summaryrefslogtreecommitdiff
path: root/macros/latex/contrib/l3kernel/l3token.dtx
diff options
context:
space:
mode:
Diffstat (limited to 'macros/latex/contrib/l3kernel/l3token.dtx')
-rw-r--r--macros/latex/contrib/l3kernel/l3token.dtx11
1 files changed, 10 insertions, 1 deletions
diff --git a/macros/latex/contrib/l3kernel/l3token.dtx b/macros/latex/contrib/l3kernel/l3token.dtx
index 3db1f706c8..ce65d87061 100644
--- a/macros/latex/contrib/l3kernel/l3token.dtx
+++ b/macros/latex/contrib/l3kernel/l3token.dtx
@@ -43,7 +43,7 @@
% }^^A
% }
%
-% \date{Released 2021-05-07}
+% \date{Released 2021-05-11}
%
% \maketitle
%
@@ -1080,6 +1080,15 @@
% \cs{peek_regex:nTF} as control sequences, while functions that
% inspect individual tokens (for instance \cs{peek_charcode:NTF})
% only take into account their meaning.
+%
+% The \cs{peek_regex:nTF} function only inspects as few tokens as
+% necessary to determine whether the regular expression matches.
+% For instance \cs{peek_regex:nTF} \verb"{ abc | [a-z] }" |{ } { }|
+% |abc| will only inspect the first token~|a| even though the first
+% branch |abc| of the alternative is preferred in functions such as
+% \cs{peek_regex_remove_once:n}. This may have an effect on
+% tokenization if the input stream has not yet been tokenized and
+% category codes are changed.
% \end{texnote}
% \end{function}
%