scripts to fetch papers from main s2 api

1 files changed, 44 insertions, 0 deletions

diff --git a/s2-raw-papers.py b/s2-raw-papers.py
new file mode 100644
index 00000000..86ec8710
--- /dev/null
+++ b/s2-raw-papers.py
@@ -0,0 +1,44 @@
+import os
+import sys
+import csv
+import subprocess
+import time
+import random
+import re
+import json
+import click
+from s2 import SemanticScholarAPI
+from util import *
+
+s2 = SemanticScholarAPI()
+
+@click.command()
+@click.option('--fn', '-i', default='missing.csv', help='Filename of CSV.')
+def fetch_raw_papers(paper_ids_fn):
+  lines = read_csv(paper_ids_fn, keys=False)
+  for line in lines:
+    paper_id = line[0]
+    fetch_raw_paper(paper_id)
+
+def fetch_raw_paper(paper_id):
+  os.makedirs(make_raw_paper_fn(paper_id), exist_ok=True)
+  paper_fn = make_raw_paper_fn(paper_id)
+  if os.path.exists(paper_fn):
+    return read_json(paper_fn)
+  print(paper_id)
+  paper = s2.raw_paper(paper_id)
+  if paper is None:
+    print("Got empty paper?? {}".format(paper_id))
+    time.sleep(random.randint(5, 10))
+    return None  
+  write_json(paper_fn, paper)
+  time.sleep(random.randint(5, 10))
+  return paper
+
+def make_raw_paper_path(paper_id):
+  return './datasets/s2/raw_papers/{}/{}'.format(paper_id[0:2], paper_id)
+def make_raw_paper_fn(paper_id):
+  return './datasets/s2/raw_papers/{}/{}/paper.json'.format(paper_id[0:2], paper_id)
+
+if __name__ == '__main__':
+  fetch_raw_papers()