jakarta-regexp-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "James Buchanan" <jam...@northnet.com.au>
Subject Nested subexpressions
Date Wed, 28 Mar 2001 21:50:32 GMT
Hi guys!

I'm interested to know what the developers where thinking when they wrote
the code for handling nested subexpressions, e.g.
(abc*(foobar(foo)+))){3,5}. Did they go through some ideas, like having a
list of these "atomic" parts which cannot be broken up, like list node #1
having "ab", #2 having "c*", #3 having "foobar" and #4 having "foo+", and
then #5 saying "the previous 4 nodes having to be matched at least 3 times
but no more than 5." (or whatever). What kind of designs did they go
through, which ones were rejected for what reasons?

Is anyone interested in a perl-like s/regexp/replacement/args kind of search
and replace construct?

And almost off-topic, how did they search posix.org (which I think is
redirected to the IEEE site) for the POSIX docs on regexps? I'd like to grab
a copy and have a look.

James Buchanan

View raw message