s2-dump-db-pdf-urls.py

1 files changed, 37 insertions, 0 deletions

diff --git a/s2-dump-db-pdf-urls.py b/s2-dump-db-pdf-urls.py
new file mode 100644
index 00000000..520b513e
--- /dev/null
+++ b/s2-dump-db-pdf-urls.py
@@ -0,0 +1,37 @@
+import os
+import glob
+import simplejson as json
+import click
+from util import *
+
+PAPER_JSON_DIR = 'datasets/s2/db_papers'
+
+@click.command()
+def s2_dump_pdf_urls():
+  # loop over all the papers in db_papers
+  # get all the PDF urls, pick the best one
+  # store it and the paper id
+  # another script will fetch the urls from this process
+  rows = [process_paper(fn) for fn in glob.iglob('{}/**/paper.json'.format(PAPER_JSON_DIR), recursive=True)]
+  print("Wrote {} rows".format(len(rows)))
+  write_csv('db_paper_pdf_list.csv', keys=['Paper ID', 'PDF URL', 'IEEE URL', 'Extra URL'], rows=rows)
+
+def process_paper(fn, lookups):
+  paper = read_json(fn)
+  paper_id = paper['id']
+  pdf_url = None
+  ieee_url = None
+  extra_url = None
+  if paper['s2PdfUrl']:
+    pdf_url = paper['s2PdfUrl']
+  for url in paper['pdfUrls']:
+    if 'ieeexplore.ieee.org' in url:
+      ieee_url = url
+    elif pdf_url is None and 'pdf' in url:
+      pdf_url = url
+    else:
+      extra_url = url
+  return [paper_id, pdf_url, ieee_url, extra_url]
+
+if __name__ == '__main__':
+  s2_dump_pdf_urls()