classifiers

A document processing system written in Go that extracts data from documents and classifies them based on configurable rules.

How It Works

1. Classification Rules

Rules define document types and the keywords that identify them.

2. Document Processing

Extract → Classify → Organize

• Extract: Pull text from various document formats
• Classify: Apply rules to determine document type
• Organize: Sort documents by classification

3. File Browser Interface

• 📁 Browse directories with visual representation
• 📄 Select files for processing
• 🔍 Filter by supported types

4. Classification Rules Management

• 📋 View and edit rules
• 🔄 Reload from external files
• 🔀 Select different rule sets

Supported Document Types

• 📊 Invoices
• 📜 Contracts
• 🧾 Receipts
• 📝 Reports

Project Structure

• /models - Data structure definitions
• /services - Core business logic
• /services/extractors - Document data extraction components
• /services/classifiers - Document classification logic
• /ui - User interface components

Dependencies

Tesseract OCR

Required for image text extraction:

Windows:

• Download from UB-Mannheim/tesseract
• During installation, select the languages you need (English is 'eng', Portuguese is 'por')

macOS:

brew install tesseract tesseract-lang

• For specific languages: brew install tesseract-lang-eng tesseract-lang-por

Linux (Ubuntu/Debian):

# For English language support
sudo apt update && sudo apt install -y tesseract-ocr tesseract-ocr-eng

# For Portuguese language support
sudo apt install -y tesseract-ocr-por

Language Packs:

• 'eng': English
• 'por': Portuguese
• 'spa': Spanish
• 'fra': French
• 'deu': German

Install the language packs appropriate for your documents' content.

Usage

Run with path: ./classifiers