EAAE,sBAAsB;gBAC7B,QAAQ,EAAE,OAAO;gBACjB,KAAK;gBACL,UAAU,EAAE,QAAQ,CAAC,KAAK;gBAC1B,WAAW,EAAE,SAAS,CAAC,KAAK;gBAC5B,UAAU,EAAE,QAAQ,CAAC,KAAK;gBAC1B,WAAW,EAAE,SAAS,CAAC,KAAK;gBAC5B,MAAM,EAAE,UAAU,KAAK,wBAAwB,QAAQ,CAAC,KAAK,KAAK,QAAQ,CAAC,KAAK,IAAI,QAAQ,CAAC,KAAK,KAAK,CAAC,QAAQ,CAAC,UAAU,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,0BAA0B,SAAS,CAAC,KAAK,KAAK,SAAS,CAAC,KAAK,IAAI,SAAS,CAAC,KAAK,KAAK,CAAC,SAAS,CAAC,UAAU,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK;aAC9Q,CAAC,CAAA;QACH,CAAC;IACF,CAAC;IAED,OAAO,KAAK,CAAA;AACb,CAAC;AAED,SAAS,gBAAgB,CAAC,KAAiB,EAAE,MAAmB;IAC/D,MAAM,KAAK,GAAoB,EAAE,CAAA;IAEjC,KAAK,MAAM,CAAC,KAAK,EAAE,aAAa,CAAC,IAAI,KAAK,CAAC,MAAM,EAAE,CAAC;QACnD,MAAM,cAAc,GAAG,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAA;QAE3C,IAAI,CAAC,cAAc;YAAE,SAAQ;QAC7B,MAAM,WAAW,GAAG,MAAM,CAAC,MAAM,CAAC,cAAc,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAA;QAE5E,IAAI,WAAW,GAAG,uBAAuB;YAAE,SAAQ;QAEnD,KAAK,MAAM,CAAC,KAAK,EAAE,UAAU,CAAC,IAAI,aAAa,EAAE,CAAC;YACjD,IAAI,UAAU,GAAG,sBAAsB;gBAAE,SAAQ;YAEjD,IAAI,cAAc,CAAC,KAAK,CAAC,KAAK,SAAS,IAAI,cAAc,CAAC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC;gBACxE,KAAK,CAAC,IAAI,CAAC;oBACV,KAAK,EAAE,cAAc;oBACrB,QAAQ,EAAE,OAAO;oBACjB,KAAK;oBACL,UAAU,EAAE,KAAK;oBACjB,UAAU;oBACV,WAAW,EAAE,WAAW;oBACxB,MAAM,EAAE,UAAU,KAAK,sBAAsB,KAAK,IAAI,UAAU,2BAA2B,WAAW,6DAA6D;iBACnK,CAAC,CAAA;YACH,CAAC;QACF,CAAC;IACF,CAAC;IAED,OAAO,KAAK,CAAA;AACb,CAAC;AAED,SAAS,qBAAqB,CAAC,KAAiB,EAAE,MAAmB;IACpE,MAAM,KAAK,GAAoB,EAAE,CAAA;IAEjC,KAAK,MAAM,CAAC,MAAM,EAAE,aAAa,CAAC,IAAI,KAAK,CAAC,OAAO,EAAE,CAAC;QACrD,MAAM,cAAc,GAAG,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,CAAA;QAE7C,IAAI,CAAC,cAAc;YAAE,SAAQ;QAC7B,MAAM,QAAQ,GAAG,aAAa,CAAC,aAAa,CAAC,CAAA;QAC7C,MAAM,SAAS,GAAG,aAAa,CAAC,cAAc,CAAC,CAAA;QAE/C,IACC,QAAQ,CAAC,KAAK,KAAK,SAAS,CAAC,KAAK;YAClC,QAAQ,CAAC,KAAK,IAAI,gBAAgB;YAClC,SAAS,CAAC,KAAK,IAAI,gBAAgB,EAClC,CAAC;YACF,MAAM,YAAY,GAAG,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;YAChD,MAAM,gBAAgB,GAAG,QAAQ,CAAC,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAA;YAC9D,MAAM,iBAAiB,GAAG,SAAS,CAAC,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAA;YAEhE,KAAK,CAAC,IAAI,CAAC;gBACV,KAAK,EAAE,kBAAkB;gBACzB,QAAQ,EAAE,OAAO;gBACjB,MAAM,EAAE,YAAY;gBACpB,UAAU,EAAE,gBAAgB;gBAC5B,WAAW,EAAE,iBAAiB;gBAC9B,UAAU,EAAE,QAAQ,CAAC,KAAK;gBAC1B,WAAW,EAAE,SAAS,CAAC,KAAK;gBAC5B,MAAM,EAAE,WAAW,YAAY,6BAA6B,gBAAgB,MAAM,QAAQ,CAAC,KAAK,+BAA+B,iBAAiB,MAAM,SAAS,CAAC,KAAK,sDAAsD;aAC3N,CAAC,CAAA;QACH,CAAC;IACF,CAAC;IAED,OAAO,KAAK,CAAA;AACb,CAAC;AAED,SAAS,UAAU,CAAC,KAAiB,EAAE,SAAwB;IAC9D,MAAM,KAAK,GAAoB,EAAE,CAAA;IAEjC,MAAM,QAAQ,GAAG,SAAS,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QAC5C,IAAI,EAAE,CAAC;QACP,KAAK,EAAE,IAAI,MAAM,CAAC,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,sBAAsB,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC;KACjE,CAAC,CAAC,CAAA;IAEH,KAAK,MAAM,CAAC,KAAK,EAAE,QAAQ,CAAC,IAAI,KAAK,CAAC,MAAM,EAAE,CAAC;QAC9C,KAAK,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,QAAQ,EAAE,CAAC;YACxC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC;gBAAE,SAAQ;YAEhC,KAAK,MAAM,CAAC,KAAK,EAAE,KAAK,CAAC,IAAI,QAAQ,EAAE,CAAC;gBACvC,IAAI,IAAI,CAAC,gBAAgB,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,IAAI,gCAAgC,EAAE,CAAC;oBACxF,KAAK,CAAC,IAAI,CAAC;wBACV,KAAK,EAAE,mBAAmB;wBAC1B,QAAQ,EAAE,IAAI,CAAC,QAAQ;wBACvB,MAAM,EAAE,IAAI,CAAC,EAAE;wBACf,KAAK;wBACL,UAAU,EAAE,KAAK;wBACjB,UAAU,EAAE,KAAK;wBACjB,MAAM,EAAE,UAAU,KAAK,kBAAkB,IAAI,CAAC,EAAE,mBAAmB,KAAK,IAAI,KAAK,2BAA2B,IAAI,CAAC,OAAO,EAAE;qBAC1H,CAAC,CAAA;gBACH,CAAC;YACF,CAAC;QACF,CAAC;IACF,CAAC;IAED,OAAO,KAAK,CAAA;AACb,CAAC;AAED,SAAS,WAAW,CAAC,KAAiB;IACrC,MAAM,KAAK,GAAoB,EAAE,CAAA;IAEjC,IAAI,KAAK,CAAC,aAAa,GAAG,CAAC,EAAE,CAAC;QAC7B,KAAK,CAAC,IAAI,CAAC;YACV,KAAK,EAAE,gBAAgB;YACvB,QAAQ,EAAE,OAAO;YACjB,MAAM,EAAE,GAAG,KAAK,CAAC,aAAa,uEAAuE;SACrG,CAAC,CAAA;IACH,CAAC;IAED,MAAM,SAAS,GAAG,KAAK,CAAC,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,KAAK,CAAC,QAAQ,CAAC,CAAA;IAE9D,IAAI,SAAS,IAAI,mBAAmB,EAAE,CAAC;QACtC,KAAK,CAAC,IAAI,CAAC;YACV,KAAK,EAAE,aAAa;YACpB,QAAQ,EAAE,MAAM;YAChB,MAAM,EAAE,GAAG,KAAK,CAAC,QAAQ,IAAI,KAAK,CAAC,QAAQ,UAAU,CAAC,SAAS,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,yDAAyD;SAC1I,CAAC,CAAA;IACH,CAAC;IAED,OAAO,KAAK,CAAA;AACb,CAAC;AAED,SAAS,YAAY,CACpB,IAAiE,EACjE,KAAiB,EACjB,KAAsB;IAEtB,MAAM,MAAM,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAA;IAC1D,MAAM,KAAK,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,KAAK,MAAM,CAAC,CAAA;IACxD,MAAM,OAAO,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,eAAe,CAAA;IAE/D,MAAM,KAAK,GAAa;QACvB,kBAAkB,OAAO,EAAE;QAC3B,EAAE;QACF,kBAAkB,IAAI,CAAC,SAAS,IAAI;QACpC,yBAAyB,IAAI,CAAC,SAAS,IAAI;QAC3C,kBAAkB,IAAI,CAAC,SAAS,IAAI;QACpC,qBAAqB,KAAK,CAAC,QAAQ,EAAE;QACrC,wBAAwB,KAAK,CAAC,MAAM,CAAC,IAAI,EAAE;QAC3C,yBAAyB,KAAK,CAAC,OAAO,CAAC,IAAI,EAAE;QAC7C,EAAE;QACF,eAAe,MAAM,CAAC,MAAM,kFAAkF;QAC9G,iBAAiB,KAAK,CAAC,MAAM,aAAa;QAC1C,EAAE;KACF,CAAA;IAED,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC;QACnB,KAAK,CAAC,IAAI,CAAC,wBAAwB,CAAC,CAAA;QAEpC,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;IACxB,CAAC;IAED,MAAM,OAAO,GAAG,IAAI,GAAG,EAA2B,CAAA;IAElD,KAAK,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC;QACvB,MAAM,GAAG,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,EAAE,CAAA;QACtC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACX,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,KAAK,EAAE,GAAG,CAAC,CAAA;IAC1B,CAAC;IAED,KAAK,MAAM,CAAC,KAAK,EAAE,IAAI,CAAC,IAAI,OAAO,EAAE,CAAC;QACrC,KAAK,CAAC,IAAI,CAAC,MAAM,KAAK,KAAK,IAAI,CAAC,MAAM,GAAG,CAAC,CAAA;QAC1C,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAA;QACd,wDAAwD;QACxD,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,UAAU,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,UAAU,IAAI,CAAC,CAAC,CAAC,CAAA;QAE9D,KAAK,MAAM,CAAC,IAAI,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAC;YACnC,KAAK,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,QAAQ,CAAC,WAAW,EAAE,OAAO,CAAC,CAAC,MAAM,EAAE,CAAC,CAAA;QAC9D,CAAC;QAED,IAAI,IAAI,CAAC,MAAM,GAAG,mBAAmB,EAAE,CAAC;YACvC,KAAK,CAAC,IAAI,CAAC,aAAa,IAAI,CAAC,MAAM,GAAG,EAAE,OAAO,CAAC,CAAA;QACjD,CAAC;QAED,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAA;IACf,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;AACxB,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,eAAe,CACpC,OAA+B,EAC/B,MAA+B;IAE/B,MAAM,SAAS,GAAG,OAAO,CAAC,SAAS,IAAI,gBAAgB,EAAE,CAAA;IACzD,MAAM,EAAE,CAAC,6BAA6B,OAAO,CAAC,SAAS,KAAK,CAAC,CAAA;IAC7D,MAAM,MAAM,GAAG,MAAM,iBAAiB,CAAc,OAAO,CAAC,SAAS,CAAC,CAAA;IAEtE,MAAM,EAAE,CACP,KAAK,MAAM,CAAC,SAAS,cAAc,MAAM,CAAC,WAAW,CAAC,MAAM,cAAc,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,MAAM,YAAY,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,MAAM,UAAU,CACnK,CAAA;IAED,MAAM,EAAE,CAAC,sBAAsB,OAAO,CAAC,SAAS,KAAK,CAAC,CAAA;IAEtD,MAAM,KAAK,GAAG,MAAM,cAAc,CAAC,oBAAoB,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,CAAA;IAE3E,MAAM,EAAE,CAAC,KAAK,KAAK,CAAC,QAAQ,OAAO,CAAC,CAAA;IAEpC,MAAM,EAAE,CAAC,sBAAsB,SAAS,KAAK,CAAC,CAAA;IAC9C,MAAM,SAAS,GAAG,MAAM,iBAAiB,CAAgB,SAAS,CAAC,CAAA;IAEnE,MAAM,EAAE,CAAC,mBAAmB,CAAC,CAAA;IAE7B,MAAM,KAAK,GAAoB;QAC9B,GAAG,yBAAyB,CAAC,KAAK,EAAE,MAAM,CAAC;QAC3C,GAAG,gBAAgB,CAAC,KAAK,EAAE,MAAM,CAAC;QAClC,GAAG,qBAAqB,CAAC,KAAK,EAAE,MAAM,CAAC;QACvC,GAAG,UAAU,CAAC,KAAK,EAAE,SAAS,CAAC;QAC/B,GAAG,WAAW,CAAC,KAAK,CAAC;KACrB,CAAA;IAED,MAAM,QAAQ,GAAG,YAAY,CAAC,EAAE,SAAS,EAAE,OAAO,CAAC,SAAS,EAAE,SAAS,EAAE,OAAO,CAAC,SAAS,EAAE,SAAS,EAAE,EAAE,KAAK,EAAE,KAAK,CAAC,CAAA;IAEtH,OAAO,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAA;IAErB,IAAI,OAAO,CAAC,KAAK,EAAE,CAAC;QACnB,MAAM,cAAc,CAAC,QAAQ,EAAE,OAAO,CAAC,KAAK,CAAC,CAAA;IAC9C,CAAC;IAED,IAAI,OAAO,CAAC,OAAO,EAAE,CAAC;QACrB,MAAM,kBAAkB,CACvB;YACC,KAAK,EAAE,OAAO,CAAC,SAAS;YACxB,KAAK,EAAE,OAAO,CAAC,SAAS;YACxB,KAAK;YACL,OAAO,EAAE;gBACR,MAAM,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,MAAM;gBAC1D,QAAQ,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,KAAK,MAAM,CAAC,CAAC,MAAM;aAC3D;SACD,EACD,OAAO,CAAC,OAAO,CACf,CAAA;IACF,CAAC;IAED,MAAM,UAAU,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,MAAM,CAAA;IACrE,MAAM,YAAY,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,KAAK,MAAM,CAAC,CAAC,MAAM,CAAA;IAEtE,IAAI,UAAU,GAAG,CAAC,EAAE,CAAC;QACpB,MAAM,EAAE,CAAC,gBAAgB,UAAU,YAAY,CAAC,CAAA;IACjD,CAAC;SAAM,CAAC;QACP,MAAM,EAAE,CAAC,cAAc,CAAC,CAAA;IACzB,CAAC;IAED,OAAO,EAAE,MAAM,EAAE,UAAU,EAAE,QAAQ,EAAE,YAAY,EAAE,QAAQ,EAAE,KAAK,EAAE,MAAM,EAAE,QAAQ,EAAE,CAAA;AACzF,CAAC"}
|
|
@@ -1,32 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* @copyright Sister Software
|
|
3
|
-
* @license AGPL-3.0
|
|
4
|
-
* @author Teffen Ellis, et al.
|
|
5
|
-
*
|
|
6
|
-
* Build a parquet slice from the DeepSeek-generated kryptonite JSONL and emit the corpus-v0.4.0
|
|
7
|
-
* MANIFEST. corpus-v0.4.0 is a pure adapter-addition revision: it points at every slice from
|
|
8
|
-
* v0.3.0 plus the new kryptonite slice(s). No v0.3.0 bytes are touched or re-shuffled.
|
|
9
|
-
*
|
|
10
|
-
* See docs/engineering/reference/CORPUS_V0_4_0_GENERATION.md for the why; that doc also pins the
|
|
11
|
-
* DeepSeek model version + prompt versions used to produce the JSONL.
|
|
12
|
-
*
|
|
13
|
-
* Invoke via `mailwoman corpus slice kryptonite \
|
|
14
|
-
* --jsonl /data/corpus/versioned/v0.4.0/kryptonite/canonical-kryptonite.jsonl \
|
|
15
|
-
* --base-manifest /data/corpus/versioned/v0.3.0/corpus-v0.3.0/MANIFEST.json \
|
|
16
|
-
* --out-dir /data/corpus/versioned/v0.4.0`
|
|
17
|
-
*/
|
|
18
|
-
export interface SliceKryptoniteOptions {
|
|
19
|
-
jsonl: string;
|
|
20
|
-
baseManifest: string;
|
|
21
|
-
outDir: string;
|
|
22
|
-
/**
|
|
23
|
-
* Default `"0.4.0"`.
|
|
24
|
-
*/
|
|
25
|
-
corpusVersion?: string;
|
|
26
|
-
/**
|
|
27
|
-
* Default `"deepseek-kryptonite"`.
|
|
28
|
-
*/
|
|
29
|
-
source?: string;
|
|
30
|
-
}
|
|
31
|
-
export declare function buildKryptoniteSlice(options: SliceKryptoniteOptions, report?: (line: string) => void): Promise<void>;
|
|
32
|
-
//# sourceMappingURL=slice-kryptonite.d.ts.map
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
{"version":3,"file":"slice-kryptonite.d.ts","sourceRoot":"","sources":["../../lib/tools/slice-kryptonite.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAWH,MAAM,WAAW,sBAAsB;IACtC,KAAK,EAAE,MAAM,CAAA;IACb,YAAY,EAAE,MAAM,CAAA;IACpB,MAAM,EAAE,MAAM,CAAA;IACd;;OAEG;IACH,aAAa,CAAC,EAAE,MAAM,CAAA;IACtB;;OAEG;IACH,MAAM,CAAC,EAAE,MAAM,CAAA;CACf;AAiCD,wBAAsB,oBAAoB,CACzC,OAAO,EAAE,sBAAsB,EAC/B,MAAM,CAAC,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,IAAI,GAC7B,OAAO,CAAC,IAAI,CAAC,CA6Df"}
|
|
@@ -1,99 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* @copyright Sister Software
|
|
3
|
-
* @license AGPL-3.0
|
|
4
|
-
* @author Teffen Ellis, et al.
|
|
5
|
-
*
|
|
6
|
-
* Build a parquet slice from the DeepSeek-generated kryptonite JSONL and emit the corpus-v0.4.0
|
|
7
|
-
* MANIFEST. corpus-v0.4.0 is a pure adapter-addition revision: it points at every slice from
|
|
8
|
-
* v0.3.0 plus the new kryptonite slice(s). No v0.3.0 bytes are touched or re-shuffled.
|
|
9
|
-
*
|
|
10
|
-
* See docs/engineering/reference/CORPUS_V0_4_0_GENERATION.md for the why; that doc also pins the
|
|
11
|
-
* DeepSeek model version + prompt versions used to produce the JSONL.
|
|
12
|
-
*
|
|
13
|
-
* Invoke via `mailwoman corpus slice kryptonite \
|
|
14
|
-
* --jsonl /data/corpus/versioned/v0.4.0/kryptonite/canonical-kryptonite.jsonl \
|
|
15
|
-
* --base-manifest /data/corpus/versioned/v0.3.0/corpus-v0.3.0/MANIFEST.json \
|
|
16
|
-
* --out-dir /data/corpus/versioned/v0.4.0`
|
|
17
|
-
*/
|
|
18
|
-
import { pathExists, readLocalJSONFile } from "@mailwoman/core/fs/readers";
|
|
19
|
-
import { writeLocalJSONFile, writeLocalTextFile, makeDirectories } from "@mailwoman/core/fs/writers";
|
|
20
|
-
import { join } from "path-ts";
|
|
21
|
-
import { JSONSpliterator } from "spliterator";
|
|
22
|
-
import { alignRow, PARQUET_COLUMNS, ROW_GROUP_SIZE, SLICE_COMPRESSION, writeSlices } from "#utils";
|
|
23
|
-
async function* canonicalRows(jsonl, corpusVersion) {
|
|
24
|
-
for await (const raw of JSONSpliterator.fromAsync(jsonl)) {
|
|
25
|
-
// Strip sidecar underscore-prefixed fields the generator left behind for debugging.
|
|
26
|
-
const components = raw["components"];
|
|
27
|
-
yield {
|
|
28
|
-
raw: raw["raw"],
|
|
29
|
-
components,
|
|
30
|
-
country: raw["country"] ?? "US",
|
|
31
|
-
locale: raw["locale"] ?? undefined,
|
|
32
|
-
source: raw["source"] ?? "deepseek-kryptonite",
|
|
33
|
-
source_id: raw["source_id"],
|
|
34
|
-
corpus_version: corpusVersion,
|
|
35
|
-
license: raw["license"] ?? "Synthetic (DeepSeek-v4-flash, AGPL-compatible)",
|
|
36
|
-
synth: raw["synth"],
|
|
37
|
-
};
|
|
38
|
-
}
|
|
39
|
-
}
|
|
40
|
-
async function* labeledRows(jsonl, corpusVersion, quarantineLog) {
|
|
41
|
-
for await (const row of canonicalRows(jsonl, corpusVersion)) {
|
|
42
|
-
const result = alignRow(row);
|
|
43
|
-
if (result.kind === "labeled") {
|
|
44
|
-
yield result.row;
|
|
45
|
-
}
|
|
46
|
-
else {
|
|
47
|
-
quarantineLog.push(`${row.source_id}\t${result.row.reason}`);
|
|
48
|
-
}
|
|
49
|
-
}
|
|
50
|
-
}
|
|
51
|
-
export async function buildKryptoniteSlice(options, report) {
|
|
52
|
-
const corpusVersion = options.corpusVersion ?? "0.4.0";
|
|
53
|
-
const source = options.source ?? "deepseek-kryptonite";
|
|
54
|
-
if (!(await pathExists(options.jsonl)))
|
|
55
|
-
throw new Error(`jsonl not found: ${options.jsonl}`);
|
|
56
|
-
if (!(await pathExists(options.baseManifest)))
|
|
57
|
-
throw new Error(`base-manifest not found: ${options.baseManifest}`);
|
|
58
|
-
await makeDirectories(options.outDir);
|
|
59
|
-
const quarantine = [];
|
|
60
|
-
const newManifest = await writeSlices({ train: labeledRows(options.jsonl, corpusVersion, quarantine) }, { outputDir: options.outDir, corpusVersion });
|
|
61
|
-
report?.(`wrote ${newManifest.total_rows} rows into ${newManifest.slices.length} slice(s); ` +
|
|
62
|
-
`quarantined ${quarantine.length}`);
|
|
63
|
-
if (quarantine.length) {
|
|
64
|
-
const qPath = join(options.outDir, `corpus-v${corpusVersion}`, "quarantine-kryptonite.tsv");
|
|
65
|
-
await writeLocalTextFile(quarantine.join("\n") + "\n", qPath);
|
|
66
|
-
report?.(`quarantine log → ${qPath}`);
|
|
67
|
-
}
|
|
68
|
-
// Stamp the new slice's source field for audit.ts (which prefers slice.source over
|
|
69
|
-
// first_source_id-prefix inference). Without this, deepseek-kryptonite IDs would have
|
|
70
|
-
// to match a prefix in KNOWN_SOURCE_PREFIXES — we add it there too as a belt-and-braces.
|
|
71
|
-
for (const sh of newManifest.slices) {
|
|
72
|
-
sh.source = source;
|
|
73
|
-
}
|
|
74
|
-
// Compose the final corpus-v0.4.0 manifest: every slice from base + the new slice(s).
|
|
75
|
-
const base = await readLocalJSONFile(options.baseManifest);
|
|
76
|
-
const combined = {
|
|
77
|
-
corpus_version: corpusVersion,
|
|
78
|
-
schema: PARQUET_COLUMNS,
|
|
79
|
-
rows_per_slice: base.rows_per_slice,
|
|
80
|
-
row_group_size: base.row_group_size ?? ROW_GROUP_SIZE,
|
|
81
|
-
slices: [...base.slices, ...newManifest.slices],
|
|
82
|
-
counts: {
|
|
83
|
-
train: base.counts.train + (newManifest.counts.train ?? 0),
|
|
84
|
-
val: base.counts.val,
|
|
85
|
-
test: base.counts.test,
|
|
86
|
-
},
|
|
87
|
-
total_rows: base.total_rows + newManifest.total_rows,
|
|
88
|
-
};
|
|
89
|
-
// Stamp source on the legacy v0.3.0 slices too, so audit's slice.source path is the
|
|
90
|
-
// authoritative one. v0.3.0 slices mix sources; we use the first_source_id-prefix
|
|
91
|
-
// inference for them (audit.ts will re-derive on its own when slice.source is absent).
|
|
92
|
-
const combinedPath = join(options.outDir, `corpus-v${corpusVersion}`, "MANIFEST.json");
|
|
93
|
-
await writeLocalJSONFile(combined, combinedPath);
|
|
94
|
-
report?.(`wrote combined manifest → ${combinedPath}`);
|
|
95
|
-
report?.(` total_rows=${combined.total_rows} (base=${base.total_rows}, added=${newManifest.total_rows})`);
|
|
96
|
-
report?.(` slices=${combined.slices.length} (base=${base.slices.length}, added=${newManifest.slices.length})`);
|
|
97
|
-
report?.(` compression=${SLICE_COMPRESSION}`);
|
|
98
|
-
}
|
|
99
|
-
//# sourceMappingURL=slice-kryptonite.js.map
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
{"version":3,"file":"slice-kryptonite.js","sourceRoot":"","sources":["../../lib/tools/slice-kryptonite.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAEH,OAAO,EAAE,UAAU,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAA;AAC1E,OAAO,EAAE,kBAAkB,EAAE,kBAAkB,EAAE,eAAe,EAAE,MAAM,4BAA4B,CAAA;AACpG,OAAO,EAAE,IAAI,EAAE,MAAM,SAAS,CAAA;AAC9B,OAAO,EAAE,eAAe,EAAE,MAAM,aAAa,CAAA;AAG7C,OAAO,EAAE,QAAQ,EAAE,eAAe,EAAE,cAAc,EAAE,iBAAiB,EAAE,WAAW,EAAE,MAAM,QAAQ,CAAA;AAiBlG,KAAK,SAAS,CAAC,CAAC,aAAa,CAAC,KAAa,EAAE,aAAqB;IACjE,IAAI,KAAK,EAAE,MAAM,GAAG,IAAI,eAAe,CAAC,SAAS,CAA0B,KAAK,CAAC,EAAE,CAAC;QACnF,oFAAoF;QACpF,MAAM,UAAU,GAAG,GAAG,CAAC,YAAY,CAA2B,CAAA;QAE9D,MAAM;YACL,GAAG,EAAE,GAAG,CAAC,KAAK,CAAW;YACzB,UAAU;YACV,OAAO,EAAG,GAAG,CAAC,SAAS,CAAY,IAAI,IAAI;YAC3C,MAAM,EAAG,GAAG,CAAC,QAAQ,CAAY,IAAI,SAAS;YAC9C,MAAM,EAAG,GAAG,CAAC,QAAQ,CAAY,IAAI,qBAAqB;YAC1D,SAAS,EAAE,GAAG,CAAC,WAAW,CAAW;YACrC,cAAc,EAAE,aAAa;YAC7B,OAAO,EAAG,GAAG,CAAC,SAAS,CAAY,IAAI,gDAAgD;YACvF,KAAK,EAAE,GAAG,CAAC,OAAO,CAA0B;SAC5C,CAAA;IACF,CAAC;AACF,CAAC;AAED,KAAK,SAAS,CAAC,CAAC,WAAW,CAAC,KAAa,EAAE,aAAqB,EAAE,aAAuB;IACxF,IAAI,KAAK,EAAE,MAAM,GAAG,IAAI,aAAa,CAAC,KAAK,EAAE,aAAa,CAAC,EAAE,CAAC;QAC7D,MAAM,MAAM,GAAG,QAAQ,CAAC,GAAG,CAAC,CAAA;QAE5B,IAAI,MAAM,CAAC,IAAI,KAAK,SAAS,EAAE,CAAC;YAC/B,MAAM,MAAM,CAAC,GAAG,CAAA;QACjB,CAAC;aAAM,CAAC;YACP,aAAa,CAAC,IAAI,CAAC,GAAG,GAAG,CAAC,SAAS,KAAK,MAAM,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC,CAAA;QAC7D,CAAC;IACF,CAAC;AACF,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,oBAAoB,CACzC,OAA+B,EAC/B,MAA+B;IAE/B,MAAM,aAAa,GAAG,OAAO,CAAC,aAAa,IAAI,OAAO,CAAA;IACtD,MAAM,MAAM,GAAG,OAAO,CAAC,MAAM,IAAI,qBAAqB,CAAA;IAEtD,IAAI,CAAC,CAAC,MAAM,UAAU,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,oBAAoB,OAAO,CAAC,KAAK,EAAE,CAAC,CAAA;IAE5F,IAAI,CAAC,CAAC,MAAM,UAAU,CAAC,OAAO,CAAC,YAAY,CAAC,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,4BAA4B,OAAO,CAAC,YAAY,EAAE,CAAC,CAAA;IAElH,MAAM,eAAe,CAAC,OAAO,CAAC,MAAM,CAAC,CAAA;IAErC,MAAM,UAAU,GAAa,EAAE,CAAA;IAE/B,MAAM,WAAW,GAAG,MAAM,WAAW,CACpC,EAAE,KAAK,EAAE,WAAW,CAAC,OAAO,CAAC,KAAK,EAAE,aAAa,EAAE,UAAU,CAAC,EAAE,EAChE,EAAE,SAAS,EAAE,OAAO,CAAC,MAAM,EAAE,aAAa,EAAE,CAC5C,CAAA;IAED,MAAM,EAAE,CACP,SAAS,WAAW,CAAC,UAAU,cAAc,WAAW,CAAC,MAAM,CAAC,MAAM,aAAa;QAClF,eAAe,UAAU,CAAC,MAAM,EAAE,CACnC,CAAA;IAED,IAAI,UAAU,CAAC,MAAM,EAAE,CAAC;QACvB,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,WAAW,aAAa,EAAE,EAAE,2BAA2B,CAAC,CAAA;QAC3F,MAAM,kBAAkB,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,EAAE,KAAK,CAAC,CAAA;QAC7D,MAAM,EAAE,CAAC,oBAAoB,KAAK,EAAE,CAAC,CAAA;IACtC,CAAC;IAED,mFAAmF;IACnF,sFAAsF;IACtF,yFAAyF;IACzF,KAAK,MAAM,EAAE,IAAI,WAAW,CAAC,MAAM,EAAE,CAAC;QACrC,EAAE,CAAC,MAAM,GAAG,MAAM,CAAA;IACnB,CAAC;IAED,sFAAsF;IACtF,MAAM,IAAI,GAAG,MAAM,iBAAiB,CAAgB,OAAO,CAAC,YAAY,CAAC,CAAA;IAEzE,MAAM,QAAQ,GAAkB;QAC/B,cAAc,EAAE,aAAa;QAC7B,MAAM,EAAE,eAAe;QACvB,cAAc,EAAE,IAAI,CAAC,cAAc;QACnC,cAAc,EAAE,IAAI,CAAC,cAAc,IAAI,cAAc;QACrD,MAAM,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,GAAG,WAAW,CAAC,MAAM,CAAC;QAC/C,MAAM,EAAE;YACP,KAAK,EAAE,IAAI,CAAC,MAAM,CAAC,KAAK,GAAG,CAAC,WAAW,CAAC,MAAM,CAAC,KAAK,IAAI,CAAC,CAAC;YAC1D,GAAG,EAAE,IAAI,CAAC,MAAM,CAAC,GAAG;YACpB,IAAI,EAAE,IAAI,CAAC,MAAM,CAAC,IAAI;SACtB;QACD,UAAU,EAAE,IAAI,CAAC,UAAU,GAAG,WAAW,CAAC,UAAU;KACpD,CAAA;IAED,oFAAoF;IACpF,kFAAkF;IAClF,uFAAuF;IACvF,MAAM,YAAY,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,WAAW,aAAa,EAAE,EAAE,eAAe,CAAC,CAAA;IACtF,MAAM,kBAAkB,CAAC,QAAQ,EAAE,YAAY,CAAC,CAAA;IAChD,MAAM,EAAE,CAAC,6BAA6B,YAAY,EAAE,CAAC,CAAA;IACrD,MAAM,EAAE,CAAC,gBAAgB,QAAQ,CAAC,UAAU,UAAU,IAAI,CAAC,UAAU,WAAW,WAAW,CAAC,UAAU,GAAG,CAAC,CAAA;IAC1G,MAAM,EAAE,CAAC,YAAY,QAAQ,CAAC,MAAM,CAAC,MAAM,UAAU,IAAI,CAAC,MAAM,CAAC,MAAM,WAAW,WAAW,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAA;IAC/G,MAAM,EAAE,CAAC,iBAAiB,iBAAiB,EAAE,CAAC,CAAA;AAC/C,CAAC"}
|
|
@@ -1,51 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* @copyright Sister Software
|
|
3
|
-
* @license AGPL-3.0
|
|
4
|
-
* @author Teffen Ellis, et al.
|
|
5
|
-
*
|
|
6
|
-
* Build per-script parquet slices from the DeepSeek-generated transliteration JSONL and emit the
|
|
7
|
-
* corpus-v0.4.0 MANIFEST that combines them with the existing kryptonite + v0.3.0 slices.
|
|
8
|
-
*
|
|
9
|
-
* Sibling to `slice-kryptonite.ts`. The two modules share the same composition pattern: take a
|
|
10
|
-
* base MANIFEST, append new slices, write a combined MANIFEST. Differences specific to
|
|
11
|
-
* transliteration:
|
|
12
|
-
*
|
|
13
|
-
* - One JSONL contains rows from N target scripts (source = `deepseek-translit-<slug>`). We bucket by
|
|
14
|
-
* `source` and write one slice per script so `audit.ts` can attribute each slice to its
|
|
15
|
-
* synthetic source without relying on filename-prefix inference.
|
|
16
|
-
* - Each slice is written to `train/part-translit-<slug>.parquet` (distinct from kryptonite's
|
|
17
|
-
* `part-0000.parquet`, which v0.4.0's first builder already produced).
|
|
18
|
-
* - Inherits the path-canonicalization fix flagged in Thread B's postmortem: v0.3.0 slice paths are
|
|
19
|
-
* rewritten from `$MAILWOMAN_DATA_ROOT/...` to `/data/...` in the combined MANIFEST so
|
|
20
|
-
* all paths share one container-friendly form.
|
|
21
|
-
*
|
|
22
|
-
* See docs/engineering/reference/CORPUS_V0_4_0_GENERATION.md for prompts, model, and the
|
|
23
|
-
* reproducibility contract.
|
|
24
|
-
*
|
|
25
|
-
* Invoke via `mailwoman corpus slice translit \
|
|
26
|
-
* --jsonl /data/corpus/versioned/v0.4.0/transliteration/canonical-transliteration.jsonl \
|
|
27
|
-
* --base-manifest /data/corpus/versioned/v0.4.0/corpus-v0.4.0/MANIFEST.json \
|
|
28
|
-
* --out-dir /data/corpus/versioned/v0.4.0`
|
|
29
|
-
*/
|
|
30
|
-
export interface SliceTranslitOptions {
|
|
31
|
-
jsonl: string;
|
|
32
|
-
baseManifest: string;
|
|
33
|
-
outDir: string;
|
|
34
|
-
/**
|
|
35
|
-
* Default `"0.4.0"`.
|
|
36
|
-
*/
|
|
37
|
-
corpusVersion?: string;
|
|
38
|
-
/**
|
|
39
|
-
* Default `"/data/"`.
|
|
40
|
-
*/
|
|
41
|
-
canonicalPathPrefix?: string;
|
|
42
|
-
/**
|
|
43
|
-
* Prefix the base manifest's slice paths currently carry, to be rewritten to
|
|
44
|
-
* {@link SliceTranslitOptions.canonicalPathPrefix}. Defaults to `mailwomanDataRoot()` with a trailing slash — the
|
|
45
|
-
* root that WROTE those paths. Pass it explicitly when translating a manifest generated under a different
|
|
46
|
-
* `$MAILWOMAN_DATA_ROOT` than the one you are running with.
|
|
47
|
-
*/
|
|
48
|
-
legacyPathPrefix?: string;
|
|
49
|
-
}
|
|
50
|
-
export declare function buildTranslitSlice(options: SliceTranslitOptions, report?: (line: string) => void): Promise<void>;
|
|
51
|
-
//# sourceMappingURL=slice-translit.d.ts.map
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
{"version":3,"file":"slice-translit.d.ts","sourceRoot":"","sources":["../../lib/tools/slice-translit.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA4BG;AAaH,MAAM,WAAW,oBAAoB;IACpC,KAAK,EAAE,MAAM,CAAA;IACb,YAAY,EAAE,MAAM,CAAA;IACpB,MAAM,EAAE,MAAM,CAAA;IACd;;OAEG;IACH,aAAa,CAAC,EAAE,MAAM,CAAA;IACtB;;OAEG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAA;IAC5B;;;;;OAKG;IACH,gBAAgB,CAAC,EAAE,MAAM,CAAA;CACzB;AA0DD,wBAAsB,kBAAkB,CACvC,OAAO,EAAE,oBAAoB,EAC7B,MAAM,CAAC,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,IAAI,GAC7B,OAAO,CAAC,IAAI,CAAC,CAgGf"}
|