Re: Help to split

[davido]

in reply to Help to split

I prefer 2teez's solution; it's the way to do it. But you specified that it should be done with one or more calls to split.

It's usually a better practice when asking for help here to explain what you've tried, and how it's failing, but not to dictate what tools a proper solution should employ. You're likely to get more elegant solutions if you don't limit people such. Nevertheless, if one or more calls to split is the solution you would prefer, you could do something like this:

use strict;
use warnings;
use feature qw( say );

# Warning: Big slurp! ;)

say for map { ( split /\s/, $_, 2 )[0] } # 3: split on whitespace.
        map { ( split /:/,  $_, 5 )[4] } # 2: split on colons.
        <DATA>;                          # 1: auto-split on newlines.




__DATA__
bsulli03:*:32452:5002:barry sullivan,l230,555-6666,:/students/bsulli03
+:/usr/bin/ksh
sbarto22:*:32453:7990:sally bartok,l134,444-3333,:/students/sbarto22:/
+usr/bin/ksh
sbarto25:*:32453:7990:sparky bartok,l135,444-4444,:/students/sbarto25:
+/usr/bin/ksh
sbarto35:*:32453:7990:sammy bartok,l137,444-5555,:/students/sbarto35:/
+usr/bin/ksh
sdemar03:*:32454:7990:sally demartra,S23,233-5234,:/students/sdemar03:
+/usr/bin/ksh
sdemar05:*:32454:7990:bill demartra,l123,555-1234,:/students/sdemar05:
+/usr/bin/ksh
singl04:*:32455:5002:bart singleton,l134,555-1235,:/students/singl:/us
+r/bin/ksh
gomez:*:32456:7990:bill gomez,l433,535-6666,:/students/gomez:/usr/bin/
+ksh
bolde01:*:32457:7990:bart bolden,l444,444-1234,:/students/bolde:/usr/b
+in/ksh
reaga09:*:32458:7990:cal reagan,l455,333-1234,:/students/reaga:/usr/bi
+n/ksh
liu44:*:32460:5002:tom liu,s213,222-1234,:/students/liu:/usr/bin/ksh
ulano:*:32462:7990:sally ulanouskay,s233,222-9999,:/students/ulano:/us
+r/bin/ksh
tran56:*:32463:5002:phuong tran,s111,444-8888,:/students/tran:/usr/bin
+/ksh
mehr:*:32464:5003:cindy mehr,c113,555-8888,:/students/mehr:/usr/bin/ks
+h
ramos:*:32449:7990:olivo ramos,l454,999-0000,:/students/ramos:/usr/bin
+/ksh
brown:*:32459:7990:bart bro,l477,888-7654,:/students/brown:/usr/bin/ks
+h
mojic:*:32461:7990:celia mojica,c242,555-7654,:/students/mojic:/usr/bi
+n/ksh
fisher:*:2915:208:cindy fisher,l312,478-0371,:/users/fisher:/usr/bin/k
+sh
bsulli02:*:32452:5002:bart sullivan,l130,555-5555,:/students/bsulli02:
+/usr/bin/ksh

[Athanasius]

Or you can dispense with map by chaining the calls to split:

#! perl
use strict;
use warnings;
print +(split / /, (split /,/, (split /:/)[4])[0])[0], "\n" while <DAT
+A>;

Or as a one-liner:

16:05 >perl -pwe "$_ = (split / /, (split /,/, (split /:/)[4])[0])[0] 
+. qq[\n];" passwd
barry
sally
sparky
sammy
sally
bill
bart
bill
bart
cal
tom
sally
phuong
cindy
olivo
bart
celia
cindy
bart

16:05 >

:-)

Hope that helps,

[semenych]

Thanks Athanasius

[Athanasius]

You’re welcome!

Here is the same solution broken into separate steps, and with comments added:

#! perl
use strict;
use warnings;

my $file = 'passwd';

open(my $fh, '<', $file) or die "Cannot open file '$file' for reading:
+ $!";

# The while loop reads one line of the data file at a time by implicit
+ly
# splitting on newlines. So, on the first iteration, $line =
# bsulli03:*:32452:5002:barry sullivan,l230,555-6666,:/students/bsulli
+03:/usr/bin/ksh

while (my $line = <$fh>)
{
    chomp $line;    # Remove the trailing newline

    # Split the line on the colon character. On the first loop iterati
+on,
    # @fields = ('bsulli03',                        # $fields[0]
    #            '*',                               # $fields[1]
    #            '32452',                           # $fields[2]
    #            '5002',                            # $fields[3]
    #            'barry sullivan,l230,555-6666,',   # $fields[4]
    #            '/students/bsulli03',              # $fields[5]
    #            '/usr/bin/ksh');                   # $fields[6]
    # (Note that array indexes start at zero.)

    my @fields = split /:/, $line;

    # Now split the fifth field (index 4) on the comma character. On t
+he first
    # loop iteration, @names = ('barry sullivan', 'l230', '555-6666');

    my @names  = split /,/, $fields[4];

    # Now split the first field (index 0) on the space character. On t
+he first
    # loop iteration, @first = ('barry', 'sullivan');

    my @first  = split / /, $names [0];

    # Print the first name, which on the first loop iteration is 'barr
+y'

    print "$first[0]\n";
}

close $fh or die "Cannot close file '$file': $!";

Hope that helps,

[semenych]

Thanks, Dave. But unfortunately I have not learnt map yet. Could you please advice to do it the easiest way, probably, just using one or more split calls. Thanks again

[semenych]

Thanks Dave for your advice. But unfortunately I have not learnt map yet. As I understand I am supposed to do it with while or for loops and using one or more split calls. Could you please help to write it in the easiest way. Thanks again

[dsheroh]

It's usually a better practice when asking for help here to explain what you've tried, and how it's failing, but not to dictate what tools a proper solution should employ.

He may not have explicitly stated it, but this is clearly a homework assignment. The "solve by using split one or more times" requirement is part of the assignment. Even if he knew other ways to do it, the instructor has declared them to be out of bounds.

Beyond that, he's also provided his source code in the question, which qualifies as "explain[ing] what you've tried" in my book. The only thing missing from the question is a statement of how it fails, although I expect that was obvious to most experienced monks pretty quickly (as CountZero said, he forgot to read from the file after opening it).

[grizzley]

Pliiiiz, don't use regexps in your split. They are scary and OP doesn't want to be scared :)

[Anonymous Monk]

Pliiiiz, don't use regexps in your split. They are scary and OP doesn't want to be scared :)
by grizzley

Eeeep! A talking grizzley!