@mailwoman/corpus 10.0.0 → 10.1.0

package/out/adapters/index.js

@@ -1,97 +0,0 @@
-/**
- * @copyright Sister Software
- * @license AGPL-3.0
- * @author Teffen Ellis, et al.
- *
- *   Adapter registry bootstrap.
- *
- *   Importing this module registers every built-in adapter with `defaultAdapterRegistry`. The CLI
- *   (`commands/corpus/list.tsx`, `commands/corpus/run.tsx`) imports it once at startup.
- *
- *   Adapters under construction live in their own subdirectories (`./wof-admin-json/`, `./ban/`, ...)
- *   and are added to the `BUILTIN_ADAPTERS` list here as they come online. Tests that need a
- *   pristine registry should construct their own `InMemoryAdapterRegistry` instead of mutating the
- *   default.
- *
- *   The WOF adapters export their canonical ids — `wof-admin` and `wof-postalcode` — so existing
- *   `mailwoman corpus build` callsites do not need to change despite the Phase 1.5.1 SQLite →
- *   JSON-bundle pivot (`./wof-admin-json/` and `./wof-postalcode-json/` directories hold the
- *   implementations; the registered ids are unchanged).
- */
-import { geonamesAdapter } from "#adapters/geonames/adapter";
-import { geonamesPostalAdapter } from "#adapters/geonames/postal/adapter";
-import { openaddressesAdapter } from "#adapters/openaddresses/adapter";
-import { osmAdapter } from "#adapters/osm/adapter";
-import { overtureAdapter } from "#adapters/overture/adapter";
-import { defaultAdapterRegistry } from "#adapters/utils";
-import { wofAdminAdapter } from "#adapters/wof/admin/json/adapter";
-import { wofPostalcodeAdapter } from "#adapters/wof/postalcode-json/adapter";
-import { gnafAdapter } from "#au/adapters/gnaf/adapter";
-import { banAdapter } from "#fr/adapters/ban/adapter";
-import { fccBdcAdapter } from "#us/adapters/fcc-bdc/adapter";
-import { stateHiSchoolsAdapter } from "#us/adapters/state/hi-schools/adapter";
-import { stateIaContractorsAdapter } from "#us/adapters/state/ia-contractors/adapter";
-import { stateNyNotariesAdapter } from "#us/adapters/state/ny-notaries/adapter";
-import { stateTxNotariesAdapter } from "#us/adapters/state/tx-notaries/adapter";
-import { tigerAdapter } from "#us/adapters/tiger/adapter";
-import { usgovHrsaFqhcAdapter } from "#us/adapters/usgov/hrsa-fqhc/adapter";
-import { usgovImlsPlsAdapter } from "#us/adapters/usgov/imls-pls/adapter";
-import { USGovIRSBMFAdapter } from "#us/adapters/usgov/irs-bmf/adapter";
-import { usgovNADAdapter } from "#us/adapters/usgov/nad/adapter";
-import { usgovNPPESAdapter } from "#us/adapters/usgov/nppes/adapter";
-/**
- * Built-in adapters. Order is significant: `corpus build` iterates this list to drive every adapter in turn.
- * Coarse-first (admin → postcode), then street-level (BAN FR, TIGER US, OpenAddresses global), then adversarial-source
- * (FCC BDC US, HRSA FQHC US).
- *
- * The `usgov-samhsa-treatment-locator` adapter is intentionally absent from this list — the SAMHSA Open Data Foundry
- * bulk CSV the adapter was written against is no longer publicly distributed (see issue #33, 2026-05-17 investigation).
- * The factory + named export remain available so the adapter can be hand-registered if an operator obtains a compatible
- * CSV (FOIA, partner channel, upstream restoration). Re-add it here once a stable public source returns.
- */
-export const BUILTIN_ADAPTERS = [
-    wofAdminAdapter,
-    wofPostalcodeAdapter,
-    geonamesAdapter,
-    geonamesPostalAdapter,
-    banAdapter,
-    tigerAdapter,
-    openaddressesAdapter,
-    overtureAdapter,
-    osmAdapter,
-    gnafAdapter,
-    fccBdcAdapter,
-    usgovHrsaFqhcAdapter,
-    usgovNPPESAdapter,
-    usgovNADAdapter,
-    usgovImlsPlsAdapter,
-    USGovIRSBMFAdapter,
-    stateIaContractorsAdapter,
-    stateTxNotariesAdapter,
-    stateNyNotariesAdapter,
-    stateHiSchoolsAdapter,
-];
-for (const adapter of BUILTIN_ADAPTERS) {
-    if (!defaultAdapterRegistry.get(adapter.id)) {
-        defaultAdapterRegistry.register(adapter);
-    }
-}
-export { BAN_ADAPTER_ID, banAdapter } from "#fr/adapters/ban/adapter";
-export { FCC_BDC_ADAPTER_ID, FCC_BDC_DEFAULT_LICENSE, fccBdcAdapter } from "#us/adapters/fcc-bdc/adapter";
-export { GEONAMES_POSTAL_ADAPTER_ID, GEONAMES_POSTAL_DEFAULT_LICENSE, geonamesPostalAdapter, } from "#adapters/geonames/postal/adapter";
-export { GEONAMES_ADAPTER_ID, GEONAMES_DEFAULT_LICENSE, geonamesAdapter } from "#adapters/geonames/adapter";
-export { OPENADDRESSES_ADAPTER_ID, OPENADDRESSES_DEFAULT_LICENSE, openaddressesAdapter, } from "#adapters/openaddresses/adapter";
-export { STATE_HI_SCHOOLS_ADAPTER_ID, STATE_HI_SCHOOLS_DEFAULT_LICENSE, stateHiSchoolsAdapter, } from "#us/adapters/state/hi-schools/adapter";
-export { STATE_IA_CONTRACTORS_ADAPTER_ID, STATE_IA_CONTRACTORS_DEFAULT_LICENSE, stateIaContractorsAdapter, } from "#us/adapters/state/ia-contractors/adapter";
-export { STATE_NY_NOTARIES_ADAPTER_ID, STATE_NY_NOTARIES_DEFAULT_LICENSE, stateNyNotariesAdapter, } from "#us/adapters/state/ny-notaries/adapter";
-export { STATE_TX_NOTARIES_ADAPTER_ID, STATE_TX_NOTARIES_DEFAULT_LICENSE, stateTxNotariesAdapter, } from "#us/adapters/state/tx-notaries/adapter";
-export { TIGER_ADAPTER_ID, TIGER_DEFAULT_LICENSE, tigerAdapter } from "#us/adapters/tiger/adapter";
-export { USGOV_HRSA_FQHC_ADAPTER_ID, USGOV_HRSA_FQHC_DEFAULT_LICENSE, usgovHrsaFqhcAdapter, } from "#us/adapters/usgov/hrsa-fqhc/adapter";
-export { USGOV_IMLS_PLS_ADAPTER_ID, USGOV_IMLS_PLS_DEFAULT_LICENSE, usgovImlsPlsAdapter, } from "#us/adapters/usgov/imls-pls/adapter";
-export { USGOV_IRS_BMF_ADAPTER_ID, USGOV_IRS_BMF_DEFAULT_LICENSE, USGovIRSBMFAdapter, } from "#us/adapters/usgov/irs-bmf/adapter";
-export { USGOV_NAD_ADAPTER_ID, USGOV_NAD_DEFAULT_LICENSE, usgovNADAdapter } from "#us/adapters/usgov/nad/adapter";
-export { USGOV_NPPES_ADAPTER_ID, USGOV_NPPES_DEFAULT_LICENSE, usgovNPPESAdapter, } from "#us/adapters/usgov/nppes/adapter";
-export { USGOV_SAMHSA_ADAPTER_ID, USGOV_SAMHSA_DEFAULT_LICENSE, usgovSamhsaTreatmentLocatorAdapter, } from "#us/adapters/usgov/samhsa-treatment-locator/adapter";
-export { WOF_ADMIN_ADAPTER_ID, wofAdminAdapter } from "#adapters/wof/admin/json/adapter";
-export { WOF_POSTALCODE_ADAPTER_ID, wofPostalcodeAdapter } from "#adapters/wof/postalcode-json/adapter";
-//# sourceMappingURL=index.js.map

package/out/adapters/index.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../../lib/adapters/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;GAmBG;AAEH,OAAO,EAAE,eAAe,EAAE,MAAM,4BAA4B,CAAA;AAC5D,OAAO,EAAE,qBAAqB,EAAE,MAAM,mCAAmC,CAAA;AACzE,OAAO,EAAE,oBAAoB,EAAE,MAAM,iCAAiC,CAAA;AACtE,OAAO,EAAE,UAAU,EAAE,MAAM,uBAAuB,CAAA;AAClD,OAAO,EAAE,eAAe,EAAE,MAAM,4BAA4B,CAAA;AAC5D,OAAO,EAAE,sBAAsB,EAAE,MAAM,iBAAiB,CAAA;AACxD,OAAO,EAAE,eAAe,EAAE,MAAM,kCAAkC,CAAA;AAClE,OAAO,EAAE,oBAAoB,EAAE,MAAM,uCAAuC,CAAA;AAC5E,OAAO,EAAE,WAAW,EAAE,MAAM,2BAA2B,CAAA;AACvD,OAAO,EAAE,UAAU,EAAE,MAAM,0BAA0B,CAAA;AAErD,OAAO,EAAE,aAAa,EAAE,MAAM,8BAA8B,CAAA;AAC5D,OAAO,EAAE,qBAAqB,EAAE,MAAM,uCAAuC,CAAA;AAC7E,OAAO,EAAE,yBAAyB,EAAE,MAAM,2CAA2C,CAAA;AACrF,OAAO,EAAE,sBAAsB,EAAE,MAAM,wCAAwC,CAAA;AAC/E,OAAO,EAAE,sBAAsB,EAAE,MAAM,wCAAwC,CAAA;AAC/E,OAAO,EAAE,YAAY,EAAE,MAAM,4BAA4B,CAAA;AACzD,OAAO,EAAE,oBAAoB,EAAE,MAAM,sCAAsC,CAAA;AAC3E,OAAO,EAAE,mBAAmB,EAAE,MAAM,qCAAqC,CAAA;AACzE,OAAO,EAAE,kBAAkB,EAAE,MAAM,oCAAoC,CAAA;AACvE,OAAO,EAAE,eAAe,EAAE,MAAM,gCAAgC,CAAA;AAChE,OAAO,EAAE,iBAAiB,EAAE,MAAM,kCAAkC,CAAA;AAEpE;;;;;;;;;GASG;AACH,MAAM,CAAC,MAAM,gBAAgB,GAA6B;IACzD,eAAe;IACf,oBAAoB;IACpB,eAAe;IACf,qBAAqB;IACrB,UAAU;IACV,YAAY;IACZ,oBAAoB;IACpB,eAAe;IACf,UAAU;IACV,WAAW;IACX,aAAa;IACb,oBAAoB;IACpB,iBAAiB;IACjB,eAAe;IACf,mBAAmB;IACnB,kBAAkB;IAClB,yBAAyB;IACzB,sBAAsB;IACtB,sBAAsB;IACtB,qBAAqB;CACrB,CAAA;AAED,KAAK,MAAM,OAAO,IAAI,gBAAgB,EAAE,CAAC;IACxC,IAAI,CAAC,sBAAsB,CAAC,GAAG,CAAC,OAAO,CAAC,EAAE,CAAC,EAAE,CAAC;QAC7C,sBAAsB,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAA;IACzC,CAAC;AACF,CAAC;AAED,OAAO,EAAE,cAAc,EAAE,UAAU,EAAE,MAAM,0BAA0B,CAAA;AACrE,OAAO,EAAE,kBAAkB,EAAE,uBAAuB,EAAE,aAAa,EAAE,MAAM,8BAA8B,CAAA;AAEzG,OAAO,EACN,0BAA0B,EAC1B,+BAA+B,EAC/B,qBAAqB,GACrB,MAAM,mCAAmC,CAAA;AAE1C,OAAO,EAAE,mBAAmB,EAAE,wBAAwB,EAAE,eAAe,EAAE,MAAM,4BAA4B,CAAA;AAE3G,OAAO,EACN,wBAAwB,EACxB,6BAA6B,EAC7B,oBAAoB,GACpB,MAAM,iCAAiC,CAAA;AAExC,OAAO,EACN,2BAA2B,EAC3B,gCAAgC,EAChC,qBAAqB,GACrB,MAAM,uCAAuC,CAAA;AAE9C,OAAO,EACN,+BAA+B,EAC/B,oCAAoC,EACpC,yBAAyB,GACzB,MAAM,2CAA2C,CAAA;AAElD,OAAO,EACN,4BAA4B,EAC5B,iCAAiC,EACjC,sBAAsB,GACtB,MAAM,wCAAwC,CAAA;AAE/C,OAAO,EACN,4BAA4B,EAC5B,iCAAiC,EACjC,sBAAsB,GACtB,MAAM,wCAAwC,CAAA;AAE/C,OAAO,EAAE,gBAAgB,EAAE,qBAAqB,EAAE,YAAY,EAAE,MAAM,4BAA4B,CAAA;AAElG,OAAO,EACN,0BAA0B,EAC1B,+BAA+B,EAC/B,oBAAoB,GACpB,MAAM,sCAAsC,CAAA;AAE7C,OAAO,EACN,yBAAyB,EACzB,8BAA8B,EAC9B,mBAAmB,GACnB,MAAM,qCAAqC,CAAA;AAE5C,OAAO,EACN,wBAAwB,EACxB,6BAA6B,EAC7B,kBAAkB,GAClB,MAAM,oCAAoC,CAAA;AAE3C,OAAO,EAAE,oBAAoB,EAAE,yBAAyB,EAAE,eAAe,EAAE,MAAM,gCAAgC,CAAA;AAEjH,OAAO,EACN,sBAAsB,EACtB,2BAA2B,EAC3B,iBAAiB,GACjB,MAAM,kCAAkC,CAAA;AAEzC,OAAO,EACN,uBAAuB,EACvB,4BAA4B,EAC5B,kCAAkC,GAClC,MAAM,qDAAqD,CAAA;AAE5D,OAAO,EAAE,oBAAoB,EAAE,eAAe,EAAE,MAAM,kCAAkC,CAAA;AACxF,OAAO,EAAE,yBAAyB,EAAE,oBAAoB,EAAE,MAAM,uCAAuC,CAAA"}

package/out/adapters/synth-po-box/adapter.d.ts

@@ -1,67 +0,0 @@
-/**
- * @copyright Sister Software
- * @license AGPL-3.0
- * @author Teffen Ellis, et al.
- *
- *   `synth-po-box`: PO box / PMB / Apartado / BP synthesizer adapter.
- *
- *   Consumes a JSONL stream of (locality, region, postcode, country) tuples — typically extracted
- *   from existing corpus output (TIGER/NAD/BAN/WOF) — and emits synthetic PO box training rows. See
- *   `@mailwoman/corpus/synthesizers/po-box` for the per-locale templates and number-noise logic.
- *
- *   Why an adapter and not an augmenter:
- *
- *   - Per USPS Pub 28 / DMM 508, a PO box delivery line is mutually exclusive with a street line.
- *       Synthesizing PO boxes by mutating a street row would teach the model an invalid pattern.
- *       The clean shape is: read just (locality, region, postcode, country) and produce a fresh
- *       PO-box-shaped row.
- *   - Per-DeepSeek (3-turn consult, 2026-05-28): PMB rows that COMBINE a street line with a PMB number
- *       ARE valid (CMRA addresses). Those are produced when `pmbRatio > 0` AND the input tuple
- *       carries a `street` field.
- */
-import { type PoBoxBaseTuple } from "#synthesizers/po-box";
-import type { CorpusAdapter } from "#types";
-/**
- * Registry id for this adapter. Stamped into every row it emits, so a corpus record can be traced back to the dataset
- * it came from.
- */
-export declare const SYNTH_PO_BOX_ADAPTER_ID = "synth-po-box";
-/**
- * License for the synthetic PO-box rows. The output is generated, but it inherits the terms of the real tuples it is
- * derived from, so the attribution travels with it.
- */
-export declare const SYNTH_PO_BOX_LICENSE = "Synthetic \u2014 derived from CC-BY / public-domain input tuples";
-export interface PoBoxInputRow extends PoBoxBaseTuple {
-    street?: string;
-    houseNumber?: string;
-}
-export interface SynthPoBoxAdapterOptions {
-    /**
-     * How many PO box variants to emit per input tuple. Each variant picks a different leader (and possibly a different
-     * number / noise level). Default 1.
-     */
-    variantsPerInput?: number;
-    /**
-     * Probability (0..1) of emitting a PMB-with-street variant when both the input has a street and the locale supports
-     * PMB. Default 0.15.
-     */
-    pmbRatio?: number;
-    /**
-     * Deterministic seed for reproducible synthesis. Default Date.now().
-     */
-    seed?: number;
-    /**
-     * Probability (0..1), evaluated per input tuple, of ALSO emitting one US military/diplomatic PO-box row
-     * (`PSC/CMR/Unit <id> Box <box>, APO/FPO/DPO AA/AE/AP <zip>`, #517). These rows are self-contained — they draw no
-     * field from the input tuple, so military volume scales with the input stream size. Default 0 (off) — the adapter's
-     * contract is "one row per input"; the corpus build recipe opts in to seed the rare-but-real military class without
-     * changing the default.
-     */
-    militaryRatio?: number;
-}
-export declare function createSynthPoBoxAdapter(opts?: SynthPoBoxAdapterOptions): CorpusAdapter;
-/**
- * The configured adapter instance registered with the corpus builder.
- */
-export declare const synthPoBoxAdapter: CorpusAdapter;
-//# sourceMappingURL=adapter.d.ts.map

package/out/adapters/synth-po-box/adapter.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../lib/adapters/synth-po-box/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AAOH,OAAO,EAKN,KAAK,cAAc,EACnB,MAAM,sBAAsB,CAAA;AAC7B,OAAO,KAAK,EAAgC,aAAa,EAAE,MAAM,QAAQ,CAAA;AAEzE;;;GAGG;AACH,eAAO,MAAM,uBAAuB,iBAAiB,CAAA;AACrD;;;GAGG;AACH,eAAO,MAAM,oBAAoB,qEAAgE,CAAA;AAEjG,MAAM,WAAW,aAAc,SAAQ,cAAc;IACpD,MAAM,CAAC,EAAE,MAAM,CAAA;IACf,WAAW,CAAC,EAAE,MAAM,CAAA;CACpB;AAED,MAAM,WAAW,wBAAwB;IACxC;;;OAGG;IACH,gBAAgB,CAAC,EAAE,MAAM,CAAA;IACzB;;;OAGG;IACH,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB;;OAEG;IACH,IAAI,CAAC,EAAE,MAAM,CAAA;IACb;;;;;;OAMG;IACH,aAAa,CAAC,EAAE,MAAM,CAAA;CACtB;AAED,wBAAgB,uBAAuB,CAAC,IAAI,GAAE,wBAA6B,GAAG,aAAa,CAyH1F;AAED;;GAEG;AACH,eAAO,MAAM,iBAAiB,eAA4B,CAAA"}

package/out/adapters/synth-po-box/adapter.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../lib/adapters/synth-po-box/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AAEH,OAAO,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAA;AACrD,OAAO,EAAE,OAAO,EAAE,MAAM,wBAAwB,CAAA;AAChD,OAAO,EAAE,eAAe,EAAE,MAAM,aAAa,CAAA;AAE7C,OAAO,EAAE,cAAc,EAAE,MAAM,iBAAiB,CAAA;AAChD,OAAO,EACN,eAAe,EACf,uBAAuB,EACvB,0BAA0B,EAC1B,kBAAkB,GAElB,MAAM,sBAAsB,CAAA;AAG7B;;;GAGG;AACH,MAAM,CAAC,MAAM,uBAAuB,GAAG,cAAc,CAAA;AACrD;;;GAGG;AACH,MAAM,CAAC,MAAM,oBAAoB,GAAG,6DAA6D,CAAA;AAgCjG,MAAM,UAAU,uBAAuB,CAAC,IAAI,GAA6B,EAAE;IAC1E,MAAM,gBAAgB,GAAG,IAAI,CAAC,gBAAgB,IAAI,CAAC,CAAA;IACnD,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,IAAI,IAAI,CAAA;IACtC,MAAM,aAAa,GAAG,IAAI,CAAC,aAAa,IAAI,CAAC,CAAA;IAE7C,OAAO;QACN,EAAE,EAAE,uBAAuB;QAC3B,cAAc,EAAE,oBAAoB;QACpC,WAAW,EACV,sKAAsK;QAEvK,KAAK,CAAC,CAAC,IAAI,CAAC,OAAuB;YAClC,MAAM,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;YAE/C,4FAA4F;YAC5F,0FAA0F;YAC1F,4DAA4D;YAC5D,MAAM,KAAK,GAAG,eAAe,CAAC,SAAS,CAAC,OAAO,CAAC,SAAS,CAAC,CAAA;YAE1D,IAAI,OAAO,GAAG,CAAC,CAAA;YACf,IAAI,OAAO,GAAG,CAAC,CAAA;YACf,IAAI,WAAW,GAAG,CAAC,CAAA;YAEnB,IAAI,KAAK,EAAE,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;gBAChC,IAAI,OAAO,CAAC,MAAM,EAAE,OAAO;oBAAE,MAAK;gBAElC,IAAI,OAAO,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,OAAO,CAAC,KAAK;oBAAE,MAAK;gBAElE,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAA;gBAE3B,IAAI,CAAC,OAAO;oBAAE,SAAQ;gBAEtB,MAAM,KAAK,GAAG,cAAc,CAAgB,OAAO,CAAC,CAAA;gBAEpD,IAAI,KAAK,KAAK,IAAI,EAAE,CAAC;oBACpB,OAAO,EAAE,CAAA;oBAET,SAAQ;gBACT,CAAC;gBAED,yFAAyF;gBACzF,wFAAwF;gBACxF,qDAAqD;gBACrD,MAAM,cAAc,GAAG,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,uBAAuB,CAAC,GAAG,CAAC,eAAe,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAA;gBAE1G,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,IAAI,CAAC,CAAC,KAAK,CAAC,MAAM,IAAI,CAAC,cAAc,CAAC,EAAE,CAAC;oBAChG,OAAO,EAAE,CAAA;oBAET,SAAQ;gBACT,CAAC;gBAED,IAAI,OAAO,CAAC,OAAO,IAAI,OAAO,CAAC,OAAO,KAAK,KAAK,CAAC,OAAO;oBAAE,SAAQ;gBAElE,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,gBAAgB,EAAE,CAAC,EAAE,EAAE,CAAC;oBAC3C,MAAM,KAAK,GAAG,kBAAkB,CAAC,KAAK,EAAE,EAAE,MAAM,EAAE,QAAQ,EAAE,CAAC,CAAA;oBAE7D,IAAI,CAAC,KAAK;wBAAE,SAAQ;oBAEpB,6EAA6E;oBAC7E,iDAAiD;oBACjD,MAAM,QAAQ,GAAG,cAAc,CAAC,uBAAuB,EAAE;wBACxD,QAAQ,EAAE,GAAG,KAAK,CAAC,QAAQ,IAAI,CAAC,EAAE;wBAClC,MAAM,EAAE,KAAK,CAAC,MAAM;wBACpB,QAAQ,EAAE,KAAK,CAAC,QAAQ;wBACxB,OAAO,EAAE,KAAK,CAAC,OAAO;qBACtB,CAAC,CAAA;oBAEF,MAAM;wBACL,GAAG,EAAE,KAAK,CAAC,GAAG;wBACd,UAAU,EAAE,KAAK,CAAC,UAAU;wBAC5B,OAAO,EAAE,KAAK,CAAC,OAAO;wBACtB,MAAM,EAAE,KAAK,CAAC,MAAM;wBACpB,MAAM,EAAE,uBAAuB;wBAC/B,SAAS,EAAE,QAAQ;wBACnB,cAAc,EAAE,EAAE;wBAClB,OAAO,EAAE,oBAAoB;qBAC7B,CAAA;oBAED,OAAO,EAAE,CAAA;oBAET,IAAI,OAAO,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,OAAO,CAAC,KAAK;wBAAE,MAAK;gBACnE,CAAC;gBAED,0FAA0F;gBAC1F,0FAA0F;gBAC1F,wFAAwF;gBACxF,2FAA2F;gBAC3F,kDAAkD;gBAClD,MAAM,eAAe,GAAG,CAAC,OAAO,CAAC,OAAO,IAAI,OAAO,CAAC,OAAO,KAAK,IAAI,CAAA;gBAEpE,IACC,aAAa,GAAG,CAAC;oBACjB,eAAe;oBACf,CAAC,OAAO,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC;oBACxD,MAAM,EAAE,GAAG,aAAa,EACvB,CAAC;oBACF,MAAM,GAAG,GAAG,0BAA0B,CAAC,EAAE,MAAM,EAAE,CAAC,CAAA;oBAElD,MAAM,QAAQ,GAAG,cAAc,CAAC,uBAAuB,EAAE;wBACxD,MAAM,EAAE,GAAG,GAAG,CAAC,UAAU,CAAC,MAAM,OAAO,WAAW,EAAE,EAAE;wBACtD,QAAQ,EAAE,GAAG,CAAC,UAAU,CAAC,QAAS;wBAClC,MAAM,EAAE,GAAG,CAAC,UAAU,CAAC,MAAO;wBAC9B,QAAQ,EAAE,GAAG,CAAC,UAAU,CAAC,QAAS;qBAClC,CAAC,CAAA;oBAEF,MAAM;wBACL,GAAG,EAAE,GAAG,CAAC,GAAG;wBACZ,UAAU,EAAE,GAAG,CAAC,UAAU;wBAC1B,OAAO,EAAE,IAAI;wBACb,MAAM,EAAE,GAAG,CAAC,MAAM;wBAClB,MAAM,EAAE,uBAAuB;wBAC/B,SAAS,EAAE,QAAQ;wBACnB,cAAc,EAAE,EAAE;wBAClB,OAAO,EAAE,oBAAoB;qBAC7B,CAAA;oBAED,OAAO,EAAE,CAAA;gBACV,CAAC;YACF,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,iBAAiB,GAAG,uBAAuB,EAAE,CAAA"}

package/out/adapters/utils/index.d.ts

@@ -1,191 +0,0 @@
-/**
- * @copyright Sister Software
- * @license AGPL-3.0
- * @author Teffen Ellis, et al.
- *
- *   Adapter framework helpers — the pieces every corpus adapter and the runner share.
- *
- *   This file does **not** define `CorpusAdapter` (that lives in `./types.ts`, which is the single
- *   canonical schema module). It exposes:
- *
- *   - `AdapterRegistry`: a tiny lookup table the CLI + build pipeline use to find adapters by id.
- *   - `InMemoryAdapterRegistry`: the default implementation.
- *   - `stableSourceID(adapterID, components)`: deterministic content-addressed id for adapters whose
- *       source data has no native primary key (CSV, GeoJSON).
- *   - `HOUSE_NUMBER_PREFIX` + `splitStreetLine(line)`: the one house-number/street split every
- *       US CSV adapter uses.
- *   - `composeRaw(parts)`: the envelope-style raw line every US CSV adapter renders.
- *   - `canonicalDedupKey(row)`: normalized signature used to drop near-identical rows during a run.
- *       Adapter-internal dedup; cross-adapter dedup is the runner's job.
- *   - `streamingSha256()`: thin wrapper around `node:crypto` so the runner can hash JSONL output as it
- *       streams (avoids re-reading the slice for the manifest checksum).
- *   - `loadLibpostalDictionary(language, filename)`: the curated libpostal dictionary reader the
- *       street-decompose modules share — the one read this module performs; everything else here is
- *       pure, and other side-effecting code goes in `./runner.ts`.
- */
-import type { ComponentTag } from "@mailwoman/codex/component";
-import type { CanonicalRow, CorpusAdapter } from "#types";
-/**
- * Lookup table for corpus adapters.
- *
- * The CLI's `npx mailwoman corpus run <adapter-id>` resolves `<adapter-id>` against this registry; the same registry is
- * iterated by the `corpus build` pipeline. Adapters do not self-register at module load — they're added explicitly so
- * the dependency graph stays traceable.
- */
-export interface AdapterRegistry {
-    /**
-     * Add an adapter. Throws if `adapter.id` is already registered.
-     */
-    register(adapter: CorpusAdapter): void;
-    /**
-     * Return the adapter for `id`, or `undefined`.
-     */
-    get(id: string): CorpusAdapter | undefined;
-    /**
-     * All registered adapters, in insertion order.
-     */
-    list(): readonly CorpusAdapter[];
-    /**
-     * Convenience: ids only, in insertion order.
-     */
-    ids(): readonly string[];
-}
-/**
- * Default in-memory registry. The runner constructs one per invocation; the CLI re-uses a shared singleton
- * (`defaultAdapterRegistry`) populated by `./adapters/index.ts` as adapters come online.
- */
-export declare class InMemoryAdapterRegistry implements AdapterRegistry {
-    #private;
-    register(adapter: CorpusAdapter): void;
-    get(id: string): CorpusAdapter | undefined;
-    list(): readonly CorpusAdapter[];
-    ids(): readonly string[];
-}
-/**
- * Process-wide default registry. Populated by `./adapters/index.ts` as adapters are built; imported by the CLI. Tests
- * should construct their own `InMemoryAdapterRegistry` to avoid cross-test pollution.
- */
-export declare const defaultAdapterRegistry: InMemoryAdapterRegistry;
-/**
- * Deterministic content-addressed source id.
- *
- * For adapters whose upstream source has no native primary key (CSV rows, GeoJSON features), the runner expects a
- * stable id so dedup, holdout manifests, and resumability work across reruns. This helper produces one by hashing the
- * adapter id and a canonical serialization of the components dict (keys sorted, values verbatim).
- *
- * Output format: `<adapterID>-<first-12-hex-chars-of-sha256>`. 48 bits of entropy is enough for ~17M rows per adapter
- * before the expected collision count exceeds 1 (birthday paradox); adapters with more rows should extend the prefix
- * length.
- */
-export declare function stableSourceID(adapterID: string, components: Partial<Record<ComponentTag, string>>): string;
-/**
- * {@link stableSourceID} over arbitrary disambiguator keys — a variant index, a slot number, anything that is not a
- * `ComponentTag`. Every key handed in is sorted and hashed either way; only the key vocabulary differs, and the narrow
- * signature above is what stops an adapter hashing a misspelled component name.
- */
-export declare function stableSourceIDFromParts(adapterID: string, parts: Readonly<Record<string, string | undefined>>): string;
-/**
- * Leading house number on a US-style street line: digits, an optional hyphenated range (Queens `40-12`), an optional
- * single alpha suffix (`101A`), then whitespace, then the rest.
- *
- * This decides where the house number ends and the street begins for every US CSV adapter, so an edge case fixed here
- * is fixed for all of them.
- *
- * The remainder must stay `(\S.*)` and must NOT become `(.+)`: `\s+` and `.` both match a tab, so `\s+(.+)$` lets the
- * engine split a run of tabs between the two groups every possible way before failing — quadratic backtracking on
- * attacker-shaped input. Requiring a non-space start removes the overlap. Group 2 is trimmed by the caller either way,
- * so the two forms are indistinguishable on real input; only the failure cost differs.
- */
-export declare const HOUSE_NUMBER_PREFIX: RegExp;
-/**
- * A street line split into its house number and the remainder.
- */
-export interface SplitStreetLine {
-    house_number?: string;
-    street: string;
-}
-/**
- * Split a US-style street line on {@link HOUSE_NUMBER_PREFIX}.
- *
- * The US CSV sources follow USPS Publication 28 conventions with hand-entry drift. The leading digit run is the house
- * number (`"123 Main St"`, `"6450 W Indian School Rd"`), and the regex tolerates one trailing letter (`"123A Main St"`)
- * plus an optional hyphenated half (`"40-12 Bell Blvd"`, common in NYC and suburban garden-apartment numbering; Hawaii
- * uses it island-wide — `"47-470 Hui Aeko Place"`).
- *
- * Returns `null` for blank input. Anything that does not match the prefix shape (`"PO Box 1234"`, `"RR 2 Box 67"`, `"HC
- * 1"`) becomes a single `street` value rather than being mangled — the model sees the original surface form and
- * downstream classifiers pick it up. Callers that need those forms recognized as something other than a street (see
- * `usgov-irs-bmf`) test for them BEFORE calling this.
- */
-export declare function splitStreetLine(line: string): SplitStreetLine | null;
-/**
- * Parts of the envelope-style raw line {@link composeRaw} renders. Every field is dropped when empty, so a caller only
- * supplies what its source carries.
- */
-export interface ComposeRawParts {
-    venue?: string;
-    houseNumber?: string;
-    street?: string;
-    /**
-     * A street line already joined by the caller (e.g. a PO-box phrase) — used INSTEAD of `houseNumber`/`street`/`unit`
-     * when present.
-     */
-    streetLine?: string;
-    unit?: string;
-    locality: string;
-    region?: string;
-    postcode: string;
-}
-/**
- * Compose the envelope-style raw address line the US CSV adapters share:
- *
- * `"<venue>, <house> <street> <unit>, <city>, <state> <postcode>"`
- *
- * The venue leads (US conventional addressee-then-address ordering) so a downstream model sees the
- * venue-prefix-then-address shape users actually type into geocoders. Empty segments drop out rather than leaving a
- * dangling separator.
- */
-export declare function composeRaw(parts: ComposeRawParts): string;
-/**
- * Load one curated libpostal dictionary (`core/data/libpostal/dictionaries/<language>/<filename>`) as a lower-cased
- * form set. libpostal format: `canonical|abbr|abbr|...` — every form is indexed.
- *
- * `resourceDictionaryPath` already resolves both layouts — `core/data/...` from source and from the packaged `out/`
- * tree. The candidate list this replaced named it TWICE and then guessed a third path off `process.cwd()`, and
- * swallowed every error while probing, so a corrupt dictionary reported as a missing one.
- *
- * The largest libpostal dictionary is 8.4 KB, and each caller runs this once per process at module load.
- */
-export declare function loadLibpostalDictionary(language: string, filename: string): Promise<Set<string>>;
-/**
- * Canonical dedup key for a row.
- *
- * Two rows that share this key are treated as duplicates and only the first wins. The key is built from `country`, the
- * sorted `components` dict, and a normalized `raw` (lower-cased, whitespace collapsed). License and provenance fields
- * are intentionally excluded so the same address from multiple adapters is recognized as a duplicate.
- *
- * Synthetic rows are never deduplicated against natural rows: `synth.method` is folded into the key when present,
- * ensuring each augmentation variant survives.
- */
-export declare function canonicalDedupKey(row: CanonicalRow): string;
-/**
- * Streaming SHA-256 hasher.
- *
- * The runner feeds every JSONL line into one of these so the per-slice checksum can be recorded in `MANIFEST.json`
- * without a second pass over the slice. Implementation is a one-line wrapper, but giving it a name keeps the runner's
- * hash-tracking intent obvious.
- */
-export interface StreamingHasher {
-    update(chunk: string | Uint8Array): void;
-    digest(): string;
-}
-/**
- * Default `StreamingHasher` (SHA-256, hex).
- */
-/**
- * @see {@link StreamingHasher} — this is the incremental counterpart to `sha256Hex`, not a copy of
- *   it. `@mailwoman/core/utils` exposes one-shot digests (`sha256Hex`) and whole-file digests
- *   (`sha256File`); the runner needs neither, because it hashes JSONL lines as they stream past.
- */
-export declare function streamingSha256(): StreamingHasher;
-//# sourceMappingURL=index.d.ts.map

package/out/adapters/utils/index.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../lib/adapters/utils/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;GAwBG;AAEH,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,4BAA4B,CAAA;AAO9D,OAAO,KAAK,EAAE,YAAY,EAAE,aAAa,EAAE,MAAM,QAAQ,CAAA;AAEzD;;;;;;GAMG;AACH,MAAM,WAAW,eAAe;IAC/B;;OAEG;IACH,QAAQ,CAAC,OAAO,EAAE,aAAa,GAAG,IAAI,CAAA;IAEtC;;OAEG;IACH,GAAG,CAAC,EAAE,EAAE,MAAM,GAAG,aAAa,GAAG,SAAS,CAAA;IAE1C;;OAEG;IACH,IAAI,IAAI,SAAS,aAAa,EAAE,CAAA;IAEhC;;OAEG;IACH,GAAG,IAAI,SAAS,MAAM,EAAE,CAAA;CACxB;AAED;;;GAGG;AACH,qBAAa,uBAAwB,YAAW,eAAe;;IAG9D,QAAQ,CAAC,OAAO,EAAE,aAAa,GAAG,IAAI,CAMrC;IAED,GAAG,CAAC,EAAE,EAAE,MAAM,GAAG,aAAa,GAAG,SAAS,CAEzC;IAED,IAAI,IAAI,SAAS,aAAa,EAAE,CAE/B;IAED,GAAG,IAAI,SAAS,MAAM,EAAE,CAEvB;CACD;AAED;;;GAGG;AACH,eAAO,MAAM,sBAAsB,yBAAgC,CAAA;AAEnE;;;;;;;;;;GAUG;AACH,wBAAgB,cAAc,CAAC,SAAS,EAAE,MAAM,EAAE,UAAU,EAAE,OAAO,CAAC,MAAM,CAAC,YAAY,EAAE,MAAM,CAAC,CAAC,GAAG,MAAM,CAE3G;AAED;;;;GAIG;AACH,wBAAgB,uBAAuB,CACtC,SAAS,EAAE,MAAM,EACjB,KAAK,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC,CAAC,GACjD,MAAM,CAQR;AAED;;;;;;;;;;;GAWG;AACH,eAAO,MAAM,mBAAmB,QAAuC,CAAA;AAEvE;;GAEG;AACH,MAAM,WAAW,eAAe;IAC/B,YAAY,CAAC,EAAE,MAAM,CAAA;IACrB,MAAM,EAAE,MAAM,CAAA;CACd;AAED;;;;;;;;;;;;GAYG;AACH,wBAAgB,eAAe,CAAC,IAAI,EAAE,MAAM,GAAG,eAAe,GAAG,IAAI,CAUpE;AAED;;;GAGG;AACH,MAAM,WAAW,eAAe;IAC/B,KAAK,CAAC,EAAE,MAAM,CAAA;IACd,WAAW,CAAC,EAAE,MAAM,CAAA;IACpB,MAAM,CAAC,EAAE,MAAM,CAAA;IACf;;;OAGG;IACH,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB,IAAI,CAAC,EAAE,MAAM,CAAA;IACb,QAAQ,EAAE,MAAM,CAAA;IAChB,MAAM,CAAC,EAAE,MAAM,CAAA;IACf,QAAQ,EAAE,MAAM,CAAA;CAChB;AAED;;;;;;;;GAQG;AACH,wBAAgB,UAAU,CAAC,KAAK,EAAE,eAAe,GAAG,MAAM,CASzD;AAED;;;;;;;;;GASG;AACH,wBAAsB,uBAAuB,CAAC,QAAQ,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAmBtG;AAED;;;;;;;;;GASG;AACH,wBAAgB,iBAAiB,CAAC,GAAG,EAAE,YAAY,GAAG,MAAM,CAO3D;AAED;;;;;;GAMG;AACH,MAAM,WAAW,eAAe;IAC/B,MAAM,CAAC,KAAK,EAAE,MAAM,GAAG,UAAU,GAAG,IAAI,CAAA;IACxC,MAAM,IAAI,MAAM,CAAA;CAChB;AAED;;GAEG;AACH;;;;GAIG;AACH,wBAAgB,eAAe,IAAI,eAAe,CAmBjD"}

package/out/adapters/utils/index.js

@@ -1,206 +0,0 @@
-/**
- * @copyright Sister Software
- * @license AGPL-3.0
- * @author Teffen Ellis, et al.
- *
- *   Adapter framework helpers — the pieces every corpus adapter and the runner share.
- *
- *   This file does **not** define `CorpusAdapter` (that lives in `./types.ts`, which is the single
- *   canonical schema module). It exposes:
- *
- *   - `AdapterRegistry`: a tiny lookup table the CLI + build pipeline use to find adapters by id.
- *   - `InMemoryAdapterRegistry`: the default implementation.
- *   - `stableSourceID(adapterID, components)`: deterministic content-addressed id for adapters whose
- *       source data has no native primary key (CSV, GeoJSON).
- *   - `HOUSE_NUMBER_PREFIX` + `splitStreetLine(line)`: the one house-number/street split every
- *       US CSV adapter uses.
- *   - `composeRaw(parts)`: the envelope-style raw line every US CSV adapter renders.
- *   - `canonicalDedupKey(row)`: normalized signature used to drop near-identical rows during a run.
- *       Adapter-internal dedup; cross-adapter dedup is the runner's job.
- *   - `streamingSha256()`: thin wrapper around `node:crypto` so the runner can hash JSONL output as it
- *       streams (avoids re-reading the slice for the manifest checksum).
- *   - `loadLibpostalDictionary(language, filename)`: the curated libpostal dictionary reader the
- *       street-decompose modules share — the one read this module performs; everything else here is
- *       pure, and other side-effecting code goes in `./runner.ts`.
- */
-import { readLocalTextFile } from "@mailwoman/core/fs/readers";
-import { sha256Hex, createHash } from "@mailwoman/core/hash";
-import { isPresent } from "@mailwoman/core/objects";
-import { resourceDictionaryPath } from "@mailwoman/core/paths";
-import { TextSpliterator } from "spliterator";
-/**
- * Default in-memory registry. The runner constructs one per invocation; the CLI re-uses a shared singleton
- * (`defaultAdapterRegistry`) populated by `./adapters/index.ts` as adapters come online.
- */
-export class InMemoryAdapterRegistry {
-    #byID = new Map();
-    register(adapter) {
-        if (this.#byID.has(adapter.id)) {
-            throw new Error(`AdapterRegistry: id ${JSON.stringify(adapter.id)} already registered`);
-        }
-        this.#byID.set(adapter.id, adapter);
-    }
-    get(id) {
-        return this.#byID.get(id);
-    }
-    list() {
-        return Array.from(this.#byID.values());
-    }
-    ids() {
-        return Array.from(this.#byID.keys());
-    }
-}
-/**
- * Process-wide default registry. Populated by `./adapters/index.ts` as adapters are built; imported by the CLI. Tests
- * should construct their own `InMemoryAdapterRegistry` to avoid cross-test pollution.
- */
-export const defaultAdapterRegistry = new InMemoryAdapterRegistry();
-/**
- * Deterministic content-addressed source id.
- *
- * For adapters whose upstream source has no native primary key (CSV rows, GeoJSON features), the runner expects a
- * stable id so dedup, holdout manifests, and resumability work across reruns. This helper produces one by hashing the
- * adapter id and a canonical serialization of the components dict (keys sorted, values verbatim).
- *
- * Output format: `<adapterID>-<first-12-hex-chars-of-sha256>`. 48 bits of entropy is enough for ~17M rows per adapter
- * before the expected collision count exceeds 1 (birthday paradox); adapters with more rows should extend the prefix
- * length.
- */
-export function stableSourceID(adapterID, components) {
-    return stableSourceIDFromParts(adapterID, components);
-}
-/**
- * {@link stableSourceID} over arbitrary disambiguator keys — a variant index, a slot number, anything that is not a
- * `ComponentTag`. Every key handed in is sorted and hashed either way; only the key vocabulary differs, and the narrow
- * signature above is what stops an adapter hashing a misspelled component name.
- */
-export function stableSourceIDFromParts(adapterID, parts) {
-    const sortedKeys = Object.keys(parts).toSorted();
-    const payload = sortedKeys.map((k) => `${k}=${parts[k] ?? ""}`).join("\u001F");
-    // One `update` over the joined string, not three chained ones — identical byte stream either way,
-    // but the separator has to stay inline or every id in every slice changes.
-    const digest = sha256Hex(`${adapterID}\u001E${payload}`);
-    return `${adapterID}-${digest.slice(0, 12)}`;
-}
-/**
- * Leading house number on a US-style street line: digits, an optional hyphenated range (Queens `40-12`), an optional
- * single alpha suffix (`101A`), then whitespace, then the rest.
- *
- * This decides where the house number ends and the street begins for every US CSV adapter, so an edge case fixed here
- * is fixed for all of them.
- *
- * The remainder must stay `(\S.*)` and must NOT become `(.+)`: `\s+` and `.` both match a tab, so `\s+(.+)$` lets the
- * engine split a run of tabs between the two groups every possible way before failing — quadratic backtracking on
- * attacker-shaped input. Requiring a non-space start removes the overlap. Group 2 is trimmed by the caller either way,
- * so the two forms are indistinguishable on real input; only the failure cost differs.
- */
-export const HOUSE_NUMBER_PREFIX = /^(\d+(?:-\d+)?[A-Za-z]?)\s+(\S.*)$/;
-/**
- * Split a US-style street line on {@link HOUSE_NUMBER_PREFIX}.
- *
- * The US CSV sources follow USPS Publication 28 conventions with hand-entry drift. The leading digit run is the house
- * number (`"123 Main St"`, `"6450 W Indian School Rd"`), and the regex tolerates one trailing letter (`"123A Main St"`)
- * plus an optional hyphenated half (`"40-12 Bell Blvd"`, common in NYC and suburban garden-apartment numbering; Hawaii
- * uses it island-wide — `"47-470 Hui Aeko Place"`).
- *
- * Returns `null` for blank input. Anything that does not match the prefix shape (`"PO Box 1234"`, `"RR 2 Box 67"`, `"HC
- * 1"`) becomes a single `street` value rather than being mangled — the model sees the original surface form and
- * downstream classifiers pick it up. Callers that need those forms recognized as something other than a street (see
- * `usgov-irs-bmf`) test for them BEFORE calling this.
- */
-export function splitStreetLine(line) {
-    const trimmed = line.trim();
-    if (!trimmed)
-        return null;
-    const match = HOUSE_NUMBER_PREFIX.exec(trimmed);
-    if (match)
-        return { house_number: match[1], street: match[2].trim() };
-    return { street: trimmed };
-}
-/**
- * Compose the envelope-style raw address line the US CSV adapters share:
- *
- * `"<venue>, <house> <street> <unit>, <city>, <state> <postcode>"`
- *
- * The venue leads (US conventional addressee-then-address ordering) so a downstream model sees the
- * venue-prefix-then-address shape users actually type into geocoders. Empty segments drop out rather than leaving a
- * dangling separator.
- */
-export function composeRaw(parts) {
-    const streetLine = parts.streetLine ?? [parts.houseNumber, parts.street, parts.unit].filter(isPresent).join(" ").trim();
-    const tail = [parts.locality.trim(), [parts.region, parts.postcode].filter(isPresent).join(" ").trim()]
-        .filter(isPresent)
-        .join(", ");
-    return [parts.venue, streetLine || undefined, tail].filter(isPresent).join(", ");
-}
-/**
- * Load one curated libpostal dictionary (`core/data/libpostal/dictionaries/<language>/<filename>`) as a lower-cased
- * form set. libpostal format: `canonical|abbr|abbr|...` — every form is indexed.
- *
- * `resourceDictionaryPath` already resolves both layouts — `core/data/...` from source and from the packaged `out/`
- * tree. The candidate list this replaced named it TWICE and then guessed a third path off `process.cwd()`, and
- * swallowed every error while probing, so a corrupt dictionary reported as a missing one.
- *
- * The largest libpostal dictionary is 8.4 KB, and each caller runs this once per process at module load.
- */
-export async function loadLibpostalDictionary(language, filename) {
-    const text = await readLocalTextFile(resourceDictionaryPath("libpostal", language, filename));
-    const set = new Set();
-    for (const line of TextSpliterator.from(text)) {
-        const trimmed = line.trim();
-        if (!trimmed || trimmed.startsWith("#"))
-            continue;
-        for (const form of trimmed.split("|")) {
-            const f = form.trim().toLowerCase();
-            if (f) {
-                set.add(f);
-            }
-        }
-    }
-    return set;
-}
-/**
- * Canonical dedup key for a row.
- *
- * Two rows that share this key are treated as duplicates and only the first wins. The key is built from `country`, the
- * sorted `components` dict, and a normalized `raw` (lower-cased, whitespace collapsed). License and provenance fields
- * are intentionally excluded so the same address from multiple adapters is recognized as a duplicate.
- *
- * Synthetic rows are never deduplicated against natural rows: `synth.method` is folded into the key when present,
- * ensuring each augmentation variant survives.
- */
-export function canonicalDedupKey(row) {
-    const sortedKeys = Object.keys(row.components).toSorted();
-    const compPart = sortedKeys.map((k) => `${k}=${row.components[k] ?? ""}`).join("\u001F");
-    const rawNorm = row.raw.toLowerCase().replaceAll(/\s+/g, " ").trim();
-    const synthPart = row.synth ? `\u001E${row.synth.method}` : "";
-    return `${row.country}\u001E${rawNorm}\u001E${compPart}${synthPart}`;
-}
-/**
- * Default `StreamingHasher` (SHA-256, hex).
- */
-/**
- * @see {@link StreamingHasher} — this is the incremental counterpart to `sha256Hex`, not a copy of
- *   it. `@mailwoman/core/utils` exposes one-shot digests (`sha256Hex`) and whole-file digests
- *   (`sha256File`); the runner needs neither, because it hashes JSONL lines as they stream past.
- */
-export function streamingSha256() {
-    const h = createHash("sha256");
-    let finalized = false;
-    let digestHex = "";
-    return {
-        update(chunk) {
-            if (finalized)
-                throw new Error("streamingSha256: update() called after digest()");
-            h.update(typeof chunk === "string" ? chunk : chunk);
-        },
-        digest() {
-            if (!finalized) {
-                digestHex = h.digest("hex");
-                finalized = true;
-            }
-            return digestHex;
-        },
-    };
-}
-//# sourceMappingURL=index.js.map

package/out/adapters/utils/index.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../../../lib/adapters/utils/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;GAwBG;AAGH,OAAO,EAAE,iBAAiB,EAAE,MAAM,4BAA4B,CAAA;AAC9D,OAAO,EAAE,SAAS,EAAE,UAAU,EAAa,MAAM,sBAAsB,CAAA;AACvE,OAAO,EAAE,SAAS,EAAE,MAAM,yBAAyB,CAAA;AACnD,OAAO,EAAE,sBAAsB,EAAE,MAAM,uBAAuB,CAAA;AAC9D,OAAO,EAAE,eAAe,EAAE,MAAM,aAAa,CAAA;AAiC7C;;;GAGG;AACH,MAAM,OAAO,uBAAuB;IACnC,KAAK,GAAG,IAAI,GAAG,EAAyB,CAAA;IAExC,QAAQ,CAAC,OAAsB;QAC9B,IAAI,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,OAAO,CAAC,EAAE,CAAC,EAAE,CAAC;YAChC,MAAM,IAAI,KAAK,CAAC,uBAAuB,IAAI,CAAC,SAAS,CAAC,OAAO,CAAC,EAAE,CAAC,qBAAqB,CAAC,CAAA;QACxF,CAAC;QAED,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,OAAO,CAAC,EAAE,EAAE,OAAO,CAAC,CAAA;IACpC,CAAC;IAED,GAAG,CAAC,EAAU;QACb,OAAO,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,CAAC,CAAA;IAC1B,CAAC;IAED,IAAI;QACH,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC,CAAA;IACvC,CAAC;IAED,GAAG;QACF,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE,CAAC,CAAA;IACrC,CAAC;CACD;AAED;;;GAGG;AACH,MAAM,CAAC,MAAM,sBAAsB,GAAG,IAAI,uBAAuB,EAAE,CAAA;AAEnE;;;;;;;;;;GAUG;AACH,MAAM,UAAU,cAAc,CAAC,SAAiB,EAAE,UAAiD;IAClG,OAAO,uBAAuB,CAAC,SAAS,EAAE,UAAU,CAAC,CAAA;AACtD,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,uBAAuB,CACtC,SAAiB,EACjB,KAAmD;IAEnD,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,QAAQ,EAAE,CAAA;IAChD,MAAM,OAAO,GAAG,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;IAC9E,kGAAkG;IAClG,2EAA2E;IAC3E,MAAM,MAAM,GAAG,SAAS,CAAC,GAAG,SAAS,SAAS,OAAO,EAAE,CAAC,CAAA;IAExD,OAAO,GAAG,SAAS,IAAI,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAA;AAC7C,CAAC;AAED;;;;;;;;;;;GAWG;AACH,MAAM,CAAC,MAAM,mBAAmB,GAAG,oCAAoC,CAAA;AAUvE;;;;;;;;;;;;GAYG;AACH,MAAM,UAAU,eAAe,CAAC,IAAY;IAC3C,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAA;IAE3B,IAAI,CAAC,OAAO;QAAE,OAAO,IAAI,CAAA;IAEzB,MAAM,KAAK,GAAG,mBAAmB,CAAC,IAAI,CAAC,OAAO,CAAC,CAAA;IAE/C,IAAI,KAAK;QAAE,OAAO,EAAE,YAAY,EAAE,KAAK,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,KAAK,CAAC,CAAC,CAAE,CAAC,IAAI,EAAE,EAAE,CAAA;IAEtE,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,CAAA;AAC3B,CAAC;AAqBD;;;;;;;;GAQG;AACH,MAAM,UAAU,UAAU,CAAC,KAAsB;IAChD,MAAM,UAAU,GACf,KAAK,CAAC,UAAU,IAAI,CAAC,KAAK,CAAC,WAAW,EAAE,KAAK,CAAC,MAAM,EAAE,KAAK,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;IAErG,MAAM,IAAI,GAAG,CAAC,KAAK,CAAC,QAAQ,CAAC,IAAI,EAAE,EAAE,CAAC,KAAK,CAAC,MAAM,EAAE,KAAK,CAAC,QAAQ,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;SACrG,MAAM,CAAC,SAAS,CAAC;SACjB,IAAI,CAAC,IAAI,CAAC,CAAA;IAEZ,OAAO,CAAC,KAAK,CAAC,KAAK,EAAE,UAAU,IAAI,SAAS,EAAE,IAAI,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;AACjF,CAAC;AAED;;;;;;;;;GASG;AACH,MAAM,CAAC,KAAK,UAAU,uBAAuB,CAAC,QAAgB,EAAE,QAAgB;IAC/E,MAAM,IAAI,GAAG,MAAM,iBAAiB,CAAC,sBAAsB,CAAC,WAAW,EAAE,QAAQ,EAAE,QAAQ,CAAC,CAAC,CAAA;IAC7F,MAAM,GAAG,GAAG,IAAI,GAAG,EAAU,CAAA;IAE7B,KAAK,MAAM,IAAI,IAAI,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;QAC/C,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,CAAA;QAE3B,IAAI,CAAC,OAAO,IAAI,OAAO,CAAC,UAAU,CAAC,GAAG,CAAC;YAAE,SAAQ;QAEjD,KAAK,MAAM,IAAI,IAAI,OAAO,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,CAAC;YACvC,MAAM,CAAC,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAA;YAEnC,IAAI,CAAC,EAAE,CAAC;gBACP,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAA;YACX,CAAC;QACF,CAAC;IACF,CAAC;IAED,OAAO,GAAG,CAAA;AACX,CAAC;AAED;;;;;;;;;GASG;AACH,MAAM,UAAU,iBAAiB,CAAC,GAAiB;IAClD,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,QAAQ,EAAoB,CAAA;IAC3E,MAAM,QAAQ,GAAG,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,IAAI,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;IACxF,MAAM,OAAO,GAAG,GAAG,CAAC,GAAG,CAAC,WAAW,EAAE,CAAC,UAAU,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;IACpE,MAAM,SAAS,GAAG,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,SAAS,GAAG,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAA;IAE9D,OAAO,GAAG,GAAG,CAAC,OAAO,SAAS,OAAO,SAAS,QAAQ,GAAG,SAAS,EAAE,CAAA;AACrE,CAAC;AAcD;;GAEG;AACH;;;;GAIG;AACH,MAAM,UAAU,eAAe;IAC9B,MAAM,CAAC,GAAS,UAAU,CAAC,QAAQ,CAAC,CAAA;IACpC,IAAI,SAAS,GAAG,KAAK,CAAA;IACrB,IAAI,SAAS,GAAG,EAAE,CAAA;IAElB,OAAO;QACN,MAAM,CAAC,KAAK;YACX,IAAI,SAAS;gBAAE,MAAM,IAAI,KAAK,CAAC,iDAAiD,CAAC,CAAA;YACjF,CAAC,CAAC,MAAM,CAAC,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,CAAA;QACpD,CAAC;QACD,MAAM;YACL,IAAI,CAAC,SAAS,EAAE,CAAC;gBAChB,SAAS,GAAG,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAA;gBAC3B,SAAS,GAAG,IAAI,CAAA;YACjB,CAAC;YAED,OAAO,SAAS,CAAA;QACjB,CAAC;KACD,CAAA;AACF,CAAC"}

package/out/fr/adapters/ban/street-decompose.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"street-decompose.d.ts","sourceRoot":"","sources":["../../../../lib/fr/adapters/ban/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AAWH,MAAM,WAAW,kBAAkB;IAClC,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;IACrB,MAAM,EAAE,MAAM,CAAA;CACd;AAED;;;;;GAKG;AACH,wBAAgB,iBAAiB,CAAC,QAAQ,EAAE,MAAM,GAAG,kBAAkB,CA4BtE"}

package/out/fr/adapters/ban/street-decompose.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"street-decompose.js","sourceRoot":"","sources":["../../../../lib/fr/adapters/ban/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AAEH,OAAO,EAAE,uBAAuB,EAAE,MAAM,iBAAiB,CAAA;AAEzD;;GAEG;AACH,MAAM,wBAAwB,GAAG,CAAC,CAAA;AAElC,MAAM,eAAe,GAAG,MAAM,uBAAuB,CAAC,IAAI,EAAE,kBAAkB,CAAC,CAAA;AAO/E;;;;;GAKG;AACH,MAAM,UAAU,iBAAiB,CAAC,QAAgB;IACjD,MAAM,OAAO,GAAG,QAAQ,CAAC,IAAI,EAAE,CAAA;IAE/B,IAAI,CAAC,OAAO;QAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,EAAE,EAAE,CAAA;IAEjD,MAAM,MAAM,GAAG,OAAO,CAAC,KAAK,CAAC,KAAK,CAAC,CAAA;IAEnC,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,CAAA;IAE/D,MAAM,IAAI,GAAG,CAAC,CAAS,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAA;IAEjE,iDAAiD;IACjD,IAAI,MAAM,CAAC,MAAM,IAAI,wBAAwB,EAAE,CAAC;QAC/C,MAAM,OAAO,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAE,CAAC,GAAG,GAAG,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAE,CAAC,CAAA;QAEzD,IAAI,eAAe,CAAC,GAAG,CAAC,OAAO,CAAC,EAAE,CAAC;YAClC,OAAO,EAAE,MAAM,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAA;QACnF,CAAC;IACF,CAAC;IAED,yBAAyB;IACzB,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAE,CAAC,CAAA;IAE9B,IAAI,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC;QAChC,OAAO,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAE,EAAE,MAAM,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAA;IACjE,CAAC;IAED,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,CAAA;AACzC,CAAC"}

package/out/no/recipes/street/index.d.ts

@@ -1,18 +0,0 @@
-/**
- * @copyright Sister Software
- * @license AGPL-3.0
- * @author Teffen Ellis, et al.
- *
- *   `no-street` slice recipe — synthetic NO-street counter-example rows: tuples →
- *   {@link synthesizeNoStreetRow} → aligned LabeledRow. The corpus-side counterweight to the
- *   synth-street slice that drove v0.6.1's `dependent_locality` regression — venue+admin and
- *   admin-only rows with explicit absence of any street-side component. Ported from
- *   scripts/build-no-street-slice.mjs.
- */
-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 noStreetRecipe: CorpusRecipe;
-//# sourceMappingURL=index.d.ts.map

package/out/no/recipes/street/index.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../../lib/no/recipes/street/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAIH,OAAO,EAKN,KAAK,YAAY,EACjB,MAAM,mBAAmB,CAAA;AAG1B;;;GAGG;AACH,eAAO,MAAM,cAAc,EAAE,YA+D5B,CAAA"}

package/out/no/recipes/street/index.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../../../../lib/no/recipes/street/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAEH,OAAO,EAAE,OAAO,EAAE,MAAM,uBAAuB,CAAA;AAE/C,OAAO,EACN,aAAa,EACb,UAAU,EACV,aAAa,EACb,uBAAuB,IAAI,OAAO,GAElC,MAAM,mBAAmB,CAAA;AAC1B,OAAO,EAAE,qBAAqB,EAA0B,MAAM,yBAAyB,CAAA;AAEvF;;;GAGG;AACH,MAAM,CAAC,MAAM,cAAc,GAAiB;IAC3C,IAAI,EAAE,WAAW;IACjB,WAAW,EAAE,qFAAqF;IAClG,IAAI,EAAE,QAAQ;IACd,KAAK,CAAC,GAAG,CAAC,IAAI,EAAE,KAAK;QACpB,IAAI,CAAC,IAAI,CAAC,KAAK;YAAE,MAAM,IAAI,KAAK,CAAC,kDAAkD,CAAC,CAAA;QACpF,wFAAwF;QACxF,MAAM,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;QACjC,MAAM,MAAM,GAAG,IAAI,CAAC,UAAU,IAAI,iBAAiB,CAAA;QACnD,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,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,MAAM,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,EAAE,CAAC;gBAC3E,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,qBAAqB,CAAC,KAA0B,EAAE,EAAE,MAAM,EAAE,CAAC,CAAA;gBAE3E,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"}