Fixed #129: Adapted DOM Integrity Checker to EN's Top categories change

openzim · May 25, 2022 · 8c5f52c · 8c5f52c
1 parent 7648d5b
commit 8c5f52c
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 2 deletions.
diff --git a/wikihow2zim/rewriter.py b/wikihow2zim/rewriter.py
@@ -181,7 +181,9 @@ def rewrite_links_for_excludes(self, soup, to_root):
             # categorlyListing thumbnail link (english)
             seldef(".cat_container a[href]", True, True, False),
             # categorylisting link to category
-            seldef("#catlist_container #catlist a[href]", True, True, False),
+            seldef("#catlist a[href]", True, True, False),
+            # EN Top categories categorylisting link to category
+            seldef("#catlist_container .catlist a[href]", True, True, False),
             # top breadcrumb in article page
             seldef(".breadcrumbs a[href]", True, True, False),
             # top breadcrumb in article page

diff --git a/wikihow2zim/scraper.py b/wikihow2zim/scraper.py
@@ -849,7 +849,7 @@ def check_dom_integrity(self):
         soup, _ = get_soup("/Special:CategoryListing")
         if not soup.select("#content_wrapper"):
             raise DomIntegrityError("#content_wrapper not found")
-        category_links = soup.select("#catlist_container #catlist a")
+        category_links = soup.select("#catlist a")
 
         if not category_links:
             raise DomIntegrityError("No links in #catlist_container")