On Friday 21 September 2007 09:44, Joe Smith wrote:
> Mark Phillips wrote:
> > I am using Dcount() to count the occurrences of some strings in a
> > spreadsheet in OpenOffice Calc 2.0.3. I am using a regular expression to
> > find the string I want.
> >
>  > ...
> >
> > ... is there a way to modify my regular expression to handle this case.
> >
> > BTW, I tried using [:alpha:] instead of [a-zA-Z], and dcount() did not
> > find any matches. Is [:alpha:] not working - it is listed in the
> > documentation as a valid regular expression.
>
> I'm not sure I follow what it is you're trying to count--what a 'match'
> should be.
Sorry, - I hoped the example test cases were sufficient. Basically, each cell 
has a list of "terms", and I want to count how many of each one. 
1. A term can be 1 or more words with spaces in between
2. If there are multiple terms in the cell, then each term has a single quote 
around it, and all the terms are enclosed in brackets "[" and "]".
3. If there is only one term in the cell, then there will only be the one 
term - no brackets or single quotes

>
> If you just want to count cells that have the string "Board Member", then
>
> .*Board Member.* or .*'Board Member'.*

Yup, that is what I am trying to do. You greatly simplified the solution - 
thanks!
>
> should be adequate.
>
> If there's something more complicated and you need to stay with your
> current approach, try removing the single quotes from the character
> classes. All that does is include the single quote as part of the class,
> and I think that's where it fails in your example.
>
> \[[a-zA-Z, ]*'Board Member'(, )?[a-zA-Z]*(, )?\]|Board Member
>
> The named classes, e.g. [:alpha:] have some nasty problems. Usually they
> can be tweaked to work, but sometimes not. Unless you anticipate needing to
> scan non-Western text, they aren't particularly useful.
Hmmm...wish that were documented somewhere. Thanks!
>
> If you're still stuck, post some verbatim examples of your data, and
> indicate which ones should match and what portion of each should match, and
> I'm sure we can find something that will work.

Thanks, your solution is working!
>
> <Joe
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]

-- 
Mark Phillips
[EMAIL PROTECTED]
602 524-0376

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to