Re: Quasi-infinite recursion in Turkish stemmer
Martin Porter <[email protected]> Tue, 30 Aug 2022 18:14:00 +0100
| Newsgroups | gmane.comp.search.snowball |
|---|---|
| Message-ID | <CALc6duzvh9zF1gK7pYQRGgiF=22gBiTqnUyUaYXRHY82YmbYQA@mail.gmail.com> |
I really think this is not an issue. The old Porter stemmer, for example, run on pure binary data with the occasional hex 20 marking a word break, runs happily to completion, and similarly for the other stemmers. Suggesting a limit on input length, it seems to me, is to ring unnecessary alarm bells.