@mailwoman/corpus 9.2.0 → 9.3.0

package/out/recipes/house-venue.js

@@ -0,0 +1,73 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   `house-venue` slice recipe — synthetic house_number+venue+street co-occurrence rows: tuples →
+ *   {@link synthesizeHouseVenueRow} → aligned LabeledRow. The v0.6.3 corrective companion to the
+ *   no-street slice: every row carries BOTH house_number AND venue, restoring the house_number
+ *   signal that no-street's distributional shift cost the model. Ported from
+ *   scripts/build-house-venue-slice.mjs.
+ */
+import { alignAndWrite, makeLcg, readTuples, sliceSourceID, SYNTHETIC_TUPLE_LICENSE as LICENSE, } from "#recipes/scaffold";
+import { synthesizeHouseVenueRow } from "#synthesizers/house-venue";
+/**
+ * Slice recipe registered with the corpus builder — see the file header for the parse behaviour it exists to exercise,
+ * and `description` below for the surface form it generates.
+ */
+export const houseVenueRecipe = {
+    name: "house-venue",
+    description: "House_number+venue co-occurrence rows: tuples → synthesizeHouseVenueRow → aligned LabeledRow",
+    mode: "tuples",
+    async run(opts, write) {
+        if (!opts.input)
+            throw new Error("house-venue recipe requires --input <tuples.jsonl>");
+        // Legacy build-house-venue-slice.mjs seeded the LCG via makeRandom(opts.seed) (s = seed).
+        const random = makeLcg(opts.seed);
+        const source = opts.sourceName ?? "synth-house-venue";
+        let read = 0;
+        let emitted = 0;
+        let skipped = 0;
+        for await (const tuple of readTuples(opts.input)) {
+            read++;
+            // FR renders without a region (postcode-before-locality tail — the run-2 contingency), so
+            // an empty region is valid there and stays required everywhere else.
+            if (!tuple.locality || !tuple.postcode || !tuple.country || (!tuple.region && tuple.country !== "FR")) {
+                skipped++;
+                continue;
+            }
+            for (let v = 0; v < opts.variants; v++) {
+                const synth = synthesizeHouseVenueRow(tuple, { random });
+                if (!synth) {
+                    skipped++;
+                    continue;
+                }
+                const ok = alignAndWrite(write, {
+                    raw: synth.raw,
+                    components: synth.components,
+                    country: tuple.country,
+                    locale: synth.locale,
+                    source,
+                    source_id: sliceSourceID(source, {
+                        locality: tuple.locality,
+                        region: tuple.region,
+                        postcode: tuple.postcode,
+                        country: tuple.country,
+                        template: synth.template,
+                        v: String(v),
+                    }),
+                    corpus_version: "0.4.0",
+                    license: LICENSE,
+                }, synth.template);
+                if (ok) {
+                    emitted++;
+                }
+                else {
+                    skipped++;
+                }
+            }
+        }
+        return { read, emitted, skipped };
+    },
+};
+//# sourceMappingURL=house-venue.js.map

package/out/recipes/house-venue.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"house-venue.js","sourceRoot":"","sources":["../../lib/recipes/house-venue.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAEH,OAAO,EACN,aAAa,EACb,OAAO,EACP,UAAU,EACV,aAAa,EACb,uBAAuB,IAAI,OAAO,GAElC,MAAM,mBAAmB,CAAA;AAC1B,OAAO,EAAE,uBAAuB,EAA4B,MAAM,2BAA2B,CAAA;AAE7F;;;GAGG;AACH,MAAM,CAAC,MAAM,gBAAgB,GAAiB;IAC7C,IAAI,EAAE,aAAa;IACnB,WAAW,EAAE,8FAA8F;IAC3G,IAAI,EAAE,QAAQ;IACd,KAAK,CAAC,GAAG,CAAC,IAAI,EAAE,KAAK;QACpB,IAAI,CAAC,IAAI,CAAC,KAAK;YAAE,MAAM,IAAI,KAAK,CAAC,oDAAoD,CAAC,CAAA;QACtF,0FAA0F;QAC1F,MAAM,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;QACjC,MAAM,MAAM,GAAG,IAAI,CAAC,UAAU,IAAI,mBAAmB,CAAA;QACrD,IAAI,IAAI,GAAG,CAAC,CAAA;QACZ,IAAI,OAAO,GAAG,CAAC,CAAA;QACf,IAAI,OAAO,GAAG,CAAC,CAAA;QAEf,IAAI,KAAK,EAAE,MAAM,KAAK,IAAI,UAAU,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;YAClD,IAAI,EAAE,CAAA;YAEN,0FAA0F;YAC1F,qEAAqE;YACrE,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,IAAI,CAAC,CAAC,KAAK,CAAC,MAAM,IAAI,KAAK,CAAC,OAAO,KAAK,IAAI,CAAC,EAAE,CAAC;gBACvG,OAAO,EAAE,CAAA;gBAET,SAAQ;YACT,CAAC;YAED,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,QAAQ,EAAE,CAAC,EAAE,EAAE,CAAC;gBACxC,MAAM,KAAK,GAAG,uBAAuB,CAAC,KAA4B,EAAE,EAAE,MAAM,EAAE,CAAC,CAAA;gBAE/E,IAAI,CAAC,KAAK,EAAE,CAAC;oBACZ,OAAO,EAAE,CAAA;oBAET,SAAQ;gBACT,CAAC;gBAED,MAAM,EAAE,GAAG,aAAa,CACvB,KAAK,EACL;oBACC,GAAG,EAAE,KAAK,CAAC,GAAG;oBACd,UAAU,EAAE,KAAK,CAAC,UAAU;oBAC5B,OAAO,EAAE,KAAK,CAAC,OAAO;oBACtB,MAAM,EAAE,KAAK,CAAC,MAAM;oBACpB,MAAM;oBACN,SAAS,EAAE,aAAa,CAAC,MAAM,EAAE;wBAChC,QAAQ,EAAE,KAAK,CAAC,QAAQ;wBACxB,MAAM,EAAE,KAAK,CAAC,MAAM;wBACpB,QAAQ,EAAE,KAAK,CAAC,QAAQ;wBACxB,OAAO,EAAE,KAAK,CAAC,OAAO;wBACtB,QAAQ,EAAE,KAAK,CAAC,QAAQ;wBACxB,CAAC,EAAE,MAAM,CAAC,CAAC,CAAC;qBACZ,CAAC;oBACF,cAAc,EAAE,OAAO;oBACvB,OAAO,EAAE,OAAO;iBAChB,EACD,KAAK,CAAC,QAAQ,CACd,CAAA;gBAED,IAAI,EAAE,EAAE,CAAC;oBACR,OAAO,EAAE,CAAA;gBACV,CAAC;qBAAM,CAAC;oBACP,OAAO,EAAE,CAAA;gBACV,CAAC;YACF,CAAC;QACF,CAAC;QAED,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,OAAO,EAAE,CAAA;IAClC,CAAC;CACD,CAAA"}

package/out/recipes/index.d.ts

@@ -0,0 +1,24 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   The synthetic-corpus SLICE RECIPE registry. Each recipe ({@link CorpusRecipe}) is what one legacy
+ *   `build-*-slice.mjs` script used to be; the `mailwoman corpus slice <recipe>` command dispatches
+ *   to them. A new slice is a new entry here, not a new script.
+ */
+import type { CorpusRecipe } from "#recipes/scaffold";
+export * from "#recipes/scaffold";
+/**
+ * Recipe name → recipe.
+ */
+export declare const SLICE_RECIPES: ReadonlyMap<string, CorpusRecipe>;
+/**
+ * Look up a recipe by its `<recipe>` name.
+ */
+export declare function getSliceRecipe(name: string): CorpusRecipe | undefined;
+/**
+ * All recipes (for `--list` / help).
+ */
+export declare function listSliceRecipes(): readonly CorpusRecipe[];
+//# sourceMappingURL=index.d.ts.map

package/out/recipes/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../lib/recipes/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAwBH,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,mBAAmB,CAAA;AASrD,cAAc,mBAAmB,CAAA;AAsCjC;;GAEG;AACH,eAAO,MAAM,aAAa,EAAE,WAAW,CAAC,MAAM,EAAE,YAAY,CAA4C,CAAA;AAExG;;GAEG;AACH,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,GAAG,YAAY,GAAG,SAAS,CAErE;AAED;;GAEG;AACH,wBAAgB,gBAAgB,IAAI,SAAS,YAAY,EAAE,CAE1D"}

package/out/recipes/index.js

@@ -0,0 +1,91 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   The synthetic-corpus SLICE RECIPE registry. Each recipe ({@link CorpusRecipe}) is what one legacy
+ *   `build-*-slice.mjs` script used to be; the `mailwoman corpus slice <recipe>` command dispatches
+ *   to them. A new slice is a new entry here, not a new script.
+ */
+import { anchorAbsorptionRecipe } from "#recipes/anchor-absorption";
+import { bareCountryRecipe } from "#recipes/bare-country";
+import { boundaryStressRecipe } from "#recipes/boundary-stress";
+import { cnOrganizationalUnitsRecipe } from "#recipes/cn-organizational-units";
+import { countryBalancedRecipe } from "#recipes/country-balanced";
+import { czPcFirstPrepositionRecipe } from "#recipes/cz-pcfirst-preposition";
+import { frAdminSplitRecipe } from "#recipes/fr-admin-split";
+import { frBareStreetRecipe } from "#recipes/fr-bare-street";
+import { frFragmentRecipe } from "#recipes/fr-fragment";
+import { frLieuditRecipe } from "#recipes/fr-lieudit";
+import { frOrderRecipe } from "#recipes/fr-order";
+import { germanRecipe } from "#recipes/german";
+import { houseVenueRecipe } from "#recipes/house-venue";
+import { intersectionRecipe } from "#recipes/intersection";
+import { localeRecipe } from "#recipes/locale";
+import { nlPostcodeRecipe } from "#recipes/nl-postcode";
+import { noFragmentRecipe } from "#recipes/no-fragment";
+import { noStreetRecipe } from "#recipes/no-street";
+import { noStreetLedRecipe } from "#recipes/no-street-led";
+import { poBoxRecipe } from "#recipes/po-box";
+import { poBoxCedexRecipe } from "#recipes/po-box-cedex";
+import { reviewedPostcodeTailRecipe } from "#recipes/reviewed-postcode-tail";
+import { siBareVillageRecipe } from "#recipes/si-bare-village";
+import { streetRecipe } from "#recipes/street";
+import { streetAffixRecipe, suffixBoundaryRecipe } from "#recipes/street-affix";
+import { streetBareRecipe } from "#recipes/street-bare";
+import { subVenueRecipe } from "#recipes/sub-venue";
+import { trailingRegionRecipe } from "#recipes/trailing-region";
+import { unitRecipe } from "#recipes/unit";
+export * from "#recipes/scaffold";
+/**
+ * Every registered recipe, in display order.
+ */
+const RECIPES = [
+    bareCountryRecipe,
+    trailingRegionRecipe,
+    reviewedPostcodeTailRecipe,
+    streetRecipe,
+    streetBareRecipe,
+    streetAffixRecipe,
+    suffixBoundaryRecipe,
+    noStreetRecipe,
+    houseVenueRecipe,
+    poBoxRecipe,
+    poBoxCedexRecipe,
+    unitRecipe,
+    intersectionRecipe,
+    germanRecipe,
+    cnOrganizationalUnitsRecipe,
+    localeRecipe,
+    frOrderRecipe,
+    frAdminSplitRecipe,
+    frBareStreetRecipe,
+    frFragmentRecipe,
+    frLieuditRecipe,
+    czPcFirstPrepositionRecipe,
+    nlPostcodeRecipe,
+    noStreetLedRecipe,
+    noFragmentRecipe,
+    siBareVillageRecipe,
+    countryBalancedRecipe,
+    boundaryStressRecipe,
+    anchorAbsorptionRecipe,
+    subVenueRecipe,
+];
+/**
+ * Recipe name → recipe.
+ */
+export const SLICE_RECIPES = new Map(RECIPES.map((r) => [r.name, r]));
+/**
+ * Look up a recipe by its `<recipe>` name.
+ */
+export function getSliceRecipe(name) {
+    return SLICE_RECIPES.get(name);
+}
+/**
+ * All recipes (for `--list` / help).
+ */
+export function listSliceRecipes() {
+    return RECIPES;
+}
+//# sourceMappingURL=index.js.map

package/out/recipes/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../lib/recipes/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,EAAE,sBAAsB,EAAE,MAAM,4BAA4B,CAAA;AACnE,OAAO,EAAE,iBAAiB,EAAE,MAAM,uBAAuB,CAAA;AACzD,OAAO,EAAE,oBAAoB,EAAE,MAAM,0BAA0B,CAAA;AAC/D,OAAO,EAAE,2BAA2B,EAAE,MAAM,kCAAkC,CAAA;AAC9E,OAAO,EAAE,qBAAqB,EAAE,MAAM,2BAA2B,CAAA;AACjE,OAAO,EAAE,0BAA0B,EAAE,MAAM,iCAAiC,CAAA;AAC5E,OAAO,EAAE,kBAAkB,EAAE,MAAM,yBAAyB,CAAA;AAC5D,OAAO,EAAE,kBAAkB,EAAE,MAAM,yBAAyB,CAAA;AAC5D,OAAO,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAA;AACvD,OAAO,EAAE,eAAe,EAAE,MAAM,qBAAqB,CAAA;AACrD,OAAO,EAAE,aAAa,EAAE,MAAM,mBAAmB,CAAA;AACjD,OAAO,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAA;AAC9C,OAAO,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAA;AACvD,OAAO,EAAE,kBAAkB,EAAE,MAAM,uBAAuB,CAAA;AAC1D,OAAO,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAA;AAC9C,OAAO,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAA;AACvD,OAAO,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAA;AACvD,OAAO,EAAE,cAAc,EAAE,MAAM,oBAAoB,CAAA;AACnD,OAAO,EAAE,iBAAiB,EAAE,MAAM,wBAAwB,CAAA;AAC1D,OAAO,EAAE,WAAW,EAAE,MAAM,iBAAiB,CAAA;AAC7C,OAAO,EAAE,gBAAgB,EAAE,MAAM,uBAAuB,CAAA;AACxD,OAAO,EAAE,0BAA0B,EAAE,MAAM,iCAAiC,CAAA;AAE5E,OAAO,EAAE,mBAAmB,EAAE,MAAM,0BAA0B,CAAA;AAC9D,OAAO,EAAE,YAAY,EAAE,MAAM,iBAAiB,CAAA;AAC9C,OAAO,EAAE,iBAAiB,EAAE,oBAAoB,EAAE,MAAM,uBAAuB,CAAA;AAC/E,OAAO,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAA;AACvD,OAAO,EAAE,cAAc,EAAE,MAAM,oBAAoB,CAAA;AACnD,OAAO,EAAE,oBAAoB,EAAE,MAAM,0BAA0B,CAAA;AAC/D,OAAO,EAAE,UAAU,EAAE,MAAM,eAAe,CAAA;AAE1C,cAAc,mBAAmB,CAAA;AAEjC;;GAEG;AACH,MAAM,OAAO,GAA4B;IACxC,iBAAiB;IACjB,oBAAoB;IACpB,0BAA0B;IAC1B,YAAY;IACZ,gBAAgB;IAChB,iBAAiB;IACjB,oBAAoB;IACpB,cAAc;IACd,gBAAgB;IAChB,WAAW;IACX,gBAAgB;IAChB,UAAU;IACV,kBAAkB;IAClB,YAAY;IACZ,2BAA2B;IAC3B,YAAY;IACZ,aAAa;IACb,kBAAkB;IAClB,kBAAkB;IAClB,gBAAgB;IAChB,eAAe;IACf,0BAA0B;IAC1B,gBAAgB;IAChB,iBAAiB;IACjB,gBAAgB;IAChB,mBAAmB;IACnB,qBAAqB;IACrB,oBAAoB;IACpB,sBAAsB;IACtB,cAAc;CACd,CAAA;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,aAAa,GAAsC,IAAI,GAAG,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,CAAA;AAExG;;GAEG;AACH,MAAM,UAAU,cAAc,CAAC,IAAY;IAC1C,OAAO,aAAa,CAAC,GAAG,CAAC,IAAI,CAAC,CAAA;AAC/B,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,gBAAgB;IAC/B,OAAO,OAAO,CAAA;AACf,CAAC"}

package/out/recipes/intersection.d.ts

@@ -0,0 +1,42 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   `intersection` slice recipe — the REAL-pair intersection training slice (#487). The model scored
+ *   0.0 on intersection_a/b because the training mix had ZERO intersection-labeled rows. This is
+ *   the missing data. Ported from scripts/build-intersection-slice.mjs.
+ *
+ *   STREET PAIRS ARE REAL: the same TIGER 2023 EDGES extraction as the eval builder
+ *   (scripts/eval/build-intersection-real.ts) — a node where two road edges (MTFCC S1*) with
+ *   distinct FULLNAMEs meet is a real crossing. Real pairs avoid teaching fake street-street
+ *   co-occurrences.
+ *
+ *   LEAKAGE POLICY (mirrors the affix slice's VT discipline):
+ *
+ *   - TRAIN counties: Cook IL (grid city) + Morris NJ (suburb).
+ *   - GOLDEN (`--golden`) county: Washington VT (rural) ONLY — the corpus defaultHoldout state.
+ *   - Every crossing in data/eval/external/intersection-real.jsonl is excluded from BOTH modes, by node
+ *       id AND by order-insensitive name pair (the eval shares all three counties).
+ *
+ *   RENDERING: junction-format variety — padded/TIGHT `&` and `/`, `and`, `at`, `@`, leading-phrase
+ *   `corner of` / `intersection of` — crossed with tails (bare / `, ST` / `, ST ZIP` / `, City, ST
+ *   [ZIP]`) and case variants. ZIPs are the crossing's own TIGER edge ZIPL (real); the locality
+ *   tail comes from the OA Cook-county ZIP→city majority map.
+ *
+ *   AUDIT: every emitted row is label-checked on the RAW SURFACE via the #519 char-offset span
+ *   triple. Any violation FAILS the build (throws). A JSON audit report lands next to the output.
+ *
+ *   External inputs (`--edges-dir`, opts.edgesDir; both already on disk — do not re-download):
+ *
+ *   - <edges-dir>/tl_2023_{17031,34027,50023}_edges.shp (unzipped TIGER 2023 EDGES; default
+ *       `$MAILWOMAN_DATA_ROOT/census/tiger2023-edges`, where `mailwoman situs interpolation` puts them)
+ *   - `$MAILWOMAN_DATA_ROOT/oa-cache/us__il__cook.zip` (ZIP→city tails)
+ */
+import { type CorpusRecipe } from "#recipes/scaffold";
+/**
+ * Slice recipe registered with the corpus builder — see the file header for the parse behaviour it exists to exercise,
+ * and `description` below for the surface form it generates.
+ */
+export declare const intersectionRecipe: CorpusRecipe;
+//# sourceMappingURL=intersection.d.ts.map

package/out/recipes/intersection.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"intersection.d.ts","sourceRoot":"","sources":["../../lib/recipes/intersection.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAkCG;AAYH,OAAO,EAAwC,KAAK,YAAY,EAAE,MAAM,mBAAmB,CAAA;AA+X3F;;;GAGG;AACH,eAAO,MAAM,kBAAkB,EAAE,YAwLhC,CAAA"}