Merge Mark-Hetherington-xml-parse-warn with slight modifications.

linkcheck/checker/const.py

@@ -100,6 +100,7 @@ WARN_IGNORE_URL = "ignore-url"
 WARN_MAIL_NO_MX_HOST = "mail-no-mx-host"
 WARN_NNTP_NO_SERVER = "nntp-no-server"
 WARN_NNTP_NO_NEWSGROUP = "nntp-no-newsgroup"
+WARN_XML_PARSE_ERROR = "xml-parse-error"
 
 # registered warnings
 Warnings = {
@@ -123,4 +124,5 @@ Warnings = {
     WARN_NNTP_NO_SERVER: _("No NNTP server was found."),
     WARN_NNTP_NO_NEWSGROUP: _("The NNTP newsgroup could not be found."),
     WARN_URL_OBFUSCATED_IP: _("The IP is obfuscated."),
+    WARN_XML_PARSE_ERROR: _("XML could not be parsed."),
 }

linkcheck/director/checker.py

@@ -59,6 +59,7 @@ def check_url(url_data, logger):
                     # redirect aliases
                     cache.add_result(alias, result)
                 # parse content recursively
+                # XXX this could add new warnings which should be cached.
                 if do_parse:
                     parser.parse_url(url_data)
             finally:

linkcheck/parser/sitemap.py

@@ -18,7 +18,8 @@
 Main functions for link parsing
 """
 from xml.parsers.expat import ParserCreate
-
+from xml.parsers.expat import ExpatError
+from ..checker.const import (WARN_XML_PARSE_ERROR)
 
 class XmlTagUrlParser(object):
     """Parse XML files and find URLs in text content of a tag name."""
@@ -40,8 +41,10 @@ class XmlTagUrlParser(object):
         self.url = u""
         data = url_data.get_content()
         isfinal = True
-        self.parser.Parse(data, isfinal)
-
+        try:
+            self.parser.Parse(data, isfinal)
+        except ExpatError as expaterr:
+            self.url_data.add_warning(expaterr.message,tag=WARN_XML_PARSE_ERROR)
     def start_element(self, name, attrs):
         """Set tag status for start element."""
         self.in_tag = (name == self.tag)