Aryeh M. Friedman wrote:
I am attempting to make (without the perl expansions) a regular expansion that when used as a delim will split words on any punction/whitespace character *EXCEPT* "$" (for java people I want to feed it into something like this:


for(String foo:input.split([insert regex here])
   ...

Well, there's no way to say "all foo except bar" using standard regexes, so
you can't use the [:punct:] character class. You'll have to roll your own
class.

If your input is ASCII then see ispunct(3) for a handy list of all the
ascii punctuation characters.  I guess you'll need a RE something like this:

  []!"#%&'\(\)\*\+,\./:;<=>?...@[\\^_`{\|}~-[:space:]]+

although that's completely untried, quite likely to not have all the
metacharacters properly escaped (exactly what is or isn't a metacharacter
depends on the RE implementation you're using) and is probably horribly
confused due to the inclusion of '[' '-' and ']' amongst the characters
matched in the range.
If you're using anything other than ascii, then I suspect you're going
to have problems with RE libs anyhow, unless you can somehow use PCRE. The \p{isPunct} and \p{isWhite} escapes for matching unicode punctuation
or whitespace is probably what you need.

Even so, your best choice would probably be to separately check strings
for the presence of $ characters -- maybe transform those $ characters to
something else -- and then split on any remaining punctuation characters.

        Cheers,

        Matthew

--
Dr Matthew J Seaman MA, D.Phil.                   7 Priory Courtyard
                                                 Flat 3
PGP: http://www.infracaninophile.co.uk/pgpkey     Ramsgate
                                                 Kent, CT11 9PW

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to