documentation_crawler: Stop treating all portico URLs as external.

Signed-off-by: Anders Kaseorg <anders@zulipchat.com>

tools/documentation_crawler/documentation_crawler/spiders/check_help_documentation.py

@@ -55,6 +55,14 @@ class APIDocumentationSpider(UnusedImagesLinterSpider):
     images_path = "static/images/api"
 
 class PorticoDocumentationSpider(BaseDocumentationSpider):
+    def _is_external_url(self, url: str) -> bool:
+        return (
+            not url.startswith('http://localhost:9981')
+            or url.startswith('http://localhost:9981/help')
+            or url.startswith('http://localhost:9981/api')
+            or self._has_extension(url)
+        )
+
     name = 'portico_documentation_crawler'
     start_urls = ['http://localhost:9981/hello',
                   'http://localhost:9981/history',

tools/documentation_crawler/documentation_crawler/spiders/common/spiders.py

@@ -31,6 +31,10 @@ VNU_IGNORE = re.compile(r'|'.join([
     r'The first occurrence of ID “[^”]*” was here\.',
     r'Attribute “markdown” not allowed on element “div” at this point\.',
     r'No “p” element in scope but a “p” end tag seen\.',
+    r'Element “div” not allowed as child of element “ul” in this context\. '
+    + r'\(Suppressing further errors from this subtree\.\)',
+    r'The element “button” must not appear as a descendant of the “a” element\.',
+    r'Attribute “href” not allowed on element “button” at this point\.',
 
     # Warnings that are probably less important.
     r'The “type” attribute is unnecessary for JavaScript resources\.',