I think this works:
\makeatletter
\AtEveryCitekey{%
\blx@langsetup\abx@field@hyphenation%
\blx@hyphenreset%
}
\makeatother
In \abx@field@hyphenation is hyphenation value. Probably it needs some testing, if it is set.
edit:
Some time ago, I was looking at biblatex language switching because of problem with my citation style[1]. I didn't understand this code at all, so I ended with ugly hack[2].
But when I looked yesterday on biblatex's patch of babel you posted, I have found line
\blx@langsetup\languagename\select@language
Function \blx@langsetup uses edef to define \blx@languagename, which is used by macro \blx@hyphenreset to load hyphenation patterns, and then loads localization strings for given language. So in fact, instead of
\blx@langsetup\abx@field@hyphenation%
it is possible to use just
\edef\blx@languagename{\abx@field@hyphenation}%
Then there is problem with French language. When used as the main document language, instead of
... (Fis-
el 1985)...
there is
...(FISCHEL
1985)...
I think there is issue only with French, I tried Czech, Russian and Spanish and they worked correctly.
With babel, we can solve this issue with
\select@language\abx@field@hyphenation%
but, polyglossia in xelatex has same issue and this trick is there not working, I don't know how to fix that.
Anyway, if you don't need French with polyglossia, this is the current solution
\makeatletter
\AtEveryCitekey{%
\ifcsdef{abx@field@hyphenation}{%
\edef\blx@languagename{\abx@field@hyphenation}%
\select@language\abx@field@hyphenation%
\blx@hyphenreset%
}{}%
}
\makeatother
[1] Biblatex - using two languages in one reference entry
[2] Biblatex - using two languages in one reference entry
It seems that the acronym package is putting stuff in front of the first mention of an acronym (some code writing to file to make a label; I didn't analyse it further) which in effect inhibits hyphenation of the first word of the acronym text.
The following change of the preamble will patch the faulty command so that hyphenation is re-established. Sorry for the strange patching pattern, but \patchcmd seems to always replace only the first occurrence (but there are two for \AC@acl).
\usepackage{etoolbox}
\makeatletter
\patchcmd\@acf{\AC@acl}{\AC@foo}{}{}
\patchcmd\@acf{\AC@acl}{\AC@foo}{}{}
\patchcmd\@acf{\AC@foo}{\hskip\z@\AC@acl}{}{}
\patchcmd\@acf{\AC@foo}{\hskip\z@\AC@acl}{}{}
\makeatother
\begin{document}
Please also consider egregs advice: You'll get better hyphenation of german text if you properly mark it as german.
]1
Best Answer
With
\textbf{"Aquivalenz\-umformungen}you define only one possible hyphenation. But the interword space will be too big if TeX tries to hyphenate at this point. UseThe possible stretching of the interwordspace is saved in
\fontdimen3\font:However, you should always use
\emergencystetchor thesloppyparenvironment!