专刊

scripts/alias-hooks.mjs

@@ -0,0 +1,15 @@
+import path from 'path';
+import fs from 'fs';
+import { pathToFileURL } from 'url';
+
+export async function resolve(specifier, context, nextResolve) {
+  if (specifier.startsWith('@/')) {
+    let abs = path.join(process.cwd(), 'src', specifier.slice(2));
+    if (!fs.existsSync(abs) && !path.extname(abs)) {
+      if (fs.existsSync(abs + '.js')) abs += '.js';
+      else if (fs.existsSync(path.join(abs, 'index.js'))) abs = path.join(abs, 'index.js');
+    }
+    return { shortCircuit: true, url: pathToFileURL(abs).href };
+  }
+  return nextResolve(specifier, context);
+}

scripts/alias-loader.mjs

@@ -0,0 +1,19 @@
+import path from 'path';
+import fs from 'fs';
+import { pathToFileURL } from 'url';
+
+export async function resolve(specifier, context, nextResolve) {
+  if (specifier.startsWith('@/')) {
+    let rel = specifier.slice(2);
+    let abs = path.join(process.cwd(), 'src', rel);
+    if (!fs.existsSync(abs) && !path.extname(abs)) {
+      if (fs.existsSync(abs + '.js')) abs = abs + '.js';
+      else if (fs.existsSync(path.join(abs, 'index.js'))) abs = path.join(abs, 'index.js');
+    }
+    return {
+      shortCircuit: true,
+      url: pathToFileURL(abs).href
+    };
+  }
+  return nextResolve(specifier, context);
+}

scripts/alias-register.mjs

@@ -0,0 +1,5 @@
+import { register } from 'node:module';
+import { pathToFileURL } from 'url';
+import path from 'path';
+
+register(pathToFileURL(path.join(process.cwd(), 'scripts/alias-hooks.mjs')).href);

scripts/inspect-word-fig-pairing.mjs

@@ -0,0 +1,45 @@
+import FS from 'fs';
+import path from 'path';
+import JSZip from 'jszip';
+
+const file = path.join(
+  'C:/Users/Administrator/Desktop/IN2026A0418001_final_20260713 1.docx'
+);
+const xml = await (await JSZip.loadAsync(FS.readFileSync(file))).file('word/document.xml').async('string');
+const paras = xml.split(/<\/w:p>/);
+
+function plainOf(p) {
+  return [...p.matchAll(/<w:t(?:\s[^>]*)?>([\s\S]*?)<\/w:t>/g)]
+    .map((x) => x[1])
+    .join('')
+    .replace(/&amp;/g, '&')
+    .replace(/\s+/g, ' ')
+    .trim();
+}
+
+function blipCount(p) {
+  return (p.match(/a:blip/g) || []).length;
+}
+
+function drawingCount(p) {
+  return (p.match(/<w:drawing/g) || []).length;
+}
+
+for (let i = 0; i < paras.length; i++) {
+  const plain = plainOf(paras[i]);
+  if (!/^Figure\s+\d+/i.test(plain)) continue;
+  // look back up to 8 paras for images / A B labels
+  const back = [];
+  for (let j = Math.max(0, i - 8); j < i; j++) {
+    const t = plainOf(paras[j]);
+    const b = blipCount(paras[j]);
+    const d = drawingCount(paras[j]);
+    if (b || d || t) {
+      back.push({ j, blips: b, drawings: d, text: t.slice(0, 80) });
+    }
+  }
+  const next = plainOf(paras[i + 1] || '').slice(0, 180);
+  console.log('\nCAPTION', plain.slice(0, 100));
+  console.log('BACK', JSON.stringify(back, null, 0));
+  console.log('NOTE?', next);
+}

scripts/inspect-word-figures.mjs

@@ -0,0 +1,82 @@
+import FS from 'fs';
+import path from 'path';
+import JSZip from 'jszip';
+
+const file = path.join(
+  'C:/Users/Administrator/Desktop/IN2026A0418001_final_20260713 1.docx'
+);
+
+const buf = FS.readFileSync(file);
+const zip = await JSZip.loadAsync(buf);
+const xml = await zip.file('word/document.xml').async('string');
+
+const drawings = (xml.match(/<w:drawing/g) || []).length;
+const blips = (xml.match(/a:blip/g) || []).length;
+console.log('drawings', drawings, 'blips', blips);
+
+const colors = {};
+for (const m of xml.matchAll(/w:color\s+w:val="([^"]+)"/g)) {
+  colors[m[1]] = (colors[m[1]] || 0) + 1;
+}
+console.log(
+  'colors top',
+  Object.entries(colors)
+    .sort((a, b) => b[1] - a[1])
+    .slice(0, 20)
+);
+
+// Extract paragraphs containing "Figure" with nearby text
+const paras = xml.split(/<\/w:p>/);
+let figParaCount = 0;
+for (let i = 0; i < paras.length; i++) {
+  const p = paras[i];
+  const texts = [...p.matchAll(/<w:t(?:\s[^>]*)?>([\s\S]*?)<\/w:t>/g)].map((x) => x[1]);
+  const plain = texts.join('').replace(/&amp;/g, '&').trim();
+  if (!/^Figure\s+\d+/i.test(plain) && !/^Fig\.?\s*\d+/i.test(plain)) continue;
+  figParaCount += 1;
+  const hasDrawing = /<w:drawing/.test(p);
+  const colorVals = [...p.matchAll(/w:color\s+w:val="([^"]+)"/g)].map((x) => x[1]);
+  // next non-empty para
+  let nextPlain = '';
+  for (let j = i + 1; j < Math.min(i + 5, paras.length); j++) {
+    const nt = [...paras[j].matchAll(/<w:t(?:\s[^>]*)?>([\s\S]*?)<\/w:t>/g)]
+      .map((x) => x[1])
+      .join('')
+      .replace(/&amp;/g, '&')
+      .trim();
+    if (nt) {
+      nextPlain = nt.slice(0, 220);
+      break;
+    }
+  }
+  // prev: look back for drawing-only paragraphs
+  let prevHasDrawing = false;
+  for (let j = i - 1; j >= Math.max(0, i - 6); j--) {
+    if (/<w:drawing/.test(paras[j])) {
+      prevHasDrawing = true;
+      break;
+    }
+    const pt = [...paras[j].matchAll(/<w:t(?:\s[^>]*)?>([\s\S]*?)<\/w:t>/g)]
+      .map((x) => x[1])
+      .join('')
+      .trim();
+    if (pt && !/^[AB]$/.test(pt)) break;
+  }
+  console.log('---');
+  console.log('caption:', plain.slice(0, 160));
+  console.log('colors:', [...new Set(colorVals)]);
+  console.log('sameParaDrawing:', hasDrawing, 'prevHasDrawing:', prevHasDrawing);
+  console.log('next:', nextPlain);
+}
+console.log('figParaCount', figParaCount);
+
+// Check standalone A/B label paragraphs near figures
+let abCount = 0;
+for (const p of paras) {
+  const plain = [...p.matchAll(/<w:t(?:\s[^>]*)?>([\s\S]*?)<\/w:t>/g)]
+    .map((x) => x[1])
+    .join('')
+    .trim();
+  if (/^[AB]$/.test(plain)) abCount += 1;
+}
+console.log('standalone A/B paras', abCount);

scripts/test-extract-figures.mjs

@@ -0,0 +1,29 @@
+import FS from 'fs';
+import path from 'path';
+import { JSDOM } from 'jsdom';
+
+const { window } = new JSDOM('<!DOCTYPE html><html><body></body></html>');
+global.window = window;
+global.document = window.document;
+global.DOMParser = window.DOMParser;
+global.Node = window.Node;
+global.HTMLElement = window.HTMLElement;
+global.FileReader = window.FileReader;
+global.Blob = window.Blob;
+
+const { parseWordFigures } = await import('../src/utils/wordFigureImport.js');
+
+const filePath = path.join(
+  'C:/Users/Administrator/Desktop/IN2026A0418001_final_20260713 1.docx'
+);
+const buf = FS.readFileSync(filePath);
+const result = await parseWordFigures(buf);
+console.log('parsedCount', result.parsedCount);
+result.figures.forEach((f, idx) => {
+  console.log('----', idx + 1);
+  console.log('num', f.figureNumber);
+  console.log('title', f.title.slice(0, 130));
+  console.log('note', (f.note || '').slice(0, 150));
+  console.log('imgs', f.imageCount);
+  console.log('red', /data-word-caption-color="red"|#d25a5a/i.test(f.titleHtml || ''));
+});

scripts/test-fig-match.mjs

@@ -0,0 +1,35 @@
+import {
+  extractFigureNumbersFromText,
+  paragraphMentionsFigureNumber,
+  parseFigureNumberFromTitle,
+  findFirstAmIdForFigureNumber
+} from '../src/utils/manuscriptMediaReferences.js';
+
+const samples = [
+  'Figure 3A and Figure 3B.',
+  'as shown in Figure 3A and Figure 3B.',
+  'See Fig. 3 and Fig. 3A',
+  'as shown in <blue>Figure 3A</blue> and <blue>Figure 3B</blue>.'
+];
+for (const s of samples) {
+  console.log(JSON.stringify(s), '=>', extractFigureNumbersFromText(s), 'mentions3', paragraphMentionsFigureNumber(s, 3));
+}
+console.log(
+  'title',
+  parseFigureNumberFromTitle('<span style="color:#d25a5a">Figure 3 Co-occurrence map</span>')
+);
+console.log(
+  'find',
+  findFirstAmIdForFigureNumber(
+    [
+      { am_id: 1, type: 0, content: 'hello' },
+      {
+        am_id: 146120,
+        type: 0,
+        content: 'as shown in <blue>Figure 3A</blue> and <blue>Figure 3B</blue>.'
+      },
+      { am_id: 3, type: 1, content: '', ami_id: 99 }
+    ],
+    3
+  )
+);