kordoc 4.0.5 → 4.0.7

package/dist/index.js

@@ -20,10 +20,10 @@ import {
   sanitizeHref,
   stripDtd,
   toArrayBuffer
-} from "./chunk-FKK4K53F.js";
+} from "./chunk-5OKHAJ62.js";
 import {
   parsePageRange
-} from "./chunk-GE43BE46.js";
+} from "./chunk-AIQ3ISQU.js";
 
 // src/index.ts
 import { readFile } from "fs/promises";
@@ -2229,8 +2229,15 @@ function collectSubListText(el, ctx, depth = 0) {
     if (tag === "p" || tag === "para") {
       const t = extractParagraphInfo(ch, ctx.styleMap, ctx).text;
       if (t) parts.push(t);
+      const tbls = [];
+      findTopLevelTbls(ch, tbls);
+      for (const tbl2 of tbls) {
+        const flat = flattenSubListTable(tbl2, ctx, depth);
+        if (flat) parts.push(flat);
+      }
     } else if (tag === "tbl") {
-      continue;
+      const flat = flattenSubListTable(ch, ctx, depth);
+      if (flat) parts.push(flat);
     } else {
       const t = collectSubListText(ch, ctx, depth + 1);
       if (t) parts.push(t);
@@ -2238,6 +2245,29 @@ function collectSubListText(el, ctx, depth = 0) {
   }
   return parts.join("\n").trim();
 }
+function flattenSubListTable(el, ctx, depth) {
+  const sink = [];
+  const st = { rows: [], currentRow: [], cell: null };
+  walkSection(el, sink, st, [], ctx, depth + 1);
+  let flat = convertTableToText(st.rows);
+  if (st.caption) flat = st.caption + (flat ? "\n" + flat : "");
+  return flat;
+}
+function findTopLevelTbls(el, out, depth = 0) {
+  if (depth > MAX_XML_DEPTH) return;
+  const kids2 = el.childNodes;
+  if (!kids2) return;
+  for (let i = 0; i < kids2.length; i++) {
+    const ch = kids2[i];
+    if (ch.nodeType !== 1) continue;
+    const tag = (ch.tagName || ch.localName || "").replace(/^[^:]+:/, "");
+    if (tag === "tbl") {
+      out.push(ch);
+      continue;
+    }
+    findTopLevelTbls(ch, out, depth + 1);
+  }
+}
 function walkParagraphChildren(node, blocks, tableCtx, tableStack, ctx, depth = 0) {
   if (depth > MAX_XML_DEPTH) return tableCtx;
   const children = node.childNodes;
@@ -18027,7 +18057,7 @@ async function parseXlsxDocument(buffer, options) {
   }
   let pageFilter = null;
   if (options?.pages) {
-    const { parsePageRange: parsePageRange2 } = await import("./page-range-LNMXJU6W.js");
+    const { parsePageRange: parsePageRange2 } = await import("./page-range-CIRRJPXJ.js");
     pageFilter = parsePageRange2(options.pages, sheets.length);
   }
   const blocks = [];
@@ -18665,7 +18695,7 @@ async function parseXlsDocument(buffer, options) {
   const totalSheets = Math.min(globals.sheets.length, MAX_SHEETS2);
   let pageFilter = null;
   if (options?.pages) {
-    const { parsePageRange: parsePageRange2 } = await import("./page-range-LNMXJU6W.js");
+    const { parsePageRange: parsePageRange2 } = await import("./page-range-CIRRJPXJ.js");
     pageFilter = parsePageRange2(options.pages, totalSheets);
   }
   const allBlocks = [];
@@ -19267,6 +19297,102 @@ function extractRun(r) {
   }
   return { text, bold, italic };
 }
+function fieldUrlFromInstr(instr) {
+  if (!/HYPERLINK\b/i.test(instr)) return null;
+  const urlM = instr.match(/HYPERLINK\s+"([^"]*)"/i);
+  const anchorM = instr.match(/\\l\s+"([^"]*)"/i);
+  if (urlM && urlM[1]) {
+    return anchorM && anchorM[1] ? `${urlM[1]}#${anchorM[1]}` : urlM[1];
+  }
+  if (anchorM && anchorM[1]) return `#${anchorM[1]}`;
+  return null;
+}
+function makeLink(text, rawHref) {
+  if (!text) return "";
+  if (!rawHref) return text;
+  const href = sanitizeHref(rawHref);
+  return href ? `[${text}](${href})` : text;
+}
+function collectInline(p, footnotes, rels) {
+  const parts = [];
+  let bold = false;
+  let italic = false;
+  let footnoteText;
+  let fieldActive = false;
+  let fieldStage = "instr";
+  let fieldInstr = "";
+  let fieldDisplay = "";
+  const flushField = () => {
+    if (!fieldActive) return;
+    if (fieldDisplay) {
+      const url = fieldUrlFromInstr(fieldInstr);
+      parts.push(url ? makeLink(fieldDisplay, url) : fieldDisplay);
+    }
+    fieldActive = false;
+    fieldStage = "instr";
+    fieldInstr = "";
+    fieldDisplay = "";
+  };
+  for (const el of effectiveChildElements(p)) {
+    if (matchesLocal(el, "hyperlink")) {
+      const rId = getAttr(el, "id");
+      const anchor = getAttr(el, "anchor");
+      const t = findElements(el, "r").map((r) => extractRun(r).text).join("");
+      if (!t) continue;
+      let raw = null;
+      if (rId && rels.has(rId)) raw = rels.get(rId);
+      else if (anchor) raw = `#${anchor}`;
+      if (fieldActive && fieldStage === "display") fieldDisplay += t;
+      else parts.push(makeLink(t, raw));
+      continue;
+    }
+    if (matchesLocal(el, "fldSimple")) {
+      const t = findElements(el, "r").map((r) => extractRun(r).text).join("");
+      if (!t) continue;
+      const url = fieldUrlFromInstr(getAttr(el, "instr") ?? "");
+      parts.push(url ? makeLink(t, url) : t);
+      continue;
+    }
+    if (matchesLocal(el, "r")) {
+      for (const fc of getChildElements(el, "fldChar")) {
+        const ty = getAttr(fc, "fldCharType");
+        if (ty === "begin") {
+          flushField();
+          fieldActive = true;
+          fieldStage = "instr";
+        } else if (ty === "separate") {
+          if (fieldActive) fieldStage = "display";
+        } else if (ty === "end") flushField();
+      }
+      if (fieldActive && fieldStage === "instr") {
+        for (const it of getChildElements(el, "instrText")) fieldInstr += it.textContent ?? "";
+      }
+      const rr = extractRun(el);
+      if (rr.bold) bold = true;
+      if (rr.italic) italic = true;
+      const fnRefEls = getChildElements(el, "footnoteReference");
+      if (fnRefEls.length > 0) {
+        const fnId = getAttr(fnRefEls[0], "id");
+        if (fnId && footnotes.has(fnId)) footnoteText = footnotes.get(fnId);
+      }
+      if (rr.text) {
+        if (fieldActive && fieldStage === "display") fieldDisplay += rr.text;
+        else if (fieldActive && fieldStage === "instr") {
+        } else parts.push(rr.text);
+      }
+      continue;
+    }
+  }
+  flushField();
+  for (const om of collectOmmlRoots(p)) {
+    const latex = ommlElementToLatex(om);
+    if (!latex) continue;
+    if (isDisplayMath(om)) parts.push(" $$" + latex + "$$ ");
+    else parts.push(" $" + latex + "$ ");
+  }
+  const text = parts.join("").replace(/[ \t]{2,}/g, " ").trim();
+  return { text, bold, italic, footnoteText };
+}
 function parseParagraph2(p, styles, numbering, footnotes, rels) {
   const pPrEls = getChildElements(p, "pPr");
   let styleId = "";
@@ -19283,54 +19409,7 @@ function parseParagraph2(p, styles, numbering, footnotes, rels) {
       ilvl = ilvlEls.length > 0 ? parseInt(getAttr(ilvlEls[0], "val") ?? "0", 10) : 0;
     }
   }
-  const parts = [];
-  let hasBold = false;
-  let hasItalic = false;
-  let href;
-  let footnoteText;
-  const hyperlinks = getChildElements(p, "hyperlink");
-  const hyperlinkTexts = /* @__PURE__ */ new Set();
-  for (const hl of hyperlinks) {
-    const rId = getAttr(hl, "id");
-    const hlText = [];
-    const runs2 = findElements(hl, "r");
-    for (const r of runs2) {
-      const result = extractRun(r);
-      hlText.push(result.text);
-    }
-    const text2 = hlText.join("");
-    if (text2) {
-      hyperlinkTexts.add(text2);
-      if (rId && rels.has(rId)) {
-        href = rels.get(rId);
-        parts.push(text2);
-      } else {
-        parts.push(text2);
-      }
-    }
-  }
-  const runs = getChildElements(p, "r");
-  for (const r of runs) {
-    if (r.parentNode && r.parentNode.localName === "hyperlink") continue;
-    const result = extractRun(r);
-    if (result.bold) hasBold = true;
-    if (result.italic) hasItalic = true;
-    const fnRefEls = getChildElements(r, "footnoteReference");
-    if (fnRefEls.length > 0) {
-      const fnId = getAttr(fnRefEls[0], "id");
-      if (fnId && footnotes.has(fnId)) {
-        footnoteText = footnotes.get(fnId);
-      }
-    }
-    if (result.text) parts.push(result.text);
-  }
-  for (const om of collectOmmlRoots(p)) {
-    const latex = ommlElementToLatex(om);
-    if (!latex) continue;
-    if (isDisplayMath(om)) parts.push(" $$" + latex + "$$ ");
-    else parts.push(" $" + latex + "$ ");
-  }
-  const text = parts.join("").replace(/[ \t]{2,}/g, " ").trim();
+  const { text, bold: hasBold, italic: hasItalic, footnoteText } = collectInline(p, footnotes, rels);
   if (!text) return null;
   const style = styles.get(styleId);
   if (style?.outlineLevel !== void 0 && style.outlineLevel >= 0 && style.outlineLevel <= 5) {
@@ -19350,7 +19429,6 @@ function parseParagraph2(p, styles, numbering, footnotes, rels) {
   if (hasBold || hasItalic) {
     block.style = { bold: hasBold || void 0, italic: hasItalic || void 0 };
   }
-  if (href) block.href = href;
   if (footnoteText) block.footnoteText = footnoteText;
   return block;
 }
@@ -19451,46 +19529,61 @@ function collectCellText(tc2, styles, numbering, footnotes, rels, depth) {
   }
   return parts;
 }
-async function extractImages(zip, rels, doc, warnings) {
-  const blocks = [];
+async function buildImageMap(zip, rels, doc, warnings) {
+  const imageMap = /* @__PURE__ */ new Map();
   const images = [];
-  const drawingElements = findElements(doc.documentElement, "drawing");
   let imgIdx = 0;
-  for (const drawing of drawingElements) {
-    const blips = findElements(drawing, "blip");
-    for (const blip of blips) {
-      const embedId = getAttr(blip, "embed");
-      if (!embedId) continue;
-      const target = rels.get(embedId);
-      if (!target) continue;
-      const imgPath = target.startsWith("/") ? target.slice(1) : target.startsWith("word/") ? target : `word/${target}`;
-      const imgFile = zip.file(imgPath);
-      if (!imgFile) continue;
-      try {
-        const data = await imgFile.async("uint8array");
-        imgIdx++;
-        const ext = imgPath.split(".").pop()?.toLowerCase() ?? "png";
-        const mimeMap = {
-          png: "image/png",
-          jpg: "image/jpeg",
-          jpeg: "image/jpeg",
-          gif: "image/gif",
-          bmp: "image/bmp",
-          wmf: "image/wmf",
-          emf: "image/emf"
-        };
-        const filename = `image_${String(imgIdx).padStart(3, "0")}.${ext}`;
-        images.push({ filename, data, mimeType: mimeMap[ext] ?? "image/png" });
-        blocks.push({ type: "image", text: filename });
-      } catch (err) {
-        warnings.push({
-          code: "SKIPPED_IMAGE",
-          message: `DOCX \uC774\uBBF8\uC9C0 \uCD94\uCD9C \uC2E4\uD328 (${imgPath}): ${err instanceof Error ? err.message : String(err)}`
-        });
-      }
+  for (const blip of findElements(doc.documentElement, "blip")) {
+    const embedId = getAttr(blip, "embed");
+    if (!embedId || imageMap.has(embedId)) continue;
+    const target = rels.get(embedId);
+    if (!target) continue;
+    const imgPath = target.startsWith("/") ? target.slice(1) : target.startsWith("word/") ? target : `word/${target}`;
+    const imgFile = zip.file(imgPath);
+    if (!imgFile) continue;
+    try {
+      const data = await imgFile.async("uint8array");
+      imgIdx++;
+      const ext = imgPath.split(".").pop()?.toLowerCase() ?? "png";
+      const mimeMap = {
+        png: "image/png",
+        jpg: "image/jpeg",
+        jpeg: "image/jpeg",
+        gif: "image/gif",
+        bmp: "image/bmp",
+        wmf: "image/wmf",
+        emf: "image/emf"
+      };
+      const filename = `image_${String(imgIdx).padStart(3, "0")}.${ext}`;
+      images.push({ filename, data, mimeType: mimeMap[ext] ?? "image/png" });
+      imageMap.set(embedId, filename);
+    } catch (err) {
+      warnings.push({
+        code: "SKIPPED_IMAGE",
+        message: `DOCX \uC774\uBBF8\uC9C0 \uCD94\uCD9C \uC2E4\uD328 (${imgPath}): ${err instanceof Error ? err.message : String(err)}`
+      });
     }
   }
-  return { blocks, images };
+  return { imageMap, images };
+}
+function collectParagraphBlips(node, out = [], depth = 0) {
+  if (depth > 40) return out;
+  for (const el of effectiveChildElements(node)) {
+    if (matchesLocal(el, "txbxContent") || matchesLocal(el, "Fallback")) continue;
+    if (matchesLocal(el, "blip")) out.push(el);
+    else collectParagraphBlips(el, out, depth + 1);
+  }
+  return out;
+}
+function emitParagraphImages(p, imageMap, linked, out) {
+  for (const blip of collectParagraphBlips(p)) {
+    const embedId = getAttr(blip, "embed");
+    if (!embedId) continue;
+    const filename = imageMap.get(embedId);
+    if (!filename) continue;
+    linked.add(embedId);
+    out.push({ type: "image", text: filename });
+  }
 }
 async function parseDocxDocument(buffer, options) {
   precheckZipSize(buffer, MAX_DECOMPRESS_SIZE4);
@@ -19547,6 +19640,8 @@ async function parseDocxDocument(buffer, options) {
   if (body.length === 0) {
     throw new KordocError("DOCX \uBCF8\uBB38(w:body)\uC744 \uCC3E\uC744 \uC218 \uC5C6\uC2B5\uB2C8\uB2E4");
   }
+  const { imageMap, images } = await buildImageMap(zip, rels, doc, warnings);
+  const linkedImages = /* @__PURE__ */ new Set();
   const blocks = [];
   const bodyEl = body[0];
   const topLevel = effectiveChildElements(bodyEl);
@@ -19555,16 +19650,20 @@ async function parseDocxDocument(buffer, options) {
     if (localName2 === "p") {
       const block = parseParagraph2(el, styles, numbering, footnotes, rels);
       if (block) blocks.push(block);
+      if (imageMap.size > 0) emitParagraphImages(el, imageMap, linkedImages, blocks);
       for (const tp of collectTextboxParagraphs(el)) {
         const tb = parseParagraph2(tp, styles, numbering, footnotes, rels);
         if (tb) blocks.push(tb);
+        if (imageMap.size > 0) emitParagraphImages(tp, imageMap, linkedImages, blocks);
       }
     } else if (localName2 === "tbl") {
       const block = parseTable(el, styles, numbering, footnotes, rels);
       if (block) blocks.push(block);
     }
   }
-  const { blocks: imgBlocks, images } = await extractImages(zip, rels, doc, warnings);
+  for (const [embedId, filename] of imageMap) {
+    if (!linkedImages.has(embedId)) blocks.push({ type: "image", text: filename });
+  }
   const metadata = {};
   const coreFile = zip.file("docProps/core.xml");
   if (coreFile) {
@@ -29784,7 +29883,7 @@ async function parseHwp(buffer, options) {
 async function parsePdf(buffer, options) {
   let parsePdfDocument;
   try {
-    const mod = await import("./parser-YSHZC5FQ.js");
+    const mod = await import("./parser-UGB3NIVH.js");
     parsePdfDocument = mod.parsePdfDocument;
   } catch {
     return {