4

I am using BibDesk to organise my bibliography which includes a paper with the author Caitríona.

BibDesk automatically converts the í to {\'\i}. However, the biber command translates {\'\i} to the dotless i (U+0131) and the accute accent (U+0301) which latex subsequently struggles with. In particular, it complains about the accent character:

Package inputenc Error: Unicode char ́ (U+301) not set up for use with LaTeX.

Manually changing {\'\i} to {\'i} makes biber translate to a single character (U+00ED) which latex is happy with.

Unfortunately, I cannot redefine the character translation that BibDesk uses internally. How do I work around this problem without having to manually replace {\'\i} by {\'i} before I can compile my document?

Till Hoffmann
  • 297
  • 1
    Related: https://tex.stackexchange.com/questions/345476/ I'm not sure why BibDesk is so stubborn in using \'\i. – egreg Jan 11 '18 at 18:28
  • Tell the BibDesk maintainers that they should change the conversion: \'i for “i with acute” has been in LaTeX for 20+ years. – egreg Jan 11 '18 at 18:32
  • https://sourceforge.net/p/bibdesk/bugs/1523/ – Till Hoffmann Jan 11 '18 at 18:38
  • There is also 1492, I left a comment there. – egreg Jan 11 '18 at 18:43
  • 1
    If you use JabRef instead, you won't have these problems. – Bernard Jan 11 '18 at 19:07
  • If you're using biber it means you're working with biblatex that uses unicode, so what is the need of struggle with the ascii codification such as in traditional bibtex when you can avoid it using the biblatex eases? You can follow the advice of @Bernard and use JabRef with support for biblatex. – Aradnix Jan 11 '18 at 22:31
  • @egreg BibDesk's character conversion is designed for ASCII BibTeX workflows, and it's been using {'\i} since the beginning (and both work fine for me with pdflatex). Is BibDesk wrong or is biber wrong here? If you're using biber, disable BibDesk's character conversion and save your file as UTF-8. – Adam Maxwell Jan 12 '18 at 04:59
  • @AdamMaxwell \'i may only worry plain TeX users. Why isn’t it customizable in the preferences? – egreg Jan 12 '18 at 09:16
  • Oddly, new character encodings can be added in the preferences, but existing ones can’t be modified. – Till Hoffmann Jan 12 '18 at 09:18
  • @egreg Because BibDesk has too many preferences already. In addition, working around other people's broken stuff isn't a high priority, and biber/biblatex is a never-ending stream of complaints. You can modify CharacterConversion.plist inside BibDesk.app if you want this, but you'll have to do it after every update. – Adam Maxwell Jan 12 '18 at 15:14
  • Related https://tex.stackexchange.com/q/251261/35864 – moewe Jan 12 '18 at 22:54

1 Answers1

5

Combining the answer from @egreg's comment with this suggestion for adding arguments to biber in latexmk, I created a .latexmkrc file in the project directory with the following content

$biber='biber --output-safechars %O %S';

which solved the problem.

egreg
  • 1,121,712
Till Hoffmann
  • 297
  • 2
    I've been using UTF-8 Unicode with biblatex/biber and the \usepackage[utf8]{inputenc} under latexmk and have had no problems with direct input of í. I use BibDesk to create the .bib file and enter these characters directly. This won't work with bibtex but biber is fine. – Herb Schulz Jan 11 '18 at 20:03