@ncukondo/search-hub 0.12.0 → 0.12.2

package/dist/node_modules/@ncukondo/academic-fulltext/dist/convert/arxiv-html-parser.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"arxiv-html-parser.js","sources":["../../../../../../node_modules/@ncukondo/academic-fulltext/dist/convert/arxiv-html-parser.js"],"sourcesContent":["/**\n * arXiv HTML to JatsDocument parser.\n *\n * Parses LaTeXML-generated HTML from arXiv and converts it to the intermediate\n * JatsDocument representation, which can then be rendered to Markdown via writeMarkdown().\n */\nimport { parse as parseHtml } from \"node-html-parser\";\n// ---------------------------------------------------------------------------\n// Helpers\n// ---------------------------------------------------------------------------\n/** Query a single element, returning null if not found. */\nfunction q(root, selector) {\n    return root.querySelector(selector);\n}\n/** Query all matching elements. */\nfunction qa(root, selector) {\n    return root.querySelectorAll(selector);\n}\n/** Check if an element has a given CSS class. */\nfunction hasClass(el, cls) {\n    return el.classList.contains(cls);\n}\nfunction escapeRegex(str) {\n    return str.replace(/[.*+?^${}()|[\\]\\\\]/g, \"\\\\$&\");\n}\n/** Clone an element and remove <annotation> elements, then return textContent. */\nfunction stripAnnotations(el) {\n    const clone = parseHtml(el.outerHTML);\n    for (const ann of qa(clone, \"annotation\")) {\n        ann.remove();\n    }\n    return clone.textContent.trim();\n}\n// ---------------------------------------------------------------------------\n// Inline content parsing — class-based handlers\n// ---------------------------------------------------------------------------\n/** Try matching a class-based inline element. Returns null if no match. */\nfunction matchClassInline(el) {\n    if (hasClass(el, \"ltx_font_bold\")) {\n        return { type: \"bold\", children: parseInlineChildren(el) };\n    }\n    if (hasClass(el, \"ltx_font_italic\")) {\n        return { type: \"italic\", children: parseInlineChildren(el) };\n    }\n    if (hasClass(el, \"ltx_font_typewriter\")) {\n        return { type: \"code\", text: el.textContent.trim() };\n    }\n    return null;\n}\nfunction handleMath(el) {\n    const alt = el.getAttribute(\"alttext\");\n    const text = stripAnnotations(el);\n    const result = { type: \"inline-formula\", text };\n    if (alt)\n        result.tex = alt;\n    return result;\n}\nfunction handleAnchor(el) {\n    const href = el.getAttribute(\"href\") ?? \"\";\n    if (href.startsWith(\"#bib\")) {\n        return { type: \"citation\", refId: href.slice(1), text: el.textContent.trim() };\n    }\n    if (href.startsWith(\"http://\") || href.startsWith(\"https://\")) {\n        return { type: \"link\", url: href, children: parseInlineChildren(el) };\n    }\n    return { type: \"text\", text: el.textContent.trim() };\n}\n/** Dispatch map for tag names → inline handlers. */\nconst TAG_INLINE_HANDLERS = {\n    math: handleMath,\n    a: handleAnchor,\n    b: (el) => ({ type: \"bold\", children: parseInlineChildren(el) }),\n    strong: (el) => ({ type: \"bold\", children: parseInlineChildren(el) }),\n    i: (el) => ({ type: \"italic\", children: parseInlineChildren(el) }),\n    em: (el) => ({ type: \"italic\", children: parseInlineChildren(el) }),\n    code: (el) => ({ type: \"code\", text: el.textContent.trim() }),\n    sup: (el) => ({ type: \"superscript\", text: el.textContent.trim() }),\n    sub: (el) => ({ type: \"subscript\", text: el.textContent.trim() }),\n};\n/** Tags that should be recursed into as inline containers. */\nconst INLINE_CONTAINER_TAGS = new Set([\"span\", \"cite\"]);\n/** Process a single element node into inline content. */\nfunction processInlineElement(el) {\n    // Class-based detection first\n    const classMatch = matchClassInline(el);\n    if (classMatch)\n        return [classMatch];\n    // Tag-based dispatch\n    const tag = el.tagName?.toLowerCase() ?? \"\";\n    const handler = TAG_INLINE_HANDLERS[tag];\n    if (handler) {\n        const content = handler(el);\n        return content ? [content] : [];\n    }\n    // Inline containers: recurse\n    if (INLINE_CONTAINER_TAGS.has(tag)) {\n        return parseInlineChildren(el);\n    }\n    // Fallback: extract text\n    const text = el.textContent.trim();\n    return text ? [{ type: \"text\", text }] : [];\n}\n/** Parse inline content from child nodes of an element. */\nfunction parseInlineChildren(parent) {\n    const result = [];\n    for (const node of parent.childNodes) {\n        if (node.nodeType === 3) {\n            const text = node.textContent;\n            if (text)\n                result.push({ type: \"text\", text });\n        }\n        else if (node.nodeType === 1) {\n            result.push(...processInlineElement(node));\n        }\n    }\n    return result;\n}\n// ---------------------------------------------------------------------------\n// Block content parsing — individual block handlers\n// ---------------------------------------------------------------------------\nfunction parseParagraph(el) {\n    const innerP = q(el, \"p\");\n    return { type: \"paragraph\", content: parseInlineChildren(innerP ?? el) };\n}\n/**\n * Parse an ltx_para element which may contain a mix of <p> text and\n * block-level elements (equations, tables, etc.).\n */\nfunction parseLtxPara(el) {\n    const blocks = [];\n    for (const child of el.childNodes) {\n        if (child.nodeType !== 1)\n            continue;\n        const childEl = child;\n        const tag = childEl.tagName?.toLowerCase() ?? \"\";\n        if (tag === \"section\" || /^h[1-6]$/.test(tag))\n            continue;\n        const block = matchBlockByClass(childEl) ?? matchBlockByTag(childEl, tag);\n        if (block) {\n            blocks.push(block);\n        }\n    }\n    if (blocks.length === 0) {\n        blocks.push(parseParagraph(el));\n    }\n    return blocks;\n}\nfunction parseList(el) {\n    const tag = el.tagName?.toLowerCase() ?? \"\";\n    const ordered = tag === \"ol\" || hasClass(el, \"ltx_enumerate\");\n    const items = [];\n    for (const li of qa(el, \"li\")) {\n        items.push(parseInlineChildren(li));\n    }\n    return { type: \"list\", ordered, items };\n}\nfunction parseFigure(el) {\n    const labelEl = q(el, \".ltx_caption .ltx_tag_figure\");\n    const captionEl = q(el, \".ltx_caption\");\n    const id = el.getAttribute(\"id\");\n    let label;\n    if (labelEl) {\n        label = labelEl.textContent.trim().replace(/:$/, \"\");\n    }\n    let caption;\n    if (captionEl) {\n        const captionText = captionEl.textContent.trim();\n        caption = label\n            ? captionText.replace(new RegExp(`^${escapeRegex(label)}[:\\\\s]*`), \"\").trim()\n            : captionText;\n    }\n    const block = { type: \"figure\" };\n    if (id)\n        block.id = id;\n    if (label)\n        block.label = label;\n    if (caption)\n        block.caption = caption;\n    return block;\n}\nfunction parseTableHeaders(thead) {\n    const headers = [];\n    const headerRow = q(thead, \"tr\") ?? q(thead, \".ltx_tr\");\n    if (headerRow) {\n        for (const th of qa(headerRow, \"th, td, .ltx_td\")) {\n            headers.push(stripAnnotations(th));\n        }\n    }\n    return headers;\n}\nfunction parseTableRows(el, thead) {\n    const rows = [];\n    const tbody = q(el, \"tbody\") ?? q(el, \".ltx_tbody\") ?? el;\n    for (const tr of qa(tbody, \"tr, .ltx_tr\")) {\n        if (thead && tr.parentNode === thead)\n            continue;\n        const cells = [];\n        for (const td of qa(tr, \"td, th, .ltx_td\")) {\n            cells.push(stripAnnotations(td));\n        }\n        if (cells.length > 0)\n            rows.push(cells);\n    }\n    return rows;\n}\nfunction parseTable(el) {\n    const captionEl = q(el, \".ltx_caption\");\n    const thead = q(el, \"thead\") ?? q(el, \".ltx_thead\");\n    const headers = thead ? parseTableHeaders(thead) : [];\n    const rows = parseTableRows(el, thead);\n    if (captionEl) {\n        return { type: \"table\", caption: captionEl.textContent.trim(), headers, rows };\n    }\n    return { type: \"table\", headers, rows };\n}\nfunction parseFormula(el) {\n    const mathEl = q(el, \"math\");\n    const id = el.getAttribute(\"id\");\n    const labelEl = q(el, \".ltx_tag_equation\");\n    const label = labelEl ? labelEl.textContent.trim() : undefined;\n    let tex;\n    let text;\n    if (mathEl) {\n        const alt = mathEl.getAttribute(\"alttext\");\n        if (alt)\n            tex = alt;\n        text = stripAnnotations(mathEl);\n    }\n    else {\n        text = el.textContent.trim();\n    }\n    const block = { type: \"formula\" };\n    if (id)\n        block.id = id;\n    if (label)\n        block.label = label;\n    if (tex)\n        block.tex = tex;\n    if (text && !tex)\n        block.text = text;\n    return block;\n}\n// ---------------------------------------------------------------------------\n// Block content parsing — dispatch\n// ---------------------------------------------------------------------------\n/** Match block elements by CSS class (checked before tag-based matching). */\nfunction matchBlockByClass(el) {\n    if (hasClass(el, \"ltx_equation\") || hasClass(el, \"ltx_eqn_table\"))\n        return parseFormula(el);\n    if (hasClass(el, \"ltx_tabular\"))\n        return parseTable(el);\n    return null;\n}\n/** Match block elements by tag name. */\nfunction matchBlockByTag(el, tag) {\n    if (tag === \"p\")\n        return { type: \"paragraph\", content: parseInlineChildren(el) };\n    if (tag === \"ol\" || tag === \"ul\")\n        return parseList(el);\n    if (tag === \"figure\" && hasClass(el, \"ltx_table\"))\n        return parseTable(el);\n    if (tag === \"table\")\n        return parseTable(el);\n    if (tag === \"figure\")\n        return parseFigure(el);\n    if (tag === \"blockquote\")\n        return { type: \"blockquote\", content: parseInlineChildren(el) };\n    if (tag === \"pre\")\n        return { type: \"preformat\", text: el.textContent.trim() };\n    return null;\n}\n/** Try to parse an element as a specific block type. Returns null if not matched. */\nfunction matchBlock(el) {\n    const tag = el.tagName?.toLowerCase() ?? \"\";\n    if (tag === \"section\" || /^h[1-6]$/.test(tag))\n        return null;\n    return matchBlockByClass(el) ?? matchBlockByTag(el, tag);\n}\n/** Parse block content from a section's child elements. */\nfunction parseBlockContent(parent) {\n    const blocks = [];\n    for (const child of parent.childNodes) {\n        if (child.nodeType !== 1)\n            continue;\n        const el = child;\n        // ltx_para may contain mixed content (text + embedded equations/tables)\n        if (hasClass(el, \"ltx_para\")) {\n            blocks.push(...parseLtxPara(el));\n            continue;\n        }\n        const block = matchBlock(el);\n        if (block) {\n            blocks.push(block);\n            continue;\n        }\n        // Fallback: treat as paragraph if it has text and is not a skipped element\n        const tag = el.tagName?.toLowerCase() ?? \"\";\n        if (tag !== \"section\" && tag !== \"nav\" && !/^h[1-6]$/.test(tag)) {\n            const text = el.textContent.trim();\n            if (text)\n                blocks.push({ type: \"paragraph\", content: parseInlineChildren(el) });\n        }\n    }\n    return blocks;\n}\n// ---------------------------------------------------------------------------\n// Section parsing\n// ---------------------------------------------------------------------------\n/** Determine section level from CSS class. */\nfunction getSectionLevel(el) {\n    if (hasClass(el, \"ltx_subsection\"))\n        return 3;\n    if (hasClass(el, \"ltx_subsubsection\"))\n        return 4;\n    if (hasClass(el, \"ltx_paragraph\"))\n        return 5;\n    return 2;\n}\n/** Find the heading element for a section. */\nfunction findHeading(el, level) {\n    return (q(el, `:scope > h${level}`) ??\n        q(el, \":scope > h2\") ??\n        q(el, \":scope > h3\") ??\n        q(el, \":scope > h4\") ??\n        q(el, \":scope > h5\") ??\n        q(el, \":scope > h6\"));\n}\n/** Parse a section element and its subsections recursively. */\nfunction parseSection(el) {\n    const level = getSectionLevel(el);\n    const headingEl = findHeading(el, level);\n    const title = headingEl ? headingEl.textContent.trim() : \"\";\n    const content = parseBlockContent(el);\n    const subsections = [];\n    for (const child of el.childNodes) {\n        if (child.nodeType !== 1)\n            continue;\n        const childEl = child;\n        if (childEl.tagName?.toLowerCase() === \"section\") {\n            subsections.push(parseSection(childEl));\n        }\n    }\n    return { title, level, content, subsections };\n}\n// ---------------------------------------------------------------------------\n// Metadata parsing\n// ---------------------------------------------------------------------------\n/** Extract document title from the HTML. */\nfunction parseTitle(root) {\n    const titleEl = q(root, \".ltx_title.ltx_title_document\");\n    if (!titleEl)\n        return \"\";\n    const titleClone = parseHtml(titleEl.outerHTML);\n    for (const authorInTitle of qa(titleClone, \".ltx_authors\")) {\n        authorInTitle.remove();\n    }\n    return titleClone.textContent.trim();\n}\n/** Try to parse a comma-split part as an author. Returns null for affiliations. */\nfunction parseAuthorName(commaPart) {\n    // Take the first non-empty line (affiliations appear on subsequent lines after <br>)\n    const lines = commaPart.split(/\\n/);\n    const firstLine = lines.find((l) => l.trim() !== \"\")?.trim() ?? \"\";\n    if (!firstLine)\n        return null;\n    // Skip affiliation entries: start with digit (e.g. \"1IBM Research\", \"2 NASA MFSC\")\n    if (/^\\d/.test(firstLine))\n        return null;\n    // Remove trailing superscript digits (affiliation markers attached to names)\n    const name = firstLine.replace(/\\d+$/, \"\").trim();\n    if (!name)\n        return null;\n    const words = name.split(/\\s+/);\n    const surname = words.pop() ?? name;\n    const author = { surname };\n    if (words.length > 0)\n        author.givenNames = words.join(\" \");\n    return author;\n}\n/** Extract authors from the HTML. */\nfunction parseAuthors(root) {\n    const authors = [];\n    for (const authorEl of qa(root, \".ltx_authors .ltx_personname\")) {\n        const fullText = authorEl.textContent.trim();\n        if (!fullText)\n            continue;\n        // Split by comma to handle multiple authors in a single ltx_personname\n        for (const part of fullText.split(\",\")) {\n            const author = parseAuthorName(part);\n            if (author)\n                authors.push(author);\n        }\n    }\n    return authors;\n}\n/** Extract abstract text from the HTML. */\nfunction parseAbstract(root) {\n    const abstractEl = q(root, \".ltx_abstract\");\n    if (!abstractEl)\n        return undefined;\n    const clone = parseHtml(abstractEl.outerHTML);\n    const absTitle = q(clone, \".ltx_title\");\n    if (absTitle)\n        absTitle.remove();\n    for (const note of qa(clone, \".ltx_note\")) {\n        note.remove();\n    }\n    const text = clone.textContent.trim();\n    return text || undefined;\n}\n/** Parse document metadata from the HTML. */\nfunction parseMetadata(root) {\n    const title = parseTitle(root);\n    const authors = parseAuthors(root);\n    const abstract = parseAbstract(root);\n    const keywordEls = qa(root, \".ltx_keywords .ltx_text\");\n    const keywords = keywordEls.map((kw) => kw.textContent.trim()).filter(Boolean);\n    const metadata = { title, authors };\n    if (abstract)\n        metadata.abstract = abstract;\n    if (keywords.length > 0)\n        metadata.keywords = keywords;\n    return metadata;\n}\n// ---------------------------------------------------------------------------\n// References parsing\n// ---------------------------------------------------------------------------\n/** Extract reference text from a bibitem element. */\nfunction extractRefText(item) {\n    const bibBlock = q(item, \".ltx_bibblock\");\n    if (bibBlock)\n        return bibBlock.textContent.trim();\n    let text = item.textContent.trim();\n    const labelEl = q(item, \".ltx_tag_bibitem\");\n    if (labelEl) {\n        const labelText = labelEl.textContent.trim();\n        if (text.startsWith(labelText)) {\n            text = text.slice(labelText.length).trim();\n        }\n    }\n    return text;\n}\n/** Extract DOI from links in a bibitem. */\nfunction extractDoi(item) {\n    for (const link of qa(item, \"a\")) {\n        const href = link.getAttribute(\"href\") ?? \"\";\n        const doiMatch = href.match(/doi\\.org\\/(.+)/);\n        if (doiMatch)\n            return doiMatch[1];\n    }\n    return undefined;\n}\n/** Parse bibliography references. */\nfunction parseReferences(root) {\n    const refs = [];\n    for (const item of qa(root, \".ltx_bibitem\")) {\n        const id = item.getAttribute(\"id\") ?? `ref${refs.length + 1}`;\n        const text = extractRefText(item);\n        const doi = extractDoi(item);\n        const ref = { id, text };\n        if (doi)\n            ref.doi = doi;\n        refs.push(ref);\n    }\n    return refs;\n}\n// ---------------------------------------------------------------------------\n// Acknowledgments parsing\n// ---------------------------------------------------------------------------\nfunction parseAcknowledgments(root) {\n    const ackEl = q(root, \".ltx_acknowledgement\");\n    if (!ackEl)\n        return undefined;\n    const clone = parseHtml(ackEl.outerHTML);\n    const heading = q(clone, \".ltx_title\");\n    if (heading)\n        heading.remove();\n    const text = clone.textContent.trim();\n    return text || undefined;\n}\n// ---------------------------------------------------------------------------\n// Body sections parsing (shared logic)\n// ---------------------------------------------------------------------------\n/** Check if a section element is a body section (not bibliography or acknowledgements). */\nfunction isBodySection(el) {\n    return (el.tagName?.toLowerCase() === \"section\" &&\n        !hasClass(el, \"ltx_bibliography\") &&\n        !hasClass(el, \"ltx_acknowledgement\"));\n}\n/** Parse body sections from an article root element. */\nfunction parseBodySections(article) {\n    const sections = [];\n    for (const child of article.childNodes) {\n        if (child.nodeType !== 1)\n            continue;\n        const el = child;\n        if (isBodySection(el)) {\n            sections.push(parseSection(el));\n        }\n    }\n    return sections;\n}\n// ---------------------------------------------------------------------------\n// Public API\n// ---------------------------------------------------------------------------\n/**\n * Parse arXiv LaTeXML HTML into a JatsDocument.\n * @param html - Full HTML string from arXiv\n * @returns Complete parsed document\n */\nexport function parseArxivHtml(html) {\n    const root = parseHtml(html);\n    const metadata = parseMetadata(root);\n    const references = parseReferences(root);\n    const article = q(root, \"article.ltx_document\") ?? root;\n    const sections = parseBodySections(article);\n    const acknowledgments = parseAcknowledgments(root);\n    const doc = { metadata, sections, references };\n    if (acknowledgments)\n        doc.acknowledgments = acknowledgments;\n    return doc;\n}\n/**\n * Parse only metadata from arXiv HTML.\n */\nexport function parseArxivHtmlMetadata(html) {\n    return parseMetadata(parseHtml(html));\n}\n/**\n * Parse only body sections from arXiv HTML.\n */\nexport function parseArxivHtmlBody(html) {\n    const root = parseHtml(html);\n    const article = q(root, \"article.ltx_document\") ?? root;\n    return parseBodySections(article);\n}\n/**\n * Parse only references from arXiv HTML.\n */\nexport function parseArxivHtmlReferences(html) {\n    return parseReferences(parseHtml(html));\n}\n//# sourceMappingURL=arxiv-html-parser.js.map"],"names":["parseHtml"],"mappings":";AAWA,SAAS,EAAE,MAAM,UAAU;AACvB,SAAO,KAAK,cAAc,QAAQ;AACtC;AAEA,SAAS,GAAG,MAAM,UAAU;AACxB,SAAO,KAAK,iBAAiB,QAAQ;AACzC;AAEA,SAAS,SAAS,IAAI,KAAK;AACvB,SAAO,GAAG,UAAU,SAAS,GAAG;AACpC;AACA,SAAS,YAAY,KAAK;AACtB,SAAO,IAAI,QAAQ,uBAAuB,MAAM;AACpD;AAEA,SAAS,iBAAiB,IAAI;AAC1B,QAAM,QAAQA,YAAAA,MAAU,GAAG,SAAS;AACpC,aAAW,OAAO,GAAG,OAAO,YAAY,GAAG;AACvC,QAAI,OAAM;AAAA,EACd;AACA,SAAO,MAAM,YAAY,KAAI;AACjC;AAKA,SAAS,iBAAiB,IAAI;AAC1B,MAAI,SAAS,IAAI,eAAe,GAAG;AAC/B,WAAO,EAAE,MAAM,QAAQ,UAAU,oBAAoB,EAAE,EAAC;AAAA,EAC5D;AACA,MAAI,SAAS,IAAI,iBAAiB,GAAG;AACjC,WAAO,EAAE,MAAM,UAAU,UAAU,oBAAoB,EAAE,EAAC;AAAA,EAC9D;AACA,MAAI,SAAS,IAAI,qBAAqB,GAAG;AACrC,WAAO,EAAE,MAAM,QAAQ,MAAM,GAAG,YAAY,OAAM;AAAA,EACtD;AACA,SAAO;AACX;AACA,SAAS,WAAW,IAAI;AACpB,QAAM,MAAM,GAAG,aAAa,SAAS;AACrC,QAAM,OAAO,iBAAiB,EAAE;AAChC,QAAM,SAAS,EAAE,MAAM,kBAAkB,KAAI;AAC7C,MAAI;AACA,WAAO,MAAM;AACjB,SAAO;AACX;AACA,SAAS,aAAa,IAAI;AACtB,QAAM,OAAO,GAAG,aAAa,MAAM,KAAK;AACxC,MAAI,KAAK,WAAW,MAAM,GAAG;AACzB,WAAO,EAAE,MAAM,YAAY,OAAO,KAAK,MAAM,CAAC,GAAG,MAAM,GAAG,YAAY,KAAI,EAAE;AAAA,EAChF;AACA,MAAI,KAAK,WAAW,SAAS,KAAK,KAAK,WAAW,UAAU,GAAG;AAC3D,WAAO,EAAE,MAAM,QAAQ,KAAK,MAAM,UAAU,oBAAoB,EAAE,EAAC;AAAA,EACvE;AACA,SAAO,EAAE,MAAM,QAAQ,MAAM,GAAG,YAAY,OAAM;AACtD;AAEA,MAAM,sBAAsB;AAAA,EACxB,MAAM;AAAA,EACN,GAAG;AAAA,EACH,GAAG,CAAC,QAAQ,EAAE,MAAM,QAAQ,UAAU,oBAAoB,EAAE;EAC5D,QAAQ,CAAC,QAAQ,EAAE,MAAM,QAAQ,UAAU,oBAAoB,EAAE;EACjE,GAAG,CAAC,QAAQ,EAAE,MAAM,UAAU,UAAU,oBAAoB,EAAE;EAC9D,IAAI,CAAC,QAAQ,EAAE,MAAM,UAAU,UAAU,oBAAoB,EAAE;EAC/D,MAAM,CAAC,QAAQ,EAAE,MAAM,QAAQ,MAAM,GAAG,YAAY,KAAI;EACxD,KAAK,CAAC,QAAQ,EAAE,MAAM,eAAe,MAAM,GAAG,YAAY,KAAI;EAC9D,KAAK,CAAC,QAAQ,EAAE,MAAM,aAAa,MAAM,GAAG,YAAY,KAAI;AAChE;AAEA,MAAM,wBAAwB,oBAAI,IAAI,CAAC,QAAQ,MAAM,CAAC;AAEtD,SAAS,qBAAqB,IAAI;AAE9B,QAAM,aAAa,iBAAiB,EAAE;AACtC,MAAI;AACA,WAAO,CAAC,UAAU;AAEtB,QAAM,MAAM,GAAG,SAAS,YAAW,KAAM;AACzC,QAAM,UAAU,oBAAoB,GAAG;AACvC,MAAI,SAAS;AACT,UAAM,UAAU,QAAQ,EAAE;AAC1B,WAAO,UAAU,CAAC,OAAO,IAAI,CAAA;AAAA,EACjC;AAEA,MAAI,sBAAsB,IAAI,GAAG,GAAG;AAChC,WAAO,oBAAoB,EAAE;AAAA,EACjC;AAEA,QAAM,OAAO,GAAG,YAAY,KAAI;AAChC,SAAO,OAAO,CAAC,EAAE,MAAM,QAAQ,KAAI,CAAE,IAAI,CAAA;AAC7C;AAEA,SAAS,oBAAoB,QAAQ;AACjC,QAAM,SAAS,CAAA;AACf,aAAW,QAAQ,OAAO,YAAY;AAClC,QAAI,KAAK,aAAa,GAAG;AACrB,YAAM,OAAO,KAAK;AAClB,UAAI;AACA,eAAO,KAAK,EAAE,MAAM,QAAQ,KAAI,CAAE;AAAA,IAC1C,WACS,KAAK,aAAa,GAAG;AAC1B,aAAO,KAAK,GAAG,qBAAqB,IAAI,CAAC;AAAA,IAC7C;AAAA,EACJ;AACA,SAAO;AACX;AAIA,SAAS,eAAe,IAAI;AACxB,QAAM,SAAS,EAAE,IAAI,GAAG;AACxB,SAAO,EAAE,MAAM,aAAa,SAAS,oBAAoB,UAAU,EAAE,EAAC;AAC1E;AAKA,SAAS,aAAa,IAAI;AACtB,QAAM,SAAS,CAAA;AACf,aAAW,SAAS,GAAG,YAAY;AAC/B,QAAI,MAAM,aAAa;AACnB;AACJ,UAAM,UAAU;AAChB,UAAM,MAAM,QAAQ,SAAS,YAAW,KAAM;AAC9C,QAAI,QAAQ,aAAa,WAAW,KAAK,GAAG;AACxC;AACJ,UAAM,QAAQ,kBAAkB,OAAO,KAAK,gBAAgB,SAAS,GAAG;AACxE,QAAI,OAAO;AACP,aAAO,KAAK,KAAK;AAAA,IACrB;AAAA,EACJ;AACA,MAAI,OAAO,WAAW,GAAG;AACrB,WAAO,KAAK,eAAe,EAAE,CAAC;AAAA,EAClC;AACA,SAAO;AACX;AACA,SAAS,UAAU,IAAI;AACnB,QAAM,MAAM,GAAG,SAAS,YAAW,KAAM;AACzC,QAAM,UAAU,QAAQ,QAAQ,SAAS,IAAI,eAAe;AAC5D,QAAM,QAAQ,CAAA;AACd,aAAW,MAAM,GAAG,IAAI,IAAI,GAAG;AAC3B,UAAM,KAAK,oBAAoB,EAAE,CAAC;AAAA,EACtC;AACA,SAAO,EAAE,MAAM,QAAQ,SAAS,MAAK;AACzC;AACA,SAAS,YAAY,IAAI;AACrB,QAAM,UAAU,EAAE,IAAI,8BAA8B;AACpD,QAAM,YAAY,EAAE,IAAI,cAAc;AACtC,QAAM,KAAK,GAAG,aAAa,IAAI;AAC/B,MAAI;AACJ,MAAI,SAAS;AACT,YAAQ,QAAQ,YAAY,KAAI,EAAG,QAAQ,MAAM,EAAE;AAAA,EACvD;AACA,MAAI;AACJ,MAAI,WAAW;AACX,UAAM,cAAc,UAAU,YAAY,KAAI;AAC9C,cAAU,QACJ,YAAY,QAAQ,IAAI,OAAO,IAAI,YAAY,KAAK,CAAC,SAAS,GAAG,EAAE,EAAE,KAAI,IACzE;AAAA,EACV;AACA,QAAM,QAAQ,EAAE,MAAM,SAAQ;AAC9B,MAAI;AACA,UAAM,KAAK;AACf,MAAI;AACA,UAAM,QAAQ;AAClB,MAAI;AACA,UAAM,UAAU;AACpB,SAAO;AACX;AACA,SAAS,kBAAkB,OAAO;AAC9B,QAAM,UAAU,CAAA;AAChB,QAAM,YAAY,EAAE,OAAO,IAAI,KAAK,EAAE,OAAO,SAAS;AACtD,MAAI,WAAW;AACX,eAAW,MAAM,GAAG,WAAW,iBAAiB,GAAG;AAC/C,cAAQ,KAAK,iBAAiB,EAAE,CAAC;AAAA,IACrC;AAAA,EACJ;AACA,SAAO;AACX;AACA,SAAS,eAAe,IAAI,OAAO;AAC/B,QAAM,OAAO,CAAA;AACb,QAAM,QAAQ,EAAE,IAAI,OAAO,KAAK,EAAE,IAAI,YAAY,KAAK;AACvD,aAAW,MAAM,GAAG,OAAO,aAAa,GAAG;AACvC,QAAI,SAAS,GAAG,eAAe;AAC3B;AACJ,UAAM,QAAQ,CAAA;AACd,eAAW,MAAM,GAAG,IAAI,iBAAiB,GAAG;AACxC,YAAM,KAAK,iBAAiB,EAAE,CAAC;AAAA,IACnC;AACA,QAAI,MAAM,SAAS;AACf,WAAK,KAAK,KAAK;AAAA,EACvB;AACA,SAAO;AACX;AACA,SAAS,WAAW,IAAI;AACpB,QAAM,YAAY,EAAE,IAAI,cAAc;AACtC,QAAM,QAAQ,EAAE,IAAI,OAAO,KAAK,EAAE,IAAI,YAAY;AAClD,QAAM,UAAU,QAAQ,kBAAkB,KAAK,IAAI,CAAA;AACnD,QAAM,OAAO,eAAe,IAAI,KAAK;AACrC,MAAI,WAAW;AACX,WAAO,EAAE,MAAM,SAAS,SAAS,UAAU,YAAY,KAAI,GAAI,SAAS,KAAI;AAAA,EAChF;AACA,SAAO,EAAE,MAAM,SAAS,SAAS,KAAI;AACzC;AACA,SAAS,aAAa,IAAI;AACtB,QAAM,SAAS,EAAE,IAAI,MAAM;AAC3B,QAAM,KAAK,GAAG,aAAa,IAAI;AAC/B,QAAM,UAAU,EAAE,IAAI,mBAAmB;AACzC,QAAM,QAAQ,UAAU,QAAQ,YAAY,KAAI,IAAK;AACrD,MAAI;AACJ,MAAI;AACJ,MAAI,QAAQ;AACR,UAAM,MAAM,OAAO,aAAa,SAAS;AACzC,QAAI;AACA,YAAM;AACV,WAAO,iBAAiB,MAAM;AAAA,EAClC,OACK;AACD,WAAO,GAAG,YAAY,KAAI;AAAA,EAC9B;AACA,QAAM,QAAQ,EAAE,MAAM,UAAS;AAC/B,MAAI;AACA,UAAM,KAAK;AACf,MAAI;AACA,UAAM,QAAQ;AAClB,MAAI;AACA,UAAM,MAAM;AAChB,MAAI,QAAQ,CAAC;AACT,UAAM,OAAO;AACjB,SAAO;AACX;AAKA,SAAS,kBAAkB,IAAI;AAC3B,MAAI,SAAS,IAAI,cAAc,KAAK,SAAS,IAAI,eAAe;AAC5D,WAAO,aAAa,EAAE;AAC1B,MAAI,SAAS,IAAI,aAAa;AAC1B,WAAO,WAAW,EAAE;AACxB,SAAO;AACX;AAEA,SAAS,gBAAgB,IAAI,KAAK;AAC9B,MAAI,QAAQ;AACR,WAAO,EAAE,MAAM,aAAa,SAAS,oBAAoB,EAAE,EAAC;AAChE,MAAI,QAAQ,QAAQ,QAAQ;AACxB,WAAO,UAAU,EAAE;AACvB,MAAI,QAAQ,YAAY,SAAS,IAAI,WAAW;AAC5C,WAAO,WAAW,EAAE;AACxB,MAAI,QAAQ;AACR,WAAO,WAAW,EAAE;AACxB,MAAI,QAAQ;AACR,WAAO,YAAY,EAAE;AACzB,MAAI,QAAQ;AACR,WAAO,EAAE,MAAM,cAAc,SAAS,oBAAoB,EAAE,EAAC;AACjE,MAAI,QAAQ;AACR,WAAO,EAAE,MAAM,aAAa,MAAM,GAAG,YAAY,OAAM;AAC3D,SAAO;AACX;AAEA,SAAS,WAAW,IAAI;AACpB,QAAM,MAAM,GAAG,SAAS,YAAW,KAAM;AACzC,MAAI,QAAQ,aAAa,WAAW,KAAK,GAAG;AACxC,WAAO;AACX,SAAO,kBAAkB,EAAE,KAAK,gBAAgB,IAAI,GAAG;AAC3D;AAEA,SAAS,kBAAkB,QAAQ;AAC/B,QAAM,SAAS,CAAA;AACf,aAAW,SAAS,OAAO,YAAY;AACnC,QAAI,MAAM,aAAa;AACnB;AACJ,UAAM,KAAK;AAEX,QAAI,SAAS,IAAI,UAAU,GAAG;AAC1B,aAAO,KAAK,GAAG,aAAa,EAAE,CAAC;AAC/B;AAAA,IACJ;AACA,UAAM,QAAQ,WAAW,EAAE;AAC3B,QAAI,OAAO;AACP,aAAO,KAAK,KAAK;AACjB;AAAA,IACJ;AAEA,UAAM,MAAM,GAAG,SAAS,YAAW,KAAM;AACzC,QAAI,QAAQ,aAAa,QAAQ,SAAS,CAAC,WAAW,KAAK,GAAG,GAAG;AAC7D,YAAM,OAAO,GAAG,YAAY,KAAI;AAChC,UAAI;AACA,eAAO,KAAK,EAAE,MAAM,aAAa,SAAS,oBAAoB,EAAE,GAAG;AAAA,IAC3E;AAAA,EACJ;AACA,SAAO;AACX;AAKA,SAAS,gBAAgB,IAAI;AACzB,MAAI,SAAS,IAAI,gBAAgB;AAC7B,WAAO;AACX,MAAI,SAAS,IAAI,mBAAmB;AAChC,WAAO;AACX,MAAI,SAAS,IAAI,eAAe;AAC5B,WAAO;AACX,SAAO;AACX;AAEA,SAAS,YAAY,IAAI,OAAO;AAC5B,SAAQ,EAAE,IAAI,aAAa,KAAK,EAAE,KAC9B,EAAE,IAAI,aAAa,KACnB,EAAE,IAAI,aAAa,KACnB,EAAE,IAAI,aAAa,KACnB,EAAE,IAAI,aAAa,KACnB,EAAE,IAAI,aAAa;AAC3B;AAEA,SAAS,aAAa,IAAI;AACtB,QAAM,QAAQ,gBAAgB,EAAE;AAChC,QAAM,YAAY,YAAY,IAAI,KAAK;AACvC,QAAM,QAAQ,YAAY,UAAU,YAAY,KAAI,IAAK;AACzD,QAAM,UAAU,kBAAkB,EAAE;AACpC,QAAM,cAAc,CAAA;AACpB,aAAW,SAAS,GAAG,YAAY;AAC/B,QAAI,MAAM,aAAa;AACnB;AACJ,UAAM,UAAU;AAChB,QAAI,QAAQ,SAAS,YAAW,MAAO,WAAW;AAC9C,kBAAY,KAAK,aAAa,OAAO,CAAC;AAAA,IAC1C;AAAA,EACJ;AACA,SAAO,EAAE,OAAO,OAAO,SAAS,YAAW;AAC/C;AAKA,SAAS,WAAW,MAAM;AACtB,QAAM,UAAU,EAAE,MAAM,+BAA+B;AACvD,MAAI,CAAC;AACD,WAAO;AACX,QAAM,aAAaA,YAAAA,MAAU,QAAQ,SAAS;AAC9C,aAAW,iBAAiB,GAAG,YAAY,cAAc,GAAG;AACxD,kBAAc,OAAM;AAAA,EACxB;AACA,SAAO,WAAW,YAAY,KAAI;AACtC;AAEA,SAAS,gBAAgB,WAAW;AAEhC,QAAM,QAAQ,UAAU,MAAM,IAAI;AAClC,QAAM,YAAY,MAAM,KAAK,CAAC,MAAM,EAAE,WAAW,EAAE,GAAG,KAAI,KAAM;AAChE,MAAI,CAAC;AACD,WAAO;AAEX,MAAI,MAAM,KAAK,SAAS;AACpB,WAAO;AAEX,QAAM,OAAO,UAAU,QAAQ,QAAQ,EAAE,EAAE,KAAI;AAC/C,MAAI,CAAC;AACD,WAAO;AACX,QAAM,QAAQ,KAAK,MAAM,KAAK;AAC9B,QAAM,UAAU,MAAM,IAAG,KAAM;AAC/B,QAAM,SAAS,EAAE,QAAO;AACxB,MAAI,MAAM,SAAS;AACf,WAAO,aAAa,MAAM,KAAK,GAAG;AACtC,SAAO;AACX;AAEA,SAAS,aAAa,MAAM;AACxB,QAAM,UAAU,CAAA;AAChB,aAAW,YAAY,GAAG,MAAM,8BAA8B,GAAG;AAC7D,UAAM,WAAW,SAAS,YAAY,KAAI;AAC1C,QAAI,CAAC;AACD;AAEJ,eAAW,QAAQ,SAAS,MAAM,GAAG,GAAG;AACpC,YAAM,SAAS,gBAAgB,IAAI;AACnC,UAAI;AACA,gBAAQ,KAAK,MAAM;AAAA,IAC3B;AAAA,EACJ;AACA,SAAO;AACX;AAEA,SAAS,cAAc,MAAM;AACzB,QAAM,aAAa,EAAE,MAAM,eAAe;AAC1C,MAAI,CAAC;AACD,WAAO;AACX,QAAM,QAAQA,YAAAA,MAAU,WAAW,SAAS;AAC5C,QAAM,WAAW,EAAE,OAAO,YAAY;AACtC,MAAI;AACA,aAAS,OAAM;AACnB,aAAW,QAAQ,GAAG,OAAO,WAAW,GAAG;AACvC,SAAK,OAAM;AAAA,EACf;AACA,QAAM,OAAO,MAAM,YAAY,KAAI;AACnC,SAAO,QAAQ;AACnB;AAEA,SAAS,cAAc,MAAM;AACzB,QAAM,QAAQ,WAAW,IAAI;AAC7B,QAAM,UAAU,aAAa,IAAI;AACjC,QAAM,WAAW,cAAc,IAAI;AACnC,QAAM,aAAa,GAAG,MAAM,yBAAyB;AACrD,QAAM,WAAW,WAAW,IAAI,CAAC,OAAO,GAAG,YAAY,KAAI,CAAE,EAAE,OAAO,OAAO;AAC7E,QAAM,WAAW,EAAE,OAAO,QAAO;AACjC,MAAI;AACA,aAAS,WAAW;AACxB,MAAI,SAAS,SAAS;AAClB,aAAS,WAAW;AACxB,SAAO;AACX;AAKA,SAAS,eAAe,MAAM;AAC1B,QAAM,WAAW,EAAE,MAAM,eAAe;AACxC,MAAI;AACA,WAAO,SAAS,YAAY,KAAI;AACpC,MAAI,OAAO,KAAK,YAAY,KAAI;AAChC,QAAM,UAAU,EAAE,MAAM,kBAAkB;AAC1C,MAAI,SAAS;AACT,UAAM,YAAY,QAAQ,YAAY,KAAI;AAC1C,QAAI,KAAK,WAAW,SAAS,GAAG;AAC5B,aAAO,KAAK,MAAM,UAAU,MAAM,EAAE,KAAI;AAAA,IAC5C;AAAA,EACJ;AACA,SAAO;AACX;AAEA,SAAS,WAAW,MAAM;AACtB,aAAW,QAAQ,GAAG,MAAM,GAAG,GAAG;AAC9B,UAAM,OAAO,KAAK,aAAa,MAAM,KAAK;AAC1C,UAAM,WAAW,KAAK,MAAM,gBAAgB;AAC5C,QAAI;AACA,aAAO,SAAS,CAAC;AAAA,EACzB;AACA,SAAO;AACX;AAEA,SAAS,gBAAgB,MAAM;AAC3B,QAAM,OAAO,CAAA;AACb,aAAW,QAAQ,GAAG,MAAM,cAAc,GAAG;AACzC,UAAM,KAAK,KAAK,aAAa,IAAI,KAAK,MAAM,KAAK,SAAS,CAAC;AAC3D,UAAM,OAAO,eAAe,IAAI;AAChC,UAAM,MAAM,WAAW,IAAI;AAC3B,UAAM,MAAM,EAAE,IAAI,KAAI;AACtB,QAAI;AACA,UAAI,MAAM;AACd,SAAK,KAAK,GAAG;AAAA,EACjB;AACA,SAAO;AACX;AAIA,SAAS,qBAAqB,MAAM;AAChC,QAAM,QAAQ,EAAE,MAAM,sBAAsB;AAC5C,MAAI,CAAC;AACD,WAAO;AACX,QAAM,QAAQA,YAAAA,MAAU,MAAM,SAAS;AACvC,QAAM,UAAU,EAAE,OAAO,YAAY;AACrC,MAAI;AACA,YAAQ,OAAM;AAClB,QAAM,OAAO,MAAM,YAAY,KAAI;AACnC,SAAO,QAAQ;AACnB;AAKA,SAAS,cAAc,IAAI;AACvB,SAAQ,GAAG,SAAS,YAAW,MAAO,aAClC,CAAC,SAAS,IAAI,kBAAkB,KAChC,CAAC,SAAS,IAAI,qBAAqB;AAC3C;AAEA,SAAS,kBAAkB,SAAS;AAChC,QAAM,WAAW,CAAA;AACjB,aAAW,SAAS,QAAQ,YAAY;AACpC,QAAI,MAAM,aAAa;AACnB;AACJ,UAAM,KAAK;AACX,QAAI,cAAc,EAAE,GAAG;AACnB,eAAS,KAAK,aAAa,EAAE,CAAC;AAAA,IAClC;AAAA,EACJ;AACA,SAAO;AACX;AASO,SAAS,eAAe,MAAM;AACjC,QAAM,OAAOA,YAAAA,MAAU,IAAI;AAC3B,QAAM,WAAW,cAAc,IAAI;AACnC,QAAM,aAAa,gBAAgB,IAAI;AACvC,QAAM,UAAU,EAAE,MAAM,sBAAsB,KAAK;AACnD,QAAM,WAAW,kBAAkB,OAAO;AAC1C,QAAM,kBAAkB,qBAAqB,IAAI;AACjD,QAAM,MAAM,EAAE,UAAU,UAAU,WAAU;AAC5C,MAAI;AACA,QAAI,kBAAkB;AAC1B,SAAO;AACX;","x_google_ignoreList":[0]}

package/dist/node_modules/@ncukondo/academic-fulltext/dist/convert/index.js

@@ -0,0 +1,93 @@
+import { readFile, writeFile, stat } from "node:fs/promises";
+import { parseArxivHtml } from "./arxiv-html-parser.js";
+import { parseJatsMetadata, parseJatsBody, parseJatsReferences, parseJatsBackMatter } from "./jats-parser.js";
+import { writeMarkdown } from "./markdown-writer.js";
+async function convertPmcXmlToMarkdown(xmlPath, mdPath, metaPath) {
+  try {
+    const xml = await readFile(xmlPath, "utf-8");
+    const metadata = parseJatsMetadata(xml);
+    const sections = parseJatsBody(xml);
+    const references = parseJatsReferences(xml);
+    const backMatter = parseJatsBackMatter(xml);
+    const doc = { metadata, sections, references };
+    if (backMatter.acknowledgments)
+      doc.acknowledgments = backMatter.acknowledgments;
+    if (backMatter.appendices)
+      doc.appendices = backMatter.appendices;
+    if (backMatter.footnotes)
+      doc.footnotes = backMatter.footnotes;
+    if (backMatter.floats)
+      doc.floats = backMatter.floats;
+    if (backMatter.notes)
+      doc.notes = backMatter.notes;
+    const md = writeMarkdown(doc);
+    await writeFile(mdPath, md, "utf-8");
+    if (metaPath) {
+      try {
+        await stat(metaPath);
+        const metaRaw = await readFile(metaPath, "utf-8");
+        const meta = JSON.parse(metaRaw);
+        const mdStat = await stat(mdPath);
+        meta.files.markdown = {
+          filename: "fulltext.md",
+          source: "conversion",
+          retrievedAt: (/* @__PURE__ */ new Date()).toISOString(),
+          size: mdStat.size,
+          convertedFrom: "fulltext.xml"
+        };
+        await writeFile(metaPath, `${JSON.stringify(meta, null, 2)}
+`, "utf-8");
+      } catch {
+      }
+    }
+    const result = { success: true };
+    result.title = metadata.title;
+    result.sections = sections.length;
+    result.references = references.length;
+    return result;
+  } catch (err) {
+    const result = { success: false };
+    result.error = err instanceof Error ? err.message : String(err);
+    return result;
+  }
+}
+async function convertArxivHtmlToMarkdown(htmlPath, mdPath, metaPath) {
+  try {
+    const html = await readFile(htmlPath, "utf-8");
+    const doc = parseArxivHtml(html);
+    const md = writeMarkdown(doc);
+    await writeFile(mdPath, md, "utf-8");
+    if (metaPath) {
+      try {
+        await stat(metaPath);
+        const metaRaw = await readFile(metaPath, "utf-8");
+        const meta = JSON.parse(metaRaw);
+        const mdStat = await stat(mdPath);
+        meta.files.markdown = {
+          filename: "fulltext.md",
+          source: "conversion",
+          retrievedAt: (/* @__PURE__ */ new Date()).toISOString(),
+          size: mdStat.size,
+          convertedFrom: "fulltext.html"
+        };
+        await writeFile(metaPath, `${JSON.stringify(meta, null, 2)}
+`, "utf-8");
+      } catch {
+      }
+    }
+    const result = { success: true };
+    result.title = doc.metadata.title;
+    result.sections = doc.sections.length;
+    result.references = doc.references.length;
+    return result;
+  } catch (err) {
+    const result = { success: false };
+    result.error = err instanceof Error ? err.message : String(err);
+    return result;
+  }
+}
+export {
+  convertArxivHtmlToMarkdown,
+  convertPmcXmlToMarkdown
+};
+//# sourceMappingURL=index.js.map

package/dist/node_modules/@ncukondo/academic-fulltext/dist/convert/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sources":["../../../../../../node_modules/@ncukondo/academic-fulltext/dist/convert/index.js"],"sourcesContent":["/**\n * Conversion orchestrator for PMC XML and arXiv HTML to Markdown.\n *\n * Ties together the parsers and Markdown writer with file I/O.\n */\nimport { readFile, stat, writeFile } from \"node:fs/promises\";\nimport { parseArxivHtml } from \"./arxiv-html-parser.js\";\nimport { parseJatsBackMatter, parseJatsBody, parseJatsMetadata, parseJatsReferences, } from \"./jats-parser.js\";\nimport { writeMarkdown } from \"./markdown-writer.js\";\n/**\n * Convert a PMC JATS XML file to Markdown.\n *\n * Reads the XML, parses it into a JatsDocument, writes Markdown,\n * and optionally updates meta.json.\n */\nexport async function convertPmcXmlToMarkdown(xmlPath, mdPath, metaPath) {\n    try {\n        const xml = await readFile(xmlPath, \"utf-8\");\n        // Parse\n        const metadata = parseJatsMetadata(xml);\n        const sections = parseJatsBody(xml);\n        const references = parseJatsReferences(xml);\n        const backMatter = parseJatsBackMatter(xml);\n        const doc = { metadata, sections, references };\n        if (backMatter.acknowledgments)\n            doc.acknowledgments = backMatter.acknowledgments;\n        if (backMatter.appendices)\n            doc.appendices = backMatter.appendices;\n        if (backMatter.footnotes)\n            doc.footnotes = backMatter.footnotes;\n        if (backMatter.floats)\n            doc.floats = backMatter.floats;\n        if (backMatter.notes)\n            doc.notes = backMatter.notes;\n        // Write Markdown\n        const md = writeMarkdown(doc);\n        await writeFile(mdPath, md, \"utf-8\");\n        // Update meta.json if path provided and file exists\n        if (metaPath) {\n            try {\n                await stat(metaPath);\n                const metaRaw = await readFile(metaPath, \"utf-8\");\n                const meta = JSON.parse(metaRaw);\n                const mdStat = await stat(mdPath);\n                meta.files.markdown = {\n                    filename: \"fulltext.md\",\n                    source: \"conversion\",\n                    retrievedAt: new Date().toISOString(),\n                    size: mdStat.size,\n                    convertedFrom: \"fulltext.xml\",\n                };\n                await writeFile(metaPath, `${JSON.stringify(meta, null, 2)}\\n`, \"utf-8\");\n            }\n            catch {\n                // meta.json doesn't exist or can't be read, skip update\n            }\n        }\n        const result = { success: true };\n        result.title = metadata.title;\n        result.sections = sections.length;\n        result.references = references.length;\n        return result;\n    }\n    catch (err) {\n        const result = { success: false };\n        result.error = err instanceof Error ? err.message : String(err);\n        return result;\n    }\n}\n/**\n * Convert an arXiv HTML file to Markdown.\n *\n * Reads the HTML, parses it into a JatsDocument via the arXiv HTML parser,\n * writes Markdown using the shared writer, and optionally updates meta.json.\n */\nexport async function convertArxivHtmlToMarkdown(htmlPath, mdPath, metaPath) {\n    try {\n        const html = await readFile(htmlPath, \"utf-8\");\n        const doc = parseArxivHtml(html);\n        // Write Markdown\n        const md = writeMarkdown(doc);\n        await writeFile(mdPath, md, \"utf-8\");\n        // Update meta.json if path provided and file exists\n        if (metaPath) {\n            try {\n                await stat(metaPath);\n                const metaRaw = await readFile(metaPath, \"utf-8\");\n                const meta = JSON.parse(metaRaw);\n                const mdStat = await stat(mdPath);\n                meta.files.markdown = {\n                    filename: \"fulltext.md\",\n                    source: \"conversion\",\n                    retrievedAt: new Date().toISOString(),\n                    size: mdStat.size,\n                    convertedFrom: \"fulltext.html\",\n                };\n                await writeFile(metaPath, `${JSON.stringify(meta, null, 2)}\\n`, \"utf-8\");\n            }\n            catch {\n                // meta.json doesn't exist or can't be read, skip update\n            }\n        }\n        const result = { success: true };\n        result.title = doc.metadata.title;\n        result.sections = doc.sections.length;\n        result.references = doc.references.length;\n        return result;\n    }\n    catch (err) {\n        const result = { success: false };\n        result.error = err instanceof Error ? err.message : String(err);\n        return result;\n    }\n}\n//# sourceMappingURL=index.js.map"],"names":[],"mappings":";;;;AAeO,eAAe,wBAAwB,SAAS,QAAQ,UAAU;AACrE,MAAI;AACA,UAAM,MAAM,MAAM,SAAS,SAAS,OAAO;AAE3C,UAAM,WAAW,kBAAkB,GAAG;AACtC,UAAM,WAAW,cAAc,GAAG;AAClC,UAAM,aAAa,oBAAoB,GAAG;AAC1C,UAAM,aAAa,oBAAoB,GAAG;AAC1C,UAAM,MAAM,EAAE,UAAU,UAAU,WAAU;AAC5C,QAAI,WAAW;AACX,UAAI,kBAAkB,WAAW;AACrC,QAAI,WAAW;AACX,UAAI,aAAa,WAAW;AAChC,QAAI,WAAW;AACX,UAAI,YAAY,WAAW;AAC/B,QAAI,WAAW;AACX,UAAI,SAAS,WAAW;AAC5B,QAAI,WAAW;AACX,UAAI,QAAQ,WAAW;AAE3B,UAAM,KAAK,cAAc,GAAG;AAC5B,UAAM,UAAU,QAAQ,IAAI,OAAO;AAEnC,QAAI,UAAU;AACV,UAAI;AACA,cAAM,KAAK,QAAQ;AACnB,cAAM,UAAU,MAAM,SAAS,UAAU,OAAO;AAChD,cAAM,OAAO,KAAK,MAAM,OAAO;AAC/B,cAAM,SAAS,MAAM,KAAK,MAAM;AAChC,aAAK,MAAM,WAAW;AAAA,UAClB,UAAU;AAAA,UACV,QAAQ;AAAA,UACR,cAAa,oBAAI,KAAI,GAAG,YAAW;AAAA,UACnC,MAAM,OAAO;AAAA,UACb,eAAe;AAAA,QACnC;AACgB,cAAM,UAAU,UAAU,GAAG,KAAK,UAAU,MAAM,MAAM,CAAC,CAAC;AAAA,GAAM,OAAO;AAAA,MAC3E,QACM;AAAA,MAEN;AAAA,IACJ;AACA,UAAM,SAAS,EAAE,SAAS,KAAI;AAC9B,WAAO,QAAQ,SAAS;AACxB,WAAO,WAAW,SAAS;AAC3B,WAAO,aAAa,WAAW;AAC/B,WAAO;AAAA,EACX,SACO,KAAK;AACR,UAAM,SAAS,EAAE,SAAS,MAAK;AAC/B,WAAO,QAAQ,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC9D,WAAO;AAAA,EACX;AACJ;AAOO,eAAe,2BAA2B,UAAU,QAAQ,UAAU;AACzE,MAAI;AACA,UAAM,OAAO,MAAM,SAAS,UAAU,OAAO;AAC7C,UAAM,MAAM,eAAe,IAAI;AAE/B,UAAM,KAAK,cAAc,GAAG;AAC5B,UAAM,UAAU,QAAQ,IAAI,OAAO;AAEnC,QAAI,UAAU;AACV,UAAI;AACA,cAAM,KAAK,QAAQ;AACnB,cAAM,UAAU,MAAM,SAAS,UAAU,OAAO;AAChD,cAAM,OAAO,KAAK,MAAM,OAAO;AAC/B,cAAM,SAAS,MAAM,KAAK,MAAM;AAChC,aAAK,MAAM,WAAW;AAAA,UAClB,UAAU;AAAA,UACV,QAAQ;AAAA,UACR,cAAa,oBAAI,KAAI,GAAG,YAAW;AAAA,UACnC,MAAM,OAAO;AAAA,UACb,eAAe;AAAA,QACnC;AACgB,cAAM,UAAU,UAAU,GAAG,KAAK,UAAU,MAAM,MAAM,CAAC,CAAC;AAAA,GAAM,OAAO;AAAA,MAC3E,QACM;AAAA,MAEN;AAAA,IACJ;AACA,UAAM,SAAS,EAAE,SAAS,KAAI;AAC9B,WAAO,QAAQ,IAAI,SAAS;AAC5B,WAAO,WAAW,IAAI,SAAS;AAC/B,WAAO,aAAa,IAAI,WAAW;AACnC,WAAO;AAAA,EACX,SACO,KAAK;AACR,UAAM,SAAS,EAAE,SAAS,MAAK;AAC/B,WAAO,QAAQ,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC9D,WAAO;AAAA,EACX;AACJ;","x_google_ignoreList":[0]}