home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.theory
- Path: sparky!uunet!cs.utexas.edu!usc!news.service.uci.edu!ucivax!ucla-cs!ucla-ma!sonia!aps
- From: aps@sonia.math.ucla.edu (Alexei Stolboushkin)
- Subject: Re: solution or information needed
- Message-ID: <1993Jan24.042945.6448@math.ucla.edu>
- Sender: news@math.ucla.edu
- Organization: UCLA Mathematics Department
- References: <1993Jan19.151627.21525@matrix.com> <1993Jan21.153028.7679@cs.yale.edu> <1993Jan22.214830.26496@math.ucla.edu>
- Distribution: usa
- Date: Sun, 24 Jan 93 04:29:45 GMT
- Lines: 24
-
- In article <1993Jan22.214830.26496@math.ucla.edu> aps@sonia.math.ucla.edu (Alexei Stolboushkin) writes:
- >In article <1993Jan21.153028.7679@cs.yale.edu> fischer-michael@cs.yale.edu (Michael Fischer) writes:
- >
- >...
- >
- >>limitations of these devices. To see why {ww | w E {0,1}*} is not
- >>regular, for example, you need to know about the pumping lemma for
- >>regular sets and other techniques for proving non-regularity. To see
- >>why it is not accepted by a pushdown store machine, it is useful to
- > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- >certainly it IS context-free, and therefore accepted by a PDA.
-
- ooops, I thought it is ww' (' for reversion), sorry. Yes, you're right,
- for any CF language there is n such that every word in the language
- of length >n can be represented as xyzuv, with |yzu|<n, |yu|>0,
- and all words xy^k zu^k v are also in the language. So it seems
- enough to consider one word of the form:
- 0^m 1^m 0^m 1^m, for some m>n,
- and all four situations, when yzu consists of 1's only, of 0's only,
- of 1's, and 0's, and of 0's and 1's.
-
- A much more interesting exercise (and an instructive one), is to
- give an example of a non-CF language satisfying the above pumping
- lemma.
-