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

Hello all, My objective is to remove the following code from the two strings below: 
set SRVBUILD=\\srvbuild\e
want to remove: \\srvbuild\e

set INCLUDE=%SRVBUILD%\ADW61\VENDOR\MICROSOFT\PLATFORMSDK\INCLUDE;\%SR
+VBUILD%\ADW61\VENDOR\MICROSOFT\PLATFORMSDK\WTL\INCLUDE;%SRVBUILD%\KEY
+BASE\SRC32\H;%MSVCDir%\ATL\INCLUDE;%MSVCDir%\INCLUDE;

want to remove: everything from %SRVBUILD% to the first %MSVCDir% not 
+including %MSVCDir%

this is the code I have so far, but no luck:

 foreach my $line (@vcvars)
    {
    if ($line =~ /set SRVBUILD=/)
    {
        $line =~ s/^\=(\W*)/\=INI_CODE_HERE/;
        print VCVARS32 $line;
        next;

        }

    if ($line =~ /set INCLUDE=/)
    {
        $line =~ s/set INCLUDE=/TEST/;
        my $result =~ s/^\%SRVBuild(\w+)^\%MSVCDir\%/INI_CODE_HERE/;
        print "New Line: $result\n";
        print VCVARS32 $line; 
        next;
    }

    else
    {
        print VCVARS32 $line; 
    }
    }

[download]
any ideas?

Replies are listed 'Best First'.
Re: How to remove middle of string?
by Aristotle (Chancellor) on Jun 06, 2002 at 15:14 UTC
    See comments.
    foreach my $line (@vcvars) {
    if ($line =~ /^set INCLUDE=(.*)/) { ## use the ^ and $ anchors whe
+never possible
        my @paths = split(/;/, $1); # $1 is the part in brackets in th
+e regex above
        print VCVARS32 "set INCLUDE=";
        print VCVARS32 join(";",  # concatenate list of strings using 
+semicolons
            grep( !/^\%SRVBUILD\%/, @paths) # return list of the strin
+gs that do not start with %SRVBUILD%
        );
        print VCVARS32 "\n";
        next;
    }
    # no else{} necessary; next() will prevent you falling through to 
+here

    # in you "set SRVBUILD" case:
    # $line =~ s/^\=(\W*)/\=/;
    # this will not work since you're using an ^ anchor;
    # the "=" is not at the beginning of the string, is it?
    # also, \W matches any NONword-character; anything NOT a letter, d
+igit or underscore
    # but if I understood correctly you want to change everything afte
+r the "="

    # you can simply use a s///ubstitution here
    # since if the left side does not match, nothing gets replaced;
    # no need for an if(){}
    s/^(set SRVBUILD)=.*/$1=INI_CODE_HERE/;

    # and then we can just let the boilerplate print handle the modifi
+ed $line
    print VCVARS32 $line; 
}

    [download]
    ____________
    Makeshifts last the longest.
[reply]
[d/l]
Re: How to remove middle of string?
by atopolc (Pilgrim) on Jun 06, 2002 at 14:03 UTC
    This was covered fairly well in a node about extracting a range.

    this node

    Update: changed from an absolute link, thanks broquaint

[reply]
      Thanks, but I actually am trying to replace the specified text with Different text
[reply]
        The four-argument usage of perlfunc:substr will do what you are looking for. For example:

        my $string = "foo bar baz";
my $replacement = "moo";

#   This will print "foo moo baz"
print substr($string, 4, 3, $replacement), "\n";

        [download]

         

[reply]
[d/l]

Back to Seekers of Perl Wisdom