jenkins-bot merged this change.

View Change

[IMPR] Fix Wikibase aliases handling

Individual fixes:
- user can now remove all aliases of a single language be either removing
  the language key or making the list empty (T194512)
- compare lists by sorting them instead of using Counter to handle
  duplicate aliases
- improve WikibasePage._normalizeData, so that user can provide both
  strings and structured dicts
- call WikibasePage._normalizeLanguages on aliases in
  WikibasePage._normalizeData

Bug: T194512
Change-Id: I578b08d91cabbf69f071dfddf4afe19329df1e0c
---
M pywikibot/page.py
M tests/wikibase_tests.py
2 files changed, 50 insertions(+), 19 deletions(-)

diff --git a/pywikibot/page.py b/pywikibot/page.py
index dd4f49a..4de9d02 100644
--- a/pywikibot/page.py
+++ b/pywikibot/page.py
@@ -3983,19 +3983,23 @@
         self._diff_to('descriptions', 'language', 'value', diffto, data)
 
         aliases = self._normalizeLanguages(self.aliases).copy()
-        if diffto and 'aliases' in diffto:
-            for lang in set(diffto['aliases'].keys()) - set(aliases.keys()):
-                aliases[lang] = []
-        for lang, strings in list(aliases.items()):
-            if diffto and 'aliases' in diffto and lang in diffto['aliases']:
-                empty = len(diffto['aliases'][lang]) - len(strings)
-                if empty > 0:
-                    strings += [''] * empty
-                elif Counter(val['value'] for val
-                             in diffto['aliases'][lang]) == Counter(strings):
-                    del aliases[lang]
-            if lang in aliases:
-                aliases[lang] = [{'language': lang, 'value': i} for i in strings]
+        if diffto:
+            for lang, strings in diffto.get('aliases', {}).items():
+                if len(aliases.get(lang, [])) > 0:
+                    if tuple(sorted(val['value'] for val in strings)) != tuple(
+                            sorted(aliases[lang])):
+                        aliases[lang] = [{'language': lang, 'value': i}
+                                         for i in aliases[lang]]
+                    else:
+                        del aliases[lang]
+                else:
+                    aliases[lang] = [
+                        {'language': lang, 'value': i['value'], 'remove': ''}
+                        for i in strings]
+        else:
+            for lang, values in aliases.items():
+                aliases[lang] = [{'language': lang, 'value': i}
+                                 for i in values]
 
         if aliases:
             data['aliases'] = aliases
@@ -4117,11 +4121,16 @@
                     data[prop][key] = {'language': key, 'value': value}
 
         if 'aliases' in data:
+            data['aliases'] = cls._normalizeLanguages(data['aliases'])
             for key, values in data['aliases'].items():
-                if (isinstance(values, list) and
-                        isinstance(values[0], basestring)):
-                    data['aliases'][key] = [{'language': key, 'value': value}
-                                            for value in values]
+                if isinstance(values, list):
+                    strings = []
+                    for value in values:
+                        if isinstance(value, basestring):
+                            strings.append({'language': key, 'value': value})
+                        else:
+                            strings.append(value)
+                    data['aliases'][key] = strings
 
         return data
 
diff --git a/tests/wikibase_tests.py b/tests/wikibase_tests.py
index d02b01e..04a4332 100644
--- a/tests/wikibase_tests.py
+++ b/tests/wikibase_tests.py
@@ -1626,14 +1626,20 @@
         """Setup tests."""
         super(TestWriteNormalizeData, self).setUp()
         self.data_out = {
-            'aliases': {'en': [{'language': 'en', 'value': 'Bah'}]},
+            'aliases': {'en': [
+                {'language': 'en', 'value': 'Bah'},
+                {'language': 'en', 'value': 'Bar', 'remove': ''},
+            ]},
             'labels': {'en': {'language': 'en', 'value': 'Foo'}},
         }
 
     def test_normalize_data(self):
         """Test _normalizeData() method."""
         data_in = {
-            'aliases': {'en': ['Bah']},
+            'aliases': {'en': [
+                'Bah',
+                {'language': 'en', 'value': 'Bar', 'remove': ''},
+            ]},
             'labels': {'en': 'Foo'},
         }
 
@@ -1957,6 +1963,10 @@
     def test_json_diff(self):
         """Test json diff."""
         del self.wdp.labels['en']
+        self.wdp.aliases['de'].append('New York')
+        self.wdp.aliases['de'].append('foo')
+        self.wdp.aliases['de'].remove('NYC')
+        del self.wdp.aliases['nl']
         del self.wdp.claims['P213']
         expected = {
             'labels': {
@@ -1965,6 +1975,18 @@
                     'value': ''
                 }
             },
+            'aliases': {
+                'de': [
+                    {'language': 'de', 'value': 'City of New York'},
+                    {'language': 'de', 'value': 'The Big Apple'},
+                    {'language': 'de', 'value': 'New York'},
+                    {'language': 'de', 'value': 'New York'},
+                    {'language': 'de', 'value': 'foo'},
+                ],
+                'nl': [
+                    {'language': 'nl', 'value': 'New York', 'remove': ''},
+                ],
+            },
             'claims': {
                 'P213': [
                     {

To view, visit change 460742. To unsubscribe, or for help writing mail filters, visit settings.