From acc16d8f35a3b10021ff75db06503851feb8efde Mon Sep 17 00:00:00 2001 From: Jules Laplace Date: Tue, 6 Nov 2018 01:30:46 +0100 Subject: reports --- pdf_report_first_pages.py | 39 --------------------------------------- 1 file changed, 39 deletions(-) delete mode 100644 pdf_report_first_pages.py (limited to 'pdf_report_first_pages.py') diff --git a/pdf_report_first_pages.py b/pdf_report_first_pages.py deleted file mode 100644 index ae080539..00000000 --- a/pdf_report_first_pages.py +++ /dev/null @@ -1,39 +0,0 @@ -import os -import gzip -import glob -import json -import click -from util import * - -PDF_DIR = 'datasets/s2/pdf' -FIRST_PAGES_KEYS = '' - -@click.command() -def pdf_report_first_pages(): - rows = [] - for fn in glob.iglob('{}/**/*.txt'.format(PDF_DIR), recursive=True): - row = process_paper(fn) - rows.append(row) - write_report('reports/first_pages.html', title='First pages', keys=FIRST_PAGES_KEYS, rows=rows) - print("Wrote {} rows".format(len(rows))) - -def process_paper(fn): - index = fn.replace(PDF_DIR, '').split('/')[2] - with open(fn, 'r') as f: - lines = '' - for line in f.readlines(): - if 'abstract' in line.lower(): - break - if len(line) < 3: - continue - lines += line + '
' - return [ - index, - lines - ] - -def paper_path(paper_id): - return '{}/{}/{}'.format(DATA_DIR, paper_id[0:2], paper_id) - -if __name__ == '__main__': - pdf_report_first_pages() -- cgit v1.2.3-70-g09d2