PDF2TXT

A powerful Deno library and CLI tool for extracting text and detailed layout information from PDF files using Mozilla PDF.js. Features advanced layout reconstruction that preserves column alignment and table structures.

Features

• 📄 Simple Text Extraction: Extract plain text from PDF files
• 🎯 Layout Preservation: Maintain visual layout with proper spacing and column alignment
• 📊 Table Structure: Accurately reconstruct table layouts with column headers and data rows
• 🔧 Customizable: Configurable options for fine-tuning layout reconstruction
• 🚀 High Performance: Built on Mozilla PDF.js for reliable PDF parsing
• 📦 Dual Usage: Available as both a library and CLI tool
• 🦕 Deno Native: Built specifically for the Deno runtime

Installation

As a Library

import { extractText, extractLayoutData } from "https://deno.land/x/pdf2txt/src/mod.ts";

As a CLI Tool

# Install globally
deno install --allow-read --allow-write -n pdf2txt https://deno.land/x/pdf2txt/src/main.ts

# Or run directly
deno run --allow-read --allow-write https://deno.land/x/pdf2txt/src/main.ts --input document.pdf

Library Usage

Simple Text Extraction

import { extractText } from "./src/mod.ts";

// Extract plain text
const text = await extractText("document.pdf");
console.log(text);

// Extract from buffer
const buffer = await Deno.readFile("document.pdf");
const text2 = await extractText(buffer);

Advanced Layout Extraction

import { extractLayoutData, extractTextFromPdf } from "./src/mod.ts";

// Extract with layout information
const layoutData = await extractLayoutData("document.pdf");
console.log("Metadata:", layoutData.metadata);
console.log("Reconstructed layout:", layoutData.layoutText);

// Custom options for better table handling
const result = await extractTextFromPdf("document.pdf", {
  includeLayout: true,
  yTolerance: 1.5,              // Tighter row grouping
  characterWidthDivisor: 3.5,   // Different column spacing
  enableDebug: true             // Show debug output
});

Working with Layout Data

import { extractLayoutData } from "./src/mod.ts";

const data = await extractLayoutData("table-document.pdf");

// Access metadata
console.log(`Title: ${data.metadata.title}`);
console.log(`Pages: ${data.metadata.pages}`);
console.log(`Author: ${data.metadata.author}`);

// Get reconstructed layout (preserves table structure)
console.log("Layout with preserved columns:");
console.log(data.layoutText);

// Access raw positioning data
data.pageLayouts.forEach(page => {
  console.log(`Page ${page.pageNumber} has ${page.textItems.length} text items`);
  
  // Find table-like content
  const tableItems = page.textItems.filter(item => 
    item.str.match(/^[0-9]+$/) || item.str.includes("$")
  );
  
  tableItems.forEach(item => {
    console.log(`"${item.str}" at position (${item.x}, ${item.y})`);
  });
});

CLI Usage

Basic Commands

# Extract text to stdout
pdf2txt --input document.pdf

# Save to file
pdf2txt --input document.pdf --output extracted.txt

# Extract with layout information
pdf2txt --input document.pdf --layout --output layout.txt

# Output as JSON
pdf2txt --input document.pdf --layout --json --output data.json

Advanced Options

# Custom layout options for better table alignment
pdf2txt --input document.pdf --layout \
  --y-tolerance 1.5 \
  --character-width-divisor 3.5 \
  --debug

# Help and version
pdf2txt --help
pdf2txt --version

CLI Options

Option	Description
-i, --input <FILE>	Input PDF file path
-o, --output <FILE>	Output file path (optional)
-l, --layout	Include layout and metadata information
-j, --json	Output in JSON format
-d, --debug	Enable debug output for layout analysis
--y-tolerance <NUMBER>	Y-coordinate tolerance for row grouping (default: 2)
--character-width-divisor <NUMBER>	Character width divisor for column positioning (default: 4.0)
-h, --help	Show help message
-v, --version	Show version information

API Reference

Types

interface PdfExtractionOptions {
  includeLayout?: boolean;           // Include detailed layout information
  yTolerance?: number;              // Y-coordinate tolerance for grouping (default: 2)
  characterWidthDivisor?: number;   // Character width divisor for positioning (default: 4.0)
  enableDebug?: boolean;            // Enable debug output to console
}

interface PdfMetadata {
  title: string | null;
  author: string | null;
  subject: string | null;
  creator: string | null;
  producer: string | null;
  creationDate: string | null;
  modificationDate: string | null;
  pages: number;
}

interface TextItem {
  str: string;        // Text content
  x: number;          // X coordinate
  y: number;          // Y coordinate
  width: number;      // Item width
  height: number;     // Item height
  fontName: string;   // Font name
  // ... additional properties
}

interface PdfLayoutData {
  text: string;                    // Raw extracted text
  layoutText: string;              // Reconstructed layout text
  metadata: PdfMetadata;           // PDF metadata
  pageLayouts: PageLayoutData[];   // Detailed page layouts
  // ... additional properties
}

Functions

// Main extraction function with full options
extractTextFromPdf(filePath: string | Uint8Array, options?: PdfExtractionOptions): Promise<string | PdfLayoutData>

// Convenience function for simple text extraction
extractText(filePath: string | Uint8Array): Promise<string>

// Convenience function for layout extraction
extractLayoutData(filePath: string | Uint8Array, options?: Omit<PdfExtractionOptions, 'includeLayout'>): Promise<PdfLayoutData>

Examples

See examples.ts for comprehensive usage examples including:

• Basic text extraction
• Layout extraction with default settings
• Custom options for table handling
• Buffer-based extraction
• Debug output for troubleshooting
• Error handling patterns

Run the examples:

deno run --allow-read examples.ts

Development

Setup

# Clone the repository
git clone <repository-url>
cd pdf2txt

# Check TypeScript
deno check src/**/*.ts

# Format code
deno fmt

# Lint code
deno lint

# Run tests
deno test --allow-read --allow-write

Build

# Build CLI executable
deno task build

# The executable will be created at ./bin/pdf2txt

Project Structure

pdf2txt/
├── src/
│   ├── lib.ts      # Core library functions
│   ├── main.ts     # CLI interface
│   └── mod.ts      # Library exports
├── tests/          # Test files
├── examples.ts     # Usage examples
├── deno.json       # Deno configuration
└── README.md       # This file

Customization for Different Documents

// For documents with tightly packed tables
const result = await extractTextFromPdf("document.pdf", {
  includeLayout: true,
  yTolerance: 1.0,              // Very tight row grouping
  characterWidthDivisor: 2.5    // More precise column spacing
});

// For documents with loose spacing
const result = await extractTextFromPdf("document.pdf", {
  includeLayout: true,
  yTolerance: 3.0,              // Looser row grouping
  characterWidthDivisor: 5.0    // Wider column spacing
});

License

AGPL License - see LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit issues, feature requests, or pull requests.

Changelog

v1.0.0

• Initial release
• Core text extraction functionality
• Advanced layout reconstruction
• Table structure preservation
• CLI and library interfaces
• Comprehensive documentation and examples