global: fix DOI unicode issues

* Handles unicode characters in DOI. * URL-encodes generated DOI URLs
inveniosoftware · Nov 22, 2017 · 84ab6fa · 84ab6fa
1 parent a62bf1a
commit 84ab6fa
Show file tree

Hide file tree

Showing 2 changed files with 20 additions and 10 deletions.
diff --git a/idutils/__init__.py b/idutils/__init__.py
@@ -18,7 +18,8 @@
 import re
 
 from isbn import ISBN
-from six.moves.urllib.parse import urlparse
+from six.moves.urllib.parse import urlparse, quote_plus
+
 
 from .version import __version__
 
@@ -501,15 +502,15 @@ def normalize_pid(val, scheme):
 
 
 LANDING_URLS = {
-    'doi': 'https://doi.org/{pid}',
-    'handle': '{scheme}://hdl.handle.net/{pid}',
-    'arxiv': '{scheme}://arxiv.org/abs/{pid}',
-    'orcid': '{scheme}://orcid.org/{pid}',
-    'pmid': '{scheme}://www.ncbi.nlm.nih.gov/pubmed/{pid}',
-    'ads': 'http://adsabs.harvard.edu/abs/{pid}',
-    'pmcid': '{scheme}://www.ncbi.nlm.nih.gov/pmc/{pid}',
-    'gnd': 'http://d-nb.info/gnd/{pid}',
-    'urn': '{scheme}://nbn-resolving.org/{pid}',
+    'doi': u'https://doi.org/{pid}',
+    'handle': u'{scheme}://hdl.handle.net/{pid}',
+    'arxiv': u'{scheme}://arxiv.org/abs/{pid}',
+    'orcid': u'{scheme}://orcid.org/{pid}',
+    'pmid': u'{scheme}://www.ncbi.nlm.nih.gov/pubmed/{pid}',
+    'ads': u'http://adsabs.harvard.edu/abs/{pid}',
+    'pmcid': u'{scheme}://www.ncbi.nlm.nih.gov/pmc/{pid}',
+    'gnd': u'http://d-nb.info/gnd/{pid}',
+    'urn': u'{scheme}://nbn-resolving.org/{pid}',
 }
 """URL generation configuration for the supported PID providers."""
 

diff --git a/tests/test_idutils.py b/tests/test_idutils.py
@@ -51,6 +51,15 @@
     ('doi.org/10.1016/j.epsl.2011.11.037', ['doi', 'handle'],
         '10.1016/j.epsl.2011.11.037',
         'https://doi.org/10.1016/j.epsl.2011.11.037'),
+    (u'10.1016/üникóδé-дôΐ', ['doi', 'handle'],
+        u'10.1016/üникóδé-дôΐ', u'https://doi.org/10.1016/üникóδé-дôΐ'),
+    (u'10.1016/སྦ་བཞེད་', ['doi', 'handle'],
+        u'10.1016/སྦ་བཞེད་', u'https://doi.org/10.1016/སྦ་བཞེད་'),
+    ('10.1002/(SICI)1521-3978(199806)46:4/5<493::AID-PROP493>3.0.CO;2-P',
+        ['doi', 'handle'],
+        '10.1002/(SICI)1521-3978(199806)46:4/5<493::AID-PROP493>3.0.CO;2-P',
+        ('https://doi.org/'
+         '10.1002/(SICI)1521-3978(199806)46:4/5<493::AID-PROP493>3.0.CO;2-P')),
     ('9783468111242', ['isbn', 'ean13'], '978-3-468-11124-2', ''),
     ('4006381333931', ['ean13'], '', ''),
     ('73513537', ['ean8'], '', ''),