Don't understand behavior of this split

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

Howdy Bros. I am trying to split some html paragraphs into a list and keep only the text. So say the html is

<p>Paragraph 1</p>
<p>Paragraph 2</p>
<p>Paragraph 3</p>
[download]

I try to split it with the regex (<\/p>)?\s*<p> and what I get is a list with elements:

(null)
Paragraph 1
</p>
Paragraph 2
</p>
Paragraph 3</p>
[download]

I understand why the last element has the close paragraph tag in there, but why are there elements between the paragraph ones with a close paragraph tag? I thought split was supposed to split on whatever it matches, and I know it matches the </p><p> sequence.

TIA

Steve

Comment on Don't understand behavior of this split Select or Download Code

Replies are listed 'Best First'.
Re: Don't understand behavior of this split by moritz (Cardinal) on Aug 26, 2010 at 14:25 UTC
If the regex has a capturing group (and yours has, the `(\/p>)`), then the result of that capture (here $1) is interleaved with the split chunks. To avoid that, use the regex `(?:<\/p>)?\s*<p>` instead. See split and perlre for more details. Perl 6 - links to (nearly) everything that is Perl 6.	[reply] [d/l] [select]
Re: Don't understand behavior of this split by ikegami (Patriarch) on Aug 26, 2010 at 14:31 UTC
You're trying to 1) separate the paragraphs, then 2) extract the data from those paragraphs. `split` could be use for the former, but it won't be sufficient. This will do both: `my @matches = m{<p>\s(.?)\s*</p>}sg;` [download]	[reply] [d/l] [select]
Re: Don't understand behavior of this split by kennethk (Abbot) on Aug 26, 2010 at 14:26 UTC
As documented in split: If the PATTERN contains parentheses, additional list elements are created from each matching substring in the delimiter. You can get your expected behavior by changing to a non-capturing group: `(?:<\/p>)?\s*<p>`	[reply] [d/l]
Re: Don't understand behavior of this split by TomDLux (Vicar) on Aug 26, 2010 at 14:28 UTC
From `perldoc -f split`: `If the PATTERN contains parentheses, additional list elements are created from each matching substring in the delimiter. split(/([,-])/, "1-10,20", 3); produces the list value (1, '-', 10, ',', 20)` [download] Try using the non-capturing parentheses`(?:)` update: Boy you gotta type fast around here! update: changes from 'pre' to 'code', cause square brackets were being misinterpreted As Occam said: Entia non sunt multiplicanda praeter necessitatem.	[reply] [d/l] [select]
Re: Don't understand behavior of this split by Corion (Patriarch) on Aug 26, 2010 at 14:26 UTC
See split. You don't show us the code you're using, but `split` returns captured items in the regex as well. Instead of splitting, maybe matching what you want to keep works better for you?	[reply] [d/l]
Re: Don't understand behavior of this split by psini (Deacon) on Aug 26, 2010 at 14:28 UTC
From perlfunc for split operator: " If the PATTERN contains parentheses, additional list elements are created from each matching substring in the delimiter." So what you get is the expected behaviour Rule One: "Do not act incautiously when confronting a little bald wrinkly smiling man."	[reply]