diff options
Diffstat (limited to 'scraper/s2-citation-report.py')
| -rw-r--r-- | scraper/s2-citation-report.py | 7 |
1 files changed, 6 insertions, 1 deletions
diff --git a/scraper/s2-citation-report.py b/scraper/s2-citation-report.py index e4e41478..c515ec37 100644 --- a/scraper/s2-citation-report.py +++ b/scraper/s2-citation-report.py @@ -184,6 +184,10 @@ def process_paper(row, addresses, success): address_count = 0 fn = file_path('papers', row['paper_id'], 'paper.json') + if not os.path.exists(fn): + print("not found: {}".format(fn)) + print(row) + return with open(fn, 'r') as f: data = json.load(f) @@ -364,7 +368,8 @@ def load_megapixels_queries(): rec = {} for index, key in enumerate(keys): rec[key] = row[index] - recs.append(rec) + if rec['verified'] == '1': + recs.append(rec) return recs #def load_institutions(paperId): |