@monoes/monomindcli 2.14.0 → 2.15.0

package/dist/src/knowledge/library.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"library.d.ts","sourceRoot":"","sources":["../../../src/knowledge/library.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;GAmBG;AAGH,OAAO,KAAK,EAAE,iBAAiB,EAAE,MAAM,uBAAuB,CAAC;AAC/D,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,wBAAwB,CAAC;AAE3D,MAAM,WAAW,aAAa;IAC5B,sCAAsC;IACtC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC;IAChB,GAAG,CAAC,EAAE,MAAM,EAAE,CAAC;IACf,UAAU,CAAC,EAAE,MAAM,EAAE,CAAC;IACtB,4CAA4C;IAC5C,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,4DAA4D;IAC5D,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,wDAAwD;IACxD,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,0CAA0C;IAC1C,IAAI,CAAC,EAAE,MAAM,CAAC;CACf;AAED,MAAM,MAAM,WAAW,GAAG,UAAU,GAAG,SAAS,GAAG,OAAO,GAAG,MAAM,GAAG,MAAM,CAAC;AAE7E,MAAM,WAAW,YAAY;IAC3B,QAAQ,EAAE,MAAM,CAAC;IACjB,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,MAAM,CAAC;IAClB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,IAAI,EAAE,MAAM,EAAE,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,6EAA6E;IAC7E,QAAQ,EAAE,OAAO,CAAC;CACnB;AAID,4EAA4E;AAC5E,wBAAgB,YAAY,CAAC,IAAI,EAAE;IACjC,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,UAAU,CAAC,EAAE,iBAAiB,CAAC;CAChC,GAAG,MAAM,GAAG,SAAS,CAQrB;AAED,qEAAqE;AACrE,wBAAgB,kBAAkB,CAAC,IAAI,EAAE;IACvC,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,UAAU,CAAC,EAAE,iBAAiB,CAAC;CAChC,GAAG,MAAM,CAET;AAUD;;;;GAIG;AACH,wBAAgB,SAAS,CAAC,KAAK,EAAE,MAAM,GAAG,SAAS,EAAE,GAAG,SAAa,GAAG,MAAM,GAAG,IAAI,CAOpF;AAkBD,iDAAiD;AACjD,wBAAgB,oBAAoB,CAClC,IAAI,EAAE,YAAY,EAClB,MAAM,GAAE,aAAkB,EAC1B,GAAG,SAAa,GACf,OAAO,CAqCT;AAED,wBAAgB,eAAe,CAC7B,IAAI,EAAE,YAAY,EAAE,EACpB,MAAM,GAAE,aAAkB,EAC1B,GAAG,SAAa,GACf,YAAY,EAAE,CAEhB;AAED,wBAAgB,aAAa,CAC3B,IAAI,EAAE,YAAY,EAAE,EACpB,IAAI,GAAE,WAAwB,EAC9B,KAAK,GAAE,KAAK,GAAG,MAAe,GAC7B,YAAY,EAAE,CAsBhB;AAED,uEAAuE;AACvE,wBAAgB,YAAY,CAAC,IAAI,EAAE,YAAY,GAAG,YAAY,CAmB7D;AAED,MAAM,WAAW,kBAAmB,SAAQ,aAAa;IACvD,IAAI,CAAC,EAAE,WAAW,CAAC;IACnB,KAAK,CAAC,EAAE,KAAK,GAAG,MAAM,CAAC;IACvB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,GAAG,CAAC,EAAE,MAAM,CAAC;CACd;AAED,6EAA6E;AAC7E,wBAAgB,WAAW,CAAC,IAAI,EAAE,YAAY,EAAE,EAAE,IAAI,GAAE,kBAAuB,GAAG,YAAY,EAAE,CAM/F;AAED,kEAAkE;AAClE,wBAAgB,aAAa,CAAC,IAAI,EAAE,YAAY,EAAE,GAAG;IACnD,KAAK,EAAE,KAAK,CAAC;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IAC/C,IAAI,EAAE,KAAK,CAAC;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IAC9C,WAAW,EAAE,KAAK,CAAC;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IACrD,OAAO,EAAE,KAAK,CAAC;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;CAClD,CAiBA"}

package/dist/src/knowledge/library.js

@@ -0,0 +1,195 @@
+/**
+ * The capture library (RCL-09) — browsing what has been captured, rather than
+ * searching it.
+ *
+ * `listDocuments` answers "what is indexed". A library answers a different
+ * question: everything from this site, captured last week, tagged `reading`,
+ * newest first. The facets come from the capture envelope's `meta.json`
+ * (RCL-07), which is already on every record — this module is the query layer
+ * over them, and the data layer the desktop GUI renders.
+ *
+ * Rules, once, here:
+ *  - within a facet values are OR'd, across facets AND'd (`--site a --site b
+ *    --tag x` = "from a or b, tagged x");
+ *  - `--site example.com` matches `www.example.com` and `docs.example.com`,
+ *    because nobody browsing their own library means the bare host only;
+ *  - a document with no provenance (an ordinary file on disk) sorts by its
+ *    index time and matches no capture facet — it is not a capture.
+ *
+ * @module v1/cli/knowledge/library
+ */
+import * as path from 'node:path';
+// ── Facet extraction ───────────────────────────────────────────────
+/** The host a document was captured from, or undefined for a local file. */
+export function documentSite(meta) {
+    const url = meta.canonicalUrl ?? meta.provenance?.canonicalUrl ?? meta.provenance?.url;
+    if (!url)
+        return undefined;
+    try {
+        return new URL(url).hostname.toLowerCase().replace(/^www\./, '');
+    }
+    catch {
+        return undefined;
+    }
+}
+/** When the page was captured; falls back to when it was indexed. */
+export function documentCapturedAt(meta) {
+    return meta.provenance?.capturedAt ?? meta.indexedAt ?? '';
+}
+const RELATIVE_RE = /^(\d+(?:\.\d+)?)\s*(m|h|d|w)$/i;
+const UNIT_MS = {
+    m: 60_000,
+    h: 3_600_000,
+    d: 86_400_000,
+    w: 604_800_000,
+};
+/**
+ * A point in time from `2026-09-01`, a full ISO stamp, or an AGE like `7d`
+ * (meaning "7 days ago"). Returns null for anything unparseable so a typo
+ * cannot silently filter the whole library away.
+ */
+export function parseWhen(value, now = Date.now()) {
+    const raw = (value ?? '').trim();
+    if (!raw)
+        return null;
+    const rel = RELATIVE_RE.exec(raw);
+    if (rel)
+        return now - Number(rel[1]) * UNIT_MS[rel[2].toLowerCase()];
+    const parsed = Date.parse(raw.length === 10 ? `${raw}T00:00:00.000Z` : raw);
+    return Number.isNaN(parsed) ? null : parsed;
+}
+function siteMatches(host, wanted) {
+    if (!host)
+        return false;
+    return wanted.some((w) => {
+        const want = w
+            .trim()
+            .toLowerCase()
+            .replace(/^www\./, '');
+        return !!want && (host === want || host.endsWith(`.${want}`));
+    });
+}
+function anyOf(values, wanted) {
+    const have = (values ?? []).map((v) => v.toLowerCase());
+    return wanted.some((w) => have.includes(w.trim().toLowerCase()));
+}
+/** True when a document belongs in this view. */
+export function matchesLibraryFilter(meta, filter = {}, now = Date.now()) {
+    if (filter.scope && meta.scope !== filter.scope)
+        return false;
+    const prov = meta.provenance;
+    const isCapture = !!prov || !!meta.canonicalUrl;
+    if (filter.capturedOnly && !isCapture)
+        return false;
+    if (filter.site?.length && !siteMatches(documentSite(meta), filter.site))
+        return false;
+    if (filter.tag?.length && !anyOf(prov?.tags, filter.tag))
+        return false;
+    if (filter.collection?.length &&
+        !anyOf(prov?.collection ? [prov.collection] : [], filter.collection))
+        return false;
+    if (filter.source?.length && !anyOf(prov?.source ? [prov.source] : [], filter.source))
+        return false;
+    if (filter.since || filter.until) {
+        const at = Date.parse(documentCapturedAt(meta));
+        const since = parseWhen(filter.since, now);
+        const until = parseWhen(filter.until, now);
+        if (Number.isNaN(at))
+            return false;
+        if (since !== null && at < since)
+            return false;
+        if (until !== null && at > until)
+            return false;
+    }
+    if (filter.text) {
+        const needle = filter.text.toLowerCase();
+        const hay = [
+            prov?.title ?? '',
+            meta.canonicalUrl ?? prov?.url ?? '',
+            meta.filePath,
+            prov?.note ?? '',
+        ]
+            .join(' ')
+            .toLowerCase();
+        if (!hay.includes(needle))
+            return false;
+    }
+    return true;
+}
+export function filterDocuments(docs, filter = {}, now = Date.now()) {
+    return docs.filter((d) => matchesLibraryFilter(d, filter, now));
+}
+export function sortDocuments(docs, sort = 'captured', order = 'desc') {
+    const dir = order === 'asc' ? 1 : -1;
+    const key = (d) => {
+        switch (sort) {
+            case 'indexed':
+                return d.indexedAt ?? '';
+            case 'title':
+                return (d.provenance?.title ?? path.basename(d.filePath)).toLowerCase();
+            case 'path':
+                return d.filePath;
+            case 'size':
+                return d.size ?? 0;
+            default:
+                return documentCapturedAt(d);
+        }
+    };
+    return [...docs].sort((a, b) => {
+        const ka = key(a);
+        const kb = key(b);
+        if (ka === kb)
+            return a.filePath.localeCompare(b.filePath);
+        return ka < kb ? -dir : dir;
+    });
+}
+/** The flat, JSON-friendly row the GUI and `--json` consumers read. */
+export function libraryEntry(meta) {
+    const prov = meta.provenance;
+    const url = meta.canonicalUrl ?? prov?.canonicalUrl ?? prov?.url;
+    return {
+        filePath: meta.filePath,
+        scope: meta.scope,
+        title: prov?.title ?? path.basename(meta.filePath),
+        ...(url ? { url } : {}),
+        ...(documentSite(meta) ? { site: documentSite(meta) } : {}),
+        capturedAt: documentCapturedAt(meta),
+        indexedAt: meta.indexedAt,
+        ...(prov?.source ? { source: prov.source } : {}),
+        ...(prov?.collection ? { collection: prov.collection } : {}),
+        tags: prov?.tags ?? [],
+        chunkCount: meta.chunkCount,
+        size: meta.size,
+        ...(meta.version ? { version: meta.version } : {}),
+        captured: !!prov || !!meta.canonicalUrl,
+    };
+}
+/** Browse the library: filter, sort, cap. Never throws on an empty store. */
+export function listLibrary(docs, opts = {}) {
+    const now = opts.now ?? Date.now();
+    const filtered = filterDocuments(docs, opts, now);
+    const sorted = sortDocuments(filtered, opts.sort ?? 'captured', opts.order ?? 'desc');
+    const capped = opts.limit && opts.limit > 0 ? sorted.slice(0, opts.limit) : sorted;
+    return capped.map(libraryEntry);
+}
+/** Facet counts for the current view — what a sidebar renders. */
+export function libraryFacets(docs) {
+    const tally = (values) => {
+        const counts = new Map();
+        for (const v of values) {
+            if (!v)
+                continue;
+            counts.set(v, (counts.get(v) ?? 0) + 1);
+        }
+        return [...counts.entries()]
+            .map(([value, count]) => ({ value, count }))
+            .sort((a, b) => b.count - a.count || a.value.localeCompare(b.value));
+    };
+    return {
+        sites: tally(docs.map(documentSite)),
+        tags: tally(docs.flatMap((d) => d.provenance?.tags ?? [])),
+        collections: tally(docs.map((d) => d.provenance?.collection)),
+        sources: tally(docs.map((d) => d.provenance?.source)),
+    };
+}
+//# sourceMappingURL=library.js.map

package/dist/src/knowledge/library.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"library.js","sourceRoot":"","sources":["../../../src/knowledge/library.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;GAmBG;AAEH,OAAO,KAAK,IAAI,MAAM,WAAW,CAAC;AAyClC,sEAAsE;AAEtE,4EAA4E;AAC5E,MAAM,UAAU,YAAY,CAAC,IAG5B;IACC,MAAM,GAAG,GAAG,IAAI,CAAC,YAAY,IAAI,IAAI,CAAC,UAAU,EAAE,YAAY,IAAI,IAAI,CAAC,UAAU,EAAE,GAAG,CAAC;IACvF,IAAI,CAAC,GAAG;QAAE,OAAO,SAAS,CAAC;IAC3B,IAAI,CAAC;QACH,OAAO,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAC;IACnE,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,SAAS,CAAC;IACnB,CAAC;AACH,CAAC;AAED,qEAAqE;AACrE,MAAM,UAAU,kBAAkB,CAAC,IAGlC;IACC,OAAO,IAAI,CAAC,UAAU,EAAE,UAAU,IAAI,IAAI,CAAC,SAAS,IAAI,EAAE,CAAC;AAC7D,CAAC;AAED,MAAM,WAAW,GAAG,gCAAgC,CAAC;AACrD,MAAM,OAAO,GAA2B;IACtC,CAAC,EAAE,MAAM;IACT,CAAC,EAAE,SAAS;IACZ,CAAC,EAAE,UAAU;IACb,CAAC,EAAE,WAAW;CACf,CAAC;AAEF;;;;GAIG;AACH,MAAM,UAAU,SAAS,CAAC,KAAyB,EAAE,GAAG,GAAG,IAAI,CAAC,GAAG,EAAE;IACnE,MAAM,GAAG,GAAG,CAAC,KAAK,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;IACjC,IAAI,CAAC,GAAG;QAAE,OAAO,IAAI,CAAC;IACtB,MAAM,GAAG,GAAG,WAAW,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAClC,IAAI,GAAG;QAAE,OAAO,GAAG,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,CAAC;IACrE,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,KAAK,EAAE,CAAC,CAAC,CAAC,GAAG,GAAG,gBAAgB,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;IAC5E,OAAO,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,MAAM,CAAC;AAC9C,CAAC;AAED,SAAS,WAAW,CAAC,IAAwB,EAAE,MAAgB;IAC7D,IAAI,CAAC,IAAI;QAAE,OAAO,KAAK,CAAC;IACxB,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE;QACvB,MAAM,IAAI,GAAG,CAAC;aACX,IAAI,EAAE;aACN,WAAW,EAAE;aACb,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAC;QACzB,OAAO,CAAC,CAAC,IAAI,IAAI,CAAC,IAAI,KAAK,IAAI,IAAI,IAAI,CAAC,QAAQ,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,CAAC;IAChE,CAAC,CAAC,CAAC;AACL,CAAC;AAED,SAAS,KAAK,CAAC,MAA4B,EAAE,MAAgB;IAC3D,MAAM,IAAI,GAAG,CAAC,MAAM,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,CAAC;IACxD,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC,CAAC,CAAC;AACnE,CAAC;AAED,iDAAiD;AACjD,MAAM,UAAU,oBAAoB,CAClC,IAAkB,EAClB,MAAM,GAAkB,EAAE,EAC1B,GAAG,GAAG,IAAI,CAAC,GAAG,EAAE;IAEhB,IAAI,MAAM,CAAC,KAAK,IAAI,IAAI,CAAC,KAAK,KAAK,MAAM,CAAC,KAAK;QAAE,OAAO,KAAK,CAAC;IAC9D,MAAM,IAAI,GAAG,IAAI,CAAC,UAAU,CAAC;IAC7B,MAAM,SAAS,GAAG,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,IAAI,CAAC,YAAY,CAAC;IAChD,IAAI,MAAM,CAAC,YAAY,IAAI,CAAC,SAAS;QAAE,OAAO,KAAK,CAAC;IACpD,IAAI,MAAM,CAAC,IAAI,EAAE,MAAM,IAAI,CAAC,WAAW,CAAC,YAAY,CAAC,IAAI,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC;QAAE,OAAO,KAAK,CAAC;IACvF,IAAI,MAAM,CAAC,GAAG,EAAE,MAAM,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE,IAAI,EAAE,MAAM,CAAC,GAAG,CAAC;QAAE,OAAO,KAAK,CAAC;IACvE,IACE,MAAM,CAAC,UAAU,EAAE,MAAM;QACzB,CAAC,KAAK,CAAC,IAAI,EAAE,UAAU,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,EAAE,EAAE,MAAM,CAAC,UAAU,CAAC;QAEpE,OAAO,KAAK,CAAC;IACf,IAAI,MAAM,CAAC,MAAM,EAAE,MAAM,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,EAAE,EAAE,MAAM,CAAC,MAAM,CAAC;QACnF,OAAO,KAAK,CAAC;IAEf,IAAI,MAAM,CAAC,KAAK,IAAI,MAAM,CAAC,KAAK,EAAE,CAAC;QACjC,MAAM,EAAE,GAAG,IAAI,CAAC,KAAK,CAAC,kBAAkB,CAAC,IAAI,CAAC,CAAC,CAAC;QAChD,MAAM,KAAK,GAAG,SAAS,CAAC,MAAM,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;QAC3C,MAAM,KAAK,GAAG,SAAS,CAAC,MAAM,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;QAC3C,IAAI,MAAM,CAAC,KAAK,CAAC,EAAE,CAAC;YAAE,OAAO,KAAK,CAAC;QACnC,IAAI,KAAK,KAAK,IAAI,IAAI,EAAE,GAAG,KAAK;YAAE,OAAO,KAAK,CAAC;QAC/C,IAAI,KAAK,KAAK,IAAI,IAAI,EAAE,GAAG,KAAK;YAAE,OAAO,KAAK,CAAC;IACjD,CAAC;IAED,IAAI,MAAM,CAAC,IAAI,EAAE,CAAC;QAChB,MAAM,MAAM,GAAG,MAAM,CAAC,IAAI,CAAC,WAAW,EAAE,CAAC;QACzC,MAAM,GAAG,GAAG;YACV,IAAI,EAAE,KAAK,IAAI,EAAE;YACjB,IAAI,CAAC,YAAY,IAAI,IAAI,EAAE,GAAG,IAAI,EAAE;YACpC,IAAI,CAAC,QAAQ;YACb,IAAI,EAAE,IAAI,IAAI,EAAE;SACjB;aACE,IAAI,CAAC,GAAG,CAAC;aACT,WAAW,EAAE,CAAC;QACjB,IAAI,CAAC,GAAG,CAAC,QAAQ,CAAC,MAAM,CAAC;YAAE,OAAO,KAAK,CAAC;IAC1C,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,MAAM,UAAU,eAAe,CAC7B,IAAoB,EACpB,MAAM,GAAkB,EAAE,EAC1B,GAAG,GAAG,IAAI,CAAC,GAAG,EAAE;IAEhB,OAAO,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,oBAAoB,CAAC,CAAC,EAAE,MAAM,EAAE,GAAG,CAAC,CAAC,CAAC;AAClE,CAAC;AAED,MAAM,UAAU,aAAa,CAC3B,IAAoB,EACpB,IAAI,GAAgB,UAAU,EAC9B,KAAK,GAAmB,MAAM;IAE9B,MAAM,GAAG,GAAG,KAAK,KAAK,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACrC,MAAM,GAAG,GAAG,CAAC,CAAe,EAAmB,EAAE;QAC/C,QAAQ,IAAI,EAAE,CAAC;YACb,KAAK,SAAS;gBACZ,OAAO,CAAC,CAAC,SAAS,IAAI,EAAE,CAAC;YAC3B,KAAK,OAAO;gBACV,OAAO,CAAC,CAAC,CAAC,UAAU,EAAE,KAAK,IAAI,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC;YAC1E,KAAK,MAAM;gBACT,OAAO,CAAC,CAAC,QAAQ,CAAC;YACpB,KAAK,MAAM;gBACT,OAAO,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;YACrB;gBACE,OAAO,kBAAkB,CAAC,CAAC,CAAC,CAAC;QACjC,CAAC;IACH,CAAC,CAAC;IACF,OAAO,CAAC,GAAG,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;QAC7B,MAAM,EAAE,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC;QAClB,MAAM,EAAE,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC;QAClB,IAAI,EAAE,KAAK,EAAE;YAAE,OAAO,CAAC,CAAC,QAAQ,CAAC,aAAa,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC;QAC3D,OAAO,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC;IAC9B,CAAC,CAAC,CAAC;AACL,CAAC;AAED,uEAAuE;AACvE,MAAM,UAAU,YAAY,CAAC,IAAkB;IAC7C,MAAM,IAAI,GAAG,IAAI,CAAC,UAAU,CAAC;IAC7B,MAAM,GAAG,GAAG,IAAI,CAAC,YAAY,IAAI,IAAI,EAAE,YAAY,IAAI,IAAI,EAAE,GAAG,CAAC;IACjE,OAAO;QACL,QAAQ,EAAE,IAAI,CAAC,QAAQ;QACvB,KAAK,EAAE,IAAI,CAAC,KAAK;QACjB,KAAK,EAAE,IAAI,EAAE,KAAK,IAAI,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC;QAClD,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACvB,GAAG,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,YAAY,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC3D,UAAU,EAAE,kBAAkB,CAAC,IAAI,CAAC;QACpC,SAAS,EAAE,IAAI,CAAC,SAAS;QACzB,GAAG,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAChD,GAAG,CAAC,IAAI,EAAE,UAAU,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,IAAI,CAAC,UAAU,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC5D,IAAI,EAAE,IAAI,EAAE,IAAI,IAAI,EAAE;QACtB,UAAU,EAAE,IAAI,CAAC,UAAU;QAC3B,IAAI,EAAE,IAAI,CAAC,IAAI;QACf,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAClD,QAAQ,EAAE,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,IAAI,CAAC,YAAY;KACxC,CAAC;AACJ,CAAC;AASD,6EAA6E;AAC7E,MAAM,UAAU,WAAW,CAAC,IAAoB,EAAE,IAAI,GAAuB,EAAE;IAC7E,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,IAAI,IAAI,CAAC,GAAG,EAAE,CAAC;IACnC,MAAM,QAAQ,GAAG,eAAe,CAAC,IAAI,EAAE,IAAI,EAAE,GAAG,CAAC,CAAC;IAClD,MAAM,MAAM,GAAG,aAAa,CAAC,QAAQ,EAAE,IAAI,CAAC,IAAI,IAAI,UAAU,EAAE,IAAI,CAAC,KAAK,IAAI,MAAM,CAAC,CAAC;IACtF,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,IAAI,IAAI,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;IACnF,OAAO,MAAM,CAAC,GAAG,CAAC,YAAY,CAAC,CAAC;AAClC,CAAC;AAED,kEAAkE;AAClE,MAAM,UAAU,aAAa,CAAC,IAAoB;IAMhD,MAAM,KAAK,GAAG,CAAC,MAAiC,EAAE,EAAE;QAClD,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;QACzC,KAAK,MAAM,CAAC,IAAI,MAAM,EAAE,CAAC;YACvB,IAAI,CAAC,CAAC;gBAAE,SAAS;YACjB,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QAC1C,CAAC;QACD,OAAO,CAAC,GAAG,MAAM,CAAC,OAAO,EAAE,CAAC;aACzB,GAAG,CAAC,CAAC,CAAC,KAAK,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,CAAC,CAAC;aAC3C,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,IAAI,CAAC,CAAC,KAAK,CAAC,aAAa,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC;IACzE,CAAC,CAAC;IACF,OAAO;QACL,KAAK,EAAE,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,YAAY,CAAC,CAAC;QACpC,IAAI,EAAE,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,UAAU,EAAE,IAAI,IAAI,EAAE,CAAC,CAAC;QAC1D,WAAW,EAAE,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,UAAU,EAAE,UAAU,CAAC,CAAC;QAC7D,OAAO,EAAE,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,UAAU,EAAE,MAAM,CAAC,CAAC;KACtD,CAAC;AACJ,CAAC"}

package/dist/src/knowledge/lookup.d.ts

@@ -0,0 +1,80 @@
+/**
+ * "Have I already saved this?" (RCL-02) — one URL in, one answer out.
+ *
+ * The browser extension asks this on every navigation. That is the only
+ * thing that shapes this module:
+ *
+ *  - it NEVER throws. An empty store, a store that has never been written,
+ *    a URL that is not a URL — all of them are `{saved: false}`, because a
+ *    page load must not be able to fail on a badge;
+ *  - it is a pure read. No embedder, no subprocess, no index write. The
+ *    answer comes off the metadata log, which is a file read;
+ *  - it answers the WHOLE question in one call. `doc list` can already say
+ *    that a URL is present, but not what note was written at save time —
+ *    and "you saved this, and here is why you said you were saving it" is
+ *    the entire value of the badge over a bookmark.
+ *
+ * Identity is `captureIdentityUrl`'s rule and nothing else: canonical URL
+ * when the capturer resolved one, fragment stripped, nothing else rewritten.
+ * Matching more loosely than the ingest side would make the badge claim
+ * pages are saved that cannot actually be found again.
+ *
+ * @module v1/cli/knowledge/lookup
+ */
+/** What the badge and the "already saved" panel render. */
+export interface UrlLookup {
+    saved: boolean;
+    /** The identity URL the lookup was performed against. */
+    url: string;
+    title?: string;
+    site?: string;
+    /** When the page was captured (not when it was indexed). */
+    capturedAt?: string;
+    /** RCL-06: 1 for a first save, +1 per stored revision. 0 when unsaved. */
+    versions: number;
+    /** The note written at save time (CLIP-07) — the reason it was kept. */
+    note?: string;
+    tags: string[];
+    collection?: string;
+    source?: string;
+    /** The indexed file (`…/readable.md`). */
+    filePath?: string;
+    /** The capture envelope directory holding it — what "open the archived
+     *  copy" opens. Absent when the document is an ordinary file on disk. */
+    envelope?: string;
+    scope?: string;
+    contentHash?: string;
+    /** RCL-04: how many highlight notes were saved on this page. */
+    highlights?: number;
+}
+export interface LookupOptions {
+    rootDir?: string;
+    scope?: string;
+    /** Skip the `highlights.json` read. The badge path does not need it. */
+    withHighlights?: boolean;
+}
+/**
+ * identityUrl mirrors `captureIdentityUrl` for a raw URL rather than a
+ * provenance record: drop the fragment, trim, rewrite nothing else.
+ *
+ * Duplicated deliberately rather than reshaping the URL into a fake
+ * `CaptureProvenance` to borrow that function — the two callers mean
+ * different things ("the URL a tab is on" vs. "what this capture is of")
+ * and only happen to agree on the rule.
+ */
+export declare function identityUrl(raw: string | undefined): string;
+/** The envelope directory for an indexed file, or undefined when the file
+ *  is not inside one. */
+export declare function envelopeFor(filePath: string | undefined): string | undefined;
+/**
+ * lookupUrl answers whether this URL is in the store, and what is known
+ * about it. `rootDir` is the store to look in; pass the global brain's root
+ * to ask the personal store.
+ */
+export declare function lookupUrl(rawUrl: string, opts?: LookupOptions): UrlLookup;
+/**
+ * lookupUrls answers a batch in one pass over the metadata, for a caller
+ * with several tabs to ask about. Same rules, same guarantees.
+ */
+export declare function lookupUrls(urls: readonly string[], opts?: LookupOptions): UrlLookup[];
+//# sourceMappingURL=lookup.d.ts.map

package/dist/src/knowledge/lookup.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"lookup.d.ts","sourceRoot":"","sources":["../../../src/knowledge/lookup.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;GAsBG;AAaH,2DAA2D;AAC3D,MAAM,WAAW,SAAS;IACxB,KAAK,EAAE,OAAO,CAAC;IACf,yDAAyD;IACzD,GAAG,EAAE,MAAM,CAAC;IACZ,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,4DAA4D;IAC5D,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,0EAA0E;IAC1E,QAAQ,EAAE,MAAM,CAAC;IACjB,wEAAwE;IACxE,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,EAAE,CAAC;IACf,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,0CAA0C;IAC1C,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB;6EACyE;IACzE,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,gEAAgE;IAChE,UAAU,CAAC,EAAE,MAAM,CAAC;CACrB;AAED,MAAM,WAAW,aAAa;IAC5B,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,wEAAwE;IACxE,cAAc,CAAC,EAAE,OAAO,CAAC;CAC1B;AAID;;;;;;;;GAQG;AACH,wBAAgB,WAAW,CAAC,GAAG,EAAE,MAAM,GAAG,SAAS,GAAG,MAAM,CAI3D;AAwBD;yBACyB;AACzB,wBAAgB,WAAW,CAAC,QAAQ,EAAE,MAAM,GAAG,SAAS,GAAG,MAAM,GAAG,SAAS,CAQ5E;AAMD;;;;GAIG;AACH,wBAAgB,SAAS,CAAC,MAAM,EAAE,MAAM,EAAE,IAAI,GAAE,aAAkB,GAAG,SAAS,CA8D7E;AAED;;;GAGG;AACH,wBAAgB,UAAU,CAAC,IAAI,EAAE,SAAS,MAAM,EAAE,EAAE,IAAI,GAAE,aAAkB,GAAG,SAAS,EAAE,CAEzF"}

package/dist/src/knowledge/lookup.js

@@ -0,0 +1,156 @@
+/**
+ * "Have I already saved this?" (RCL-02) — one URL in, one answer out.
+ *
+ * The browser extension asks this on every navigation. That is the only
+ * thing that shapes this module:
+ *
+ *  - it NEVER throws. An empty store, a store that has never been written,
+ *    a URL that is not a URL — all of them are `{saved: false}`, because a
+ *    page load must not be able to fail on a badge;
+ *  - it is a pure read. No embedder, no subprocess, no index write. The
+ *    answer comes off the metadata log, which is a file read;
+ *  - it answers the WHOLE question in one call. `doc list` can already say
+ *    that a URL is present, but not what note was written at save time —
+ *    and "you saved this, and here is why you said you were saving it" is
+ *    the entire value of the badge over a bookmark.
+ *
+ * Identity is `captureIdentityUrl`'s rule and nothing else: canonical URL
+ * when the capturer resolved one, fragment stripped, nothing else rewritten.
+ * Matching more loosely than the ingest side would make the badge claim
+ * pages are saved that cannot actually be found again.
+ *
+ * @module v1/cli/knowledge/lookup
+ */
+import * as fs from 'node:fs';
+import * as path from 'node:path';
+import { isCaptureEnvelope } from './capture-envelope.js';
+import { getKnowledgeRoot, listDocuments, listDocumentVersions, } from './document-pipeline.js';
+import { documentCapturedAt, documentSite } from './library.js';
+const EMPTY = (url) => ({ saved: false, url, versions: 0, tags: [] });
+/**
+ * identityUrl mirrors `captureIdentityUrl` for a raw URL rather than a
+ * provenance record: drop the fragment, trim, rewrite nothing else.
+ *
+ * Duplicated deliberately rather than reshaping the URL into a fake
+ * `CaptureProvenance` to borrow that function — the two callers mean
+ * different things ("the URL a tab is on" vs. "what this capture is of")
+ * and only happen to agree on the rule.
+ */
+export function identityUrl(raw) {
+    const s = String(raw ?? '').trim();
+    const hash = s.indexOf('#');
+    return (hash === -1 ? s : s.slice(0, hash)).trim();
+}
+/** The identity URL a record is addressable by, or undefined. */
+function recordUrl(meta) {
+    const url = meta.canonicalUrl ?? meta.provenance?.canonicalUrl ?? meta.provenance?.url;
+    return url ? identityUrl(url) : undefined;
+}
+/**
+ * countHighlights reads the `highlights.json` the extension wrote beside
+ * the document (RCL-04). Silent on every failure: the count is decoration
+ * on an answer that is already complete without it.
+ */
+function countHighlights(envelope) {
+    try {
+        const file = path.join(envelope, 'highlights.json');
+        if (!fs.existsSync(file))
+            return 0;
+        const parsed = JSON.parse(fs.readFileSync(file, 'utf-8'));
+        return Array.isArray(parsed?.highlights) ? parsed.highlights.length : 0;
+    }
+    catch {
+        return 0;
+    }
+}
+/** The envelope directory for an indexed file, or undefined when the file
+ *  is not inside one. */
+export function envelopeFor(filePath) {
+    if (!filePath)
+        return undefined;
+    const dir = path.dirname(path.resolve(filePath));
+    try {
+        return isCaptureEnvelope(dir) ? dir : undefined;
+    }
+    catch {
+        return undefined;
+    }
+}
+function provenanceOf(meta) {
+    return meta.provenance ?? {};
+}
+/**
+ * lookupUrl answers whether this URL is in the store, and what is known
+ * about it. `rootDir` is the store to look in; pass the global brain's root
+ * to ask the personal store.
+ */
+export function lookupUrl(rawUrl, opts = {}) {
+    const url = identityUrl(rawUrl);
+    if (!url)
+        return EMPTY('');
+    // The SAME routing the ingest used: scope `global` lives in the personal
+    // cross-project brain, not under `rootDir`. Reading the wrong store would
+    // report every globally-captured page as unsaved.
+    let root;
+    let docs;
+    try {
+        root = opts.scope ? getKnowledgeRoot(opts.scope, opts.rootDir) : opts.rootDir;
+        docs = listDocuments(root, opts.scope);
+    }
+    catch {
+        // A store that has never been written, or one this process cannot
+        // read. Neither is an error worth failing a page load over.
+        return EMPTY(url);
+    }
+    const matches = docs.filter((d) => recordUrl(d) === url);
+    if (!matches.length)
+        return EMPTY(url);
+    // The live record with the highest version. A re-captured page lands at
+    // a new path, so several records can match and only the newest is the
+    // one the badge should describe.
+    const best = matches.reduce((a, b) => ((b.version ?? 0) > (a.version ?? 0) ? b : a));
+    const prov = provenanceOf(best);
+    // The version COUNT is the larger of the record's own number and what
+    // the append-only log still holds: the log is compacted past 1MB, and
+    // the `version` field is the durable count. Reading only the log would
+    // under-report a long-lived document.
+    let logged = 0;
+    try {
+        logged = listDocumentVersions(root, best.canonicalUrl ?? url, opts.scope).length;
+    }
+    catch {
+        logged = 0;
+    }
+    const versions = Math.max(best.version ?? 1, logged, 1);
+    const envelope = envelopeFor(best.filePath);
+    const answer = {
+        saved: true,
+        url,
+        versions,
+        tags: prov.tags ?? [],
+        ...(prov.title ? { title: prov.title } : {}),
+        ...(documentSite(best) ? { site: documentSite(best) } : {}),
+        ...(documentCapturedAt(best) ? { capturedAt: documentCapturedAt(best) } : {}),
+        ...(prov.note ? { note: prov.note } : {}),
+        ...(prov.collection ? { collection: prov.collection } : {}),
+        ...(prov.source ? { source: prov.source } : {}),
+        filePath: best.filePath,
+        ...(envelope ? { envelope } : {}),
+        scope: best.scope,
+        ...(best.contentHash ? { contentHash: best.contentHash } : {}),
+    };
+    if (opts.withHighlights && envelope) {
+        const count = countHighlights(envelope);
+        if (count)
+            answer.highlights = count;
+    }
+    return answer;
+}
+/**
+ * lookupUrls answers a batch in one pass over the metadata, for a caller
+ * with several tabs to ask about. Same rules, same guarantees.
+ */
+export function lookupUrls(urls, opts = {}) {
+    return urls.map((u) => lookupUrl(u, opts));
+}
+//# sourceMappingURL=lookup.js.map

package/dist/src/knowledge/lookup.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"lookup.js","sourceRoot":"","sources":["../../../src/knowledge/lookup.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;GAsBG;AAEH,OAAO,KAAK,EAAE,MAAM,SAAS,CAAC;AAC9B,OAAO,KAAK,IAAI,MAAM,WAAW,CAAC;AAClC,OAAO,EAA0B,iBAAiB,EAAE,MAAM,uBAAuB,CAAC;AAClF,OAAO,EAEL,gBAAgB,EAChB,aAAa,EACb,oBAAoB,GACrB,MAAM,wBAAwB,CAAC;AAChC,OAAO,EAAE,kBAAkB,EAAE,YAAY,EAAE,MAAM,cAAc,CAAC;AAoChE,MAAM,KAAK,GAAG,CAAC,GAAW,EAAa,EAAE,CAAC,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,GAAG,EAAE,QAAQ,EAAE,CAAC,EAAE,IAAI,EAAE,EAAE,EAAE,CAAC,CAAC;AAEzF;;;;;;;;GAQG;AACH,MAAM,UAAU,WAAW,CAAC,GAAuB;IACjD,MAAM,CAAC,GAAG,MAAM,CAAC,GAAG,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;IACnC,MAAM,IAAI,GAAG,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC;IAC5B,OAAO,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;AACrD,CAAC;AAED,iEAAiE;AACjE,SAAS,SAAS,CAAC,IAAkB;IACnC,MAAM,GAAG,GAAG,IAAI,CAAC,YAAY,IAAI,IAAI,CAAC,UAAU,EAAE,YAAY,IAAI,IAAI,CAAC,UAAU,EAAE,GAAG,CAAC;IACvF,OAAO,GAAG,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;AAC5C,CAAC;AAED;;;;GAIG;AACH,SAAS,eAAe,CAAC,QAAgB;IACvC,IAAI,CAAC;QACH,MAAM,IAAI,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,iBAAiB,CAAC,CAAC;QACpD,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,IAAI,CAAC;YAAE,OAAO,CAAC,CAAC;QACnC,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,YAAY,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC,CAAC;QAC1D,OAAO,KAAK,CAAC,OAAO,CAAC,MAAM,EAAE,UAAU,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC;IAC1E,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,CAAC,CAAC;IACX,CAAC;AACH,CAAC;AAED;yBACyB;AACzB,MAAM,UAAU,WAAW,CAAC,QAA4B;IACtD,IAAI,CAAC,QAAQ;QAAE,OAAO,SAAS,CAAC;IAChC,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,CAAC;IACjD,IAAI,CAAC;QACH,OAAO,iBAAiB,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC;IAClD,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,SAAS,CAAC;IACnB,CAAC;AACH,CAAC;AAED,SAAS,YAAY,CAAC,IAAkB;IACtC,OAAO,IAAI,CAAC,UAAU,IAAI,EAAE,CAAC;AAC/B,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,SAAS,CAAC,MAAc,EAAE,IAAI,GAAkB,EAAE;IAChE,MAAM,GAAG,GAAG,WAAW,CAAC,MAAM,CAAC,CAAC;IAChC,IAAI,CAAC,GAAG;QAAE,OAAO,KAAK,CAAC,EAAE,CAAC,CAAC;IAE3B,yEAAyE;IACzE,0EAA0E;IAC1E,kDAAkD;IAClD,IAAI,IAAwB,CAAC;IAC7B,IAAI,IAAoB,CAAC;IACzB,IAAI,CAAC;QACH,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,gBAAgB,CAAC,IAAI,CAAC,KAAK,EAAE,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC;QAC9E,IAAI,GAAG,aAAa,CAAC,IAAI,EAAE,IAAI,CAAC,KAAK,CAAC,CAAC;IACzC,CAAC;IAAC,MAAM,CAAC;QACP,kEAAkE;QAClE,4DAA4D;QAC5D,OAAO,KAAK,CAAC,GAAG,CAAC,CAAC;IACpB,CAAC;IAED,MAAM,OAAO,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,SAAS,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC;IACzD,IAAI,CAAC,OAAO,CAAC,MAAM;QAAE,OAAO,KAAK,CAAC,GAAG,CAAC,CAAC;IAEvC,wEAAwE;IACxE,sEAAsE;IACtE,iCAAiC;IACjC,MAAM,IAAI,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACrF,MAAM,IAAI,GAAG,YAAY,CAAC,IAAI,CAAC,CAAC;IAEhC,sEAAsE;IACtE,sEAAsE;IACtE,uEAAuE;IACvE,sCAAsC;IACtC,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,IAAI,CAAC;QACH,MAAM,GAAG,oBAAoB,CAAC,IAAI,EAAE,IAAI,CAAC,YAAY,IAAI,GAAG,EAAE,IAAI,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC;IACnF,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,GAAG,CAAC,CAAC;IACb,CAAC;IACD,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC;IAExD,MAAM,QAAQ,GAAG,WAAW,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;IAC5C,MAAM,MAAM,GAAc;QACxB,KAAK,EAAE,IAAI;QACX,GAAG;QACH,QAAQ;QACR,IAAI,EAAE,IAAI,CAAC,IAAI,IAAI,EAAE;QACrB,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,IAAI,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC5C,GAAG,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,YAAY,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC3D,GAAG,CAAC,kBAAkB,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,kBAAkB,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC7E,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACzC,GAAG,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,IAAI,CAAC,UAAU,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC3D,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC/C,QAAQ,EAAE,IAAI,CAAC,QAAQ;QACvB,GAAG,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACjC,KAAK,EAAE,IAAI,CAAC,KAAK;QACjB,GAAG,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC,CAAC,EAAE,WAAW,EAAE,IAAI,CAAC,WAAW,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;KAC/D,CAAC;IAEF,IAAI,IAAI,CAAC,cAAc,IAAI,QAAQ,EAAE,CAAC;QACpC,MAAM,KAAK,GAAG,eAAe,CAAC,QAAQ,CAAC,CAAC;QACxC,IAAI,KAAK;YAAE,MAAM,CAAC,UAAU,GAAG,KAAK,CAAC;IACvC,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,UAAU,CAAC,IAAuB,EAAE,IAAI,GAAkB,EAAE;IAC1E,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,SAAS,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC;AAC7C,CAAC"}

package/dist/src/knowledge/mhtml.d.ts

@@ -0,0 +1,60 @@
+/**
+ * MHTML / MHT container parsing (RCL-01).
+ *
+ * An MHTML archive is an RFC 2557 `multipart/related` MIME message: the page
+ * plus every subresource it referenced, each part carrying its own transfer
+ * encoding and charset. Chrome's `Page.captureSnapshot` writes them
+ * quoted-printable; Internet Explorer and most "save as web archive" tools
+ * write base64.
+ *
+ * Only the ROOT frame is prose. Picking the first `text/html` part instead is
+ * how an ad iframe ends up indexed as the article, so the root is resolved
+ * from the container's `Snapshot-Content-Location` / `Content-Location` when
+ * one is present.
+ *
+ * Read the file as `latin1` (or pass a Buffer): byte fidelity matters before
+ * transfer decoding, and the real charset is only known per part.
+ *
+ * @module v1/cli/knowledge/mhtml
+ */
+export interface MhtmlPart {
+    /** Lowercased MIME type without parameters, e.g. `text/html`. */
+    contentType: string;
+    /** Charset from the part's Content-Type, lowercased, or null. */
+    charset: string | null;
+    /** Content-Location, or null. */
+    location: string | null;
+    /** Lowercased Content-Transfer-Encoding; `binary` when unstated. */
+    encoding: string;
+    /** Raw bytes after transfer decoding. */
+    body: Buffer;
+    headers: Record<string, string>;
+}
+export interface ParsedMhtml {
+    /** The part holding the page itself, or null when there is no HTML part. */
+    root: MhtmlPart | null;
+    parts: MhtmlPart[];
+    /** Container `Snapshot-Content-Location`/`Content-Location` header. */
+    snapshotLocation: string | null;
+    /** Container `Subject` header — Chrome writes the page title here. */
+    subject: string | null;
+}
+/** Decode a quoted-printable body to bytes: soft line breaks vanish, `=XX`
+ *  becomes one byte, everything else is its own latin1 byte. */
+export declare function decodeQuotedPrintable(input: string): Buffer;
+/** Decode bytes under a MIME charset, falling back to UTF-8 (then latin1)
+ *  rather than failing an ingest over an exotic label. */
+export declare function decodeCharset(bytes: Buffer, charset: string | null): string;
+/**
+ * Parse an MHTML container. Anything that is not a well-formed multipart
+ * message is treated as a single part, so a plain `.html` file renamed to
+ * `.mhtml` still ingests.
+ */
+export declare function parseMhtml(raw: string | Buffer): ParsedMhtml;
+/** Parse a container and return its root frame decoded to a string. */
+export declare function mhtmlToHtml(raw: string | Buffer): {
+    html: string;
+    location: string | null;
+    subject: string | null;
+};
+//# sourceMappingURL=mhtml.d.ts.map

package/dist/src/knowledge/mhtml.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"mhtml.d.ts","sourceRoot":"","sources":["../../../src/knowledge/mhtml.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;GAkBG;AAEH,MAAM,WAAW,SAAS;IACxB,iEAAiE;IACjE,WAAW,EAAE,MAAM,CAAC;IACpB,iEAAiE;IACjE,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;IACvB,iCAAiC;IACjC,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;IACxB,oEAAoE;IACpE,QAAQ,EAAE,MAAM,CAAC;IACjB,yCAAyC;IACzC,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;CACjC;AAED,MAAM,WAAW,WAAW;IAC1B,4EAA4E;IAC5E,IAAI,EAAE,SAAS,GAAG,IAAI,CAAC;IACvB,KAAK,EAAE,SAAS,EAAE,CAAC;IACnB,uEAAuE;IACvE,gBAAgB,EAAE,MAAM,GAAG,IAAI,CAAC;IAChC,sEAAsE;IACtE,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;CACxB;AAED;gEACgE;AAChE,wBAAgB,qBAAqB,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,CAyB3D;AAaD;0DAC0D;AAC1D,wBAAgB,aAAa,CAAC,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,IAAI,GAAG,MAAM,CAU3E;AA4CD;;;;GAIG;AACH,wBAAgB,UAAU,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,WAAW,CAiC5D;AAED,uEAAuE;AACvE,wBAAgB,WAAW,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG;IACjD,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;IACxB,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;CACxB,CAOA"}