PGP in lexer (was: [bogofilter-announce] Bogofilter-0.17.4 - new current release)

[Boris 'pi' Piwinger]

David Relson <relson at osagesoftware.com> wrote:

>A typical pgp signature block looks like:
>
>-----BEGIN PGP SIGNATURE-----
>Version: GnuPG v1.2.1 (GNU/Linux)
>
>iD8DBQBAHUWBaTUjGSdb2eIRAuYnAJ9NP7yUiGKL8Xp3cbqJSBVjPbrDjwCeMGZB
>SBUICjVBeMmFkmsljrgCpsI=
>=9z3S
>-----END PGP SIGNATURE-----
>
>Training with pgp signed messages adds tokens that are unlikely to every
>again be seen.  The new code recognizes this and ignors those tokens.

I just gave it a try: SBUICjVBeMmFkmsljrgCpsI is seen in
output of -vvv. What's wrong?

Also: Is it worth the complication? Will this just save some
tokens or does it boost accuracy?

pi