On Thu, 4 May 2000, Kurt Granroth wrote:

> David Faure wrote:
> > > I am indeed guilty :-)  I agree that reading it is slightly harder
> > > now.. but it makes for parsing by scripts a *LOT* easier so it's
> > > definitely needed.
> > 
> > Well [\t]+ isn't that difficult to write, is it ?
> 
> Just for fun, get the version of 'accounts' *before* the change and
> try to parse it into a file with accounts, names, and emails all
> separate.  Just try it :-)
> 
> What you'll see is what I did... the fact that the real names had
> variable number of spaces and words and *tabs* made the parsing a huge
> PITA.  It is possible, of course... I used a perl script to convert it
> to the new version.. but *boy* is it a pain.

What's wrong with:
"/^\s*(\S+)\s+(.*)\s+(\S+)\s*$/;" as perl regex. In $1 is account, $2 the
name (with space and everything) and in $3 the email. Regexp's rule ;)

> Now, all you need to do is delimit by tab and you have all three
> fields.

But it is so ugly ;)


Ciao,
Michael.