From 2116027843edad22d87e6a56269b26cd6aafb8e8 Mon Sep 17 00:00:00 2001
From: "jules@lens" <julescarbon@gmail.com>
Date: Wed, 20 Feb 2019 18:36:25 +0100
Subject: updating all reports

---
 scraper/s2-citation-report.py | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

(limited to 'scraper/s2-citation-report.py')

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):
-- 
cgit v1.2.3-70-g09d2