@cesarandreslopez/occ 0.5.0 → 0.6.0

package/dist/src/doc/discover.js

@@ -0,0 +1,65 @@
+import fg from 'fast-glob';
+import { readFile, stat } from 'node:fs/promises';
+import path from 'node:path';
+const DOCUMENT_EXTENSIONS = ['docx', 'pdf', 'pptx', 'xlsx', 'odt', 'ods', 'odp'];
+function isDocumentFormat(ext) {
+    return DOCUMENT_EXTENSIONS.includes(ext);
+}
+async function loadGitignorePatterns(rootDir) {
+    try {
+        const content = await readFile(path.join(rootDir, '.gitignore'), 'utf8');
+        return content
+            .split(/\r?\n/)
+            .map(line => line.trim())
+            .filter(line => line.length > 0 && !line.startsWith('#') && !line.startsWith('!'))
+            .map(line => {
+            const normalized = line.replace(/^\//, '').split(path.sep).join('/');
+            if (normalized.endsWith('/'))
+                return `${normalized}**`;
+            return normalized;
+        });
+    }
+    catch {
+        return [];
+    }
+}
+export async function discoverDocuments(rootDir, options) {
+    const resolvedRoot = path.resolve(rootDir);
+    const excludeDir = options?.excludeDir ?? ['node_modules', '.git', 'dist', 'vendor', 'build', 'coverage', 'target'];
+    const maxSizeBytes = options?.maxSizeBytes ?? 50 * 1024 * 1024; // 50MB
+    const ignore = [
+        ...excludeDir.flatMap(dir => [`**/${dir}/**`]),
+        ...await loadGitignorePatterns(resolvedRoot),
+    ];
+    const pattern = `**/*.{${DOCUMENT_EXTENSIONS.join(',')}}`;
+    const discovered = await fg(pattern, {
+        cwd: resolvedRoot,
+        absolute: true,
+        onlyFiles: true,
+        dot: false,
+        followSymbolicLinks: false,
+        ignore,
+    });
+    const documents = [];
+    for (const filePath of discovered.sort()) {
+        try {
+            const fileStat = await stat(filePath);
+            if (!fileStat.isFile() || fileStat.size > maxSizeBytes)
+                continue;
+            const ext = path.extname(filePath).toLowerCase().replace('.', '');
+            if (!isDocumentFormat(ext))
+                continue;
+            documents.push({
+                filePath,
+                relativePath: path.relative(resolvedRoot, filePath).split(path.sep).join('/'),
+                format: ext,
+                sizeBytes: fileStat.size,
+            });
+        }
+        catch {
+            // Skip files that disappear during discovery
+        }
+    }
+    return documents;
+}
+//# sourceMappingURL=discover.js.map

package/dist/src/doc/discover.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"discover.js","sourceRoot":"","sources":["../../../src/doc/discover.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,WAAW,CAAC;AAC3B,OAAO,EAAE,QAAQ,EAAE,IAAI,EAAE,MAAM,kBAAkB,CAAC;AAClD,OAAO,IAAI,MAAM,WAAW,CAAC;AAS7B,MAAM,mBAAmB,GAAG,CAAC,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,CAAU,CAAC;AAI1F,SAAS,gBAAgB,CAAC,GAAW;IACnC,OAAQ,mBAAyC,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;AAClE,CAAC;AAED,KAAK,UAAU,qBAAqB,CAAC,OAAe;IAClD,IAAI,CAAC;QACH,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,YAAY,CAAC,EAAE,MAAM,CAAC,CAAC;QACzE,OAAO,OAAO;aACX,KAAK,CAAC,OAAO,CAAC;aACd,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC;aACxB,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC;aACjF,GAAG,CAAC,IAAI,CAAC,EAAE;YACV,MAAM,UAAU,GAAG,IAAI,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YACrE,IAAI,UAAU,CAAC,QAAQ,CAAC,GAAG,CAAC;gBAAE,OAAO,GAAG,UAAU,IAAI,CAAC;YACvD,OAAO,UAAU,CAAC;QACpB,CAAC,CAAC,CAAC;IACP,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,EAAE,CAAC;IACZ,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,iBAAiB,CAAC,OAAe,EAAE,OAGxD;IACC,MAAM,YAAY,GAAG,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC;IAC3C,MAAM,UAAU,GAAG,OAAO,EAAE,UAAU,IAAI,CAAC,cAAc,EAAE,MAAM,EAAE,MAAM,EAAE,QAAQ,EAAE,OAAO,EAAE,UAAU,EAAE,QAAQ,CAAC,CAAC;IACpH,MAAM,YAAY,GAAG,OAAO,EAAE,YAAY,IAAI,EAAE,GAAG,IAAI,GAAG,IAAI,CAAC,CAAC,OAAO;IAEvE,MAAM,MAAM,GAAG;QACb,GAAG,UAAU,CAAC,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,MAAM,GAAG,KAAK,CAAC,CAAC;QAC9C,GAAG,MAAM,qBAAqB,CAAC,YAAY,CAAC;KAC7C,CAAC;IAEF,MAAM,OAAO,GAAG,SAAS,mBAAmB,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;IAE1D,MAAM,UAAU,GAAG,MAAM,EAAE,CAAC,OAAO,EAAE;QACnC,GAAG,EAAE,YAAY;QACjB,QAAQ,EAAE,IAAI;QACd,SAAS,EAAE,IAAI;QACf,GAAG,EAAE,KAAK;QACV,mBAAmB,EAAE,KAAK;QAC1B,MAAM;KACP,CAAC,CAAC;IAEH,MAAM,SAAS,GAAyB,EAAE,CAAC;IAC3C,KAAK,MAAM,QAAQ,IAAI,UAAU,CAAC,IAAI,EAAE,EAAE,CAAC;QACzC,IAAI,CAAC;YACH,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,QAAQ,CAAC,CAAC;YACtC,IAAI,CAAC,QAAQ,CAAC,MAAM,EAAE,IAAI,QAAQ,CAAC,IAAI,GAAG,YAAY;gBAAE,SAAS;YAEjE,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC,OAAO,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC;YAClE,IAAI,CAAC,gBAAgB,CAAC,GAAG,CAAC;gBAAE,SAAS;YAErC,SAAS,CAAC,IAAI,CAAC;gBACb,QAAQ;gBACR,YAAY,EAAE,IAAI,CAAC,QAAQ,CAAC,YAAY,EAAE,QAAQ,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC;gBAC7E,MAAM,EAAE,GAAG;gBACX,SAAS,EAAE,QAAQ,CAAC,IAAI;aACzB,CAAC,CAAC;QACL,CAAC;QAAC,MAAM,CAAC;YACP,6CAA6C;QAC/C,CAAC;IACH,CAAC;IAED,OAAO,SAAS,CAAC;AACnB,CAAC"}

package/dist/src/doc/entities.d.ts

@@ -0,0 +1,27 @@
+/**
+ * Light entity extraction — regex-based pattern matching + TF-IDF keyword extraction.
+ */
+export interface ExtractedEntity {
+    type: 'email' | 'url' | 'date' | 'phone' | 'mention';
+    value: string;
+    position: number;
+    context: string;
+}
+export interface KeywordScore {
+    term: string;
+    tfidf: number;
+    tf: number;
+    idf: number;
+}
+export interface EntityExtractionResult {
+    entities: ExtractedEntity[];
+    keywords: KeywordScore[];
+}
+/** Extract entities (emails, URLs, dates, phone numbers, @mentions) from text */
+export declare function extractEntities(text: string): ExtractedEntity[];
+/** Extract keywords using TF-IDF across a batch of documents */
+export declare function extractKeywords(texts: string[], topK?: number): KeywordScore[][];
+/** Single-document keyword extraction (IDF is 1 for all terms, falls back to TF) */
+export declare function extractKeywordsSingle(text: string, topK?: number): KeywordScore[];
+/** Full entity extraction: entities + keywords for a single document */
+export declare function extractDocumentEntities(text: string, topK?: number): EntityExtractionResult;

package/dist/src/doc/entities.js

@@ -0,0 +1,102 @@
+/**
+ * Light entity extraction — regex-based pattern matching + TF-IDF keyword extraction.
+ */
+const ENTITY_PATTERNS = [
+    { type: 'email', pattern: /[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}/g },
+    { type: 'url', pattern: /https?:\/\/[^\s)<>]+/g },
+    { type: 'date', pattern: /\b\d{4}-\d{2}-\d{2}\b/g },
+    { type: 'date', pattern: /\b\d{1,2}\/\d{1,2}\/\d{2,4}\b/g },
+    { type: 'phone', pattern: /(?:\+\d{1,3}[\s-]?)?\(?\d{3}\)?[\s.-]?\d{3}[\s.-]?\d{4}\b/g },
+    { type: 'mention', pattern: /@[a-zA-Z0-9_-]+/g },
+];
+const CONTEXT_WINDOW = 40;
+function extractContext(text, position, length) {
+    const start = Math.max(0, position - CONTEXT_WINDOW);
+    const end = Math.min(text.length, position + length + CONTEXT_WINDOW);
+    let context = text.slice(start, end).replace(/\n/g, ' ');
+    if (start > 0)
+        context = '...' + context;
+    if (end < text.length)
+        context = context + '...';
+    return context;
+}
+/** Extract entities (emails, URLs, dates, phone numbers, @mentions) from text */
+export function extractEntities(text) {
+    const entities = [];
+    const seen = new Set();
+    for (const { type, pattern } of ENTITY_PATTERNS) {
+        const regex = new RegExp(pattern.source, pattern.flags);
+        let match;
+        while ((match = regex.exec(text)) !== null) {
+            const key = `${type}:${match[0]}:${match.index}`;
+            if (seen.has(key))
+                continue;
+            seen.add(key);
+            entities.push({
+                type,
+                value: match[0],
+                position: match.index,
+                context: extractContext(text, match.index, match[0].length),
+            });
+        }
+    }
+    return entities.sort((a, b) => a.position - b.position);
+}
+const STOP_WORDS = new Set([
+    'the', 'be', 'to', 'of', 'and', 'a', 'in', 'that', 'have', 'i',
+    'it', 'for', 'not', 'on', 'with', 'he', 'as', 'you', 'do', 'at',
+    'this', 'but', 'his', 'by', 'from', 'they', 'we', 'say', 'her', 'she',
+    'or', 'an', 'will', 'my', 'one', 'all', 'would', 'there', 'their', 'what',
+    'so', 'up', 'out', 'if', 'about', 'who', 'get', 'which', 'go', 'me',
+    'when', 'make', 'can', 'like', 'time', 'no', 'just', 'him', 'know', 'take',
+    'is', 'are', 'was', 'were', 'been', 'being', 'has', 'had', 'having',
+    'does', 'did', 'doing', 'am', 'its', 'than', 'also', 'into', 'could',
+    'may', 'should', 'each', 'other', 'then', 'them', 'these', 'some',
+]);
+function tokenize(text) {
+    return text.toLowerCase().split(/[^a-zA-Z0-9]+/).filter(w => w.length > 2 && !STOP_WORDS.has(w));
+}
+/** Extract keywords using TF-IDF across a batch of documents */
+export function extractKeywords(texts, topK = 20) {
+    const docCount = texts.length;
+    const allTokenized = texts.map(tokenize);
+    // Document frequency
+    const df = new Map();
+    for (const tokens of allTokenized) {
+        const unique = new Set(tokens);
+        for (const term of unique) {
+            df.set(term, (df.get(term) ?? 0) + 1);
+        }
+    }
+    return allTokenized.map(tokens => {
+        // Term frequency for this document
+        const tf = new Map();
+        for (const term of tokens) {
+            tf.set(term, (tf.get(term) ?? 0) + 1);
+        }
+        const scores = [];
+        for (const [term, count] of tf) {
+            const termFreq = count / tokens.length;
+            const inverseDocFreq = Math.log(docCount / (df.get(term) ?? 1));
+            scores.push({
+                term,
+                tfidf: termFreq * inverseDocFreq,
+                tf: termFreq,
+                idf: inverseDocFreq,
+            });
+        }
+        return scores.sort((a, b) => b.tfidf - a.tfidf).slice(0, topK);
+    });
+}
+/** Single-document keyword extraction (IDF is 1 for all terms, falls back to TF) */
+export function extractKeywordsSingle(text, topK = 20) {
+    return extractKeywords([text], topK)[0];
+}
+/** Full entity extraction: entities + keywords for a single document */
+export function extractDocumentEntities(text, topK = 20) {
+    return {
+        entities: extractEntities(text),
+        keywords: extractKeywordsSingle(text, topK),
+    };
+}
+//# sourceMappingURL=entities.js.map

package/dist/src/doc/entities.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"entities.js","sourceRoot":"","sources":["../../../src/doc/entities.ts"],"names":[],"mappings":"AAAA;;GAEG;AAqBH,MAAM,eAAe,GAA8D;IACjF,EAAE,IAAI,EAAE,OAAO,EAAE,OAAO,EAAE,iDAAiD,EAAE;IAC7E,EAAE,IAAI,EAAE,KAAK,EAAE,OAAO,EAAE,uBAAuB,EAAE;IACjD,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,wBAAwB,EAAE;IACnD,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,gCAAgC,EAAE;IAC3D,EAAE,IAAI,EAAE,OAAO,EAAE,OAAO,EAAE,4DAA4D,EAAE;IACxF,EAAE,IAAI,EAAE,SAAS,EAAE,OAAO,EAAE,kBAAkB,EAAE;CACjD,CAAC;AAEF,MAAM,cAAc,GAAG,EAAE,CAAC;AAE1B,SAAS,cAAc,CAAC,IAAY,EAAE,QAAgB,EAAE,MAAc;IACpE,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,QAAQ,GAAG,cAAc,CAAC,CAAC;IACrD,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,EAAE,QAAQ,GAAG,MAAM,GAAG,cAAc,CAAC,CAAC;IACtE,IAAI,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;IACzD,IAAI,KAAK,GAAG,CAAC;QAAE,OAAO,GAAG,KAAK,GAAG,OAAO,CAAC;IACzC,IAAI,GAAG,GAAG,IAAI,CAAC,MAAM;QAAE,OAAO,GAAG,OAAO,GAAG,KAAK,CAAC;IACjD,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,iFAAiF;AACjF,MAAM,UAAU,eAAe,CAAC,IAAY;IAC1C,MAAM,QAAQ,GAAsB,EAAE,CAAC;IACvC,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;IAE/B,KAAK,MAAM,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,eAAe,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC;QACxD,IAAI,KAA6B,CAAC;QAClC,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YAC3C,MAAM,GAAG,GAAG,GAAG,IAAI,IAAI,KAAK,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,KAAK,EAAE,CAAC;YACjD,IAAI,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;gBAAE,SAAS;YAC5B,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;YACd,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI;gBACJ,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC;gBACf,QAAQ,EAAE,KAAK,CAAC,KAAK;gBACrB,OAAO,EAAE,cAAc,CAAC,IAAI,EAAE,KAAK,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;aAC5D,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IAED,OAAO,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC,CAAC;AAC1D,CAAC;AAED,MAAM,UAAU,GAAG,IAAI,GAAG,CAAC;IACzB,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,GAAG;IAC9D,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI;IAC/D,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK;IACrE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM;IACzE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,IAAI;IACnE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM;IAC1E,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,QAAQ;IACnE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO;IACpE,KAAK,EAAE,QAAQ,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM;CAClE,CAAC,CAAC;AAEH,SAAS,QAAQ,CAAC,IAAY;IAC5B,OAAO,IAAI,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,eAAe,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;AACnG,CAAC;AAED,gEAAgE;AAChE,MAAM,UAAU,eAAe,CAAC,KAAe,EAAE,IAAI,GAAG,EAAE;IACxD,MAAM,QAAQ,GAAG,KAAK,CAAC,MAAM,CAAC;IAC9B,MAAM,YAAY,GAAG,KAAK,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC;IAEzC,qBAAqB;IACrB,MAAM,EAAE,GAAG,IAAI,GAAG,EAAkB,CAAC;IACrC,KAAK,MAAM,MAAM,IAAI,YAAY,EAAE,CAAC;QAClC,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,CAAC;QAC/B,KAAK,MAAM,IAAI,IAAI,MAAM,EAAE,CAAC;YAC1B,EAAE,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QACxC,CAAC;IACH,CAAC;IAED,OAAO,YAAY,CAAC,GAAG,CAAC,MAAM,CAAC,EAAE;QAC/B,mCAAmC;QACnC,MAAM,EAAE,GAAG,IAAI,GAAG,EAAkB,CAAC;QACrC,KAAK,MAAM,IAAI,IAAI,MAAM,EAAE,CAAC;YAC1B,EAAE,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QACxC,CAAC;QAED,MAAM,MAAM,GAAmB,EAAE,CAAC;QAClC,KAAK,MAAM,CAAC,IAAI,EAAE,KAAK,CAAC,IAAI,EAAE,EAAE,CAAC;YAC/B,MAAM,QAAQ,GAAG,KAAK,GAAG,MAAM,CAAC,MAAM,CAAC;YACvC,MAAM,cAAc,GAAG,IAAI,CAAC,GAAG,CAAC,QAAQ,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YAChE,MAAM,CAAC,IAAI,CAAC;gBACV,IAAI;gBACJ,KAAK,EAAE,QAAQ,GAAG,cAAc;gBAChC,EAAE,EAAE,QAAQ;gBACZ,GAAG,EAAE,cAAc;aACpB,CAAC,CAAC;QACL,CAAC;QAED,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;IACjE,CAAC,CAAC,CAAC;AACL,CAAC;AAED,oFAAoF;AACpF,MAAM,UAAU,qBAAqB,CAAC,IAAY,EAAE,IAAI,GAAG,EAAE;IAC3D,OAAO,eAAe,CAAC,CAAC,IAAI,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;AAC1C,CAAC;AAED,wEAAwE;AACxE,MAAM,UAAU,uBAAuB,CAAC,IAAY,EAAE,IAAI,GAAG,EAAE;IAC7D,OAAO;QACL,QAAQ,EAAE,eAAe,CAAC,IAAI,CAAC;QAC/B,QAAQ,EAAE,qBAAqB,CAAC,IAAI,EAAE,IAAI,CAAC;KAC5C,CAAC;AACJ,CAAC"}

package/dist/src/doc/references.d.ts

@@ -0,0 +1,20 @@
+export interface CrossReference {
+    sourceFile: string;
+    targetFile: string;
+    referenceType: 'filename-mention' | 'hyperlink' | 'citation';
+    context: string;
+    line?: number;
+}
+export interface ReferenceAnalysis {
+    references: CrossReference[];
+    unresolvedMentions: Array<{
+        sourceFile: string;
+        mention: string;
+        context: string;
+    }>;
+}
+/** Detect cross-references between documents */
+export declare function detectCrossReferences(documents: Array<{
+    filePath: string;
+    content: string;
+}>): ReferenceAnalysis;

package/dist/src/doc/references.js

@@ -0,0 +1,112 @@
+/**
+ * Cross-document reference detection.
+ * Detects filename mentions, hyperlinks, and citation patterns across documents.
+ */
+import path from 'node:path';
+const CONTEXT_WINDOW = 60;
+function getContext(text, position, length) {
+    const start = Math.max(0, position - CONTEXT_WINDOW);
+    const end = Math.min(text.length, position + length + CONTEXT_WINDOW);
+    let context = text.slice(start, end).replace(/\n/g, ' ').trim();
+    if (start > 0)
+        context = '...' + context;
+    if (end < text.length)
+        context = context + '...';
+    return context;
+}
+function getLineNumber(text, position) {
+    let line = 1;
+    for (let i = 0; i < position && i < text.length; i++) {
+        if (text[i] === '\n')
+            line++;
+    }
+    return line;
+}
+/** Detect cross-references between documents */
+export function detectCrossReferences(documents) {
+    const references = [];
+    const unresolvedMentions = [];
+    // Build a set of known filenames for matching
+    const fileNames = new Map(); // basename (lower) -> full path
+    const fileNamesNoExt = new Map(); // basename without ext (lower) -> full path
+    for (const doc of documents) {
+        const base = path.basename(doc.filePath).toLowerCase();
+        const noExt = path.basename(doc.filePath, path.extname(doc.filePath)).toLowerCase();
+        fileNames.set(base, doc.filePath);
+        fileNamesNoExt.set(noExt, doc.filePath);
+    }
+    for (const doc of documents) {
+        const { filePath, content } = doc;
+        // 1. Filename mentions — look for other filenames in this document's text
+        for (const [name, targetPath] of fileNames) {
+            if (targetPath === filePath)
+                continue;
+            const regex = new RegExp(escapeRegex(name), 'gi');
+            let match;
+            while ((match = regex.exec(content)) !== null) {
+                references.push({
+                    sourceFile: filePath,
+                    targetFile: targetPath,
+                    referenceType: 'filename-mention',
+                    context: getContext(content, match.index, match[0].length),
+                    line: getLineNumber(content, match.index),
+                });
+            }
+        }
+        // 2. Hyperlink extraction — markdown links [text](url) and bare URLs with local file references
+        const linkRegex = /\[([^\]]*)\]\(([^)]+)\)/g;
+        let linkMatch;
+        while ((linkMatch = linkRegex.exec(content)) !== null) {
+            const linkTarget = linkMatch[2];
+            // Check if link target references a known file
+            const linkBasename = path.basename(linkTarget).toLowerCase();
+            const resolved = fileNames.get(linkBasename) ?? fileNamesNoExt.get(linkBasename.replace(/\.[^.]+$/, ''));
+            if (resolved && resolved !== filePath) {
+                references.push({
+                    sourceFile: filePath,
+                    targetFile: resolved,
+                    referenceType: 'hyperlink',
+                    context: getContext(content, linkMatch.index, linkMatch[0].length),
+                    line: getLineNumber(content, linkMatch.index),
+                });
+            }
+        }
+        // 3. Citation patterns — "see X", "refer to X", "as described in X"
+        const citationPatterns = [
+            /\b(?:see|refer\s+to|as\s+described\s+in|as\s+noted\s+in|per|according\s+to)\s+["']?([^"'\n,;.]+?)["']?(?=[,;.\s]|$)/gi,
+        ];
+        for (const citationRegex of citationPatterns) {
+            let citMatch;
+            while ((citMatch = citationRegex.exec(content)) !== null) {
+                const mention = citMatch[1].trim();
+                const mentionLower = mention.toLowerCase();
+                // Try to resolve against known files
+                const resolved = fileNames.get(mentionLower)
+                    ?? fileNamesNoExt.get(mentionLower)
+                    ?? fileNamesNoExt.get(mentionLower.replace(/\s+/g, '-'))
+                    ?? fileNamesNoExt.get(mentionLower.replace(/\s+/g, '_'));
+                if (resolved && resolved !== filePath) {
+                    references.push({
+                        sourceFile: filePath,
+                        targetFile: resolved,
+                        referenceType: 'citation',
+                        context: getContext(content, citMatch.index, citMatch[0].length),
+                        line: getLineNumber(content, citMatch.index),
+                    });
+                }
+                else if (mention.length > 3 && mention.length < 100) {
+                    unresolvedMentions.push({
+                        sourceFile: filePath,
+                        mention,
+                        context: getContext(content, citMatch.index, citMatch[0].length),
+                    });
+                }
+            }
+        }
+    }
+    return { references, unresolvedMentions };
+}
+function escapeRegex(str) {
+    return str.replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+}
+//# sourceMappingURL=references.js.map

package/dist/src/doc/references.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"references.js","sourceRoot":"","sources":["../../../src/doc/references.ts"],"names":[],"mappings":"AAAA;;;GAGG;AACH,OAAO,IAAI,MAAM,WAAW,CAAC;AAe7B,MAAM,cAAc,GAAG,EAAE,CAAC;AAE1B,SAAS,UAAU,CAAC,IAAY,EAAE,QAAgB,EAAE,MAAc;IAChE,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,QAAQ,GAAG,cAAc,CAAC,CAAC;IACrD,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,EAAE,QAAQ,GAAG,MAAM,GAAG,cAAc,CAAC,CAAC;IACtE,IAAI,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;IAChE,IAAI,KAAK,GAAG,CAAC;QAAE,OAAO,GAAG,KAAK,GAAG,OAAO,CAAC;IACzC,IAAI,GAAG,GAAG,IAAI,CAAC,MAAM;QAAE,OAAO,GAAG,OAAO,GAAG,KAAK,CAAC;IACjD,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,SAAS,aAAa,CAAC,IAAY,EAAE,QAAgB;IACnD,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,IAAI,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACrD,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI;YAAE,IAAI,EAAE,CAAC;IAC/B,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,gDAAgD;AAChD,MAAM,UAAU,qBAAqB,CACnC,SAAuD;IAEvD,MAAM,UAAU,GAAqB,EAAE,CAAC;IACxC,MAAM,kBAAkB,GAA4C,EAAE,CAAC;IAEvE,8CAA8C;IAC9C,MAAM,SAAS,GAAG,IAAI,GAAG,EAAkB,CAAC,CAAC,gCAAgC;IAC7E,MAAM,cAAc,GAAG,IAAI,GAAG,EAAkB,CAAC,CAAC,4CAA4C;IAC9F,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;QAC5B,MAAM,IAAI,GAAG,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;QACvD,MAAM,KAAK,GAAG,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,QAAQ,EAAE,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC;QACpF,SAAS,CAAC,GAAG,CAAC,IAAI,EAAE,GAAG,CAAC,QAAQ,CAAC,CAAC;QAClC,cAAc,CAAC,GAAG,CAAC,KAAK,EAAE,GAAG,CAAC,QAAQ,CAAC,CAAC;IAC1C,CAAC;IAED,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;QAC5B,MAAM,EAAE,QAAQ,EAAE,OAAO,EAAE,GAAG,GAAG,CAAC;QAElC,0EAA0E;QAC1E,KAAK,MAAM,CAAC,IAAI,EAAE,UAAU,CAAC,IAAI,SAAS,EAAE,CAAC;YAC3C,IAAI,UAAU,KAAK,QAAQ;gBAAE,SAAS;YACtC,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,WAAW,CAAC,IAAI,CAAC,EAAE,IAAI,CAAC,CAAC;YAClD,IAAI,KAA6B,CAAC;YAClC,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBAC9C,UAAU,CAAC,IAAI,CAAC;oBACd,UAAU,EAAE,QAAQ;oBACpB,UAAU,EAAE,UAAU;oBACtB,aAAa,EAAE,kBAAkB;oBACjC,OAAO,EAAE,UAAU,CAAC,OAAO,EAAE,KAAK,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;oBAC1D,IAAI,EAAE,aAAa,CAAC,OAAO,EAAE,KAAK,CAAC,KAAK,CAAC;iBAC1C,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,gGAAgG;QAChG,MAAM,SAAS,GAAG,0BAA0B,CAAC;QAC7C,IAAI,SAAiC,CAAC;QACtC,OAAO,CAAC,SAAS,GAAG,SAAS,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YACtD,MAAM,UAAU,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC;YAChC,+CAA+C;YAC/C,MAAM,YAAY,GAAG,IAAI,CAAC,QAAQ,CAAC,UAAU,CAAC,CAAC,WAAW,EAAE,CAAC;YAC7D,MAAM,QAAQ,GAAG,SAAS,CAAC,GAAG,CAAC,YAAY,CAAC,IAAI,cAAc,CAAC,GAAG,CAAC,YAAY,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC;YACzG,IAAI,QAAQ,IAAI,QAAQ,KAAK,QAAQ,EAAE,CAAC;gBACtC,UAAU,CAAC,IAAI,CAAC;oBACd,UAAU,EAAE,QAAQ;oBACpB,UAAU,EAAE,QAAQ;oBACpB,aAAa,EAAE,WAAW;oBAC1B,OAAO,EAAE,UAAU,CAAC,OAAO,EAAE,SAAS,CAAC,KAAK,EAAE,SAAS,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;oBAClE,IAAI,EAAE,aAAa,CAAC,OAAO,EAAE,SAAS,CAAC,KAAK,CAAC;iBAC9C,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,oEAAoE;QACpE,MAAM,gBAAgB,GAAG;YACvB,uHAAuH;SACxH,CAAC;QACF,KAAK,MAAM,aAAa,IAAI,gBAAgB,EAAE,CAAC;YAC7C,IAAI,QAAgC,CAAC;YACrC,OAAO,CAAC,QAAQ,GAAG,aAAa,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBACzD,MAAM,OAAO,GAAG,QAAQ,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;gBACnC,MAAM,YAAY,GAAG,OAAO,CAAC,WAAW,EAAE,CAAC;gBAC3C,qCAAqC;gBACrC,MAAM,QAAQ,GAAG,SAAS,CAAC,GAAG,CAAC,YAAY,CAAC;uBACvC,cAAc,CAAC,GAAG,CAAC,YAAY,CAAC;uBAChC,cAAc,CAAC,GAAG,CAAC,YAAY,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC;uBACrD,cAAc,CAAC,GAAG,CAAC,YAAY,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,CAAC;gBAE3D,IAAI,QAAQ,IAAI,QAAQ,KAAK,QAAQ,EAAE,CAAC;oBACtC,UAAU,CAAC,IAAI,CAAC;wBACd,UAAU,EAAE,QAAQ;wBACpB,UAAU,EAAE,QAAQ;wBACpB,aAAa,EAAE,UAAU;wBACzB,OAAO,EAAE,UAAU,CAAC,OAAO,EAAE,QAAQ,CAAC,KAAK,EAAE,QAAQ,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;wBAChE,IAAI,EAAE,aAAa,CAAC,OAAO,EAAE,QAAQ,CAAC,KAAK,CAAC;qBAC7C,CAAC,CAAC;gBACL,CAAC;qBAAM,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,IAAI,OAAO,CAAC,MAAM,GAAG,GAAG,EAAE,CAAC;oBACtD,kBAAkB,CAAC,IAAI,CAAC;wBACtB,UAAU,EAAE,QAAQ;wBACpB,OAAO;wBACP,OAAO,EAAE,UAAU,CAAC,OAAO,EAAE,QAAQ,CAAC,KAAK,EAAE,QAAQ,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;qBACjE,CAAC,CAAC;gBACL,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,EAAE,UAAU,EAAE,kBAAkB,EAAE,CAAC;AAC5C,CAAC;AAED,SAAS,WAAW,CAAC,GAAW;IAC9B,OAAO,GAAG,CAAC,OAAO,CAAC,qBAAqB,EAAE,MAAM,CAAC,CAAC;AACpD,CAAC"}

package/dist/src/inspect/shared.js

@@ -1,4 +1,5 @@
 import { z } from 'zod';
+import { estimateTokenCount } from '../tokens.js';
 export const DocumentPropertiesSchema = z.object({
     title: z.string().optional(),
     subject: z.string().optional(),
@@ -10,8 +11,7 @@ export const DocumentPropertiesSchema = z.object({
     keywords: z.string().optional(),
 });
 export function estimateTokens(input) {
-    const chars = typeof input === 'number' ? input : input.length;
-    return Math.max(0, Math.ceil(chars / 4));
+    return estimateTokenCount(input);
 }
 export function asOptionalString(value) {
     return typeof value === 'string' && value.trim() ? value : undefined;

package/dist/src/inspect/shared.js.map

@@ -1 +1 @@
-{"version":3,"file":"shared.js","sourceRoot":"","sources":["../../../src/inspect/shared.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,CAAC,EAAE,MAAM,KAAK,CAAC;AAExB,MAAM,CAAC,MAAM,wBAAwB,GAAG,CAAC,CAAC,MAAM,CAAC;IAC/C,KAAK,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC5B,OAAO,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC9B,MAAM,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC7B,OAAO,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC9B,WAAW,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAClC,YAAY,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IACnC,cAAc,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IACrC,QAAQ,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;CAChC,CAAC,CAAC;AAGH,MAAM,UAAU,cAAc,CAAC,KAAsB;IACnD,MAAM,KAAK,GAAG,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,MAAM,CAAC;IAC/D,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC;AAC3C,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,KAAc;IAC7C,OAAO,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,SAAS,CAAC;AACvE,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,KAAc;IAC3C,IAAI,KAAK,YAAY,IAAI;QAAE,OAAO,KAAK,CAAC,WAAW,EAAE,CAAC;IACtD,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,IAAI,EAAE;QAAE,OAAO,KAAK,CAAC;IAC5D,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,MAAM,UAAU,oBAAoB,CAAI,IAAY,EAAE,KAA8B,EAAE,OAAU;IAC9F,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC;AAClC,CAAC"}
+{"version":3,"file":"shared.js","sourceRoot":"","sources":["../../../src/inspect/shared.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,CAAC,EAAE,MAAM,KAAK,CAAC;AACxB,OAAO,EAAE,kBAAkB,EAAE,MAAM,cAAc,CAAC;AAElD,MAAM,CAAC,MAAM,wBAAwB,GAAG,CAAC,CAAC,MAAM,CAAC;IAC/C,KAAK,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC5B,OAAO,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC9B,MAAM,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC7B,OAAO,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAC9B,WAAW,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IAClC,YAAY,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IACnC,cAAc,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;IACrC,QAAQ,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE;CAChC,CAAC,CAAC;AAGH,MAAM,UAAU,cAAc,CAAC,KAAsB;IACnD,OAAO,kBAAkB,CAAC,KAAK,CAAC,CAAC;AACnC,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,KAAc;IAC7C,OAAO,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,SAAS,CAAC;AACvE,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,KAAc;IAC3C,IAAI,KAAK,YAAY,IAAI;QAAE,OAAO,KAAK,CAAC,WAAW,EAAE,CAAC;IACtD,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,IAAI,EAAE;QAAE,OAAO,KAAK,CAAC;IAC5D,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,MAAM,UAAU,oBAAoB,CAAI,IAAY,EAAE,KAA8B,EAAE,OAAU;IAC9F,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC;AAClC,CAAC"}

package/dist/src/markdown/convert.d.ts

@@ -1,2 +1,2 @@
-/** Convert a document to markdown. Returns null for unsupported formats (xlsx, ods). */
+/** Convert a document to markdown. Returns null for unsupported formats (ods). */
 export declare function documentToMarkdown(filePath: string): Promise<string | null>;

package/dist/src/markdown/convert.js

@@ -2,6 +2,7 @@ import { readFile } from 'node:fs/promises';
 import mammoth from 'mammoth';
 import pdf from 'pdf-parse';
 import JSZip from 'jszip';
+import XLSX from 'xlsx';
 import officeparser from 'officeparser';
 import TurndownService from 'turndown';
 import { z } from 'zod';
@@ -100,7 +101,58 @@ async function odpToMarkdown(filePath) {
     }
     return parts.join('\n\n');
 }
-/** Convert a document to markdown. Returns null for unsupported formats (xlsx, ods). */
+/** Convert an XLSX file to markdown with table formatting per sheet */
+async function xlsxToMarkdown(filePath, maxRows = 20) {
+    const buffer = await readFile(filePath);
+    const workbook = XLSX.read(buffer, {
+        type: 'buffer',
+        cellDates: true,
+        cellText: true,
+    });
+    const parts = [];
+    for (const sheetName of workbook.SheetNames) {
+        const sheet = workbook.Sheets[sheetName];
+        if (!sheet?.['!ref'])
+            continue;
+        const range = XLSX.utils.decode_range(sheet['!ref']);
+        const totalRows = range.e.r - range.s.r + 1;
+        const totalCols = range.e.c - range.s.c + 1;
+        parts.push(`## ${sheetName}`);
+        parts.push('');
+        if (totalCols === 0 || totalRows === 0) {
+            parts.push('*Empty sheet*');
+            parts.push('');
+            continue;
+        }
+        // Read header row
+        const headers = [];
+        for (let col = range.s.c; col <= range.e.c; col++) {
+            const cell = sheet[XLSX.utils.encode_cell({ r: range.s.r, c: col })];
+            const raw = cell?.v != null ? String(cell.w ?? cell.v) : XLSX.utils.encode_col(col);
+            headers.push(raw.replace(/\|/g, '\\|').replace(/\n/g, ' '));
+        }
+        // Build markdown table
+        parts.push(`| ${headers.join(' | ')} |`);
+        parts.push(`| ${headers.map(() => '---').join(' | ')} |`);
+        const dataRows = Math.min(maxRows, totalRows - 1);
+        for (let row = range.s.r + 1; row <= range.s.r + dataRows; row++) {
+            const values = [];
+            for (let col = range.s.c; col <= range.e.c; col++) {
+                const cell = sheet[XLSX.utils.encode_cell({ r: row, c: col })];
+                const val = cell?.v != null ? String(cell.w ?? cell.v) : '';
+                values.push(val.replace(/\|/g, '\\|').replace(/\n/g, ' '));
+            }
+            parts.push(`| ${values.join(' | ')} |`);
+        }
+        if (totalRows - 1 > maxRows) {
+            parts.push('');
+            parts.push(`*...${totalRows - 1 - maxRows} more rows*`);
+        }
+        parts.push('');
+    }
+    return parts.join('\n');
+}
+/** Convert a document to markdown. Returns null for unsupported formats (ods). */
 export async function documentToMarkdown(filePath) {
     const ext = getExtension(filePath);
     switch (ext) {
@@ -109,7 +161,7 @@ export async function documentToMarkdown(filePath) {
         case 'pptx': return pptxToMarkdown(filePath);
         case 'odt': return odtToMarkdown(filePath);
         case 'odp': return odpToMarkdown(filePath);
-        case 'xlsx':
+        case 'xlsx': return xlsxToMarkdown(filePath);
         case 'ods':
             return null;
         default:

package/dist/src/markdown/convert.js.map

@@ -1 +1 @@
-{"version":3,"file":"convert.js","sourceRoot":"","sources":["../../../src/markdown/convert.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,OAAO,MAAM,SAAS,CAAC;AAC9B,OAAO,GAAG,MAAM,WAAW,CAAC;AAC5B,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,YAAY,MAAM,cAAc,CAAC;AACxC,OAAO,eAAe,MAAM,UAAU,CAAC;AACvC,OAAO,EAAE,CAAC,EAAE,MAAM,KAAK,CAAC;AACxB,OAAO,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAE3C,MAAM,QAAQ,GAAG,IAAI,eAAe,CAAC,EAAE,YAAY,EAAE,KAAK,EAAE,CAAC,CAAC;AAC9D,MAAM,oBAAoB,GAAG,CAAC,CAAC,MAAM,CAAC,EAAE,KAAK,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,WAAW,EAAE,CAAC,EAAE,CAAC,CAAC;AAEvG,oEAAoE;AACpE,KAAK,UAAU,cAAc,CAAC,QAAgB;IAC5C,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,aAAa,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,CAAC,CAAC;IAC/D,MAAM,IAAI,GAAG,MAAM,CAAC,KAAK,IAAI,EAAE,CAAC;IAChC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE;QAAE,OAAO,EAAE,CAAC;IAC5B,OAAO,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;AACjC,CAAC;AAED,sDAAsD;AACtD,KAAK,UAAU,aAAa,CAAC,QAAgB;IAC3C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IAExC,2BAA2B;IAC3B,MAAM,WAAW,GAAG,OAAO,CAAC,GAAG,CAAC;IAChC,OAAO,CAAC,GAAG,GAAG,CAAC,GAAG,IAAe,EAAE,EAAE;QACnC,IAAI,OAAO,IAAI,CAAC,CAAC,CAAC,KAAK,QAAQ,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,WAAW,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,QAAQ,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,wBAAwB,CAAC,CAAC,EAAE,CAAC;YACrJ,OAAO;QACT,CAAC;QACD,WAAW,CAAC,KAAK,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC;IACnC,CAAC,CAAC;IAEF,IAAI,IAAwC,CAAC;IAC7C,IAAI,CAAC;QACH,IAAI,GAAG,MAAM,GAAG,CAAC,MAAM,EAAE;YACvB,UAAU,EAAE,KAAK,EAAE,QAAuE,EAAE,EAAE;gBAC5F,MAAM,WAAW,GAAG,oBAAoB,CAAC,KAAK,CAAC,MAAM,QAAQ,CAAC,cAAc,EAAE,CAAC,CAAC;gBAChF,MAAM,OAAO,GAAG,WAAW,CAAC,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;gBACxD,OAAO,SAAS,QAAQ,CAAC,SAAS,GAAG,CAAC,MAAM,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAClE,CAAC;SACF,CAAC,CAAC;IACL,CAAC;YAAS,CAAC;QACT,OAAO,CAAC,GAAG,GAAG,WAAW,CAAC;IAC5B,CAAC;IAED,OAAO,IAAI,CAAC,IAAI,CAAC;AACnB,CAAC;AAED,oDAAoD;AACpD,KAAK,UAAU,cAAc,CAAC,QAAgB;IAC5C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACxC,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IAE1C,uCAAuC;IACvC,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC;SACtC,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,8BAA8B,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;SACzD,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;QACb,MAAM,IAAI,GAAG,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,YAAY,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,GAAG,EAAE,EAAE,CAAC,CAAC;QAC7D,MAAM,IAAI,GAAG,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,YAAY,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,GAAG,EAAE,EAAE,CAAC,CAAC;QAC7D,OAAO,IAAI,GAAG,IAAI,CAAC;IACrB,CAAC,CAAC,CAAC;IAEL,iCAAiC;IACjC,MAAM,QAAQ,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,MAAM,YAAY,CAAC,WAAW,CAAC,MAAM,CAAC,CAAC,CAAC;IAC1E,MAAM,UAAU,GAAG,UAAU,CAAC,MAAM,CAAC;IAErC,IAAI,UAAU,IAAI,CAAC,EAAE,CAAC;QACpB,OAAO,gBAAgB,QAAQ,EAAE,CAAC;IACpC,CAAC;IAED,oCAAoC;IACpC,MAAM,KAAK,GAAG,QAAQ,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IACnC,MAAM,aAAa,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,UAAU,CAAC,CAAC,CAAC;IACxE,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,UAAU,EAAE,CAAC,EAAE,EAAE,CAAC;QACpC,MAAM,KAAK,GAAG,CAAC,GAAG,aAAa,CAAC;QAChC,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,aAAa,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;QAC5D,MAAM,SAAS,GAAG,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;QAC5D,KAAK,CAAC,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,OAAO,SAAS,EAAE,CAAC,CAAC;IACjD,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;AAC5B,CAAC;AAED,sEAAsE;AACtE,KAAK,UAAU,aAAa,CAAC,QAAgB;IAC3C,MAAM,IAAI,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,MAAM,YAAY,CAAC,WAAW,CAAC,QAAQ,CAAC,CAAC,CAAC;IACxE,OAAO,IAAI,CAAC;AACd,CAAC;AAED,yDAAyD;AACzD,KAAK,UAAU,aAAa,CAAC,QAAgB;IAC3C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACxC,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IAC1C,MAAM,UAAU,GAAG,MAAM,GAAG,CAAC,IAAI,CAAC,aAAa,CAAC,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;IAChE,IAAI,CAAC,UAAU;QAAE,OAAO,EAAE,CAAC;IAE3B,MAAM,MAAM,GAAG,CAAC,UAAU,CAAC,KAAK,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC;IAC/D,MAAM,IAAI,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,MAAM,YAAY,CAAC,WAAW,CAAC,MAAM,CAAC,CAAC,CAAC;IAEtE,IAAI,MAAM,IAAI,CAAC,EAAE,CAAC;QAChB,OAAO,gBAAgB,IAAI,EAAE,CAAC;IAChC,CAAC;IAED,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAC/B,MAAM,aAAa,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,MAAM,CAAC,CAAC,CAAC;IACpE,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAChC,MAAM,KAAK,GAAG,CAAC,GAAG,aAAa,CAAC;QAChC,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,aAAa,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;QAC5D,MAAM,SAAS,GAAG,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;QAC5D,KAAK,CAAC,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,OAAO,SAAS,EAAE,CAAC,CAAC;IACjD,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;AAC5B,CAAC;AAED,wFAAwF;AACxF,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IACvD,MAAM,GAAG,GAAG,YAAY,CAAC,QAAQ,CAAC,CAAC;IAEnC,QAAQ,GAAG,EAAE,CAAC;QACZ,KAAK,MAAM,CAAC,CAAC,OAAO,cAAc,CAAC,QAAQ,CAAC,CAAC;QAC7C,KAAK,KAAK,CAAC,CAAC,OAAO,aAAa,CAAC,QAAQ,CAAC,CAAC;QAC3C,KAAK,MAAM,CAAC,CAAC,OAAO,cAAc,CAAC,QAAQ,CAAC,CAAC;QAC7C,KAAK,KAAK,CAAC,CAAC,OAAO,aAAa,CAAC,QAAQ,CAAC,CAAC;QAC3C,KAAK,KAAK,CAAC,CAAC,OAAO,aAAa,CAAC,QAAQ,CAAC,CAAC;QAC3C,KAAK,MAAM,CAAC;QACZ,KAAK,KAAK;YACR,OAAO,IAAI,CAAC;QACd;YACE,OAAO,IAAI,CAAC;IAChB,CAAC;AACH,CAAC"}
+{"version":3,"file":"convert.js","sourceRoot":"","sources":["../../../src/markdown/convert.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,OAAO,MAAM,SAAS,CAAC;AAC9B,OAAO,GAAG,MAAM,WAAW,CAAC;AAC5B,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,IAAI,MAAM,MAAM,CAAC;AACxB,OAAO,YAAY,MAAM,cAAc,CAAC;AACxC,OAAO,eAAe,MAAM,UAAU,CAAC;AACvC,OAAO,EAAE,CAAC,EAAE,MAAM,KAAK,CAAC;AACxB,OAAO,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAE3C,MAAM,QAAQ,GAAG,IAAI,eAAe,CAAC,EAAE,YAAY,EAAE,KAAK,EAAE,CAAC,CAAC;AAC9D,MAAM,oBAAoB,GAAG,CAAC,CAAC,MAAM,CAAC,EAAE,KAAK,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,WAAW,EAAE,CAAC,EAAE,CAAC,CAAC;AAEvG,oEAAoE;AACpE,KAAK,UAAU,cAAc,CAAC,QAAgB;IAC5C,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,aAAa,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,CAAC,CAAC;IAC/D,MAAM,IAAI,GAAG,MAAM,CAAC,KAAK,IAAI,EAAE,CAAC;IAChC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE;QAAE,OAAO,EAAE,CAAC;IAC5B,OAAO,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;AACjC,CAAC;AAED,sDAAsD;AACtD,KAAK,UAAU,aAAa,CAAC,QAAgB;IAC3C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IAExC,2BAA2B;IAC3B,MAAM,WAAW,GAAG,OAAO,CAAC,GAAG,CAAC;IAChC,OAAO,CAAC,GAAG,GAAG,CAAC,GAAG,IAAe,EAAE,EAAE;QACnC,IAAI,OAAO,IAAI,CAAC,CAAC,CAAC,KAAK,QAAQ,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,WAAW,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,QAAQ,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,wBAAwB,CAAC,CAAC,EAAE,CAAC;YACrJ,OAAO;QACT,CAAC;QACD,WAAW,CAAC,KAAK,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC;IACnC,CAAC,CAAC;IAEF,IAAI,IAAwC,CAAC;IAC7C,IAAI,CAAC;QACH,IAAI,GAAG,MAAM,GAAG,CAAC,MAAM,EAAE;YACvB,UAAU,EAAE,KAAK,EAAE,QAAuE,EAAE,EAAE;gBAC5F,MAAM,WAAW,GAAG,oBAAoB,CAAC,KAAK,CAAC,MAAM,QAAQ,CAAC,cAAc,EAAE,CAAC,CAAC;gBAChF,MAAM,OAAO,GAAG,WAAW,CAAC,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;gBACxD,OAAO,SAAS,QAAQ,CAAC,SAAS,GAAG,CAAC,MAAM,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAClE,CAAC;SACF,CAAC,CAAC;IACL,CAAC;YAAS,CAAC;QACT,OAAO,CAAC,GAAG,GAAG,WAAW,CAAC;IAC5B,CAAC;IAED,OAAO,IAAI,CAAC,IAAI,CAAC;AACnB,CAAC;AAED,oDAAoD;AACpD,KAAK,UAAU,cAAc,CAAC,QAAgB;IAC5C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACxC,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IAE1C,uCAAuC;IACvC,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC;SACtC,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,8BAA8B,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;SACzD,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;QACb,MAAM,IAAI,GAAG,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,YAAY,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,GAAG,EAAE,EAAE,CAAC,CAAC;QAC7D,MAAM,IAAI,GAAG,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,YAAY,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,GAAG,EAAE,EAAE,CAAC,CAAC;QAC7D,OAAO,IAAI,GAAG,IAAI,CAAC;IACrB,CAAC,CAAC,CAAC;IAEL,iCAAiC;IACjC,MAAM,QAAQ,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,MAAM,YAAY,CAAC,WAAW,CAAC,MAAM,CAAC,CAAC,CAAC;IAC1E,MAAM,UAAU,GAAG,UAAU,CAAC,MAAM,CAAC;IAErC,IAAI,UAAU,IAAI,CAAC,EAAE,CAAC;QACpB,OAAO,gBAAgB,QAAQ,EAAE,CAAC;IACpC,CAAC;IAED,oCAAoC;IACpC,MAAM,KAAK,GAAG,QAAQ,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IACnC,MAAM,aAAa,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,UAAU,CAAC,CAAC,CAAC;IACxE,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,UAAU,EAAE,CAAC,EAAE,EAAE,CAAC;QACpC,MAAM,KAAK,GAAG,CAAC,GAAG,aAAa,CAAC;QAChC,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,aAAa,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;QAC5D,MAAM,SAAS,GAAG,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;QAC5D,KAAK,CAAC,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,OAAO,SAAS,EAAE,CAAC,CAAC;IACjD,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;AAC5B,CAAC;AAED,sEAAsE;AACtE,KAAK,UAAU,aAAa,CAAC,QAAgB;IAC3C,MAAM,IAAI,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,MAAM,YAAY,CAAC,WAAW,CAAC,QAAQ,CAAC,CAAC,CAAC;IACxE,OAAO,IAAI,CAAC;AACd,CAAC;AAED,yDAAyD;AACzD,KAAK,UAAU,aAAa,CAAC,QAAgB;IAC3C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACxC,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IAC1C,MAAM,UAAU,GAAG,MAAM,GAAG,CAAC,IAAI,CAAC,aAAa,CAAC,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;IAChE,IAAI,CAAC,UAAU;QAAE,OAAO,EAAE,CAAC;IAE3B,MAAM,MAAM,GAAG,CAAC,UAAU,CAAC,KAAK,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC;IAC/D,MAAM,IAAI,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,MAAM,YAAY,CAAC,WAAW,CAAC,MAAM,CAAC,CAAC,CAAC;IAEtE,IAAI,MAAM,IAAI,CAAC,EAAE,CAAC;QAChB,OAAO,gBAAgB,IAAI,EAAE,CAAC;IAChC,CAAC;IAED,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAC/B,MAAM,aAAa,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,MAAM,CAAC,CAAC,CAAC;IACpE,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAChC,MAAM,KAAK,GAAG,CAAC,GAAG,aAAa,CAAC;QAChC,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,aAAa,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;QAC5D,MAAM,SAAS,GAAG,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;QAC5D,KAAK,CAAC,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,OAAO,SAAS,EAAE,CAAC,CAAC;IACjD,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;AAC5B,CAAC;AAED,uEAAuE;AACvE,KAAK,UAAU,cAAc,CAAC,QAAgB,EAAE,OAAO,GAAG,EAAE;IAC1D,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;IACxC,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE;QACjC,IAAI,EAAE,QAAQ;QACd,SAAS,EAAE,IAAI;QACf,QAAQ,EAAE,IAAI;KACf,CAAC,CAAC;IAEH,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,KAAK,MAAM,SAAS,IAAI,QAAQ,CAAC,UAAU,EAAE,CAAC;QAC5C,MAAM,KAAK,GAAG,QAAQ,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC;QACzC,IAAI,CAAC,KAAK,EAAE,CAAC,MAAM,CAAC;YAAE,SAAS;QAE/B,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,YAAY,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;QACrD,MAAM,SAAS,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QAC5C,MAAM,SAAS,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QAE5C,KAAK,CAAC,IAAI,CAAC,MAAM,SAAS,EAAE,CAAC,CAAC;QAC9B,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAEf,IAAI,SAAS,KAAK,CAAC,IAAI,SAAS,KAAK,CAAC,EAAE,CAAC;YACvC,KAAK,CAAC,IAAI,CAAC,eAAe,CAAC,CAAC;YAC5B,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YACf,SAAS;QACX,CAAC;QAED,kBAAkB;QAClB,MAAM,OAAO,GAAa,EAAE,CAAC;QAC7B,KAAK,IAAI,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,EAAE,CAAC;YAClD,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,EAAE,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC;YACrE,MAAM,GAAG,GAAG,IAAI,EAAE,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC;YACpF,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,CAAC;QAC9D,CAAC;QAED,uBAAuB;QACvB,KAAK,CAAC,IAAI,CAAC,KAAK,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;QACzC,KAAK,CAAC,IAAI,CAAC,KAAK,OAAO,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;QAE1D,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,OAAO,EAAE,SAAS,GAAG,CAAC,CAAC,CAAC;QAClD,KAAK,IAAI,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,GAAG,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,QAAQ,EAAE,GAAG,EAAE,EAAE,CAAC;YACjE,MAAM,MAAM,GAAa,EAAE,CAAC;YAC5B,KAAK,IAAI,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,EAAE,EAAE,CAAC;gBAClD,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC;gBAC/D,MAAM,GAAG,GAAG,IAAI,EAAE,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBAC5D,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,CAAC;YAC7D,CAAC;YACD,KAAK,CAAC,IAAI,CAAC,KAAK,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;QAC1C,CAAC;QAED,IAAI,SAAS,GAAG,CAAC,GAAG,OAAO,EAAE,CAAC;YAC5B,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YACf,KAAK,CAAC,IAAI,CAAC,OAAO,SAAS,GAAG,CAAC,GAAG,OAAO,aAAa,CAAC,CAAC;QAC1D,CAAC;QAED,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACjB,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAED,kFAAkF;AAClF,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IACvD,MAAM,GAAG,GAAG,YAAY,CAAC,QAAQ,CAAC,CAAC;IAEnC,QAAQ,GAAG,EAAE,CAAC;QACZ,KAAK,MAAM,CAAC,CAAC,OAAO,cAAc,CAAC,QAAQ,CAAC,CAAC;QAC7C,KAAK,KAAK,CAAC,CAAC,OAAO,aAAa,CAAC,QAAQ,CAAC,CAAC;QAC3C,KAAK,MAAM,CAAC,CAAC,OAAO,cAAc,CAAC,QAAQ,CAAC,CAAC;QAC7C,KAAK,KAAK,CAAC,CAAC,OAAO,aAAa,CAAC,QAAQ,CAAC,CAAC;QAC3C,KAAK,KAAK,CAAC,CAAC,OAAO,aAAa,CAAC,QAAQ,CAAC,CAAC;QAC3C,KAAK,MAAM,CAAC,CAAC,OAAO,cAAc,CAAC,QAAQ,CAAC,CAAC;QAC7C,KAAK,KAAK;YACR,OAAO,IAAI,CAAC;QACd;YACE,OAAO,IAAI,CAAC;IAChB,CAAC;AACH,CAAC"}

package/dist/src/output/json.js

@@ -14,6 +14,15 @@ export function formatJson(stats, sccData = null, structureResults) {
                 entry[f] = r[f] || 0;
         }
         entry.size = r.size;
+        if (stats.showConfidence && r.confidence) {
+            const conf = {};
+            for (const f of METRIC_FIELDS) {
+                if (r[hasKey(f)] && r.confidence[f])
+                    conf[f] = r.confidence[f];
+            }
+            if (Object.keys(conf).length > 0)
+                entry.confidence = conf;
+        }
         return entry;
     };
     const mapTotals = (t) => {

package/dist/src/output/json.js.map

@@ -1 +1 @@
-{"version":3,"file":"json.js","sourceRoot":"","sources":["../../../src/output/json.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,MAAM,EAAE,MAAM,aAAa,CAAC;AACpD,OAAO,EAAE,mBAAmB,EAAE,MAAM,WAAW,CAAC;AAKhD,MAAM,UAAU,UAAU,CACxB,KAAsB,EACtB,UAAgC,IAAI,EACpC,gBAAoC;IAEpC,MAAM,EAAE,OAAO,EAAE,GAAG,KAAK,CAAC;IAE1B,MAAM,MAAM,GAAG,CAAC,CAAW,EAAE,EAAE;QAC7B,MAAM,KAAK,GAA4B;YACrC,IAAI,EAAE,CAAC,CAAC,QAAQ;YAChB,GAAG,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC3C,GAAG,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC3C,KAAK,EAAE,CAAC,CAAC,KAAK;SACf,CAAC;QACF,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;YAC9B,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;gBAAE,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QACzC,CAAC;QACD,KAAK,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC;QACpB,OAAO,KAAK,CAAC;IACf,CAAC,CAAC;IAEF,MAAM,SAAS,GAAG,CAAC,CAAW,EAAE,EAAE;QAChC,MAAM,KAAK,GAA4B,EAAE,KAAK,EAAE,CAAC,CAAC,KAAK,EAAE,CAAC;QAC1D,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;YAC9B,IAAI,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;gBAAE,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QAC1C,CAAC;QACD,KAAK,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC;QACpB,OAAO,KAAK,CAAC;IACf,CAAC,CAAC;IAEF,MAAM,MAAM,GAA4B;QACtC,SAAS,EAAE;YACT,KAAK,EAAE,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC;YAC7B,MAAM,EAAE,SAAS,CAAC,KAAK,CAAC,MAAM,CAAC;SAChC;KACF,CAAC;IAEF,IAAI,OAAO,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAClC,MAAM,CAAC,IAAI,GAAG,OAAO,CAAC;IACxB,CAAC;IAED,IAAI,gBAAgB,IAAI,gBAAgB,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACpD,MAAM,CAAC,UAAU,GAAG,mBAAmB,CAAC,gBAAgB,CAAC,CAAC;IAC5D,CAAC;IAED,OAAO,IAAI,CAAC,SAAS,CAAC,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC;AACzC,CAAC"}
+{"version":3,"file":"json.js","sourceRoot":"","sources":["../../../src/output/json.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,MAAM,EAAE,MAAM,aAAa,CAAC;AACpD,OAAO,EAAE,mBAAmB,EAAE,MAAM,WAAW,CAAC;AAKhD,MAAM,UAAU,UAAU,CACxB,KAAsB,EACtB,UAAgC,IAAI,EACpC,gBAAoC;IAEpC,MAAM,EAAE,OAAO,EAAE,GAAG,KAAK,CAAC;IAE1B,MAAM,MAAM,GAAG,CAAC,CAAW,EAAE,EAAE;QAC7B,MAAM,KAAK,GAA4B;YACrC,IAAI,EAAE,CAAC,CAAC,QAAQ;YAChB,GAAG,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC3C,GAAG,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC3C,KAAK,EAAE,CAAC,CAAC,KAAK;SACf,CAAC;QACF,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;YAC9B,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;gBAAE,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QACzC,CAAC;QACD,KAAK,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC;QACpB,IAAI,KAAK,CAAC,cAAc,IAAI,CAAC,CAAC,UAAU,EAAE,CAAC;YACzC,MAAM,IAAI,GAA2B,EAAE,CAAC;YACxC,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;gBAC9B,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC;oBAAE,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;YACjE,CAAC;YACD,IAAI,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC;gBAAE,KAAK,CAAC,UAAU,GAAG,IAAI,CAAC;QAC5D,CAAC;QACD,OAAO,KAAK,CAAC;IACf,CAAC,CAAC;IAEF,MAAM,SAAS,GAAG,CAAC,CAAW,EAAE,EAAE;QAChC,MAAM,KAAK,GAA4B,EAAE,KAAK,EAAE,CAAC,CAAC,KAAK,EAAE,CAAC;QAC1D,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;YAC9B,IAAI,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;gBAAE,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QAC1C,CAAC;QACD,KAAK,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC;QACpB,OAAO,KAAK,CAAC;IACf,CAAC,CAAC;IAEF,MAAM,MAAM,GAA4B;QACtC,SAAS,EAAE;YACT,KAAK,EAAE,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC;YAC7B,MAAM,EAAE,SAAS,CAAC,KAAK,CAAC,MAAM,CAAC;SAChC;KACF,CAAC;IAEF,IAAI,OAAO,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAClC,MAAM,CAAC,IAAI,GAAG,OAAO,CAAC;IACxB,CAAC;IAED,IAAI,gBAAgB,IAAI,gBAAgB,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACpD,MAAM,CAAC,UAAU,GAAG,mBAAmB,CAAC,gBAAgB,CAAC,CAAC;IAC5D,CAAC;IAED,OAAO,IAAI,CAAC,SAAS,CAAC,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC;AACzC,CAAC"}

package/dist/src/output/tabular.js

@@ -19,9 +19,9 @@ export function formatDocumentTable(stats, options = {}) {
         colAligns,
     });
     for (const row of stats.rows) {
-        table.push(buildRow(row, stats.columns, isByFile, c));
+        table.push(buildRow(row, stats.columns, isByFile, c, false, stats.showConfidence));
     }
-    table.push(buildRow(stats.totals, stats.columns, isByFile, c, true));
+    table.push(buildRow(stats.totals, stats.columns, isByFile, c, true, false));
     const tableStr = addSeparators(table.toString(), ci ? '-' : '─');
     const tableWidth = stripAnsi(tableStr.split('\n')[0]).length;
     const lines = [];
@@ -33,6 +33,9 @@ export function formatDocumentTable(stats, options = {}) {
     if (hasEstimatedPages) {
         lines.push(c.dim('* Word/ODT pages estimated at 250 words/page'));
     }
+    if (stats.showConfidence) {
+        lines.push(c.dim('~ estimated metric'));
+    }
     return lines.join('\n');
 }
 export function formatSccTable(sccData, options = {}) {
@@ -181,7 +184,12 @@ function buildColAligns(columns, byFile) {
     aligns.push('right');
     return aligns;
 }
-function buildRow(row, columns, byFile, c, isTotal = false) {
+function annotateConfidence(value, metric, row, showConfidence) {
+    if (!showConfidence || !row.confidence || row.confidence[metric] !== 'estimated')
+        return value;
+    return value + '~';
+}
+function buildRow(row, columns, byFile, c, isTotal = false, showConfidence = false) {
     const fmt = isTotal ? c.total : (v) => v;
     const fmtType = isTotal ? c.total : c.type;
     const fmtNum = isTotal ? c.total : c.number;
@@ -200,21 +208,21 @@ function buildRow(row, columns, byFile, c, isTotal = false) {
     if (!byFile)
         cells.push(fmt(formatNumber(row.files)));
     if (columns.hasWords)
-        cells.push(fmtNum(row.words ? formatNumber(row.words) : ''));
+        cells.push(fmtNum(row.words ? annotateConfidence(formatNumber(row.words), 'words', row, showConfidence) : ''));
     if (columns.hasPages)
-        cells.push(fmtNum(row.pages ? formatNumber(row.pages) : ''));
+        cells.push(fmtNum(row.pages ? annotateConfidence(formatNumber(row.pages), 'pages', row, showConfidence) : ''));
     if (hasExtraColumns(columns)) {
         const parts = [];
         if (row.paragraphs)
-            parts.push(`${formatNumber(row.paragraphs)} paras`);
+            parts.push(`${annotateConfidence(formatNumber(row.paragraphs), 'paragraphs', row, showConfidence)} paras`);
         if (row.sheets)
-            parts.push(`${formatNumber(row.sheets)} sheets`);
+            parts.push(`${annotateConfidence(formatNumber(row.sheets), 'sheets', row, showConfidence)} sheets`);
         if (row.rows)
-            parts.push(`${formatNumber(row.rows)} rows`);
+            parts.push(`${annotateConfidence(formatNumber(row.rows), 'rows', row, showConfidence)} rows`);
         if (row.cells)
-            parts.push(`${formatNumber(row.cells)} cells`);
+            parts.push(`${annotateConfidence(formatNumber(row.cells), 'cells', row, showConfidence)} cells`);
         if (row.slides)
-            parts.push(`${formatNumber(row.slides)} slides`);
+            parts.push(`${annotateConfidence(formatNumber(row.slides), 'slides', row, showConfidence)} slides`);
         cells.push(fmt(parts.join(', ')));
     }
     cells.push(fmt(formatBytes(row.size)));