Perl - Code review [closed] - perl

Question

Closed. This question is off-topic. It is not currently accepting answers.
Want to improve this question? Update the question so it's on-topic for Stack Overflow.
Closed 9 years ago.
Improve this question
I am working on a program that takes information from a CSV file as a source to search with through a text file that has "customer packages". I am getting odd counts on only some of the entries, and I can't seem to figure out what is causing the duplicate counts. Can anyone look through my code and tell me if my logic/syntax is off? (probably is). All i am trying to accomplish is to count the total occurances in the text file of an entry in the csv file (packageid,package_description)
Thanks for the help! im going nuts over here.
#!/usr/bin/perl
use strict;
use Text::CSV;
# Variables already declared in the other PL file ** Remove if consolidating **
my $file2 = 'master_plist.csv';
my $csv2 = Text::CSV->new(); # Create a Text::CSV object
open (CSV2, "<", $file2) or die $!; #open CSV file for parsing
while (<CSV2>) {
if ($csv2->parse($_)) {
my #columns2 = $csv2->fields(); # Parse CSV and load into an array for each row.
my $packID = $columns2[0];
my $packDESC = $columns2[1];
my $val = 'customer_packages_report.txt';
chomp ($val);
my $cnt=0;
open (HNDL, "$val") || die "wrong filename";
while ($val = <HNDL>)
{
while ($val =~ /$packID - $packDESC/ig)
{
$cnt++;
}
}
#if ($packDESC =~ /\(/g) {
# $packDESC =~ s/\(/\(/g;
#}
print "Total iterations of $packDESC: $cnt\n";
close (HNDL);
# End original code
} # Close IF
} # Close WHILE
close CSV;

Answer 1

#!/usr/bin/perl
use strict;
use warnings;
use Text::CSV;
# Variables already declared in the other PL file ** Remove if consolidating **
my $file2 = 'master_plist.csv';
my $csv2 = Text::CSV->new(); # Create a Text::CSV object
open (CSV2, "<", $file2) or die "I die while opening $file2! $!"; #open CSV file for parsing
while ($each_csv2_line=<CSV2>) {
if ($csv2->parse($each_csv2_line)) {
my #columns2 = $csv2->fields(); # Parse CSV and load into an array for each row.
my $packID = $columns2[0];
my $packDESC = $columns2[1];
my $val = 'customer_packages_report.txt';
chomp ($val);
my $cnt=0;
open (HNDL,"<","$val") or die "wrong filename: $val! $!";
while (<HNDL>){
$cnt++ while (/$packID - $packDESC/ig);
}
#if ($packDESC =~ /\(/g) {
# $packDESC =~ s/\(/\(/g;
#}
print "Total iterations of $packDESC: $cnt\n";
close (HNDL);
# End original code
} # Close IF
} # Close WHILE
# end of script
close CSV;
My recommendations:
Use $HNDL instead of HNDL <- lexical variables for filehandles more better.
Try to catch all mistakes (by defined and ==0 and eq "")
I try to format your code and add some features that i sometimes use. Be better than me and read first Style Coding for Little Perl Monk. And you can be more impressive with this language and write not only writeonly code.
Example (and also a quote):
"The situation is exactly the same for the line-input operator, <>, although Perl does this for you automatically.
It looks like you’re testing the line from STDIN in this while:
while (<STDIN>) {
do_something($_);
}
However, this is a special case in which Perl automatically converts to check $_ for definedness:
while ( defined( $_ = <STDIN> ) ) { # implicitly done
do_something($_);
}
"
Effective Perl Programming, page 24.

Answer 2

You could do a number of things to improve your code:
use warnings;.
Use proper indentation.
Use descriptive variable names. Instead of $file2 (has no meaning, and why is there no file 1?), use $package_file or whatever makes sense.
if you are already using Text::CSV, you can use $csv->getline() to go through the file line by line. This will simplify your code. See the documentation for an example.
chomp($val) removes a newline from the end of a string. You are using it on a string literal you just declared, which has no newline. That doesn't make sense.
Never use the same variable ($val) to do two completely different things. This is extremely confusing.
Might the variables that you are interpolating in the regex contain special characters? If so, you need to escape them. For example, if $packDESC contained a period, it would match any character in the regex. To treat the contents of the variable literally, use \Q..\E, as in this example: /\Q$packID - $packDESC\E/ig.
You are opening customer_packages_report.txt and going through it line-by-line on every line of the csv file. You could simplify this by reading it in once and storing the results in an array.
You don't need a while loop to count matches: $cnt = () = /$packID - $packDESC/ig;. This puts the match in array context, returning an array of matches, then puts it back in scalar context to count the matches. A little bit tricky, but simpler.
It's hard to say exactly what is causing your problem without seeing the data. Might you have some unnecessary repetition that stems from your nested looping over both files? I would start by rewriting to improve your code, then see if the problem still exists.

Answer 3

Your code seems to compile with perl -c without errors, so that's good. If I were to guess, I would assume your problem lies in having meta characters in some of your fields. The regex /$packID - $packDESC/ is vulnerable to meta characters. For example
my $str = "foo? bar";
$str =~ /$str/; # returns false, because ? is a meta character
In the above example, the question mark ? is a quantifier which affects whatever comes before it, so that o? means "0 or 1 o". To solve the meta character problem, use the \Q ... \E escape:
$str =~ /\Q$str/; # will now match
Terminating the escape sequence with \E is optional.
Some other things to note:
It is very good that you use use strict. You should also always use warnings. Not doing so is not removing the issues with your code, only hiding them.
You create a Text::CSV object with default settings. Depending on your input, that may or may not be appropriate. Setting binary => 1 is recommended in the documentation.
Using the parse() function may not be the best option, the documentation has good things to say about getline.
As loldop points out in the comments, you are reusing $val to read from your file. While technically that should work, it is asking for trouble.
Style and practice notes and practical tips:
Using three-argument open and lexical file handles is a good thing to do. Three-argument in essence means to use an explicit open mode, which makes your script safer to use. Using lexical file handles means that you will not have global scope on your file handle, which is a good thing.
This code
my #columns2 = $csv2->fields();
my $packID = $columns2[0];
my $packDESC = $columns2[1];
Can be written like this
my ($packID, $packDESC) = $csv2->fields();
You are chomping $val right after you assign it. That is redundant, because chomp by default only removes newlines from the end of your strings, and you did not add any such. It doesn't change anything, but not required here. If you read something from stdin or a file, you would probably want to use chomp, though.
Using die without referring to the error $! is a sure way to make yourself annoyed.
Do not underestimate how much easier it becomes to write code when you use proper indentation. Use a text editor with automatic indentation and colouring. I can warmly recommend vim (gvim if you are using windows). Though it has a learning curve, is is a powerful editor that also often comes already installed on many systems.

Answer 4

Since so many people have already commented on your program itself, I'm going to talk about how you can become a better Perl programmer, and help write in such a way that will help eliminate many of your issues.
Take a look at Perl::Tidy and run your program thorough that. That will help improve your syntax and Perl and will help you catch a lot of the various issues you're having.
Also, you should get a copy of Perl Best Practices which is where most of Perl Tidy is taken from. And, as someone already referenced Effective Perl Programming is another excellent book.
The big issue with Perl is that few people learn it. Most are tossed into a situation where we had to pick it up ourselves. Plus, Perl is a fairly old and rather crufty language. Most Perl books still lean heavily on Perl 3.x ways of programming and fail to mention such basics as using use strict; and use warnings;.
You combine old programming practices, with most people learning Perl by hacking their way through old programs with old syntax (and probably written by people who learned Perl by hacking their way through even older programs), and you can see why Perl has a reputation of being a write-only language.

Answer 5

You may want to use the getline method from Text::CSV, which saves a few lines of code.
The problem is likely to be because you have regex metacharacters in the strings you are searching for. Escape them with \Q...\E in the regex so that they are taken literally. In the rewrite below I have also added \s* instead of a literal space, just in case there isn't exactly one space on either side of the hyphen.
I have also changed the filehandles to lexical ones, which have the advantage that they will be closed automatically when the handle goes out of scope.
#!/usr/bin/perl
use strict;
use warnings;
use Text::CSV;
my $file2 = 'master_plist.csv';
my $csv2 = Text::CSV->new();
open(my $csv_fh, '<', $file2) or die $!;
while (my $row = $csv2->getline($csv_fh)) {
my ($packID, $packDESC) = #$row;
my $val = 'customer_packages_report.txt';
chomp($val);
open(my $fh, '<', $val) or die "wrong filename";
my $cnt = 0;
while ($val = <$fh>) {
while ($val =~ /\Q$packID\E\s*-\s*\Q$packDESC\E/ig) {
$cnt++;
}
}
print "Total iterations of $packDESC: $cnt\n";
}