#!/usr/local/bin/perl -w # $Id: detab.pl,v 1.6 2001/01/23 22:08:12 ted Exp $ # detab quick hack to strip table tags out of pages for WAI #called as uri,detab on this server and transparently proxies so relative uris stay relative #RewriteRule ^(.+),detab$ http://cgi.w3.org/cgi-bin/tablin.pl?uri=http://%{HTTP_HOST}$1 [P,L] $ENV{TERM}="vt100"; $ENV{LYNX_TEMP_SPACE}="/tmp"; $lynx='/usr/local/bin/lynx'; print "Content-type: text/html\n\n"; if ( $ENV{'QUERY_STRING'} ) { @pairs = split(/[&;]/, $ENV{QUERY_STRING}); foreach $pair (@pairs) { ($name, $value) = split(/=/, $pair); $value =~ tr/+/ /; $value =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg; $FORM{$name} = $value; } $uri = $FORM{uri}; } else { print "no uri, this was accessed improperly"; exit; } open(URI, "$lynx -source -validate $uri |"); $baseuri =~ $uri; #what to do about table tags spanning multi-lines? #perhaps something more efficient than mashing into one huge buffer while() { $page .= $_; } $page =~ s/<\/?(td|tr|table|tbody|thead|tfoot|th|caption|colgroup|col)[^>]*>//ig; $page =~ s/,detab//ig; # our full_uri#frag may be wrapped in ' or " and if there's one the other can appear within the uri or frag # we might just be relative as only a frag too #couldn't get back ref within a [^] :-( so i split into the 3 choices explicitly - wanted to keep 1-liner # if you're reading this and see how to make it work, let me know ted@w3.org #$page =~ s!(href *= *(\'|\")?($uri)?\#[^\2]+)!\1,detab!ig; #$page =~ s!(href *= *\'($uri)?\#[^\']+)!\1,detab!ig; #$page =~ s!(href *= *\"($uri)?\#[^\"]+)!\1,detab!ig; #$page =~ s!(href *= *($uri)?\#[^> ]+)!\1,detab!ig; print $page;