Set bozo bit on non-fatal errors

maksverver · maksverver · commit 7ce41769807e · 2024-08-30T00:01:39.000+02:00
Previously these errors were ignored, since an exception is raised only on fatal errors. With this change, when a non-fatal error occurs, the bozo bit is still set, but the feed is not reparsed with the loose parser. Background and discussion here: lemon24/reader#350
diff --git a/feedparser/api.py b/feedparser/api.py
@@ -327,8 +327,12 @@ def _parse_file_inplace(
             saxparser.parse(source)
         except xml.sax.SAXException as e:
             result["bozo"] = 1
-            result["bozo_exception"] = feed_parser.exc or e
+            result["bozo_exception"] = e
             use_strict_parser = False
+        else:
+            if feed_parser.bozo:
+                result["bozo"] = 1
+                result["bozo_exception"] = feed_parser.exc
 
     # The loose XML parser will be tried if the strict XML parser was not used
     # (or if it failed to parse the feed).
diff --git a/tests/illformed/always_strip_doctype.xml b/tests/illformed/always_strip_doctype.xml
@@ -1,7 +1,11 @@
 <?xml version="1.0" encoding="utf-8"?><!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://site.invalid/">
 <!--
 Description: unstripped invalid doctype
-Expect:      not bozo and feed['title'] == 'found'
+Expect:      feed['title'] == 'found'
+--><!--
+The Expect-line above doesn't check if the bozo bit is set, because it depends
+on the XML parser used: libxml2 issues an error for the invalid doctype, while
+Expat does not.
 -->
 <rss>
     <channel>
