Added TLDExtract cache files so that internet connection is not required

Page conversion now page by page calls hopefully to avoid fastapi timeouts on AWS. gunicorn keep_alive parameter extended to 60 seconds just in case that helps too.

correctly spelled --no-cache-dir this time

Unspecifying gradio and spacy in requirements, then reinstalling latest gradio afterwards in Dockerfile. All to try to avoid typer conflict

Created output folder specifically in Dockerfile

Specify GRADIO_SERVER_NAME variable in Dockerfile as 0.0.0.0

Modified Dockerfile to run with user 1000. Changed port to standard 7860 and removed server name specification.

Added opencv installation to dockerfile and reverted to slim-bookworm

Changed base python distribution to (hopefully) have access to tesseract-ocr package

Added -y to tesseract-ocr installation in Dockerfile

Added -y to poppler-utils installation in Dockerfile. Added support for image files in image-based redaction.

Initial commit