parsing a file

tcf03 has asked for the wisdom of the Perl Monks concerning the following question:

Replies are listed 'Best First'.
Re: parsing a file by Limbic~Region (Chancellor) on Jan 25, 2005 at 15:44 UTC
tcf03, Your question was a little confusing to me so please forgive me if this solution doesn't do what you want. You said that the file is in the format of not having the __BEGIN__/__END__ blocks, so presumably you want to add them. You also want to process each users information before adding them. Something like the following should work `#!/usr/bin/perl use strict; use warnings; { local $/ = ""; # paragraph mode while ( <DATA> ) { chomp; process_user( $_ ); print "__BEGIN__\n$_\n__END__\n"; } } sub process_user { my @lines = split /\n/, shift(); for ( @lines ) { # do something with line; print "$_\n"; } } __DATA__ USER 1 LOGIN date MORE info MORE info USER 2 LOGIN date MORE info MORE info USER3 LOGIN data MORE info MORE info` [download] Cheers - L~R	[reply] [d/l]
Re: parsing a file by friedo (Prior) on Jan 25, 2005 at 15:36 UTC
You could use a regex to grab each set of everything between __BEGIN__ and __END__. `# untested my @users = ( $file =~ /__BEGIN__\n(.*?)\n__END__/gs );` [download] But if I may make a reccomendation, for dealing with this kind of data, I find YAML to be extremely useful and eady.	[reply] [d/l]
Re: parsing a file by holli (Abbot) on Jan 25, 2005 at 16:37 UTC
Limbic`s code is well written, but i take a different approach. Assumptions: a) USER, LOGIN and MORE is a kind of markup. b) You want to process the data somehow before writing it back. c) You are capable to write file in/output by yourself. That leads to the following code: #!/usr/bin/perl use strict; use warnings; { local $/ = ""; # paragraph mode while ( <DATA> ) { chomp; print process_user($_); } } sub process_user { my ($user, $login, @info) = split ("\n", shift); $user =~ s/USER //; $login =~ s/LOGIN //; @info = map { s/^MORE //; $_ } @info; #sample processing: add a star everywhere return join "\n", "__BEGIN__", "USER $user", "LOGIN $login", (map { "MORE $_" } @info), "__END__\n"; } __DATA__ USER 1 LOGIN date MORE info MORE info USER 2 LOGIN date MORE info MORE info [download] That produces this output: `__BEGIN__ USER 1 LOGIN date MORE info MORE info __END__ __BEGIN__ USER 2 LOGIN date MORE info MORE info __END__` [download] Update:* Corrected minor typo and Limbic`s "s". holli, regexed monk	[reply] [d/l] [select]
Re: parsing a file by dragonchild (Archbishop) on Jan 25, 2005 at 15:34 UTC
Have your parsing code call a function that does the processing once it finally has all the data for user N. When building your parsing code, have your processing function print out the Data::Dumper output of the input to the function. You might want to look at the flip-flop operator within nested while-loops. Check out This code is just freaky... Being right, does not endow the right to be rude; politeness costs nothing. Being unknowing, is not the same as being stupid. Expressing a contrary opinion, whether to the individual or the group, is more often a sign of deeper thought than of cantankerous belligerence. Do not mistake your goals as the only goals; your opinion as the only opinion; your confidence as correctness. Saying you know better is not the same as explaining you know better.	[reply]
Re: parsing a file by Elijah (Hermit) on Jan 25, 2005 at 18:26 UTC
Unless you really want to for some other reason, there is really no need to add the __BEGIN__ and __END__'s. Here is an example of processing each block of info without adding these. `#!/usr/bin/perl -w use strict; my $block; while (<DATA>) { if (!/^\n/) { $block .= $_; }else{ process($block); undef($block); } } process($block) if ($block); sub process { my $info = shift; print "\n",$info if ($info); } __DATA__ USER 1 LOGIN date MORE info MORE info USER 2 LOGIN date MORE info MORE info` [download] www.perlskripts.com	[reply] [d/l]
Re: parsing a file by Random_Walk (Prior) on Jan 25, 2005 at 17:57 UTC
I am rather late to this party but assuming your records are all seperated by a single blank line and there are no superfluous blank lines then this rather obfu looking one liner will get you there `perl -pe'BEGIN{print"__BEGIN__\n"}s/^\s$/__BEGIN__\n__END__\n/;END{pr +int"__END__\n"}' <file>` [download] Or the mildly shorter but fiercly uglier `perl -pe'$e="__END__\n";BEGIN{$b="__BEGIN__\n";print$b}s/^\s$/$b$e/;E +ND{print$e}' <file>` [download] Cheers, R. Pereant, qui ante nos nostra dixerunt!	[reply] [d/l] [select]
Re^2: parsing a file by Limbic~Region (Chancellor) on Jan 25, 2005 at 18:21 UTC
Random_Walk, I believe in addition to adding the __BEGIN__/__END__ tags, tcf03 also wanted to "process" each user's information. The trouble I had was in figuring out what was meant by "process", so I just added a sub that could be changed according to suite one's needs. Cheers - L~R	[reply]
Re: parsing a file by borisz (Canon) on Jan 25, 2005 at 16:39 UTC
What about this: `local $/ = "\n\n"; @users = <DATA>; __DATA__ USER 1 LOGIN date MORE info MORE info USER 2 LOGIN date MORE info MORE info USER3 LOGIN data MORE info MORE info` [download] Boris	[reply] [d/l]
Re^2: parsing a file by holli (Abbot) on Jan 25, 2005 at 16:45 UTC
what should it profit the OP to get this `$VAR1 = [ 'USER 1 LOGIN date MORE info MORE info ', 'USER 2 LOGIN date MORE info MORE info ', 'USER3 LOGIN data MORE info MORE info ', ' ' ];` [download] structure? holli, regexed monk	[reply] [d/l]
Re^3: parsing a file by borisz (Canon) on Jan 25, 2005 at 17:03 UTC
The data is split into peaces that start with every user. As the op requested. Boris	[reply]
Re: parsing a file by TedPride (Priest) on Jan 25, 2005 at 17:00 UTC
`use strict; my $in = 'sample.txt'; my $out = 'sampleout.txt'; my ($handle, $key, $val); open($handle, $in); # Messy, but I'm assuming a small file... split(/\n\n/, join('', <$handle>)); close($handle); open($handle, ">$out"); for (@_) { print $handle "__BEGIN__\n$_\n__END__\n"; my %hash; for (split(/\n/)) { ($key, $val) = split(/ /); $hash{$key} = $val; } # Do whatever with record data, now inside # %hash as key / value pairs } close($handle);` [download]	[reply] [d/l]