[Date Prev][Date Next][Thread Prev][Thread Next]
[Search]
[Date Index]
[Thread Index]
[MacPerl] Columns
Hallo,
Some people asked me last week to summerize the
answers given to my columns question. I promissed
to to do so, so here it is.
First, this was my question:
>Hallo,
>
>I have a (text) file with the following format:
>
>123420010110010010101 (and so on this way => )
>234531111101100101010
>678940000010010111010
>340151010101110101010
>and so on...
>
>Now let's say the first four characters are a PIN number,
>the following two characters the subject's age, and so on.
>How can I read such a file split in (fixed) columns ?
>I would like to insert tabs at the end of each field, creating
>a tab-delimited text file.
To all who replied, of course (as someone said yesterday)
'Thanks for the lightbulb over the head'
Here are the answers given (in random order)
=============================================================
I saw most of the answers and I believe that the fastest way
to do it is:
while (<>) {
$pin = substr ($_, 0, 6);
$age = substr ($_, 6, 2);
etc...
}
=============================================================
This is my solution:
I assume you are reading the text file through a filehandle: <F>
while (<F>) {
$_ =~ s/(\d{4})(\d{2})(\d{15})/$1\t$2\t$3/;
$fHolder .= $_;
}
The regular expression groups each line into three sets of 4, 2 and 15
digits respectively and then reconstructs the line with the tab character
between grouped sets. Each line thus processed is appended to the variable
$fHolder for further processing (or for writing out to disk).
=============================================================
Try this (untested) code:
##---------------------------- start sample
open(INPUT, "filename") || die;
open(OUTPUT, ">filewithtabs") || die;
while (<INPUT>) {
## Try to replace 4 digits and 2 digits at the beginning of a line
## with the matched 4 digits, tab, the matched 2 digits, tab.
if (s/^(\d{4})(\d\d)/$1\t$2\t/) { ## line 6 (see below)
print OUTPUT;
} else {
## The match failed, there were not 6 digits at the
## beginning of the input line
print "warning: line of unexpected format ($_)\n";
}
}
close(INPUT);
close(OUTPUT);
##---------------------------- end sample
If you are certain your input data are correct, more minimal code could
replace the "if" code at line 6 with:
s/^(.{4})(..)/$1\t$2\t/;
print;
If you actually want to get values out to manipulate in your program,
change the "if" code starting at line 6 to:
if ( ($pin, $age, $the_rest) = /^(\d{4})(\d\d)(.*)/ ) {
## Now you can work with $age, $pin, or $the_rest
## Example: $max_age = $age if ($age > $max_age);
} else {
## The match failed.
print "warning: line of unexpected format ($_)\n";
## If you prefer, print bad lines to some error file
## instead of the screen--good for large data sets.
}
=============================================================
Try something along these lines:
$template = "a4 a2 a2";
($pin, $age, $so-on) = unpack($template, $_);
The "a4 a2 ..." is defining alpha fields for the unpack instruction. So
it'll take the first four chars and make it a field, the next 2 and make it
a field, etc. It's quite handy, and with Perl's loose type checking, quite
flexible.
templates can do a whole lot more, if you want, and are discusses in the
programming perl books.
=============================================================
while(<>)
{
print join("\t",unpack("a4a2...",$_));
}
where the unpack format describes the fixed format fields...
=============================================================
See the Llama book Ch. 17 under Fixed-length Random Access Databases.
Try something like this:
my ($pin, $age) = unpack ("A4 A2", <>);
print ($pin, "\t", $age, "\n");
If you just want to insert tabs (and not process the data) then try:
print (join ("\t", unpack ("A4 A2", <>)), "\n");
Don't use regexps because this is faster and easier to understand.
=============================================================
Somthing like
print "$1\t$2" if /^(\d{4})(\d\d)/;
=============================================================
$s = '123420010110010010101';
$s =~ /^(\d{4})(\d{2})(\d)(\d{14})/; #read the regular expressions
$pin = $1; #man pages for more info
$age = $2;
$oth = $3;
$misc = $4;
print "$pin\t$age\t$oth\t$misc"; #"\t" = tab
Or, on the whole file:
open(FILE,$filename) || die "$!\n";
open(NEWFILE,">$newfilename") || die "$!\n";
while($s = <FILE>) {
$s = '123420010110010010101';
$s =~ /^(\d{4})(\d{2})(\d)(\d{14})/;
$pin = $1;
$age = $2;
$oth = $3;
$misc = $4;
print NEWFILE "$pin\t$age\t$oth\t$misc";
}
close(NEWFILE);
close(FILE);
=============================================================
You probaly want to look at unpack() for reading the fixed width fields in
a while(<FILEHANDLE>) loop and some variant of 'print join()' to write them
back out.
=============================================================
If I had to do this, I'd use a regular expression. Reading in the file
line by line, I'd do something like
($pin,$age,....) = /(\d{4})(\d{2})..../;
and if I wanted to tab delimit the results and print them out, simply
print "$pin\t$age\t....\n";
=============================================================
The alternative to using a 'regex' is to use 'substr'. It is worth try:
$pin = substr($_, 0, 4);
$age = substr($_, 4, 2);
You could of course do the thing in one hit:
$new = substr($_, 0, 4)."\t".
substr($_, 4, 2)."\t".
substr($_, 6, 2)."\t".
etc
=============================================================
open(INFILE, "/in/file/name") or die("no in: $!, stopped");
open(OUTFILE, ">/out/file/name") or die("no out: $!, stopped");
while (<INFILE>) {
# the numbers following the "a"s indicate field width
my($pin, $age, $etc) = unpack("a4a2a99", $_);
print OUTFILE "$pin\t$age\t$etc\t\n";
}
According to the Camel, 2nd ed., p. 66, using unpack() is more efficient
than patterns like /(...)(..)(.....)/ .
=============================================================
You'll probably get lots of responses telling you to use "PACK", which
is certainly more efficient. Here's how to do it if you want to
understand what you're doing :*)
Assuming the input file is open for read on INFILE and the output file is
open for write on OUTFILE:
$INFILE="TheFileToRead";
$OUTFILE=">TheFileToWrite"; #the ">" says we're allowed to write to the file
open INFILE | die "Can't open $INFILE to read.\n";
open OUTFILE | die "Can't open $OUTFILE to write.\n";
while (<INFILE>) {
#everything inside this WHILE will automatically use each
#each successive line of INFILE unless we tell it otherwise
#...which we won't
@TheData = split(""); #split the line into single character parts
$ThePIN=join("",@TheData[1..4]); join the first four into a string
$TheAge=join("",@TheData[5..6]);
$TheBlah=join("",@TheData[7.blah]);
etc etc
#finally, print the outfile.
print OUTFILE "$ThePIN\t$TheAge\t$TheBlah\tetc\n";
}
close OUTFILE;
close INFILE;
=============================================================
If I were doing this, I'd read in the line of text, and use
substr() a few times to pick off the individual items and push
them into an array. Like this:
while(<>) {
@array = ();
push (@array, substr($_, 0, 4));
push (@array, substr($_, 4, 4));
[...etc...]
print join("\t", @array), "\n";
}
=============================================================
# Nico Rozendaal, #
# University of Maastricht, #
# Dept. of Psychiatry & Neuropsychology, #
# Maastricht, The Netherlands. #
# e-mail: nico.rozendaal@np.unimaas.nl #