Fix encoding of keywords.

Also introduces ustr instead of unicode for python2 and str for python3.

pubs/bibstruct.py

@@ -1,6 +1,8 @@
 import unicodedata
 import re
 
+from .p3 import ustr
+
     # citekey stuff
 
 CONTROL_CHARS = ''.join(map(unichr, range(0, 32) + range(127, 160)))
@@ -10,14 +12,14 @@ CITEKEY_EXCLUDE_RE = re.compile('[%s]'
         % re.escape(CONTROL_CHARS + CITEKEY_FORBIDDEN_CHARS))
 
 def str2citekey(s):
-    key = unicodedata.normalize('NFKD', unicode(s)).encode('ascii', 'ignore')
+    key = unicodedata.normalize('NFKD', ustr(s)).encode('ascii', 'ignore')
     key = CITEKEY_EXCLUDE_RE.sub('', key)
     # Normalize chars and remove non-ascii
     return key
 
 def check_citekey(citekey):
     # TODO This is not the right way to test that (17/12/2012)
-    if unicode(citekey) != str2citekey(citekey):
+    if ustr(citekey) != str2citekey(citekey):
         raise ValueError("Invalid citekey: %s" % citekey)
 
 def verify_bibdata(bibdata):

pubs/endecoder.py

@@ -20,6 +20,7 @@ def sanitize_citekey(record):
     record['id'] = record['id'].strip('\n')
     return record
 
+
 def customizations(record):
     """ Use some functions delivered by the library
 
@@ -75,6 +76,8 @@ class EnDecoder(object):
             return ' and '.join(editor['name'] for editor in value)
         elif key == 'journal':
             return value['name']
+        elif key == 'keyword':
+            return ', '.join(keyword for keyword in value)
         else:
             return value
 

pubs/p3.py

@@ -4,10 +4,12 @@ if sys.version_info[0] == 2:
     import ConfigParser as configparser
     import StringIO as io
     input = raw_input
+    ustr = unicode
 else:
     import configparser
     import io
+    ustr = str
 
 configparser = configparser
 io = io
-input = input
+input = input

pubs/uis.py

@@ -4,6 +4,7 @@ import sys
 
 from .beets_ui import _encoding, input_
 from .content import editor_input
+from .p3 import ustr
 from . import color
 
 # package-shared ui that can be accessed using :
@@ -39,7 +40,7 @@ class UI:
         replaces it.
         """
         txt = [s.encode(self.encoding, 'replace')
-               if isinstance(s, unicode) else s
+               if isinstance(s, ustr) else s
                for s in strings]
         print(' '.join(txt))
 

tests/test_endecoder.py

@@ -51,15 +51,39 @@ class TestEnDecode(unittest.TestCase):
             bibentry2 = entry2[citekey]
             for key, value in bibentry1.items():
                 self.assertEqual(bibentry1[key], bibentry2[key])
-
         self.assertEqual(bibraw1, bibraw2)
 
-    def test_endecode_metadata(self):
+    def test_endecode_keyword(self):
+        decoder = endecoder.EnDecoder()
+        entry = decoder.decode_bibdata(turing_bib)
+        keywords = ['artificial intelligence', 'Turing test']
+        entry['turing1950computing']['keyword'] = keywords
+        bibraw = decoder.encode_bibdata(entry)
+        entry1 = decoder.decode_bibdata(bibraw)
+        self.assertIn('keyword', entry1['turing1950computing'])
+        self.assertEqual(set(keywords),
+                         set(entry1['turing1950computing']['keyword']))
+
+    def test_endecode_keyword_as_keywords(self):
+        decoder = endecoder.EnDecoder()
+        keywords = [u'artificial intelligence', u'Turing test']
+        # Add keywords to bibraw
+        keyword_str = 'keywords = {artificial intelligence, Turing test},\n'
+        biblines = turing_bib.splitlines()
+        biblines.insert(-3, keyword_str)
+        bibsrc = '\n'.join(biblines)
+        print bibsrc
+        entry = decoder.decode_bibdata(bibsrc)['turing1950computing']
+        print entry
+        self.assertNotIn(u'keywords', entry)
+        self.assertIn(u'keyword', entry)
+        self.assertEqual(set(keywords), set(entry[u'keyword']))
+
 
+    def test_endecode_metadata(self):
         decoder = endecoder.EnDecoder()
         entry = decoder.decode_metadata(metadata_raw0)
         metadata_output0 = decoder.encode_metadata(entry)
-
         self.assertEqual(set(metadata_raw0.split('\n')), set(metadata_output0.split('\n')))