Listings in Latex with UTF-8 (or at least german umlauts)

Question

Trying to include a source-file into my latex document using the listings package, i got problems with german umlauts inside of the comments in the code. Using

\lstset{
extendedchars=\true,
inputencoding=utf8x
}

Umlauts in the source files (encoded in UTF-8 without BOM) are processed, but they are somehow moved to the beginning of the word they are contained in. So

// die Größe muss berücksichtigt werden

in the input source file, becomes

// die ößGre muss übercksichtigt werden

in the output file.

NOTE: since i found errors in my initial setup, i heavily edited this question

Solution 1

ok, found kinda workaround now:

1. instead of listings package, use listingsutf8

   \usepackage{listingsutf8}

2. copy listings.sty to the folder the document resides

3. find the following lines

   \lst@CCPutMacro
       \lst@ProcessOther {"23}\#
       \lst@ProcessLetter{"24}\textdollar
       \lst@ProcessOther {"25}\%
       \lst@ProcessOther {"26}\&

4. Enter there the following lines (each "registers" one umlaut)

   \lst@ProcessLetter{"E4}{\"a}
   \lst@ProcessLetter{"F6}{\"o}
   \lst@ProcessLetter{"FC}{\"u}
   \lst@ProcessLetter{"C4}{\"A}
   \lst@ProcessLetter{"D6}{\"O}
   \lst@ProcessLetter{"DC}{\"U}
   \lst@ProcessLetter{"DF}{\ss{}}

5. Save the file

6. Use

   \lstset{
       extendedchars=\true,
       inputencoding=utf8/latin1
   }

to enable utf8 character to latin1 character mapping

7. Convert line endings of your source file from windows (\r\n) to unix (\n)
8. enjoy

I know this is ugly in many way, but its the only solution that works for me so far.

Solution 2

I found a simpler approach, which works for me:

\usepackage{listings}

\lstset{
  literate={ö}{{\"o}}1
           {ä}{{\"a}}1
           {ü}{{\"u}}1
}

Solution 3

For comments only, you can use the texcl option:

\lstset{language=C++,texcl=true}

Than your comments become Latex and you can use "special" characters

\begin{lstlisting}
int iLink = 0x01; // Paramètre entrée
\end{lstlisting}

Solution 4

This should work for other languages (Spanish, Danish) as well:

\documentclass[
a4paper, %% defines the paper size: a4paper (default), a5paper, letterpaper, ...
12pt %% set default font size to 12 point
]{scrartcl} %% article, see KOMA documentation (scrguide.dvi)

\usepackage[utf8]{inputenc}

\usepackage[T1]{fontenc}
\usepackage{listings}

\lstset{language=Pascal}
\lstset{literate=%
{Ö}{{\"O}}1
{Ä}{{\"A}}1
{Ü}{{\"U}}1
{ß}{{\ss}}2
{ü}{{\"u}}1
{ä}{{\"a}}1
{ö}{{\"o}}1
}

\begin{document}

[Latex: kann man Umlaute in lstlisting verwenden?]
\begin{lstlisting}
Test für Umlaut äöü ÄÖÜ ß So geht es
\end{lstlisting}

\end{document} 

Solution 5

My contribution for Czech language.

\lstset{
    inputencoding=utf8,
    extendedchars=true,
    literate=%
    {á}{{\'a}}1
    {č}{{\v{c}}}1
    {ď}{{\v{d}}}1
    {é}{{\'e}}1
    {ě}{{\v{e}}}1
    {í}{{\'i}}1
    {ň}{{\v{n}}}1
    {ó}{{\'o}}1
    {ř}{{\v{r}}}1
    {š}{{\v{s}}}1
    {ť}{{\v{t}}}1
    {ú}{{\'u}}1
    {ů}{{\r{u}}}1
    {ý}{{\'y}}1
    {ž}{{\v{z}}}1
    {Á}{{\'A}}1
    {Č}{{\v{C}}}1
    {Ď}{{\v{D}}}1
    {É}{{\'E}}1
    {Ě}{{\v{E}}}1
    {Í}{{\'I}}1
    {Ň}{{\v{N}}}1
    {Ó}{{\'O}}1
    {Ř}{{\v{R}}}1
    {Š}{{\v{S}}}1
    {Ť}{{\v{T}}}1
    {Ú}{{\'U}}1
    {Ů}{{\r{U}}}1
    {Ý}{{\'Y}}1
    {Ž}{{\v{Z}}}1
}

View more solutions

Comments

[scrub]

My main document is in utf8. (and it works, i can even write äöü in the main document)

[James]

listings does its character processing differently than the main document. So inputenc doesn't help, here; the listings packages needs to support utf8 input explicitly (hence listingsutf8).

[Vanuan]

I think, 'extendedchars=\true' is equal to 'extendedchars=false'.

[GDR]

Thank you - it worked! The same for Polish language: \lstset{literate={ą}{{\k{a}}}1 {ł}{{\l{}}}1 {ń}{{\'n}}1 {ę}{{\k{e}}}1 {ś}{{\'s}}1 {ż}{{\.z}}1 {ó}{{\'o}}1 {ź}{{\'z}}1 {Ą}{{\k{A}}}1 {Ł}{{\L{}}}1 {Ń}{{\'N}}1 {Ę}{{\k{E}}}1 {Ś}{{\'S}}1 {Ż}{{\.Z}}1 {Ó}{{\'O}}1 {Ź}{{\'Z}}1 }

[Martin Thoma]

I copied listings.sty to listingsutf8.sty in /usr/share/texmf-texlive/tex/latex/listings/ on Ubuntu 10.10. I edited the file, but my listings don't work.

[Chielus]

this works fine for me, package listingsutf8 is not needed. Best workaround!

[petrichor]

It also works for Turkish. Here is the related code snippet: \lstset{ literate={â}{{\^{a}}}1 {Â}{{\^{A}}}1 {ç}{{\c{c}}}1 {Ç}{{\c{C}}}1 {ğ}{{\u{g}}}1 {Ğ}{{\u{G}}}1 {ı}{{\i}}1 {İ}{{\.{I}}}1 {ö}{{\"o}}1 {Ö}{{\"O}}1 {ş}{{\c{s}}}1 {Ş}{{\c{S}}}1 {ü}{{\"u}}1 {Ü}{{\"U}}1 }

[przemoc]

And thank you, GDR! It was a time saver. You only forgot ć and Ć. Here is the full list (bonus: sorted) for quick Ctrl+C + Ctrl+V for others: \lstset{literate=% {ą}{{\k{a}}}1 {ć}{{\'c}}1 {ę}{{\k{e}}}1 {ł}{{\l{}}}1 {ń}{{\'n}}1 {ó}{{\'o}}1 {ś}{{\'s}}1 {ż}{{\.z}}1 {ź}{{\'z}}1 {Ą}{{\k{A}}}1 {Ć}{{\'C}}1 {Ę}{{\k{E}}}1 {Ł}{{\L{}}}1 {Ń}{{\'N}}1 {Ó}{{\'O}}1 {Ś}{{\'S}}1 {Ż}{{\.Z}}1 {Ź}{{\'Z}}1 } (obviously comments don't have newlines, so after pasting you have to fix it (e.g. in vim: :.s/ /\r/g)

[Simon]

Thank you - good solution! Anyway, it should be {ß}{{\ss}}1, because "ß" takes only 1 character in the output ;)

[Jan Špaček]

This is one of the most elegant solutions here, needs more upvotes! :)

[Jan Špaček]

The solution using texcl=true described in another answer seems to be more elegant.

[Eduardo Santana]

I have the same problem, I want to have a keyword with accents. Did any one did it?

[Njaal Gjerde]

Someone knows how this solution would look like for the norwegian characters æøå?

[Elmar Zander]

Wow! Then you can even put math formulas into the listings, e.g. double pi = 3.141; // This is $\pi$ or double d = 1.0 // $3 \int_0^1 x^2 dx$. This is really cool!

[zbr]

Thank you! :) Please note that I had to remove the inputencoding=utf8, and extendedchars=true, lines and also the % after literate= for it to work in my case.

[WerWet]

Awesome! The other option (with literate) don't work with XeLaTex.

[anion]

i do not think this is a solution or workaround because this does not help if you have long (maybe generated) content with arbitrary special characters.

[user202729]

It seems that if there's $ or similar in some comment, they will be interpreted as math formula (which may be unintended, and causes many bugs if they're malformed)