DocumentLens

Text Analysis & Academic Intelligence Microservice

Transform text content into actionable insights through comprehensive linguistic analysis, writing quality assessment, and academic integrity checking.

🚀 Quick Start

# Docker deployment (recommended)
docker-compose up -d

# Or raw deployment
./deploy.sh

# API available at: http://localhost:8002
# Documentation: http://localhost:8002/docs

📊 API Endpoints

Core Analysis

• GET /health - Service health check
• POST /text - Text analysis (readability, quality, word frequency)
• POST /academic - Academic analysis (citations, DOI resolution, integrity)
• POST /files - File upload + analysis (PDF, DOCX, TXT, MD)

Advanced Text Analysis

• POST /advanced/ngrams - N-gram extraction with optional filter terms
• POST /advanced/ner - Named entity recognition
• POST /advanced/search/keywords - Batch keyword search across multiple terms

Document Intelligence

• POST /files/infer-metadata - Infer year, company, industry, document type from content
• POST /text/infer-metadata - Metadata inference from raw text
• Page-level text extraction (via include_extracted_text=true on /files)

Integration

• Root endpoint: GET / - Service info and available endpoints
• For presentations: Use PresentationLens
• For recordings: Use RecordingLens

🎯 Use Cases

• Text Analysis: Readability, writing quality, word frequency for any text content
• Academic Analysis: Citation verification, DOI resolution, AI detection, integrity checking
• Document Intelligence: Extract and analyze text from PDFs and Word documents
• Sustainability Research: Batch keyword analysis for TCFD, GRI, SDGs, SASB frameworks
• Corporate Report Analysis: Auto-detect metadata (year, company, industry) from annual reports
• Multi-Service Workflows: Integrate with specialized analysis services

Desktop Application Support

DocumentLens powers the document-lens-desktop Electron application for researchers analyzing corporate sustainability reports. Features include:

• Smart metadata inference (company name, year, industry, document type)
• Framework keyword analysis (TCFD, GRI, SDGs, SASB)
• Batch processing with SQLite storage
• Offline operation via bundled Python backend

🏗️ Microservices Ecosystem

DocumentLens is part of a focused microservices architecture:

Integration Pattern

graph LR
    A[Student Submission] --> B[SubmissionLens Frontend]
    B --> C{File Type Router}
    C -->|Text/PDF/DOCX| D[DocumentLens]
    C -->|PPTX| E[PresentationLens]
    C -->|Video/Audio| F[RecordingLens]
    C -->|Source Code| G[CodeLens]
    E --> D
    F --> D
    G --> D
    D --> H[Combined Feedback]
    H --> B
    B --> I[Student Dashboard]

🚀 Deployment

Docker Deployment (Recommended)

git clone https://github.com/michael-borck/document-lens.git
cd document-lens
docker-compose up -d  # Single container deployment

Raw/Native Deployment

git clone https://github.com/michael-borck/document-lens.git
cd document-lens
./deploy.sh  # Handles venv, dependencies, and production server

🧪 Testing

# Install dev dependencies
uv sync --extra dev

# Run all tests
uv run pytest tests/ -v

# Run specific test file
uv run pytest tests/test_files.py -v

# Run only PDF tests
uv run pytest tests/ -m pdf -v

# Skip slow tests
uv run pytest tests/ -m "not slow" -v

# Run with coverage report
uv run pytest tests/

Test Structure

• tests/conftest.py - Shared fixtures and test client setup
• tests/test_health.py - Health/smoke tests
• tests/test_text_analysis.py - Text analysis endpoint tests
• tests/test_academic_analysis.py - Academic analysis endpoint tests
• tests/test_files.py - PDF file upload tests

Test Data

Place test files (PDF, DOCX, etc.) in the test-data/ directory. The test suite automatically discovers and uses these files for parameterized tests.

📚 Documentation

• DEPLOYMENT.md - Deployment guide for Docker and raw installations
• DOCUMENTLENS_SETUP.md - Setup and usage instructions
• .env.example - Configuration template
• docs/ - Additional architecture and integration documentation

DocumentLens: Pure text intelligence at the heart of content analysis