While processing LilyPond documentation in Catalan I noticed that I
get messages like
```
translation 1 error(s)
translated string: Inici del cap@'@dotless{i}tol actual o previ
Error messages:
@' expected braces
translation 1 error(s)
translated string: Cap@'@dotless{i}tol seg@"uent
Error messag
[texinfo.tex 2025-02-03.17]
> We should document that node names should not contain @macro-defined
> macros, and texi2any could give a warning if a document does this.
Yes, please do.
> One possibility is to expand to the entire @node line instead, as in:
>
> @macro Node {name}
> @node \n
Building texinfo-7.2 on OS X 10.13 with support libraries supplied by fink, all
of these related self-tests fail:
FAIL: test_scripts/encoded_non_ascii_command_line.sh
FAIL: test_scripts/encoded_non_ascii_test_epub.sh
FAIL: test_scripts/encoded_non_ascii_test_rawtext.sh
FAIL: test_scripts/encoded_
On Wed, Feb 05, 2025 at 02:56:29PM +, Werner LEMBERG wrote:
>
> > The reason, AFAICS, is that user-defined macros are not immediately
> > expanded while writing an entry to the `.toc` file. Is there a
> > reason for it?
>
> Note that it works just fine if I use `@set` and `@value` for the
>
> The reason, AFAICS, is that user-defined macros are not immediately
> expanded while writing an entry to the `.toc` file. Is there a
> reason for it?
Note that it works just fine if I use `@set` and `@value` for the
purpose shown in the file, and this is probably even a better solution
for th
[texinfo.tex 2025-02-03.17]
> Please consider the attached example.
Oops, here is the right example.
Werner
\input texinfo
@macro A
urgh
@end macro
@contents
@node Top
@top Top
Top
@unmacro A
@macro A
foo
@end macro
@node @A{}
@chapter @A{}
@A{}
@page
@unmacro A
@macro A
bar
Please consider the attached example. It works as expected if
processed with `texi2any`. However, the output of `texi2pdf` is
incorrect, producing a bad ToC.
The reason, AFAICS, is that user-defined macros are not immediately
expanded while writing an entry to the `.toc` file. Is there a reaso
On Tue, Feb 04, 2025 at 11:11:40PM +0100, Patrice Dumas wrote:
> > Yes, I think this makes sense. If there is any more testing we could do
> > with real-life manuals this could help to throw up any bugs or
> > incompatibilies (maybe you have done this already).
>
> Yes, I already did it for the c
