[RESOLVED] Regex difficulties

scrupul0us

$pattern = '/(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?:\s\/\s|\s)?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})\s?(?:\[|\(|\{)\s?(\d{1,3})(?:\sips\s?)(?:\]|\)|\})\s?(?:\*)*\s?(?:OPEN)\s?(?:\*)*\s?\r/i';

$string = '192.168.1.0 / 255.255.255.0 [256 ips] *** OPEN ***';

for the LIFE of me I can't figure out why this doesn't work in PHP 5.0.4 (it's a 'legacy' system)

i'd been using: http://gskinner.com/RegExr/

...to help build my regexs for my dataset and it matches this string (of which there are a couple hundred), but, PHP won't match it with preg_match_all

rather confused

bradgrafelman

Worked fine for me after I removed the '\r' at the end of the pattern. Why do you have that there, anyway?

jazz_snob

Please explain what is the expected result is you are trying to get.

scrupul0us

@... I have a few other patterns that match that string with some other information coming after OPEN ... if I don't put anything at the end of the regex (line return, etc) it globally matches EVERY instance (what i want the other pattens to capture) and I get dupes in my output...

meanwhile, all those other patterns are based off the same base (as posted above, sans the \r) and work just fine...

idk it boggles my mind

@jazzsnob: expected results is that pattern to match properly and not greedily match anything else (any other data that may follow OPEN )

scrupul0us

so given:

192.168.1.0 / 255.255.255.0 [256 ips] *** OPEN ***
192.168.1.0 / 255.255.255.0 [256 ips] *** OPEN *** some kind of desc
192.168.1.0 / 255.255.255.0 [256 ips] *** OPEN *** [alert]
192.168.1.0 / 255.255.255.0 [256 ips] *** OPEN *** (notation of some sort)

what i want is a pattern that will match JUST that first line (capturing IP / MASK / # of ips) with the following in mind:

-spacing is variable
-the '/' between ip and mask is optional
-the brackets around the ips can be (), [], {}
-the num of * around OPEN is variable
-case-insensitive

... i already have the patterns to grab the other lines (which work perfectly), but if my pattern is:

$pattern = '/(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?:\s\/\s|\s)?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})\s?(?:\[|\(|\{)\s?(\d{1,3})(?:\sips\s?)(?:\]|\)|\})\s?(?:\*)*\s?(?:OPEN)\s?(?:\*)*/i';

which works on the website i linked in my first post... in code it matches ALL the conditions (too greedy somehow?) which causes 100's of dupes

bradgrafelman

Try this:

$pattern = '/(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?:\s\/\s|\s)?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})\s?(?:\[|\(|\{)\s?(\d{1,3})\sips\s?(?:\]|\)|\})\s?\**\s?OPEN\s?\**\s*$/i';

I cleaned up the end of the pattern a bit (why so many (?🙂 groups?!) and had it match the end of the line (with the possibility of whitespace, but nothing else, after * OPEN *).

scrupul0us

skips right over them 🙁

as for the large qty of non-returning groups e.g. (?: I'm not really well versed in regex yet so i was using them to define "groups" but didnt want the data to be returned

bradgrafelman

How about this:

$pattern = '/(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})(?:\s\/\s|\s)?(\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})\s?(?:\[|\(|\{)\s?(\d{1,3})\sips\s?(?:\]|\)|\})\s?\**\s?OPEN\s?\**\s*$/im';

Seems to work for me... I'd forgotten about the 'm' multi-line modifier making '$' match a new line character rather than end-of-string. My bad!

scrupul0us

yes that worked!!! now why did the multi-line parameter caus it to work?

bradgrafelman

In my patterns, I used a $ at the end of the pattern indicating that nothing else should be on that line.

Well, I meant "line".. but without the 'm' modifier, the $ only matches the end of the entire string.

scrupul0us

ahhh... i had tried to use $ myself which borked things up... good to know about using 'm'

thanks brad!