simple regexp problem

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

Replies are listed 'Best First'.
Re: simple regexp problem by davidrw (Prior) on Dec 05, 2005 at 17:09 UTC
what about just using `perldoc -f split`? `($part1, $part2) = split(/_/, $str); ($part1, $part2) = split(/_/, $str1); # or possibly better depending on context: @parts = split(/_/, $str);` [download]	[reply] [d/l] [select]
Re^2: simple regexp problem by pKai (Priest) on Dec 05, 2005 at 18:04 UTC
I think you have to add a limit parameter to split here, to not lose additional "underscore delimited fields" `D:\temp>perl -e "$x='a_b_c'; ($j, $k) = split /_/, $x; print $k" b D:\temp>perl -e "$x='a_b_c'; ($j, $k) = split /_/, $x, 2; print $k" b_c` [download]	[reply] [d/l]
Re^3: simple regexp problem by davidrw (Prior) on Dec 05, 2005 at 18:21 UTC
true, but is totally dependent upon OP's requirements .. should "a_b_c" result in part1=a, part2=b and c getting lost, or should "a_b_c" result in part1=a and part2=b_c ?	[reply]
Re: simple regexp problem by choedebeck (Beadle) on Dec 05, 2005 at 17:08 UTC
how about `my ($part1, $part2) = split /_/, $str;` [download]	[reply] [d/l]
Re^2: simple regexp problem by Celada (Monk) on Dec 05, 2005 at 17:41 UTC
Sometimes you don't want to use split because it's too lenient! Your example will be happy to accept `AAAA_BBBB_CCCC`. The original poster probably wants this string to be rejected, but split will eat `AAAA`, `BBBB`, and discard `CCCC`.	[reply]
Re^3: simple regexp problem by jeffa (Bishop) on Dec 05, 2005 at 18:07 UTC
Which is exactly why the original poster should be capturing the output into an array instead of multiple sequentially numbered scalars. I think split is a very good and acceptable solution for this problem. jeffa L-LL-L--L-LL-L--L-LL-L-- -R--R-RR-R--R-RR-R--R-RR B--B--B--B--B--B--B--B-- H---H---H---H---H---H--- (the triplet paradiddle with high-hat)	[reply]
Re^3: simple regexp problem by Tanktalus (Canon) on Dec 05, 2005 at 19:09 UTC
In that case, split still works: `my ($part1, $part2) = split /_/, $str, 2;` [download] That extra parameter asks for no more than two items - more _'s will not be even looked at.	[reply] [d/l]
Re: simple regexp problem by ikegami (Patriarch) on Dec 05, 2005 at 17:05 UTC
Square brackets don't mean what you think they mean. Use `(?:...)` to group things without capturing, and use the `?` suffix to make something optional. `/^(\w+)(?:_(\w+))?$/` [download]	[reply] [d/l] [select]
Re^2: simple regexp problem by ikegami (Patriarch) on Dec 05, 2005 at 17:46 UTC
Ah shoot! While the advice in the above post is sound, the code won't work because "`\w`" includes "`_`"! Fix: `/^([a-zA-Z0-9+])(?:_(\w+))?$/` [download] or `/^((?:(?!_)\w)+)(?:_(\w+))?$/` [download]	[reply] [d/l] [select]
Re^3: simple regexp problem by jeanluca (Deacon) on Dec 06, 2005 at 13:36 UTC
This works too: `/^(\w+?)(?:_(\w+))?$/` But why the last ? ? If I leave it off, it doesn't parse the string "AAAAAAA" Thanks Luca	[reply] [d/l]
Re^4: simple regexp problem by ikegami (Patriarch) on Dec 06, 2005 at 15:31 UTC