tilde
David Relson
relson at osagesoftware.com
Thu Dec 11 15:11:45 CET 2003
On Thu, 11 Dec 2003 14:54:22 +0100
Boris 'pi' Piwinger <3.14 at logic.univie.ac.at> wrote:
> Boris 'pi' Piwinger wrote:
>
> >>> I believe the reason are TOKENs like F~R~E~E. That makes
> >>> sense, but I don't have a strong opinion here.
> >>
> >> Unfortunately there are also "F*R*E*E", "F R E E", etc.
> >
> > Right, we allow some of the and some not.
>
> While we are at it. Right now we have:
> TOKENFRONT
> [^[:blank:][:cntrl:][:digit:][:punct:]]
> TOKENMID
> [^[:blank:][:cntrl:]<>;=():&%$#@+|/~\\{}^\"?*,\[\]]+
> TOKENBACK
> [^[:blank:][:cntrl:]<>;=():&%$#@+|/~\\{}^\"?*,\[\]._\'\`\-]
>
> So TOKENBACK is TOKENFRONT plus digits and exclamation point.
>
> TOKENMID is TOKENBACK plus "._'`-". There is certainly
> justification for them, well ` would be a typo in a word,
> but it is very common. So I think we can live with that. But
> one could also argue that _ does not show up in "real
> words", . only in abbreviations like "i.e.". But looking at
> it ~ would not really fit, so it seems reasonable to take it
> out.
>
> pi
Works for me. tilde will be allowed in the middle of tokens, but not at
the end.
More information about the bogofilter-dev
mailing list