sxng-cli 1.0.5 → 1.0.7

package/dist/deep/extractor.d.ts

@@ -0,0 +1,29 @@
+/**
+ * Content extraction using linkedom + @mozilla/readability
+ */
+export interface ExtractorOptions {
+    timeoutMs?: number;
+    concurrency?: number;
+    maxResponseBytes?: number;
+}
+export interface ExtractedContent {
+    title: string;
+    content: string;
+    excerpt: string;
+    url: string;
+    byline?: string;
+    siteName?: string;
+    length: number;
+    extractedAt: number;
+    error?: string;
+}
+export declare class ContentExtractor {
+    private timeoutMs;
+    private concurrency;
+    private maxResponseBytes;
+    constructor(options?: ExtractorOptions);
+    extract(url: string): Promise<ExtractedContent>;
+    extractFromHtml(html: string, url: string): ExtractedContent;
+    extractBatch(urls: string[]): Promise<ExtractedContent[]>;
+}
+//# sourceMappingURL=extractor.d.ts.map

package/dist/deep/extractor.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"extractor.d.ts","sourceRoot":"","sources":["../../src/deep/extractor.ts"],"names":[],"mappings":"AAAA;;GAEG;AASH,MAAM,WAAW,gBAAgB;IAC7B,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,gBAAgB,CAAC,EAAE,MAAM,CAAC;CAC7B;AAED,MAAM,WAAW,gBAAgB;IAC7B,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,EAAE,MAAM,CAAC;IAChB,OAAO,EAAE,MAAM,CAAC;IAChB,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,MAAM,EAAE,MAAM,CAAC;IACf,WAAW,EAAE,MAAM,CAAC;IACpB,KAAK,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,qBAAa,gBAAgB;IACzB,OAAO,CAAC,SAAS,CAAS;IAC1B,OAAO,CAAC,WAAW,CAAS;IAC5B,OAAO,CAAC,gBAAgB,CAAS;gBAErB,OAAO,CAAC,EAAE,gBAAgB;IAMhC,OAAO,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC,gBAAgB,CAAC;IA8DrD,eAAe,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,EAAE,MAAM,GAAG,gBAAgB;IAyCtD,YAAY,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,gBAAgB,EAAE,CAAC;CASlE"}

package/dist/deep/extractor.js

@@ -0,0 +1,125 @@
+/**
+ * Content extraction using linkedom + @mozilla/readability
+ */
+import { parseHTML } from 'linkedom';
+import { Readability } from '@mozilla/readability';
+const DEFAULT_TIMEOUT_MS = 10_000;
+const DEFAULT_CONCURRENCY = 3;
+const MAX_RESPONSE_BYTES = 5 * 1024 * 1024; // 5MB
+export class ContentExtractor {
+    timeoutMs;
+    concurrency;
+    maxResponseBytes;
+    constructor(options) {
+        this.timeoutMs = options?.timeoutMs ?? DEFAULT_TIMEOUT_MS;
+        this.concurrency = options?.concurrency ?? DEFAULT_CONCURRENCY;
+        this.maxResponseBytes = options?.maxResponseBytes ?? MAX_RESPONSE_BYTES;
+    }
+    async extract(url) {
+        try {
+            const response = await fetch(url, {
+                headers: {
+                    'User-Agent': 'Mozilla/5.0 (compatible; SxngDeepSearch/1.0)',
+                    'Accept': 'text/html,application/xhtml+xml',
+                },
+                signal: AbortSignal.timeout(this.timeoutMs),
+            });
+            if (!response.ok) {
+                return {
+                    title: '',
+                    content: '',
+                    excerpt: '',
+                    url,
+                    length: 0,
+                    extractedAt: Date.now(),
+                    error: `HTTP ${response.status}: ${response.statusText}`,
+                };
+            }
+            const contentLength = response.headers.get('content-length');
+            if (contentLength && parseInt(contentLength, 10) > this.maxResponseBytes) {
+                return {
+                    title: '',
+                    content: '',
+                    excerpt: '',
+                    url,
+                    length: 0,
+                    extractedAt: Date.now(),
+                    error: `Response too large: ${contentLength} bytes (max ${this.maxResponseBytes})`,
+                };
+            }
+            const html = await response.text();
+            if (html.length > this.maxResponseBytes) {
+                return {
+                    title: '',
+                    content: '',
+                    excerpt: '',
+                    url,
+                    length: 0,
+                    extractedAt: Date.now(),
+                    error: `Response too large: ${html.length} bytes (max ${this.maxResponseBytes})`,
+                };
+            }
+            return this.extractFromHtml(html, url);
+        }
+        catch (error) {
+            return {
+                title: '',
+                content: '',
+                excerpt: '',
+                url,
+                length: 0,
+                extractedAt: Date.now(),
+                error: error instanceof Error ? error.message : String(error),
+            };
+        }
+    }
+    extractFromHtml(html, url) {
+        try {
+            const { document } = parseHTML(html);
+            const reader = new Readability(document, { charThreshold: 100 });
+            const article = reader.parse();
+            if (!article || !article.textContent) {
+                return {
+                    title: '',
+                    content: '',
+                    excerpt: '',
+                    url,
+                    length: 0,
+                    extractedAt: Date.now(),
+                    error: 'Readability could not extract content',
+                };
+            }
+            return {
+                title: article.title || '',
+                content: article.textContent.replace(/\n{3,}/g, '\n\n').trim(),
+                excerpt: article.excerpt || '',
+                url,
+                byline: article.byline || undefined,
+                siteName: article.siteName || undefined,
+                length: article.length || article.textContent.length,
+                extractedAt: Date.now(),
+            };
+        }
+        catch (error) {
+            return {
+                title: '',
+                content: '',
+                excerpt: '',
+                url,
+                length: 0,
+                extractedAt: Date.now(),
+                error: error instanceof Error ? error.message : 'HTML parsing failed',
+            };
+        }
+    }
+    async extractBatch(urls) {
+        const results = [];
+        for (let i = 0; i < urls.length; i += this.concurrency) {
+            const batch = urls.slice(i, i + this.concurrency);
+            const batchResults = await Promise.all(batch.map(u => this.extract(u)));
+            results.push(...batchResults);
+        }
+        return results;
+    }
+}
+//# sourceMappingURL=extractor.js.map

package/dist/deep/extractor.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"extractor.js","sourceRoot":"","sources":["../../src/deep/extractor.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,EAAE,SAAS,EAAE,MAAM,UAAU,CAAC;AACrC,OAAO,EAAE,WAAW,EAAE,MAAM,sBAAsB,CAAC;AAEnD,MAAM,kBAAkB,GAAG,MAAM,CAAC;AAClC,MAAM,mBAAmB,GAAG,CAAC,CAAC;AAC9B,MAAM,kBAAkB,GAAG,CAAC,GAAG,IAAI,GAAG,IAAI,CAAC,CAAC,MAAM;AAoBlD,MAAM,OAAO,gBAAgB;IACjB,SAAS,CAAS;IAClB,WAAW,CAAS;IACpB,gBAAgB,CAAS;IAEjC,YAAY,OAA0B;QAClC,IAAI,CAAC,SAAS,GAAG,OAAO,EAAE,SAAS,IAAI,kBAAkB,CAAC;QAC1D,IAAI,CAAC,WAAW,GAAG,OAAO,EAAE,WAAW,IAAI,mBAAmB,CAAC;QAC/D,IAAI,CAAC,gBAAgB,GAAG,OAAO,EAAE,gBAAgB,IAAI,kBAAkB,CAAC;IAC5E,CAAC;IAED,KAAK,CAAC,OAAO,CAAC,GAAW;QACrB,IAAI,CAAC;YACD,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,GAAG,EAAE;gBAC9B,OAAO,EAAE;oBACL,YAAY,EAAE,8CAA8C;oBAC5D,QAAQ,EAAE,iCAAiC;iBAC9C;gBACD,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,IAAI,CAAC,SAAS,CAAC;aAC9C,CAAC,CAAC;YAEH,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,CAAC;gBACf,OAAO;oBACH,KAAK,EAAE,EAAE;oBACT,OAAO,EAAE,EAAE;oBACX,OAAO,EAAE,EAAE;oBACX,GAAG;oBACH,MAAM,EAAE,CAAC;oBACT,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE;oBACvB,KAAK,EAAE,QAAQ,QAAQ,CAAC,MAAM,KAAK,QAAQ,CAAC,UAAU,EAAE;iBAC3D,CAAC;YACN,CAAC;YAED,MAAM,aAAa,GAAG,QAAQ,CAAC,OAAO,CAAC,GAAG,CAAC,gBAAgB,CAAC,CAAC;YAC7D,IAAI,aAAa,IAAI,QAAQ,CAAC,aAAa,EAAE,EAAE,CAAC,GAAG,IAAI,CAAC,gBAAgB,EAAE,CAAC;gBACvE,OAAO;oBACH,KAAK,EAAE,EAAE;oBACT,OAAO,EAAE,EAAE;oBACX,OAAO,EAAE,EAAE;oBACX,GAAG;oBACH,MAAM,EAAE,CAAC;oBACT,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE;oBACvB,KAAK,EAAE,uBAAuB,aAAa,eAAe,IAAI,CAAC,gBAAgB,GAAG;iBACrF,CAAC;YACN,CAAC;YAED,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;YACnC,IAAI,IAAI,CAAC,MAAM,GAAG,IAAI,CAAC,gBAAgB,EAAE,CAAC;gBACtC,OAAO;oBACH,KAAK,EAAE,EAAE;oBACT,OAAO,EAAE,EAAE;oBACX,OAAO,EAAE,EAAE;oBACX,GAAG;oBACH,MAAM,EAAE,CAAC;oBACT,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE;oBACvB,KAAK,EAAE,uBAAuB,IAAI,CAAC,MAAM,eAAe,IAAI,CAAC,gBAAgB,GAAG;iBACnF,CAAC;YACN,CAAC;YAED,OAAO,IAAI,CAAC,eAAe,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC;QAC3C,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACb,OAAO;gBACH,KAAK,EAAE,EAAE;gBACT,OAAO,EAAE,EAAE;gBACX,OAAO,EAAE,EAAE;gBACX,GAAG;gBACH,MAAM,EAAE,CAAC;gBACT,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE;gBACvB,KAAK,EAAE,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC;aAChE,CAAC;QACN,CAAC;IACL,CAAC;IAED,eAAe,CAAC,IAAY,EAAE,GAAW;QACrC,IAAI,CAAC;YACD,MAAM,EAAE,QAAQ,EAAE,GAAG,SAAS,CAAC,IAAI,CAAC,CAAC;YACrC,MAAM,MAAM,GAAG,IAAI,WAAW,CAAC,QAAe,EAAE,EAAE,aAAa,EAAE,GAAG,EAAE,CAAC,CAAC;YACxE,MAAM,OAAO,GAAG,MAAM,CAAC,KAAK,EAAE,CAAC;YAE/B,IAAI,CAAC,OAAO,IAAI,CAAC,OAAO,CAAC,WAAW,EAAE,CAAC;gBACnC,OAAO;oBACH,KAAK,EAAE,EAAE;oBACT,OAAO,EAAE,EAAE;oBACX,OAAO,EAAE,EAAE;oBACX,GAAG;oBACH,MAAM,EAAE,CAAC;oBACT,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE;oBACvB,KAAK,EAAE,uCAAuC;iBACjD,CAAC;YACN,CAAC;YAED,OAAO;gBACH,KAAK,EAAE,OAAO,CAAC,KAAK,IAAI,EAAE;gBAC1B,OAAO,EAAE,OAAO,CAAC,WAAW,CAAC,OAAO,CAAC,SAAS,EAAE,MAAM,CAAC,CAAC,IAAI,EAAE;gBAC9D,OAAO,EAAE,OAAO,CAAC,OAAO,IAAI,EAAE;gBAC9B,GAAG;gBACH,MAAM,EAAE,OAAO,CAAC,MAAM,IAAI,SAAS;gBACnC,QAAQ,EAAE,OAAO,CAAC,QAAQ,IAAI,SAAS;gBACvC,MAAM,EAAE,OAAO,CAAC,MAAM,IAAI,OAAO,CAAC,WAAW,CAAC,MAAM;gBACpD,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE;aAC1B,CAAC;QACN,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACb,OAAO;gBACH,KAAK,EAAE,EAAE;gBACT,OAAO,EAAE,EAAE;gBACX,OAAO,EAAE,EAAE;gBACX,GAAG;gBACH,MAAM,EAAE,CAAC;gBACT,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE;gBACvB,KAAK,EAAE,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,qBAAqB;aACxE,CAAC;QACN,CAAC;IACL,CAAC;IAED,KAAK,CAAC,YAAY,CAAC,IAAc;QAC7B,MAAM,OAAO,GAAuB,EAAE,CAAC;QACvC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,IAAI,IAAI,CAAC,WAAW,EAAE,CAAC;YACrD,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,WAAW,CAAC,CAAC;YAClD,MAAM,YAAY,GAAG,MAAM,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACxE,OAAO,CAAC,IAAI,CAAC,GAAG,YAAY,CAAC,CAAC;QAClC,CAAC;QACD,OAAO,OAAO,CAAC;IACnB,CAAC;CACJ"}

package/dist/deep/graph.d.ts

@@ -0,0 +1,55 @@
+/**
+ * Entity-centric knowledge graph using graphology
+ *
+ * The graph stores entities and their relationships.
+ * Search results are stored as lightweight metadata nodes (title, url, rank)
+ * connected to entities via "mentions" edges.
+ *
+ * Entity extraction is done by the Agent (LLM), not by the CLI.
+ * The CLI only stores what the Agent sends via graph-add.
+ */
+import { DirectedGraph } from 'graphology';
+export interface GraphNodeAttrs {
+    type: 'entity' | 'result' | 'query' | 'domain';
+    label: string;
+    entityType?: string;
+    score?: number;
+    url?: string;
+    rank?: number;
+    title?: string;
+    query?: string;
+    round?: number;
+    domain?: string;
+}
+export interface GraphEdgeAttrs {
+    relation: string;
+    weight: number;
+}
+/** Create an empty entity-centric knowledge graph */
+export declare function createGraph(): DirectedGraph<GraphNodeAttrs, GraphEdgeAttrs>;
+/** Generate a node ID for an entity */
+export declare function entityId(label: string): string;
+/** Generate a node ID for a result */
+export declare function resultId(url: string): string;
+/** Generate a node ID for a query */
+export declare function queryId(query: string): string;
+/** Generate a node ID for a domain */
+export declare function domainId(domain: string): string;
+/** Auto-build structural edges: query→result→domain */
+export declare function buildStructuralEdges(graph: DirectedGraph<GraphNodeAttrs, GraphEdgeAttrs>, query: string, results: Array<{
+    url: string;
+    title: string;
+    rank?: number;
+}>, round?: number): void;
+export declare function bfsSubgraph(graph: DirectedGraph<GraphNodeAttrs, GraphEdgeAttrs>, seeds: string[], depth?: number): DirectedGraph<GraphNodeAttrs, GraphEdgeAttrs>;
+export declare function serializeGraph(graph: DirectedGraph<GraphNodeAttrs, GraphEdgeAttrs>): object;
+export declare function deserializeGraph(data: any): DirectedGraph<GraphNodeAttrs, GraphEdgeAttrs>;
+export declare function graphStats(graph: DirectedGraph<GraphNodeAttrs, GraphEdgeAttrs>): {
+    nodes: number;
+    edges: number;
+    entities: number;
+    results: number;
+    queries: number;
+    domains: number;
+};
+//# sourceMappingURL=graph.d.ts.map

package/dist/deep/graph.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"graph.d.ts","sourceRoot":"","sources":["../../src/deep/graph.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAEH,OAAO,EAAE,aAAa,EAAE,MAAM,YAAY,CAAC;AAE3C,MAAM,WAAW,cAAc;IAC3B,IAAI,EAAE,QAAQ,GAAG,QAAQ,GAAG,OAAO,GAAG,QAAQ,CAAC;IAC/C,KAAK,EAAE,MAAM,CAAC;IAEd,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,KAAK,CAAC,EAAE,MAAM,CAAC;IAEf,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,KAAK,CAAC,EAAE,MAAM,CAAC;IAEf,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;IAEf,MAAM,CAAC,EAAE,MAAM,CAAC;CACnB;AAED,MAAM,WAAW,cAAc;IAC3B,QAAQ,EAAE,MAAM,CAAC;IACjB,MAAM,EAAE,MAAM,CAAC;CAClB;AAMD,qDAAqD;AACrD,wBAAgB,WAAW,IAAI,aAAa,CAAC,cAAc,EAAE,cAAc,CAAC,CAE3E;AAED,uCAAuC;AACvC,wBAAgB,QAAQ,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,CAE9C;AAED,sCAAsC;AACtC,wBAAgB,QAAQ,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAE5C;AAED,qCAAqC;AACrC,wBAAgB,OAAO,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,CAE7C;AAED,sCAAsC;AACtC,wBAAgB,QAAQ,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,CAE/C;AAWD,uDAAuD;AACvD,wBAAgB,oBAAoB,CAChC,KAAK,EAAE,aAAa,CAAC,cAAc,EAAE,cAAc,CAAC,EACpD,KAAK,EAAE,MAAM,EACb,OAAO,EAAE,KAAK,CAAC;IAAE,GAAG,EAAE,MAAM,CAAC;IAAC,KAAK,EAAE,MAAM,CAAC;IAAC,IAAI,CAAC,EAAE,MAAM,CAAA;CAAE,CAAC,EAC7D,KAAK,CAAC,EAAE,MAAM,GACf,IAAI,CAuDN;AAED,wBAAgB,WAAW,CACvB,KAAK,EAAE,aAAa,CAAC,cAAc,EAAE,cAAc,CAAC,EACpD,KAAK,EAAE,MAAM,EAAE,EACf,KAAK,GAAE,MAAU,GAClB,aAAa,CAAC,cAAc,EAAE,cAAc,CAAC,CA+B/C;AAED,wBAAgB,cAAc,CAAC,KAAK,EAAE,aAAa,CAAC,cAAc,EAAE,cAAc,CAAC,GAAG,MAAM,CAE3F;AAED,wBAAgB,gBAAgB,CAAC,IAAI,EAAE,GAAG,GAAG,aAAa,CAAC,cAAc,EAAE,cAAc,CAAC,CAIzF;AAED,wBAAgB,UAAU,CAAC,KAAK,EAAE,aAAa,CAAC,cAAc,EAAE,cAAc,CAAC,GAAG;IAC9E,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,QAAQ,EAAE,MAAM,CAAC;IACjB,OAAO,EAAE,MAAM,CAAC;IAChB,OAAO,EAAE,MAAM,CAAC;IAChB,OAAO,EAAE,MAAM,CAAC;CACnB,CAiBA"}

package/dist/deep/graph.js

@@ -0,0 +1,153 @@
+/**
+ * Entity-centric knowledge graph using graphology
+ *
+ * The graph stores entities and their relationships.
+ * Search results are stored as lightweight metadata nodes (title, url, rank)
+ * connected to entities via "mentions" edges.
+ *
+ * Entity extraction is done by the Agent (LLM), not by the CLI.
+ * The CLI only stores what the Agent sends via graph-add.
+ */
+import { DirectedGraph } from 'graphology';
+function generateId(prefix, value) {
+    return `${prefix}:${value.toLowerCase().replace(/[^\w]+/g, '_').slice(0, 60)}`;
+}
+/** Create an empty entity-centric knowledge graph */
+export function createGraph() {
+    return new DirectedGraph();
+}
+/** Generate a node ID for an entity */
+export function entityId(label) {
+    return generateId('e', label);
+}
+/** Generate a node ID for a result */
+export function resultId(url) {
+    return generateId('r', url);
+}
+/** Generate a node ID for a query */
+export function queryId(query) {
+    return generateId('q', query);
+}
+/** Generate a node ID for a domain */
+export function domainId(domain) {
+    return generateId('d', domain);
+}
+/** Extract domain from URL */
+function extractDomain(url) {
+    try {
+        return new URL(url).hostname;
+    }
+    catch {
+        return url;
+    }
+}
+/** Auto-build structural edges: query→result→domain */
+export function buildStructuralEdges(graph, query, results, round) {
+    const qId = queryId(query);
+    // Ensure query node exists
+    if (!graph.hasNode(qId)) {
+        graph.mergeNode(qId, {
+            type: 'query',
+            label: query,
+            query,
+            round,
+        });
+    }
+    for (let i = 0; i < results.length; i++) {
+        const r = results[i];
+        const rId = resultId(r.url);
+        // Ensure result node exists
+        if (!graph.hasNode(rId)) {
+            graph.mergeNode(rId, {
+                type: 'result',
+                label: r.title,
+                url: r.url,
+                title: r.title,
+                rank: r.rank ?? i + 1,
+            });
+        }
+        // query→result edge
+        if (!graph.hasEdge(qId, rId)) {
+            graph.addEdge(qId, rId, {
+                relation: 'yields',
+                weight: 1 / (i + 1),
+            });
+        }
+        // domain node and result→domain edge
+        const domain = extractDomain(r.url);
+        if (domain) {
+            const dId = domainId(domain);
+            if (!graph.hasNode(dId)) {
+                graph.mergeNode(dId, {
+                    type: 'domain',
+                    label: domain,
+                    domain,
+                });
+            }
+            if (!graph.hasEdge(rId, dId)) {
+                graph.addEdge(rId, dId, {
+                    relation: 'belongs_to',
+                    weight: 1,
+                });
+            }
+        }
+    }
+}
+export function bfsSubgraph(graph, seeds, depth = 2) {
+    const visited = new Set();
+    const queue = seeds
+        .filter(id => graph.hasNode(id))
+        .map(id => ({ id, depth: 0 }));
+    while (queue.length > 0) {
+        const { id, depth: d } = queue.shift();
+        if (d > depth || visited.has(id))
+            continue;
+        visited.add(id);
+        graph.forEachNeighbor(id, (neighbor) => {
+            if (!visited.has(neighbor)) {
+                queue.push({ id: neighbor, depth: d + 1 });
+            }
+        });
+    }
+    const subgraph = new DirectedGraph();
+    for (const nodeId of visited) {
+        const attrs = graph.getNodeAttributes(nodeId);
+        subgraph.mergeNode(nodeId, attrs);
+    }
+    graph.forEachEdge((edge, attrs, source, target) => {
+        if (visited.has(source) && visited.has(target)) {
+            subgraph.mergeEdge(source, target, attrs);
+        }
+    });
+    return subgraph;
+}
+export function serializeGraph(graph) {
+    return graph.export();
+}
+export function deserializeGraph(data) {
+    const graph = new DirectedGraph();
+    graph.import(data);
+    return graph;
+}
+export function graphStats(graph) {
+    let entities = 0, results = 0, queries = 0, domains = 0;
+    graph.forEachNode((node, attrs) => {
+        if (attrs.type === 'entity')
+            entities++;
+        else if (attrs.type === 'result')
+            results++;
+        else if (attrs.type === 'query')
+            queries++;
+        else if (attrs.type === 'domain')
+            domains++;
+    });
+    return {
+        nodes: graph.order,
+        edges: graph.size,
+        entities,
+        results,
+        queries,
+        domains,
+    };
+}
+//# sourceMappingURL=graph.js.map

package/dist/deep/graph.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"graph.js","sourceRoot":"","sources":["../../src/deep/graph.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AAEH,OAAO,EAAE,aAAa,EAAE,MAAM,YAAY,CAAC;AAwB3C,SAAS,UAAU,CAAC,MAAc,EAAE,KAAa;IAC7C,OAAO,GAAG,MAAM,IAAI,KAAK,CAAC,WAAW,EAAE,CAAC,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAC;AACnF,CAAC;AAED,qDAAqD;AACrD,MAAM,UAAU,WAAW;IACvB,OAAO,IAAI,aAAa,EAAkC,CAAC;AAC/D,CAAC;AAED,uCAAuC;AACvC,MAAM,UAAU,QAAQ,CAAC,KAAa;IAClC,OAAO,UAAU,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;AAClC,CAAC;AAED,sCAAsC;AACtC,MAAM,UAAU,QAAQ,CAAC,GAAW;IAChC,OAAO,UAAU,CAAC,GAAG,EAAE,GAAG,CAAC,CAAC;AAChC,CAAC;AAED,qCAAqC;AACrC,MAAM,UAAU,OAAO,CAAC,KAAa;IACjC,OAAO,UAAU,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;AAClC,CAAC;AAED,sCAAsC;AACtC,MAAM,UAAU,QAAQ,CAAC,MAAc;IACnC,OAAO,UAAU,CAAC,GAAG,EAAE,MAAM,CAAC,CAAC;AACnC,CAAC;AAED,8BAA8B;AAC9B,SAAS,aAAa,CAAC,GAAW;IAC9B,IAAI,CAAC;QACD,OAAO,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC,QAAQ,CAAC;IACjC,CAAC;IAAC,MAAM,CAAC;QACL,OAAO,GAAG,CAAC;IACf,CAAC;AACL,CAAC;AAED,uDAAuD;AACvD,MAAM,UAAU,oBAAoB,CAChC,KAAoD,EACpD,KAAa,EACb,OAA6D,EAC7D,KAAc;IAEd,MAAM,GAAG,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC;IAE3B,2BAA2B;IAC3B,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC;QACtB,KAAK,CAAC,SAAS,CAAC,GAAG,EAAE;YACjB,IAAI,EAAE,OAAO;YACb,KAAK,EAAE,KAAK;YACZ,KAAK;YACL,KAAK;SACR,CAAC,CAAC;IACP,CAAC;IAED,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,MAAM,CAAC,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC;QACrB,MAAM,GAAG,GAAG,QAAQ,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QAE5B,4BAA4B;QAC5B,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC;YACtB,KAAK,CAAC,SAAS,CAAC,GAAG,EAAE;gBACjB,IAAI,EAAE,QAAQ;gBACd,KAAK,EAAE,CAAC,CAAC,KAAK;gBACd,GAAG,EAAE,CAAC,CAAC,GAAG;gBACV,KAAK,EAAE,CAAC,CAAC,KAAK;gBACd,IAAI,EAAE,CAAC,CAAC,IAAI,IAAI,CAAC,GAAG,CAAC;aACxB,CAAC,CAAC;QACP,CAAC;QAED,oBAAoB;QACpB,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,EAAE,GAAG,CAAC,EAAE,CAAC;YAC3B,KAAK,CAAC,OAAO,CAAC,GAAG,EAAE,GAAG,EAAE;gBACpB,QAAQ,EAAE,QAAQ;gBAClB,MAAM,EAAE,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC;aACtB,CAAC,CAAC;QACP,CAAC;QAED,qCAAqC;QACrC,MAAM,MAAM,GAAG,aAAa,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QACpC,IAAI,MAAM,EAAE,CAAC;YACT,MAAM,GAAG,GAAG,QAAQ,CAAC,MAAM,CAAC,CAAC;YAC7B,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC;gBACtB,KAAK,CAAC,SAAS,CAAC,GAAG,EAAE;oBACjB,IAAI,EAAE,QAAQ;oBACd,KAAK,EAAE,MAAM;oBACb,MAAM;iBACT,CAAC,CAAC;YACP,CAAC;YACD,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,EAAE,GAAG,CAAC,EAAE,CAAC;gBAC3B,KAAK,CAAC,OAAO,CAAC,GAAG,EAAE,GAAG,EAAE;oBACpB,QAAQ,EAAE,YAAY;oBACtB,MAAM,EAAE,CAAC;iBACZ,CAAC,CAAC;YACP,CAAC;QACL,CAAC;IACL,CAAC;AACL,CAAC;AAED,MAAM,UAAU,WAAW,CACvB,KAAoD,EACpD,KAAe,EACf,QAAgB,CAAC;IAEjB,MAAM,OAAO,GAAG,IAAI,GAAG,EAAU,CAAC;IAClC,MAAM,KAAK,GAAyC,KAAK;SACpD,MAAM,CAAC,EAAE,CAAC,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC;SAC/B,GAAG,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,EAAE,EAAE,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC;IAEnC,OAAO,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACtB,MAAM,EAAE,EAAE,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,KAAK,CAAC,KAAK,EAAG,CAAC;QACxC,IAAI,CAAC,GAAG,KAAK,IAAI,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC;YAAE,SAAS;QAE3C,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QAChB,KAAK,CAAC,eAAe,CAAC,EAAE,EAAE,CAAC,QAAgB,EAAE,EAAE;YAC3C,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,QAAQ,CAAC,EAAE,CAAC;gBACzB,KAAK,CAAC,IAAI,CAAC,EAAE,EAAE,EAAE,QAAQ,EAAE,KAAK,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;YAC/C,CAAC;QACL,CAAC,CAAC,CAAC;IACP,CAAC;IAED,MAAM,QAAQ,GAAG,IAAI,aAAa,EAAkC,CAAC;IACrE,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE,CAAC;QAC3B,MAAM,KAAK,GAAG,KAAK,CAAC,iBAAiB,CAAC,MAAM,CAAC,CAAC;QAC9C,QAAQ,CAAC,SAAS,CAAC,MAAM,EAAE,KAAK,CAAC,CAAC;IACtC,CAAC;IAED,KAAK,CAAC,WAAW,CAAC,CAAC,IAAY,EAAE,KAAqB,EAAE,MAAc,EAAE,MAAc,EAAE,EAAE;QACtF,IAAI,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC;YAC7C,QAAQ,CAAC,SAAS,CAAC,MAAM,EAAE,MAAM,EAAE,KAAK,CAAC,CAAC;QAC9C,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,OAAO,QAAQ,CAAC;AACpB,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,KAAoD;IAC/E,OAAO,KAAK,CAAC,MAAM,EAAE,CAAC;AAC1B,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,IAAS;IACtC,MAAM,KAAK,GAAG,IAAI,aAAa,EAAkC,CAAC;IAClE,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC;IACnB,OAAO,KAAK,CAAC;AACjB,CAAC;AAED,MAAM,UAAU,UAAU,CAAC,KAAoD;IAQ3E,IAAI,QAAQ,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,CAAC;IACxD,KAAK,CAAC,WAAW,CAAC,CAAC,IAAY,EAAE,KAAqB,EAAE,EAAE;QACtD,IAAI,KAAK,CAAC,IAAI,KAAK,QAAQ;YAAE,QAAQ,EAAE,CAAC;aACnC,IAAI,KAAK,CAAC,IAAI,KAAK,QAAQ;YAAE,OAAO,EAAE,CAAC;aACvC,IAAI,KAAK,CAAC,IAAI,KAAK,OAAO;YAAE,OAAO,EAAE,CAAC;aACtC,IAAI,KAAK,CAAC,IAAI,KAAK,QAAQ;YAAE,OAAO,EAAE,CAAC;IAChD,CAAC,CAAC,CAAC;IAEH,OAAO;QACH,KAAK,EAAE,KAAK,CAAC,KAAK;QAClB,KAAK,EAAE,KAAK,CAAC,IAAI;QACjB,QAAQ;QACR,OAAO;QACP,OAAO;QACP,OAAO;KACV,CAAC;AACN,CAAC"}

package/dist/deep/index.d.ts

@@ -0,0 +1,10 @@
+/**
+ * Deep search module exports
+ */
+export { rrf, RrfInputItem, RrfOutputItem } from './rrf.js';
+export { SimHash } from './simhash.js';
+export { dedupe, dedupeByUrl, dedupeBySimHash, normalizeUrl, DedupItem } from './dedupe.js';
+export { ContentExtractor, ExtractedContent } from './extractor.js';
+export { createGraph, entityId, resultId, queryId, domainId, buildStructuralEdges, bfsSubgraph, serializeGraph, deserializeGraph, graphStats, GraphNodeAttrs, GraphEdgeAttrs } from './graph.js';
+export { initSessionDir, resolveSessionPath, appendSessionResults, updateSessionGraph, loadSessionResults, mergeExtractedContent, SessionResult } from './session.js';
+//# sourceMappingURL=index.d.ts.map

package/dist/deep/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/deep/index.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,EAAE,GAAG,EAAE,YAAY,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AAC5D,OAAO,EAAE,OAAO,EAAE,MAAM,cAAc,CAAC;AACvC,OAAO,EAAE,MAAM,EAAE,WAAW,EAAE,eAAe,EAAE,YAAY,EAAE,SAAS,EAAE,MAAM,aAAa,CAAC;AAC5F,OAAO,EAAE,gBAAgB,EAAE,gBAAgB,EAAE,MAAM,gBAAgB,CAAC;AACpE,OAAO,EAAE,WAAW,EAAE,QAAQ,EAAE,QAAQ,EAAE,OAAO,EAAE,QAAQ,EAAE,oBAAoB,EAAE,WAAW,EAAE,cAAc,EAAE,gBAAgB,EAAE,UAAU,EAAE,cAAc,EAAE,cAAc,EAAE,MAAM,YAAY,CAAC;AACjM,OAAO,EAAE,cAAc,EAAE,kBAAkB,EAAE,oBAAoB,EAAE,kBAAkB,EAAE,kBAAkB,EAAE,qBAAqB,EAAE,aAAa,EAAE,MAAM,cAAc,CAAC"}

package/dist/deep/index.js

@@ -0,0 +1,10 @@
+/**
+ * Deep search module exports
+ */
+export { rrf } from './rrf.js';
+export { SimHash } from './simhash.js';
+export { dedupe, dedupeByUrl, dedupeBySimHash, normalizeUrl } from './dedupe.js';
+export { ContentExtractor } from './extractor.js';
+export { createGraph, entityId, resultId, queryId, domainId, buildStructuralEdges, bfsSubgraph, serializeGraph, deserializeGraph, graphStats } from './graph.js';
+export { initSessionDir, resolveSessionPath, appendSessionResults, updateSessionGraph, loadSessionResults, mergeExtractedContent } from './session.js';
+//# sourceMappingURL=index.js.map

package/dist/deep/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/deep/index.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,EAAE,GAAG,EAA+B,MAAM,UAAU,CAAC;AAC5D,OAAO,EAAE,OAAO,EAAE,MAAM,cAAc,CAAC;AACvC,OAAO,EAAE,MAAM,EAAE,WAAW,EAAE,eAAe,EAAE,YAAY,EAAa,MAAM,aAAa,CAAC;AAC5F,OAAO,EAAE,gBAAgB,EAAoB,MAAM,gBAAgB,CAAC;AACpE,OAAO,EAAE,WAAW,EAAE,QAAQ,EAAE,QAAQ,EAAE,OAAO,EAAE,QAAQ,EAAE,oBAAoB,EAAE,WAAW,EAAE,cAAc,EAAE,gBAAgB,EAAE,UAAU,EAAkC,MAAM,YAAY,CAAC;AACjM,OAAO,EAAE,cAAc,EAAE,kBAAkB,EAAE,oBAAoB,EAAE,kBAAkB,EAAE,kBAAkB,EAAE,qBAAqB,EAAiB,MAAM,cAAc,CAAC"}

package/dist/deep/keywords.d.ts

@@ -0,0 +1,27 @@
+/**
+ * Entity extraction using compromise NLP (English) + BM25 bigrams (CJK)
+ *
+ * English: compromise extracts named entities + noun phrases (POS-aware)
+ * CJK: character bigrams with BM25 scoring (compromise doesn't support CJK)
+ * Both phases use BM25 cross-document scoring for relevance ranking.
+ */
+export interface KeywordResult {
+    term: string;
+    score: number;
+    type: 'named' | 'proper' | 'phrase';
+}
+/**
+ * Extract keywords from a collection of documents.
+ *
+ * English: compromise for POS-aware candidate extraction (named entities,
+ * proper nouns, noun phrases), filtered for validity.
+ * CJK: character bigrams with stop word filtering.
+ * Both: BM25 cross-document scoring for relevance ranking.
+ */
+export declare function extractKeywords(documents: string[], topN?: number): KeywordResult[][];
+/**
+ * Extract global keywords across all documents.
+ * Aggregates scores and boosts terms appearing in multiple documents.
+ */
+export declare function extractGlobalKeywords(documents: string[], topN?: number): KeywordResult[];
+//# sourceMappingURL=keywords.d.ts.map

package/dist/deep/keywords.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"keywords.d.ts","sourceRoot":"","sources":["../../src/deep/keywords.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAIH,MAAM,WAAW,aAAa;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,OAAO,GAAG,QAAQ,GAAG,QAAQ,CAAC;CACvC;AAkQD;;;;;;;GAOG;AACH,wBAAgB,eAAe,CAC3B,SAAS,EAAE,MAAM,EAAE,EACnB,IAAI,GAAE,MAAU,GACjB,aAAa,EAAE,EAAE,CAwBnB;AAED;;;GAGG;AACH,wBAAgB,qBAAqB,CACjC,SAAS,EAAE,MAAM,EAAE,EACnB,IAAI,GAAE,MAAW,GAClB,aAAa,EAAE,CAwBjB"}