scancab

Documentation

Overview

Scancab is an implementation of a document scanning cabinet. A scanning cabinet is like a file cabinet but stores digital scans of documents instead of paper copies.

Scancab watches the given directory for PDF files to include in the cabinet. Each PDF is assumed to be a separate document: there is no facility for grouping multiple PDFs into a single document or for splitting a large PDF into multiple documents.

When a new PDF is found, it shows up in the incoming list and can be filed by adding metadata. The per-document metadata is:

file name (immutable)
SHA1 hash of content (immutable)
file creation time (immutable)
document time
due date
tags
description
ocr text
physical location

Scancab stores metadata and page thumbnails in two sqlite3 databases named metadata.db and thumbnail.db, in the same directory as the PDFs.

This program was inspired by github.com/bradfitz/scanningcabinet, but the implementation is new. The most significant difference is the use of local disk and a local web server instead of App Engine.