# HG changeset patch # User ymh # Date 1326075583 -3600 # Node ID 4b73a767a6c0bc44a9565808ed341d5438df7052 # Parent 9170357f7e90a2b90050b2d8d46bc0e8a5fef291 backport changes made on model for hdabo_sf diff -r 9170357f7e90 -r 4b73a767a6c0 .settings/org.eclipse.core.resources.prefs --- a/.settings/org.eclipse.core.resources.prefs Fri Jul 15 15:23:35 2011 +0200 +++ b/.settings/org.eclipse.core.resources.prefs Mon Jan 09 03:19:43 2012 +0100 @@ -1,4 +1,4 @@ -#Fri Jul 15 14:15:04 CEST 2011 +#Mon Jan 09 03:01:43 CET 2012 eclipse.preferences.version=1 encoding//virtualenv/web/env/hdabo/lib/python2.6/site-packages/haystack/backends/__init__.py=utf-8 encoding//virtualenv/web/env/hdabo/lib/python2.6/site-packages/sortedm2m/fields.py=utf-8 @@ -11,6 +11,7 @@ encoding//web/hdabo/management/commands/import_tag_popularity.py=utf-8 encoding//web/hdabo/management/commands/query_wikipedia.py=utf-8 encoding//web/hdabo/migrations/0001_initial.py=utf-8 +encoding//web/hdabo/migrations/0002_backport_hdabo_sf.py=utf-8 encoding//web/hdabo/models.py=utf-8 encoding//web/hdabo/search/french_whoosh_backend.py=utf-8 encoding//web/hdabo/tests/models.py=utf-8 diff -r 9170357f7e90 -r 4b73a767a6c0 .settings/org.eclipse.php.core.prefs --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/.settings/org.eclipse.php.core.prefs Mon Jan 09 03:19:43 2012 +0100 @@ -0,0 +1,3 @@ +#Sat Sep 24 15:42:33 CEST 2011 +eclipse.preferences.version=1 +include_path= diff -r 9170357f7e90 -r 4b73a767a6c0 web/hdabo/management/commands/query_wikipedia.py --- a/web/hdabo/management/commands/query_wikipedia.py Fri Jul 15 15:23:35 2011 +0200 +++ b/web/hdabo/management/commands/query_wikipedia.py Mon Jan 09 03:19:43 2012 +0100 @@ -97,8 +97,8 @@ spaces = math.floor(width - marks) loader = u'[' + (u'=' * int(marks)) + (u' ' * int(spaces)) + u']' - - sys.stdout.write(u"%s %d%% %d/%d - %s\r" % (loader, percent, current_line - 1, total_line - 1, repr(label))) #takes the header into account + + sys.stdout.write(u"%s %d%% %d/%d - %r\r" % (loader, percent, current_line - 1, total_line - 1, label[:50].rjust(50))) #takes the header into account if percent >= 100: sys.stdout.write("\n") sys.stdout.flush() diff -r 9170357f7e90 -r 4b73a767a6c0 web/hdabo/migrations/0002_backport_hdabo_sf.py --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/web/hdabo/migrations/0002_backport_hdabo_sf.py Mon Jan 09 03:19:43 2012 +0100 @@ -0,0 +1,231 @@ +# encoding: utf-8 +import datetime +from south.db import db +from south.v2 import SchemaMigration + +class Migration(SchemaMigration): + + def forwards(self, orm): + + # Adding field 'Tag.alternative_label' + db.add_column('hdabo_tag', 'alternative_label', self.gf('django.db.models.fields.CharField')(max_length=1024, null=True, blank=True), keep_default=False) #@UndefinedVariable + + # Adding field 'Tag.created_at' + db.add_column('hdabo_tag', 'created_at', self.gf('django.db.models.fields.DateTimeField')(auto_now_add=True, default=datetime.datetime.utcnow, blank=True), keep_default=False) #@UndefinedVariable + + # Adding field 'Tag.alternative_wikipedia_url' + db.add_column('hdabo_tag', 'alternative_wikipedia_url', self.gf('django.db.models.fields.URLField')(max_length=2048, null=True, blank=True), keep_default=False) #@UndefinedVariable + + # Adding field 'Tag.alternative_wikipedia_pageid' + db.add_column('hdabo_tag', 'alternative_wikipedia_pageid', self.gf('django.db.models.fields.BigIntegerField')(null=True, blank=True), keep_default=False) #@UndefinedVariable + + # Adding field 'TaggedSheet.created_at' + db.add_column('hdabo_taggedsheet', 'created_at', self.gf('django.db.models.fields.DateTimeField')(auto_now_add=True, default=datetime.datetime.utcnow, blank=True), keep_default=False) #@UndefinedVariable + + + def backwards(self, orm): + + # Deleting field 'Tag.alternative_label' + db.delete_column('hdabo_tag', 'alternative_label') #@UndefinedVariable + + # Deleting field 'Tag.created_at' + db.delete_column('hdabo_tag', 'created_at') #@UndefinedVariable + + # Deleting field 'Tag.alternative_wikipedia_url' + db.delete_column('hdabo_tag', 'alternative_wikipedia_url') #@UndefinedVariable + + # Deleting field 'Tag.alternative_wikipedia_pageid' + db.delete_column('hdabo_tag', 'alternative_wikipedia_pageid') #@UndefinedVariable + + # Deleting field 'TaggedSheet.created_at' + db.delete_column('hdabo_taggedsheet', 'created_at') #@UndefinedVariable + + + models = { + 'auth.group': { + 'Meta': {'object_name': 'Group'}, + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'name': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '80'}), + 'permissions': ('django.db.models.fields.related.ManyToManyField', [], {'to': "orm['auth.Permission']", 'symmetrical': 'False', 'blank': 'True'}) + }, + 'auth.permission': { + 'Meta': {'ordering': "('content_type__app_label', 'content_type__model', 'codename')", 'unique_together': "(('content_type', 'codename'),)", 'object_name': 'Permission'}, + 'codename': ('django.db.models.fields.CharField', [], {'max_length': '100'}), + 'content_type': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['contenttypes.ContentType']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'name': ('django.db.models.fields.CharField', [], {'max_length': '50'}) + }, + 'auth.user': { + 'Meta': {'object_name': 'User'}, + 'date_joined': ('django.db.models.fields.DateTimeField', [], {'default': 'datetime.datetime.now'}), + 'email': ('django.db.models.fields.EmailField', [], {'max_length': '75', 'blank': 'True'}), + 'first_name': ('django.db.models.fields.CharField', [], {'max_length': '30', 'blank': 'True'}), + 'groups': ('django.db.models.fields.related.ManyToManyField', [], {'to': "orm['auth.Group']", 'symmetrical': 'False', 'blank': 'True'}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'is_active': ('django.db.models.fields.BooleanField', [], {'default': 'True'}), + 'is_staff': ('django.db.models.fields.BooleanField', [], {'default': 'False'}), + 'is_superuser': ('django.db.models.fields.BooleanField', [], {'default': 'False'}), + 'last_login': ('django.db.models.fields.DateTimeField', [], {'default': 'datetime.datetime.now'}), + 'last_name': ('django.db.models.fields.CharField', [], {'max_length': '30', 'blank': 'True'}), + 'password': ('django.db.models.fields.CharField', [], {'max_length': '128'}), + 'user_permissions': ('django.db.models.fields.related.ManyToManyField', [], {'to': "orm['auth.Permission']", 'symmetrical': 'False', 'blank': 'True'}), + 'username': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '30'}) + }, + 'contenttypes.contenttype': { + 'Meta': {'ordering': "('name',)", 'unique_together': "(('app_label', 'model'),)", 'object_name': 'ContentType', 'db_table': "'django_content_type'"}, + 'app_label': ('django.db.models.fields.CharField', [], {'max_length': '100'}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'model': ('django.db.models.fields.CharField', [], {'max_length': '100'}), + 'name': ('django.db.models.fields.CharField', [], {'max_length': '100'}) + }, + 'hdabo.author': { + 'Meta': {'object_name': 'Author'}, + 'firstname': ('django.db.models.fields.CharField', [], {'max_length': '512', 'null': 'True', 'blank': 'True'}), + 'hda_id': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '512'}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'lastname': ('django.db.models.fields.CharField', [], {'max_length': '512', 'null': 'True', 'blank': 'True'}) + }, + 'hdabo.datasheet': { + 'Meta': {'object_name': 'Datasheet'}, + 'author': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Author']", 'null': 'True', 'blank': 'True'}), + 'college_periods': ('django.db.models.fields.related.ManyToManyField', [], {'related_name': "'college_periods_datasheets'", 'symmetrical': 'False', 'through': "orm['hdabo.Datasheet_college_periods']", 'to': "orm['hdabo.TimePeriod']"}), + 'college_themes': ('django.db.models.fields.related.ManyToManyField', [], {'related_name': "'college_themes_datasheets'", 'symmetrical': 'False', 'through': "orm['hdabo.Datasheet_college_themes']", 'to': "orm['hdabo.Domain']"}), + 'description': ('django.db.models.fields.TextField', [], {'null': 'True', 'blank': 'True'}), + 'domains': ('django.db.models.fields.related.ManyToManyField', [], {'related_name': "'datasheets'", 'symmetrical': 'False', 'through': "orm['hdabo.Datasheet_domains']", 'to': "orm['hdabo.Domain']"}), + 'format': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.DocumentFormat']", 'null': 'True', 'blank': 'True'}), + 'hda_id': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '512'}), + 'highschool_periods': ('django.db.models.fields.related.ManyToManyField', [], {'related_name': "'highschool_periods_datasheets'", 'symmetrical': 'False', 'through': "orm['hdabo.Datasheet_highschool_periods']", 'to': "orm['hdabo.TimePeriod']"}), + 'highschool_themes': ('django.db.models.fields.related.ManyToManyField', [], {'related_name': "'highschool_themes_datasheets'", 'symmetrical': 'False', 'through': "orm['hdabo.Datasheet_highschool_themes']", 'to': "orm['hdabo.Domain']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'manual_order': ('django.db.models.fields.BooleanField', [], {'default': 'False', 'db_index': 'True'}), + 'modification_datetime': ('django.db.models.fields.DateTimeField', [], {'auto_now': 'True', 'blank': 'True'}), + 'organisation': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Organisation']"}), + 'original_creation_date': ('django.db.models.fields.DateField', [], {}), + 'original_modification_date': ('django.db.models.fields.DateField', [], {}), + 'primary_periods': ('django.db.models.fields.related.ManyToManyField', [], {'related_name': "'primary_periods_datasheets'", 'symmetrical': 'False', 'through': "orm['hdabo.Datasheet_primary_periods']", 'to': "orm['hdabo.TimePeriod']"}), + 'primary_themes': ('django.db.models.fields.related.ManyToManyField', [], {'related_name': "'primary_themes_datasheets'", 'symmetrical': 'False', 'through': "orm['hdabo.Datasheet_primary_themes']", 'to': "orm['hdabo.Domain']"}), + 'tags': ('django.db.models.fields.related.ManyToManyField', [], {'to': "orm['hdabo.Tag']", 'through': "orm['hdabo.TaggedSheet']", 'symmetrical': 'False'}), + 'title': ('django.db.models.fields.CharField', [], {'max_length': '2048'}), + 'town': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Location']", 'null': 'True', 'blank': 'True'}), + 'url': ('django.db.models.fields.URLField', [], {'max_length': '2048', 'null': 'True', 'blank': 'True'}), + 'validated': ('django.db.models.fields.BooleanField', [], {'default': 'False', 'db_index': 'True'}), + 'validation_date': ('django.db.models.fields.DateTimeField', [], {'null': 'True', 'blank': 'True'}), + 'validator': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['auth.User']", 'null': 'True', 'blank': 'True'}) + }, + 'hdabo.datasheet_college_periods': { + 'Meta': {'ordering': "['sort_value']", 'object_name': 'Datasheet_college_periods'}, + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'sort_value': ('django.db.models.fields.IntegerField', [], {}), + 'timeperiod': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.TimePeriod']"}) + }, + 'hdabo.datasheet_college_themes': { + 'Meta': {'ordering': "['sort_value']", 'object_name': 'Datasheet_college_themes'}, + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'domain': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Domain']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'sort_value': ('django.db.models.fields.IntegerField', [], {}) + }, + 'hdabo.datasheet_domains': { + 'Meta': {'ordering': "['sort_value']", 'object_name': 'Datasheet_domains'}, + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'domain': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Domain']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'sort_value': ('django.db.models.fields.IntegerField', [], {}) + }, + 'hdabo.datasheet_highschool_periods': { + 'Meta': {'ordering': "['sort_value']", 'object_name': 'Datasheet_highschool_periods'}, + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'sort_value': ('django.db.models.fields.IntegerField', [], {}), + 'timeperiod': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.TimePeriod']"}) + }, + 'hdabo.datasheet_highschool_themes': { + 'Meta': {'ordering': "['sort_value']", 'object_name': 'Datasheet_highschool_themes'}, + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'domain': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Domain']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'sort_value': ('django.db.models.fields.IntegerField', [], {}) + }, + 'hdabo.datasheet_primary_periods': { + 'Meta': {'ordering': "['sort_value']", 'object_name': 'Datasheet_primary_periods'}, + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'sort_value': ('django.db.models.fields.IntegerField', [], {}), + 'timeperiod': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.TimePeriod']"}) + }, + 'hdabo.datasheet_primary_themes': { + 'Meta': {'ordering': "['sort_value']", 'object_name': 'Datasheet_primary_themes'}, + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'domain': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Domain']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'sort_value': ('django.db.models.fields.IntegerField', [], {}) + }, + 'hdabo.documentformat': { + 'Meta': {'object_name': 'DocumentFormat'}, + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'label': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '512'}) + }, + 'hdabo.domain': { + 'Meta': {'unique_together': "(('label', 'school_period'),)", 'object_name': 'Domain'}, + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'label': ('django.db.models.fields.CharField', [], {'max_length': '512'}), + 'school_period': ('django.db.models.fields.IntegerField', [], {}) + }, + 'hdabo.location': { + 'Meta': {'object_name': 'Location'}, + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'insee': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '5'}), + 'name': ('django.db.models.fields.CharField', [], {'max_length': '512'}) + }, + 'hdabo.organisation': { + 'Meta': {'object_name': 'Organisation'}, + 'hda_id': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '512'}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'location': ('django.db.models.fields.CharField', [], {'max_length': '512', 'null': 'True', 'blank': 'True'}), + 'name': ('django.db.models.fields.CharField', [], {'max_length': '512'}), + 'website': ('django.db.models.fields.CharField', [], {'max_length': '2048', 'null': 'True', 'blank': 'True'}) + }, + 'hdabo.tag': { + 'Meta': {'unique_together': "(('label', 'original_label', 'url_status'),)", 'object_name': 'Tag'}, + 'alias': ('django.db.models.fields.CharField', [], {'max_length': '1024', 'null': 'True', 'blank': 'True'}), + 'alternative_label': ('django.db.models.fields.CharField', [], {'max_length': '1024', 'null': 'True', 'blank': 'True'}), + 'alternative_wikipedia_pageid': ('django.db.models.fields.BigIntegerField', [], {'null': 'True', 'blank': 'True'}), + 'alternative_wikipedia_url': ('django.db.models.fields.URLField', [], {'max_length': '2048', 'null': 'True', 'blank': 'True'}), + 'category': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.TagCategory']", 'null': 'True', 'blank': 'True'}), + 'created_at': ('django.db.models.fields.DateTimeField', [], {'auto_now_add': 'True', 'blank': 'True'}), + 'dbpedia_uri': ('django.db.models.fields.URLField', [], {'max_length': '2048', 'null': 'True', 'blank': 'True'}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'label': ('django.db.models.fields.CharField', [], {'max_length': '1024', 'db_index': 'True'}), + 'normalized_label': ('django.db.models.fields.CharField', [], {'max_length': '1024', 'db_index': 'True'}), + 'original_label': ('django.db.models.fields.CharField', [], {'max_length': '1024'}), + 'popularity': ('django.db.models.fields.IntegerField', [], {'default': '0', 'db_index': 'True'}), + 'url_status': ('django.db.models.fields.IntegerField', [], {'default': 'None', 'null': 'True', 'blank': 'True'}), + 'wikipedia_pageid': ('django.db.models.fields.BigIntegerField', [], {'null': 'True', 'blank': 'True'}), + 'wikipedia_url': ('django.db.models.fields.URLField', [], {'max_length': '2048', 'null': 'True', 'blank': 'True'}) + }, + 'hdabo.tagcategory': { + 'Meta': {'object_name': 'TagCategory'}, + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'label': ('django.db.models.fields.CharField', [], {'unique': 'True', 'max_length': '512'}) + }, + 'hdabo.taggedsheet': { + 'Meta': {'object_name': 'TaggedSheet'}, + 'created_at': ('django.db.models.fields.DateTimeField', [], {'auto_now_add': 'True', 'blank': 'True'}), + 'datasheet': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Datasheet']"}), + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'index_note': ('django.db.models.fields.FloatField', [], {'default': '0.0', 'db_index': 'True'}), + 'order': ('django.db.models.fields.IntegerField', [], {'default': '0', 'db_index': 'True'}), + 'original_order': ('django.db.models.fields.IntegerField', [], {'default': '0'}), + 'tag': ('django.db.models.fields.related.ForeignKey', [], {'to': "orm['hdabo.Tag']"}), + 'wikipedia_revision_id': ('django.db.models.fields.BigIntegerField', [], {'null': 'True', 'blank': 'True'}) + }, + 'hdabo.timeperiod': { + 'Meta': {'unique_together': "(('label', 'school_period'),)", 'object_name': 'TimePeriod'}, + 'id': ('django.db.models.fields.AutoField', [], {'primary_key': 'True'}), + 'label': ('django.db.models.fields.CharField', [], {'max_length': '512'}), + 'school_period': ('django.db.models.fields.IntegerField', [], {}) + } + } + + complete_apps = ['hdabo'] diff -r 9170357f7e90 -r 4b73a767a6c0 web/hdabo/models.py --- a/web/hdabo/models.py Fri Jul 15 15:23:35 2011 +0200 +++ b/web/hdabo/models.py Mon Jan 09 03:19:43 2012 +0100 @@ -96,6 +96,7 @@ (1, "redirection"), (2, "homonyme"), (3, "match"), + (4, "unsematized"), ) TAG_URL_STATUS_DICT = { @@ -103,15 +104,20 @@ "redirection":1, "homonyme":2, "match":3, + "unsemantized":4, } label = models.CharField(max_length=1024, unique=False, blank=False, null=False, db_index=True) + alternative_label = models.CharField(max_length=1024, unique=False, blank=True, null=True) normalized_label = models.CharField(max_length=1024, unique=False, blank=False, null=False, db_index=True, editable=False) + created_at = models.DateTimeField(auto_now_add=True) original_label = models.CharField(max_length=1024, unique=False, blank=False, null=False, editable=False) alias = models.CharField(max_length=1024, unique=False, blank=True, null=True) category = models.ForeignKey(TagCategory, null=True, blank=True) wikipedia_url = models.URLField(verify_exists=False, max_length=2048, blank=True, null=True) wikipedia_pageid = models.BigIntegerField(unique=False, blank=True, null=True) + alternative_wikipedia_url = models.URLField(verify_exists=False, max_length=2048, blank=True, null=True) + alternative_wikipedia_pageid = models.BigIntegerField(unique=False, blank=True, null=True) url_status = models.IntegerField(choices=TAG_URL_STATUS_CHOICES, blank=True, null=True, default=None) dbpedia_uri = models.URLField(verify_exists=False, max_length=2048, blank=True, null=True) popularity = models.IntegerField(blank=False, null=False, default=0, db_index=True) @@ -347,6 +353,7 @@ class TaggedSheet(models.Model): datasheet = models.ForeignKey(Datasheet) tag = models.ForeignKey(Tag) + created_at = models.DateTimeField(auto_now_add=True) original_order = models.IntegerField(null=False, blank=False, default=0) order = models.IntegerField(null=False, blank=False, default=0, db_index=True) index_note = models.FloatField(null=False, blank=False, default=0.0, db_index=True) diff -r 9170357f7e90 -r 4b73a767a6c0 web/hdabo/wp_utils.py --- a/web/hdabo/wp_utils.py Fri Jul 15 15:23:35 2011 +0200 +++ b/web/hdabo/wp_utils.py Mon Jan 09 03:19:43 2012 +0100 @@ -56,6 +56,9 @@ url = page.get(u'fullurl', None) pageid = page.get(u'pageid', None) new_label = page[u'title'] + alternative_label = None + alternative_url = None + alternative_pageid = None if __is_homonymie(page): status = Tag.TAG_URL_STATUS_DICT["homonyme"] @@ -74,6 +77,9 @@ if len(pages) > 1 or len(pages) == 0: return return_null_result() page = pages.values()[0] + alternative_label = page.get('title', None) + alternative_url = page.get('fullurl', None) + alternative_pageid = page.get('pageid',None) revision_id = page.get('lastrevid', None) @@ -94,7 +100,7 @@ dbpedia_uri = None - return { 'new_label': new_label, 'status': status, 'wikipedia_url': url, 'pageid': pageid, 'dbpedia_uri': dbpedia_uri, 'revision_id': revision_id, 'response': original_response } + return { 'new_label': new_label, 'alternative_label': alternative_label, 'status': status, 'wikipedia_url': url, 'pageid': pageid, 'alternative_wikipedia_url': alternative_url, 'alternative_pageid': alternative_pageid, 'dbpedia_uri': dbpedia_uri, 'revision_id': revision_id, 'response': original_response } @@ -121,7 +127,15 @@ if created: wp_res = query_wikipedia_title(site, label=tag_label_normalized) - new_label, status, url, pageid, dbpedia_uri, wikipedia_revision_id = wp_res['new_label'], wp_res['status'], wp_res['wikipedia_url'], wp_res['pageid'], wp_res["dbpedia_uri"], wp_res['revision_id'] + new_label = wp_res['new_label'] + alternative_label= wp_res['alternative_label'] + status = wp_res['status'] + url = wp_res['wikipedia_url'] + alternative_url = wp_res['alternative_wikipedia_url'] + pageid = wp_res['pageid'] + alternative_pageid = wp_res['alternative_pageid'] + dbpedia_uri = wp_res["dbpedia_uri"] + wikipedia_revision_id = wp_res['revision_id'] # We save the datas @@ -129,6 +143,9 @@ tag.label = new_label if status is not None: tag.url_status = status + tag.alternative_label = alternative_label + tag.alternative_wikipedia_url = alternative_url + tag.alternative_wikipedia_pageid = alternative_pageid tag.wikipedia_url = url tag.wikipedia_pageid = pageid tag.dbpedia_uri = dbpedia_uri