[Trennmuster] Hyphenation patterns
Werner LEMBERG
wl at gnu.org
Mi Mär 14 18:20:00 CET 2012
> I propose '|' for suffixes (Vorsilben), z.B.
... ah, another category :-)
> [about `.' vs. `_']
>
> I am not sure, whether we need (or can do) this conceptual
> distinction: Both, low line '_' and full stop '.' indicate an
> "undesired" hyphenation point (i.e. a hyphenation that point that
> should not be used but nevertheless marked as existent to make clear
> we know it exists).
This should be discussed, indeed.
> The full stop '.' is the "legacy" character, used in words that have
> non-categorized marks '·'.
AFAIK, this is not correct; `.' is not restricted to non-categorized
words. For example, we have
An-den.ken
> Some of the full stops can be converted to "normal" hyphenation
> marks when hyphenation points are marked according to classes:
>
> Pres-se=in-for-ma-tio-nen
>
> or (when also marking the suffix "in") to '|' or '='
>
> Pres-se=in|for-ma-tio-nen Pres-se==in=for-ma-tio-nen
>
> A special marker for the "undesired" hyphenation point after "in" is
> no longer required, as '=' separates parts of a compound word and we
> can suppress hyphenations "near" '=' by a rule.
[For non-wortliste users: a hyphenation point is `near' if the
distance is smaller than three characters.]
As you correctly write, some of the full stops could be avoided.
However, in some cases `=.' could be helpful for clarity.
> The to-be-suppressed hyphenation point in An·den.ken or the famous
> Ur·in.stink, on the other hand, might still need a special mark.
>
> Proposals are Ur|in_stinkt or Ur=in-.stinkt.
As mentioned in other mails `.' is essentially a shorthand for `-.'.
> [...] correct automatic hyphenation of words like
>
> Nord-S-Bahn (Nord- S-Bahn) vs.
> Schluss-S-Schreibung (Schluss-S -Schreibung)
>
> is not possible with the German hyphenation patterns.
Hmm, good examples – better than the Dutch ones :-) But patgen can't
handle this. What about libhyphen?
Werner
Mehr Informationen über die Mailingliste Trennmuster