Moves to using bibtexparser's writer for bibdata. [Fixes #33]

pubs/endecoder.py

@@ -53,9 +53,10 @@ def customizations(record):
 
     return record
 
+
 bibfield_order = ['author', 'title', 'journal', 'institution', 'publisher',
-                  'year', 'month', 'number', 'volume', 'pages', 'link', 'doi', 'note',
+                  'year', 'month', 'number', 'volume', 'pages', 'link', 'doi',
-                  'abstract']
+                  'note', 'abstract']
 
 
 class EnDecoder(object):
@@ -69,6 +70,9 @@ class EnDecoder(object):
         * encode_bibdata will try to recognize exceptions
     """
 
+    bwriter = bp.bwriter.BibTexWriter()
+    bwriter.display_order = bibfield_order
+
     def encode_metadata(self, metadata):
         return yaml.safe_dump(metadata, allow_unicode=True,
                               encoding=None, indent=4)
@@ -76,41 +80,35 @@ class EnDecoder(object):
     def decode_metadata(self, metadata_raw):
         return yaml.safe_load(metadata_raw)
 
-    def encode_bibdata(self, bibdata):
+    def encode_bibdata(self, bibdata, ignore_fields=[]):
         """Encode bibdata """
-        return '\n'.join(self._encode_bibentry(citekey, entry)
+        bpdata = bp.bibdatabase.BibDatabase()
-                         for citekey, entry in bibdata.items())
+        bpdata.entries = [self._entry_to_bp_entry(k, copy.copy(bibdata[k]),
+                                                  ignore_fields=ignore_fields)
+                          for k in bibdata]
+        return self.bwriter.write(bpdata)
 
-    @staticmethod
+    def _entry_to_bp_entry(self, key, entry, ignore_fields=[]):
-    def _encode_field(key, value):
+        """Convert back entries to the format expected by bibtexparser."""
-        if key == 'link':
+        entry[BP_ID_KEY] = key
-            return ', '.join(link['url'] for link in value)
+        # Convert internal 'type' to bibtexparser entrytype key
-        elif key == 'author':
+        entry[BP_ENTRYTYPE_KEY] = entry.pop(TYPE_KEY)
-            return ' and '.join(author for author in value)
+        for f in ignore_fields:
-        elif key == 'editor':
+            entry.pop(f, None)
-            return ' and '.join(editor['name'] for editor in value)
+        if 'link' in entry:
-        elif key == 'journal':
+            entry['link'] = ', '.join(link['url'] for link in entry['link'])
-            return value['name']
+        if 'author' in entry:
-        elif key == 'keyword':
+            entry['author'] = ' and '.join(
-            return ', '.join(keyword for keyword in value)
+                author for author in entry['author'])
-        else:
+        if 'editor' in entry:
-            return value
+            entry['editor'] = ' and '.join(
-
+                editor['name'] for editor in entry['editor'])
-    @staticmethod
+        if 'journal' in entry:
-    def _encode_bibentry(citekey, bibentry):
+            entry['journal'] = entry['journal']['name']
-        bibraw = '@{}{{{},\n'.format(bibentry[TYPE_KEY], citekey)
+        if 'keyword' in entry:
-        bibentry = copy.copy(bibentry)
+            entry['keyword'] = ', '.join(
-        for key in bibfield_order:
+                keyword for keyword in entry['keyword'])
-            if key in bibentry:
+        return entry
-                value = bibentry.pop(key)
-                bibraw += '    {} = {{{}}},\n'.format(
-                    key, EnDecoder._encode_field(key, value))
-        for key, value in bibentry.items():
-            if key != TYPE_KEY:
-                bibraw += '    {} = {{{}}},\n'.format(
-                    key, EnDecoder._encode_field(key, value))
-        bibraw += '}\n'
-        return bibraw
 
     def decode_bibdata(self, bibdata):
         """"""

tests/test_endecoder.py

@@ -91,13 +91,41 @@ class TestEnDecode(unittest.TestCase):
         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')))
 
+    def test_endecode_bibtex_field_order(self):
+        decoder = endecoder.EnDecoder()
+        entry = decoder.decode_bibdata(bibtex_raw0)
+        lines = decoder.encode_bibdata(entry).splitlines()
+        self.assertEqual(lines[1].split('=')[0].strip(), u'author')
+        self.assertEqual(lines[2].split('=')[0].strip(), u'title')
+        self.assertEqual(lines[3].split('=')[0].strip(), u'institution')
+        self.assertEqual(lines[4].split('=')[0].strip(), u'publisher')
+        self.assertEqual(lines[5].split('=')[0].strip(), u'year')
+        self.assertEqual(lines[6].split('=')[0].strip(), u'month')
+        self.assertEqual(lines[7].split('=')[0].strip(), u'number')
+        self.assertEqual(lines[8].split('=')[0].strip(), u'link')
+        self.assertEqual(lines[9].split('=')[0].strip(), u'note')
+        self.assertEqual(lines[10].split('=')[0].strip(), u'abstract')
+
+    def test_endecode_bibtex_ignores_fields(self):
+        decoder = endecoder.EnDecoder()
+        entry = decoder.decode_bibdata(bibtex_raw0)
+
+        bibraw1 = decoder.encode_bibdata(
+            entry, ignore_fields=['title', 'note', 'abstract', 'journal'])
+        entry1 = list(decoder.decode_bibdata(bibraw1).values())[0]
+
+        self.assertNotIn('title', entry1)
+        self.assertNotIn('note', entry1)
+        self.assertNotIn('abtract', entry1)
+        self.assertIn('author', entry1)
+        self.assertIn('institution', entry1)
+
 
 if __name__ == '__main__':
     unittest.main()