Do not require arxiv id

Author: Marcin Kardas

parse_references.py

@@ -28,7 +28,7 @@ def __init__(self):
     def parse_refs(self, p):
         for d in extract_refs(p):
             if not d["ref_id"].startswith("pwc-"):
-                key = d["paper_arxiv_id"] + d["ref_id"]
+                key = p.arxiv_no_version + d["ref_id"]
                 if key not in self.cache:
                     new_id = self.refsdb.add_reference_string(d['ref_str'])
                     if new_id is not None:

sota_extractor2/data/references.py

@@ -370,7 +370,6 @@ def extract_refs(p):
         if m:
             ref_id, ref_str = m.groups()
             yield {
-                "paper_arxiv_id": p.arxiv_no_version,
                 "ref_id": ref_id,
                 "ref_str": ref_str.strip(r'\s')
             }