import/mediawiki: more improvements

2022-07-11 19:30:57 +02:00 · 2022-07-11 19:30:57 +02:00 · 74b26cecfa
commit 74b26cecfa
parent 5689b79e13
1 changed files with 7 additions and 0 deletions
--- a/import/mediawiki/main.py
+++ b/import/mediawiki/main.py
@ -51,6 +51,13 @@ for _, elem in parser:
        if (page.model != "wikitext"
                or page.redirect is not None
                or page.markup is None):
+            elem.clear()
+            del elem
+            continue
+        if (page.title.startswith("Wikipedia:")
+                or page.title.startswith("Template:")):
+            elem.clear()
+            del elem
            continue
        content = mw.parse(page.markup).strip_code()
        sha = hashlib.sha512()