Recently, I had the same problem. BibTeX generated wrong characters or
errors, due to bad encoding. It was complaining that it can not
understand \h{o} as a control sequence! Why? I thought it is a general
way to get accented letters.

I have switched in the .bib file from LateX encoding to UTF8 and then to
my local ISO8859-2 encoding. The last switch solved the problem. LyX
file has UTF8 encoding.

I have just read the other e-mails. When I finish with the current,
writing I will also try biber.


Stefano Franchi írta:
> Problem solved,  but I am even more confused than before.
> I did install biber and tried it out (after a rather unpleasant time spent 
> fighting with perl dependencies..). The problem remained.
> As Richard guessed, there was a problem with the encoding of the LyX file. 
> Switching to Unicode UTF* in the Document>>Settings>>Language pane solved all 
> the problems. After the switch to UTF8, the references were formatted 
> correctly, both by standard bibtex and by biber.
> I am happy but confused. UTF8 (for bib files) should not be supported by 
> bibtex and Lyx (since it just calls bibtex). Yet it works. I am wondering if 
> there is some magic going on behind the scenes. 
> Cheers,
> S. 
> On Friday 09 October 2009 07:35:32 am rgheck wrote:
>> On 10/09/2009 01:02 AM, Stefano Franchi wrote:
>>> Thanks Richard,
>>>     I think I understand better now. I suppose I'll have to try out biber
>>> (I have switched to BibLatex already) already and check the encoding of
>>> the LyX file.
>> Let me know how biber goes. If it works well, we'll add it to the list
>> of BibTeX options. In 2.0/1.7, or whatever it is going to be, you can
>> select which bib-file processor you want to use. At the moment, it's
>> bibtex and bibtex8, with a "custom" option as well.
>> rh
>>> Cheers,
>>> Stefano
>>> On Thursday 08 October 2009 05:09:42 pm you wrote:
>>>> On 10/08/2009 05:55 PM, Stefano Franchi wrote:
>>>>> Dear all,
>>>>> is there any special trick to using a .bib file encoded as UTF-8 with
>>>>> LyX/Latex? If there is, I would certainly appreciate knowing about
>>>>> it...
>>>>> Here is my problem/use-case:
>>>>> I use Jabref as my bib files editor. After inserting a reference which
>>>>> contained the character Ž (Latin capital Z with caron), Jabref
>>>>> suggested I switched the encoding to UTF-8 because the current one
>>>>> (8859-1, I suppose), did not contain the requested character. I
>>>>> accepted the kind offer. Alas, now all the references containing
>>>>> diacritic marks are screwed up in the Latex output. for instance 
>>>>> Schöpfungs- has become Schöpfungs- and so on.
>>>>> Help or pointers to approriate documentation greatly appreciated. I am
>>>>> sure the day will come when I will master these  encoding issues.
>>>>> Unfortunately I do not seem to be there yet.
>>>> It sounds like there may be a couple issues here.
>>>> First, unless I'm mistaken, standard BibTeX simply does not support
>>>> UTF-8, or any other sort of mutli-byte encoding. Perhaps this has
>>>> changed, but Philip Lehman wrote just a year and a bit ago: "In contrast
>>>> to what a lot of users think, it is not and has never been possible to
>>>> use UTF-8 in bib files. Neither traditional Bibtex nor Bibtex8 support
>>>> multibyte encodings such as UTF-8. If it seems to work with some files,
>>>> it only does so by chance." And because BibLaTeX also relies upon BibTeX
>>>> for some of its work, this won't change even there.
>>>> There's also this project, though:
>>>> Basically, it replaces BibTeX
>>>> with a program written in Perl. I haven't tried it, but it should be
>>>> possible to use it with LyX if you're also using BibLaTeX. It may even
>>>> work without it. I don't know.
>>>> The weirdness with Schöpfungsis probably due to a conflict between the
>>>> encoding of your document and the encoding of the .bib file. E.g, the
>>>> document isn't in UTF-8.
>>>> Richard

Reply via email to