perlstu has asked for the wisdom of the Perl Monks concerning the following question:

I'm trying to extract lowercase characters out of a html tag convert it to uppercase then overwrite the code to the same file. It should look something like this: 
if ( $fold =~m/.html$/)
    {
    open(FILE, $fold) || die "Cannot open File\n";
    @lines=<FILE>;
    close (FILE);
                
    foreach $lines(@lines)
        { 
        $lines=~s/http/https/g;
    #------------upper extent of HELP Here-------------               
+ 
    #convert html tags fromlower to upper
        #Whats wrong with match and extraction here !
              
               if ($lines=~m/(<+)(\/?)(.+)(>+)/)
                  {$tag=$3;
                   $tag=~tr/a-z/A-Z/;
                  }
           }
 #------------------------lower extent of help -----------
          @log=@lines;
      open(FILE,">$fold") || die "cannot open file";
      print FILE @lines;
      open(LOG,"+>>logfile.txt") || die "log file will not open!";
    print LOG @log,"\n","#------newlog---------","\n";
                close (LOG);
                close (FILE);
                
    }
        
}

[download]

Not all code is present, some bits maybe missing from copy and paste

Replies are listed 'Best First'.
Re: perl html tag editing to uppercase
by grep (Monsignor) on Mar 31, 2002 at 07:34 UTC
    $lines=~m/(<+)(\/?)(.+)(>+)/

    This unfortuately is not the right approach. You regex will match quite a bit more than you expect.

    You should look into a module like HTML::TokeParser.

    Now getting back to why your regex wasn't helping:
    m/(<+)(\/?)(.+)(>+)/
  • (<+) - capture one or more occurances of '<' so << or <<<<<<<<< will match and be captured to $1
  • (\/?) - capture zero or one '/' to $2. Not that this matters because if you use the uc function instead of your tr/// the '/' will be ignored
  • (.+) - capture 1 or more of anything but a line ending to $3. NOTE: this is greedy and will gobble everything in site unless it's required to finish the match
  • (>+) - capture one or more '>' (like >> or >>>>) to $4. NOTE: because the the '.+' is greedy you can have some very strange results with 2 '>' on a line whether or not it is in a tag.

    • imagine this line of HTML

    look at this <<<----- <b>It is really neat</b> I swear >>

    What will be captured to $3 is

    ----- <b>It is really neat</b> I swear

    One tool to add to your regex arsenal is negated character classes like [^ ]. Rewriting your regex like /<([^>]+)>/ would get you closer, but still is not what you want. You can read more about negated character classes in perlre

    Writing a HTML parser with a regex is a somewhat like hammering a nail with a flounder. You should really look for right tool in perl, in this case a true parser is going to help you out a lot.



    grep
    grep> cd /pub
    grep> more beer
[reply]
[d/l]
[select]
      Thankyou for your help :) I found a oneliner that doesn't do the whole job but still
      $lines =~ s/<(.+?)>/'<'.uc($1).'>'/ge;
      This will convert src attributes such as blah.gif to uppercase which would wreak havock on some servers but its only an exercise to drive me nuts :) Else I could:
      
if ($lines=~m/(<)(.+)(>)/)
    {
                            $tag=$1;
                            $tag2=$2;
                            $tag3=$3;
                            $tag2=~tr/a-z/A-Z/;
                            $new=$tag.$tag2.$tag3;
    }                        
    $upper=~s/(<)(.+)(>)/$new/g;

      [download]
      Boy am I hangin to learn about modules now :)
[reply]
[d/l]
[select]

Back to Seekers of Perl Wisdom