Another tough one. Pls... help

[josephgigi]

I am trying to construct a pattern that extracts all the dates in a file that appear in a sentence (a sentence can end in by . or ; or ,). These dates comes after one of the words, cutoff, deadline or extend which must be preceeded by the word "Discovery".

There can be any word or letters between "Discovery" and "cutoff". Simiarly there can be any word or letters between the "cutoff" and Date. These dates can appear multiple times in the a file. The content of the searched file (it is an HTML file) is read into one string.And the string is to be searched for the pattern and all the occurances of the pattern should be returned.

Here is the pattern I am working on now. And it does not work correct as it should be. It does not find the multiple occurances and some time it does not work at all.

"((discovery)((extend)|(extension)|((cut)(\s)(off))|((dead)(\s)(line))))((\w|\s))*((\d{2})|(\d))\/((\d{2})|(\d))\/((\d{4})|(\d{2}))"

Any help is greatly appreciated.

Thanks
-Gigi JK

[Anon]

Sorry wish I could help. Not familiar with the regexps, but I should probably look into those soon

[Anon]

preg_match_all("%
(?:^|(?<=[.,;])) # The end of the previous sentence
\s # Maybe some whitespace
discovery # Discovery
\s # space
.? # stuff
(extend|extension|cut\soff|dead\sline) # keyword
\s # space
.*? # more stuff
(\d{1,2}/\d{1,2}/\d{2,4}) # A date.
[.,;] # end of sentence
%six", $text, $matches,PREG_SET_ORDER);

"six": Match "\n" as a whitespace character; case-independent matching;
ignore whitespace and #comments in pattern.

$matches[n][1] will contain the keyword for the nth matching sentence,
$matches[n][2] will contain the date. $matches[n][0] will match the
entire sentence. To swap array indices so that $matches[1][n] is the
keyword for the nth matching sentence, either remove ",PREG_SET_ORDER"
or replace it with ",PREG_PATTERN_ORDER".