@zaowu-labs/doc 1.0.6 → 1.1.0

package/README.md

@@ -19,7 +19,7 @@ canvas warning during the dynamic import. This is a contained, documented
 workaround.
 
 Long-term we would prefer a pure-JS text-only PDF extractor that does not
-require these shims. See the source comments in `src/index.ts` (functions
+require these shims. See the source comments in `src/pdf.ts` (functions
 `ensurePdfTextRuntime` and `importPdfParse`) for the full explanation and
 the class names used for the shims.
 
@@ -28,3 +28,30 @@ Safety:
 - Document conversion previews by default when an output path is supplied.
 - Confirmed writes refuse unsafe overwrites.
 - Format support should stay explicit in command help and docs.
+
+## Extraction contract and limits
+
+DOCX uses the existing focused JSZip and xmldom libraries directly, removing
+Mammoth and its unused CLI dependency chain. It preserves main-body paragraphs,
+heading styles (including inherited/strict namespaces), list structure, tables,
+and HTTP(S)/mailto hyperlinks. It never follows external relationships, extracts
+archives to disk, or runs macros. Literal Word text is escaped for Markdown.
+Footnotes, images/OCR, complex layout, embedded objects, and tracked deletions
+are outside this text extraction contract. Re-export complex documents as text
+when exact visual fidelity matters.
+
+Input files are limited to 20 MiB; requested DOCX XML parts to 8 MiB each and
+2,000 archive entries; extracted text to two million characters. DTD/entity
+XML and excessive nesting are rejected. XML budgets apply to actual inflated
+bytes. PDF extraction with no text layer explains how to use OCR rather than
+returning a misleading successful empty summary. Concurrent PDF extraction
+serializes temporary runtime shims and restores the original globals.
+
+`summary` explicitly reports `method: "extractive"`: it takes the first three
+prose lines, without calling a model. Frontmatter, fenced code, and headings do
+not become summary prose; fenced examples do not create false outline headings.
+See [the report-to-model recipe](../../docs/WORKFLOWS.md).
+
+Confirmation execution also checks the freshly prepared content against the
+already verified plan immediately before the write/request. Changes between
+reads fail with `OPERATION_PLAN_MISMATCH` and produce no write or network call.

package/dist/docx.d.ts

@@ -0,0 +1,2 @@
+export declare const extractDocx: (buffer: Buffer, format: "markdown" | "text") => Promise<string>;
+//# sourceMappingURL=docx.d.ts.map

package/dist/docx.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"docx.d.ts","sourceRoot":"","sources":["../src/docx.ts"],"names":[],"mappings":"AA+FA,eAAO,MAAM,WAAW,GAAU,QAAQ,MAAM,EAAE,QAAQ,UAAU,GAAG,MAAM,KAAG,OAAO,CAAC,MAAM,CAmI7F,CAAC"}

package/dist/docx.js

@@ -0,0 +1,226 @@
+import JSZip from 'jszip';
+import { DOMParser } from '@xmldom/xmldom';
+import { ZaoWuError } from '@zaowu-labs/core';
+const WORD_NAMESPACES = new Set([
+    'http://schemas.openxmlformats.org/wordprocessingml/2006/main',
+    'http://purl.oclc.org/ooxml/wordprocessingml/main',
+]);
+const MAX_XML_BYTES = 8 * 1024 * 1024;
+const MAX_DEPTH = 100;
+const invalidDocx = () => new Error('Invalid or unsupported DOCX XML.');
+const tooLarge = () => new ZaoWuError({
+    code: 'DOCUMENT_INPUT_TOO_LARGE',
+    message: 'Document is too large to extract safely.',
+    why: 'A DOCX XML part exceeds 8 MiB, the archive exceeds 2,000 entries, or XML nesting exceeds 100 levels.',
+    fix: 'Split the document or export a smaller text-only copy.',
+});
+const children = (element) => {
+    const result = [];
+    for (let node = element.firstChild; node; node = node.nextSibling) {
+        if (node.nodeType === 1)
+            result.push(node);
+    }
+    return result;
+};
+const word = (node, name) => WORD_NAMESPACES.has(node.namespaceURI ?? '') && node.localName === name;
+const child = (node, name) => node && children(node).find((item) => word(item, name));
+const attr = (node, name) => {
+    if (!node)
+        return '';
+    for (const ns of WORD_NAMESPACES) {
+        if (node.hasAttributeNS(ns, name))
+            return node.getAttributeNS(ns, name) ?? '';
+    }
+    return '';
+};
+/** Pause the decompressor on actual inflated bytes, not ZIP size claims. */
+const readXml = async (zip, name, required = false) => {
+    const entry = zip.file(name);
+    if (!entry) {
+        if (required)
+            throw invalidDocx();
+        return undefined;
+    }
+    const xml = await new Promise((resolve, reject) => {
+        let size = 0;
+        const chunks = [];
+        const stream = entry.nodeStream('nodebuffer');
+        stream.on('data', (chunk) => {
+            size += chunk.length;
+            if (size > MAX_XML_BYTES) {
+                stream.pause();
+                reject(tooLarge());
+                return;
+            }
+            chunks.push(chunk);
+        });
+        stream.on('error', reject);
+        stream.on('end', () => resolve(Buffer.concat(chunks).toString('utf8')));
+        stream.resume();
+    });
+    if (/<!\s*(?:DOCTYPE|ENTITY)\b/i.test(xml))
+        throw invalidDocx();
+    let malformed = false;
+    const document = new DOMParser({
+        errorHandler: () => {
+            malformed = true;
+        },
+    }).parseFromString(xml, 'application/xml');
+    if (malformed || !document.documentElement)
+        throw invalidDocx();
+    const stack = [[document.documentElement, 1]];
+    while (stack.length) {
+        const [element, depth] = stack.pop();
+        if (depth > MAX_DEPTH)
+            throw tooLarge();
+        for (const elementChild of children(element))
+            stack.push([elementChild, depth + 1]);
+    }
+    return document;
+};
+const safeLink = (target) => {
+    try {
+        const url = new URL(target);
+        if (!['https:', 'http:', 'mailto:'].includes(url.protocol))
+            return undefined;
+        if (url.username || url.password)
+            return undefined;
+        return url.toString().replaceAll('(', '%28').replaceAll(')', '%29');
+    }
+    catch {
+        return undefined;
+    }
+};
+export const extractDocx = async (buffer, format) => {
+    const zip = await JSZip.loadAsync(buffer);
+    if (Object.keys(zip.files).length > 2_000)
+        throw tooLarge();
+    // Read only known parts. Never follow internal paths or external relationships.
+    const document = await readXml(zip, 'word/document.xml', true);
+    const styles = await readXml(zip, 'word/styles.xml');
+    const numbering = await readXml(zip, 'word/numbering.xml');
+    const relationships = await readXml(zip, 'word/_rels/document.xml.rels');
+    const root = document.documentElement;
+    if (!word(root, 'document'))
+        throw invalidDocx();
+    const body = child(root, 'body');
+    if (!body)
+        throw invalidDocx();
+    const links = new Map();
+    if (relationships) {
+        for (const rel of children(relationships.documentElement)) {
+            if (rel.localName !== 'Relationship' || rel.getAttribute('TargetMode') !== 'External')
+                continue;
+            if (!(rel.getAttribute('Type') ?? '').endsWith('/hyperlink'))
+                continue;
+            const target = safeLink(rel.getAttribute('Target') ?? '');
+            if (target)
+                links.set(rel.getAttribute('Id') ?? '', target);
+        }
+    }
+    const styleMap = new Map();
+    if (styles) {
+        for (const style of children(styles.documentElement)) {
+            if (!word(style, 'style'))
+                continue;
+            const id = attr(style, 'styleId');
+            const outline = attr(child(child(style, 'pPr'), 'outlineLvl'), 'val');
+            const name = attr(child(style, 'name'), 'val');
+            const match = /^(?:heading|标题)\s*([1-6])$/i.exec(name || id);
+            const level = outline && /^[0-5]$/.test(outline)
+                ? Number(outline) + 1
+                : match
+                    ? Number(match[1])
+                    : undefined;
+            styleMap.set(id, { level, basedOn: attr(child(style, 'basedOn'), 'val') });
+        }
+    }
+    const styleLevel = (id, seen = new Set()) => {
+        if (seen.has(id) || seen.size >= MAX_DEPTH)
+            return undefined;
+        seen.add(id);
+        const style = styleMap.get(id);
+        return (style?.level ??
+            (style?.basedOn ? styleLevel(style.basedOn, seen) : undefined) ??
+            (/^Heading([1-6])$/i.test(id) ? Number(id.slice(-1)) : undefined));
+    };
+    const numIds = new Map();
+    const numFormats = new Map();
+    if (numbering) {
+        for (const item of children(numbering.documentElement)) {
+            if (word(item, 'num'))
+                numIds.set(attr(item, 'numId'), attr(child(item, 'abstractNumId'), 'val'));
+            if (word(item, 'abstractNum')) {
+                for (const level of children(item).filter((node) => word(node, 'lvl'))) {
+                    numFormats.set(`${attr(item, 'abstractNumId')}:${attr(level, 'ilvl')}`, attr(child(level, 'numFmt'), 'val'));
+                }
+            }
+        }
+    }
+    const inline = (element) => {
+        if (word(element, 'del') || word(element, 'instrText') || word(element, 'pPr'))
+            return '';
+        if (word(element, 't')) {
+            const text = element.textContent ?? '';
+            return format === 'markdown' ? text.replace(/[\\`*_[\]<>#]/g, '\\$&') : text;
+        }
+        if (word(element, 'tab'))
+            return '\t';
+        if (word(element, 'br') || word(element, 'cr'))
+            return '\n';
+        let text = children(element).map(inline).join('');
+        if (format === 'markdown' && word(element, 'hyperlink')) {
+            const id = element.getAttribute('r:id') ||
+                element.getAttributeNS('http://schemas.openxmlformats.org/officeDocument/2006/relationships', 'id');
+            const target = links.get(id ?? '');
+            if (target && text)
+                text = `[${text}](${target})`;
+        }
+        return text;
+    };
+    const paragraph = (element) => {
+        const text = inline(element).trim();
+        if (!text || format === 'text')
+            return text;
+        const properties = child(element, 'pPr');
+        const outline = attr(child(properties, 'outlineLvl'), 'val');
+        const level = /^[0-5]$/.test(outline)
+            ? Number(outline) + 1
+            : styleLevel(attr(child(properties, 'pStyle'), 'val'));
+        if (level)
+            return `${'#'.repeat(level)} ${text}`;
+        const numPr = child(properties, 'numPr');
+        const numId = attr(child(numPr, 'numId'), 'val');
+        if (numId && numId !== '0') {
+            const ilvl = attr(child(numPr, 'ilvl'), 'val') || '0';
+            const indent = '  '.repeat(Math.min(8, Number(ilvl) || 0));
+            const marker = numFormats.get(`${numIds.get(numId)}:${ilvl}`) === 'bullet' ? '-' : '1.';
+            return `${indent}${marker} ${text}`;
+        }
+        return text;
+    };
+    const blocks = (element) => children(element).flatMap((item) => {
+        if (word(item, 'p'))
+            return [paragraph(item)];
+        if (word(item, 'del'))
+            return [];
+        if (word(item, 'tbl')) {
+            const rows = children(item)
+                .filter((node) => word(node, 'tr'))
+                .map((row) => children(row)
+                .filter((node) => word(node, 'tc'))
+                .map((cell) => blocks(cell).filter(Boolean).join(' ')));
+            if (format === 'text')
+                return rows.map((row) => row.join('\t'));
+            const width = rows.reduce((max, row) => Math.max(max, row.length), 0);
+            const lines = rows.map((row) => `| ${Array.from({ length: width }, (_, i) => (row[i] ?? '').replaceAll('|', '\\|').replaceAll('\n', ' ')).join(' | ')} |`);
+            if (lines.length)
+                lines.splice(1, 0, `| ${Array(width).fill('---').join(' | ')} |`);
+            return lines;
+        }
+        // Structured document tags and inserted content may wrap normal blocks.
+        return blocks(item);
+    });
+    return blocks(body).join('\n').trim();
+};
+//# sourceMappingURL=docx.js.map

package/dist/docx.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"docx.js","sourceRoot":"","sources":["../src/docx.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,EAAE,SAAS,EAAE,MAAM,gBAAgB,CAAC;AAC3C,OAAO,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAE9C,MAAM,eAAe,GAAG,IAAI,GAAG,CAAC;IAC9B,8DAA8D;IAC9D,kDAAkD;CACnD,CAAC,CAAC;AACH,MAAM,aAAa,GAAG,CAAC,GAAG,IAAI,GAAG,IAAI,CAAC;AACtC,MAAM,SAAS,GAAG,GAAG,CAAC;AAEtB,MAAM,WAAW,GAAG,GAAU,EAAE,CAAC,IAAI,KAAK,CAAC,kCAAkC,CAAC,CAAC;AAC/E,MAAM,QAAQ,GAAG,GAAe,EAAE,CAChC,IAAI,UAAU,CAAC;IACb,IAAI,EAAE,0BAA0B;IAChC,OAAO,EAAE,0CAA0C;IACnD,GAAG,EAAE,sGAAsG;IAC3G,GAAG,EAAE,wDAAwD;CAC9D,CAAC,CAAC;AAEL,MAAM,QAAQ,GAAG,CAAC,OAAgB,EAAa,EAAE;IAC/C,MAAM,MAAM,GAAc,EAAE,CAAC;IAC7B,KAAK,IAAI,IAAI,GAAG,OAAO,CAAC,UAAU,EAAE,IAAI,EAAE,IAAI,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC;QAClE,IAAI,IAAI,CAAC,QAAQ,KAAK,CAAC;YAAE,MAAM,CAAC,IAAI,CAAC,IAAe,CAAC,CAAC;IACxD,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC,CAAC;AACF,MAAM,IAAI,GAAG,CAAC,IAAa,EAAE,IAAY,EAAW,EAAE,CACpD,eAAe,CAAC,GAAG,CAAC,IAAI,CAAC,YAAY,IAAI,EAAE,CAAC,IAAI,IAAI,CAAC,SAAS,KAAK,IAAI,CAAC;AAC1E,MAAM,KAAK,GAAG,CAAC,IAAyB,EAAE,IAAY,EAAuB,EAAE,CAC7E,IAAI,IAAI,QAAQ,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC,CAAC;AAC1D,MAAM,IAAI,GAAG,CAAC,IAAyB,EAAE,IAAY,EAAU,EAAE;IAC/D,IAAI,CAAC,IAAI;QAAE,OAAO,EAAE,CAAC;IACrB,KAAK,MAAM,EAAE,IAAI,eAAe,EAAE,CAAC;QACjC,IAAI,IAAI,CAAC,cAAc,CAAC,EAAE,EAAE,IAAI,CAAC;YAAE,OAAO,IAAI,CAAC,cAAc,CAAC,EAAE,EAAE,IAAI,CAAC,IAAI,EAAE,CAAC;IAChF,CAAC;IACD,OAAO,EAAE,CAAC;AACZ,CAAC,CAAC;AAEF,4EAA4E;AAC5E,MAAM,OAAO,GAAG,KAAK,EACnB,GAAU,EACV,IAAY,EACZ,QAAQ,GAAG,KAAK,EACe,EAAE;IACjC,MAAM,KAAK,GAAG,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAC7B,IAAI,CAAC,KAAK,EAAE,CAAC;QACX,IAAI,QAAQ;YAAE,MAAM,WAAW,EAAE,CAAC;QAClC,OAAO,SAAS,CAAC;IACnB,CAAC;IACD,MAAM,GAAG,GAAG,MAAM,IAAI,OAAO,CAAS,CAAC,OAAO,EAAE,MAAM,EAAE,EAAE;QACxD,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,MAAM,MAAM,GAAa,EAAE,CAAC;QAC5B,MAAM,MAAM,GAAG,KAAK,CAAC,UAAU,CAAC,YAAY,CAAC,CAAC;QAC9C,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,CAAC,KAAa,EAAE,EAAE;YAClC,IAAI,IAAI,KAAK,CAAC,MAAM,CAAC;YACrB,IAAI,IAAI,GAAG,aAAa,EAAE,CAAC;gBACzB,MAAM,CAAC,KAAK,EAAE,CAAC;gBACf,MAAM,CAAC,QAAQ,EAAE,CAAC,CAAC;gBACnB,OAAO;YACT,CAAC;YACD,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACrB,CAAC,CAAC,CAAC;QACH,MAAM,CAAC,EAAE,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;QAC3B,MAAM,CAAC,EAAE,CAAC,KAAK,EAAE,GAAG,EAAE,CAAC,OAAO,CAAC,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;QACxE,MAAM,CAAC,MAAM,EAAE,CAAC;IAClB,CAAC,CAAC,CAAC;IACH,IAAI,4BAA4B,CAAC,IAAI,CAAC,GAAG,CAAC;QAAE,MAAM,WAAW,EAAE,CAAC;IAChE,IAAI,SAAS,GAAG,KAAK,CAAC;IACtB,MAAM,QAAQ,GAAG,IAAI,SAAS,CAAC;QAC7B,YAAY,EAAE,GAAG,EAAE;YACjB,SAAS,GAAG,IAAI,CAAC;QACnB,CAAC;KACF,CAAC,CAAC,eAAe,CAAC,GAAG,EAAE,iBAAiB,CAAC,CAAC;IAC3C,IAAI,SAAS,IAAI,CAAC,QAAQ,CAAC,eAAe;QAAE,MAAM,WAAW,EAAE,CAAC;IAChE,MAAM,KAAK,GAA6B,CAAC,CAAC,QAAQ,CAAC,eAAe,EAAE,CAAC,CAAC,CAAC,CAAC;IACxE,OAAO,KAAK,CAAC,MAAM,EAAE,CAAC;QACpB,MAAM,CAAC,OAAO,EAAE,KAAK,CAAC,GAAG,KAAK,CAAC,GAAG,EAAG,CAAC;QACtC,IAAI,KAAK,GAAG,SAAS;YAAE,MAAM,QAAQ,EAAE,CAAC;QACxC,KAAK,MAAM,YAAY,IAAI,QAAQ,CAAC,OAAO,CAAC;YAAE,KAAK,CAAC,IAAI,CAAC,CAAC,YAAY,EAAE,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC;IACtF,CAAC;IACD,OAAO,QAAQ,CAAC;AAClB,CAAC,CAAC;AAEF,MAAM,QAAQ,GAAG,CAAC,MAAc,EAAsB,EAAE;IACtD,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,CAAC;QAC5B,IAAI,CAAC,CAAC,QAAQ,EAAE,OAAO,EAAE,SAAS,CAAC,CAAC,QAAQ,CAAC,GAAG,CAAC,QAAQ,CAAC;YAAE,OAAO,SAAS,CAAC;QAC7E,IAAI,GAAG,CAAC,QAAQ,IAAI,GAAG,CAAC,QAAQ;YAAE,OAAO,SAAS,CAAC;QACnD,OAAO,GAAG,CAAC,QAAQ,EAAE,CAAC,UAAU,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC,UAAU,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;IACtE,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,SAAS,CAAC;IACnB,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,WAAW,GAAG,KAAK,EAAE,MAAc,EAAE,MAA2B,EAAmB,EAAE;IAChG,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IAC1C,IAAI,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,MAAM,GAAG,KAAK;QAAE,MAAM,QAAQ,EAAE,CAAC;IAC5D,gFAAgF;IAChF,MAAM,QAAQ,GAAG,MAAM,OAAO,CAAC,GAAG,EAAE,mBAAmB,EAAE,IAAI,CAAC,CAAC;IAC/D,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,GAAG,EAAE,iBAAiB,CAAC,CAAC;IACrD,MAAM,SAAS,GAAG,MAAM,OAAO,CAAC,GAAG,EAAE,oBAAoB,CAAC,CAAC;IAC3D,MAAM,aAAa,GAAG,MAAM,OAAO,CAAC,GAAG,EAAE,8BAA8B,CAAC,CAAC;IACzE,MAAM,IAAI,GAAG,QAAS,CAAC,eAAe,CAAC;IACvC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,UAAU,CAAC;QAAE,MAAM,WAAW,EAAE,CAAC;IACjD,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;IACjC,IAAI,CAAC,IAAI;QAAE,MAAM,WAAW,EAAE,CAAC;IAE/B,MAAM,KAAK,GAAG,IAAI,GAAG,EAAkB,CAAC;IACxC,IAAI,aAAa,EAAE,CAAC;QAClB,KAAK,MAAM,GAAG,IAAI,QAAQ,CAAC,aAAa,CAAC,eAAe,CAAC,EAAE,CAAC;YAC1D,IAAI,GAAG,CAAC,SAAS,KAAK,cAAc,IAAI,GAAG,CAAC,YAAY,CAAC,YAAY,CAAC,KAAK,UAAU;gBACnF,SAAS;YACX,IAAI,CAAC,CAAC,GAAG,CAAC,YAAY,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,QAAQ,CAAC,YAAY,CAAC;gBAAE,SAAS;YACvE,MAAM,MAAM,GAAG,QAAQ,CAAC,GAAG,CAAC,YAAY,CAAC,QAAQ,CAAC,IAAI,EAAE,CAAC,CAAC;YAC1D,IAAI,MAAM;gBAAE,KAAK,CAAC,GAAG,CAAC,GAAG,CAAC,YAAY,CAAC,IAAI,CAAC,IAAI,EAAE,EAAE,MAAM,CAAC,CAAC;QAC9D,CAAC;IACH,CAAC;IACD,MAAM,QAAQ,GAAG,IAAI,GAAG,EAA+C,CAAC;IACxE,IAAI,MAAM,EAAE,CAAC;QACX,KAAK,MAAM,KAAK,IAAI,QAAQ,CAAC,MAAM,CAAC,eAAe,CAAC,EAAE,CAAC;YACrD,IAAI,CAAC,IAAI,CAAC,KAAK,EAAE,OAAO,CAAC;gBAAE,SAAS;YACpC,MAAM,EAAE,GAAG,IAAI,CAAC,KAAK,EAAE,SAAS,CAAC,CAAC;YAClC,MAAM,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,KAAK,CAAC,EAAE,YAAY,CAAC,EAAE,KAAK,CAAC,CAAC;YACtE,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,MAAM,CAAC,EAAE,KAAK,CAAC,CAAC;YAC/C,MAAM,KAAK,GAAG,6BAA6B,CAAC,IAAI,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC;YAC7D,MAAM,KAAK,GACT,OAAO,IAAI,SAAS,CAAC,IAAI,CAAC,OAAO,CAAC;gBAChC,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,GAAG,CAAC;gBACrB,CAAC,CAAC,KAAK;oBACL,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;oBAClB,CAAC,CAAC,SAAS,CAAC;YAClB,QAAQ,CAAC,GAAG,CAAC,EAAE,EAAE,EAAE,KAAK,EAAE,OAAO,EAAE,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,SAAS,CAAC,EAAE,KAAK,CAAC,EAAE,CAAC,CAAC;QAC7E,CAAC;IACH,CAAC;IACD,MAAM,UAAU,GAAG,CAAC,EAAU,EAAE,OAAO,IAAI,GAAG,EAAU,EAAsB,EAAE;QAC9E,IAAI,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,IAAI,CAAC,IAAI,IAAI,SAAS;YAAE,OAAO,SAAS,CAAC;QAC7D,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QACb,MAAM,KAAK,GAAG,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QAC/B,OAAO,CACL,KAAK,EAAE,KAAK;YACZ,CAAC,KAAK,EAAE,OAAO,CAAC,CAAC,CAAC,UAAU,CAAC,KAAK,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;YAC9D,CAAC,mBAAmB,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAClE,CAAC;IACJ,CAAC,CAAC;IACF,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;IACzC,MAAM,UAAU,GAAG,IAAI,GAAG,EAAkB,CAAC;IAC7C,IAAI,SAAS,EAAE,CAAC;QACd,KAAK,MAAM,IAAI,IAAI,QAAQ,CAAC,SAAS,CAAC,eAAe,CAAC,EAAE,CAAC;YACvD,IAAI,IAAI,CAAC,IAAI,EAAE,KAAK,CAAC;gBACnB,MAAM,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,EAAE,OAAO,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE,eAAe,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC;YAC7E,IAAI,IAAI,CAAC,IAAI,EAAE,aAAa,CAAC,EAAE,CAAC;gBAC9B,KAAK,MAAM,KAAK,IAAI,QAAQ,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC,EAAE,CAAC;oBACvE,UAAU,CAAC,GAAG,CACZ,GAAG,IAAI,CAAC,IAAI,EAAE,eAAe,CAAC,IAAI,IAAI,CAAC,KAAK,EAAE,MAAM,CAAC,EAAE,EACvD,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,QAAQ,CAAC,EAAE,KAAK,CAAC,CACpC,CAAC;gBACJ,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IACD,MAAM,MAAM,GAAG,CAAC,OAAgB,EAAU,EAAE;QAC1C,IAAI,IAAI,CAAC,OAAO,EAAE,KAAK,CAAC,IAAI,IAAI,CAAC,OAAO,EAAE,WAAW,CAAC,IAAI,IAAI,CAAC,OAAO,EAAE,KAAK,CAAC;YAAE,OAAO,EAAE,CAAC;QAC1F,IAAI,IAAI,CAAC,OAAO,EAAE,GAAG,CAAC,EAAE,CAAC;YACvB,MAAM,IAAI,GAAG,OAAO,CAAC,WAAW,IAAI,EAAE,CAAC;YACvC,OAAO,MAAM,KAAK,UAAU,CAAC,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,gBAAgB,EAAE,MAAM,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;QAC/E,CAAC;QACD,IAAI,IAAI,CAAC,OAAO,EAAE,KAAK,CAAC;YAAE,OAAO,IAAI,CAAC;QACtC,IAAI,IAAI,CAAC,OAAO,EAAE,IAAI,CAAC,IAAI,IAAI,CAAC,OAAO,EAAE,IAAI,CAAC;YAAE,OAAO,IAAI,CAAC;QAC5D,IAAI,IAAI,GAAG,QAAQ,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAClD,IAAI,MAAM,KAAK,UAAU,IAAI,IAAI,CAAC,OAAO,EAAE,WAAW,CAAC,EAAE,CAAC;YACxD,MAAM,EAAE,GACN,OAAO,CAAC,YAAY,CAAC,MAAM,CAAC;gBAC5B,OAAO,CAAC,cAAc,CACpB,qEAAqE,EACrE,IAAI,CACL,CAAC;YACJ,MAAM,MAAM,GAAG,KAAK,CAAC,GAAG,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC;YACnC,IAAI,MAAM,IAAI,IAAI;gBAAE,IAAI,GAAG,IAAI,IAAI,KAAK,MAAM,GAAG,CAAC;QACpD,CAAC;QACD,OAAO,IAAI,CAAC;IACd,CAAC,CAAC;IACF,MAAM,SAAS,GAAG,CAAC,OAAgB,EAAU,EAAE;QAC7C,MAAM,IAAI,GAAG,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,CAAC;QACpC,IAAI,CAAC,IAAI,IAAI,MAAM,KAAK,MAAM;YAAE,OAAO,IAAI,CAAC;QAC5C,MAAM,UAAU,GAAG,KAAK,CAAC,OAAO,EAAE,KAAK,CAAC,CAAC;QACzC,MAAM,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,UAAU,EAAE,YAAY,CAAC,EAAE,KAAK,CAAC,CAAC;QAC7D,MAAM,KAAK,GAAG,SAAS,CAAC,IAAI,CAAC,OAAO,CAAC;YACnC,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,GAAG,CAAC;YACrB,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,KAAK,CAAC,UAAU,EAAE,QAAQ,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC;QACzD,IAAI,KAAK;YAAE,OAAO,GAAG,GAAG,CAAC,MAAM,CAAC,KAAK,CAAC,IAAI,IAAI,EAAE,CAAC;QACjD,MAAM,KAAK,GAAG,KAAK,CAAC,UAAU,EAAE,OAAO,CAAC,CAAC;QACzC,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,OAAO,CAAC,EAAE,KAAK,CAAC,CAAC;QACjD,IAAI,KAAK,IAAI,KAAK,KAAK,GAAG,EAAE,CAAC;YAC3B,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,MAAM,CAAC,EAAE,KAAK,CAAC,IAAI,GAAG,CAAC;YACtD,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YAC3D,MAAM,MAAM,GAAG,UAAU,CAAC,GAAG,CAAC,GAAG,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,IAAI,EAAE,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC;YACxF,OAAO,GAAG,MAAM,GAAG,MAAM,IAAI,IAAI,EAAE,CAAC;QACtC,CAAC;QACD,OAAO,IAAI,CAAC;IACd,CAAC,CAAC;IACF,MAAM,MAAM,GAAG,CAAC,OAAgB,EAAY,EAAE,CAC5C,QAAQ,CAAC,OAAO,CAAC,CAAC,OAAO,CAAC,CAAC,IAAI,EAAY,EAAE;QAC3C,IAAI,IAAI,CAAC,IAAI,EAAE,GAAG,CAAC;YAAE,OAAO,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC,CAAC;QAC9C,IAAI,IAAI,CAAC,IAAI,EAAE,KAAK,CAAC;YAAE,OAAO,EAAE,CAAC;QACjC,IAAI,IAAI,CAAC,IAAI,EAAE,KAAK,CAAC,EAAE,CAAC;YACtB,MAAM,IAAI,GAAG,QAAQ,CAAC,IAAI,CAAC;iBACxB,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;iBAClC,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CACX,QAAQ,CAAC,GAAG,CAAC;iBACV,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;iBAClC,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CACzD,CAAC;YACJ,IAAI,MAAM,KAAK,MAAM;gBAAE,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC;YAChE,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,GAAG,EAAE,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,EAAE,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC;YACtE,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CACpB,CAAC,GAAG,EAAE,EAAE,CACN,KAAK,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,UAAU,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC,UAAU,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAC5H,CAAC;YACF,IAAI,KAAK,CAAC,MAAM;gBAAE,KAAK,CAAC,MAAM,CAAC,CAAC,EAAE,CAAC,EAAE,KAAK,KAAK,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;YACpF,OAAO,KAAK,CAAC;QACf,CAAC;QACD,wEAAwE;QACxE,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC;IACtB,CAAC,CAAC,CAAC;IACL,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;AACxC,CAAC,CAAC"}

package/dist/index.d.ts

@@ -1,5 +1,6 @@
 import { type DomainDefinition } from '@zaowu-labs/core';
 export interface DocumentSummary {
+    method: 'extractive';
     schemaVersion: 1;
     status: 'ok';
     filePath: string;
@@ -57,5 +58,6 @@ export declare const convertDocument: (inputPath: string, options?: {
     outputPath?: string;
     format?: "markdown" | "text";
     yes?: boolean;
+    expectedContent?: string;
 }) => Promise<DocumentConversion>;
 //# sourceMappingURL=index.d.ts.map

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAEA,OAAO,EAML,KAAK,gBAAgB,EACtB,MAAM,kBAAkB,CAAC;AAE1B,MAAM,WAAW,eAAe;IAC9B,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,KAAK,EAAE,MAAM,CAAC;IACd,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,eAAe;IAC9B,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,MAAM,EAAE,CAAC;IACnB,KAAK,EAAE,MAAM,EAAE,CAAC;IAChB,cAAc,EAAE,MAAM,CAAC;IACvB,WAAW,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;CACrC;AAED,MAAM,WAAW,kBAAkB;IACjC,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,GAAG,SAAS,CAAC;IACzB,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC;IAC5B,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,OAAO,CAAC;CACpB;AAED,MAAM,WAAW,mBAAmB;IAClC,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,eAAe;IAC9B,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,OAAO,EAAE,mBAAmB,EAAE,CAAC;CAChC;AAED,MAAM,WAAW,mBAAmB;IAClC,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,oBAAoB;IACnC,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,OAAO,EAAE,MAAM,CAAC;IAChB,OAAO,EAAE,mBAAmB,EAAE,CAAC;CAChC;AAED,eAAO,MAAM,UAAU,EAAE,gBAmCxB,CAAC;AA4XF,eAAO,MAAM,iBAAiB,GAAU,UAAU,MAAM,KAAG,OAAO,CAAC,eAAe,CAejF,CAAC;AAEF,eAAO,MAAM,eAAe,GAAU,UAAU,MAAM,KAAG,OAAO,CAAC,eAAe,CAe/E,CAAC;AAEF,eAAO,MAAM,eAAe,GAAU,UAAU,MAAM,KAAG,OAAO,CAAC,eAAe,CAS/E,CAAC;AAEF,eAAO,MAAM,cAAc,GACzB,UAAU,MAAM,EAChB,SAAS,MAAM,KACd,OAAO,CAAC,oBAAoB,CA6B9B,CAAC;AAEF,eAAO,MAAM,eAAe,GAC1B,WAAW,MAAM,EACjB,UAAS;IAAE,UAAU,CAAC,EAAE,MAAM,CAAC;IAAC,MAAM,CAAC,EAAE,UAAU,GAAG,MAAM,CAAC;IAAC,GAAG,CAAC,EAAE,OAAO,CAAA;CAAO,KACjF,OAAO,CAAC,kBAAkB,CAqC5B,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAIA,OAAO,EAOL,KAAK,gBAAgB,EACtB,MAAM,kBAAkB,CAAC;AAE1B,MAAM,WAAW,eAAe;IAC9B,MAAM,EAAE,YAAY,CAAC;IACrB,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,KAAK,EAAE,MAAM,CAAC;IACd,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,eAAe;IAC9B,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,MAAM,EAAE,CAAC;IACnB,KAAK,EAAE,MAAM,EAAE,CAAC;IAChB,cAAc,EAAE,MAAM,CAAC;IACvB,WAAW,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;CACrC;AAED,MAAM,WAAW,kBAAkB;IACjC,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,GAAG,SAAS,CAAC;IACzB,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,EAAE,UAAU,GAAG,MAAM,CAAC;IAC5B,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,OAAO,CAAC;CACpB;AAED,MAAM,WAAW,mBAAmB;IAClC,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,eAAe;IAC9B,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,OAAO,EAAE,mBAAmB,EAAE,CAAC;CAChC;AAED,MAAM,WAAW,mBAAmB;IAClC,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,oBAAoB;IACnC,aAAa,EAAE,CAAC,CAAC;IACjB,MAAM,EAAE,IAAI,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,OAAO,EAAE,MAAM,CAAC;IAChB,OAAO,EAAE,mBAAmB,EAAE,CAAC;CAChC;AAED,eAAO,MAAM,UAAU,EAAE,gBAmCxB,CAAC;AA6NF,eAAO,MAAM,iBAAiB,GAAU,UAAU,MAAM,KAAG,OAAO,CAAC,eAAe,CAgBjF,CAAC;AAEF,eAAO,MAAM,eAAe,GAAU,UAAU,MAAM,KAAG,OAAO,CAAC,eAAe,CAkB/E,CAAC;AAEF,eAAO,MAAM,eAAe,GAAU,UAAU,MAAM,KAAG,OAAO,CAAC,eAAe,CAS/E,CAAC;AAEF,eAAO,MAAM,cAAc,GACzB,UAAU,MAAM,EAChB,SAAS,MAAM,KACd,OAAO,CAAC,oBAAoB,CA6B9B,CAAC;AAEF,eAAO,MAAM,eAAe,GAC1B,WAAW,MAAM,EACjB,UAAS;IACP,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,CAAC,EAAE,UAAU,GAAG,MAAM,CAAC;IAC7B,GAAG,CAAC,EAAE,OAAO,CAAC;IACd,eAAe,CAAC,EAAE,MAAM,CAAC;CACrB,KACL,OAAO,CAAC,kBAAkB,CA+C5B,CAAC"}

package/dist/index.js

@@ -1,6 +1,8 @@
-import { readFile } from 'node:fs/promises';
 import path from 'node:path';
-import { assertSafeOutputPath, createCapabilityLedger, stripUtf8Bom, writeTextFileExclusive, ZaoWuError, } from '@zaowu-labs/core';
+import { extractDocx } from './docx.js';
+import { extractPdf } from './pdf.js';
+import { scanMarkdown } from './markdown.js';
+import { assertSafeOutputPath, createCapabilityLedger, stripUtf8Bom, readBoundedFile, writeTextFileExclusive, ZaoWuError, } from '@zaowu-labs/core';
 export const DOC_DOMAIN = {
     name: 'doc',
     summary: 'Document workflows for summary, extraction, and conversion',
@@ -50,93 +52,6 @@ const SUPPORTED_TEXT_EXTENSIONS = new Set([
     '.html',
     '.htm',
 ]);
-const PDF_SHIM_KEYS = ['DOMMatrix', 'ImageData', 'Path2D'];
-let pdfParseModulePromise;
-const ensurePdfTextRuntime = () => {
-    const runtime = globalThis;
-    // Minimal DOMMatrix shim sufficient for pdf-parse text extraction.
-    runtime.DOMMatrix ??= class ZaoWuPdfDomMatrix {
-        a = 1;
-        b = 0;
-        c = 0;
-        d = 1;
-        e = 0;
-        f = 0;
-        constructor(init) {
-            if (Array.isArray(init) && init.length >= 6) {
-                [this.a, this.b, this.c, this.d, this.e, this.f] = init.map(Number).slice(0, 6);
-            }
-        }
-        multiplySelf() {
-            return this;
-        }
-        preMultiplySelf() {
-            return this;
-        }
-        translateSelf() {
-            return this;
-        }
-        scaleSelf() {
-            return this;
-        }
-        rotateSelf() {
-            return this;
-        }
-        invertSelf() {
-            return this;
-        }
-        transformPoint(point = {}) {
-            return {
-                x: point.x ?? 0,
-                y: point.y ?? 0,
-                z: point.z ?? 0,
-                w: point.w ?? 1,
-            };
-        }
-    };
-    // Minimal ImageData shim.
-    runtime.ImageData ??= class ZaoWuPdfImageData {
-        data;
-        width;
-        height;
-        constructor(dataOrWidth, width = 0, height = 0) {
-            this.width = typeof dataOrWidth === 'number' ? dataOrWidth : width;
-            this.height = height;
-            this.data =
-                dataOrWidth instanceof Uint8ClampedArray
-                    ? dataOrWidth
-                    : new Uint8ClampedArray(dataOrWidth * width * 4);
-        }
-    };
-    // Path2D is referenced but never actually used in our text-only path.
-    runtime.Path2D ??= class ZaoWuPdfPath2D {
-    };
-};
-const importPdfParse = async () => {
-    ensurePdfTextRuntime();
-    if (!pdfParseModulePromise) {
-        const originalWarn = console.warn;
-        // Only suppress the specific canvas-missing warning that pdf-parse emits
-        // when it can't find its optional native module. Everything else is passed
-        // through so we don't accidentally hide real problems.
-        console.warn = (...values) => {
-            const message = values.map(String).join(' ');
-            if (message.includes('Cannot load "@napi-rs/canvas" package')) {
-                return;
-            }
-            originalWarn(...values);
-        };
-        pdfParseModulePromise = import('pdf-parse')
-            .catch((error) => {
-            pdfParseModulePromise = undefined;
-            throw error;
-        })
-            .finally(() => {
-            console.warn = originalWarn;
-        });
-    }
-    return pdfParseModulePromise;
-};
 const assertSupportedDocument = (filePath) => {
     const extension = path.extname(filePath).toLowerCase();
     if (!SUPPORTED_TEXT_EXTENSIONS.has(extension)) {
@@ -148,16 +63,48 @@ const assertSupportedDocument = (filePath) => {
         });
     }
 };
+const stripHtmlBlocks = (html) => {
+    const uncommented = [];
+    let cursor = 0;
+    while (true) {
+        const start = html.indexOf('<!--', cursor);
+        if (start < 0) {
+            uncommented.push(html.slice(cursor));
+            break;
+        }
+        uncommented.push(html.slice(cursor, start));
+        const end = html.indexOf('-->', start + 4);
+        if (end < 0)
+            break;
+        cursor = end + 3;
+    }
+    const content = uncommented.join('');
+    const chunks = [];
+    const open = /<(head|script|style)\b[^<>]*>/gi;
+    cursor = 0;
+    let match;
+    while ((match = open.exec(content))) {
+        chunks.push(content.slice(cursor, match.index));
+        const close = new RegExp(`</${match[1]}\\s*>`, 'gi');
+        close.lastIndex = open.lastIndex;
+        const end = close.exec(content);
+        cursor = end ? close.lastIndex : content.length;
+        open.lastIndex = cursor;
+        if (!end)
+            break;
+    }
+    chunks.push(content.slice(cursor));
+    return chunks.join('');
+};
 const stripHtmlTags = (html) => {
-    let clean = html.replace(/<!--[\s\S]*?-->/g, '');
-    clean = clean.replace(/<(head|script|style)\b[^>]*>([\s\S]*?)<\/\1>/gi, '');
-    clean = clean.replace(/<h([1-6])\b[^>]*>/gi, (_match, level) => `${'#'.repeat(Number(level))} `);
+    let clean = stripHtmlBlocks(html);
+    clean = clean.replace(/<h([1-6])\b[^<>]*>/gi, (_match, level) => `${'#'.repeat(Number(level))} `);
     clean = clean.replace(/<\/h[1-6]>\s*/gi, '\n');
     clean = clean.replace(/<\/t[dh]>\s*/gi, ' | ');
     clean = clean.replace(/<\/tr>\s*/gi, '\n');
     clean = clean.replace(/<\/(p|div|li)>\s*/gi, '\n');
     clean = clean.replace(/<br\s*\/?>\s*/gi, '\n');
-    clean = clean.replace(/<[^>]+>/g, ' ');
+    clean = clean.replace(/<[^<>]*>/g, ' ');
     const decodeCodePoint = (match, codePoint) => {
         if (!Number.isInteger(codePoint) || codePoint <= 0 || codePoint > 0x10ffff) {
             return match;
@@ -193,75 +140,56 @@ const stripHtmlTags = (html) => {
 const readTextDocument = async (filePath, format) => {
     assertSupportedDocument(filePath);
     const extension = path.extname(filePath).toLowerCase();
+    const buffer = await readBoundedFile(filePath, 20 * 1024 * 1024, {
+        failed: {
+            code: 'DOCUMENT_READ_FAILED',
+            message: 'Could not read document.',
+            why: `ZaoWu could not read \`${filePath}\` as a regular file.`,
+            fix: 'Check the path and file permissions, then run the command again.',
+        },
+        tooLarge: {
+            code: 'DOCUMENT_INPUT_TOO_LARGE',
+            message: 'Document is too large to extract safely.',
+            why: 'Document inputs are limited to 20 MiB.',
+            fix: 'Split the document or export a smaller text-only copy.',
+        },
+    });
     try {
-        if (extension === '.pdf') {
-            const runtime = globalThis;
-            // Snapshot each shim slot before parsing: restore the previous value if
-            // one existed, and delete the key entirely if it was absent — a shim key
-            // left behind as `undefined` would still shadow `'key' in runtime`.
-            const saved = PDF_SHIM_KEYS.map((key) => ({
-                key,
-                had: key in runtime,
-                old: runtime[key],
-            }));
-            try {
-                const { PDFParse } = await importPdfParse();
-                const parser = new PDFParse({
-                    data: new Uint8Array(await readFile(filePath)),
-                });
-                try {
-                    const parsed = await parser.getText();
-                    return stripUtf8Bom(parsed.text ?? '');
-                }
-                finally {
-                    await parser.destroy();
-                }
-            }
-            finally {
-                for (const { key, had, old } of saved) {
-                    if (!had) {
-                        delete runtime[key];
-                    }
-                    else {
-                        runtime[key] = old;
-                    }
-                }
-            }
-        }
-        if (extension === '.docx') {
-            const mammoth = await import('mammoth');
-            const mammothModule = mammoth;
-            // convertToMarkdown is a version-dependent mammoth export. Distinguish a
-            // missing capability from an unreadable file so the error is actionable.
-            if (format !== 'text' && typeof mammothModule.convertToMarkdown !== 'function') {
-                throw new ZaoWuError({
-                    code: 'DOCUMENT_FORMAT_UNSUPPORTED',
-                    message: 'DOCX markdown conversion is not available.',
-                    why: 'The installed mammoth build does not provide `convertToMarkdown`, so ZaoWu cannot produce Markdown from this DOCX.',
-                    fix: 'Use `--format text` for raw text extraction, or update the `mammoth` dependency to a build with markdown support.',
-                });
-            }
-            const parsed = format === 'text'
-                ? await mammothModule.extractRawText({
-                    buffer: await readFile(filePath),
-                })
-                : await mammothModule.convertToMarkdown({
-                    buffer: await readFile(filePath),
-                });
-            return stripUtf8Bom(parsed.value);
+        let content;
+        if (extension === '.pdf')
+            content = await extractPdf(buffer);
+        else if (extension === '.docx')
+            content = await extractDocx(buffer, format ?? 'markdown');
+        else if (extension === '.html' || extension === '.htm')
+            content = stripHtmlTags(buffer.toString('utf8'));
+        else
+            content = buffer.toString('utf8');
+        if (content.length > 2_000_000) {
+            throw new ZaoWuError({
+                code: 'DOCUMENT_INPUT_TOO_LARGE',
+                message: 'Extracted document text is too large.',
+                why: 'Extracted text exceeds two million characters.',
+                fix: 'Split the document into smaller parts.',
+            });
         }
-        if (extension === '.html' || extension === '.htm') {
-            const rawHtml = await readFile(filePath, 'utf8');
-            return stripUtf8Bom(stripHtmlTags(rawHtml));
+        if (extension === '.pdf' && !content.replace(/-- \d+ of \d+ --/g, '').trim()) {
+            throw new ZaoWuError({
+                code: 'DOCUMENT_TEXT_UNAVAILABLE',
+                message: 'PDF has no extractable text.',
+                why: 'The PDF may contain scanned images rather than a text layer. ZaoWu does not perform OCR.',
+                fix: 'Run OCR in your PDF editor and export a searchable PDF or text file, then try again.',
+            });
         }
-        return stripUtf8Bom(await readFile(filePath, 'utf8'));
+        return stripUtf8Bom(content);
     }
-    catch {
+    catch (error) {
+        if (error instanceof ZaoWuError)
+            throw error;
         throw new ZaoWuError({
             code: 'DOCUMENT_READ_FAILED',
             message: 'Could not read document.',
-            why: `ZaoWu tried to read \`${filePath}\`, but the file was not readable.`,
-            fix: 'Check the path and file permissions, then run the command again.',
+            why: `ZaoWu could not extract text from \`${filePath}\`. The file may be corrupt, encrypted, or contain unsupported document structures.`,
+            fix: 'Open the file in its original editor, remove password protection if appropriate, and re-export it as PDF, DOCX, or plain text.',
         });
     }
 };
@@ -281,11 +209,7 @@ const assertSafeDocumentOutputPath = (inputPath, outputPath) => assertSafeOutput
 });
 const getTitle = (content, filePath) => parseHeadings(content).find((heading) => heading.level === 1)?.title ?? path.basename(filePath);
 const getWords = (content) => content.match(/[\p{L}\p{N}_-]+/gu) ?? [];
-const getFirstMeaningfulLines = (content, count) => content
-    .split(/\r?\n/)
-    .map((line) => line.trim())
-    .filter((line) => line && !line.startsWith('#'))
-    .slice(0, count);
+const getFirstMeaningfulLines = (content, count) => scanMarkdown(content).prose.slice(0, count);
 const getFrontmatter = (content) => {
     const lines = content.split(/\r?\n/);
     if (lines[0] !== '---') {
@@ -304,19 +228,7 @@ const getFrontmatter = (content) => {
     }
     return frontmatter;
 };
-const parseHeadings = (content) => content
-    .split(/\r?\n/)
-    .map((line, index) => {
-    const match = /^(#{1,6})\s+(.+)$/.exec(line.trim());
-    return match
-        ? {
-            level: match[1].length,
-            title: match[2].trim(),
-            line: index + 1,
-        }
-        : null;
-})
-    .filter((item) => Boolean(item));
+const parseHeadings = (content) => scanMarkdown(content).headings;
 export const summarizeDocument = async (filePath) => {
     const content = await readTextDocument(filePath);
     const lines = content.split(/\r?\n/);
@@ -326,6 +238,7 @@ export const summarizeDocument = async (filePath) => {
         schemaVersion: 1,
         status: 'ok',
         filePath,
+        method: 'extractive',
         title: getTitle(content, filePath),
         lineCount: lines.length,
         wordCount: words.length,
@@ -334,9 +247,10 @@ export const summarizeDocument = async (filePath) => {
 };
 export const extractDocument = async (filePath) => {
     const content = await readTextDocument(filePath);
-    const headings = [...content.matchAll(/^#{1,6}\s+(.+)$/gmu)].map((match) => match[1].trim());
-    const links = [...content.matchAll(/\[[^\]]+\]\(([^)]+)\)/gmu)].map((match) => match[1].trim());
-    const codeBlockCount = [...content.matchAll(/^```/gmu)].length / 2;
+    const scan = scanMarkdown(content);
+    const headings = scan.headings.map((heading) => heading.title);
+    const links = [...content.matchAll(/(?<!\\)\[(?:\\.|[^[\]\\])+\]\(([^)]+)\)/gmu)].map((match) => match[1].trim());
+    const codeBlockCount = scan.codeBlockCount;
     return {
         schemaVersion: 1,
         status: 'ok',
@@ -387,10 +301,20 @@ export const convertDocument = async (inputPath, options = {}) => {
     const format = options.format ??
         (path.extname(options.outputPath ?? '').toLowerCase() === '.txt' ? 'text' : 'markdown');
     const content = await readTextDocument(inputPath, format);
-    const converted = format === 'text'
-        ? content.replace(/^#{1,6}\s+/gmu, '').replace(/\[([^\]]+)\]\([^)]+\)/gmu, '$1')
+    const converted = format === 'text' &&
+        ['.md', '.markdown', '.html', '.htm'].includes(path.extname(inputPath).toLowerCase())
+        ? content
+            .replace(/^#{1,6}\s+/gmu, '')
+            .replace(/(?<!\\)\[([^[\]\r\n]+)\]\([^)\r\n]+\)/gmu, '$1')
         : content;
     if (options.outputPath && options.yes) {
+        if (options.expectedContent !== undefined && converted !== options.expectedContent)
+            throw new ZaoWuError({
+                code: 'OPERATION_PLAN_MISMATCH',
+                message: 'Document content changed after plan verification.',
+                why: 'The converted output no longer matches the content authorized by the operation plan. No file was written.',
+                fix: 'Run the preview again and confirm its new fingerprint.',
+            });
         await assertSafeDocumentOutputPath(inputPath, options.outputPath);
         await writeTextFileExclusive(options.outputPath, converted, {
             exists: {

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EACL,oBAAoB,EACpB,sBAAsB,EACtB,YAAY,EACZ,sBAAsB,EACtB,UAAU,GAEX,MAAM,kBAAkB,CAAC;AA0D1B,MAAM,CAAC,MAAM,UAAU,GAAqB;IAC1C,IAAI,EAAE,KAAK;IACX,OAAO,EAAE,4DAA4D;IACrE,YAAY,EAAE,sBAAsB,CAAC;QACnC,UAAU,EAAE,IAAI;QAChB,WAAW,EAAE,IAAI;KAClB,CAAC;IACF,QAAQ,EAAE;QACR;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,sBAAsB;YAC/B,MAAM,EAAE,WAAW;SACpB;QACD;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,4CAA4C;YACrD,MAAM,EAAE,WAAW;SACpB;QACD;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,iDAAiD;YAC1D,MAAM,EAAE,WAAW;YACnB,SAAS,EAAE,IAAI;SAChB;QACD;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,0CAA0C;YACnD,MAAM,EAAE,WAAW;SACpB;QACD;YACE,IAAI,EAAE,QAAQ;YACd,OAAO,EAAE,2CAA2C;YACpD,MAAM,EAAE,WAAW;SACpB;KACF;CACF,CAAC;AAEF,MAAM,yBAAyB,GAAG,IAAI,GAAG,CAAC;IACxC,MAAM;IACN,KAAK;IACL,WAAW;IACX,MAAM;IACN,OAAO;IACP,MAAM;IACN,OAAO;IACP,MAAM;IACN,OAAO;IACP,OAAO;IACP,MAAM;CACP,CAAC,CAAC;AAsCH,MAAM,aAAa,GAAG,CAAC,WAAW,EAAE,WAAW,EAAE,QAAQ,CAAU,CAAC;AAIpE,IAAI,qBAA0D,CAAC;AAE/D,MAAM,oBAAoB,GAAG,GAAS,EAAE;IACtC,MAAM,OAAO,GAAG,UAA8B,CAAC;IAE/C,mEAAmE;IACnE,OAAO,CAAC,SAAS,KAAK,MAAM,iBAAiB;QAC3C,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QAEN,YAAY,IAAc;YACxB,IAAI,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,MAAM,IAAI,CAAC,EAAE,CAAC;gBAC5C,CAAC,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;YAClF,CAAC;QACH,CAAC;QAED,YAAY;YACV,OAAO,IAAI,CAAC;QACd,CAAC;QACD,eAAe;YACb,OAAO,IAAI,CAAC;QACd,CAAC;QACD,aAAa;YACX,OAAO,IAAI,CAAC;QACd,CAAC;QACD,SAAS;YACP,OAAO,IAAI,CAAC;QACd,CAAC;QACD,UAAU;YACR,OAAO,IAAI,CAAC;QACd,CAAC;QACD,UAAU;YACR,OAAO,IAAI,CAAC;QACd,CAAC;QAED,cAAc,CAAC,QAA4D,EAAE;YAC3E,OAAO;gBACL,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;gBACf,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;gBACf,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;gBACf,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;aAChB,CAAC;QACJ,CAAC;KACF,CAAC;IAEF,0BAA0B;IAC1B,OAAO,CAAC,SAAS,KAAK,MAAM,iBAAiB;QAC3C,IAAI,CAAoB;QACxB,KAAK,CAAS;QACd,MAAM,CAAS;QAEf,YAAY,WAAuC,EAAE,KAAK,GAAG,CAAC,EAAE,MAAM,GAAG,CAAC;YACxE,IAAI,CAAC,KAAK,GAAG,OAAO,WAAW,KAAK,QAAQ,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,KAAK,CAAC;YACnE,IAAI,CAAC,MAAM,GAAG,MAAM,CAAC;YACrB,IAAI,CAAC,IAAI;gBACP,WAAW,YAAY,iBAAiB;oBACtC,CAAC,CAAC,WAAW;oBACb,CAAC,CAAC,IAAI,iBAAiB,CAAC,WAAW,GAAG,KAAK,GAAG,CAAC,CAAC,CAAC;QACvD,CAAC;KACF,CAAC;IAEF,sEAAsE;IACtE,OAAO,CAAC,MAAM,KAAK,MAAM,cAAc;KAAG,CAAC;AAC7C,CAAC,CAAC;AAEF,MAAM,cAAc,GAAG,KAAK,IAA6B,EAAE;IACzD,oBAAoB,EAAE,CAAC;IAEvB,IAAI,CAAC,qBAAqB,EAAE,CAAC;QAC3B,MAAM,YAAY,GAAG,OAAO,CAAC,IAAI,CAAC;QAElC,yEAAyE;QACzE,2EAA2E;QAC3E,uDAAuD;QACvD,OAAO,CAAC,IAAI,GAAG,CAAC,GAAG,MAAiB,EAAQ,EAAE;YAC5C,MAAM,OAAO,GAAG,MAAM,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC7C,IAAI,OAAO,CAAC,QAAQ,CAAC,uCAAuC,CAAC,EAAE,CAAC;gBAC9D,OAAO;YACT,CAAC;YACD,YAAY,CAAC,GAAG,MAAM,CAAC,CAAC;QAC1B,CAAC,CAAC;QAEF,qBAAqB,GAAG,MAAM,CAAC,WAAW,CAAC;aACxC,KAAK,CAAC,CAAC,KAAc,EAAE,EAAE;YACxB,qBAAqB,GAAG,SAAS,CAAC;YAClC,MAAM,KAAK,CAAC;QACd,CAAC,CAAC;aACD,OAAO,CAAC,GAAG,EAAE;YACZ,OAAO,CAAC,IAAI,GAAG,YAAY,CAAC;QAC9B,CAAC,CAAC,CAAC;IACP,CAAC;IAED,OAAO,qBAAqB,CAAC;AAC/B,CAAC,CAAC;AAEF,MAAM,uBAAuB,GAAG,CAAC,QAAgB,EAAQ,EAAE;IACzD,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IAEvD,IAAI,CAAC,yBAAyB,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC;QAC9C,MAAM,IAAI,UAAU,CAAC;YACnB,IAAI,EAAE,6BAA6B;YACnC,OAAO,EAAE,uCAAuC;YAChD,GAAG,EAAE,oFAAoF,SAAS,IAAI,SAAS,eAAe;YAC9H,GAAG,EAAE,yHAAyH;SAC/H,CAAC,CAAC;IACL,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,aAAa,GAAG,CAAC,IAAY,EAAU,EAAE;IAC7C,IAAI,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,kBAAkB,EAAE,EAAE,CAAC,CAAC;IACjD,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,gDAAgD,EAAE,EAAE,CAAC,CAAC;IAC5E,KAAK,GAAG,KAAK,CAAC,OAAO,CACnB,qBAAqB,EACrB,CAAC,MAAM,EAAE,KAAa,EAAE,EAAE,CAAC,GAAG,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,GAAG,CAC3D,CAAC;IACF,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,iBAAiB,EAAE,IAAI,CAAC,CAAC;IAC/C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,gBAAgB,EAAE,KAAK,CAAC,CAAC;IAC/C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,aAAa,EAAE,IAAI,CAAC,CAAC;IAC3C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,qBAAqB,EAAE,IAAI,CAAC,CAAC;IACnD,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,iBAAiB,EAAE,IAAI,CAAC,CAAC;IAC/C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC,CAAC;IACvC,MAAM,eAAe,GAAG,CAAC,KAAa,EAAE,SAAiB,EAAU,EAAE;QACnE,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,SAAS,CAAC,IAAI,SAAS,IAAI,CAAC,IAAI,SAAS,GAAG,QAAQ,EAAE,CAAC;YAC3E,OAAO,KAAK,CAAC;QACf,CAAC;QAED,IAAI,CAAC;YACH,OAAO,MAAM,CAAC,aAAa,CAAC,SAAS,CAAC,CAAC;QACzC,CAAC;QAAC,MAAM,CAAC;YACP,OAAO,KAAK,CAAC;QACf,CAAC;IACH,CAAC,CAAC;IAEF,uEAAuE;IACvE,0EAA0E;IAC1E,KAAK,GAAG,KAAK;SACV,UAAU,CAAC,QAAQ,EAAE,GAAG,CAAC;SACzB,OAAO,CAAC,mBAAmB,EAAE,CAAC,KAAK,EAAE,GAAW,EAAE,EAAE,CACnD,eAAe,CAAC,KAAK,EAAE,MAAM,CAAC,QAAQ,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC,CACjD;SACA,OAAO,CAAC,WAAW,EAAE,CAAC,KAAK,EAAE,OAAe,EAAE,EAAE,CAC/C,eAAe,CAAC,KAAK,EAAE,MAAM,CAAC,QAAQ,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,CACrD;SACA,UAAU,CAAC,MAAM,EAAE,GAAG,CAAC;SACvB,UAAU,CAAC,MAAM,EAAE,GAAG,CAAC;SACvB,UAAU,CAAC,QAAQ,EAAE,GAAG,CAAC;SACzB,UAAU,CAAC,QAAQ,EAAE,GAAG,CAAC;SACzB,UAAU,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC;IAC5B,OAAO,KAAK;SACT,KAAK,CAAC,OAAO,CAAC;SACd,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CACZ,IAAI;SACD,OAAO,CAAC,YAAY,EAAE,EAAE,CAAC;SACzB,OAAO,CAAC,WAAW,EAAE,KAAK,CAAC;SAC3B,IAAI,EAAE,CACV;SACA,MAAM,CAAC,CAAC,IAAI,EAAE,GAAG,EAAE,GAAG,EAAE,EAAE,CAAC,IAAI,IAAI,CAAC,GAAG,GAAG,CAAC,IAAI,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC;SAC7D,IAAI,CAAC,IAAI,CAAC;SACV,IAAI,EAAE,CAAC;AACZ,CAAC,CAAC;AAEF,MAAM,gBAAgB,GAAG,KAAK,EAC5B,QAAgB,EAChB,MAA4B,EACX,EAAE;IACnB,uBAAuB,CAAC,QAAQ,CAAC,CAAC;IAClC,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IAEvD,IAAI,CAAC;QACH,IAAI,SAAS,KAAK,MAAM,EAAE,CAAC;YACzB,MAAM,OAAO,GAAG,UAA8B,CAAC;YAC/C,wEAAwE;YACxE,yEAAyE;YACzE,oEAAoE;YACpE,MAAM,KAAK,GAAG,aAAa,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC;gBACxC,GAAG;gBACH,GAAG,EAAE,GAAG,IAAI,OAAO;gBACnB,GAAG,EAAE,OAAO,CAAC,GAAG,CAAC;aAClB,CAAC,CAAC,CAAC;YAEJ,IAAI,CAAC;gBACH,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,cAAc,EAAE,CAAC;gBAC5C,MAAM,MAAM,GAAG,IAAI,QAAQ,CAAC;oBAC1B,IAAI,EAAE,IAAI,UAAU,CAAC,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;iBAC/C,CAAC,CAAC;gBAEH,IAAI,CAAC;oBACH,MAAM,MAAM,GAAG,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;oBAEtC,OAAO,YAAY,CAAC,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC;gBACzC,CAAC;wBAAS,CAAC;oBACT,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;gBACzB,CAAC;YACH,CAAC;oBAAS,CAAC;gBACT,KAAK,MAAM,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI,KAAK,EAAE,CAAC;oBACtC,IAAI,CAAC,GAAG,EAAE,CAAC;wBACT,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC;oBACtB,CAAC;yBAAM,CAAC;wBACL,OAAO,CAAC,GAAG,CAAa,GAAG,GAAG,CAAC;oBAClC,CAAC;gBACH,CAAC;YACH,CAAC;QACH,CAAC;QAED,IAAI,SAAS,KAAK,OAAO,EAAE,CAAC;YAC1B,MAAM,OAAO,GAAG,MAAM,MAAM,CAAC,SAAS,CAAC,CAAC;YACxC,MAAM,aAAa,GAAG,OAGrB,CAAC;YAEF,yEAAyE;YACzE,yEAAyE;YACzE,IAAI,MAAM,KAAK,MAAM,IAAI,OAAO,aAAa,CAAC,iBAAiB,KAAK,UAAU,EAAE,CAAC;gBAC/E,MAAM,IAAI,UAAU,CAAC;oBACnB,IAAI,EAAE,6BAA6B;oBACnC,OAAO,EAAE,4CAA4C;oBACrD,GAAG,EAAE,oHAAoH;oBACzH,GAAG,EAAE,mHAAmH;iBACzH,CAAC,CAAC;YACL,CAAC;YAED,MAAM,MAAM,GACV,MAAM,KAAK,MAAM;gBACf,CAAC,CAAC,MAAM,aAAa,CAAC,cAAc,CAAC;oBACjC,MAAM,EAAE,MAAM,QAAQ,CAAC,QAAQ,CAAC;iBACjC,CAAC;gBACJ,CAAC,CAAC,MAAM,aAAa,CAAC,iBAAkB,CAAC;oBACrC,MAAM,EAAE,MAAM,QAAQ,CAAC,QAAQ,CAAC;iBACjC,CAAC,CAAC;YAET,OAAO,YAAY,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;QACpC,CAAC;QAED,IAAI,SAAS,KAAK,OAAO,IAAI,SAAS,KAAK,MAAM,EAAE,CAAC;YAClD,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;YACjD,OAAO,YAAY,CAAC,aAAa,CAAC,OAAO,CAAC,CAAC,CAAC;QAC9C,CAAC;QAED,OAAO,YAAY,CAAC,MAAM,QAAQ,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC,CAAC;IACxD,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,UAAU,CAAC;YACnB,IAAI,EAAE,sBAAsB;YAC5B,OAAO,EAAE,0BAA0B;YACnC,GAAG,EAAE,yBAAyB,QAAQ,oCAAoC;YAC1E,GAAG,EAAE,kEAAkE;SACxE,CAAC,CAAC;IACL,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,4BAA4B,GAAG,CAAC,SAAiB,EAAE,UAAkB,EAAiB,EAAE,CAC5F,oBAAoB,CAAC,SAAS,EAAE,UAAU,EAAE;IAC1C,aAAa,EAAE;QACb,IAAI,EAAE,0BAA0B;QAChC,OAAO,EAAE,iDAAiD;QAC1D,GAAG,EAAE,2DAA2D;QAChE,GAAG,EAAE,qCAAqC;KAC3C;IACD,aAAa,EAAE;QACb,IAAI,EAAE,0BAA0B;QAChC,OAAO,EAAE,sCAAsC;QAC/C,GAAG,EAAE,uCAAuC,UAAU,KAAK;QAC3D,GAAG,EAAE,2EAA2E;KACjF;CACF,CAAC,CAAC;AAEL,MAAM,QAAQ,GAAG,CAAC,OAAe,EAAE,QAAgB,EAAU,EAAE,CAC7D,aAAa,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,CAAC,OAAO,EAAE,EAAE,CAAC,OAAO,CAAC,KAAK,KAAK,CAAC,CAAC,EAAE,KAAK,IAAI,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;AAElG,MAAM,QAAQ,GAAG,CAAC,OAAe,EAAY,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,mBAAmB,CAAC,IAAI,EAAE,CAAC;AAEzF,MAAM,uBAAuB,GAAG,CAAC,OAAe,EAAE,KAAa,EAAY,EAAE,CAC3E,OAAO;KACJ,KAAK,CAAC,OAAO,CAAC;KACd,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC;KAC1B,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC;KAC/C,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC;AAErB,MAAM,cAAc,GAAG,CAAC,OAAe,EAA0B,EAAE;IACjE,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IAErC,IAAI,KAAK,CAAC,CAAC,CAAC,KAAK,KAAK,EAAE,CAAC;QACvB,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,QAAQ,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,KAAK,KAAK,CAAC,CAAC;IAEpE,IAAI,QAAQ,GAAG,CAAC,EAAE,CAAC;QACjB,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,WAAW,GAA2B,EAAE,CAAC;IAE/C,KAAK,MAAM,IAAI,IAAI,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,QAAQ,GAAG,CAAC,CAAC,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,4BAA4B,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAEtD,IAAI,KAAK,EAAE,CAAC;YACV,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAC1C,CAAC;IACH,CAAC;IAED,OAAO,WAAW,CAAC;AACrB,CAAC,CAAC;AAEF,MAAM,aAAa,GAAG,CAAC,OAAe,EAAyB,EAAE,CAC/D,OAAO;KACJ,KAAK,CAAC,OAAO,CAAC;KACd,GAAG,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE;IACnB,MAAM,KAAK,GAAG,mBAAmB,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC;IAEpD,OAAO,KAAK;QACV,CAAC,CAAC;YACE,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,MAAM;YACtB,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE;YACtB,IAAI,EAAE,KAAK,GAAG,CAAC;SAChB;QACH,CAAC,CAAC,IAAI,CAAC;AACX,CAAC,CAAC;KACD,MAAM,CAAC,CAAC,IAAI,EAA+B,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC;AAElE,MAAM,CAAC,MAAM,iBAAiB,GAAG,KAAK,EAAE,QAAgB,EAA4B,EAAE;IACpF,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IACjD,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IACrC,MAAM,KAAK,GAAG,QAAQ,CAAC,OAAO,CAAC,CAAC;IAChC,MAAM,YAAY,GAAG,uBAAuB,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;IAEzD,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,KAAK,EAAE,QAAQ,CAAC,OAAO,EAAE,QAAQ,CAAC;QAClC,SAAS,EAAE,KAAK,CAAC,MAAM;QACvB,SAAS,EAAE,KAAK,CAAC,MAAM;QACvB,OAAO,EAAE,YAAY,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,YAAY,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,wBAAwB;KACrF,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,EAAE,QAAgB,EAA4B,EAAE;IAClF,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IACjD,MAAM,QAAQ,GAAG,CAAC,GAAG,OAAO,CAAC,QAAQ,CAAC,oBAAoB,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;IAC7F,MAAM,KAAK,GAAG,CAAC,GAAG,OAAO,CAAC,QAAQ,CAAC,0BAA0B,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;IAChG,MAAM,cAAc,GAAG,CAAC,GAAG,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;IAEnE,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,QAAQ;QACR,KAAK;QACL,cAAc,EAAE,IAAI,CAAC,KAAK,CAAC,cAAc,CAAC;QAC1C,WAAW,EAAE,cAAc,CAAC,OAAO,CAAC;KACrC,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,EAAE,QAAgB,EAA4B,EAAE;IAClF,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IAEjD,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,OAAO,EAAE,aAAa,CAAC,OAAO,CAAC;KAChC,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,cAAc,GAAG,KAAK,EACjC,QAAgB,EAChB,OAAe,EACgB,EAAE;IACjC,MAAM,cAAc,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC;IAEtC,IAAI,CAAC,cAAc,EAAE,CAAC;QACpB,MAAM,IAAI,UAAU,CAAC;YACnB,IAAI,EAAE,kCAAkC;YACxC,OAAO,EAAE,6BAA6B;YACtC,GAAG,EAAE,4CAA4C;YACjD,GAAG,EAAE,wCAAwC;SAC9C,CAAC,CAAC;IACL,CAAC;IAED,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IACjD,MAAM,YAAY,GAAG,cAAc,CAAC,WAAW,EAAE,CAAC;IAClD,MAAM,OAAO,GAAG,OAAO;SACpB,KAAK,CAAC,OAAO,CAAC;SACd,GAAG,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,CAAC;QACrB,IAAI,EAAE,KAAK,GAAG,CAAC;QACf,IAAI,EAAE,IAAI,CAAC,IAAI,EAAE;KAClB,CAAC,CAAC;SACF,MAAM,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,IAAI,CAAC,WAAW,EAAE,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC,CAAC;IAEtE,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,OAAO,EAAE,cAAc;QACvB,OAAO;KACR,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,EAClC,SAAiB,EACjB,UAAgF,EAAE,EACrD,EAAE;IAC/B,MAAM,MAAM,GACV,OAAO,CAAC,MAAM;QACd,CAAC,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,UAAU,IAAI,EAAE,CAAC,CAAC,WAAW,EAAE,KAAK,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC;IAC1F,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,SAAS,EAAE,MAAM,CAAC,CAAC;IAC1D,MAAM,SAAS,GACb,MAAM,KAAK,MAAM;QACf,CAAC,CAAC,OAAO,CAAC,OAAO,CAAC,eAAe,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,0BAA0B,EAAE,IAAI,CAAC;QAChF,CAAC,CAAC,OAAO,CAAC;IAEd,IAAI,OAAO,CAAC,UAAU,IAAI,OAAO,CAAC,GAAG,EAAE,CAAC;QACtC,MAAM,4BAA4B,CAAC,SAAS,EAAE,OAAO,CAAC,UAAU,CAAC,CAAC;QAClE,MAAM,sBAAsB,CAAC,OAAO,CAAC,UAAU,EAAE,SAAS,EAAE;YAC1D,MAAM,EAAE;gBACN,IAAI,EAAE,0BAA0B;gBAChC,OAAO,EAAE,sCAAsC;gBAC/C,GAAG,EAAE,uCAAuC,OAAO,CAAC,UAAU,KAAK;gBACnE,GAAG,EAAE,2EAA2E;aACjF;YACD,MAAM,EAAE;gBACN,IAAI,EAAE,uBAAuB;gBAC7B,OAAO,EAAE,qCAAqC;gBAC9C,GAAG,EAAE,0BAA0B,OAAO,CAAC,UAAU,6CAA6C;gBAC9F,GAAG,EAAE,yEAAyE;aAC/E;SACF,CAAC,CAAC;IACL,CAAC;IAED,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,OAAO,CAAC,UAAU,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI;QAC7D,SAAS;QACT,UAAU,EAAE,OAAO,CAAC,UAAU;QAC9B,MAAM;QACN,OAAO,EAAE,SAAS;QAClB,SAAS,EAAE,OAAO,CAAC,OAAO,CAAC,UAAU,IAAI,OAAO,CAAC,GAAG,CAAC;KACtD,CAAC;AACJ,CAAC,CAAC"}
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,WAAW,EAAE,MAAM,WAAW,CAAC;AACxC,OAAO,EAAE,UAAU,EAAE,MAAM,UAAU,CAAC;AACtC,OAAO,EAAE,YAAY,EAAE,MAAM,eAAe,CAAC;AAC7C,OAAO,EACL,oBAAoB,EACpB,sBAAsB,EACtB,YAAY,EACZ,eAAe,EACf,sBAAsB,EACtB,UAAU,GAEX,MAAM,kBAAkB,CAAC;AA2D1B,MAAM,CAAC,MAAM,UAAU,GAAqB;IAC1C,IAAI,EAAE,KAAK;IACX,OAAO,EAAE,4DAA4D;IACrE,YAAY,EAAE,sBAAsB,CAAC;QACnC,UAAU,EAAE,IAAI;QAChB,WAAW,EAAE,IAAI;KAClB,CAAC;IACF,QAAQ,EAAE;QACR;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,sBAAsB;YAC/B,MAAM,EAAE,WAAW;SACpB;QACD;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,4CAA4C;YACrD,MAAM,EAAE,WAAW;SACpB;QACD;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,iDAAiD;YAC1D,MAAM,EAAE,WAAW;YACnB,SAAS,EAAE,IAAI;SAChB;QACD;YACE,IAAI,EAAE,SAAS;YACf,OAAO,EAAE,0CAA0C;YACnD,MAAM,EAAE,WAAW;SACpB;QACD;YACE,IAAI,EAAE,QAAQ;YACd,OAAO,EAAE,2CAA2C;YACpD,MAAM,EAAE,WAAW;SACpB;KACF;CACF,CAAC;AAEF,MAAM,yBAAyB,GAAG,IAAI,GAAG,CAAC;IACxC,MAAM;IACN,KAAK;IACL,WAAW;IACX,MAAM;IACN,OAAO;IACP,MAAM;IACN,OAAO;IACP,MAAM;IACN,OAAO;IACP,OAAO;IACP,MAAM;CACP,CAAC,CAAC;AAEH,MAAM,uBAAuB,GAAG,CAAC,QAAgB,EAAQ,EAAE;IACzD,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IAEvD,IAAI,CAAC,yBAAyB,CAAC,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC;QAC9C,MAAM,IAAI,UAAU,CAAC;YACnB,IAAI,EAAE,6BAA6B;YACnC,OAAO,EAAE,uCAAuC;YAChD,GAAG,EAAE,oFAAoF,SAAS,IAAI,SAAS,eAAe;YAC9H,GAAG,EAAE,yHAAyH;SAC/H,CAAC,CAAC;IACL,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,eAAe,GAAG,CAAC,IAAY,EAAU,EAAE;IAC/C,MAAM,WAAW,GAAa,EAAE,CAAC;IACjC,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,OAAO,IAAI,EAAE,CAAC;QACZ,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;QAC3C,IAAI,KAAK,GAAG,CAAC,EAAE,CAAC;YACd,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;YACrC,MAAM;QACR,CAAC;QACD,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,KAAK,CAAC,CAAC,CAAC;QAC5C,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,KAAK,EAAE,KAAK,GAAG,CAAC,CAAC,CAAC;QAC3C,IAAI,GAAG,GAAG,CAAC;YAAE,MAAM;QACnB,MAAM,GAAG,GAAG,GAAG,CAAC,CAAC;IACnB,CAAC;IACD,MAAM,OAAO,GAAG,WAAW,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACrC,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,MAAM,IAAI,GAAG,iCAAiC,CAAC;IAC/C,MAAM,GAAG,CAAC,CAAC;IACX,IAAI,KAA6B,CAAC;IAClC,OAAO,CAAC,KAAK,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,EAAE,CAAC;QACpC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,MAAM,EAAE,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC;QAChD,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,KAAK,KAAK,CAAC,CAAC,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC;QACrD,KAAK,CAAC,SAAS,GAAG,IAAI,CAAC,SAAS,CAAC;QACjC,MAAM,GAAG,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QAChC,MAAM,GAAG,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC;QAChD,IAAI,CAAC,SAAS,GAAG,MAAM,CAAC;QACxB,IAAI,CAAC,GAAG;YAAE,MAAM;IAClB,CAAC;IACD,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;IACnC,OAAO,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;AACzB,CAAC,CAAC;AAEF,MAAM,aAAa,GAAG,CAAC,IAAY,EAAU,EAAE;IAC7C,IAAI,KAAK,GAAG,eAAe,CAAC,IAAI,CAAC,CAAC;IAClC,KAAK,GAAG,KAAK,CAAC,OAAO,CACnB,sBAAsB,EACtB,CAAC,MAAM,EAAE,KAAa,EAAE,EAAE,CAAC,GAAG,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,GAAG,CAC3D,CAAC;IACF,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,iBAAiB,EAAE,IAAI,CAAC,CAAC;IAC/C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,gBAAgB,EAAE,KAAK,CAAC,CAAC;IAC/C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,aAAa,EAAE,IAAI,CAAC,CAAC;IAC3C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,qBAAqB,EAAE,IAAI,CAAC,CAAC;IACnD,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,iBAAiB,EAAE,IAAI,CAAC,CAAC;IAC/C,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,WAAW,EAAE,GAAG,CAAC,CAAC;IACxC,MAAM,eAAe,GAAG,CAAC,KAAa,EAAE,SAAiB,EAAU,EAAE;QACnE,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,SAAS,CAAC,IAAI,SAAS,IAAI,CAAC,IAAI,SAAS,GAAG,QAAQ,EAAE,CAAC;YAC3E,OAAO,KAAK,CAAC;QACf,CAAC;QAED,IAAI,CAAC;YACH,OAAO,MAAM,CAAC,aAAa,CAAC,SAAS,CAAC,CAAC;QACzC,CAAC;QAAC,MAAM,CAAC;YACP,OAAO,KAAK,CAAC;QACf,CAAC;IACH,CAAC,CAAC;IAEF,uEAAuE;IACvE,0EAA0E;IAC1E,KAAK,GAAG,KAAK;SACV,UAAU,CAAC,QAAQ,EAAE,GAAG,CAAC;SACzB,OAAO,CAAC,mBAAmB,EAAE,CAAC,KAAK,EAAE,GAAW,EAAE,EAAE,CACnD,eAAe,CAAC,KAAK,EAAE,MAAM,CAAC,QAAQ,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC,CACjD;SACA,OAAO,CAAC,WAAW,EAAE,CAAC,KAAK,EAAE,OAAe,EAAE,EAAE,CAC/C,eAAe,CAAC,KAAK,EAAE,MAAM,CAAC,QAAQ,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,CACrD;SACA,UAAU,CAAC,MAAM,EAAE,GAAG,CAAC;SACvB,UAAU,CAAC,MAAM,EAAE,GAAG,CAAC;SACvB,UAAU,CAAC,QAAQ,EAAE,GAAG,CAAC;SACzB,UAAU,CAAC,QAAQ,EAAE,GAAG,CAAC;SACzB,UAAU,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC;IAC5B,OAAO,KAAK;SACT,KAAK,CAAC,OAAO,CAAC;SACd,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CACZ,IAAI;SACD,OAAO,CAAC,YAAY,EAAE,EAAE,CAAC;SACzB,OAAO,CAAC,WAAW,EAAE,KAAK,CAAC;SAC3B,IAAI,EAAE,CACV;SACA,MAAM,CAAC,CAAC,IAAI,EAAE,GAAG,EAAE,GAAG,EAAE,EAAE,CAAC,IAAI,IAAI,CAAC,GAAG,GAAG,CAAC,IAAI,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC;SAC7D,IAAI,CAAC,IAAI,CAAC;SACV,IAAI,EAAE,CAAC;AACZ,CAAC,CAAC;AAEF,MAAM,gBAAgB,GAAG,KAAK,EAC5B,QAAgB,EAChB,MAA4B,EACX,EAAE;IACnB,uBAAuB,CAAC,QAAQ,CAAC,CAAC;IAClC,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IAEvD,MAAM,MAAM,GAAG,MAAM,eAAe,CAAC,QAAQ,EAAE,EAAE,GAAG,IAAI,GAAG,IAAI,EAAE;QAC/D,MAAM,EAAE;YACN,IAAI,EAAE,sBAAsB;YAC5B,OAAO,EAAE,0BAA0B;YACnC,GAAG,EAAE,0BAA0B,QAAQ,uBAAuB;YAC9D,GAAG,EAAE,kEAAkE;SACxE;QACD,QAAQ,EAAE;YACR,IAAI,EAAE,0BAA0B;YAChC,OAAO,EAAE,0CAA0C;YACnD,GAAG,EAAE,wCAAwC;YAC7C,GAAG,EAAE,wDAAwD;SAC9D;KACF,CAAC,CAAC;IACH,IAAI,CAAC;QACH,IAAI,OAAe,CAAC;QACpB,IAAI,SAAS,KAAK,MAAM;YAAE,OAAO,GAAG,MAAM,UAAU,CAAC,MAAM,CAAC,CAAC;aACxD,IAAI,SAAS,KAAK,OAAO;YAAE,OAAO,GAAG,MAAM,WAAW,CAAC,MAAM,EAAE,MAAM,IAAI,UAAU,CAAC,CAAC;aACrF,IAAI,SAAS,KAAK,OAAO,IAAI,SAAS,KAAK,MAAM;YACpD,OAAO,GAAG,aAAa,CAAC,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC;;YAC9C,OAAO,GAAG,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;QACvC,IAAI,OAAO,CAAC,MAAM,GAAG,SAAS,EAAE,CAAC;YAC/B,MAAM,IAAI,UAAU,CAAC;gBACnB,IAAI,EAAE,0BAA0B;gBAChC,OAAO,EAAE,uCAAuC;gBAChD,GAAG,EAAE,gDAAgD;gBACrD,GAAG,EAAE,wCAAwC;aAC9C,CAAC,CAAC;QACL,CAAC;QACD,IAAI,SAAS,KAAK,MAAM,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,mBAAmB,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC;YAC7E,MAAM,IAAI,UAAU,CAAC;gBACnB,IAAI,EAAE,2BAA2B;gBACjC,OAAO,EAAE,8BAA8B;gBACvC,GAAG,EAAE,0FAA0F;gBAC/F,GAAG,EAAE,sFAAsF;aAC5F,CAAC,CAAC;QACL,CAAC;QACD,OAAO,YAAY,CAAC,OAAO,CAAC,CAAC;IAC/B,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,IAAI,KAAK,YAAY,UAAU;YAAE,MAAM,KAAK,CAAC;QAC7C,MAAM,IAAI,UAAU,CAAC;YACnB,IAAI,EAAE,sBAAsB;YAC5B,OAAO,EAAE,0BAA0B;YACnC,GAAG,EAAE,uCAAuC,QAAQ,qFAAqF;YACzI,GAAG,EAAE,gIAAgI;SACtI,CAAC,CAAC;IACL,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,4BAA4B,GAAG,CAAC,SAAiB,EAAE,UAAkB,EAAiB,EAAE,CAC5F,oBAAoB,CAAC,SAAS,EAAE,UAAU,EAAE;IAC1C,aAAa,EAAE;QACb,IAAI,EAAE,0BAA0B;QAChC,OAAO,EAAE,iDAAiD;QAC1D,GAAG,EAAE,2DAA2D;QAChE,GAAG,EAAE,qCAAqC;KAC3C;IACD,aAAa,EAAE;QACb,IAAI,EAAE,0BAA0B;QAChC,OAAO,EAAE,sCAAsC;QAC/C,GAAG,EAAE,uCAAuC,UAAU,KAAK;QAC3D,GAAG,EAAE,2EAA2E;KACjF;CACF,CAAC,CAAC;AAEL,MAAM,QAAQ,GAAG,CAAC,OAAe,EAAE,QAAgB,EAAU,EAAE,CAC7D,aAAa,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,CAAC,OAAO,EAAE,EAAE,CAAC,OAAO,CAAC,KAAK,KAAK,CAAC,CAAC,EAAE,KAAK,IAAI,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC;AAElG,MAAM,QAAQ,GAAG,CAAC,OAAe,EAAY,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,mBAAmB,CAAC,IAAI,EAAE,CAAC;AAEzF,MAAM,uBAAuB,GAAG,CAAC,OAAe,EAAE,KAAa,EAAY,EAAE,CAC3E,YAAY,CAAC,OAAO,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC;AAE9C,MAAM,cAAc,GAAG,CAAC,OAAe,EAA0B,EAAE;IACjE,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IAErC,IAAI,KAAK,CAAC,CAAC,CAAC,KAAK,KAAK,EAAE,CAAC;QACvB,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,QAAQ,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,KAAK,KAAK,CAAC,CAAC;IAEpE,IAAI,QAAQ,GAAG,CAAC,EAAE,CAAC;QACjB,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,WAAW,GAA2B,EAAE,CAAC;IAE/C,KAAK,MAAM,IAAI,IAAI,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,QAAQ,GAAG,CAAC,CAAC,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,4BAA4B,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAEtD,IAAI,KAAK,EAAE,CAAC;YACV,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAC1C,CAAC;IACH,CAAC;IAED,OAAO,WAAW,CAAC;AACrB,CAAC,CAAC;AAEF,MAAM,aAAa,GAAG,CAAC,OAAe,EAAyB,EAAE,CAAC,YAAY,CAAC,OAAO,CAAC,CAAC,QAAQ,CAAC;AAEjG,MAAM,CAAC,MAAM,iBAAiB,GAAG,KAAK,EAAE,QAAgB,EAA4B,EAAE;IACpF,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IACjD,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IACrC,MAAM,KAAK,GAAG,QAAQ,CAAC,OAAO,CAAC,CAAC;IAChC,MAAM,YAAY,GAAG,uBAAuB,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;IAEzD,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,MAAM,EAAE,YAAY;QACpB,KAAK,EAAE,QAAQ,CAAC,OAAO,EAAE,QAAQ,CAAC;QAClC,SAAS,EAAE,KAAK,CAAC,MAAM;QACvB,SAAS,EAAE,KAAK,CAAC,MAAM;QACvB,OAAO,EAAE,YAAY,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,YAAY,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,wBAAwB;KACrF,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,EAAE,QAAgB,EAA4B,EAAE;IAClF,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IACjD,MAAM,IAAI,GAAG,YAAY,CAAC,OAAO,CAAC,CAAC;IACnC,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,OAAO,EAAE,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;IAC/D,MAAM,KAAK,GAAG,CAAC,GAAG,OAAO,CAAC,QAAQ,CAAC,4CAA4C,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAC9F,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAChB,CAAC;IACF,MAAM,cAAc,GAAG,IAAI,CAAC,cAAc,CAAC;IAE3C,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,QAAQ;QACR,KAAK;QACL,cAAc,EAAE,IAAI,CAAC,KAAK,CAAC,cAAc,CAAC;QAC1C,WAAW,EAAE,cAAc,CAAC,OAAO,CAAC;KACrC,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,EAAE,QAAgB,EAA4B,EAAE;IAClF,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IAEjD,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,OAAO,EAAE,aAAa,CAAC,OAAO,CAAC;KAChC,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,cAAc,GAAG,KAAK,EACjC,QAAgB,EAChB,OAAe,EACgB,EAAE;IACjC,MAAM,cAAc,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC;IAEtC,IAAI,CAAC,cAAc,EAAE,CAAC;QACpB,MAAM,IAAI,UAAU,CAAC;YACnB,IAAI,EAAE,kCAAkC;YACxC,OAAO,EAAE,6BAA6B;YACtC,GAAG,EAAE,4CAA4C;YACjD,GAAG,EAAE,wCAAwC;SAC9C,CAAC,CAAC;IACL,CAAC;IAED,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,QAAQ,CAAC,CAAC;IACjD,MAAM,YAAY,GAAG,cAAc,CAAC,WAAW,EAAE,CAAC;IAClD,MAAM,OAAO,GAAG,OAAO;SACpB,KAAK,CAAC,OAAO,CAAC;SACd,GAAG,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,CAAC;QACrB,IAAI,EAAE,KAAK,GAAG,CAAC;QACf,IAAI,EAAE,IAAI,CAAC,IAAI,EAAE;KAClB,CAAC,CAAC;SACF,MAAM,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,IAAI,CAAC,WAAW,EAAE,CAAC,QAAQ,CAAC,YAAY,CAAC,CAAC,CAAC;IAEtE,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,IAAI;QACZ,QAAQ;QACR,OAAO,EAAE,cAAc;QACvB,OAAO;KACR,CAAC;AACJ,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,EAClC,SAAiB,EACjB,UAKI,EAAE,EACuB,EAAE;IAC/B,MAAM,MAAM,GACV,OAAO,CAAC,MAAM;QACd,CAAC,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,UAAU,IAAI,EAAE,CAAC,CAAC,WAAW,EAAE,KAAK,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC;IAC1F,MAAM,OAAO,GAAG,MAAM,gBAAgB,CAAC,SAAS,EAAE,MAAM,CAAC,CAAC;IAC1D,MAAM,SAAS,GACb,MAAM,KAAK,MAAM;QACjB,CAAC,KAAK,EAAE,WAAW,EAAE,OAAO,EAAE,MAAM,CAAC,CAAC,QAAQ,CAAC,IAAI,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,WAAW,EAAE,CAAC;QACnF,CAAC,CAAC,OAAO;aACJ,OAAO,CAAC,eAAe,EAAE,EAAE,CAAC;aAC5B,OAAO,CAAC,0CAA0C,EAAE,IAAI,CAAC;QAC9D,CAAC,CAAC,OAAO,CAAC;IAEd,IAAI,OAAO,CAAC,UAAU,IAAI,OAAO,CAAC,GAAG,EAAE,CAAC;QACtC,IAAI,OAAO,CAAC,eAAe,KAAK,SAAS,IAAI,SAAS,KAAK,OAAO,CAAC,eAAe;YAChF,MAAM,IAAI,UAAU,CAAC;gBACnB,IAAI,EAAE,yBAAyB;gBAC/B,OAAO,EAAE,mDAAmD;gBAC5D,GAAG,EAAE,2GAA2G;gBAChH,GAAG,EAAE,wDAAwD;aAC9D,CAAC,CAAC;QACL,MAAM,4BAA4B,CAAC,SAAS,EAAE,OAAO,CAAC,UAAU,CAAC,CAAC;QAClE,MAAM,sBAAsB,CAAC,OAAO,CAAC,UAAU,EAAE,SAAS,EAAE;YAC1D,MAAM,EAAE;gBACN,IAAI,EAAE,0BAA0B;gBAChC,OAAO,EAAE,sCAAsC;gBAC/C,GAAG,EAAE,uCAAuC,OAAO,CAAC,UAAU,KAAK;gBACnE,GAAG,EAAE,2EAA2E;aACjF;YACD,MAAM,EAAE;gBACN,IAAI,EAAE,uBAAuB;gBAC7B,OAAO,EAAE,qCAAqC;gBAC9C,GAAG,EAAE,0BAA0B,OAAO,CAAC,UAAU,6CAA6C;gBAC9F,GAAG,EAAE,yEAAyE;aAC/E;SACF,CAAC,CAAC;IACL,CAAC;IAED,OAAO;QACL,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,OAAO,CAAC,UAAU,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI;QAC7D,SAAS;QACT,UAAU,EAAE,OAAO,CAAC,UAAU;QAC9B,MAAM;QACN,OAAO,EAAE,SAAS;QAClB,SAAS,EAAE,OAAO,CAAC,OAAO,CAAC,UAAU,IAAI,OAAO,CAAC,GAAG,CAAC;KACtD,CAAC;AACJ,CAAC,CAAC"}

package/dist/markdown.d.ts

@@ -0,0 +1,12 @@
+export interface MarkdownScan {
+    headings: Array<{
+        level: number;
+        title: string;
+        line: number;
+    }>;
+    prose: string[];
+    codeBlockCount: number;
+}
+/** Only real headings/prose participate: YAML metadata and fenced code do not. */
+export declare const scanMarkdown: (content: string) => MarkdownScan;
+//# sourceMappingURL=markdown.d.ts.map

package/dist/markdown.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"markdown.d.ts","sourceRoot":"","sources":["../src/markdown.ts"],"names":[],"mappings":"AAAA,MAAM,WAAW,YAAY;IAC3B,QAAQ,EAAE,KAAK,CAAC;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IAChE,KAAK,EAAE,MAAM,EAAE,CAAC;IAChB,cAAc,EAAE,MAAM,CAAC;CACxB;AAED,kFAAkF;AAClF,eAAO,MAAM,YAAY,GAAI,SAAS,MAAM,KAAG,YA+B9C,CAAC"}

package/dist/markdown.js

@@ -0,0 +1,35 @@
+/** Only real headings/prose participate: YAML metadata and fenced code do not. */
+export const scanMarkdown = (content) => {
+    const result = { headings: [], prose: [], codeBlockCount: 0 };
+    const lines = content.split(/\r?\n/);
+    const frontmatterEnd = lines[0] === '---' ? lines.indexOf('---', 1) : -1;
+    let fence;
+    for (let index = 0; index < lines.length; index++) {
+        if (frontmatterEnd > 0 && index <= frontmatterEnd)
+            continue;
+        const line = lines[index];
+        const marker = /^ {0,3}(`{3,}|~{3,})(.*)$/.exec(line);
+        if (fence) {
+            if (marker &&
+                marker[1][0] === fence.marker &&
+                marker[1].length >= fence.length &&
+                !marker[2].trim())
+                fence = undefined;
+            continue;
+        }
+        if (marker) {
+            fence = { marker: marker[1][0], length: marker[1].length };
+            result.codeBlockCount++;
+            continue;
+        }
+        if (/^(?: {4}|\t)/.test(line))
+            continue;
+        const heading = /^ {0,3}(#{1,6})\s+(.+?)(?:\s+#+)?\s*$/.exec(line);
+        if (heading)
+            result.headings.push({ level: heading[1].length, title: heading[2].trim(), line: index + 1 });
+        else if (line.trim())
+            result.prose.push(line.trim());
+    }
+    return result;
+};
+//# sourceMappingURL=markdown.js.map

package/dist/markdown.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"markdown.js","sourceRoot":"","sources":["../src/markdown.ts"],"names":[],"mappings":"AAMA,kFAAkF;AAClF,MAAM,CAAC,MAAM,YAAY,GAAG,CAAC,OAAe,EAAgB,EAAE;IAC5D,MAAM,MAAM,GAAiB,EAAE,QAAQ,EAAE,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE,cAAc,EAAE,CAAC,EAAE,CAAC;IAC5E,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IACrC,MAAM,cAAc,GAAG,KAAK,CAAC,CAAC,CAAC,KAAK,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACzE,IAAI,KAAqD,CAAC;IAC1D,KAAK,IAAI,KAAK,GAAG,CAAC,EAAE,KAAK,GAAG,KAAK,CAAC,MAAM,EAAE,KAAK,EAAE,EAAE,CAAC;QAClD,IAAI,cAAc,GAAG,CAAC,IAAI,KAAK,IAAI,cAAc;YAAE,SAAS;QAC5D,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC;QAC1B,MAAM,MAAM,GAAG,2BAA2B,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACtD,IAAI,KAAK,EAAE,CAAC;YACV,IACE,MAAM;gBACN,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,KAAK,CAAC,MAAM;gBAC7B,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,IAAI,KAAK,CAAC,MAAM;gBAChC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE;gBAEjB,KAAK,GAAG,SAAS,CAAC;YACpB,SAAS;QACX,CAAC;QACD,IAAI,MAAM,EAAE,CAAC;YACX,KAAK,GAAG,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC;YAC3D,MAAM,CAAC,cAAc,EAAE,CAAC;YACxB,SAAS;QACX,CAAC;QACD,IAAI,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC;YAAE,SAAS;QACxC,MAAM,OAAO,GAAG,uCAAuC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACnE,IAAI,OAAO;YACT,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,MAAM,EAAE,KAAK,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,EAAE,IAAI,EAAE,KAAK,GAAG,CAAC,EAAE,CAAC,CAAC;aAC3F,IAAI,IAAI,CAAC,IAAI,EAAE;YAAE,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC;IACvD,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC,CAAC"}

package/dist/pdf.d.ts

@@ -0,0 +1,27 @@
+/**
+ * PDF text extraction runtime shims (foundation-phase workaround).
+ *
+ * pdf-parse (the library we use for .pdf -> text) has a hard/optional dependency
+ * on @napi-rs/canvas for its full rendering pipeline. Even when we only request
+ * text extraction, in many Node.js environments the module will:
+ *   - log a noisy warning about the missing canvas package, and/or
+ *   - assume certain DOM types (DOMMatrix, ImageData, Path2D) exist on the global.
+ *
+ * We therefore:
+ *   1. Provide minimal, no-op shims for those three types **only if they are
+ *      not already present** on globalThis. This is the smallest possible
+ *      intervention that makes pdf-parse's text path work.
+ *   2. Temporarily silence the exact canvas warning during the dynamic import.
+ *
+ * This is deliberately a "clever" but contained hack. It violates the spirit of
+ * "readable code over clever code" and "no magic without control". A proper
+ * long-term fix would be to replace pdf-parse with a pure-JS text extractor
+ * (pdf.js node build, unpdf, or similar) that has no canvas requirement.
+ *
+ * The shims are intentionally named with a ZaoWu prefix inside the class
+ * definitions so that stack traces / debug output make the origin obvious.
+ *
+ * See also: packages/doc/README.md (workaround documented).
+ */
+export declare const extractPdf: (buffer: Buffer) => Promise<string>;
+//# sourceMappingURL=pdf.d.ts.map

package/dist/pdf.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"pdf.d.ts","sourceRoot":"","sources":["../src/pdf.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;GAwBG;AA0IH,eAAO,MAAM,UAAU,GAAU,QAAQ,MAAM,KAAG,OAAO,CAAC,MAAM,CAY/D,CAAC"}

package/dist/pdf.js

@@ -0,0 +1,153 @@
+/**
+ * PDF text extraction runtime shims (foundation-phase workaround).
+ *
+ * pdf-parse (the library we use for .pdf -> text) has a hard/optional dependency
+ * on @napi-rs/canvas for its full rendering pipeline. Even when we only request
+ * text extraction, in many Node.js environments the module will:
+ *   - log a noisy warning about the missing canvas package, and/or
+ *   - assume certain DOM types (DOMMatrix, ImageData, Path2D) exist on the global.
+ *
+ * We therefore:
+ *   1. Provide minimal, no-op shims for those three types **only if they are
+ *      not already present** on globalThis. This is the smallest possible
+ *      intervention that makes pdf-parse's text path work.
+ *   2. Temporarily silence the exact canvas warning during the dynamic import.
+ *
+ * This is deliberately a "clever" but contained hack. It violates the spirit of
+ * "readable code over clever code" and "no magic without control". A proper
+ * long-term fix would be to replace pdf-parse with a pure-JS text extractor
+ * (pdf.js node build, unpdf, or similar) that has no canvas requirement.
+ *
+ * The shims are intentionally named with a ZaoWu prefix inside the class
+ * definitions so that stack traces / debug output make the origin obvious.
+ *
+ * See also: packages/doc/README.md (workaround documented).
+ */
+const PDF_SHIM_KEYS = ['DOMMatrix', 'ImageData', 'Path2D'];
+let pdfParseModulePromise;
+const ensurePdfTextRuntime = () => {
+    const runtime = globalThis;
+    // Minimal DOMMatrix shim sufficient for pdf-parse text extraction.
+    runtime.DOMMatrix ??= class ZaoWuPdfDomMatrix {
+        a = 1;
+        b = 0;
+        c = 0;
+        d = 1;
+        e = 0;
+        f = 0;
+        constructor(init) {
+            if (Array.isArray(init) && init.length >= 6) {
+                [this.a, this.b, this.c, this.d, this.e, this.f] = init.map(Number).slice(0, 6);
+            }
+        }
+        multiplySelf() {
+            return this;
+        }
+        preMultiplySelf() {
+            return this;
+        }
+        translateSelf() {
+            return this;
+        }
+        scaleSelf() {
+            return this;
+        }
+        rotateSelf() {
+            return this;
+        }
+        invertSelf() {
+            return this;
+        }
+        transformPoint(point = {}) {
+            return {
+                x: point.x ?? 0,
+                y: point.y ?? 0,
+                z: point.z ?? 0,
+                w: point.w ?? 1,
+            };
+        }
+    };
+    // Minimal ImageData shim.
+    runtime.ImageData ??= class ZaoWuPdfImageData {
+        data;
+        width;
+        height;
+        constructor(dataOrWidth, width = 0, height = 0) {
+            this.width = typeof dataOrWidth === 'number' ? dataOrWidth : width;
+            this.height = height;
+            this.data =
+                dataOrWidth instanceof Uint8ClampedArray
+                    ? dataOrWidth
+                    : new Uint8ClampedArray(dataOrWidth * width * 4);
+        }
+    };
+    // Path2D is referenced but never actually used in our text-only path.
+    runtime.Path2D ??= class ZaoWuPdfPath2D {
+    };
+};
+const importPdfParse = async () => {
+    ensurePdfTextRuntime();
+    if (!pdfParseModulePromise) {
+        const originalWarn = console.warn;
+        // Only suppress the specific canvas-missing warning that pdf-parse emits
+        // when it can't find its optional native module. Everything else is passed
+        // through so we don't accidentally hide real problems.
+        console.warn = (...values) => {
+            const message = values.map(String).join(' ');
+            if (message.includes('Cannot load "@napi-rs/canvas" package')) {
+                return;
+            }
+            originalWarn(...values);
+        };
+        pdfParseModulePromise = import('pdf-parse')
+            .catch((error) => {
+            pdfParseModulePromise = undefined;
+            throw error;
+        })
+            .finally(() => {
+            console.warn = originalWarn;
+        });
+    }
+    return pdfParseModulePromise;
+};
+const extractPdfInternal = async (buffer) => {
+    const runtime = globalThis;
+    const saved = PDF_SHIM_KEYS.map((key) => ({ key, had: key in runtime, old: runtime[key] }));
+    try {
+        const { PDFParse } = await importPdfParse();
+        const parser = new PDFParse({ data: new Uint8Array(buffer) });
+        try {
+            const parsed = await parser.getText();
+            return parsed.text ?? '';
+        }
+        finally {
+            await parser.destroy();
+        }
+    }
+    finally {
+        for (const { key, had, old } of saved) {
+            if (!had)
+                delete runtime[key];
+            else
+                runtime[key] = old;
+        }
+    }
+};
+// Global runtime shims belong to one extraction at a time. Concurrent library
+// callers must not snapshot another extraction's temporary globals as originals.
+let extractionQueue = Promise.resolve();
+export const extractPdf = async (buffer) => {
+    const previous = extractionQueue;
+    let release;
+    extractionQueue = new Promise((resolve) => {
+        release = resolve;
+    });
+    await previous;
+    try {
+        return await extractPdfInternal(buffer);
+    }
+    finally {
+        release();
+    }
+};
+//# sourceMappingURL=pdf.js.map

package/dist/pdf.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"pdf.js","sourceRoot":"","sources":["../src/pdf.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;GAwBG;AAYH,MAAM,aAAa,GAAG,CAAC,WAAW,EAAE,WAAW,EAAE,QAAQ,CAAU,CAAC;AAIpE,IAAI,qBAA0D,CAAC;AAE/D,MAAM,oBAAoB,GAAG,GAAS,EAAE;IACtC,MAAM,OAAO,GAAG,UAAyC,CAAC;IAE1D,mEAAmE;IACnE,OAAO,CAAC,SAAS,KAAK,MAAM,iBAAiB;QAC3C,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QACN,CAAC,GAAG,CAAC,CAAC;QAEN,YAAY,IAAc;YACxB,IAAI,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,MAAM,IAAI,CAAC,EAAE,CAAC;gBAC5C,CAAC,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;YAClF,CAAC;QACH,CAAC;QAED,YAAY;YACV,OAAO,IAAI,CAAC;QACd,CAAC;QACD,eAAe;YACb,OAAO,IAAI,CAAC;QACd,CAAC;QACD,aAAa;YACX,OAAO,IAAI,CAAC;QACd,CAAC;QACD,SAAS;YACP,OAAO,IAAI,CAAC;QACd,CAAC;QACD,UAAU;YACR,OAAO,IAAI,CAAC;QACd,CAAC;QACD,UAAU;YACR,OAAO,IAAI,CAAC;QACd,CAAC;QAED,cAAc,CAAC,QAA4D,EAAE;YAC3E,OAAO;gBACL,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;gBACf,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;gBACf,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;gBACf,CAAC,EAAE,KAAK,CAAC,CAAC,IAAI,CAAC;aAChB,CAAC;QACJ,CAAC;KACF,CAAC;IAEF,0BAA0B;IAC1B,OAAO,CAAC,SAAS,KAAK,MAAM,iBAAiB;QAC3C,IAAI,CAAoB;QACxB,KAAK,CAAS;QACd,MAAM,CAAS;QAEf,YAAY,WAAuC,EAAE,KAAK,GAAG,CAAC,EAAE,MAAM,GAAG,CAAC;YACxE,IAAI,CAAC,KAAK,GAAG,OAAO,WAAW,KAAK,QAAQ,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,KAAK,CAAC;YACnE,IAAI,CAAC,MAAM,GAAG,MAAM,CAAC;YACrB,IAAI,CAAC,IAAI;gBACP,WAAW,YAAY,iBAAiB;oBACtC,CAAC,CAAC,WAAW;oBACb,CAAC,CAAC,IAAI,iBAAiB,CAAC,WAAW,GAAG,KAAK,GAAG,CAAC,CAAC,CAAC;QACvD,CAAC;KACF,CAAC;IAEF,sEAAsE;IACtE,OAAO,CAAC,MAAM,KAAK,MAAM,cAAc;KAAG,CAAC;AAC7C,CAAC,CAAC;AAEF,MAAM,cAAc,GAAG,KAAK,IAA6B,EAAE;IACzD,oBAAoB,EAAE,CAAC;IAEvB,IAAI,CAAC,qBAAqB,EAAE,CAAC;QAC3B,MAAM,YAAY,GAAG,OAAO,CAAC,IAAI,CAAC;QAElC,yEAAyE;QACzE,2EAA2E;QAC3E,uDAAuD;QACvD,OAAO,CAAC,IAAI,GAAG,CAAC,GAAG,MAAiB,EAAQ,EAAE;YAC5C,MAAM,OAAO,GAAG,MAAM,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC7C,IAAI,OAAO,CAAC,QAAQ,CAAC,uCAAuC,CAAC,EAAE,CAAC;gBAC9D,OAAO;YACT,CAAC;YACD,YAAY,CAAC,GAAG,MAAM,CAAC,CAAC;QAC1B,CAAC,CAAC;QAEF,qBAAqB,GAAG,MAAM,CAAC,WAAW,CAAC;aACxC,KAAK,CAAC,CAAC,KAAc,EAAE,EAAE;YACxB,qBAAqB,GAAG,SAAS,CAAC;YAClC,MAAM,KAAK,CAAC;QACd,CAAC,CAAC;aACD,OAAO,CAAC,GAAG,EAAE;YACZ,OAAO,CAAC,IAAI,GAAG,YAAY,CAAC;QAC9B,CAAC,CAAC,CAAC;IACP,CAAC;IAED,OAAO,qBAAqB,CAAC;AAC/B,CAAC,CAAC;AAEF,MAAM,kBAAkB,GAAG,KAAK,EAAE,MAAc,EAAmB,EAAE;IACnE,MAAM,OAAO,GAAG,UAAyC,CAAC;IAC1D,MAAM,KAAK,GAAG,aAAa,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,IAAI,OAAO,EAAE,GAAG,EAAE,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;IAC5F,IAAI,CAAC;QACH,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,cAAc,EAAE,CAAC;QAC5C,MAAM,MAAM,GAAG,IAAI,QAAQ,CAAC,EAAE,IAAI,EAAE,IAAI,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC;QAC9D,IAAI,CAAC;YACH,MAAM,MAAM,GAAG,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;YACtC,OAAO,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC;QAC3B,CAAC;gBAAS,CAAC;YACT,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;QACzB,CAAC;IACH,CAAC;YAAS,CAAC;QACT,KAAK,MAAM,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI,KAAK,EAAE,CAAC;YACtC,IAAI,CAAC,GAAG;gBAAE,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC;;gBACxB,OAAO,CAAC,GAAG,CAAa,GAAG,GAAG,CAAC;QACvC,CAAC;IACH,CAAC;AACH,CAAC,CAAC;AAEF,8EAA8E;AAC9E,iFAAiF;AACjF,IAAI,eAAe,GAAG,OAAO,CAAC,OAAO,EAAE,CAAC;AACxC,MAAM,CAAC,MAAM,UAAU,GAAG,KAAK,EAAE,MAAc,EAAmB,EAAE;IAClE,MAAM,QAAQ,GAAG,eAAe,CAAC;IACjC,IAAI,OAAoB,CAAC;IACzB,eAAe,GAAG,IAAI,OAAO,CAAO,CAAC,OAAO,EAAE,EAAE;QAC9C,OAAO,GAAG,OAAO,CAAC;IACpB,CAAC,CAAC,CAAC;IACH,MAAM,QAAQ,CAAC;IACf,IAAI,CAAC;QACH,OAAO,MAAM,kBAAkB,CAAC,MAAM,CAAC,CAAC;IAC1C,CAAC;YAAS,CAAC;QACT,OAAO,EAAE,CAAC;IACZ,CAAC;AACH,CAAC,CAAC"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@zaowu-labs/doc",
-  "version": "1.0.6",
+  "version": "1.1.0",
   "description": "ZaoWu document workflows",
   "type": "module",
   "license": "Apache-2.0",
@@ -20,17 +20,15 @@
     }
   },
   "dependencies": {
-    "mammoth": "1.12.0",
+    "@xmldom/xmldom": "^0.8.15",
+    "jszip": "^3.10.1",
     "pdf-parse": "^2.4.5",
-    "@zaowu-labs/core": "1.0.6"
+    "@zaowu-labs/core": "1.1.0"
   },
   "keywords": [
     "zaowu",
     "documents"
   ],
-  "devDependencies": {
-    "jszip": "^3.10.1"
-  },
   "repository": {
     "type": "git",
     "url": "https://github.com/zaowu-labs/zaowu.git",