Make sure ignore URL patterns take precedence over nofollow URL patterns.

doc/changelog.txt

@@ -5,6 +5,9 @@ Fixes:
 - checking: Correctly detect empty FTP paths as directories.
 - checking: Reuse connections more than once and ensure they are
   closed before expiring.
+- checking: Make sure "ignore" URL patterns are checked before
+  "nofollow" URL patterns.
+  Closes: SF bug #3184973
 - gui: Remove old context menu action to view URL properties.
 - gui: Disable viewing of parent URL source if it's a directory.
 

linkcheck/__init__.py

@@ -103,7 +103,7 @@ def get_link_pat (arg, strict=False):
     @return: dictionary with keys 'pattern', 'negate' and 'strict'
     @rtype: dict
     """
-    log.debug(LOG_CHECK, "Link pattern %r", arg)
+    log.debug(LOG_CHECK, "Link pattern %r strict=%s", arg, strict)
     if arg.startswith('!'):
         pattern = arg[1:]
         negate = True

linkcheck/configuration/confparse.py

@@ -191,10 +191,6 @@ class LCConfigParser (ConfigParser.RawConfigParser, object):
         Read configuration options in section "filtering".
         """
         section = "filtering"
-        if self.has_option(section, "nofollow"):
-            for line in read_multiline(self.get(section, "nofollow")):
-                pat = get_link_pat(line, strict=0)
-                self.config["externlinks"].append(pat)
         if self.has_option(section, "ignorewarnings"):
             self.config['ignorewarnings'] = [f.strip() for f in \
                  self.get(section, 'ignorewarnings').split(',')]
@@ -202,6 +198,10 @@ class LCConfigParser (ConfigParser.RawConfigParser, object):
             for line in read_multiline(self.get(section, "ignore")):
                 pat = get_link_pat(line, strict=1)
                 self.config["externlinks"].append(pat)
+        if self.has_option(section, "nofollow"):
+            for line in read_multiline(self.get(section, "nofollow")):
+                pat = get_link_pat(line, strict=0)
+                self.config["externlinks"].append(pat)
         if self.has_option(section, "internlinks"):
             pat = get_link_pat(self.get(section, "internlinks"))
             self.config["internlinks"].append(pat)

linkchecker

@@ -605,13 +605,13 @@ if options.warnings is not None:
     config["warnings"] = options.warnings
 if options.anchors is not None:
     config["anchors"] = options.anchors
-if options.extern:
-    pats = [linkcheck.get_link_pat(arg) for arg in options.extern]
-    config["externlinks"].extend(pats)
 if options.externstrict:
     pats = [linkcheck.get_link_pat(arg, strict=True) \
             for arg in options.externstrict]
     config["externlinks"].extend(pats)
+if options.extern:
+    pats = [linkcheck.get_link_pat(arg) for arg in options.extern]
+    config["externlinks"].extend(pats)
 if options.output:
     if "/" in options.output:
         logtype, encoding = options.output.split("/", 1)