@mailwoman/corpus 9.2.0 → 9.4.0

package/out/adapters/tiger/adapter.js

@@ -0,0 +1,201 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   `tiger`: US Census TIGER/Line consumer adapter.
+ *
+ *   TIGER/Line is the canonical US street + locality dataset published by the Census Bureau as a
+ *   **public-domain** product (no ODbL share-alike concerns for US-only corpora). Coverage extends
+ *   to every named street segment + every incorporated place + CDP across the 50 states + DC + the
+ *   five primary territories — substantially better US street-name coverage than OSM, especially in
+ *   rural areas.
+ *
+ *   Following the `wof-admin` / `wof-postalcode` pattern, this adapter consumes a SQLite database the
+ *   operator pre-builds from the raw TIGER shapefiles (see the README for the schema and a
+ *   suggested `ogr2ogr` pipeline). The mailwoman side does not parse Shapefile binary directly —
+ *   keeping the adapter narrow lets the operator pick their own ingestion tool (ogr2ogr / shp2pgsql
+ *   / a custom Python script / etc.) without forcing a heavy native dep into `@mailwoman/corpus`.
+ *
+ *   Two row classes are emitted:
+ *
+ *   - **Street-level** (`tiger_streets`): one row per segment, optionally with up to two postcode
+ *       variants if `zipl` / `zipr` differ. Components: `{ street, region, postcode? }`. Streets
+ *       without a recognized state FIPS are dropped — there's no useful row without `region`.
+ *   - **Locality-level** (`tiger_places`): up to three variants per place: locality-only,
+ *       locality-with-region, locality-with-region-country (mirrors `wof-admin`'s fan-out for
+ *       consistency).
+ *   - `packages/corpus/lib/codex/us-fips-state.ts` — the FIPS → `{abbreviation, name}` lookup table
+ *       (originally `tiger/state.ts`, AGPL-3.0 → AGPL-3.0). The full isp-nexus TIGER module ships a
+ *       TypeORM-backed service layer; mailwoman only needs the lookup data so we don't carry the
+ *       service layer over.
+ *
+ *   License: stamped `"Public Domain"` per Census Bureau guidance on TIGER/Line. No per-row override
+ *   needed — every row in TIGER is the same license.
+ */
+import { formatAddress, reconcileComponents } from "@mailwoman/formatter";
+import { DatabaseClient } from "@mailwoman/sqlite/client";
+import { decomposeStreet } from "#adapters/tiger/street-decompose";
+import { lookupFipsState } from "#codex/us-fips-state";
+/**
+ * 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 const TIGER_ADAPTER_ID = "tiger";
+/**
+ * License carried by this source (Public Domain), attached to each row so downstream consumers inherit the terms rather
+ * than having to look them up.
+ */
+export const TIGER_DEFAULT_LICENSE = "Public Domain";
+/**
+ * The country surface form used in `formatAddress` for US. Matches the canonical OpenCage US template output so
+ * reconciliation doesn't strip it when the row carries `country` explicitly.
+ */
+const US_COUNTRY_DISPLAY = "United States of America";
+/**
+ * Yield one or more `CanonicalRow`s per street segment. Postcode variants:
+ *
+ * - No ZIP set → one row, street + region.
+ * - `zipl === zipr` → one row, street + region + postcode.
+ * - `zipl !== zipr` → two rows (one per side's ZIP).
+ */
+function* streetVariants(row) {
+    const fullname = row.fullname.trim();
+    if (!fullname)
+        return;
+    const state = lookupFipsState(row.statefp);
+    if (!state)
+        return;
+    const zipl = row.zipl?.trim() ?? "";
+    const zipr = row.zipr?.trim() ?? "";
+    const decomposed = decomposeStreet(fullname);
+    const baseComponents = {
+        region: state.abbreviation,
+        street: decomposed.street,
+    };
+    if (decomposed.prefix) {
+        baseComponents.street_prefix = decomposed.prefix;
+    }
+    if (decomposed.suffix) {
+        baseComponents.street_suffix = decomposed.suffix;
+    }
+    if (!zipl && !zipr) {
+        yield { components: baseComponents, variantKey: "no-zip" };
+        return;
+    }
+    if (zipl && zipr && zipl === zipr) {
+        yield {
+            components: { ...baseComponents, postcode: zipl },
+            variantKey: `zip-${zipl}`,
+        };
+        return;
+    }
+    if (zipl) {
+        yield { components: { ...baseComponents, postcode: zipl }, variantKey: `zipl-${zipl}` };
+    }
+    if (zipr && zipr !== zipl) {
+        yield { components: { ...baseComponents, postcode: zipr }, variantKey: `zipr-${zipr}` };
+    }
+}
+/**
+ * Three locality-level variants, mirroring `wof-admin`'s fan-out.
+ */
+function* placeVariants(row) {
+    const name = row.name.trim();
+    if (!name)
+        return;
+    const state = lookupFipsState(row.statefp);
+    if (!state)
+        return;
+    yield {
+        components: { locality: name },
+        variantKey: "locality-only",
+    };
+    yield {
+        components: { locality: name, region: state.abbreviation },
+        variantKey: "with-region",
+    };
+    yield {
+        components: { locality: name, region: state.abbreviation, country: US_COUNTRY_DISPLAY },
+        variantKey: "with-region-country",
+    };
+}
+/**
+ * Build a TIGER adapter. Pure factory so multiple instances can be created in tests.
+ */
+export function createTigerAdapter() {
+    return {
+        id: TIGER_ADAPTER_ID,
+        defaultLicense: TIGER_DEFAULT_LICENSE,
+        description: "US Census TIGER/Line streets + places consumer (public-domain); SQLite DB built via `mailwoman tiger fetch`.",
+        async *rows(opts) {
+            if (opts.country && opts.country !== "US") {
+                throw new Error(`tiger adapter: only US supported, got country=${opts.country}`);
+            }
+            const db = new DatabaseClient(opts.inputPath, { readOnly: true });
+            let emitted = 0;
+            try {
+                const streetStmt = db.prepare(`SELECT linearid, fullname, zipl, zipr, statefp FROM tiger_streets`);
+                const placeStmt = db.prepare(`SELECT geoid, name, statefp, lsad FROM tiger_places`);
+                for (const row of streetStmt.iterate()) {
+                    if (opts.signal?.aborted)
+                        return;
+                    for (const variant of streetVariants(row)) {
+                        if (opts.limit !== undefined && emitted >= opts.limit)
+                            return;
+                        const raw = formatAddress(variant.components, "US", { separator: ", " });
+                        if (!raw)
+                            continue;
+                        const aligned = reconcileComponents(variant.components, raw);
+                        if (!Object.keys(aligned).length)
+                            continue;
+                        yield {
+                            raw,
+                            components: aligned,
+                            country: "US",
+                            locale: "en-US",
+                            source: TIGER_ADAPTER_ID,
+                            source_id: `${TIGER_ADAPTER_ID}-st-${row.linearid}-${variant.variantKey}`,
+                            corpus_version: "",
+                            license: TIGER_DEFAULT_LICENSE,
+                        };
+                        emitted++;
+                    }
+                }
+                for (const row of placeStmt.iterate()) {
+                    if (opts.signal?.aborted)
+                        return;
+                    for (const variant of placeVariants(row)) {
+                        if (opts.limit !== undefined && emitted >= opts.limit)
+                            return;
+                        const raw = formatAddress(variant.components, "US", { separator: ", " });
+                        if (!raw)
+                            continue;
+                        const aligned = reconcileComponents(variant.components, raw);
+                        if (!Object.keys(aligned).length)
+                            continue;
+                        yield {
+                            raw,
+                            components: aligned,
+                            country: "US",
+                            locale: "en-US",
+                            source: TIGER_ADAPTER_ID,
+                            source_id: `${TIGER_ADAPTER_ID}-pl-${row.geoid}-${variant.variantKey}`,
+                            corpus_version: "",
+                            license: TIGER_DEFAULT_LICENSE,
+                        };
+                        emitted++;
+                    }
+                }
+            }
+            finally {
+                await db.destroy();
+            }
+        },
+    };
+}
+/**
+ * The configured adapter instance registered with the corpus builder.
+ */
+export const tigerAdapter = createTigerAdapter();
+//# sourceMappingURL=adapter.js.map

package/out/adapters/tiger/adapter.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../lib/adapters/tiger/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GAkCG;AAEH,OAAO,EAAE,aAAa,EAAE,mBAAmB,EAAE,MAAM,sBAAsB,CAAA;AACzE,OAAO,EAAE,cAAc,EAAE,MAAM,0BAA0B,CAAA;AAGzD,OAAO,EAAE,eAAe,EAAE,MAAM,kCAAkC,CAAA;AAClE,OAAO,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAA;AAGtD;;;GAGG;AACH,MAAM,CAAC,MAAM,gBAAgB,GAAG,OAAO,CAAA;AACvC;;;GAGG;AACH,MAAM,CAAC,MAAM,qBAAqB,GAAG,eAAe,CAAA;AAEpD;;;GAGG;AACH,MAAM,kBAAkB,GAAG,0BAA0B,CAAA;AAiBrD;;;;;;GAMG;AACH,QAAQ,CAAC,CAAC,cAAc,CAAC,GAAmB;IAI3C,MAAM,QAAQ,GAAG,GAAG,CAAC,QAAQ,CAAC,IAAI,EAAE,CAAA;IAEpC,IAAI,CAAC,QAAQ;QAAE,OAAM;IACrB,MAAM,KAAK,GAAG,eAAe,CAAC,GAAG,CAAC,OAAO,CAAC,CAAA;IAE1C,IAAI,CAAC,KAAK;QAAE,OAAM;IAElB,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;IACnC,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;IAEnC,MAAM,UAAU,GAAG,eAAe,CAAC,QAAQ,CAAC,CAAA;IAE5C,MAAM,cAAc,GAA+B;QAClD,MAAM,EAAE,KAAK,CAAC,YAAY;QAC1B,MAAM,EAAE,UAAU,CAAC,MAAM;KACzB,CAAA;IAED,IAAI,UAAU,CAAC,MAAM,EAAE,CAAC;QACvB,cAAc,CAAC,aAAa,GAAG,UAAU,CAAC,MAAM,CAAA;IACjD,CAAC;IAED,IAAI,UAAU,CAAC,MAAM,EAAE,CAAC;QACvB,cAAc,CAAC,aAAa,GAAG,UAAU,CAAC,MAAM,CAAA;IACjD,CAAC;IAED,IAAI,CAAC,IAAI,IAAI,CAAC,IAAI,EAAE,CAAC;QACpB,MAAM,EAAE,UAAU,EAAE,cAAc,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAA;QAE1D,OAAM;IACP,CAAC;IAED,IAAI,IAAI,IAAI,IAAI,IAAI,IAAI,KAAK,IAAI,EAAE,CAAC;QACnC,MAAM;YACL,UAAU,EAAE,EAAE,GAAG,cAAc,EAAE,QAAQ,EAAE,IAAI,EAAE;YACjD,UAAU,EAAE,OAAO,IAAI,EAAE;SACzB,CAAA;QAED,OAAM;IACP,CAAC;IAED,IAAI,IAAI,EAAE,CAAC;QACV,MAAM,EAAE,UAAU,EAAE,EAAE,GAAG,cAAc,EAAE,QAAQ,EAAE,IAAI,EAAE,EAAE,UAAU,EAAE,QAAQ,IAAI,EAAE,EAAE,CAAA;IACxF,CAAC;IAED,IAAI,IAAI,IAAI,IAAI,KAAK,IAAI,EAAE,CAAC;QAC3B,MAAM,EAAE,UAAU,EAAE,EAAE,GAAG,cAAc,EAAE,QAAQ,EAAE,IAAI,EAAE,EAAE,UAAU,EAAE,QAAQ,IAAI,EAAE,EAAE,CAAA;IACxF,CAAC;AACF,CAAC;AAED;;GAEG;AACH,QAAQ,CAAC,CAAC,aAAa,CAAC,GAAkB;IAIzC,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,CAAC,IAAI,EAAE,CAAA;IAE5B,IAAI,CAAC,IAAI;QAAE,OAAM;IACjB,MAAM,KAAK,GAAG,eAAe,CAAC,GAAG,CAAC,OAAO,CAAC,CAAA;IAE1C,IAAI,CAAC,KAAK;QAAE,OAAM;IAElB,MAAM;QACL,UAAU,EAAE,EAAE,QAAQ,EAAE,IAAI,EAAE;QAC9B,UAAU,EAAE,eAAe;KAC3B,CAAA;IAED,MAAM;QACL,UAAU,EAAE,EAAE,QAAQ,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,CAAC,YAAY,EAAE;QAC1D,UAAU,EAAE,aAAa;KACzB,CAAA;IAED,MAAM;QACL,UAAU,EAAE,EAAE,QAAQ,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,CAAC,YAAY,EAAE,OAAO,EAAE,kBAAkB,EAAE;QACvF,UAAU,EAAE,qBAAqB;KACjC,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,kBAAkB;IACjC,OAAO;QACN,EAAE,EAAE,gBAAgB;QACpB,cAAc,EAAE,qBAAqB;QACrC,WAAW,EACV,8GAA8G;QAE/G,KAAK,CAAC,CAAC,IAAI,CAAC,IAAoB;YAC/B,IAAI,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,OAAO,KAAK,IAAI,EAAE,CAAC;gBAC3C,MAAM,IAAI,KAAK,CAAC,iDAAiD,IAAI,CAAC,OAAO,EAAE,CAAC,CAAA;YACjF,CAAC;YAED,MAAM,EAAE,GAAG,IAAI,cAAc,CAAgB,IAAI,CAAC,SAAS,EAAE,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAA;YAChF,IAAI,OAAO,GAAG,CAAC,CAAA;YAEf,IAAI,CAAC;gBACJ,MAAM,UAAU,GAAG,EAAE,CAAC,OAAO,CAAC,mEAAmE,CAAC,CAAA;gBAClG,MAAM,SAAS,GAAG,EAAE,CAAC,OAAO,CAAC,qDAAqD,CAAC,CAAA;gBAEnF,KAAK,MAAM,GAAG,IAAI,UAAU,CAAC,OAAO,EAAsC,EAAE,CAAC;oBAC5E,IAAI,IAAI,CAAC,MAAM,EAAE,OAAO;wBAAE,OAAM;oBAEhC,KAAK,MAAM,OAAO,IAAI,cAAc,CAAC,GAAG,CAAC,EAAE,CAAC;wBAC3C,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,IAAI,CAAC,KAAK;4BAAE,OAAM;wBAC7D,MAAM,GAAG,GAAG,aAAa,CAAC,OAAO,CAAC,UAAU,EAAE,IAAI,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAA;wBAExE,IAAI,CAAC,GAAG;4BAAE,SAAQ;wBAClB,MAAM,OAAO,GAAG,mBAAmB,CAAC,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC,CAAA;wBAE5D,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,MAAM;4BAAE,SAAQ;wBAE1C,MAAM;4BACL,GAAG;4BACH,UAAU,EAAE,OAAO;4BACnB,OAAO,EAAE,IAAI;4BACb,MAAM,EAAE,OAAO;4BACf,MAAM,EAAE,gBAAgB;4BACxB,SAAS,EAAE,GAAG,gBAAgB,OAAO,GAAG,CAAC,QAAQ,IAAI,OAAO,CAAC,UAAU,EAAE;4BACzE,cAAc,EAAE,EAAE;4BAClB,OAAO,EAAE,qBAAqB;yBAC9B,CAAA;wBAED,OAAO,EAAE,CAAA;oBACV,CAAC;gBACF,CAAC;gBAED,KAAK,MAAM,GAAG,IAAI,SAAS,CAAC,OAAO,EAAqC,EAAE,CAAC;oBAC1E,IAAI,IAAI,CAAC,MAAM,EAAE,OAAO;wBAAE,OAAM;oBAEhC,KAAK,MAAM,OAAO,IAAI,aAAa,CAAC,GAAG,CAAC,EAAE,CAAC;wBAC1C,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,IAAI,CAAC,KAAK;4BAAE,OAAM;wBAC7D,MAAM,GAAG,GAAG,aAAa,CAAC,OAAO,CAAC,UAAU,EAAE,IAAI,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAA;wBAExE,IAAI,CAAC,GAAG;4BAAE,SAAQ;wBAClB,MAAM,OAAO,GAAG,mBAAmB,CAAC,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC,CAAA;wBAE5D,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,MAAM;4BAAE,SAAQ;wBAE1C,MAAM;4BACL,GAAG;4BACH,UAAU,EAAE,OAAO;4BACnB,OAAO,EAAE,IAAI;4BACb,MAAM,EAAE,OAAO;4BACf,MAAM,EAAE,gBAAgB;4BACxB,SAAS,EAAE,GAAG,gBAAgB,OAAO,GAAG,CAAC,KAAK,IAAI,OAAO,CAAC,UAAU,EAAE;4BACtE,cAAc,EAAE,EAAE;4BAClB,OAAO,EAAE,qBAAqB;yBAC9B,CAAA;wBAED,OAAO,EAAE,CAAA;oBACV,CAAC;gBACF,CAAC;YACF,CAAC;oBAAS,CAAC;gBACV,MAAM,EAAE,CAAC,OAAO,EAAE,CAAA;YACnB,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,YAAY,GAAG,kBAAkB,EAAE,CAAA"}

package/out/adapters/tiger/street-decompose.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"street-decompose.d.ts","sourceRoot":"","sources":["../../../lib/adapters/tiger/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AASH,MAAM,WAAW,gBAAgB;IAChC,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;IACrB,MAAM,EAAE,MAAM,CAAA;IACd,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;CACrB;AAED;;;;;GAKG;AACH,wBAAgB,eAAe,CAAC,QAAQ,EAAE,MAAM,GAAG,gBAAgB,CA6ClE"}

package/out/adapters/tiger/street-decompose.js

@@ -0,0 +1,68 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   Decompose a US street name into Stage 3 components: street_prefix, street, street_suffix.
+ *
+ *   Sources directionals and street types from the curated libpostal/en dictionaries
+ *   (`core/data/libpostal/dictionaries/en/{directionals,street_types}.txt`). These are the same
+ *   dictionaries the runtime classifiers (StreetPrefixClassifier, StreetSuffixClassifier) use, so
+ *   corpus labels and runtime classifications agree on the vocabulary.
+ *
+ *   Examples: "N Main St" → { prefix: "N", street: "Main", suffix: "St" } "Pennsylvania Avenue NW" →
+ *   { prefix: null, street: "Pennsylvania", suffix: "Avenue NW" } "Salmon St" → { prefix: null,
+ *   street: "Salmon", suffix: "St" } "SE Hawthorne Blvd" → { prefix: "SE", street: "Hawthorne",
+ *   suffix: "Blvd" }
+ */
+import { loadLibpostalDictionary } from "#adapters/utils";
+const [DIRECTIONALS, STREET_TYPES] = await Promise.all([
+    loadLibpostalDictionary("en", "directionals.txt"),
+    loadLibpostalDictionary("en", "street_types.txt"),
+]);
+/**
+ * Decompose a US street name into prefix/name/suffix components.
+ *
+ * Conservative — only emits prefix/suffix when there's a clear directional or street-type keyword. Returns the original
+ * as `street` if nothing matches.
+ */
+export function decomposeStreet(fullname) {
+    const trimmed = fullname.trim();
+    if (!trimmed)
+        return { prefix: null, street: "", suffix: null };
+    const tokens = trimmed.split(/\s+/);
+    if (tokens.length === 1)
+        return { prefix: null, street: trimmed, suffix: null };
+    const norm = (s) => s.toLowerCase().replace(/\.$/, "");
+    let prefix = null;
+    let suffix = null;
+    let startIdx = 0;
+    let endIdx = tokens.length;
+    // Leading directional prefix
+    if (DIRECTIONALS.has(norm(tokens[0])) && tokens.length >= 2) {
+        prefix = tokens[0];
+        startIdx = 1;
+    }
+    // Trailing post-directional combined with street type (e.g. "Pennsylvania Ave NW")
+    const last = norm(tokens[endIdx - 1]);
+    const secondLast = endIdx >= 2 ? norm(tokens[endIdx - 2]) : "";
+    if (DIRECTIONALS.has(last) && STREET_TYPES.has(secondLast)) {
+        suffix = tokens.slice(endIdx - 2, endIdx).join(" ");
+        endIdx -= 2;
+    }
+    else if (STREET_TYPES.has(last) && endIdx - startIdx >= 2) {
+        suffix = tokens[endIdx - 1];
+        endIdx -= 1;
+    }
+    else if (DIRECTIONALS.has(last) && endIdx - startIdx >= 2) {
+        // Post-directional without type
+        suffix = tokens[endIdx - 1];
+        endIdx -= 1;
+    }
+    const street = tokens.slice(startIdx, endIdx).join(" ").trim();
+    if (!street) {
+        return { prefix: null, street: trimmed, suffix: null };
+    }
+    return { prefix, street, suffix };
+}
+//# sourceMappingURL=street-decompose.js.map

package/out/adapters/tiger/street-decompose.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"street-decompose.js","sourceRoot":"","sources":["../../../lib/adapters/tiger/street-decompose.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAEH,OAAO,EAAE,uBAAuB,EAAE,MAAM,iBAAiB,CAAA;AAEzD,MAAM,CAAC,YAAY,EAAE,YAAY,CAAC,GAAG,MAAM,OAAO,CAAC,GAAG,CAAC;IACtD,uBAAuB,CAAC,IAAI,EAAE,kBAAkB,CAAC;IACjD,uBAAuB,CAAC,IAAI,EAAE,kBAAkB,CAAC;CACjD,CAAC,CAAA;AAQF;;;;;GAKG;AACH,MAAM,UAAU,eAAe,CAAC,QAAgB;IAC/C,MAAM,OAAO,GAAG,QAAQ,CAAC,IAAI,EAAE,CAAA;IAE/B,IAAI,CAAC,OAAO;QAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,EAAE,IAAI,EAAE,CAAA;IAE/D,MAAM,MAAM,GAAG,OAAO,CAAC,KAAK,CAAC,KAAK,CAAC,CAAA;IAEnC,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,CAAA;IAE/E,MAAM,IAAI,GAAG,CAAC,CAAS,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAA;IAE9D,IAAI,MAAM,GAAkB,IAAI,CAAA;IAChC,IAAI,MAAM,GAAkB,IAAI,CAAA;IAChC,IAAI,QAAQ,GAAG,CAAC,CAAA;IAChB,IAAI,MAAM,GAAG,MAAM,CAAC,MAAM,CAAA;IAE1B,6BAA6B;IAC7B,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAE,CAAC,CAAC,IAAI,MAAM,CAAC,MAAM,IAAI,CAAC,EAAE,CAAC;QAC9D,MAAM,GAAG,MAAM,CAAC,CAAC,CAAE,CAAA;QACnB,QAAQ,GAAG,CAAC,CAAA;IACb,CAAC;IAED,mFAAmF;IACnF,MAAM,IAAI,GAAG,IAAI,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAC,CAAA;IACtC,MAAM,UAAU,GAAG,MAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAC,CAAC,CAAC,CAAC,EAAE,CAAA;IAE/D,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,YAAY,CAAC,GAAG,CAAC,UAAU,CAAC,EAAE,CAAC;QAC5D,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;QACnD,MAAM,IAAI,CAAC,CAAA;IACZ,CAAC;SAAM,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,MAAM,GAAG,QAAQ,IAAI,CAAC,EAAE,CAAC;QAC7D,MAAM,GAAG,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAA;QAC5B,MAAM,IAAI,CAAC,CAAA;IACZ,CAAC;SAAM,IAAI,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,MAAM,GAAG,QAAQ,IAAI,CAAC,EAAE,CAAC;QAC7D,gCAAgC;QAChC,MAAM,GAAG,MAAM,CAAC,MAAM,GAAG,CAAC,CAAE,CAAA;QAC5B,MAAM,IAAI,CAAC,CAAA;IACZ,CAAC;IAED,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;IAE9D,IAAI,CAAC,MAAM,EAAE,CAAC;QACb,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,CAAA;IACvD,CAAC;IAED,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,CAAA;AAClC,CAAC"}

package/out/adapters/usgov-hrsa-fqhc/adapter.d.ts

@@ -0,0 +1,56 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   `usgov-hrsa-fqhc`: HRSA "Health Center Service Delivery Site Locations" CSV consumer.
+ *
+ *   Federally Qualified Health Centers (FQHCs) are HRSA-funded community health programs that
+ *   self-report site addresses to the HRSA Data Warehouse. The published CSV (`data.hrsa.gov`)
+ *   carries the site name, the postal-formatted street address, and the locality/region/postcode
+ *   quad. The source earns its place on adversarial-value-per-row (#22): every facility name is a
+ *   human-typed venue string and the addresses pass through enough hands to accumulate the
+ *   abbreviation drift + suite designator chaos that pure gazetteer data does not.
+ *
+ *   The adapter consumes a CSV file the operator pre-downloads. The HRSA data is published as a
+ *   single national CSV (~10K rows), small enough that the operator can re-fetch on every corpus
+ *   rebuild without an intermediate SQLite step. Column names below match the HRSA Data Warehouse's
+ *   "Health Center Service Delivery Site" public dataset. Operators substituting the
+ *   closely-related "Site Address" or "Health Center" public extracts may need to remap columns;
+ *   the README documents the expected set.
+ *
+ *   Output: one row per CSV record, with `venue` component carrying the site name and the address
+ *   quad on `(house_number, street, locality, region, postcode)`. Component order is critical:
+ *   `venue` is inserted FIRST so alignment claims its surface span before `locality` searches for
+ *   its own (the kryptonite case "Buffalo Health Clinic, …, Buffalo, NY" relies on `venue`
+ *   consuming the first "Buffalo" so locality lands on the second).
+ *
+ *   License: stamped `"Public Domain"` per the HRSA Data Warehouse's federal government distribution
+ *   terms.
+ */
+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 USGOV_HRSA_FQHC_ADAPTER_ID = "usgov-hrsa-fqhc";
+/**
+ * License carried by this source (Public Domain), attached to each row so downstream consumers inherit the terms rather
+ * than having to look them up.
+ */
+export declare const USGOV_HRSA_FQHC_DEFAULT_LICENSE = "Public Domain";
+/**
+ * Split a "123 Main St Suite 4" surface form into `(house_number, street)`. The regex tolerates one trailing letter on
+ * the number (`"123A Main St"`) and a hyphenated form (`"40-12 Bell Blvd"`); anything else falls back to street-only.
+ *
+ * Suite / Apt / Unit designators stay on `street` — Mailwoman's `unit` component exists but the address-formatter does
+ * not have a clean slot for it, and HRSA addresses do not separate the suite into its own column. Leaving the surface
+ * form intact in `street` preserves the adversarial training signal (the model learns that a trailing "Suite 4" is part
+ * of the road line in this distribution).
+ */
+export declare function createUSGovHRSAFQHCAdapter(): CorpusAdapter;
+/**
+ * The configured adapter instance registered with the corpus builder.
+ */
+export declare const usgovHrsaFqhcAdapter: CorpusAdapter;
+//# sourceMappingURL=adapter.d.ts.map

package/out/adapters/usgov-hrsa-fqhc/adapter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../lib/adapters/usgov-hrsa-fqhc/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;AAOH,OAAO,KAAK,EAAgC,aAAa,EAAE,MAAM,QAAQ,CAAA;AAEzE;;;GAGG;AACH,eAAO,MAAM,0BAA0B,oBAAoB,CAAA;AAC3D;;;GAGG;AACH,eAAO,MAAM,+BAA+B,kBAAkB,CAAA;AAmB9D;;;;;;;;GAQG;AAEH,wBAAgB,0BAA0B,IAAI,aAAa,CAqF1D;AAED;;GAEG;AACH,eAAO,MAAM,oBAAoB,eAA+B,CAAA"}

package/out/adapters/usgov-hrsa-fqhc/adapter.js

@@ -0,0 +1,132 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   `usgov-hrsa-fqhc`: HRSA "Health Center Service Delivery Site Locations" CSV consumer.
+ *
+ *   Federally Qualified Health Centers (FQHCs) are HRSA-funded community health programs that
+ *   self-report site addresses to the HRSA Data Warehouse. The published CSV (`data.hrsa.gov`)
+ *   carries the site name, the postal-formatted street address, and the locality/region/postcode
+ *   quad. The source earns its place on adversarial-value-per-row (#22): every facility name is a
+ *   human-typed venue string and the addresses pass through enough hands to accumulate the
+ *   abbreviation drift + suite designator chaos that pure gazetteer data does not.
+ *
+ *   The adapter consumes a CSV file the operator pre-downloads. The HRSA data is published as a
+ *   single national CSV (~10K rows), small enough that the operator can re-fetch on every corpus
+ *   rebuild without an intermediate SQLite step. Column names below match the HRSA Data Warehouse's
+ *   "Health Center Service Delivery Site" public dataset. Operators substituting the
+ *   closely-related "Site Address" or "Health Center" public extracts may need to remap columns;
+ *   the README documents the expected set.
+ *
+ *   Output: one row per CSV record, with `venue` component carrying the site name and the address
+ *   quad on `(house_number, street, locality, region, postcode)`. Component order is critical:
+ *   `venue` is inserted FIRST so alignment claims its surface span before `locality` searches for
+ *   its own (the kryptonite case "Buffalo Health Clinic, …, Buffalo, NY" relies on `venue`
+ *   consuming the first "Buffalo" so locality lands on the second).
+ *
+ *   License: stamped `"Public Domain"` per the HRSA Data Warehouse's federal government distribution
+ *   terms.
+ */
+import { reconcileComponents } from "@mailwoman/formatter";
+import { CSVSpliterator } from "spliterator";
+import { composeRaw, splitStreetLine, stableSourceID } from "#adapters/utils";
+import { lookupStateAbbreviation } from "#codex/us-fips-state";
+/**
+ * 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 const USGOV_HRSA_FQHC_ADAPTER_ID = "usgov-hrsa-fqhc";
+/**
+ * License carried by this source (Public Domain), attached to each row so downstream consumers inherit the terms rather
+ * than having to look them up.
+ */
+export const USGOV_HRSA_FQHC_DEFAULT_LICENSE = "Public Domain";
+/**
+ * Split a "123 Main St Suite 4" surface form into `(house_number, street)`. The regex tolerates one trailing letter on
+ * the number (`"123A Main St"`) and a hyphenated form (`"40-12 Bell Blvd"`); anything else falls back to street-only.
+ *
+ * Suite / Apt / Unit designators stay on `street` — Mailwoman's `unit` component exists but the address-formatter does
+ * not have a clean slot for it, and HRSA addresses do not separate the suite into its own column. Leaving the surface
+ * form intact in `street` preserves the adversarial training signal (the model learns that a trailing "Suite 4" is part
+ * of the road line in this distribution).
+ */
+export function createUSGovHRSAFQHCAdapter() {
+    return {
+        id: USGOV_HRSA_FQHC_ADAPTER_ID,
+        defaultLicense: USGOV_HRSA_FQHC_DEFAULT_LICENSE,
+        description: "HRSA Federally Qualified Health Center site locations (public-domain). Adversarial source: venue + address co-occurrence, hand-entered.",
+        async *rows(opts) {
+            if (opts.country && opts.country !== "US") {
+                throw new Error(`usgov-hrsa-fqhc adapter: only US supported, got country=${opts.country}`);
+            }
+            const rows = CSVSpliterator.fromAsync(opts.inputPath, {
+                mode: "object",
+                normalizeKeys: false,
+                enableQuoteHandling: true,
+            });
+            let emitted = 0;
+            for await (const record of rows) {
+                if (opts.signal?.aborted)
+                    break;
+                if (opts.limit !== undefined && emitted >= opts.limit)
+                    break;
+                const venue = (record["Site Name"] ?? "").trim();
+                const split = splitStreetLine(record["Site Address"] ?? "");
+                const city = (record["Site City"] ?? "").trim();
+                const stateAbbr = (record["Site State Abbreviation"] ?? "").trim();
+                const postcode = (record["Site Postal Code"] ?? "").trim();
+                if (!venue || !split || !city || !postcode)
+                    continue;
+                const state = lookupStateAbbreviation(stateAbbr);
+                if (!state)
+                    continue;
+                // Insertion order matters here. `venue` first so alignment claims its span
+                // (which may contain a token like "Buffalo") before `locality` runs its
+                // search — the kryptonite case `Buffalo Health Clinic, Buffalo NY`
+                // otherwise mis-labels the venue's "Buffalo" as locality.
+                const components = {
+                    venue,
+                    ...(split.house_number ? { house_number: split.house_number } : {}),
+                    street: split.street,
+                    locality: city,
+                    region: state.abbreviation,
+                    postcode,
+                };
+                const raw = composeRaw({
+                    venue,
+                    houseNumber: split.house_number,
+                    street: split.street,
+                    locality: city,
+                    region: state.abbreviation,
+                    postcode,
+                });
+                if (!raw)
+                    continue;
+                const aligned = reconcileComponents(components, raw);
+                if (!Object.keys(aligned).length)
+                    continue;
+                const siteID = (record["Site ID"] ?? "").trim();
+                const sourceID = siteID
+                    ? `${USGOV_HRSA_FQHC_ADAPTER_ID}-${siteID}`
+                    : stableSourceID(USGOV_HRSA_FQHC_ADAPTER_ID, aligned);
+                yield {
+                    raw,
+                    components: aligned,
+                    country: "US",
+                    locale: "en-US",
+                    source: USGOV_HRSA_FQHC_ADAPTER_ID,
+                    source_id: sourceID,
+                    corpus_version: "",
+                    license: USGOV_HRSA_FQHC_DEFAULT_LICENSE,
+                };
+                emitted++;
+            }
+        },
+    };
+}
+/**
+ * The configured adapter instance registered with the corpus builder.
+ */
+export const usgovHrsaFqhcAdapter = createUSGovHRSAFQHCAdapter();
+//# sourceMappingURL=adapter.js.map

package/out/adapters/usgov-hrsa-fqhc/adapter.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../lib/adapters/usgov-hrsa-fqhc/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA6BG;AAEH,OAAO,EAAE,mBAAmB,EAAE,MAAM,sBAAsB,CAAA;AAC1D,OAAO,EAAE,cAAc,EAAE,MAAM,aAAa,CAAA;AAE5C,OAAO,EAAE,UAAU,EAAE,eAAe,EAAE,cAAc,EAAE,MAAM,iBAAiB,CAAA;AAC7E,OAAO,EAAE,uBAAuB,EAAE,MAAM,sBAAsB,CAAA;AAG9D;;;GAGG;AACH,MAAM,CAAC,MAAM,0BAA0B,GAAG,iBAAiB,CAAA;AAC3D;;;GAGG;AACH,MAAM,CAAC,MAAM,+BAA+B,GAAG,eAAe,CAAA;AAmB9D;;;;;;;;GAQG;AAEH,MAAM,UAAU,0BAA0B;IACzC,OAAO;QACN,EAAE,EAAE,0BAA0B;QAC9B,cAAc,EAAE,+BAA+B;QAC/C,WAAW,EACV,yIAAyI;QAE1I,KAAK,CAAC,CAAC,IAAI,CAAC,IAAoB;YAC/B,IAAI,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,OAAO,KAAK,IAAI,EAAE,CAAC;gBAC3C,MAAM,IAAI,KAAK,CAAC,2DAA2D,IAAI,CAAC,OAAO,EAAE,CAAC,CAAA;YAC3F,CAAC;YAED,MAAM,IAAI,GAAG,cAAc,CAAC,SAAS,CAAC,IAAI,CAAC,SAAS,EAAE;gBACrD,IAAI,EAAE,QAAQ;gBACd,aAAa,EAAE,KAAK;gBACpB,mBAAmB,EAAE,IAAI;aACzB,CAAC,CAAA;YAEF,IAAI,OAAO,GAAG,CAAC,CAAA;YAEf,IAAI,KAAK,EAAE,MAAM,MAAM,IAAI,IAAkC,EAAE,CAAC;gBAC/D,IAAI,IAAI,CAAC,MAAM,EAAE,OAAO;oBAAE,MAAK;gBAE/B,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,IAAI,CAAC,KAAK;oBAAE,MAAK;gBAE5D,MAAM,KAAK,GAAG,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAChD,MAAM,KAAK,GAAG,eAAe,CAAC,MAAM,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC,CAAA;gBAC3D,MAAM,IAAI,GAAG,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAC/C,MAAM,SAAS,GAAG,CAAC,MAAM,CAAC,yBAAyB,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAClE,MAAM,QAAQ,GAAG,CAAC,MAAM,CAAC,kBAAkB,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAE1D,IAAI,CAAC,KAAK,IAAI,CAAC,KAAK,IAAI,CAAC,IAAI,IAAI,CAAC,QAAQ;oBAAE,SAAQ;gBACpD,MAAM,KAAK,GAAG,uBAAuB,CAAC,SAAS,CAAC,CAAA;gBAEhD,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBAEpB,2EAA2E;gBAC3E,wEAAwE;gBACxE,mEAAmE;gBACnE,0DAA0D;gBAC1D,MAAM,UAAU,GAA+B;oBAC9C,KAAK;oBACL,GAAG,CAAC,KAAK,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,YAAY,EAAE,KAAK,CAAC,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBACnE,MAAM,EAAE,KAAK,CAAC,MAAM;oBACpB,QAAQ,EAAE,IAAI;oBACd,MAAM,EAAE,KAAK,CAAC,YAAY;oBAC1B,QAAQ;iBACR,CAAA;gBAED,MAAM,GAAG,GAAG,UAAU,CAAC;oBACtB,KAAK;oBACL,WAAW,EAAE,KAAK,CAAC,YAAY;oBAC/B,MAAM,EAAE,KAAK,CAAC,MAAM;oBACpB,QAAQ,EAAE,IAAI;oBACd,MAAM,EAAE,KAAK,CAAC,YAAY;oBAC1B,QAAQ;iBACR,CAAC,CAAA;gBAEF,IAAI,CAAC,GAAG;oBAAE,SAAQ;gBAElB,MAAM,OAAO,GAAG,mBAAmB,CAAC,UAAU,EAAE,GAAG,CAAC,CAAA;gBAEpD,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,MAAM;oBAAE,SAAQ;gBAE1C,MAAM,MAAM,GAAG,CAAC,MAAM,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAE/C,MAAM,QAAQ,GAAG,MAAM;oBACtB,CAAC,CAAC,GAAG,0BAA0B,IAAI,MAAM,EAAE;oBAC3C,CAAC,CAAC,cAAc,CAAC,0BAA0B,EAAE,OAAO,CAAC,CAAA;gBAEtD,MAAM;oBACL,GAAG;oBACH,UAAU,EAAE,OAAO;oBACnB,OAAO,EAAE,IAAI;oBACb,MAAM,EAAE,OAAO;oBACf,MAAM,EAAE,0BAA0B;oBAClC,SAAS,EAAE,QAAQ;oBACnB,cAAc,EAAE,EAAE;oBAClB,OAAO,EAAE,+BAA+B;iBACxC,CAAA;gBAED,OAAO,EAAE,CAAA;YACV,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,oBAAoB,GAAG,0BAA0B,EAAE,CAAA"}

package/out/adapters/usgov-imls-pls/adapter.d.ts

@@ -0,0 +1,36 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   `usgov-imls-pls`: IMLS Public Libraries Survey outlet CSV consumer.
+ *
+ *   The Institute of Museum and Library Services publishes an annual Public Libraries Survey with one
+ *   row per library outlet (~17K rows). Each row carries the library name, street address, city,
+ *   ZIP, county, and geocoordinates.
+ *
+ *   The adapter consumes the outlet CSV the operator pre-downloads via `fetch-imls-pls.ts`. Column
+ *   names match the IMLS PLS outlet file header.
+ *
+ *   Output: one row per outlet with `venue` (library name), `(house_number, street, locality,
+ *   subregion, postcode)`, and lat/lon preserved in `source_id` stability.
+ *
+ *   License: stamped `"Public Domain"` per IMLS federal government distribution terms.
+ */
+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 USGOV_IMLS_PLS_ADAPTER_ID = "usgov-imls-pls";
+/**
+ * License carried by this source (Public Domain), attached to each row so downstream consumers inherit the terms rather
+ * than having to look them up.
+ */
+export declare const USGOV_IMLS_PLS_DEFAULT_LICENSE = "Public Domain";
+export declare function createUSGovIMLSPLSAdapter(): CorpusAdapter;
+/**
+ * The configured adapter instance registered with the corpus builder.
+ */
+export declare const usgovImlsPlsAdapter: CorpusAdapter;
+//# sourceMappingURL=adapter.d.ts.map

package/out/adapters/usgov-imls-pls/adapter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../lib/adapters/usgov-imls-pls/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAQH,OAAO,KAAK,EAAgC,aAAa,EAAE,MAAM,QAAQ,CAAA;AAEzE;;;GAGG;AACH,eAAO,MAAM,yBAAyB,mBAAmB,CAAA;AACzD;;;GAGG;AACH,eAAO,MAAM,8BAA8B,kBAAkB,CAAA;AAY7D,wBAAgB,yBAAyB,IAAI,aAAa,CAwFzD;AAED;;GAEG;AACH,eAAO,MAAM,mBAAmB,eAA8B,CAAA"}

package/out/adapters/usgov-imls-pls/adapter.js

@@ -0,0 +1,114 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   `usgov-imls-pls`: IMLS Public Libraries Survey outlet CSV consumer.
+ *
+ *   The Institute of Museum and Library Services publishes an annual Public Libraries Survey with one
+ *   row per library outlet (~17K rows). Each row carries the library name, street address, city,
+ *   ZIP, county, and geocoordinates.
+ *
+ *   The adapter consumes the outlet CSV the operator pre-downloads via `fetch-imls-pls.ts`. Column
+ *   names match the IMLS PLS outlet file header.
+ *
+ *   Output: one row per outlet with `venue` (library name), `(house_number, street, locality,
+ *   subregion, postcode)`, and lat/lon preserved in `source_id` stability.
+ *
+ *   License: stamped `"Public Domain"` per IMLS federal government distribution terms.
+ */
+import { isPresent } from "@mailwoman/core/objects";
+import { reconcileComponents } from "@mailwoman/formatter";
+import { CSVSpliterator } from "spliterator";
+import { splitStreetLine, stableSourceID } from "#adapters/utils";
+import { lookupStateAbbreviation } from "#codex/us-fips-state";
+/**
+ * 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 const USGOV_IMLS_PLS_ADAPTER_ID = "usgov-imls-pls";
+/**
+ * License carried by this source (Public Domain), attached to each row so downstream consumers inherit the terms rather
+ * than having to look them up.
+ */
+export const USGOV_IMLS_PLS_DEFAULT_LICENSE = "Public Domain";
+export function createUSGovIMLSPLSAdapter() {
+    return {
+        id: USGOV_IMLS_PLS_ADAPTER_ID,
+        defaultLicense: USGOV_IMLS_PLS_DEFAULT_LICENSE,
+        description: "IMLS Public Libraries Survey — ~17K library outlets with venue+address (public-domain).",
+        async *rows(opts) {
+            if (opts.country && opts.country !== "US") {
+                throw new Error(`usgov-imls-pls adapter: only US supported, got country=${opts.country}`);
+            }
+            const rows = CSVSpliterator.fromAsync(opts.inputPath, {
+                mode: "object",
+                normalizeKeys: false,
+                enableQuoteHandling: true,
+            });
+            let emitted = 0;
+            for await (const record of rows) {
+                if (opts.signal?.aborted)
+                    break;
+                if (opts.limit !== undefined && emitted >= opts.limit)
+                    break;
+                const libName = (record.LIBNAME ?? "").trim();
+                const address = (record.ADDRESS ?? "").trim();
+                const city = (record.CITY ?? "").trim();
+                const zip = (record.ZIP ?? "").trim();
+                const stateAbbr = (record.STABR ?? "").trim();
+                if (!libName || !city || !zip)
+                    continue;
+                const state = lookupStateAbbreviation(stateAbbr);
+                if (!state)
+                    continue;
+                const split = splitStreetLine(address);
+                if (!split)
+                    continue;
+                const components = {
+                    venue: libName,
+                    ...(split.house_number ? { house_number: split.house_number } : {}),
+                    street: split.street,
+                    locality: city,
+                    region: state.abbreviation,
+                    postcode: zip,
+                    // #552: no subregion — US postal addresses don't surface the county, so emitting
+                    // subregion creates a phantom component with no raw-span to align to, quarantining
+                    // ~21% of rows. The county is still available in the source CSV; it just isn't
+                    // a postal-surface component here.
+                };
+                const streetPart = [split.house_number, split.street].filter(isPresent).join(" ").trim();
+                const raw = [
+                    libName,
+                    streetPart,
+                    [city, [stateAbbr, zip].filter(isPresent).join(" ")].filter(isPresent).join(", "),
+                ]
+                    .filter(isPresent)
+                    .join(", ");
+                const aligned = reconcileComponents(components, raw);
+                if (Object.keys(aligned).length <= 2)
+                    continue;
+                const fscsKey = (record.FSCSKEY ?? "").trim();
+                const sourceID = fscsKey
+                    ? `${USGOV_IMLS_PLS_ADAPTER_ID}-${fscsKey}`
+                    : stableSourceID(USGOV_IMLS_PLS_ADAPTER_ID, aligned);
+                yield {
+                    raw,
+                    components: aligned,
+                    country: "US",
+                    locale: "en-US",
+                    source: USGOV_IMLS_PLS_ADAPTER_ID,
+                    source_id: sourceID,
+                    corpus_version: "",
+                    license: USGOV_IMLS_PLS_DEFAULT_LICENSE,
+                };
+                emitted++;
+            }
+        },
+    };
+}
+/**
+ * The configured adapter instance registered with the corpus builder.
+ */
+export const usgovImlsPlsAdapter = createUSGovIMLSPLSAdapter();
+//# sourceMappingURL=adapter.js.map

package/out/adapters/usgov-imls-pls/adapter.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../lib/adapters/usgov-imls-pls/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAEH,OAAO,EAAE,SAAS,EAAE,MAAM,yBAAyB,CAAA;AACnD,OAAO,EAAE,mBAAmB,EAAE,MAAM,sBAAsB,CAAA;AAC1D,OAAO,EAAE,cAAc,EAAE,MAAM,aAAa,CAAA;AAE5C,OAAO,EAAE,eAAe,EAAE,cAAc,EAAE,MAAM,iBAAiB,CAAA;AACjE,OAAO,EAAE,uBAAuB,EAAE,MAAM,sBAAsB,CAAA;AAG9D;;;GAGG;AACH,MAAM,CAAC,MAAM,yBAAyB,GAAG,gBAAgB,CAAA;AACzD;;;GAGG;AACH,MAAM,CAAC,MAAM,8BAA8B,GAAG,eAAe,CAAA;AAY7D,MAAM,UAAU,yBAAyB;IACxC,OAAO;QACN,EAAE,EAAE,yBAAyB;QAC7B,cAAc,EAAE,8BAA8B;QAC9C,WAAW,EAAE,yFAAyF;QAEtG,KAAK,CAAC,CAAC,IAAI,CAAC,IAAoB;YAC/B,IAAI,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,OAAO,KAAK,IAAI,EAAE,CAAC;gBAC3C,MAAM,IAAI,KAAK,CAAC,0DAA0D,IAAI,CAAC,OAAO,EAAE,CAAC,CAAA;YAC1F,CAAC;YAED,MAAM,IAAI,GAAG,cAAc,CAAC,SAAS,CAAC,IAAI,CAAC,SAAS,EAAE;gBACrD,IAAI,EAAE,QAAQ;gBACd,aAAa,EAAE,KAAK;gBACpB,mBAAmB,EAAE,IAAI;aACzB,CAAC,CAAA;YAEF,IAAI,OAAO,GAAG,CAAC,CAAA;YAEf,IAAI,KAAK,EAAE,MAAM,MAAM,IAAI,IAAoC,EAAE,CAAC;gBACjE,IAAI,IAAI,CAAC,MAAM,EAAE,OAAO;oBAAE,MAAK;gBAE/B,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,IAAI,CAAC,KAAK;oBAAE,MAAK;gBAE5D,MAAM,OAAO,GAAG,CAAC,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAC7C,MAAM,OAAO,GAAG,CAAC,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAC7C,MAAM,IAAI,GAAG,CAAC,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBACvC,MAAM,GAAG,GAAG,CAAC,MAAM,CAAC,GAAG,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBACrC,MAAM,SAAS,GAAG,CAAC,MAAM,CAAC,KAAK,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAE7C,IAAI,CAAC,OAAO,IAAI,CAAC,IAAI,IAAI,CAAC,GAAG;oBAAE,SAAQ;gBAEvC,MAAM,KAAK,GAAG,uBAAuB,CAAC,SAAS,CAAC,CAAA;gBAEhD,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBAEpB,MAAM,KAAK,GAAG,eAAe,CAAC,OAAO,CAAC,CAAA;gBAEtC,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBAEpB,MAAM,UAAU,GAA+B;oBAC9C,KAAK,EAAE,OAAO;oBACd,GAAG,CAAC,KAAK,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,YAAY,EAAE,KAAK,CAAC,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBACnE,MAAM,EAAE,KAAK,CAAC,MAAM;oBACpB,QAAQ,EAAE,IAAI;oBACd,MAAM,EAAE,KAAK,CAAC,YAAY;oBAC1B,QAAQ,EAAE,GAAG;oBACb,iFAAiF;oBACjF,mFAAmF;oBACnF,+EAA+E;oBAC/E,mCAAmC;iBACnC,CAAA;gBAED,MAAM,UAAU,GAAG,CAAC,KAAK,CAAC,YAAY,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;gBAExF,MAAM,GAAG,GAAG;oBACX,OAAO;oBACP,UAAU;oBACV,CAAC,IAAI,EAAE,CAAC,SAAS,EAAE,GAAG,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC;iBACjF;qBACC,MAAM,CAAC,SAAS,CAAC;qBACjB,IAAI,CAAC,IAAI,CAAC,CAAA;gBAEZ,MAAM,OAAO,GAAG,mBAAmB,CAAC,UAAU,EAAE,GAAG,CAAC,CAAA;gBAEpD,IAAI,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,MAAM,IAAI,CAAC;oBAAE,SAAQ;gBAE9C,MAAM,OAAO,GAAG,CAAC,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;gBAE7C,MAAM,QAAQ,GAAG,OAAO;oBACvB,CAAC,CAAC,GAAG,yBAAyB,IAAI,OAAO,EAAE;oBAC3C,CAAC,CAAC,cAAc,CAAC,yBAAyB,EAAE,OAAO,CAAC,CAAA;gBAErD,MAAM;oBACL,GAAG;oBACH,UAAU,EAAE,OAAO;oBACnB,OAAO,EAAE,IAAI;oBACb,MAAM,EAAE,OAAO;oBACf,MAAM,EAAE,yBAAyB;oBACjC,SAAS,EAAE,QAAQ;oBACnB,cAAc,EAAE,EAAE;oBAClB,OAAO,EAAE,8BAA8B;iBACvC,CAAA;gBAED,OAAO,EAAE,CAAA;YACV,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,mBAAmB,GAAG,yBAAyB,EAAE,CAAA"}