in reply to modify file in place in script? Regex for changing includes from SSI to PHP

Using Perl to create PHP. I love it! :)

It's not that you're "not doing that part right".
It's simply that you're "not doing that part".
You never write out your changes. 

while ( my $html_file = $rule->match ) {
   # Change file in-place.
   local @ARGV = $html_file;     
   local $^I = '.bak';  # Or '' to avoid making backups.

   while (<>) {
      ...

      # Keep (possibly edited) line
      print;
   }
}

[download]

If you want something less magical, 

while ( my $html_file = $rule->match ) {
   rename($html_file, "$html_file.bak")    or die;
   open(my $fh_in,  '<', "$html_file.bak") or die;
   open(my $fh_out, '>', $html_file)       or die;

   while (<$fh_in>) {
      ...

      # Keep (possibly edited) line
      print $fh_in $_;
   }
}

[download]

Replies are listed 'Best First'.
Re^2: modify file in place in script? Regex for changing includes from SSI to PHP
by hmbscully (Scribe) on Oct 26, 2007 at 19:23 UTC
    Let me see how much I misunderstand the "less magical" code:

    rename($html_file, "$html_file.bak")    or die;
    Makes a copy of the file with the .bak extension

    open(my $fh_in,  '<', "$html_file.bak") or die;
    Opens the copy of the file for reading

    open(my $fh_out, '>', $html_file) or die;
    Opens the original file for writing, wiping out the contents?

        while (<$fh_in>) {
    while there are file contents in the backup file, do something with them 

          # Keep (possibly edited) line
      print $fh_in $_;
   }

    [download]

    Print whatever the backup file is back into the backup file?

    Ok, yeah, I still don't get this. I know I should, I've read enough examples, but I don't.
    I read it as making a copy of the file I want to change, wiping out the contents of the original, modifying the copied file somehow and writing the modified text back into the copied file? Why am I not seeing this still?

    Also, I don't get the while(<$fh_in>) { s/// } instead of using 

    my @lines = <$fh_in>;

    for (@lines) { s/// }

    [download]
    because do I want to do the s/// on the entire file at once? Don't I want to go line by line?

    As for the using Perl to write PHP, as ignorant as I may seem about Perl, I am continuously frustrated with moving into PHP. What I find is that the things that PHP does easier than Perl does not outweight the things that I could do easily in Perl that cannot do easily in PHP.

    I learn more and more about less and less until eventually I know everything about nothing.
[reply]
[d/l]
[select]

      • rename renames, not copies.

      • Oops! My error. I did
        print $fh_in $_;
        where I meant to do
        print $fh_out $_;

      • Both
        my @lines = <$fh_in>; for (@lines) { s/// }
        and
        while(<$fh_in>) { s/// }
        work a line at a time. The only difference is that the top version needlessly keeps the entire file in memory.

        Entire file at once would be

        my $text; { local $/; $text = <$fh_in>; }
$text =~ s///g;

        [download]
[reply]
[d/l]
[select]
      Ok, I studied the code some more and tested and tried and realized the < and > were reversed. This is what I've got now and it seems to work and I think I get it: 

      while ( my $html_file = $rule->match ) {
     
       rename($html_file, "$html_file.bak")    or die;
       open(my $fh_in,  '<', "$html_file.bak") or die;
       open(my $fh_out, '>', $html_file)       or die;
    
    while (<$fh_in>) {
        my @lines = <$fh_in>;

        for (@lines) {
            #replace <!--#include virtual="[document path]"-->
            #with <?php include($_SERVER['DOCUMENT_ROOT'].'[document p
+ath]'); ?>
                if (s/<!--#include virtual="(.*)"-->/<?php include(\$\
+_SERVER['DOCUMENT_ROOT'].'$1');?>/){
                 my $result = $1;
                 #print the file changed and the document path for the
+ included file
                 print OUTFILE "$html_file: $result\n";    
            }
        print $fh_out $_;
    }

    close($fh_in);
    close($fh_out);
}

      [download]
      I learn more and more about less and less until eventually I know everything about nothing.
[reply]
[d/l]

In Section Seekers of Perl Wisdom