Installation¶

Dependencies¶

By default invoice2data extracts text with pypdfium2 — a self-contained wheel that bundles its own PDF engine, so the common path needs no system libraries. pip install invoice2data is enough to get started on Windows, macOS and Linux.

Some optional backends need extra system tools:

pdftotext (xpdf/poppler-utils) — better at preserving table layout (-layout); recommended for layout-sensitive templates. Included with macOS Homebrew, Debian and Ubuntu.
OCR — tesseract + ImageMagick, or ocrmypdf + Ghostscript, for scanned / image-only PDFs (see below).

Installation using pip¶

invoice2data is delivered by PyPI because it is a convenient way to install the latest version. However, PyPI and pip cannot address the fact that invoice2data depends on certain non-Python system libraries and programs being installed.

Install invoice2data using pip

pip install invoice2data

Important

Invoice2data uses a yaml templating system. The yaml templates are loaded with pyyaml which is a pure python implementation. (thus rather slow) As an alternative json templates can be used. Which are natively better supported by python.

Tip

The performance with yaml templates can be greatly increased 10x 🚀 by using libyaml It can be installed on most distributions by: sudo apt-get libyaml-dev

Installation using conda¶

Conda (or mamba) is the easiest route when you want the system tools handled for you: poppler, tesseract and ghostscript are all on conda-forge, so there is no separate system-library install. An environment.yml is provided in the repository:

conda env create -f environment.yml
conda activate invoice2data

This creates an environment with invoice2data, the common PDF backends and the OCR tools ready to use. invoice2data itself is installed from PyPI inside the environment (there is no dedicated conda-forge package yet); its runtime dependencies are already provided by the conda packages, so nothing extra is pulled in.

Installation of input modules¶

Most backends are optional and installed via extras. Pick what you need:

pip install "invoice2data[pdfplumber]"      # pdfplumber backend
pip install "invoice2data[doctr]"           # local deep-learning OCR (docTR)
pip install "invoice2data[paddleocr]"       # local deep-learning OCR (PaddleOCR)
pip install "invoice2data[ai]"              # LLM fallback / template generation

backend	extra	use
`pdfium`	(built-in)	default text extraction, no system deps
`text`	(built-in)	already-extracted `.txt` input
`pdftotext`	system `poppler`	layout-preserving text
`pdfplumber` / `pdfminer`	`pdfplumber` / `pdfminer-six`	pure-Python extraction
`tesseract` / `ocrmypdf`	system tools	OCR for scanned PDFs
`doctr` / `paddleocr`	`doctr` / `paddleocr`	local deep-learning OCR (privacy-friendly)
`gvision`	`googlevision`	Google Cloud Vision OCR

See How It Works for the default backend cascade, and AI features for the AI features.

docTR / PaddleOCR¶

Local deep-learning OCR backends — good for scans/photos with little pre-processing, and keep data on your machine. The model weights download on first run. Select with --input-reader doctr or --input-reader paddleocr.

tesseract¶

An tesseract wrapper is included in auto language mode. It will test your input files against the languages installed on your system. To use it, tesseract and imagemagick needs to be installed. tesseract supports multiple OCR engine modes. By default, the available engine installed on the system will be used.

Languages:

tesseract-ocr recognize more than 100 languages For Linux users, you can often find packages that provide language packs:

# Display a list of all Tesseract language packs
apt-cache search tesseract-ocr

# Debian/Ubuntu users
apt-get install tesseract-ocr-chi-sim  # Example: Install Chinese Simplified language pack

# Arch Linux users
pacman -S tesseract-data-eng tesseract-data-deu # Example: Install the English and German language packs

ocrmypdf¶

Refer to ocrmypdf documentation