Back to Details

pdf-read

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

PDF Read Skill

PDF读取技能

Extract text and metadata from PDF files.
从PDF文件中提取文本和元数据。

When to Use

适用场景

✅ USE this skill when:
  • User uploads a PDF and asks for summary
  • Extract text from a PDF document
  • Read PDF metadata (author, title, pages)
  • Analyze PDF content
✅ 适用本技能的场景：
  • 用户上传PDF并要求生成摘要
  • 从PDF文档中提取文本
  • 读取PDF元数据（作者、标题、页数）
  • 分析PDF内容

When NOT to Use

不适用场景

❌ DON'T use this skill when:
  • Creating or generating PDFs → use reporting tools
  • Editing existing PDFs → use PDF manipulation tools
  • OCR on scanned images → use OCR/tesseract tools
  • Password-protected PDFs → ask user to unlock first
❌ 请勿使用本技能的场景：
  • 创建或生成PDF → 使用报表工具
  • 编辑现有PDF → 使用PDF处理工具
  • 对扫描图像进行OCR识别 → 使用OCR/tesseract工具
  • 受密码保护的PDF → 先请用户解锁

Installation

安装

bash
cd /job
npm install pdf-parse
bash
cd /job
npm install pdf-parse

Usage

使用方法

javascript
const fs = require('fs');
const pdf = require('pdf-parse');

async function readPDF(filePath) {
  const dataBuffer = fs.readFileSync(filePath);
  const data = await pdf(dataBuffer);
  
  return {
    text: data.text,
    pages: data.numpages,
    info: data.info, // metadata
    version: data.version,
    metadata: data.metadata
  };
}

// Example
const result = await readPDF('/path/to/document.pdf');
console.log(`Pages: ${result.pages}`);
console.log(`Text preview: ${result.text.substring(0, 500)}...`);
javascript
const fs = require('fs');
const pdf = require('pdf-parse');

async function readPDF(filePath) {
  const dataBuffer = fs.readFileSync(filePath);
  const data = await pdf(dataBuffer);
  
  return {
    text: data.text,
    pages: data.numpages,
    info: data.info, // metadata
    version: data.version,
    metadata: data.metadata
  };
}

// Example
const result = await readPDF('/path/to/document.pdf');
console.log(`Pages: ${result.pages}`);
console.log(`Text preview: ${result.text.substring(0, 500)}...`);

Extract Text by Page Range

按页码范围提取文本

javascript
const pdf = require('pdf-parse');
const fs = require('fs');

async function readPDFPages(filePath, startPage, endPage) {
  const dataBuffer = fs.readFileSync(filePath);
  const data = await pdf(dataBuffer, {
    max: endPage,
    version: 'v2.0.550',
    normalizeWhitespace: true,
  });
  
  return data.text;
}
javascript
const pdf = require('pdf-parse');
const fs = require('fs');

async function readPDFPages(filePath, startPage, endPage) {
  const dataBuffer = fs.readFileSync(filePath);
  const data = await pdf(dataBuffer, {
    max: endPage,
    version: 'v2.0.550',
    normalizeWhitespace: true,
  });
  
  return data.text;
}

Get Metadata

获取元数据

javascript
const result = await readPDF('/path/to/document.pdf');
console.log('Author:', result.info?.Author);
console.log('Title:', result.info?.Title);
console.log('Subject:', result.info?.Subject);
console.log('Keywords:', result.info?.Keywords);
console.log('Creator:', result.info?.Creator);
console.log('Producer:', result.info?.Producer);
console.log('Creation Date:', result.info?.CreationDate);
console.log('Mod Date:', result.info?.ModDate);
javascript
const result = await readPDF('/path/to/document.pdf');
console.log('Author:', result.info?.Author);
console.log('Title:', result.info?.Title);
console.log('Subject:', result.info?.Subject);
console.log('Keywords:', result.info?.Keywords);
console.log('Creator:', result.info?.Creator);
console.log('Producer:', result.info?.Producer);
console.log('Creation Date:', result.info?.CreationDate);
console.log('Mod Date:', result.info?.ModDate);

Search Text in PDF

在PDF中搜索文本

javascript
async function searchInPDF(filePath, searchTerm) {
  const result = await readPDF(filePath);
  const text = result.text;
  const lines = text.split('\n');
  
  const matches = [];
  lines.forEach((line, index) => {
    if (line.toLowerCase().includes(searchTerm.toLowerCase())) {
      matches.push({
        line: index + 1,
        content: line.trim()
      });
    }
  });
  
  return {
    total_matches: matches.length,
    matches: matches.slice(0, 20) // limit results
  };
}
javascript
async function searchInPDF(filePath, searchTerm) {
  const result = await readPDF(filePath);
  const text = result.text;
  const lines = text.split('\n');
  
  const matches = [];
  lines.forEach((line, index) => {
    if (line.toLowerCase().includes(searchTerm.toLowerCase())) {
      matches.push({
        line: index + 1,
        content: line.trim()
      });
    }
  });
  
  return {
    total_matches: matches.length,
    matches: matches.slice(0, 20) // limit results
  };
}

Extract All Text as Single String

将所有文本提取为单个字符串

javascript
async function extractFullText(filePath) {
  const result = await readPDF(filePath);
  // Normalize whitespace for cleaner output
  return result.text.replace(/\s+/g, ' ').trim();
}
javascript
async function extractFullText(filePath) {
  const result = await readPDF(filePath);
  // Normalize whitespace for cleaner output
  return result.text.replace(/\s+/g, ' ').trim();
}

Handling Large PDFs

处理大体积PDF

For PDFs with many pages, process in chunks:
javascript
async function readPDFFirstNPages(filePath, maxPages = 10) {
  const dataBuffer = fs.readFileSync(filePath);
  const data = await pdf(dataBuffer, { max: maxPages });
  
  return {
    text: data.text,
    total_pages: data.numpages,
    pages_read: Math.min(maxPages, data.numpages)
  };
}
对于页数较多的PDF，分块处理：
javascript
async function readPDFFirstNPages(filePath, maxPages = 10) {
  const dataBuffer = fs.readFileSync(filePath);
  const data = await pdf(dataBuffer, { max: maxPages });
  
  return {
    text: data.text,
    total_pages: data.numpages,
    pages_read: Math.min(maxPages, data.numpages)
  };
}

Error Handling

错误处理

javascript
async function safeReadPDF(filePath) {
  try {
    const result = await readPDF(filePath);
    return { success: true, ...result };
  } catch (error) {
    if (error.message.includes('password')) {
      return { success: false, error: 'PDF is password-protected' };
    }
    if (error.message.includes('parse')) {
      return { success: false, error: 'Invalid or corrupted PDF' };
    }
    return { success: false, error: error.message };
  }
}
javascript
async function safeReadPDF(filePath) {
  try {
    const result = await readPDF(filePath);
    return { success: true, ...result };
  } catch (error) {
    if (error.message.includes('password')) {
      return { success: false, error: 'PDF is password-protected' };
    }
    if (error.message.includes('parse')) {
      return { success: false, error: 'Invalid or corrupted PDF' };
    }
    return { success: false, error: error.message };
  }
}

Quick Response Template

快速响应模板

"Read this PDF"
javascript
const result = await readPDF(filePath);
return `📄 **PDF Summary**

**Pages:** ${result.pages}
**Title:** ${result.info?.Title || 'N/A'}
**Author:** ${result.info?.Author || 'N/A'}

**Preview (first 500 chars):**
${result.text.substring(0, 500)}...
`;
"读取此PDF"
javascript
const result = await readPDF(filePath);
return `📄 **PDF摘要**

**页数:** ${result.pages}
**标题:** ${result.info?.Title || '无'}
**作者:** ${result.info?.Author || '无'}

**预览（前500字符）:**
${result.text.substring(0, 500)}...
`;

Notes

注意事项

  • pdf-parse works on most standard PDFs
  • Does NOT support OCR for scanned documents
  • Does NOT handle password-protected PDFs
  • For image-heavy PDFs, text extraction may be limited
  • Large PDFs (>100 pages) should be read in chunks
  • pdf-parse适用于大多数标准PDF
  • 不支持对扫描文档进行OCR识别
  • 无法处理受密码保护的PDF
  • 对于图片较多的PDF，文本提取效果可能受限
  • 大体积PDF（>100页）应分块读取