Change Details

During the process of extracting all images for a campaign, we ignore categories that have already been seen, but no not ignore images that have already been seen. The only way this error shows up at present is that the count of images is larger than it should be (in cases where duplicate images have been added). But obviously this also increases the load on the database. Example: For the following categories with depth=0 - Turkey photographs taken on 2020-06-14 - Taxis in Ankara Petscan shows number of files in this combination of categories as [[ https://petscan.wmflabs.org/?labels_any=&sparql=&search_max_results=500&show_disambiguation_pages=both&after=&outlinks_any=&cb_labels_any_l=1&show_soft_redirects=both&before=&sortby=none&search_query=&outlinks_yes=&combination=union&max_age=&ores_prob_from=%C2%AEexp_filter&labels_yes=&cb_labels_yes_l=1&common_wiki=auto&langs_labels_yes=&links_to_any=&search_wiki=&edits%5Bflagged%5D=both&output_limit=&cb_labels_no_l=1&interface_language=en&templates_any=&common_wiki_other=&ns%5B6%5D=1&active_tab=tab_categories&categories=Turkey+photographs+taken+on+2020-06-14%0D%0ATaxis+in+Ankara&subpage_filter=either&wikidata_source_sites=&pagepile=&search_filter=&language=commons&source_combination=&referrer_url=&sortorder=ascending&links_to_all=&ores_type=any&project=wikimedia&doit= | 38 files in total]], while ISA shows 39. The extra one image is a duplicate, because there is [[ https://petscan.wmflabs.org/?sortorder=ascending&cb_labels_yes_l=1&after=&links_to_any=&langs_labels_no=&ores_prob_from=%C2%AEexp_filter&namespace_conversion=keep&templates_any=&edits%5Bflagged%5D=both&labels_yes=&templates_yes=&outlinks_yes=&wikidata_item=no&search_query=&regexp_filter=&cb_labels_any_l=1&outlinks_no=&max_age=&max_sitelink_count=&sitelinks_yes=&manual_list_wiki=&project=wikimedia&categories=Turkey+photographs+taken+on+2020-06-14%0D%0ATaxis+in+Ankara&before=&links_to_no=&wikidata_prop_item_use=&labels_no=&language=commons&ores_type=any&sitelinks_no=&active_tab=tab_categories&ns%5B6%5D=1&show_soft_redirects=both&edits%5Bbots%5D=both&cb_labels_no_l=1&common_wiki=auto&interface_language=en&templates_no=&wikidata_label_language=&search_max_results=500&search_filter=&doit= | one image which appears in both categories ]].

During the process of extracting all images for a campaign, we ignore categories that have already been seen, but do not ignore images that have already been seen. The only way this error shows up at present is that the count of images is larger than it should be (in cases where duplicate images have been added). But obviously this also increases the load on the database. Example: For the following categories with depth=0 - Turkey photographs taken on 2020-06-14 - Taxis in Ankara Petscan shows number of files in this combination of categories as [[ https://petscan.wmflabs.org/?labels_any=&sparql=&search_max_results=500&show_disambiguation_pages=both&after=&outlinks_any=&cb_labels_any_l=1&show_soft_redirects=both&before=&sortby=none&search_query=&outlinks_yes=&combination=union&max_age=&ores_prob_from=%C2%AEexp_filter&labels_yes=&cb_labels_yes_l=1&common_wiki=auto&langs_labels_yes=&links_to_any=&search_wiki=&edits%5Bflagged%5D=both&output_limit=&cb_labels_no_l=1&interface_language=en&templates_any=&common_wiki_other=&ns%5B6%5D=1&active_tab=tab_categories&categories=Turkey+photographs+taken+on+2020-06-14%0D%0ATaxis+in+Ankara&subpage_filter=either&wikidata_source_sites=&pagepile=&search_filter=&language=commons&source_combination=&referrer_url=&sortorder=ascending&links_to_all=&ores_type=any&project=wikimedia&doit= | 38 files in total]], while ISA shows 39. The extra one image is a duplicate, because there is [[ https://petscan.wmflabs.org/?sortorder=ascending&cb_labels_yes_l=1&after=&links_to_any=&langs_labels_no=&ores_prob_from=%C2%AEexp_filter&namespace_conversion=keep&templates_any=&edits%5Bflagged%5D=both&labels_yes=&templates_yes=&outlinks_yes=&wikidata_item=no&search_query=&regexp_filter=&cb_labels_any_l=1&outlinks_no=&max_age=&max_sitelink_count=&sitelinks_yes=&manual_list_wiki=&project=wikimedia&categories=Turkey+photographs+taken+on+2020-06-14%0D%0ATaxis+in+Ankara&before=&links_to_no=&wikidata_prop_item_use=&labels_no=&language=commons&ores_type=any&sitelinks_no=&active_tab=tab_categories&ns%5B6%5D=1&show_soft_redirects=both&edits%5Bbots%5D=both&cb_labels_no_l=1&common_wiki=auto&interface_language=en&templates_no=&wikidata_label_language=&search_max_results=500&search_filter=&doit= | one image which appears in both categories ]].

During the process of extracting all images for a campaign, we ignore categories that have already been seen, but nodo not ignore images that have already been seen. The only way this error shows up at present is that the count of images is larger than it should be (in cases where duplicate images have been added). But obviously this also increases the load on the database. Example: For the following categories with depth=0 - Turkey photographs taken on 2020-06-14 - Taxis in Ankara Petscan shows number of files in this combination of categories as [[ https://petscan.wmflabs.org/?labels_any=&sparql=&search_max_results=500&show_disambiguation_pages=both&after=&outlinks_any=&cb_labels_any_l=1&show_soft_redirects=both&before=&sortby=none&search_query=&outlinks_yes=&combination=union&max_age=&ores_prob_from=%C2%AEexp_filter&labels_yes=&cb_labels_yes_l=1&common_wiki=auto&langs_labels_yes=&links_to_any=&search_wiki=&edits%5Bflagged%5D=both&output_limit=&cb_labels_no_l=1&interface_language=en&templates_any=&common_wiki_other=&ns%5B6%5D=1&active_tab=tab_categories&categories=Turkey+photographs+taken+on+2020-06-14%0D%0ATaxis+in+Ankara&subpage_filter=either&wikidata_source_sites=&pagepile=&search_filter=&language=commons&source_combination=&referrer_url=&sortorder=ascending&links_to_all=&ores_type=any&project=wikimedia&doit= | 38 files in total]], while ISA shows 39. The extra one image is a duplicate, because there is [[ https://petscan.wmflabs.org/?sortorder=ascending&cb_labels_yes_l=1&after=&links_to_any=&langs_labels_no=&ores_prob_from=%C2%AEexp_filter&namespace_conversion=keep&templates_any=&edits%5Bflagged%5D=both&labels_yes=&templates_yes=&outlinks_yes=&wikidata_item=no&search_query=&regexp_filter=&cb_labels_any_l=1&outlinks_no=&max_age=&max_sitelink_count=&sitelinks_yes=&manual_list_wiki=&project=wikimedia&categories=Turkey+photographs+taken+on+2020-06-14%0D%0ATaxis+in+Ankara&before=&links_to_no=&wikidata_prop_item_use=&labels_no=&language=commons&ores_type=any&sitelinks_no=&active_tab=tab_categories&ns%5B6%5D=1&show_soft_redirects=both&edits%5Bbots%5D=both&cb_labels_no_l=1&common_wiki=auto&interface_language=en&templates_no=&wikidata_label_language=&search_max_results=500&search_filter=&doit= | one image which appears in both categories ]].