On 5/6/05, macromedia wrote:
>
> Hello,
>
> I'm not sure about the login or how I should approach what I want to
> achieve. I require a script that does the following.
>
> 1. Search a directory and all sub-directories for a certain file
> extension. (Ex. .txt)
>
> 2. Get the results of the above search and check "inside" each file for
> a string match. (Ex. "123").
>
> 3. List the matches found into an external file for further processing.
>
> What features should I be looking at to achieve this? Anyone know of
> some good examples that I could pick away at?
>
I don't know if it's good (haven't tested the code!) but here's one example:
################# begin code
use strict;
use warnings;
use File::Find;
my $log_file = "log_file.txt";
open(OUT,">$log_file") or
die "Couldn't open $log_file for writing: $!\n";
find(\&wanted, "dir_name");
close(OUT) or
die "Couldn't close $log_file after writing: $!\n";
sub wanted {
if (m/\.txt$/) {
open(IN,$File::Find::name) or
die "Couldn't open $File::Find::name for reading: $!\n";
while(defined(my $line=<IN>)) {
if ($line =~ m/123/) {
print OUT "$File::Find::name : $line";
}
}
close(IN) or
die "Couldn't close $File::Find::name after reading: $!\n";
}
}
################# end code
Copy-paste this code (not including the "########" lines) into a text
editor that can show you line numbers, since I'll be mentioning them
in the following explanation:
* lines 1-2 are a must in every Perl program you write. Don't you dare
write another script without them, or someone from this list will have
to whip you with a differential SCSI cable ;-)
Read more about these pragmas here:
http://perldoc.perl.org/strict.html
http://perldoc.perl.org/warnings.html
Basically they help you write better code which is (hopefully) less
buggy. Get used to writing them at the top of every Perl program.
* Line 3 "use's" the File::Find module, allowing us to use the "find"
function that it supplies (this is used later on, in line 7). Read
here about "use":
http://perldoc.perl.org/functions/use.html
and here about the File::Find module:
http://perldoc.perl.org/File/Find.html
* Line 4 defines a variable holding the name of the log file, the
external one you mentioned in point (3). Feel free to change the file
name :-)
* Lines 5-6 are actually one Perl command (notice there is only a ";"
at the end of line 6) that is built of 3 basic parts:
1. An "open" command. Read:
http://perldoc.perl.org/functions/open.html
for more info about this function.
2. An "or" operator. If the left side of it (the "open" function)
succeeds, it will return true, and the right side will not be
evaluated at all. However, if the "open" fails, it will return zero or
false, and the right side will be evaluated by the Perl interpreter.
3. A "die" statement, because if the "open" fails, we want to punish
the script by causing it to die a horrible death :-) You can read more
about "die" here:
http://perldoc.perl.org/functions/die.html
* Line 7 does the actual work, calling the "wanted" subroutine
("subroutine == function" in Perl, BTW) which we wrote to do the job.
It will start at the root directory we provide (again, change this to
your actuall root directory, full path and all) and run the wanted sub
for every file below that directory.
* Lines 8-9 - once we finish doing the work, we want to close the
output filehandle. It will close anyway when the script ends, but I
like to keep things neat :-) Read more about "close" here:
http://perldoc.perl.org/functions/close.html
* Line 10 - defines the wanted subroutine. Read about Perl subroutines here:
http://perldoc.perl.org/perlsub.html
Line 22 is where the closing brace that matches the opening one on line 10 is.
* Line 11 - Uses the "if" flow control statement, the matching
operator "m//" and a simple RE inside this operator. You can read
about "if" at:
http://perldoc.perl.org/perlsyn.html
You can read about the "m//" operator at:
http://perldoc.perl.org/perlop.html
You can read about regular expressions here:
http://perldoc.perl.org/perlre.html
Basically, what happens is that File::Find passes the current file
name into the special variable "$_" (read
http://perldoc.perl.org/perlvar.html) each time it calls "wanted". We
match against this file name using "m//" and look for the string "a
literal period followed by 't' followed by 'x' followed by 't' which
is at the end of the string".
* Lines 12-13 - if the file is a ".txt" file, we want to open it for
reading, look at its contents for the required string and print each
line containing such a string to the log file. These 2 lines take care
of the "open" part. The idea is the same as lines 5-6, except for the
file name and the fact that we are opening it for reading, not writing
(hence no ">" part). I'll let you figure out the other details for
yourself.
* Line 14 - reads every line from the current file associated with the
IN filehandle and saves this line to the $line variable. For each such
line, it will run the code delimited by the curly braces which start
at line 14 and close at line 18. Read the following pages for details:
http://perldoc.perl.org/perlsyn.html (about the "while" flow control keyword)
http://perldoc.perl.org/functions/defined.html (about the "defined" function)
http://perldoc.perl.org/perlop.html#I-O-Operators (about the
"<FILEHANDLE>" construct)
* Line 15 - this time we want to match against the $line variable, not
$_, so we use the binding operator "=~".Read:
http://perldoc.perl.org/perlop.html
for details. The RE itself is the simplest possible - a simple string
* Line 16 - finally, if the current file was a ".txt" file and the
current line matched the string "123", we "print" the string
"$File::Find::name : $line" to the file associated with the OUT
filehandle. For more about "print" read:
http://perldoc.perl.org/functions/print.html
HTH,
--
Offer Kaye
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
<http://learn.perl.org/> <http://learn.perl.org/first-response>
