This probably doesn't work ... somewhere? But on the all the most pathological pages I could find (sample of 5 in the DATA section), it seems to do fine.
#! perl -slw use strict; use LWP::Simple; while( <DATA> ) { chomp; my $html = get $_; $html =~ s[( <FONT \s+ (?: (?:"[^"]+") | (?:'[^']+') | [^>]* ) > ) +] #" { my $tag = $1; $tag =~ s[ face \s* = \s* (?: (?:"[^"]+") | (?:'[^']+') | [^\s +]+ ) ][]ixm; #" $tag }eximg; my $out = $ENV{TMP} . '\\' .time() . '.htm'; open OUT, '>', $out or warn $!; print OUT $html; close OUT; system( $out ); } __DATA__ http://www.webdiner.com/annexe/font/font.htm http://www.electricearl.com/fonttest.html http://www.ilovethisplace.com/webdesign/fonts.html http://www.york.ac.uk/depts/maths/symbchrc.htm http://www.tedmontgomery.com/tutorial/style.html#face
Note: system probably won't load the modifed sample directly into the browser on systems who's commands line don't know what do with .htm files.
In reply to Re: Stripping font "face" values from font tags
by BrowserUk
in thread Stripping font "face" values from font tags
by Anonymous Monk
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |