Bug Tracker 
| ID | 309🔗 |
|---|---|
| Submitted by: | Nicola Talbot 🦜 |
| Date: | 2025-08-11 12:55:01 |
| Status | Open Sign in if you want to bump this report. |
| Category | datatool |
| Version | 3.4.1 |
| Summary | \datatool_get_first_grapheme:nN has stopped working |
Sign in to subscribe to notifications about this report.
Description
\datatool_get_first_grapheme:nN internally uses \text_map_inline:nN (which should map over each grapheme in the first argument) to fetch the first grapheme (by breaking the loop at the end of the first iteration). Unfortunately a recent update to the LaTeX kernel has changed the behaviour of \text_map_inline:nN so that it iterates over the tokens rather than the graphemes. This means that with pdfLaTeX (as opposed to LuaLaTeX and XeLaTeX) is now picking up the first octet of a UTF-8 character rather than all the octets that make up the character.
The minimal example produces an "Invalid UTF-8 byte sequence" error. This issue is outside of my control but I have reported it ([Link]).
Note that since \datatool_get_first_grapheme:nN is internally used by \printnoidxglossary provided by the glossaries package, that command also affected.
MWE
Download (218B)
% arara: pdflatex
\documentclass{article}
\usepackage{datatool-base}
\begin{document}
\ExplSyntaxOn
\datatool_get_first_grapheme:nN { élan } \l_tmpa_tl
\tl_show:N \l_tmpa_tl
\l_tmpa_tl
\ExplSyntaxOff
\end{document}
Evaluation
Comments
0 comments.
Add Comment
Page permalink: https://www.dickimaw-books.com/bugtracker.php?key=309