Accents in .bib file with biblatex (worked with natbib)

Question

I'm trying to switch from natbib to biblatex. I have trouble with the character {\`y} within an author field in my .bib file. It worked with natbib fine. The error I get is "Package inputenc: Unicode character ỳ (U+1EF3) (inputenc) not set up for use with LaTeX."

MWE:

main.tex

\documentclass[a4paper,11pt]{report}
\usepackage[utf8]{inputenc}
\usepackage[style=authoryear,natbib=true]{biblatex}
\addbibresource{main.bib}
\begin{document}
\parencite{mutny2019efficient}
\end{document}

main.bib

@article{mutny2019efficient,
  title={Efficient high dimensional bayesian optimization with additivity and quadrature fourier features},
  author={Mutn{\`y}, Mojm{\'\i}r and Krause, Andreas},
  journal={Advances in Neural Information Processing Systems 31},
  pages={9005--9016},
  year={2019}
}

It 's probably a good idea if you provide a code example and example bib file that shows that problem. Then others have something to test at their end — daleif, Jun 21 '21 at 14:20
Aren't you using the wrong accent? The actual proceeding (https://proceedings.neurips.cc/paper/2018/file/4e5046fc8d6a97d18a5f54beaed54dea-Paper.pdf) lists that name as Mojmír Mutný, that is both are \'y and \'\i not \`y — daleif, Jun 21 '21 at 14:37
Ah, thank you, not sure how that happened... Have a couple other similar errors, hopefully those can be fixed similarly. — Oxonon, Jun 21 '21 at 16:05
Also an issue has been registered (by the latex team), to get several more of the exotic accented letter supported (some might only be used in latinfication of Russian) — daleif, Jun 21 '21 at 16:56

score 1 · Accepted Answer · answered Jun 21 '21 at 16:17

Biber transforms LaTeX-ASCII-escapes like {\`y}, {\"a}, {\ss} into the corresponding Unicode character (ỳ, ä, ß) in order to be able to sort them properly.

That means that Mutn{\`y}, Mojm{\'\i}r is transformed into

Mutnỳ, Mojmı́r

With pdfTeX, LaTeX has issues with both ỳ and ı́.

It doesn't like ỳ, because it does not have the character pre-defined in its list of supported Unicode characters.

This could be resolved by adding
```
\DeclareUnicodeCharacter{1EF3}{\`y}
```
to your document preamble to let LaTeX know about ỳ.
It also doesn't like ı́. Here the issue is more subtle. The ı́ generated by Biber is not actually a single codepoint, it is a combination of a dotless i followed by the acute accent. With pdfTeX LaTeX does not support combining accents, so this breaks as well. Here the correct solution is to either use the single-codepoint glyph í in the input or {\'i} instead of {\'\i}.

See also How to put an acute on an i using Biber: issues with "\'{\i}", Input encoding error after upgrading from Biber 1.9 to Biber 2.1, Unicode -(U+301) error in biblatex, but not in main text: {\'{\i}}.

Taking both of the solutions into account, the following will compile

\documentclass[a4paper,11pt]{article}
\usepackage[T1]{fontenc}
\usepackage[utf8]{inputenc}
\usepackage[style=authoryear]{biblatex}
\DeclareUnicodeCharacter{1EF3}{`y}
\begin{filecontents}{\jobname.bib}
@article{mutny2019efficient,
  title   = {Efficient High Dimensional {Bayesian} Optimization
             With Additivity and Quadrature {Fourier} Features},
  author  = {Mutn{`y}, Mojm{'i}r and Krause, Andreas},
  journal = {Advances in Neural Information Processing Systems 31},
  pages   = {9005--9016},
  year    = {2019}
}
\end{filecontents}
\addbibresource{\jobname.bib}
\begin{document}
\parencite{mutny2019efficient}
\printbibliography
\end{document}

But as daleif points out in the comments, a closer look at the paper in question (available via https://proceedings.neurips.cc/paper/2018/hash/4e5046fc8d6a97d18a5f54beaed54dea-Abstract.html) shows that {\`y} is the wrong accent. You want ý, not ỳ.

So the following would be a better .bib entry for this particular paper.

\documentclass[a4paper,11pt]{article}
\usepackage[T1]{fontenc}
\usepackage[utf8]{inputenc}
\usepackage[style=authoryear]{biblatex}
\begin{filecontents}{\jobname.bib}
@inproceedings{mutny2019efficient,
  title          = {Efficient High Dimensional {Bayesian} Optimization
                    With Additivity and Quadrature {Fourier} Features},
  author         = {Mutný, Mojmír and Krause, Andreas},
  maintitle      = {Advances in Neural Information Processing Systems},
  booktitleaddon = {Proceedings of the 32nd Conference
                    on Neural Information Processing Systems
                    (NeurIPS 2018)},
  venue          = {Montréal, Canada},
  volume         = {31},
  editor         = {S. Bengio and H. Wallach and H. Larochelle and K. Grauman
                    and N. Cesa-Bianchi and R. Garnett},
  pages          = {9005--9016},
  year           = {2019},
}
\end{filecontents}
\addbibresource{\jobname.bib}
\begin{document}
\parencite{mutny2019efficient}
\printbibliography
\end{document}

Wasn't the í thing already solved? I only had issues in the mwe with \`y not the other one — daleif, Jun 21 '21 at 16:58
@daleif When I run the first example with {\'\i} instead of the preferable {\'i} pdfTeX complains Unicode character ́ (U+0301) not set up for use with LaTeX. on my machine. I think it was decided that Biber would continue translating {\'\i} to dotless i+combining accent because making an exception to the encoding rules would make things too messy. In the MWE from the question the first name gives no trouble because it is never printed (the document only contains the citation, not the full bibliography entry), but I thought it likely that the given name would be relevant as well. — moewe, Jun 21 '21 at 18:24
I should probably say that by "the first example" I mean the first code block in my answer. As mentioned at the end of the comment, the example in the question indeed does not reproduce an issue with the {\'\i}. — moewe, Jun 21 '21 at 19:39

Accents in .bib file with biblatex (worked with natbib)

1 Answers1