summaryrefslogtreecommitdiff
path: root/macros/latex-dev/base/utf8ienc.dtx
diff options
context:
space:
mode:
Diffstat (limited to 'macros/latex-dev/base/utf8ienc.dtx')
-rw-r--r--macros/latex-dev/base/utf8ienc.dtx46
1 files changed, 29 insertions, 17 deletions
diff --git a/macros/latex-dev/base/utf8ienc.dtx b/macros/latex-dev/base/utf8ienc.dtx
index a0c96357a0..8789dd35ef 100644
--- a/macros/latex-dev/base/utf8ienc.dtx
+++ b/macros/latex-dev/base/utf8ienc.dtx
@@ -216,7 +216,7 @@
%<+ts1> \ProvidesFile{ts1enc.dfu}
%<+x2> \ProvidesFile{x2enc.dfu}
%<+all> \ProvidesFile{utf8enc.dfu}
-%<-utf8-2018> [2021/12/13 v1.2o UTF-8 support]
+%<-utf8-2018> [2022/02/13 v1.3a UTF-8 support]
% \end{macrocode}
%
% \begin{macrocode}
@@ -363,11 +363,12 @@
% \begin{macro}{\UTFviii@two@octets@noexpand}
% \begin{macro}{\UTFviii@three@octets@noexpand}
% \begin{macro}{\UTFviii@four@octets@noexpand}
+% \changes{v1.3a}{2022/02/19}{Use \cs{unexpanded} rather than \cs{noexpand} so only takes one expansion step}%
% These temporarily prevent the active chars from expanding.
% \begin{macrocode}
-\long\def\UTF@two@octets@noexpand#1#2{\noexpand#2\noexpand}
-\long\def\UTF@three@octets@noexpand#1#2#3{\noexpand#2\noexpand#3\noexpand}
-\long\def\UTF@four@octets@noexpand#1#2#3#4{\noexpand#2\noexpand#3\noexpand#4\noexpand}
+\long\def\UTF@two@octets@noexpand#1#2#3{\unexpanded{#2#3}}
+\long\def\UTF@three@octets@noexpand#1#2#3#4{\unexpanded{#2#3#4}}
+\long\def\UTF@four@octets@noexpand#1#2#3#4#5{\unexpanded{#2#3#4#5}}
% \end{macrocode}
% \end{macro}
% \end{macro}
@@ -376,11 +377,12 @@
% \begin{macro}{\UTFviii@two@octets@string}
% \begin{macro}{\UTFviii@three@octets@string}
% \begin{macro}{\UTFviii@four@octets@string}
+% \changes{v1.3a}{2022/02/19}{Use \cs{detokenize} rather than \cs{string} so only takes one expansion step}%
% And the same with \cs{string} for use in \cs{csname} constructions.
% \begin{macrocode}
-\long\def\UTF@two@octets@string#1#2{\string#2\string}
-\long\def\UTF@three@octets@string#1#2#3{\string#2\string#3\string}
-\long\def\UTF@four@octets@string#1#2#3#4{\string#2\string#3\string#4\string}
+\long\def\UTF@two@octets@string#1#2#3{\detokenize{#2#3}}
+\long\def\UTF@three@octets@string#1#2#3#4{\detokenize{#2#3#4}}
+\long\def\UTF@four@octets@string#1#2#3#4#5{\detokenize{#2#3#4#5}}
% \end{macrocode}
% \end{macro}
% \end{macro}
@@ -542,15 +544,24 @@
% |\count@| and ending at |\@tempcnta|${} - 1$, each time executing
% the code in |\UTFviii@tmp|.
%
-% All this is done in a group so that temporary catcode changes
-% etc.~vanish after everything is set up.
+% Store current settings so can restore after the loops without using a group (gh/762).
+%
+% \begin{macrocode}
+\edef\reserved@a{%
+\catcode`\noexpand\~=\the\catcode`\~\relax
+\catcode`\noexpand\"=\the\catcode`\"\relax
+\uccode`\noexpand\~=\the\uccode`\~\relax
+\count@=\the\count@\relax
+\@tempcnta=\the\@tempcnta\relax
+\let\noexpand\reserved@a\relax}
+% \end{macrocode}
%
% \begin{macrocode}
-\begingroup
\catcode`\~13
\catcode`\"12
% \end{macrocode}
%
+% \changes{v1.3a}{2022/02/19}{Use local protected definitions in the character definitions}%
% \begin{macrocode}
\def\UTFviii@loop{%
\uccode`\~\count@
@@ -566,7 +577,7 @@
% C0 controls are valid UTF-8 but defined to give the ``Character not defined error''
% They may be defined with |\DeclareUnicodeCharacter|.
% \begin{macrocode}
- \def\UTFviii@tmp{\xdef~{\noexpand\UTFviii@undefined@err{:\string~}}}
+ \def\UTFviii@tmp{\protected\edef~{\noexpand\UTFviii@undefined@err{:\string~}}}
% 0 ^^@ null
\count@"1
\@tempcnta9
@@ -588,7 +599,7 @@
% \begin{macrocode}
\count@"80
\@tempcnta"C2
- \def\UTFviii@tmp{\xdef~{\noexpand\UTFviii@invalid@err\string~}}
+ \def\UTFviii@tmp{\protected\edef~{\noexpand\UTFviii@invalid@err\string~}}
\UTFviii@loop
% \end{macrocode}
%
@@ -597,7 +608,7 @@
% \begin{macrocode}
\count@"C2
\@tempcnta"E0
- \def\UTFviii@tmp{\xdef~{\noexpand\UTFviii@two@octets\noexpand~}}
+ \def\UTFviii@tmp{\protected\edef~{\noexpand\UTFviii@two@octets\noexpand~}}
\UTFviii@loop
% \end{macrocode}
%
@@ -605,7 +616,7 @@
% \begin{macrocode}
\count@"E0
\@tempcnta"F0
- \def\UTFviii@tmp{\xdef~{\noexpand\UTFviii@three@octets\noexpand~}}
+ \def\UTFviii@tmp{\protected\edef~{\noexpand\UTFviii@three@octets\noexpand~}}
\UTFviii@loop
% \end{macrocode}
%
@@ -614,7 +625,7 @@
% \begin{macrocode}
\count@"F0
\@tempcnta"F5
- \def\UTFviii@tmp{\xdef~{\noexpand\UTFviii@four@octets\noexpand~}}
+ \def\UTFviii@tmp{\protected\edef~{\noexpand\UTFviii@four@octets\noexpand~}}
\UTFviii@loop
% \end{macrocode}
%
@@ -623,12 +634,13 @@
% \begin{macrocode}
\count@"F5
\@tempcnta"100
- \def\UTFviii@tmp{\xdef~{\noexpand\UTFviii@invalid@err\string~}}
+ \def\UTFviii@tmp{\protected\edef~{\noexpand\UTFviii@invalid@err\string~}}
\UTFviii@loop
% \end{macrocode}
%
+% Restore values after the loops.
% \begin{macrocode}
-\endgroup
+\reserved@a
% \end{macrocode}
%
% \end{macro}