Modifying warnings when a charset override has been selected for a document with no charset declared at all.

Should fix http://www.w3.org/Bugs/Public/show_bug.cgi?id=1133

2 files changed, 41 insertions, 14 deletions

diff --git a/httpd/cgi-bin/check b/httpd/cgi-bin/check
index d04d5d4..859a7ea 100755
--- a/httpd/cgi-bin/check
+++ b/httpd/cgi-bin/check
@@ -9,7 +9,7 @@
 # This source code is available under the license at:
 #     http://www.w3.org/Consortium/Legal/copyright-software
 #
-# $Id: check,v 1.407 2005-03-24 02:41:12 ot Exp $
+# $Id: check,v 1.408 2005-03-24 08:37:26 ot Exp $
 
 #
 # Disable buffering on STDOUT!
@@ -220,7 +220,7 @@ Directory not readable (permission denied): @_r
 
   #
   # Strings
-  $VERSION    =  q$Revision: 1.407 $;
+  $VERSION    =  q$Revision: 1.408 $;
   $VERSION    =~ s/Revision: ([\d\.]+) /$1/;
 
   #
@@ -454,29 +454,40 @@ if (&conflict($File->{Opt}->{Charset}, '(detect automatically)')) {
   my ($override, undef) = split(/\s/, $File->{Opt}->{Charset}, 2);
   $File->{Charset}->{Override} = lc($override);
 
-  if ($File->{Opt}->{FB}->{Charset}) {
+  if ($File->{Opt}->{FB}->{Charset}) { # charset fallback mode
     unless ($File->{Charset}->{Use}) {
       &add_warning('W02', {W02_charset => $File->{Charset}->{Override}});
 
       $File->{Tentative} |= T_ERROR; # Tag it as Invalid.
       $File->{Charset}->{Use} = $File->{Charset}->{Override};
     }
-  } else {
-    # Warn about Override unless it's the same as the real charset...
-    unless ($File->{Charset}->{Override} eq $File->{Charset}->{Use}) {
-      &add_warning('W03', {
-        W03_use => $File->{Charset}->{Use},
-        W03_opt => $File->{Charset}->{Override},
-      });
-
+  } else { # charset "hard override" mode
+    if (! $File->{Charset}->{Use}) { # overriding "nothing"
+      &add_warning('W04', {W04_charset => $File->{Charset}->{Override}, W04_override => TRUE});
       $File->{Tentative} |= T_ERROR;
       $File->{Charset}->{Use} = $File->{Charset}->{Override};
+     }
+     else { #actually overriding something 
+      # Warn about Override unless it's the same as the real charset...
+    
+      unless ($File->{Charset}->{Override} eq $File->{Charset}->{Use}) {
+        &add_warning('W03', {
+          W03_use => $File->{Charset}->{Use},
+          W03_opt => $File->{Charset}->{Override},
+        });
+
+        $File->{Tentative} |= T_ERROR;
+        $File->{Charset}->{Use} = $File->{Charset}->{Override};
+      } 
     }
+    
+    
+    
   }
 }
 
 unless ($File->{Charset}->{Use}) { # No charset given...
-  &add_warning('W04', {});
+  &add_warning('W04', {W04_charset => 'UTF-8'});
   $File->{Tentative} |= T_ERROR; # Can never be valid.
   $File->{Charset}->{Use} = 'utf-8';
 }
diff --git a/share/templates/en_US/warnings.tmpl b/share/templates/en_US/warnings.tmpl
index 319709d..110af10 100644
--- a/share/templates/en_US/warnings.tmpl
+++ b/share/templates/en_US/warnings.tmpl
@@ -48,15 +48,31 @@
 </TMPL_IF>
 <TMPL_IF NAME="W04">
   <dt id="W04">No Character Encoding Found!
-    Falling back to <code>UTF-8</code>.</dt>
+    <TMPL_IF NAME="W04_override">
+      Using
+    <TMPL_ELSE>
+      Falling back to 
+    </TMPL_IF>
+    <code><TMPL_IF NAME="W04_charset"><TMPL_VAR NAME="W04_charset" ESCAPE="HTML"><TMPL_ELSE>UTF-8</TMPL_IF></code>.
+  </dt>
   <dd>
+  <TMPL_IF NAME="W04_override">
+    <p>I was not able to extract a character encoding labeling from any of
+      the valid sources for such information. I will use the character 
+      encoding <code><TMPL_IF NAME="W04_charset"><TMPL_VAR NAME="W04_charset" ESCAPE="HTML"><TMPL_ELSE>UTF-8</TMPL_IF></code>,
+      which you have chosen via the <code>Encoding</code> override option, for 
+      tentative validation. 
+    </p>
+  <TMPL_ELSE>
     <p>
       I was not able to extract a character encoding labeling from any of
       the valid sources for such information. Without encoding information
       it is impossible to reliably validate the document. I'm falling back
-      to the "UTF-8" encoding and will attempt to perform the validation,
+      to the "<code><TMPL_IF NAME="W04_charset"><TMPL_VAR NAME="W04_charset" ESCAPE="HTML"><TMPL_ELSE>UTF-8</TMPL_IF></code>" 
+      encoding and will attempt to perform the validation,
       but this is likely to fail for all non-trivial documents.
     </p>
+  </TMPL_IF>
   <TMPL_IF NAME="opt_verbose">
     <p>The sources I tried to find encoding information include:</p>
     <ul>