[Python-checkins] python/dist/src/Tools/webchecker README, 1.2, 1.3 webchecker.py, 1.30, 1.31

[akuchling at users.sourceforge.net]

Update of /cvsroot/python/python/dist/src/Tools/webchecker
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv1228

Modified Files:
	README webchecker.py 
Log Message:
[Patch #918212] Support XHTML's 'id' attribute, which can be on any element.

Index: README
===================================================================
RCS file: /cvsroot/python/python/dist/src/Tools/webchecker/README,v
retrieving revision 1.2
retrieving revision 1.3
diff -C2 -d -r1.2 -r1.3
*** README	17 Nov 1999 15:41:47 -0000	1.2
--- README	21 Mar 2004 19:07:23 -0000	1.3
***************
*** 19,20 ****
--- 19,23 ----
  into webchecker.py, and corresponding mods to wcgui.py and
  websucker.py.
+ 
+ - Mar 2004.  Chris Herborth contributed a patch to let webchecker.py
+ handle XHTML's 'id' attribute.

Index: webchecker.py
===================================================================
RCS file: /cvsroot/python/python/dist/src/Tools/webchecker/webchecker.py,v
retrieving revision 1.30
retrieving revision 1.31
diff -C2 -d -r1.30 -r1.31
*** webchecker.py	27 Feb 2003 06:59:10 -0000	1.30
--- webchecker.py	21 Mar 2004 19:07:23 -0000	1.31
***************
*** 785,819 ****
          sgmllib.SGMLParser.__init__(self)
  
!     def start_a(self, attributes):
!         self.link_attr(attributes, 'href')
! 
!         # We must rescue the NAME
          # attributes from the anchor, in order to
          # cache the internal anchors which are made
          # available in the page.
          for name, value in attributes:
!             if name == "name":
                  if value in self.names:
!                     self.checker.message("WARNING: duplicate name %s in %s",
                                           value, self.url)
                  else: self.names.append(value)
                  break
  
      def end_a(self): pass
  
      def do_area(self, attributes):
          self.link_attr(attributes, 'href')
  
      def do_body(self, attributes):
          self.link_attr(attributes, 'background', 'bgsound')
  
      def do_img(self, attributes):
          self.link_attr(attributes, 'src', 'lowsrc')
  
      def do_frame(self, attributes):
          self.link_attr(attributes, 'src', 'longdesc')
  
      def do_iframe(self, attributes):
          self.link_attr(attributes, 'src', 'longdesc')
  
      def do_link(self, attributes):
--- 785,833 ----
          sgmllib.SGMLParser.__init__(self)
  
!     def check_name_id( self, attributes ):
!         """ Check the name or id attributes on an element.
!         """
!         # We must rescue the NAME or id (name is deprecated in XHTML)
          # attributes from the anchor, in order to
          # cache the internal anchors which are made
          # available in the page.
          for name, value in attributes:
!             if name == "name" or name == "id":
                  if value in self.names:
!                     self.checker.message("WARNING: duplicate ID name %s in %s",
                                           value, self.url)
                  else: self.names.append(value)
                  break
  
+     def unknown_starttag( self, tag, attributes ):
+         """ In XHTML, you can have id attributes on any element.
+         """
+         self.check_name_id(attributes)
+ 
+     def start_a(self, attributes):
+         self.link_attr(attributes, 'href')
+         self.check_name_id(attributes)
+ 
      def end_a(self): pass
  
      def do_area(self, attributes):
          self.link_attr(attributes, 'href')
+         self.check_name_id(attributes)
  
      def do_body(self, attributes):
          self.link_attr(attributes, 'background', 'bgsound')
+         self.check_name_id(attributes)
  
      def do_img(self, attributes):
          self.link_attr(attributes, 'src', 'lowsrc')
+         self.check_name_id(attributes)
  
      def do_frame(self, attributes):
          self.link_attr(attributes, 'src', 'longdesc')
+         self.check_name_id(attributes)
  
      def do_iframe(self, attributes):
          self.link_attr(attributes, 'src', 'longdesc')
+         self.check_name_id(attributes)
  
      def do_link(self, attributes):
***************
*** 825,846 ****
--- 839,867 ----
                      self.link_attr(attributes, "href")
                      break
+         self.check_name_id(attributes)
  
      def do_object(self, attributes):
          self.link_attr(attributes, 'data', 'usemap')
+         self.check_name_id(attributes)
  
      def do_script(self, attributes):
          self.link_attr(attributes, 'src')
+         self.check_name_id(attributes)
  
      def do_table(self, attributes):
          self.link_attr(attributes, 'background')
+         self.check_name_id(attributes)
  
      def do_td(self, attributes):
          self.link_attr(attributes, 'background')
+         self.check_name_id(attributes)
  
      def do_th(self, attributes):
          self.link_attr(attributes, 'background')
+         self.check_name_id(attributes)
  
      def do_tr(self, attributes):
          self.link_attr(attributes, 'background')
+         self.check_name_id(attributes)
  
      def link_attr(self, attributes, *args):
***************
*** 858,861 ****
--- 879,883 ----
                          self.checker.note(1, "  Base %s", value)
                      self.base = value
+         self.check_name_id(attributes)
  
      def getlinks(self):