Working with Text Files

[rdove]

My task requires me to write all my information to a text file. Now the next thing they want is for me to get certain information from the text file based on critera. Kind of like SQL in a database only I must use a text file.

Anyone have any great linksor advice that deals with searching through information in a text file with perl?

Thanks

[gwartheg]

Originally posted by rdove
My task requires me to write all my information to a text file. Now the next thing they want is for me to get certain information from the text file based on critera. Kind of like SQL in a database only I must use a text file.

Anyone have any great linksor advice that deals with searching through information in a text file with perl?

Thanks

Code Examples may be of some help. I like Perl Monks, too. In the mean time, is this what you're trying to do? Or is it more complicated?

open (IN, "file") or die "Can't read file: $!\n";
while ( <IN>) {
chomp;
my @fields = split;
    if ($fields[0] =~ /PATTERN/) {
        # do something
    }
}
close(IN);

Here's how you can simply get a list of all lines containing a certain pattern:

open (IN, "file") or die "Can't read file: $!\n";
my @lines = <IN>;
my @stuff = grep { /PATTERN/ } @lines;
print @stuff;
close(IN);

[rdove]

Well..I'm not sure...That part of the project has sort of been put on hold for a while. Right now we are trying to get the page up and the transform part working. In the next couple of weeks I will be able to give you a better idea if thats what I'm looking for.

[rdove]

I took a look at your code. And umm.... could you explain what each line does. I know nothing about perl (which is partly why this is hard for me) and your code is beyond the little bit I do know. I understnad if you don't want to explain it all, but is there a good site out there I can visit and read up on perl and text files?

[gwartheg]

No, I don't mind at all. Shame on me for throwing up code without explaining it :o

open (IN, "file") or die "Can't read file: $!\n";  # 1
while ( <IN>) {                                   # 2
    chomp;                                         # 3
    my @fields = split;                            # 4
    if ($fields[0] =~ /PATTERN/) {                 # 5
        # do something
    }
}
close(IN);                                         # 6

# 1 - Here I'm using the "open" function to attach the file handle IN to the external file "file". If it's unsuccessful, the script dies and gives an error message.
# 2 - I'm using the diamond operator (<>) with the file handle to read each line in the file. Used with "while" it means, "execute this block as long as there is a line to be read from IN". Each line gets set to the variable $_ in turn.
# 3 - Remove any trailing newline.
# 4 - Each line gets split up into a list (split uses the pattern /\s+/ to split by default). I regularly use files with tab-delimited fields, so this is convenient for me.
# 5 - This says, "execute this block if the first item in the list looks like this pattern".
# 6 - This just explicitly closes the file I was looking in.

my @lines = <IN>;                      # 1
my @stuff = grep { /PATTERN/ } @lines; # 2
print @stuff;                          # 3

You could use these lines in place of the while loop, if you wanted a list of entire lines for which a given expression is true (usually a pattern match).
# 1 - Read all lines from IN into a list, where the first element is the first line, and so on.
# 2 - This says, "if an element in the list @lines contains PATTERN, add it to the list @stuff". It's just quicker than writing a for loop with a conditional and all that.
# 3 - This just spits out whatever is in @stuff. Notice that the input wasn't chomped, and that the @stuff isn't surrounded with quotes. I'm assuming that the input file has one record on each line, and I don't want the print function to put extra spaces between the list elements.

I know there are some file access tutorials online, but I've found that they're generally limited in scope and can't compete with a good book at hand. Oh, and coderforums.net is another good forum to visit. There are plenty of people there that know a ton more about Perl than me