Skip to content

Python PDF parser for scientific publications: content and figures

License

Notifications You must be signed in to change notification settings

cubense/scipdf_parser

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

59 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SciPDF Parser

A Python parser for scientific PDF based on GROBID.

Installation

Use pip to install from this Github repository

pip install git+https://github.com/titipata/scipdf_parser

Note

  • We also need an en_core_web_sm model for spacy, where you can run python -m spacy download en_core_web_sm to download it
  • You can change GROBID version in serve_grobid.sh to test the parser on a new GROBID version

Usage

Run the GROBID using the given bash script before parsing PDF.

NOTE: the recommended way to run grobid is via docker, so make sure it's running on your machine. Update the script so that you are using latest version. Generally, at every version there are substantial improvements.

bash serve_grobid.sh

This script will run GROBID at default port 8070 (see more here). To parse a PDF provided in example_data folder or direct URL, use the following function:

import scipdf
article_dict = scipdf.parse_pdf_to_dict('example_data/futoma2017improved.pdf') # return dictionary
 
# option to parse directly from URL to PDF, if as_list is set to True, output 'text' of parsed section will be in a list of paragraphs instead
article_dict = scipdf.parse_pdf_to_dict('https://www.biorxiv.org/content/biorxiv/early/2018/11/20/463760.full.pdf', as_list=False)

# output example
>> {
    'title': 'Proceedings of Machine Learning for Healthcare',
    'abstract': '...',
    'sections': [
        {'heading': '...', 'text': '...'},
        {'heading': '...', 'text': '...'},
        ...
    ],
    'references': [
        {'title': '...', 'year': '...', 'journal': '...', 'author': '...'},
        ...
    ],
    'figures': [
        {'figure_label': '...', 'figure_type': '...', 'figure_id': '...', 'figure_caption': '...', 'figure_data': '...'},
        ...
    ],
    'doi': '...'
}

xml = scipdf.parse_pdf('example_data/futoma2017improved.pdf', soup=True) # option to parse full XML from GROBID

To parse figures from PDF using pdffigures2, you can run

scipdf.parse_figures('example_data', output_folder='figures') # folder should contain only PDF files

You can see example output figures in figures folder.

About

Python PDF parser for scientific publications: content and figures

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 97.7%
  • Shell 2.3%