read a file twice...

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

Replies are listed 'Best First'.
Re: read a file twice... by rob_au (Abbot) on Apr 15, 2002 at 04:07 UTC
In addition to the approach involving reading the file into an array as suggested by Chmrr above, I would introduce you to the perlfunc:seek function, which allows you to move the current file read/write pointer within a file. With this command, you can move back to the start of the file and re-read it in full without closing and reopening the file. Eg. `open F, 'filename.ext' or die $!; . . seek F, 0, 0; # Move pointer back to start of file` [download] This approach is advantageous where the file may be particularly large in size and resources may be constrained.	[reply] [d/l]
Re: read a file twice... by Chmrr (Vicar) on Apr 15, 2002 at 03:58 UTC
You can "read" it as many times as you want if you stuff it into an array just once: `open F, "some_file" or die $!"; my @lines = <F>; close F; for (@lines) { # .. find the max and minimum } for (@lines) { # .. do something before / after those indexes }` [download] Update: Oops. s/\@f/\@lines/g; as MeowChow pointed out to me. perl -pe '"I lo`+$^X$\"$]!$/"=~m%(.)%s;$_=$1;y^`+*^e v^#$&V"+@( NO CARRIER'	[reply] [d/l]
Re: read a file twice... by CukiMnstr (Deacon) on Apr 15, 2002 at 06:24 UTC
besides the two approaches given above, you could give Tie::File by Dominus a try. It would allow you to manipulate a text file as a perl array (each file record, line per default, corresponding to an element of the array) without loading the whole file into memory, so it is specially useful for large files. Also useful when modifying files... hope this helps,	[reply]
Re: read a file twice... by thor (Priest) on Apr 15, 2002 at 12:26 UTC
In the spirit of TMTOWTDI, why not just keep track of the text along the way? If the number and the text are on the same line, easy money. Here's what I've got (I'm going to assume the format <text>:<number>) `while(<FILE>) { @array = split ':'; if (@array[1] > $max) #number we read in is greater than current max { $max = $array[1]; $max_text = $array[0]; } #similarly for min }` [download] Of course, your file layout may be more complex.	[reply] [d/l]
Re: read a file twice... by gdnew (Acolyte) on Apr 15, 2002 at 06:50 UTC
Hi.. thanks everyone...	[reply]
Re: read a file twice... by clintp (Curate) on Apr 15, 2002 at 15:56 UTC
In the spirit of TIMTOWDI and filed under "don't do this at home": `#!/usr/bin/perl -w push(@ARGV, "filetoread", "filetoread"); while(<>) { print; } continue { # This block not necessary, but for fun only print "Second verse, same as the first!\n" if (eof ARGV && @ARGV); }` [download]	[reply] [d/l]
Re: read a file twice... by dsb (Chaplain) on Apr 15, 2002 at 16:37 UTC
`$ perldoc -f seek` [download] For what it's worth, be sure to read how to use whence properly. I misread that part and I misused 'seek' for awhile. Amel - f.k.a. - kel	[reply] [d/l]
Re: read a file twice... by wardk (Deacon) on Apr 15, 2002 at 17:36 UTC
This sort of issue comes up alot in my shop with generating log files, typically for a scheduled "batch" job. the logfile for usability purposes needs the totals, stats, etc at the top of the log with any detail info after. To avoid processing twice (it's usually processing a SQL cursor, but occasionally a flat file too), we typically use the method mentioned by thor.... read the file once, saving the lines to be printed afterward in one or more arrays, then pushing them out at the end.	[reply]