Fix absolute reference links with anchor

yqq2022 · Nov 1, 2016 · 528c37e · 528c37e
1 parent e4c28ff
commit 528c37e
Showing 1 changed file with 2 additions and 3 deletions.
diff --git a/chemdataextractor/reader/markup.py b/chemdataextractor/reader/markup.py
@@ -164,9 +164,8 @@ def _parse_table_footnotes(self, fns, refs, specials):
 
     def _parse_reference(self, el):
         """Return reference ID from href or text content."""
-        if el.get('href', '').startswith('#'):
-            # TODO: Get anchor from href with absolute url, e.g. http://pubs.acs.org/doi/full/10.1021/acs.jmedchem.6b01195#cor1
-            return [el.get('href')[1:]]
+        if '#' in el.get('href', ''):
+            return [el.get('href').split('#', 1)[1]]
         elif 'rid' in el.attrib:
             return [el.attrib['rid']]
         elif 'idref' in el.attrib: