This repository contains various Python scripts used in the creation of datasets from Robert Singerman's Judaica Americana: A Bibliography of Publications to 1900. These datasets were used as the foundation of the Digital Second Edition of Judaica Americana.
These scripts include:
- extract_singerman.py: for extracting the data from the JA draft and writing into a csv
- flip-index-headers.py: for creating a csv of Singerman IDs and corresponding index headers
- extract_singerman_serials.py: for extracting the data from the JA draft and writing into a csv re: serials
- tess.py: forked from tess, written by Jonathan Scott Enderle. tess was used to OCR from the index from the JA print publication
An extremely basic python script for converting PDFs to TIFFs and performing OCR with tesseract.
To run the script, first ensure that ImageMagick 7 and Tesseract 4 are installed and can be run from the command line. (Later versions may work but this has only been tested with the above versions.)
You'll also need to ensure that the correct language models are installed. The tesseract wiki has installation instructions for various operating systems.
Once you have the software installed, you can run the script:
tess.py [--language LANGUAGE] files [files ...]
---------
files: One or more PDF files to process
--language: The tesseract language ID code for the language model
to use. E.g. eng (English), deu (German) or
ita (Italian). The default is eng.
An Italian-language sample file is provided in the testdata folder. To
process it, run the below command:
tess.py --language ita testdata/1961_Alessandria.pdf
Judaica Digital Humanities at the Penn Libraries (also referred to as Judaica DH) is a robust program of projects and tools for experimental digital scholarship with Judaica collections, informed by digital humanities, Jewish studies, and cultural heritage approaches. Visit our website.