«

Hello,
Assume I have following string:
my $cmds = <<DOC
__begin {
abc;
def;
{foo;bar}
} __end;
__begin {
cde;
} __end;
abc;
bad;
DOC
;

I want to split it into an array, the first item is "__begin {
abc;
def;
{foo;bar}
} __end", the second item is "__begin {
cde;
} __end", and the third is "abc" and the fourth is "bad".

split obviously cannot be used here, so I use following regex:
my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);

but it does not work at all. so how can I do with this?

regards,
bingfeng

bingfeng wrote:
>
> Assume I have following string:
> my $cmds = <<DOC
> __begin {
> abc;
> def;
> {foo;bar}
> } __end;
> __begin {
> cde;
> } __end;
> abc;
> bad;
> DOC
> ;
>
> I want to split it into an array, the first item is "__begin {
> abc;
> def;
> {foo;bar}
> } __end", the second item is "__begin {
> cde;
> } __end", and the third is "abc" and the fourth is "bad".
>
> split obviously cannot be used here, so I use following regex:
> my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);
>
> but it does not work at all. so how can I do with this?

my @lines = $cmds =~ /^\s*__begin(?s:.*?)__end;$|^\s*\S+;$/mg;



John
--
Perl isn't a toolbox, but a small machine shop where you
can special-order certain sorts of tools at low cost and
in short order. -- Larry Wall

bingfeng schreef:

> Assume I have following string:
> my $cmds = <<DOC
> __begin {
> abc;
> def;
> {foo;bar}
> } __end;
> __begin {
> cde;
> } __end;
> abc;
> bad;
> DOC
> ;
>
> I want to split it into an array, the first item is "__begin {
> abc;
> def;
> {foo;bar}
> } __end", the second item is "__begin {
> cde;
> } __end", and the third is "abc" and the fourth is "bad".
>
> split obviously cannot be used here, so I use following regex:
> my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);
>
> but it does not work at all. so how can I do with this?

my @blocks;

for my $re ( qr/__begin.*?__end/s, qr/^[^;]+/m ) {
while ($cmds =~ s/\s*($re)\s*;/" "x ($+[0] - $-[0])/es) {
push @blocks, [ $-[0], $1 ];
}
}

for (sort { $a->[0] <=> $b->[0] } @blocks) {
print "<--\n", $_->[1], "\n-->\n";
}


In the end, $cmds will still have the same length, but will contain only
whitespace (and any unmatched content).

--
Affijn, Ruud

"Gewoon is een tijger."

bingfeng <> wrote:
>Hello,
>Assume I have following string:
>my $cmds = <<DOC
> __begin {
> abc;
> def;
> {foo;bar}
> } __end;
> __begin {
> cde;
> } __end;
> abc;
> bad;
>DOC
>;
>
>I want to split it into an array, the first item is "__begin {
> abc;
> def;
> {foo;bar}
> } __end", the second item is "__begin {
> cde;
> } __end", and the third is "abc" and the fourth is "bad".

This sounds suspiciously like an X-Y problem to me. Are you reading this
text from a file? If yes, then if a block starts with '__begin{' you can
read that block until the '__end;' token is reached.

And yes, you can use split() if you do it in two steps:
First split on '__end;' . And in the second step repair the now missing
'__end;' for those items, that have a leading '__begin{' and for the
others split again at ';'.

jue

On 10ÔÂ20ÈÕ, ÏÂÎç7ʱ36·Ö, "John W. Krahn" <some...@example.com> wrote:
> bingfeng wrote:
>
> > Assume I have following string:
> > my $cmds = <<DOC
> > __begin {
> > abc;
> > def;
> > {foo;bar}
> > } __end;
> > __begin {
> > cde;
> > } __end;
> > abc;
> > bad;
> > DOC
> > ;
>
> > I want to split it into an array, the first item is "__begin {
> > abc;
> > def;
> > {foo;bar}
> > } __end", the second item is "__begin {
> > cde;
> > } __end", and the third is "abc" and the fourth is "bad".
>
> > split obviously cannot be used here, so I use following regex:
> > my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);
>
> > but it does not work at all. so how can I do with this?
>
> my @lines = $cmds =~ /^\s*__begin(?s:.*?)__end;$|^\s*\S+;$/mg;
>
> John
> --
> Perl isn't a toolbox, but a small machine shop where you
> can special-order certain sorts of tools at low cost and
> in short order. -- Larry Wall- Òþ²Ø±»ÒýÓÃÎÄ×Ö -
>
> - ÏÔʾÒýÓõÄÎÄ×Ö -

Thank you, John, it works very well. You help save some hours!

On Mon, 20 Oct 2008 02:42:03 -0700 (PDT), bingfeng <> wrote:

>Hello,
>Assume I have following string:
>my $cmds = <<DOC
> __begin {
> abc;
> def;
> {foo;bar}
> } __end;
> __begin {
> cde;
> } __end;
> abc;
> bad;
>DOC
>;
>
>I want to split it into an array, the first item is "__begin {
> abc;
> def;
> {foo;bar}
> } __end", the second item is "__begin {
> cde;
> } __end", and the third is "abc" and the fourth is "bad".
>
>split obviously cannot be used here, so I use following regex:
>my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);
^^
my @lines = ($cmds =~ /__begin.*?__end|[^\s;]+/sg);

You were on the right track. [^;] however is first to match all before ';',
which means it grabs the' __begin { .. abc;' then the next, then next.
'__begin.*?__end' is never matched. By including not whitespace, [^\s;] in
the character class, begin and end have a chance.

sln

--------------------

use strict;
use warnings;

my $cmds = <<DOC
__begin {
abc;
def;
{foo;bar}
} __end;
__begin {
cde;
} __end;
abc;
bad;
DOC
;

my @lines = ($cmds =~ /__begin.*?__end|[^\s;]+/sg);

for (my $i = 0; $i < @lines; $i++) {
print "\n\$lines[$i] = \n\n\"$lines[$i]\"\n";
}

__END__

output:

$lines[0] =

"__begin {
abc;
def;
{foo;bar}
} __end"

$lines[1] =

"__begin {
cde;
} __end"

$lines[2] =

"abc"

$lines[3] =

"bad"

On Oct 21, 1:27*am, s...@netherlands.com wrote:
> On Mon, 20 Oct 2008 02:42:03 -0700 (PDT), bingfeng <bfz...@gmail.com> wrote:
> >Hello,
> >Assume I have following string:
> >my $cmds = <<DOC
> > *__begin {
> > * * abc;
> > * * def;
> > * * {foo;bar}
> > *} __end;
> > *__begin {
> > * * cde;
> > *} __end;
> > *abc;
> > *bad;
> >DOC
> >;
>
> >I want to split it into an array, the first item is "__begin {
> > * * abc;
> > * * def;
> > * * {foo;bar}
> > *} __end", the second item *is *"__begin {
> > * * cde;
> > *} __end", and the third is "abc" and the fourth is "bad".
>
> >split obviously cannot be used here, so I use following regex:
> >my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);
>
> * * * * * * * * * * * * * * * * * * * * *^^
> my @lines = ($cmds =~ /__begin.*?__end|[^\s;]+/sg);
>
> You were on the right track. [^;] however is first to match all before ';',
> which means it grabs the' * __begin { .. abc;' then the next, then next..
> '__begin.*?__end' is never matched. By including not whitespace, [^\s;] in
> the character class, begin and end have a chance.
>
You are right. Thanks for your explanation. My sample is some
oversimple. the standalone sentence may contain other word and space,
with following test message:
my $cmds = <<DOC
__begin {
abc sss;
def;
{foo;bar}
} __end;
__begin {
cde;
} __end;
abc kkk;
bad fde;
DOC
;

you solution gives following Dumper result:
$VAR1 = '__begin {
abc sss;
def;
{foo;bar}
} __end';
$VAR2 = '__begin {
cde;
} __end';
$VAR3 = 'abc';
$VAR4 = 'kkk';
$VAR5 = 'bad';
$VAR6 = 'fde';

that's not what I want. Apart from John's solution, I have no other
solution. Thank you

> sln
>
> --------------------
>
> use strict;
> use warnings;
>
> my $cmds = <<DOC
> * __begin {
> * * *abc;
> * * *def;
> * * *{foo;bar}
> * } __end;
> * __begin {
> * * *cde;
> * } __end;
> * abc;
> * bad;
> DOC
> ;
>
> my @lines = ($cmds =~ /__begin.*?__end|[^\s;]+/sg);
>
> for (my $i = 0; $i < @lines; $i++) {
> * * * * print "\n\$lines[$i] = \n\n\"$lines[$i]\"\n";
>
> }
>
> __END__
>
> output:
>
> $lines[0] =
>
> "__begin {
> * * *abc;
> * * *def;
> * * *{foo;bar}
> * } __end"
>
> $lines[1] =
>
> "__begin {
> * * *cde;
> * } __end"
>
> $lines[2] =
>
> "abc"
>
> $lines[3] =
>
> "bad"

On Mon, 20 Oct 2008 20:59:35 -0700 (PDT), bingfeng <> wrote:

>On Oct 21, 1:27*am, s...@netherlands.com wrote:
>> On Mon, 20 Oct 2008 02:42:03 -0700 (PDT), bingfeng <bfz...@gmail.com> wrote:
>> >Hello,
>> >Assume I have following string:
>> >my $cmds = <<DOC
>> > *__begin {
>> > * * abc;
>> > * * def;
>> > * * {foo;bar}
>> > *} __end;
>> > *__begin {
>> > * * cde;
>> > *} __end;
>> > *abc;
>> > *bad;
>> >DOC
>> >;
>>
>> >I want to split it into an array, the first item is "__begin {
>> > * * abc;
>> > * * def;
>> > * * {foo;bar}
>> > *} __end", the second item *is *"__begin {
>> > * * cde;
>> > *} __end", and the third is "abc" and the fourth is "bad".
>>
>> >split obviously cannot be used here, so I use following regex:
>> >my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);
>>
>> * * * * * * * * * * * * * * * * * * * * *^^
>> my @lines = ($cmds =~ /__begin.*?__end|[^\s;]+/sg);
>>
>> You were on the right track. [^;] however is first to match all before ';',
>> which means it grabs the' * __begin { .. abc;' then the next, then next.
>> '__begin.*?__end' is never matched. By including not whitespace, [^\s;] in
>> the character class, begin and end have a chance.
>>
>You are right. Thanks for your explanation. My sample is some
>oversimple. the standalone sentence may contain other word and space,
>with following test message:
>my $cmds = <<DOC
> __begin {
> abc sss;
> def;
> {foo;bar}
> } __end;
> __begin {
> cde;
> } __end;
> abc kkk;
> bad fde;
>DOC
>;
>
>you solution gives following Dumper result:
>$VAR1 = '__begin {
> abc sss;
> def;
> {foo;bar}
> } __end';
>$VAR2 = '__begin {
> cde;
> } __end';
>$VAR3 = 'abc';
>$VAR4 = 'kkk';
>$VAR5 = 'bad';
>$VAR6 = 'fde';
>

Thats too bad. You made a good attempt and I gave
you credit by saying you almost had it right the first time.
And the regex was altered slightly from why you yourself tried.

I didn't write a regex for you. Because if I did that, you could
always come back and say for example:

#>You are right. Thanks for your explanation. My sample is some
#>oversimple. the standalone sentence may contain other word and space,
#>with following test message ...

But you didn't say that in the first place.

>that's not what I want. Apart from John's solution, I have no other
>solution.
>
^^^^^^^^^^^^^
Think again ... you just invalidated his regex.

my @lines = $str =~ /^\s*__begin(?s:.*?)__end;$|^\s*\S+;$/mg;

$lines[0] =
" __begin {
abc sss;
def;
{foo;bar}
} __end;"

$lines[1] =
" __begin {
cde;
} __end;"

What are you going to do now?
We're still in the extremely simple stage.
In fact, the more you add, the simpler it gets.

sln

-------------------------------

Version 2

#################
# Misc Parse 2
#################

use strict;
use warnings;

# the old
my $cmd1 = <<DOC1
__begin {
abc;
def;
{foo;bar}
} __end;
__begin {
cde;
} __end;
abc;
bad;
DOC1
;

# the new
my $cmds2 = <<DOC2
__begin {
abc sss;
def;
{foo;bar}
} __end;
__begin {
cde;
} __end;
abc kkk;
bad fde;
DOC2
;

my $str = $cmds2;

my @lines = ($str =~ /\s*(__begin.*?__end|.*?);/sg);

for (my $i = 0; $i < @lines; $i++) {
print "\n\$lines[$i] = \n\n\"$lines[$i]\"\n";
}

__END__

output:

$lines[0] =

"__begin {
abc sss;
def;
{foo;bar}
} __end"

$lines[1] =

"__begin {
cde;
} __end"

$lines[2] =

"abc kkk"

$lines[3] =

"bad fde"

On Oct 21, 1:46*pm, s...@netherlands.com wrote:
> On Mon, 20 Oct 2008 20:59:35 -0700 (PDT), bingfeng <bfz...@gmail.com> wrote:
> >On Oct 21, 1:27*am, s...@netherlands.com wrote:
> >> On Mon, 20 Oct 2008 02:42:03 -0700 (PDT), bingfeng <bfz...@gmail.com> wrote:
> >> >Hello,
> >> >Assume I have following string:
> >> >my $cmds = <<DOC
> >> > *__begin {
> >> > * * abc;
> >> > * * def;
> >> > * * {foo;bar}
> >> > *} __end;
> >> > *__begin {
> >> > * * cde;
> >> > *} __end;
> >> > *abc;
> >> > *bad;
> >> >DOC
> >> >;
>
> >> >I want to split it into an array, the first item is "__begin {
> >> > * * abc;
> >> > * * def;
> >> > * * {foo;bar}
> >> > *} __end", the second item *is *"__begin {
> >> > * * cde;
> >> > *} __end", and the third is "abc" and the fourth is "bad".
>
> >> >split obviously cannot be used here, so I use following regex:
> >> >my @lines = ($cmds =~ /__begin.*?__end|[^;]+/sg);
>
> >> * * * * * * * * * * * * * * * * * * * * *^^
> >> my @lines = ($cmds =~ /__begin.*?__end|[^\s;]+/sg);
>
> >> You were on the right track. [^;] however is first to match all before';',
> >> which means it grabs the' * __begin { .. abc;' then the next, then next.
> >> '__begin.*?__end' is never matched. By including not whitespace, [^\s;] in
> >> the character class, begin and end have a chance.
>
> >You are right. Thanks for your explanation. My sample is some
> >oversimple. the standalone sentence may contain other word and space,
> >with following test message:
> >my $cmds = <<DOC
> > *__begin {
> > * * abc sss;
> > * * def;
> > * * {foo;bar}
> > *} __end;
> > *__begin {
> > * * cde;
> > *} __end;
> > *abc kkk;
> > *bad fde;
> >DOC
> >;
>
> >you solution gives following Dumper result:
> >$VAR1 = '__begin {
> > * * abc sss;
> > * * def;
> > * * {foo;bar}
> > *} __end';
> >$VAR2 = '__begin {
> > * * cde;
> > *} __end';
> >$VAR3 = 'abc';
> >$VAR4 = 'kkk';
> >$VAR5 = 'bad';
> >$VAR6 = 'fde';
>
> Thats too bad. You made a good attempt and I gave
> you credit by saying you almost had it right the first time.
> And the regex was altered slightly from why you yourself tried.
>
> I didn't write a regex for you. Because if I did that, you could
> always come back and say for example:
>
> * #>You are right. Thanks for your explanation. My sample is some
> * #>oversimple. the standalone sentence may contain other word and space,
> * #>with following test message ...
>
> But you didn't say that in the first place.
>
> >that's not what I want. Apart from John's solution, I have no other
> >solution.
>
> * * ^^^^^^^^^^^^^
> Think again ... you just invalidated his regex.
>
> my @lines = $str =~ /^\s*__begin(?s:.*?)__end;$|^\s*\S+;$/mg;
>
> $lines[0] =
> " *__begin {
> * * *abc sss;
> * * *def;
> * * *{foo;bar}
> * } __end;"
>
> $lines[1] =
> " *__begin {
> * * *cde;
> * } __end;"
>
> What are you going to do now?
> We're still in the extremely simple stage.
> In fact, the more you add, the simpler it gets.
>
> sln
>
> -------------------------------
>
> Version 2
>
> #################
> # Misc Parse 2
> #################
>
> use strict;
> use warnings;
>
> # the old
> my $cmd1 = <<DOC1
> * __begin {
> * * *abc;
> * * *def;
> * * *{foo;bar}
> * } __end;
> * __begin {
> * * *cde;
> * } __end;
> * abc;
> * bad;
> DOC1
> ;
>
> # the new
> my $cmds2 = <<DOC2
> * __begin {
> * * *abc sss;
> * * *def;
> * * *{foo;bar}
> * } __end;
> * __begin {
> * * *cde;
> * } __end;
> * abc kkk;
> * bad fde;
> DOC2
> ;
>
> my $str = $cmds2;
>
> my @lines = ($str =~ /\s*(__begin.*?__end|.*?);/sg);
>
> for (my $i = 0; $i < @lines; $i++) {
> * * * * print "\n\$lines[$i] = \n\n\"$lines[$i]\"\n";
>
> }
>
> __END__
>
> output:
>
> $lines[0] =
>
> "__begin {
> * * *abc sss;
> * * *def;
> * * *{foo;bar}
> * } __end"
>
> $lines[1] =
>
> "__begin {
> * * *cde;
> * } __end"
>
> $lines[2] =
>
> "abc kkk"
>
> $lines[3] =
>
> "bad fde"

Wow, I had to admit your regex is simpler, easy to understand and
elegant. I'll study what you said carefully. Anyway, thank you very
much.