@mailwoman/corpus 9.4.0 → 10.1.0

package/out/nl/tools/fetch/kadaster.d.ts

@@ -0,0 +1,155 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   Downloads the INSPIRE Addresses theme Kadaster publishes for the Netherlands.
+ *
+ *   ## One feed and one file
+ *
+ *   PDOK serves an ATOM download service with a single entry, and that entry's one data link is the
+ *   whole country. There is no service document above it and no dataset feed below it, so the
+ *   resolution is one request:
+ *
+ *       adressen_inspire_geharmoniseerd.xml  →  addresses.gml.gz
+ *
+ *   The entry's `<id>` names a sibling feed, `adressen_inspire_geharmoniseerd_epsg4258.xml`, which
+ *   answers HTTP 404. A reader that followed the `<id>` rather than the `rel="alternate"` link would
+ *   therefore raise on that 404, so this one reads the link.
+ *
+ *   ## Why the download stays compressed
+ *
+ *   `#nl/adapters/kadaster/adapter` inflates `.gml.gz` through `gunzipChunks` as it streams. The
+ *   member is 29,677,448,685 bytes against the download's 819,465,603, so inflating it here would
+ *   cost 36 times the disk for no reader that wants it.
+ *
+ *   ## Freshness
+ *
+ *   The entry's `<updated>` is the signal recorded in the manifest. A skip requires the feed to
+ *   state an `<updated>` value:
+ *   where it states none both sides read `null`, an equality test would hold, and a download of
+ *   unknown age would be kept for as long as the feed stayed silent.
+ */
+import { APIClient } from "@mailwoman/core/api";
+import { type PathBuilderLike } from "path-ts";
+import type { BaseFetchOptions, FetchSummary, SourceManifest } from "#tools/fetch/download";
+/**
+ * The ATOM download service for the theme, which holds the country's one entry.
+ */
+export declare const NL_KADASTER_FEED_URL = "https://service.pdok.nl/kadaster/ad/atom/adressen_inspire_geharmoniseerd.xml";
+/**
+ * The name the download is written under, which is what the adapter's `inputPath` states.
+ *
+ * The publisher's own name for the file.
+ * The adapter reads a `.gml.gz` by inflating it, so the suffix is load-bearing.
+ */
+export declare const NL_KADASTER_DOWNLOAD_FILENAME = "addresses.gml.gz";
+/**
+ * The attribution recorded in the manifest.
+ *
+ * CC0 1.0 reserves no act and owes no attribution clause, so this is the credit the
+ * repository chooses to carry rather than a condition of the grant.
+ */
+export declare const NL_KADASTER_ATTRIBUTION = "Kadaster";
+/**
+ * What one run recorded.
+ *
+ * `feed_updated` is the entry's own `<updated>`.
+ * A later run downloads again when that value changes, which is what the ATOM
+ * pattern offers in place of an HTTP validator.
+ *
+ * `rights` is the license URL the feed states, recorded so a change of terms is
+ * visible in the artifact rather than only in the register.
+ */
+export interface NLKadasterManifest extends SourceManifest {
+    feed_url: string;
+    feed_updated: string | null;
+    rights: string | null;
+    license: string;
+    attribution: string;
+}
+/**
+ * Per-invocation options.
+ */
+export interface DownloadNLKadasterOptions {
+    outputDir: PathBuilderLike;
+    /**
+     * Downloads the file even where the feed reports the recorded `<updated>`.
+     */
+    force?: boolean;
+    retries?: number;
+    retryDelayMs?: number;
+    signal?: AbortSignal;
+    report?: (line: string) => void;
+}
+/**
+ * Where the download is, and what the feed states about it.
+ */
+export interface NLKadasterPublication {
+    downloadURL: string;
+    /**
+     * The entry's own `<updated>`, or `null` where the feed states none.
+     *
+     * `updated` sits on the Atom entry as well as on the feed, and this reads
+     * the entry that carried the data link.
+     * The feed reported `2026-09-02T11:45:47Z` when this reader was written.
+     */
+    feedUpdated: string | null;
+    /**
+     * The entry's `<rights>`, or `null` where it states none.
+     */
+    rights: string | null;
+    /**
+     * The byte count the link claims, or `null` where it claims none.
+     *
+     * Never taken as the size of the download: the bytes recorded in the manifest
+     * are counted off the delivered body.
+     */
+    claimedBytes: number | null;
+}
+/**
+ * Reads the feed and answers the one entry's data link.
+ *
+ * Raises where the feed lists no entry, or where its entry offers no data link.
+ * The alternative would return `{fetched: 0, skipped: 0, failed: 0}`, which a caller
+ * reads as a fetch that completed and found the publisher empty.
+ *
+ * Separate from {@linkcode downloadNLKadaster} because the resolution is the part a test can drive.
+ * The download runs on global `fetch`, which a unit test cannot intercept.
+ */
+export declare function readNLKadasterPublication(client: Pick<APIClient, "fetch">, options?: {
+    signal?: AbortSignal;
+    report?: (line: string) => void;
+}): Promise<NLKadasterPublication>;
+/**
+ * Whether the file on disk is the one the feed currently offers.
+ *
+ * A skip requires the feed to state an `<updated>` value.
+ * Where it states none, both sides read `null` and an equality test would hold,
+ * which would keep a download of unknown age for as long as the feed stayed silent.
+ *
+ * Downloading 782 MiB again is the slower error and the recoverable one.
+ */
+export declare function nlKadasterPublicationIsRecorded(recorded: NLKadasterManifest, bytesOnDisk: number, publication: NLKadasterPublication): boolean;
+/**
+ * Resolves the feed to its one file and downloads it.
+ */
+export declare function downloadNLKadaster(client: Pick<APIClient, "fetch">, options: DownloadNLKadasterOptions): Promise<FetchSummary>;
+/**
+ * The path `#nl/adapters/kadaster/adapter` reads, given the root a fetch wrote under.
+ */
+export declare function nlKadasterInputPath(outRoot: BaseFetchOptions["outRoot"]): PathBuilderLike;
+/**
+ * Per-invocation options for the registry entry.
+ */
+export interface FetchNLKadasterOptions extends BaseFetchOptions {
+    force?: boolean;
+    retries?: number;
+    retryDelayMs?: number;
+    signal?: AbortSignal;
+}
+/**
+ * The registry entry.
+ */
+export declare function fetchNLKadaster(options: FetchNLKadasterOptions, report?: (line: string) => void): Promise<FetchSummary>;
+//# sourceMappingURL=kadaster.d.ts.map

package/out/nl/tools/fetch/kadaster.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"kadaster.d.ts","sourceRoot":"","sources":["../../../../lib/nl/tools/fetch/kadaster.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA+BG;AAEH,OAAO,EAAE,SAAS,EAA+B,MAAM,qBAAqB,CAAA;AAK5E,OAAO,EAAe,KAAK,eAAe,EAAE,MAAM,SAAS,CAAA;AAI3D,OAAO,KAAK,EAAE,gBAAgB,EAAE,YAAY,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAA;AAG3F;;GAEG;AACH,eAAO,MAAM,oBAAoB,iFAAiF,CAAA;AAElH;;;;;GAKG;AACH,eAAO,MAAM,6BAA6B,qBAAqB,CAAA;AAE/D;;;;;GAKG;AACH,eAAO,MAAM,uBAAuB,aAAa,CAAA;AAIjD;;;;;;;;;GASG;AACH,MAAM,WAAW,kBAAmB,SAAQ,cAAc;IACzD,QAAQ,EAAE,MAAM,CAAA;IAChB,YAAY,EAAE,MAAM,GAAG,IAAI,CAAA;IAC3B,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;IACrB,OAAO,EAAE,MAAM,CAAA;IACf,WAAW,EAAE,MAAM,CAAA;CACnB;AAED;;GAEG;AACH,MAAM,WAAW,yBAAyB;IACzC,SAAS,EAAE,eAAe,CAAA;IAC1B;;OAEG;IACH,KAAK,CAAC,EAAE,OAAO,CAAA;IACf,OAAO,CAAC,EAAE,MAAM,CAAA;IAChB,YAAY,CAAC,EAAE,MAAM,CAAA;IACrB,MAAM,CAAC,EAAE,WAAW,CAAA;IACpB,MAAM,CAAC,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,IAAI,CAAA;CAC/B;AAED;;GAEG;AACH,MAAM,WAAW,qBAAqB;IACrC,WAAW,EAAE,MAAM,CAAA;IACnB;;;;;;OAMG;IACH,WAAW,EAAE,MAAM,GAAG,IAAI,CAAA;IAC1B;;OAEG;IACH,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;IACrB;;;;;OAKG;IACH,YAAY,EAAE,MAAM,GAAG,IAAI,CAAA;CAC3B;AAED;;;;;;;;;GASG;AACH,wBAAsB,yBAAyB,CAC9C,MAAM,EAAE,IAAI,CAAC,SAAS,EAAE,OAAO,CAAC,EAChC,OAAO,GAAE;IAAE,MAAM,CAAC,EAAE,WAAW,CAAC;IAAC,MAAM,CAAC,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,IAAI,CAAA;CAAO,GACrE,OAAO,CAAC,qBAAqB,CAAC,CAkChC;AAED;;;;;;;;GAQG;AACH,wBAAgB,+BAA+B,CAC9C,QAAQ,EAAE,kBAAkB,EAC5B,WAAW,EAAE,MAAM,EACnB,WAAW,EAAE,qBAAqB,GAChC,OAAO,CAIT;AAED;;GAEG;AACH,wBAAsB,kBAAkB,CACvC,MAAM,EAAE,IAAI,CAAC,SAAS,EAAE,OAAO,CAAC,EAChC,OAAO,EAAE,yBAAyB,GAChC,OAAO,CAAC,YAAY,CAAC,CA0DvB;AAED;;GAEG;AACH,wBAAgB,mBAAmB,CAAC,OAAO,EAAE,gBAAgB,CAAC,SAAS,CAAC,GAAG,eAAe,CAEzF;AAED;;GAEG;AACH,MAAM,WAAW,sBAAuB,SAAQ,gBAAgB;IAC/D,KAAK,CAAC,EAAE,OAAO,CAAA;IACf,OAAO,CAAC,EAAE,MAAM,CAAA;IAChB,YAAY,CAAC,EAAE,MAAM,CAAA;IACrB,MAAM,CAAC,EAAE,WAAW,CAAA;CACpB;AAED;;GAEG;AACH,wBAAsB,eAAe,CACpC,OAAO,EAAE,sBAAsB,EAC/B,MAAM,CAAC,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,IAAI,GAC7B,OAAO,CAAC,YAAY,CAAC,CAavB"}

package/out/nl/tools/fetch/kadaster.js

@@ -0,0 +1,179 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ *   Downloads the INSPIRE Addresses theme Kadaster publishes for the Netherlands.
+ *
+ *   ## One feed and one file
+ *
+ *   PDOK serves an ATOM download service with a single entry, and that entry's one data link is the
+ *   whole country. There is no service document above it and no dataset feed below it, so the
+ *   resolution is one request:
+ *
+ *       adressen_inspire_geharmoniseerd.xml  →  addresses.gml.gz
+ *
+ *   The entry's `<id>` names a sibling feed, `adressen_inspire_geharmoniseerd_epsg4258.xml`, which
+ *   answers HTTP 404. A reader that followed the `<id>` rather than the `rel="alternate"` link would
+ *   therefore raise on that 404, so this one reads the link.
+ *
+ *   ## Why the download stays compressed
+ *
+ *   `#nl/adapters/kadaster/adapter` inflates `.gml.gz` through `gunzipChunks` as it streams. The
+ *   member is 29,677,448,685 bytes against the download's 819,465,603, so inflating it here would
+ *   cost 36 times the disk for no reader that wants it.
+ *
+ *   ## Freshness
+ *
+ *   The entry's `<updated>` is the signal recorded in the manifest. A skip requires the feed to
+ *   state an `<updated>` value:
+ *   where it states none both sides read `null`, an equality test would hold, and a download of
+ *   unknown age would be kept for as long as the feed stayed silent.
+ */
+import { APIClient, assertNoOGCServiceException } from "@mailwoman/core/api";
+import { ByteFormatter } from "@mailwoman/core/fs/formatters";
+import { tryStat } from "@mailwoman/core/fs/readers";
+import { makeDirectories } from "@mailwoman/core/fs/writers";
+import { sha256File } from "@mailwoman/core/hash";
+import { PathBuilder } from "path-ts";
+import { NL_KADASTER_ADAPTER_ID, NL_KADASTER_DEFAULT_LICENSE } from "#nl/adapters/kadaster/adapter";
+import { feedChunks, linkWithRel, readAtomFeed } from "#tools/fetch/atom";
+import { downloadToFile, readManifest, writeManifest } from "#tools/fetch/download";
+/**
+ * The ATOM download service for the theme, which holds the country's one entry.
+ */
+export const NL_KADASTER_FEED_URL = "https://service.pdok.nl/kadaster/ad/atom/adressen_inspire_geharmoniseerd.xml";
+/**
+ * The name the download is written under, which is what the adapter's `inputPath` states.
+ *
+ * The publisher's own name for the file.
+ * The adapter reads a `.gml.gz` by inflating it, so the suffix is load-bearing.
+ */
+export const NL_KADASTER_DOWNLOAD_FILENAME = "addresses.gml.gz";
+/**
+ * The attribution recorded in the manifest.
+ *
+ * CC0 1.0 reserves no act and owes no attribution clause, so this is the credit the
+ * repository chooses to carry rather than a condition of the grant.
+ */
+export const NL_KADASTER_ATTRIBUTION = "Kadaster";
+const SLUG = NL_KADASTER_ADAPTER_ID;
+/**
+ * Reads the feed and answers the one entry's data link.
+ *
+ * Raises where the feed lists no entry, or where its entry offers no data link.
+ * The alternative would return `{fetched: 0, skipped: 0, failed: 0}`, which a caller
+ * reads as a fetch that completed and found the publisher empty.
+ *
+ * Separate from {@linkcode downloadNLKadaster} because the resolution is the part a test can drive.
+ * The download runs on global `fetch`, which a unit test cannot intercept.
+ */
+export async function readNLKadasterPublication(client, options = {}) {
+    const { report } = options;
+    report?.(`=== ${SLUG}: reading ${NL_KADASTER_FEED_URL}`);
+    const { data: feedXML } = await client.fetch({
+        url: NL_KADASTER_FEED_URL,
+        timeout: 120_000,
+        signal: options.signal,
+    });
+    assertNoOGCServiceException(String(feedXML), `${SLUG} download service feed`);
+    const feed = await readAtomFeed(feedChunks(String(feedXML)));
+    const offered = feed.entries.flatMap((entry) => {
+        const link = linkWithRel(entry.links, "enclosure") ?? linkWithRel(entry.links, "alternate");
+        return link ? [{ entry, link }] : [];
+    })[0];
+    if (!offered) {
+        throw new Error(`${SLUG}: the feed offers no download through a rel="enclosure" or rel="alternate" link, ` +
+            `so no file can be resolved. It listed ${feed.entries.length} entries.`);
+    }
+    return {
+        downloadURL: offered.link.href,
+        feedUpdated: offered.entry.updated || null,
+        rights: offered.entry.rights,
+        claimedBytes: offered.link.length,
+    };
+}
+/**
+ * Whether the file on disk is the one the feed currently offers.
+ *
+ * A skip requires the feed to state an `<updated>` value.
+ * Where it states none, both sides read `null` and an equality test would hold,
+ * which would keep a download of unknown age for as long as the feed stayed silent.
+ *
+ * Downloading 782 MiB again is the slower error and the recoverable one.
+ */
+export function nlKadasterPublicationIsRecorded(recorded, bytesOnDisk, publication) {
+    if (publication.feedUpdated === null)
+        return false;
+    return recorded.feed_updated === publication.feedUpdated && recorded.bytes === bytesOnDisk;
+}
+/**
+ * Resolves the feed to its one file and downloads it.
+ */
+export async function downloadNLKadaster(client, options) {
+    const { report } = options;
+    const destDir = PathBuilder.from(options.outputDir);
+    const downloadPath = destDir(NL_KADASTER_DOWNLOAD_FILENAME);
+    const manifestPath = destDir("MANIFEST.json");
+    // Resolved before the directory is made, so a publisher that answers with an
+    // exception report leaves no empty source directory behind.
+    const publication = await readNLKadasterPublication(client, { signal: options.signal, report });
+    await makeDirectories(destDir);
+    const recorded = await readManifest(manifestPath);
+    const stat = await tryStat(downloadPath);
+    if (!options.force && recorded && stat && nlKadasterPublicationIsRecorded(recorded, stat.size, publication)) {
+        report?.(`  present, and the feed reports the recorded updated ${publication.feedUpdated}`);
+        return { fetched: 0, skipped: 1, failed: 0, failedCodes: [] };
+    }
+    report?.(`  download: ${publication.downloadURL}` +
+        (publication.claimedBytes === null
+            ? ""
+            : ` (the feed claims ${ByteFormatter.formatIEC(publication.claimedBytes)})`));
+    const { bytes } = await downloadToFile({
+        url: publication.downloadURL,
+        dest: downloadPath,
+        retries: options.retries,
+        retryDelayMs: options.retryDelayMs,
+        report,
+    });
+    const sha256 = await sha256File(downloadPath);
+    report?.(`  ✓ ${ByteFormatter.formatIEC(bytes)}  sha256=${sha256}`);
+    const manifest = {
+        source_url: publication.downloadURL,
+        feed_url: NL_KADASTER_FEED_URL,
+        filename: NL_KADASTER_DOWNLOAD_FILENAME,
+        bytes,
+        sha256,
+        feed_updated: publication.feedUpdated,
+        rights: publication.rights,
+        license: NL_KADASTER_DEFAULT_LICENSE,
+        attribution: NL_KADASTER_ATTRIBUTION,
+        downloaded_at: new Date().toISOString(),
+    };
+    await writeManifest(manifestPath, manifest);
+    report?.(`  MANIFEST written to ${manifestPath}`);
+    return { fetched: 1, skipped: 0, failed: 0, failedCodes: [] };
+}
+/**
+ * The path `#nl/adapters/kadaster/adapter` reads, given the root a fetch wrote under.
+ */
+export function nlKadasterInputPath(outRoot) {
+    return outRoot(SLUG, NL_KADASTER_DOWNLOAD_FILENAME);
+}
+/**
+ * The registry entry.
+ */
+export async function fetchNLKadaster(options, report) {
+    await using client = new APIClient({ displayName: SLUG, retry: true });
+    // Awaited rather than returned: `await using` disposes the client when this scope exits,
+    // and a disposed `APIClient` refuses every later request.
+    return await downloadNLKadaster(client, {
+        outputDir: options.outRoot(SLUG),
+        force: options.force,
+        retries: options.retries,
+        retryDelayMs: options.retryDelayMs,
+        signal: options.signal,
+        report,
+    });
+}
+//# sourceMappingURL=kadaster.js.map

package/out/nl/tools/fetch/kadaster.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"kadaster.js","sourceRoot":"","sources":["../../../../lib/nl/tools/fetch/kadaster.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA+BG;AAEH,OAAO,EAAE,SAAS,EAAE,2BAA2B,EAAE,MAAM,qBAAqB,CAAA;AAC5E,OAAO,EAAE,aAAa,EAAE,MAAM,+BAA+B,CAAA;AAC7D,OAAO,EAAE,OAAO,EAAE,MAAM,4BAA4B,CAAA;AACpD,OAAO,EAAE,eAAe,EAAE,MAAM,4BAA4B,CAAA;AAC5D,OAAO,EAAE,UAAU,EAAE,MAAM,sBAAsB,CAAA;AACjD,OAAO,EAAE,WAAW,EAAwB,MAAM,SAAS,CAAA;AAE3D,OAAO,EAAE,sBAAsB,EAAE,2BAA2B,EAAE,MAAM,+BAA+B,CAAA;AACnG,OAAO,EAAE,UAAU,EAAE,WAAW,EAAE,YAAY,EAAE,MAAM,mBAAmB,CAAA;AAEzE,OAAO,EAAE,cAAc,EAAE,YAAY,EAAE,aAAa,EAAE,MAAM,uBAAuB,CAAA;AAEnF;;GAEG;AACH,MAAM,CAAC,MAAM,oBAAoB,GAAG,8EAA8E,CAAA;AAElH;;;;;GAKG;AACH,MAAM,CAAC,MAAM,6BAA6B,GAAG,kBAAkB,CAAA;AAE/D;;;;;GAKG;AACH,MAAM,CAAC,MAAM,uBAAuB,GAAG,UAAU,CAAA;AAEjD,MAAM,IAAI,GAAG,sBAAsB,CAAA;AA6DnC;;;;;;;;;GASG;AACH,MAAM,CAAC,KAAK,UAAU,yBAAyB,CAC9C,MAAgC,EAChC,OAAO,GAA8D,EAAE;IAEvE,MAAM,EAAE,MAAM,EAAE,GAAG,OAAO,CAAA;IAE1B,MAAM,EAAE,CAAC,OAAO,IAAI,aAAa,oBAAoB,EAAE,CAAC,CAAA;IAExD,MAAM,EAAE,IAAI,EAAE,OAAO,EAAE,GAAG,MAAM,MAAM,CAAC,KAAK,CAAS;QACpD,GAAG,EAAE,oBAAoB;QACzB,OAAO,EAAE,OAAO;QAChB,MAAM,EAAE,OAAO,CAAC,MAAM;KACtB,CAAC,CAAA;IAEF,2BAA2B,CAAC,MAAM,CAAC,OAAO,CAAC,EAAE,GAAG,IAAI,wBAAwB,CAAC,CAAA;IAE7E,MAAM,IAAI,GAAG,MAAM,YAAY,CAAC,UAAU,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,CAAA;IAE5D,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,KAAK,EAAE,EAAE;QAC9C,MAAM,IAAI,GAAG,WAAW,CAAC,KAAK,CAAC,KAAK,EAAE,WAAW,CAAC,IAAI,WAAW,CAAC,KAAK,CAAC,KAAK,EAAE,WAAW,CAAC,CAAA;QAE3F,OAAO,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,EAAE,CAAA;IACrC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAA;IAEL,IAAI,CAAC,OAAO,EAAE,CAAC;QACd,MAAM,IAAI,KAAK,CACd,GAAG,IAAI,mFAAmF;YACzF,yCAAyC,IAAI,CAAC,OAAO,CAAC,MAAM,WAAW,CACxE,CAAA;IACF,CAAC;IAED,OAAO;QACN,WAAW,EAAE,OAAO,CAAC,IAAI,CAAC,IAAI;QAC9B,WAAW,EAAE,OAAO,CAAC,KAAK,CAAC,OAAO,IAAI,IAAI;QAC1C,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,MAAM;QAC5B,YAAY,EAAE,OAAO,CAAC,IAAI,CAAC,MAAM;KACjC,CAAA;AACF,CAAC;AAED;;;;;;;;GAQG;AACH,MAAM,UAAU,+BAA+B,CAC9C,QAA4B,EAC5B,WAAmB,EACnB,WAAkC;IAElC,IAAI,WAAW,CAAC,WAAW,KAAK,IAAI;QAAE,OAAO,KAAK,CAAA;IAElD,OAAO,QAAQ,CAAC,YAAY,KAAK,WAAW,CAAC,WAAW,IAAI,QAAQ,CAAC,KAAK,KAAK,WAAW,CAAA;AAC3F,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,kBAAkB,CACvC,MAAgC,EAChC,OAAkC;IAElC,MAAM,EAAE,MAAM,EAAE,GAAG,OAAO,CAAA;IAC1B,MAAM,OAAO,GAAG,WAAW,CAAC,IAAI,CAAC,OAAO,CAAC,SAAS,CAAC,CAAA;IACnD,MAAM,YAAY,GAAG,OAAO,CAAC,6BAA6B,CAAC,CAAA;IAC3D,MAAM,YAAY,GAAG,OAAO,CAAC,eAAe,CAAC,CAAA;IAE7C,6EAA6E;IAC7E,4DAA4D;IAC5D,MAAM,WAAW,GAAG,MAAM,yBAAyB,CAAC,MAAM,EAAE,EAAE,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,MAAM,EAAE,CAAC,CAAA;IAE/F,MAAM,eAAe,CAAC,OAAO,CAAC,CAAA;IAE9B,MAAM,QAAQ,GAAG,MAAM,YAAY,CAAqB,YAAY,CAAC,CAAA;IACrE,MAAM,IAAI,GAAG,MAAM,OAAO,CAAC,YAAY,CAAC,CAAA;IAExC,IAAI,CAAC,OAAO,CAAC,KAAK,IAAI,QAAQ,IAAI,IAAI,IAAI,+BAA+B,CAAC,QAAQ,EAAE,IAAI,CAAC,IAAI,EAAE,WAAW,CAAC,EAAE,CAAC;QAC7G,MAAM,EAAE,CAAC,wDAAwD,WAAW,CAAC,WAAW,EAAE,CAAC,CAAA;QAE3F,OAAO,EAAE,OAAO,EAAE,CAAC,EAAE,OAAO,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,WAAW,EAAE,EAAE,EAAE,CAAA;IAC9D,CAAC;IAED,MAAM,EAAE,CACP,eAAe,WAAW,CAAC,WAAW,EAAE;QACvC,CAAC,WAAW,CAAC,YAAY,KAAK,IAAI;YACjC,CAAC,CAAC,EAAE;YACJ,CAAC,CAAC,qBAAqB,aAAa,CAAC,SAAS,CAAC,WAAW,CAAC,YAAY,CAAC,GAAG,CAAC,CAC9E,CAAA;IAED,MAAM,EAAE,KAAK,EAAE,GAAG,MAAM,cAAc,CAAC;QACtC,GAAG,EAAE,WAAW,CAAC,WAAW;QAC5B,IAAI,EAAE,YAAY;QAClB,OAAO,EAAE,OAAO,CAAC,OAAO;QACxB,YAAY,EAAE,OAAO,CAAC,YAAY;QAClC,MAAM;KACN,CAAC,CAAA;IAEF,MAAM,MAAM,GAAG,MAAM,UAAU,CAAC,YAAY,CAAC,CAAA;IAE7C,MAAM,EAAE,CAAC,OAAO,aAAa,CAAC,SAAS,CAAC,KAAK,CAAC,YAAY,MAAM,EAAE,CAAC,CAAA;IAEnE,MAAM,QAAQ,GAAuB;QACpC,UAAU,EAAE,WAAW,CAAC,WAAW;QACnC,QAAQ,EAAE,oBAAoB;QAC9B,QAAQ,EAAE,6BAA6B;QACvC,KAAK;QACL,MAAM;QACN,YAAY,EAAE,WAAW,CAAC,WAAW;QACrC,MAAM,EAAE,WAAW,CAAC,MAAM;QAC1B,OAAO,EAAE,2BAA2B;QACpC,WAAW,EAAE,uBAAuB;QACpC,aAAa,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;KACvC,CAAA;IAED,MAAM,aAAa,CAAC,YAAY,EAAE,QAAQ,CAAC,CAAA;IAE3C,MAAM,EAAE,CAAC,yBAAyB,YAAY,EAAE,CAAC,CAAA;IAEjD,OAAO,EAAE,OAAO,EAAE,CAAC,EAAE,OAAO,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,WAAW,EAAE,EAAE,EAAE,CAAA;AAC9D,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,mBAAmB,CAAC,OAAoC;IACvE,OAAO,OAAO,CAAC,IAAI,EAAE,6BAA6B,CAAC,CAAA;AACpD,CAAC;AAYD;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,eAAe,CACpC,OAA+B,EAC/B,MAA+B;IAE/B,YAAY,MAAM,GAAG,IAAI,SAAS,CAAC,EAAE,WAAW,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAA;IAEtE,yFAAyF;IACzF,0DAA0D;IAC1D,OAAO,MAAM,kBAAkB,CAAC,MAAM,EAAE;QACvC,SAAS,EAAE,OAAO,CAAC,OAAO,CAAC,IAAI,CAAC;QAChC,KAAK,EAAE,OAAO,CAAC,KAAK;QACpB,OAAO,EAAE,OAAO,CAAC,OAAO;QACxB,YAAY,EAAE,OAAO,CAAC,YAAY;QAClC,MAAM,EAAE,OAAO,CAAC,MAAM;QACtB,MAAM;KACN,CAAC,CAAA;AACH,CAAC"}

package/out/no/adapters/matrikkelen/adapter.d.ts

@@ -0,0 +1,44 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ * `matrikkelen`: Kartverket's `Matrikkelen - Adresse` CSV adapter, covering Norway and Svalbard.
+ *
+ * Input is a semicolon-separated CSV from Geonorge, published per area beside a whole-country
+ * extract. Both carry the same 46 columns, so one reader serves either. The adapter reads
+ * `adressenavn` as the street, `nummer` and `bokstav` as the house number, and `postnummer` with
+ * `poststed` as the postcode and its postal place.
+ *
+ * A row's country comes from `kommunenummer` rather than from the caller. Svalbard is municipality
+ * 2100 and is its own ISO 3166-1 jurisdiction, so the dataset's own area list is what separates the
+ * two: Geonorge offers Svalbard as `kommune` 2100 and `fylke` 21 beside the `landsdekkende` whole of
+ * the mainland.
+ *
+ * Kartverket states CC BY 4.0 for this dataset across five agreeing metadata fields, so the adapter
+ * records that license on every row and the model card must attribute Kartverket.
+ *
+ * The adapter streams with `CSVSpliterator.fromAsync`, so the 2.6M-row mainland extract never sits
+ * in memory. It honors `opts.limit`, `opts.signal` and `opts.country`.
+ */
+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 MATRIKKELEN_ADAPTER_ID = "matrikkelen";
+/**
+ * Every jurisdiction this adapter emits, checked against a caller's `--country`.
+ */
+export declare const MATRIKKELEN_COUNTRIES: readonly string[];
+/**
+ * The country a row belongs to, read from its municipality number.
+ */
+export declare function countryOfMunicipality(kommunenummer: string): string;
+export declare function createMatrikkelenAdapter(): CorpusAdapter;
+/**
+ * The configured adapter instance registered with the corpus builder.
+ */
+export declare const matrikkelenAdapter: CorpusAdapter;
+//# sourceMappingURL=adapter.d.ts.map

package/out/no/adapters/matrikkelen/adapter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"adapter.d.ts","sourceRoot":"","sources":["../../../../lib/no/adapters/matrikkelen/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;GAsBG;AASH,OAAO,EAAuD,KAAK,aAAa,EAAiB,MAAM,QAAQ,CAAA;AAE/G;;;;GAIG;AACH,eAAO,MAAM,sBAAsB,gBAAgB,CAAA;AAUnD;;GAEG;AACH,eAAO,MAAM,qBAAqB,EAAE,SAAS,MAAM,EAAiB,CAAA;AAqBpE;;GAEG;AACH,wBAAgB,qBAAqB,CAAC,aAAa,EAAE,MAAM,GAAG,MAAM,CAEnE;AAED,wBAAgB,wBAAwB,IAAI,aAAa,CAsGxD;AAED;;GAEG;AACH,eAAO,MAAM,kBAAkB,eAA6B,CAAA"}

package/out/no/adapters/matrikkelen/adapter.js

@@ -0,0 +1,143 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ *
+ * `matrikkelen`: Kartverket's `Matrikkelen - Adresse` CSV adapter, covering Norway and Svalbard.
+ *
+ * Input is a semicolon-separated CSV from Geonorge, published per area beside a whole-country
+ * extract. Both carry the same 46 columns, so one reader serves either. The adapter reads
+ * `adressenavn` as the street, `nummer` and `bokstav` as the house number, and `postnummer` with
+ * `poststed` as the postcode and its postal place.
+ *
+ * A row's country comes from `kommunenummer` rather than from the caller. Svalbard is municipality
+ * 2100 and is its own ISO 3166-1 jurisdiction, so the dataset's own area list is what separates the
+ * two: Geonorge offers Svalbard as `kommune` 2100 and `fylke` 21 beside the `landsdekkende` whole of
+ * the mainland.
+ *
+ * Kartverket states CC BY 4.0 for this dataset across five agreeing metadata fields, so the adapter
+ * records that license on every row and the model card must attribute Kartverket.
+ *
+ * The adapter streams with `CSVSpliterator.fromAsync`, so the 2.6M-row mainland extract never sits
+ * in memory. It honors `opts.limit`, `opts.signal` and `opts.country`.
+ */
+import { formatAddressRow } from "@mailwoman/codex/address/format";
+import { CSVSpliterator } from "spliterator";
+import { UnsupportedCountryError } from "#adapters/errors";
+import { stableSourceID } from "#adapters/source-id";
+import { composeHouseNumber } from "#adapters/street-line";
+import { SourceRegister } from "#registers";
+import { AddressRole, SurfaceOrigin } 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 const MATRIKKELEN_ADAPTER_ID = "matrikkelen";
+/**
+ * The municipality number Kartverket gives Svalbard.
+ *
+ * Every other number in the dataset belongs to a mainland municipality.
+ * Jan Mayen has no entry, although ISO 3166-1 assigns `SJ` to it as well as to Svalbard.
+ */
+const SVALBARD_MUNICIPALITY = "2100";
+/**
+ * Every jurisdiction this adapter emits, checked against a caller's `--country`.
+ */
+export const MATRIKKELEN_COUNTRIES = ["NO", "SJ"];
+/**
+ * The country a row belongs to, read from its municipality number.
+ */
+export function countryOfMunicipality(kommunenummer) {
+    return kommunenummer.trim() === SVALBARD_MUNICIPALITY ? "SJ" : "NO";
+}
+export function createMatrikkelenAdapter() {
+    return {
+        id: MATRIKKELEN_ADAPTER_ID,
+        defaultLicense: "CC-BY-4.0",
+        addressRole: AddressRole.Premise,
+        register: SourceRegister.MatrikkelenAdresse,
+        surface: SurfaceOrigin.Rendered,
+        description: "Matrikkelen - Adresse (Kartverket): street-level addresses for Norway and Svalbard.",
+        async *rows(opts) {
+            if (opts.country && !MATRIKKELEN_COUNTRIES.includes(opts.country)) {
+                throw new UnsupportedCountryError(MATRIKKELEN_ADAPTER_ID, MATRIKKELEN_COUNTRIES, opts.country);
+            }
+            const rows = CSVSpliterator.fromAsync(opts.inputPath, {
+                normalizeKeys: false,
+                columnDelimiter: ";",
+            });
+            let emitted = 0;
+            let cadastral = 0;
+            try {
+                for await (const record of rows) {
+                    if (opts.signal?.aborted)
+                        break;
+                    if (opts.limit !== undefined && emitted >= opts.limit)
+                        break;
+                    const country = countryOfMunicipality(record.kommunenummer ?? "");
+                    if (opts.country && country !== opts.country)
+                        continue;
+                    // A `matrikkeladresse` addresses a cadastral unit rather than a street, so its
+                    // `adresseTekst` reads as a farm name and a holding number, such as `Øvrabø, 124/1`.
+                    // It carries no street to align, and the mainland file holds 34,522 of
+                    // them against 2,568,832 `vegadresse` rows.
+                    if ((record.adressetype ?? "").trim() !== "vegadresse") {
+                        cadastral += 1;
+                        continue;
+                    }
+                    const street = (record.adressenavn ?? "").trim();
+                    // Kartverket splits `12B` into `nummer` 12 and `bokstav` B, and writes
+                    // them joined with no separator in `adresseTekst`.
+                    const house = composeHouseNumber(record.nummer ?? "", record.bokstav ?? "");
+                    const postcode = (record.postnummer ?? "").trim();
+                    const locality = (record.poststed ?? "").trim();
+                    if (!street)
+                        continue;
+                    if (!postcode && !locality)
+                        continue;
+                    const components = {};
+                    if (house) {
+                        components.house_number = house;
+                    }
+                    components.street = street;
+                    if (postcode) {
+                        components.postcode = postcode;
+                    }
+                    if (locality) {
+                        components.locality = locality;
+                    }
+                    const rendered = formatAddressRow(components, country, { singleLine: true });
+                    if (!rendered)
+                        continue;
+                    const { raw, components: aligned } = rendered;
+                    const seed = (record.adresseId ?? "").trim();
+                    const sourceID = seed ? `${MATRIKKELEN_ADAPTER_ID}-${seed}` : stableSourceID(MATRIKKELEN_ADAPTER_ID, aligned);
+                    yield {
+                        raw,
+                        components: aligned,
+                        country,
+                        // Norwegian is the language of both jurisdictions' addresses,
+                        // and Bokmål is the written form Kartverket publishes.
+                        locale: `nb-${country}`,
+                        source: MATRIKKELEN_ADAPTER_ID,
+                        source_id: sourceID,
+                        corpus_version: "",
+                        license: "CC-BY-4.0",
+                    };
+                    emitted++;
+                }
+            }
+            finally {
+                if (cadastral > 0) {
+                    process.stderr.write(`  matrikkelen: ${cadastral} matrikkeladresse rows carry no street and were dropped, ${emitted} kept\n`);
+                }
+            }
+        },
+    };
+}
+/**
+ * The configured adapter instance registered with the corpus builder.
+ */
+export const matrikkelenAdapter = createMatrikkelenAdapter();
+//# sourceMappingURL=adapter.js.map

package/out/no/adapters/matrikkelen/adapter.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"adapter.js","sourceRoot":"","sources":["../../../../lib/no/adapters/matrikkelen/adapter.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;GAsBG;AAEH,OAAO,EAAE,gBAAgB,EAAE,MAAM,iCAAiC,CAAA;AAClE,OAAO,EAAE,cAAc,EAAE,MAAM,aAAa,CAAA;AAE5C,OAAO,EAAE,uBAAuB,EAAE,MAAM,kBAAkB,CAAA;AAC1D,OAAO,EAAE,cAAc,EAAE,MAAM,qBAAqB,CAAA;AACpD,OAAO,EAAE,kBAAkB,EAAE,MAAM,uBAAuB,CAAA;AAC1D,OAAO,EAAE,cAAc,EAAE,MAAM,YAAY,CAAA;AAC3C,OAAO,EAAE,WAAW,EAA8D,aAAa,EAAE,MAAM,QAAQ,CAAA;AAE/G;;;;GAIG;AACH,MAAM,CAAC,MAAM,sBAAsB,GAAG,aAAa,CAAA;AAEnD;;;;;GAKG;AACH,MAAM,qBAAqB,GAAG,MAAM,CAAA;AAEpC;;GAEG;AACH,MAAM,CAAC,MAAM,qBAAqB,GAAsB,CAAC,IAAI,EAAE,IAAI,CAAC,CAAA;AAqBpE;;GAEG;AACH,MAAM,UAAU,qBAAqB,CAAC,aAAqB;IAC1D,OAAO,aAAa,CAAC,IAAI,EAAE,KAAK,qBAAqB,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAA;AACpE,CAAC;AAED,MAAM,UAAU,wBAAwB;IACvC,OAAO;QACN,EAAE,EAAE,sBAAsB;QAC1B,cAAc,EAAE,WAAW;QAC3B,WAAW,EAAE,WAAW,CAAC,OAAO;QAChC,QAAQ,EAAE,cAAc,CAAC,kBAAkB;QAC3C,OAAO,EAAE,aAAa,CAAC,QAAQ;QAC/B,WAAW,EAAE,qFAAqF;QAElG,KAAK,CAAC,CAAC,IAAI,CAAC,IAAoB;YAC/B,IAAI,IAAI,CAAC,OAAO,IAAI,CAAC,qBAAqB,CAAC,QAAQ,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;gBACnE,MAAM,IAAI,uBAAuB,CAAC,sBAAsB,EAAE,qBAAqB,EAAE,IAAI,CAAC,OAAO,CAAC,CAAA;YAC/F,CAAC;YAED,MAAM,IAAI,GAAG,cAAc,CAAC,SAAS,CAAC,IAAI,CAAC,SAAS,EAAE;gBACrD,aAAa,EAAE,KAAK;gBACpB,eAAe,EAAE,GAAG;aACpB,CAAC,CAAA;YAEF,IAAI,OAAO,GAAG,CAAC,CAAA;YACf,IAAI,SAAS,GAAG,CAAC,CAAA;YAEjB,IAAI,CAAC;gBACJ,IAAI,KAAK,EAAE,MAAM,MAAM,IAAI,IAAqC,EAAE,CAAC;oBAClE,IAAI,IAAI,CAAC,MAAM,EAAE,OAAO;wBAAE,MAAK;oBAE/B,IAAI,IAAI,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,IAAI,IAAI,CAAC,KAAK;wBAAE,MAAK;oBAE5D,MAAM,OAAO,GAAG,qBAAqB,CAAC,MAAM,CAAC,aAAa,IAAI,EAAE,CAAC,CAAA;oBAEjE,IAAI,IAAI,CAAC,OAAO,IAAI,OAAO,KAAK,IAAI,CAAC,OAAO;wBAAE,SAAQ;oBAEtD,+EAA+E;oBAC/E,qFAAqF;oBACrF,uEAAuE;oBACvE,4CAA4C;oBAC5C,IAAI,CAAC,MAAM,CAAC,WAAW,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,KAAK,YAAY,EAAE,CAAC;wBACxD,SAAS,IAAI,CAAC,CAAA;wBAEd,SAAQ;oBACT,CAAC;oBAED,MAAM,MAAM,GAAG,CAAC,MAAM,CAAC,WAAW,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;oBAChD,uEAAuE;oBACvE,mDAAmD;oBACnD,MAAM,KAAK,GAAG,kBAAkB,CAAC,MAAM,CAAC,MAAM,IAAI,EAAE,EAAE,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC,CAAA;oBAC3E,MAAM,QAAQ,GAAG,CAAC,MAAM,CAAC,UAAU,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;oBACjD,MAAM,QAAQ,GAAG,CAAC,MAAM,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;oBAE/C,IAAI,CAAC,MAAM;wBAAE,SAAQ;oBAErB,IAAI,CAAC,QAAQ,IAAI,CAAC,QAAQ;wBAAE,SAAQ;oBAEpC,MAAM,UAAU,GAA+B,EAAE,CAAA;oBAEjD,IAAI,KAAK,EAAE,CAAC;wBACX,UAAU,CAAC,YAAY,GAAG,KAAK,CAAA;oBAChC,CAAC;oBAED,UAAU,CAAC,MAAM,GAAG,MAAM,CAAA;oBAE1B,IAAI,QAAQ,EAAE,CAAC;wBACd,UAAU,CAAC,QAAQ,GAAG,QAAQ,CAAA;oBAC/B,CAAC;oBAED,IAAI,QAAQ,EAAE,CAAC;wBACd,UAAU,CAAC,QAAQ,GAAG,QAAQ,CAAA;oBAC/B,CAAC;oBAED,MAAM,QAAQ,GAAG,gBAAgB,CAAC,UAAU,EAAE,OAAO,EAAE,EAAE,UAAU,EAAE,IAAI,EAAE,CAAC,CAAA;oBAE5E,IAAI,CAAC,QAAQ;wBAAE,SAAQ;oBAEvB,MAAM,EAAE,GAAG,EAAE,UAAU,EAAE,OAAO,EAAE,GAAG,QAAQ,CAAA;oBAC7C,MAAM,IAAI,GAAG,CAAC,MAAM,CAAC,SAAS,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAA;oBAE5C,MAAM,QAAQ,GAAG,IAAI,CAAC,CAAC,CAAC,GAAG,sBAAsB,IAAI,IAAI,EAAE,CAAC,CAAC,CAAC,cAAc,CAAC,sBAAsB,EAAE,OAAO,CAAC,CAAA;oBAE7G,MAAM;wBACL,GAAG;wBACH,UAAU,EAAE,OAAO;wBACnB,OAAO;wBACP,8DAA8D;wBAC9D,uDAAuD;wBACvD,MAAM,EAAE,MAAM,OAAO,EAAE;wBACvB,MAAM,EAAE,sBAAsB;wBAC9B,SAAS,EAAE,QAAQ;wBACnB,cAAc,EAAE,EAAE;wBAClB,OAAO,EAAE,WAAW;qBACpB,CAAA;oBAED,OAAO,EAAE,CAAA;gBACV,CAAC;YACF,CAAC;oBAAS,CAAC;gBACV,IAAI,SAAS,GAAG,CAAC,EAAE,CAAC;oBACnB,OAAO,CAAC,MAAM,CAAC,KAAK,CACnB,kBAAkB,SAAS,4DAA4D,OAAO,SAAS,CACvG,CAAA;gBACF,CAAC;YACF,CAAC;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,MAAM,kBAAkB,GAAG,wBAAwB,EAAE,CAAA"}

package/out/no/recipes/fragment.d.ts

@@ -0,0 +1,11 @@
+/**
+ * @copyright Sister Software
+ * @license AGPL-3.0
+ * @author Teffen Ellis, et al.
+ */
+import { type CorpusRecipe } from "#recipes/scaffold";
+/**
+ * Norwegian fragment recipe.
+ */
+export declare const noFragmentRecipe: CorpusRecipe;
+//# sourceMappingURL=fragment.d.ts.map

package/out/no/recipes/fragment.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"fragment.d.ts","sourceRoot":"","sources":["../../../lib/no/recipes/fragment.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAKH,OAAO,EAKN,KAAK,YAAY,EAEjB,MAAM,mBAAmB,CAAA;AAgB1B;;GAEG;AACH,eAAO,MAAM,gBAAgB,EAAE,YAyJ9B,CAAA"}