[Python-Dev] string.find() again (was Re: timsort for jython)

Andrew Koenig ark@research.att.com
07 Aug 2002 16:14:22 -0400

>> Note the distinction between the empty language and the empty
>> string.  As a language is a set of strings, the empty language is
>> one that contains no strings, not even the empty string.
>> Therefore, a regular expression that accepts the empty language is
>> one that rejects every string, even the empty string.

Tim> Sure, that's why I said "empty language" and not "empty string".
Tim> It wouldn't make *any* sense for "re1 in re2" to consider a
Tim> regexp that accepted the language {""} to be "in" all other
Tim> regexps.  But a regexp that accepts the language {} (i.e., the
Tim> empty language) clearly accepts a subset of the language accepted
Tim> by any regexp.

Right.  (I wasn't disagreeing with you, merely pointing out a
plausible miscomprehension on the part of the reader (because
I made just that mistake the first time I read it))

>> Pedantically y'rs    --ark

Tim> Not enough to matter in this case <wink>.

Whether it matters depends on whether the reader made the same
mistake I did on first reading.

Andrew Koenig, ark@research.att.com, http://www.research.att.com/info/ark