Handling anchors by ignoring them

HPC-workshops · Jul 8, 2016 · 5520175 · 5520175
1 parent 9d6512c
commit 5520175
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/bin/extract_figures.py b/bin/extract_figures.py
@@ -69,7 +69,7 @@ def find_image_links(doc, result):
 
     if ((doc['type'] == 'a') and ('attr' in doc) and ('href' in doc['attr'])) \
        or \
-       ((doc['type'] == 'html_element') and (doc['value'] == 'a')):
+       ((doc['type'] == 'html_element') and (doc['value'] == 'a') and ('href' in doc['attr'])):
         path = doc['attr']['href']
         if os.path.splitext(path)[1].lower() in IMAGE_FILE_SUFFIX:
             result.append({'alt':'', 'src': doc['attr']['href']})