pdf-extractor-engine

pdf-extractor-engine is a standalone, subject-neutral PDF source-validation and page-text extraction core extracted from the electional application repository.

Scope

• Validates PDF sources from filesystem paths or in-memory bytes.
• Produces immutable extraction contracts for page text, warnings, errors, diagnostics, and fingerprints.
• Serializes contracts in trusted, public-safe, and fingerprint profiles.
• Performs no mandatory persistence writes during validation or extraction.

Non-scope

• No electional document registration or source-document storage.
• No desktop UI, API wrapper, or project configuration loading.
• No CLI entry point or GUI launcher.
• No automatic persistence adapter, OCR, or downstream curation logic.

Packaging boundary

• The published import root is pdf_extractor_engine.
• The package is intended to install and import without the sibling electional repository.
• Technical packaging evidence does not by itself authorize publishing.
• Licensing and full supported-Python policy remain operator decisions.

Package layout

• src/pdf_extractor_engine/: standalone extractor package.
• tests/: standalone tests plus a parity runner used during Phase 21A extraction.
• docs/: dependency and integration-boundary notes plus the extraction report.

Install and run locally

$env:PYTHONPATH = (Resolve-Path .\src)
..\need-to-connectt-my-github-to\.venv\Scripts\python.exe -m unittest discover -s tests -t .

Public API root

The standalone import root is:

import pdf_extractor_engine

The embedded electional import root remains unchanged and is not redirected by this extraction.