@mailwoman/corpus 9.4.0 → 10.1.0

package/out/ve/recipes/reviewed-postcode-tail.js

@@ -0,0 +1,146 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   Renders four hand-reviewed Venezuelan postcode facts in the `locality postcode, region` order.
+ */
+import { readLocalJSONFile } from "@mailwoman/core/fs/readers";
+import { resolveModulePath } from "@mailwoman/core/module/resolvers";
+import { stripCombiningMarks } from "@mailwoman/normalize/fold";
+import { alignAndWrite, recipeSourceID } from "#recipes/scaffold";
+import { defaultRecipeSource } from "#recipes/sources";
+import { SourceRegister } from "#registers";
+import { SurfaceOrigin } from "#types";
+/**
+ * The source label for these rows.
+ *
+ * The rows get their own sampler bucket so an eval can measure them apart from other postcode placements.
+ */
+export const REVIEWED_POSTCODE_TAIL_SOURCE = defaultRecipeSource("synth-reviewed-postcode-tail");
+const REVIEWED_TUPLE_COUNT = 4;
+const ISO_DATE = /^\d{4}-\d{2}-\d{2}$/;
+/**
+ * Returns the path of the committed reviewed-tuple data file.
+ */
+export function defaultReviewedPostcodeTuplePath() {
+    return resolveModulePath("@mailwoman/corpus/data/reviewed-ve-postcode-tuples.json");
+}
+/**
+ * Reads and validates the reviewed tuples.
+ *
+ * @throws When the file does not hold exactly four complete, unique, reviewed tuples.
+ */
+export async function readReviewedPostcodeTuples(path = defaultReviewedPostcodeTuplePath()) {
+    const document = await readLocalJSONFile(path);
+    if (!document.version || !ISO_DATE.test(document.reviewedAt) || document.tuples.length !== REVIEWED_TUPLE_COUNT) {
+        throw new Error("reviewed postcode tuple file must declare a version, review date, and exactly four tuples");
+    }
+    const ids = new Set();
+    const facts = new Set();
+    for (const tuple of document.tuples) {
+        const provenance = tuple.provenance;
+        const fact = `${tuple.locality}\u0000${tuple.postcode}\u0000${tuple.region}`;
+        if (!tuple.id ||
+            !tuple.locality ||
+            !tuple.postcode ||
+            !tuple.region ||
+            tuple.country !== "Venezuela" ||
+            tuple.cc !== "VE" ||
+            tuple.locale !== "es-VE" ||
+            tuple.postcodePlacement !== "after_locality" ||
+            !provenance?.publisher ||
+            !URL.canParse(provenance.url) ||
+            !ISO_DATE.test(provenance.retrievedAt) ||
+            provenance.reviewStatus !== "reviewed" ||
+            !provenance.sourceLicenseNote) {
+            throw new Error(`invalid reviewed postcode tuple: ${tuple.id || "<missing id>"}`);
+        }
+        if (ids.has(tuple.id) || facts.has(fact))
+            throw new Error(`duplicate reviewed postcode tuple: ${tuple.id}`);
+        ids.add(tuple.id);
+        facts.add(fact);
+    }
+    return document.tuples;
+}
+const foldAccents = (value) => stripCombiningMarks(value).normalize("NFC");
+/**
+ * Returns the surface variants of one tuple.
+ *
+ * Every variant reuses the tuple's own locality, postcode and region, so no variant
+ * pairs a postcode with a place the review did not cover.
+ */
+export function reviewedPostcodeTailVariants(tuple) {
+    const tail = `${tuple.locality} ${tuple.postcode}, ${tuple.region}`;
+    const withCountry = `${tail}, ${tuple.country}`;
+    const base = { locality: tuple.locality, postcode: tuple.postcode, region: tuple.region, country: tuple.country };
+    const upper = Object.fromEntries(Object.entries(base).map(([key, value]) => [key, value.toUpperCase()]));
+    const variants = [
+        { id: "canonical-country", raw: withCountry, components: base },
+        {
+            id: "canonical-no-country",
+            raw: tail,
+            components: { locality: tuple.locality, postcode: tuple.postcode, region: tuple.region },
+        },
+        { id: "comma-free", raw: `${tuple.locality} ${tuple.postcode} ${tuple.region} ${tuple.country}`, components: base },
+        { id: "uppercase", raw: withCountry.toUpperCase(), components: upper },
+        {
+            id: "left-context",
+            raw: `Comercio Ejemplo, Calle Principal, ${withCountry}`,
+            components: { venue: "Comercio Ejemplo", street: "Calle Principal", ...base },
+        },
+    ];
+    const foldedLocality = foldAccents(tuple.locality);
+    const foldedRegion = foldAccents(tuple.region);
+    if (foldedLocality !== tuple.locality || foldedRegion !== tuple.region) {
+        variants.push({
+            id: "accent-folded",
+            raw: `${foldedLocality} ${tuple.postcode}, ${foldedRegion}, ${tuple.country}`,
+            components: { locality: foldedLocality, postcode: tuple.postcode, region: foldedRegion, country: tuple.country },
+        });
+    }
+    return variants;
+}
+/**
+ * The corpus recipe that emits every variant of every reviewed tuple.
+ */
+export const reviewedPostcodeTailRecipe = {
+    name: "reviewed-postcode-tail",
+    description: "Reviewed Venezuela locality-postcode-region tails with bounded surface variants",
+    mode: "generate",
+    async run(opts, write) {
+        const tuples = await readReviewedPostcodeTuples(opts.input);
+        let emitted = 0;
+        let skipped = 0;
+        for (const tuple of tuples) {
+            for (const variant of reviewedPostcodeTailVariants(tuple)) {
+                const sourceID = recipeSourceID(REVIEWED_POSTCODE_TAIL_SOURCE, { tuple: tuple.id, variant: variant.id });
+                const canonical = {
+                    raw: variant.raw,
+                    components: variant.components,
+                    country: tuple.cc,
+                    locale: tuple.locale,
+                    source: REVIEWED_POSTCODE_TAIL_SOURCE,
+                    source_id: sourceID,
+                    corpus_version: "0.11.0",
+                    license: `Synthetic rendering of reviewed postal facts — ${tuple.provenance.publisher}; source terms recorded in reviewed-ve-postcode-tuples.json`,
+                };
+                if (alignAndWrite(write, canonical, "reviewed-postcode-tail", {
+                    // A person read these facts from their publishers.
+                    // The data file records the publisher of each tuple.
+                    // No upstream bulk file exists.
+                    register: SourceRegister.ReviewedByHand,
+                    surface: SurfaceOrigin.Composed,
+                    baseSourceID: tuple.id,
+                })) {
+                    emitted++;
+                }
+                else {
+                    skipped++;
+                }
+            }
+        }
+        return { read: tuples.length, emitted, skipped };
+    },
+};
+//# sourceMappingURL=reviewed-postcode-tail.js.map

package/out/ve/recipes/reviewed-postcode-tail.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"reviewed-postcode-tail.js","sourceRoot":"","sources":["../../../lib/ve/recipes/reviewed-postcode-tail.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAEH,OAAO,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAA;AAC9D,OAAO,EAAE,iBAAiB,EAAE,MAAM,kCAAkC,CAAA;AACpE,OAAO,EAAE,mBAAmB,EAAE,MAAM,2BAA2B,CAAA;AAG/D,OAAO,EAAE,aAAa,EAAqB,cAAc,EAAE,MAAM,mBAAmB,CAAA;AACpF,OAAO,EAAE,mBAAmB,EAAE,MAAM,kBAAkB,CAAA;AACtD,OAAO,EAAE,cAAc,EAAE,MAAM,YAAY,CAAA;AAC3C,OAAO,EAAE,aAAa,EAAE,MAAM,QAAQ,CAAA;AAEtC;;;;GAIG;AACH,MAAM,CAAC,MAAM,6BAA6B,GAAG,mBAAmB,CAAC,8BAA8B,CAAC,CAAA;AAEhG,MAAM,oBAAoB,GAAG,CAAC,CAAA;AA6C9B,MAAM,QAAQ,GAAG,qBAAqB,CAAA;AAEtC;;GAEG;AACH,MAAM,UAAU,gCAAgC;IAC/C,OAAO,iBAAiB,CAAC,yDAAyD,CAAC,CAAA;AACpF,CAAC;AAED;;;;GAIG;AACH,MAAM,CAAC,KAAK,UAAU,0BAA0B,CAC/C,IAAI,GAAoB,gCAAgC,EAAE;IAE1D,MAAM,QAAQ,GAAG,MAAM,iBAAiB,CAA4B,IAAI,CAAC,CAAA;IAEzE,IAAI,CAAC,QAAQ,CAAC,OAAO,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC,UAAU,CAAC,IAAI,QAAQ,CAAC,MAAM,CAAC,MAAM,KAAK,oBAAoB,EAAE,CAAC;QACjH,MAAM,IAAI,KAAK,CAAC,2FAA2F,CAAC,CAAA;IAC7G,CAAC;IAED,MAAM,GAAG,GAAG,IAAI,GAAG,EAAU,CAAA;IAC7B,MAAM,KAAK,GAAG,IAAI,GAAG,EAAU,CAAA;IAE/B,KAAK,MAAM,KAAK,IAAI,QAAQ,CAAC,MAAM,EAAE,CAAC;QACrC,MAAM,UAAU,GAAG,KAAK,CAAC,UAAU,CAAA;QACnC,MAAM,IAAI,GAAG,GAAG,KAAK,CAAC,QAAQ,SAAS,KAAK,CAAC,QAAQ,SAAS,KAAK,CAAC,MAAM,EAAE,CAAA;QAE5E,IACC,CAAC,KAAK,CAAC,EAAE;YACT,CAAC,KAAK,CAAC,QAAQ;YACf,CAAC,KAAK,CAAC,QAAQ;YACf,CAAC,KAAK,CAAC,MAAM;YACb,KAAK,CAAC,OAAO,KAAK,WAAW;YAC7B,KAAK,CAAC,EAAE,KAAK,IAAI;YACjB,KAAK,CAAC,MAAM,KAAK,OAAO;YACxB,KAAK,CAAC,iBAAiB,KAAK,gBAAgB;YAC5C,CAAC,UAAU,EAAE,SAAS;YACtB,CAAC,GAAG,CAAC,QAAQ,CAAC,UAAU,CAAC,GAAG,CAAC;YAC7B,CAAC,QAAQ,CAAC,IAAI,CAAC,UAAU,CAAC,WAAW,CAAC;YACtC,UAAU,CAAC,YAAY,KAAK,UAAU;YACtC,CAAC,UAAU,CAAC,iBAAiB,EAC5B,CAAC;YACF,MAAM,IAAI,KAAK,CAAC,oCAAoC,KAAK,CAAC,EAAE,IAAI,cAAc,EAAE,CAAC,CAAA;QAClF,CAAC;QAED,IAAI,GAAG,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC,IAAI,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC;YAAE,MAAM,IAAI,KAAK,CAAC,sCAAsC,KAAK,CAAC,EAAE,EAAE,CAAC,CAAA;QAC3G,GAAG,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC,CAAA;QACjB,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,CAAA;IAChB,CAAC;IAED,OAAO,QAAQ,CAAC,MAAM,CAAA;AACvB,CAAC;AAED,MAAM,WAAW,GAAG,CAAC,KAAa,EAAU,EAAE,CAAC,mBAAmB,CAAC,KAAK,CAAC,CAAC,SAAS,CAAC,KAAK,CAAC,CAAA;AAE1F;;;;;GAKG;AACH,MAAM,UAAU,4BAA4B,CAAC,KAA4B;IACxE,MAAM,IAAI,GAAG,GAAG,KAAK,CAAC,QAAQ,IAAI,KAAK,CAAC,QAAQ,KAAK,KAAK,CAAC,MAAM,EAAE,CAAA;IACnE,MAAM,WAAW,GAAG,GAAG,IAAI,KAAK,KAAK,CAAC,OAAO,EAAE,CAAA;IAC/C,MAAM,IAAI,GAAG,EAAE,QAAQ,EAAE,KAAK,CAAC,QAAQ,EAAE,QAAQ,EAAE,KAAK,CAAC,QAAQ,EAAE,MAAM,EAAE,KAAK,CAAC,MAAM,EAAE,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,CAAA;IACjH,MAAM,KAAK,GAAG,MAAM,CAAC,WAAW,CAAC,MAAM,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,EAAE,KAAK,CAAC,WAAW,EAAE,CAAC,CAAC,CAAC,CAAA;IAExG,MAAM,QAAQ,GAAc;QAC3B,EAAE,EAAE,EAAE,mBAAmB,EAAE,GAAG,EAAE,WAAW,EAAE,UAAU,EAAE,IAAI,EAAE;QAC/D;YACC,EAAE,EAAE,sBAAsB;YAC1B,GAAG,EAAE,IAAI;YACT,UAAU,EAAE,EAAE,QAAQ,EAAE,KAAK,CAAC,QAAQ,EAAE,QAAQ,EAAE,KAAK,CAAC,QAAQ,EAAE,MAAM,EAAE,KAAK,CAAC,MAAM,EAAE;SACxF;QACD,EAAE,EAAE,EAAE,YAAY,EAAE,GAAG,EAAE,GAAG,KAAK,CAAC,QAAQ,IAAI,KAAK,CAAC,QAAQ,IAAI,KAAK,CAAC,MAAM,IAAI,KAAK,CAAC,OAAO,EAAE,EAAE,UAAU,EAAE,IAAI,EAAE;QACnH,EAAE,EAAE,EAAE,WAAW,EAAE,GAAG,EAAE,WAAW,CAAC,WAAW,EAAE,EAAE,UAAU,EAAE,KAAK,EAAE;QACtE;YACC,EAAE,EAAE,cAAc;YAClB,GAAG,EAAE,sCAAsC,WAAW,EAAE;YACxD,UAAU,EAAE,EAAE,KAAK,EAAE,kBAAkB,EAAE,MAAM,EAAE,iBAAiB,EAAE,GAAG,IAAI,EAAE;SAC7E;KACD,CAAA;IAED,MAAM,cAAc,GAAG,WAAW,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAA;IAClD,MAAM,YAAY,GAAG,WAAW,CAAC,KAAK,CAAC,MAAM,CAAC,CAAA;IAE9C,IAAI,cAAc,KAAK,KAAK,CAAC,QAAQ,IAAI,YAAY,KAAK,KAAK,CAAC,MAAM,EAAE,CAAC;QACxE,QAAQ,CAAC,IAAI,CAAC;YACb,EAAE,EAAE,eAAe;YACnB,GAAG,EAAE,GAAG,cAAc,IAAI,KAAK,CAAC,QAAQ,KAAK,YAAY,KAAK,KAAK,CAAC,OAAO,EAAE;YAC7E,UAAU,EAAE,EAAE,QAAQ,EAAE,cAAc,EAAE,QAAQ,EAAE,KAAK,CAAC,QAAQ,EAAE,MAAM,EAAE,YAAY,EAAE,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE;SAChH,CAAC,CAAA;IACH,CAAC;IAED,OAAO,QAAQ,CAAA;AAChB,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,0BAA0B,GAAiB;IACvD,IAAI,EAAE,wBAAwB;IAC9B,WAAW,EAAE,iFAAiF;IAC9F,IAAI,EAAE,UAAU;IAChB,KAAK,CAAC,GAAG,CAAC,IAAI,EAAE,KAAK;QACpB,MAAM,MAAM,GAAG,MAAM,0BAA0B,CAAC,IAAI,CAAC,KAAK,CAAC,CAAA;QAC3D,IAAI,OAAO,GAAG,CAAC,CAAA;QACf,IAAI,OAAO,GAAG,CAAC,CAAA;QAEf,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;YAC5B,KAAK,MAAM,OAAO,IAAI,4BAA4B,CAAC,KAAK,CAAC,EAAE,CAAC;gBAC3D,MAAM,QAAQ,GAAG,cAAc,CAAC,6BAA6B,EAAE,EAAE,KAAK,EAAE,KAAK,CAAC,EAAE,EAAE,OAAO,EAAE,OAAO,CAAC,EAAE,EAAE,CAAC,CAAA;gBAExG,MAAM,SAAS,GAAG;oBACjB,GAAG,EAAE,OAAO,CAAC,GAAG;oBAChB,UAAU,EAAE,OAAO,CAAC,UAAU;oBAC9B,OAAO,EAAE,KAAK,CAAC,EAAE;oBACjB,MAAM,EAAE,KAAK,CAAC,MAAM;oBACpB,MAAM,EAAE,6BAA6B;oBACrC,SAAS,EAAE,QAAQ;oBACnB,cAAc,EAAE,QAAQ;oBACxB,OAAO,EAAE,kDAAkD,KAAK,CAAC,UAAU,CAAC,SAAS,6DAA6D;iBAClJ,CAAA;gBAED,IACC,aAAa,CAAC,KAAK,EAAE,SAAS,EAAE,wBAAwB,EAAE;oBACzD,mDAAmD;oBACnD,qDAAqD;oBACrD,gCAAgC;oBAChC,QAAQ,EAAE,cAAc,CAAC,cAAc;oBACvC,OAAO,EAAE,aAAa,CAAC,QAAQ;oBAC/B,YAAY,EAAE,KAAK,CAAC,EAAE;iBACtB,CAAC,EACD,CAAC;oBACF,OAAO,EAAE,CAAA;gBACV,CAAC;qBAAM,CAAC;oBACP,OAAO,EAAE,CAAA;gBACV,CAAC;YACF,CAAC;QACF,CAAC;QAED,OAAO,EAAE,IAAI,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,CAAA;IACjD,CAAC;CACD,CAAA"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mailwoman/corpus",
-  "version": "9.4.0",
+  "version": "10.1.0",
   "description": "Mailwoman corpus pipeline: BIO-labeled dataset builder for the neural classifier.",
   "license": "AGPL-3.0-only OR LicenseRef-Commercial",
   "repository": {
@@ -14,22 +14,17 @@
     "out/**/*.js.map",
     "out/**/*.d.ts",
     "out/**/*.d.ts.map",
-    "*.ts",
-    "*.tsx",
-    "**/*.ts",
-    "**/*.tsx",
-    "!*.test.ts",
-    "!*.test.tsx",
+    "lib/**/*.ts",
+    "lib/**/*.tsx",
     "!**/*.test.ts",
-    "!**/*.test.tsx",
-    "!test/**"
+    "!**/*.test.tsx"
   ],
   "type": "module",
   "imports": {
-    "#adapters/utils": {
-      "types": "./out/adapters/utils/index.d.ts",
-      "node": "./out/adapters/utils/index.js",
-      "default": "./out/adapters/utils/index.js"
+    "#build": {
+      "types": "./out/build.d.ts",
+      "node": "./out/build.js",
+      "default": "./out/build.js"
     },
     "#runner": {
       "types": "./out/runner.d.ts",
@@ -47,14 +42,14 @@
       "default": "./out/types.js"
     },
     "#utils": {
-      "types": "./out/utils/index.d.ts",
-      "node": "./out/utils/index.js",
-      "default": "./out/utils/index.js"
+      "types": "./out/utils.d.ts",
+      "node": "./out/utils.js",
+      "default": "./out/utils.js"
     },
     "#tools": {
-      "types": "./out/tools/index.d.ts",
-      "node": "./out/tools/index.js",
-      "default": "./out/tools/index.js"
+      "types": "./out/tools.d.ts",
+      "node": "./out/tools.js",
+      "default": "./out/tools.js"
     },
     "#*": {
       "types": "./out/*.d.ts",
@@ -69,151 +64,11 @@
       "node": "./out/index.js",
       "default": "./out/index.js"
     },
-    "./env": {
-      "types": "./out/env.d.ts",
-      "node": "./out/env.js",
-      "default": "./out/env.js"
-    },
-    "./types": {
-      "types": "./out/types.d.ts",
-      "node": "./out/types.js",
-      "default": "./out/types.js"
-    },
     "./data/*": "./data/*",
-    "./utils": {
-      "types": "./out/utils/index.d.ts",
-      "node": "./out/utils/index.js",
-      "default": "./out/utils/index.js"
-    },
-    "./runner": {
-      "types": "./out/runner.d.ts",
-      "node": "./out/runner.js",
-      "default": "./out/runner.js"
-    },
-    "./build": {
-      "types": "./out/build.d.ts",
-      "node": "./out/build.js",
-      "default": "./out/build.js"
-    },
-    "./synthesizers/*": {
-      "types": "./out/synthesizers/*.d.ts",
-      "node": "./out/synthesizers/*.js",
-      "default": "./out/synthesizers/*.js"
-    },
-    "./recipes/*": {
-      "types": "./out/recipes/*.d.ts",
-      "node": "./out/recipes/*.js",
-      "default": "./out/recipes/*.js"
-    },
-    "./adapters/utils": {
-      "types": "./out/adapters/utils/index.d.ts",
-      "node": "./out/adapters/utils/index.js",
-      "default": "./out/adapters/utils/index.js"
-    },
-    "./adapters": {
-      "types": "./out/adapters/index.d.ts",
-      "node": "./out/adapters/index.js",
-      "default": "./out/adapters/index.js"
-    },
-    "./adapters/*": {
-      "types": "./out/adapters/*.d.ts",
-      "node": "./out/adapters/*.js",
-      "default": "./out/adapters/*.js"
-    },
-    "./tools": {
-      "types": "./out/tools/index.d.ts",
-      "node": "./out/tools/index.js",
-      "default": "./out/tools/index.js"
-    },
-    "./codex/*": {
-      "types": "./out/codex/*.d.ts",
-      "node": "./out/codex/*.js",
-      "default": "./out/codex/*.js"
-    },
-    "./utils/parquet": {
-      "types": "./out/utils/parquet.d.ts",
-      "node": "./out/utils/parquet.js",
-      "default": "./out/utils/parquet.js"
-    },
-    "./tools/postcode-triples": {
-      "types": "./out/tools/postcode-triples.d.ts",
-      "node": "./out/tools/postcode-triples.js",
-      "default": "./out/tools/postcode-triples.js"
-    },
-    "./tools/audit": {
-      "types": "./out/tools/audit.d.ts",
-      "node": "./out/tools/audit.js",
-      "default": "./out/tools/audit.js"
-    },
-    "./tools/gb-title-case": {
-      "types": "./out/tools/gb-title-case.d.ts",
-      "node": "./out/tools/gb-title-case.js",
-      "default": "./out/tools/gb-title-case.js"
-    },
-    "./tools/golden-relabel-street": {
-      "types": "./out/tools/golden-relabel-street.d.ts",
-      "node": "./out/tools/golden-relabel-street.js",
-      "default": "./out/tools/golden-relabel-street.js"
-    },
-    "./tools/overture-subvenue": {
-      "types": "./out/tools/overture-subvenue.d.ts",
-      "node": "./out/tools/overture-subvenue.js",
-      "default": "./out/tools/overture-subvenue.js"
-    },
-    "./tools/sub-venue-lexicon": {
-      "types": "./out/tools/sub-venue-lexicon.d.ts",
-      "node": "./out/tools/sub-venue-lexicon.js",
-      "default": "./out/tools/sub-venue-lexicon.js"
-    },
-    "./tools/sub-venue-promotions": {
-      "types": "./out/tools/sub-venue-promotions.d.ts",
-      "node": "./out/tools/sub-venue-promotions.js",
-      "default": "./out/tools/sub-venue-promotions.js"
-    },
-    "./tools/fetch/download": {
-      "types": "./out/tools/fetch/download.d.ts",
-      "node": "./out/tools/fetch/download.js",
-      "default": "./out/tools/fetch/download.js"
-    },
-    "./tools/fetch/ppd": {
-      "types": "./out/tools/fetch/ppd.d.ts",
-      "node": "./out/tools/fetch/ppd.js",
-      "default": "./out/tools/fetch/ppd.js"
-    },
-    "./utils/tokenize": {
-      "types": "./out/utils/tokenize.d.ts",
-      "node": "./out/utils/tokenize.js",
-      "default": "./out/utils/tokenize.js"
-    },
-    "./utils/golden": {
-      "types": "./out/utils/golden.d.ts",
-      "node": "./out/utils/golden.js",
-      "default": "./out/utils/golden.js"
-    },
-    "./utils/license": {
-      "types": "./out/utils/license.d.ts",
-      "node": "./out/utils/license.js",
-      "default": "./out/utils/license.js"
-    },
-    "./utils/split": {
-      "types": "./out/utils/split.d.ts",
-      "node": "./out/utils/split.js",
-      "default": "./out/utils/split.js"
-    },
-    "./utils/wof-json": {
-      "types": "./out/utils/wof-json.d.ts",
-      "node": "./out/utils/wof-json.js",
-      "default": "./out/utils/wof-json.js"
-    },
-    "./test-kit": {
-      "types": "./out/test-kit/index.d.ts",
-      "node": "./out/test-kit/index.js",
-      "default": "./out/test-kit/index.js"
-    },
-    "./test-kit/corpus-recipe": {
-      "types": "./out/test-kit/corpus-recipe.d.ts",
-      "node": "./out/test-kit/corpus-recipe.js",
-      "default": "./out/test-kit/corpus-recipe.js"
+    "./*": {
+      "types": "./out/*.d.ts",
+      "node": "./out/*.js",
+      "default": "./out/*.js"
     }
   },
   "publishConfig": {
@@ -225,158 +80,18 @@
         "node": "./out/index.js",
         "default": "./out/index.js"
       },
-      "./env": {
-        "types": "./out/env.d.ts",
-        "node": "./out/env.js",
-        "default": "./out/env.js"
-      },
-      "./types": {
-        "types": "./out/types.d.ts",
-        "node": "./out/types.js",
-        "default": "./out/types.js"
-      },
       "./data/*": "./data/*",
-      "./utils": {
-        "types": "./out/utils/index.d.ts",
-        "node": "./out/utils/index.js",
-        "default": "./out/utils/index.js"
-      },
-      "./runner": {
-        "types": "./out/runner.d.ts",
-        "node": "./out/runner.js",
-        "default": "./out/runner.js"
-      },
-      "./build": {
-        "types": "./out/build.d.ts",
-        "node": "./out/build.js",
-        "default": "./out/build.js"
-      },
-      "./synthesizers/*": {
-        "types": "./out/synthesizers/*.d.ts",
-        "node": "./out/synthesizers/*.js",
-        "default": "./out/synthesizers/*.js"
-      },
-      "./recipes/*": {
-        "types": "./out/recipes/*.d.ts",
-        "node": "./out/recipes/*.js",
-        "default": "./out/recipes/*.js"
-      },
-      "./adapters/utils": {
-        "types": "./out/adapters/utils/index.d.ts",
-        "node": "./out/adapters/utils/index.js",
-        "default": "./out/adapters/utils/index.js"
-      },
-      "./adapters": {
-        "types": "./out/adapters/index.d.ts",
-        "node": "./out/adapters/index.js",
-        "default": "./out/adapters/index.js"
-      },
-      "./adapters/*": {
-        "types": "./out/adapters/*.d.ts",
-        "node": "./out/adapters/*.js",
-        "default": "./out/adapters/*.js"
-      },
-      "./tools": {
-        "types": "./out/tools/index.d.ts",
-        "node": "./out/tools/index.js",
-        "default": "./out/tools/index.js"
-      },
-      "./codex/*": {
-        "types": "./out/codex/*.d.ts",
-        "node": "./out/codex/*.js",
-        "default": "./out/codex/*.js"
-      },
-      "./utils/parquet": {
-        "types": "./out/utils/parquet.d.ts",
-        "node": "./out/utils/parquet.js",
-        "default": "./out/utils/parquet.js"
-      },
-      "./tools/postcode-triples": {
-        "types": "./out/tools/postcode-triples.d.ts",
-        "node": "./out/tools/postcode-triples.js",
-        "default": "./out/tools/postcode-triples.js"
-      },
-      "./tools/audit": {
-        "types": "./out/tools/audit.d.ts",
-        "node": "./out/tools/audit.js",
-        "default": "./out/tools/audit.js"
-      },
-      "./tools/gb-title-case": {
-        "types": "./out/tools/gb-title-case.d.ts",
-        "node": "./out/tools/gb-title-case.js",
-        "default": "./out/tools/gb-title-case.js"
-      },
-      "./tools/golden-relabel-street": {
-        "types": "./out/tools/golden-relabel-street.d.ts",
-        "node": "./out/tools/golden-relabel-street.js",
-        "default": "./out/tools/golden-relabel-street.js"
-      },
-      "./tools/overture-subvenue": {
-        "types": "./out/tools/overture-subvenue.d.ts",
-        "node": "./out/tools/overture-subvenue.js",
-        "default": "./out/tools/overture-subvenue.js"
-      },
-      "./tools/sub-venue-lexicon": {
-        "types": "./out/tools/sub-venue-lexicon.d.ts",
-        "node": "./out/tools/sub-venue-lexicon.js",
-        "default": "./out/tools/sub-venue-lexicon.js"
-      },
-      "./tools/sub-venue-promotions": {
-        "types": "./out/tools/sub-venue-promotions.d.ts",
-        "node": "./out/tools/sub-venue-promotions.js",
-        "default": "./out/tools/sub-venue-promotions.js"
-      },
-      "./tools/fetch/download": {
-        "types": "./out/tools/fetch/download.d.ts",
-        "node": "./out/tools/fetch/download.js",
-        "default": "./out/tools/fetch/download.js"
-      },
-      "./tools/fetch/ppd": {
-        "types": "./out/tools/fetch/ppd.d.ts",
-        "node": "./out/tools/fetch/ppd.js",
-        "default": "./out/tools/fetch/ppd.js"
-      },
-      "./utils/tokenize": {
-        "types": "./out/utils/tokenize.d.ts",
-        "node": "./out/utils/tokenize.js",
-        "default": "./out/utils/tokenize.js"
-      },
-      "./utils/golden": {
-        "types": "./out/utils/golden.d.ts",
-        "node": "./out/utils/golden.js",
-        "default": "./out/utils/golden.js"
-      },
-      "./utils/license": {
-        "types": "./out/utils/license.d.ts",
-        "node": "./out/utils/license.js",
-        "default": "./out/utils/license.js"
-      },
-      "./utils/split": {
-        "types": "./out/utils/split.d.ts",
-        "node": "./out/utils/split.js",
-        "default": "./out/utils/split.js"
-      },
-      "./utils/wof-json": {
-        "types": "./out/utils/wof-json.d.ts",
-        "node": "./out/utils/wof-json.js",
-        "default": "./out/utils/wof-json.js"
-      },
-      "./test-kit": {
-        "types": "./out/test-kit/index.d.ts",
-        "node": "./out/test-kit/index.js",
-        "default": "./out/test-kit/index.js"
-      },
-      "./test-kit/corpus-recipe": {
-        "types": "./out/test-kit/corpus-recipe.d.ts",
-        "node": "./out/test-kit/corpus-recipe.js",
-        "default": "./out/test-kit/corpus-recipe.js"
+      "./*": {
+        "types": "./out/*.d.ts",
+        "node": "./out/*.js",
+        "default": "./out/*.js"
       }
     },
     "imports": {
-      "#adapters/utils": {
-        "types": "./out/adapters/utils/index.d.ts",
-        "node": "./out/adapters/utils/index.js",
-        "default": "./out/adapters/utils/index.js"
+      "#build": {
+        "types": "./out/build.d.ts",
+        "node": "./out/build.js",
+        "default": "./out/build.js"
       },
       "#runner": {
         "types": "./out/runner.d.ts",
@@ -394,14 +109,14 @@
         "default": "./out/types.js"
       },
       "#utils": {
-        "types": "./out/utils/index.d.ts",
-        "node": "./out/utils/index.js",
-        "default": "./out/utils/index.js"
+        "types": "./out/utils.d.ts",
+        "node": "./out/utils.js",
+        "default": "./out/utils.js"
       },
       "#tools": {
-        "types": "./out/tools/index.d.ts",
-        "node": "./out/tools/index.js",
-        "default": "./out/tools/index.js"
+        "types": "./out/tools.d.ts",
+        "node": "./out/tools.js",
+        "default": "./out/tools.js"
       },
       "#*": {
         "types": "./out/*.d.ts",
@@ -411,27 +126,28 @@
     }
   },
   "dependencies": {
-    "@mailwoman/ban": "9.4.0",
-    "@mailwoman/bdc": "9.4.0",
-    "@mailwoman/codex": "9.4.0",
-    "@mailwoman/core": "9.4.0",
-    "@mailwoman/formatter": "9.4.0",
-    "@mailwoman/normalize": "9.4.0",
-    "@mailwoman/resolver-wof-sqlite": "9.4.0",
-    "@mailwoman/sqlite": "9.4.0",
-    "@mailwoman/tiger": "9.4.0",
+    "@mailwoman/ban": "10.1.0",
+    "@mailwoman/bdc": "10.1.0",
+    "@mailwoman/codex": "10.1.0",
+    "@mailwoman/core": "10.1.0",
+    "@mailwoman/evidence": "10.1.0",
+    "@mailwoman/normalize": "10.1.0",
+    "@mailwoman/query-shape": "10.1.0",
+    "@mailwoman/record": "10.1.0",
+    "@mailwoman/resolver-wof-sqlite": "10.1.0",
+    "@mailwoman/sqlite": "10.1.0",
+    "@mailwoman/tiger": "10.1.0",
     "apache-arrow": "^21.2.0",
-    "fast-glob": "^3.3.3",
     "fastest-levenshtein": "^1.0.16",
     "parquet-wasm": "0.7.2",
-    "path-ts": "^2.3.0",
+    "path-ts": "^2.5.0",
     "read-excel-file": "^9.3.10",
-    "spliterator": "^6.4.1",
+    "spliterator": "^8.1.0",
     "zod": "^4.5.4"
   },
   "devDependencies": {
     "@types/node": "^26.4.1",
-    "adm-zip": "^0.6.0",
+    "adm-zip": "^0.6.1",
     "kysely": "^0.29.5",
     "write-excel-file": "^4.1.1"
   }