Skip to content

Installation & Setup

This guide details how to set up PDF Anonymizer for development or programmatic usage.

Development Installation (Monorepo)

The repository uses uv, a fast Python package installer and resolver. Ensure you have uv installed. If not, follow the official Astral installation instructions.

Steps

  1. Clone the repository: 

    git clone https://github.com/leo-gan/anonymizer.git
cd anonymizer

  2. Sync the workspace: Run uv sync to create a virtual environment and install all development tools alongside workspace members: 

    uv sync --group dev
    This sync command will install the CLI, core library, and the dependencies for all supported LLM providers (google, ollama, huggingface, openrouter, openai, anthropic).

  3. Activate the virtual environment: 

    source .venv/bin/activate

Installing as a Library Dependency

If you are using pdf-anonymizer-core or pdf-anonymizer-cli in a separate external project, you can install them via pip or uv.

To keep the installation footprint small, the core package uses PEP 508 Extras for specific LLM providers. Install only the providers you plan to use:

# Base package only (no provider dependencies)
pip install pdf-anonymizer-core

# Install specific provider support
pip install "pdf-anonymizer-core[google]"
pip install "pdf-anonymizer-core[ollama]"
pip install "pdf-anonymizer-core[openai,anthropic]"

Available Extras

  • [google]: Installs the Google GenAI SDK (google-genai).
  • [ollama]: Installs the Ollama integration library.
  • [huggingface]: Installs libraries to connect to Hugging Face Hub models.
  • [openrouter]: Installs client tools to query OpenRouter.
  • [openai]: Installs the official OpenAI API SDK.
  • [anthropic]: Installs the official Anthropic API SDK.

Environment Variables Configuration

To run the anonymizer with cloud providers, you must supply API credentials. The CLI automatically loads a .env file from the directory where the command is executed or any parent directory.

Create a .env file at the root of the repository:

# Google GenAI (Gemini Models)
GOOGLE_API_KEY="AIzaSy..."

# OpenAI
OPENAI_API_KEY="sk-proj-..."

# Anthropic (Claude Models)
ANTHROPIC_API_KEY="sk-ant-..."

# Hugging Face Hub
HUGGING_FACE_TOKEN="hf_..."

# OpenRouter
OPENROUTER_API_KEY="sk-or-v1-..."

# Ollama Local Host (Optional, defaults to localhost:11434)
OLLAMA_HOST="http://localhost:11434"

Verifying the Setup

To verify that the installation succeeded and all package internal references are valid, run the automated test suite:

uv run pytest

If the tests pass, you are ready to use the CLI tool!

See Also