[TWiki-Dev] CVS: twiki/lib TWiki.pm,1.185,1.186

[Richard D. <rd...@us...>]

Update of /cvsroot/twiki/twiki/lib
In directory sc8-pr-cvs1:/tmp/cvs-serv28466/lib

Modified Files:
	TWiki.pm 
Log Message:
Fix for non-encoding of 8-bit characters in RSS feeds - any skins
named 'rss*' are rendered in $pageMode = 'rss', i.e. 8-bit characters
are rendered as &#nnn;.  Such characters are not encoded in HTML pages.

Tested on donkin.org RSS feed since this needs to go into TWiki.org, e.g.
http://donkin.org/bin/view/Main/WebRss?skin=rss&contenttype=text/xml has 
an example, while http://donkin.org/bin/view/Test/WebChanges doesn't encode
8-bit characters.


Index: TWiki.pm
===================================================================
RCS file: /cvsroot/twiki/twiki/lib/TWiki.pm,v
retrieving revision 1.185
retrieving revision 1.186
diff -C2 -r1.185 -r1.186
*** TWiki.pm	9 Dec 2002 19:32:01 -0000	1.185
--- TWiki.pm	10 Dec 2002 21:54:37 -0000	1.186
***************
*** 85,88 ****
--- 85,89 ----
          $formatVersion $OS
          $readTopicPermissionFailed
+ 	$pageMode
      );
  
***************
*** 97,101 ****
  	$singleUpperAlphaRegex $singleLowerAlphaRegex $singleUpperAlphaNumRegex
  	$singleMixedAlphaNumRegex $singleMixedNonAlphaNumRegex 
! 	$mixedAlphaNumRegex
      );
  
--- 98,102 ----
  	$singleUpperAlphaRegex $singleLowerAlphaRegex $singleUpperAlphaNumRegex
  	$singleMixedAlphaNumRegex $singleMixedNonAlphaNumRegex 
! 	$singleMixedNonAlphaRegex $mixedAlphaNumRegex
      );
  
***************
*** 189,192 ****
--- 190,195 ----
  $basicInitDone = 0;		# basicInitialize not yet done
  
+ $pageMode = 'html';		# Default is to render as HTML
+ 
  
  # =========================
***************
*** 471,475 ****
      $singleMixedAlphaNumRegex = qr/[${upperAlpha}${lowerAlpha}${numeric}]/;
  
!     $singleMixedNonAlphaNumRegex = qr/[^${upperAlpha}${lowerAlpha}${numeric}]/;
  
      # Multi-character alpha-based regexes
--- 474,479 ----
      $singleMixedAlphaNumRegex = qr/[${upperAlpha}${lowerAlpha}${numeric}]/;
  
!     $singleMixedNonAlphaRegex = qr/[^${upperAlpha}${lowerAlpha}${numeric}]/;
!     $singleMixedNonAlphaNumRegex = qr/[^${upperAlpha}${lowerAlpha}]/;
  
      # Multi-character alpha-based regexes
***************
*** 511,514 ****
--- 515,519 ----
      my( $pluginHeaders, $coreHeaders );
  
+ 
      $contentType .= "; charset=$siteCharset";
  
***************
*** 585,588 ****
--- 590,602 ----
  
  # =========================
+ # Set page mode:
+ #   - 'rss' - encode 8-bit characters as XML entities
+ #   - 'html' - no encoding of 8-bit characters
+ sub pageMode
+ {
+     $pageMode = shift;
+ }
+ 
+ # =========================
  sub getCgiQuery
  {
***************
*** 1118,1131 ****
      $htext =~ s/\s+[\+\-]*/ /g;       # remove newlines and special chars
  
!     # limit to 162 chars
      $htext =~ s/(.{162})($mixedAlphaNumRegex)(.*?)$/$1$2 \.\.\./g;
  
!     # Commented out by RD - encoding breaks non-ISO-8859-1 character sets, and 
!     # the browsers and RSS readers work OK with unencoded 8 bit characters
!     # within RSS feeds, as long as %CHARSET% is included in template.
!     #
!     # Encode special chars into HTML &#nnn; entities for international
!     # character support 
!     # $htext =~ s/([\x7f-\xff])/"\&\#" . unpack( "C", $1 ) .";"/ge;
  
      # inline search renders text, so prevent linking of external and
--- 1132,1145 ----
      $htext =~ s/\s+[\+\-]*/ /g;       # remove newlines and special chars
  
!     # limit to 162 chars 
!     # FIXME I18N: Avoid splitting within multi-byte character sets
      $htext =~ s/(.{162})($mixedAlphaNumRegex)(.*?)$/$1$2 \.\.\./g;
  
!     # Encode special chars into XML &#nnn; entities for use in RSS feeds
!     # - no encoding for HTML pages, to avoid breaking international 
!     # characters.
!     if( $pageMode eq 'rss' ) {
! 	$htext =~ s/([\x7f-\xff])/"\&\#" . unpack( "C", $1 ) .";"/ge;
!     }
  
      # inline search renders text, so prevent linking of external and