Update image regex to handle new page format (#254)

Fix test that fails due to being unable to parse image pages, and add a new test file for this case. --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
mediawiki-client-tools · Sep 23, 2024 · fb18862 · fb18862
1 parent 701c6e4
commit fb18862
Show file tree

Hide file tree

Showing 2 changed files with 613 additions and 1 deletion.
diff --git a/wikiteam3/dumpgenerator/dump/image/html_regexs.py b/wikiteam3/dumpgenerator/dump/image/html_regexs.py
@@ -29,6 +29,6 @@
         r'(?im)<td class="TablePager_col_img_name">\s*<a href[^>]*?>(?P<filename>[^>]+)</a>[^<]*?<a href="(?P<url>[^>]+)">[^<]*?</a>[^<]*?</td>\s*'
         r'<td class="TablePager_col_thumb">[^\n\r]*?</td>\s*'
         r'<td class="TablePager_col_img_size">[^<]*?</td>\s*'
-        r'<td class="(?:TablePager_col_img_user_text|TablePager_col_img_actor)">\s*(<a href="[^>]*?" title="[^>]*?">)?(?P<uploader>[^<]+?)(</a>)?\s*</td>'
+        r'<td class="(?:TablePager_col_img_user_text|TablePager_col_img_actor)">\s*(?:<a href="[^>]*?" title="[^>]*?">)?(?:<bdi>)?(?P<uploader>[^<]+?)(?:</bdi>)?(?:</a>)?\s*(?:<span class="mw-usertoollinks">(?:(?!</span>)(?!</td>).)*?</span>)?</td>'
     ),
 ]