M.I.M.I.R - Multi-agent Intelligent Memory & Insight Repository

#!/usr/bin/env node /** * Extract translatable text units from a DOCX file to JSON. * Uses mammoth for DOCX parsing. */ import mammoth from 'mammoth'; import fs from 'fs/promises'; import path from 'path'; const CODE_PATTERN = /^\s*\d{3}-\d{5}[A-Z]?\s*$|^\s*\d{6}\s*$/; /** * Extract translation units from DOCX * mammoth extracts as HTML, we parse paragraphs from that */ async function extractUnits(docxPath) { const buffer = await fs.readFile(docxPath); // Get raw text with paragraph markers const result = await mammoth.extractRawText({ buffer }); const text = result.value; // Also get HTML for structure detection const htmlResult = await mammoth.convertToHtml({ buffer }); const html = htmlResult.value; const units = []; // Split by paragraphs (double newlines or single newlines) const paragraphs = text.split(/\n+/).filter(p => p.trim()); paragraphs.forEach((para, pi) => { const trimmed = para.trim(); // Skip empty or code-only patterns if (!trimmed || CODE_PATTERN.test(trimmed)) { return; } // Detect style from HTML (basic heuristic) let style = 'body'; if (html.includes(`<h1>${trimmed.substring(0, 20)}`)) style = 'heading1'; else if (html.includes(`<h2>${trimmed.substring(0, 20)}`)) style = 'heading2'; else if (html.includes(`<strong>${trimmed.substring(0, 20)}`)) style = 'bold'; units.push({ id: `body:p${pi}`, source: trimmed, style: style, where: 'body' }); }); return units; } async function main() { const args = process.argv.slice(2); if (args.length < 2) { console.log('Usage: node extract_docx.mjs <input.docx> <output.json>'); process.exit(1); } const [inputPath, outputPath] = args; try { const units = await extractUnits(inputPath); await fs.writeFile(outputPath, JSON.stringify(units, null, 2), 'utf-8'); console.log(`Extracted ${units.length} units -> ${outputPath}`); } catch (error) { console.error(`Error: ${error.message}`); process.exit(1); } } main();