Re: use heading like sectioning commands with USE_NEXT_HEADING_FOR_LONE_NODE

2025-03-01 Thread Patrice Dumas
On Sat, Mar 01, 2025 at 10:05:49PM +, Gavin Smith wrote: > On Fri, Feb 14, 2025 at 06:40:31PM +0100, Patrice Dumas wrote: > > Hello, > > > > This comes from the thread on anchors, but I start a new thread to avoid > > mixing the two issues. > > > > I propose that, if USE_NEXT_HEADING_FOR_LONE

Re: normalization problem with `@anchor` targets

2025-03-01 Thread pertusus
On Sat, Mar 01, 2025 at 08:09:57PM +, Gavin Smith wrote: > ADD_TRANSLITERATED_REDIRECTION_FILES would be fine. I guess it could be > taken to imply that these are additional files that are being added. This is implemented. I think that we could keep this customization option for some years,

Re: use heading like sectioning commands with USE_NEXT_HEADING_FOR_LONE_NODE

2025-03-01 Thread Gavin Smith
On Fri, Feb 14, 2025 at 06:40:31PM +0100, Patrice Dumas wrote: > Hello, > > This comes from the thread on anchors, but I start a new thread to avoid > mixing the two issues. > > I propose that, if USE_NEXT_HEADING_FOR_LONE_NODE is set, the @*heading > appearing after a @node be treated as much as

Re: normalization problem with `@anchor` targets

2025-03-01 Thread Gavin Smith
On Sat, Mar 01, 2025 at 08:57:52PM +0100, pertu...@free.fr wrote: > On Sat, Mar 01, 2025 at 06:56:28PM +, Gavin Smith wrote: > > > We need another variable, and we should document that it is transitory > > > and that we will remove it someday, while TRANSLITERATE_FILE_NAMES > > > should remain.

Re: normalization problem with `@anchor` targets

2025-03-01 Thread pertusus
On Sat, Mar 01, 2025 at 06:56:28PM +, Gavin Smith wrote: > > We need another variable, and we should document that it is transitory > > and that we will remove it someday, while TRANSLITERATE_FILE_NAMES > > should remain. > > > > What about ADD_NODES_REDIRECTIONS_TRANSLITERATED? > > That name

Re:

2025-03-01 Thread Patrice Dumas
On Sat, Mar 01, 2025 at 06:38:24PM +, Gavin Smith wrote: > I can put it back. The main reason is that the Texinfo XML format is > not used by anybody, so it seems like a waste of time talking or thinking > about it. It could be used, maybe. > If anybody did use it as an intermediate format,

Re: normalization problem with `@anchor` targets

2025-03-01 Thread Gavin Smith
On Sat, Mar 01, 2025 at 07:20:53PM +0100, pertu...@free.fr wrote: > On Sat, Mar 01, 2025 at 05:24:29PM +, Gavin Smith wrote: > > On Wed, Feb 12, 2025 at 01:45:37AM +0100, pertu...@free.fr wrote: > > > On Tue, Feb 11, 2025 at 10:06:54PM +, Gavin Smith wrote: > > > > So it seems to me that th

Re:

2025-03-01 Thread Gavin Smith
On Sat, Mar 01, 2025 at 07:14:09PM +0100, Patrice Dumas wrote: > On Sat, Mar 01, 2025 at 12:03:14PM -0500, Gavin D. Smith wrote: > > branch: master > > commit 2b13220e1d290be0696bce255c924963339cb95c > > Author: Gavin Smith > > AuthorDate: Sat Mar 1 08:53:35 2025 + > > > > * doc/texinfo.t

Re: normalization problem with `@anchor` targets

2025-03-01 Thread pertusus
On Sat, Mar 01, 2025 at 05:24:29PM +, Gavin Smith wrote: > On Wed, Feb 12, 2025 at 01:45:37AM +0100, pertu...@free.fr wrote: > > On Tue, Feb 11, 2025 at 10:06:54PM +, Gavin Smith wrote: > > > So it seems to me that this transliteration produces bad results in many > > > cases and should be

Re:

2025-03-01 Thread Patrice Dumas
On Sat, Mar 01, 2025 at 12:03:14PM -0500, Gavin D. Smith wrote: > branch: master > commit 2b13220e1d290be0696bce255c924963339cb95c > Author: Gavin Smith > AuthorDate: Sat Mar 1 08:53:35 2025 + > > * doc/texinfo.texi (Adding Output Formats): remove suggestion > to use Texinfo XML as an

Re: normalization problem with `@anchor` targets

2025-03-01 Thread Gavin Smith
On Wed, Feb 12, 2025 at 01:45:37AM +0100, pertu...@free.fr wrote: > On Tue, Feb 11, 2025 at 10:06:54PM +, Gavin Smith wrote: > > So it seems to me that this transliteration produces bad results in many > > cases and should be turned off by default. > > Ok. Should I add some backward compatibi

Re: Warnings printed when regenerating test results

2025-03-01 Thread Patrice Dumas
On Sat, Mar 01, 2025 at 03:39:36PM +, Gavin Smith wrote: > I was regenerating the test results. Under tta/perl, I ran: > > ../maintain/all_tests.sh generate > > There are warning messages printed about prototypes: > > ./t/02coverage.t > Texinfo::ManipulateTree::set_element_tree_numbers() ca

Warnings printed when regenerating test results

2025-03-01 Thread Gavin Smith
I was regenerating the test results. Under tta/perl, I ran: ../maintain/all_tests.sh generate There are warning messages printed about prototypes: ./t/02coverage.t Texinfo::ManipulateTree::set_element_tree_numbers() called too early to check prototype at t//../../perl/Texinfo/ManipulateTree.pm

Re: Build - possible bug for "multiple outputs"

2025-03-01 Thread Gavin Smith
On Sat, Mar 01, 2025 at 01:48:39PM +0200, Eli Zaretskii wrote: > > From: Gavin Smith > > Date: Sat, 1 Mar 2025 11:16:31 + > > > > A simpler solution could be to use the GNU make .NOTPARALLEL feature: > > Are we requiring GNU Make for building Texinfo? Because otherwise > '.NOTPARALLEL' is A

Re: Build - possible bug for "multiple outputs"

2025-03-01 Thread Eli Zaretskii
> From: Gavin Smith > Date: Sat, 1 Mar 2025 11:16:31 + > > A simpler solution could be to use the GNU make .NOTPARALLEL feature: Are we requiring GNU Make for building Texinfo? Because otherwise '.NOTPARALLEL' is AFAIK GNU-specific.

Re: Build - possible bug for "multiple outputs"

2025-03-01 Thread Gavin Smith
On Sat, Mar 01, 2025 at 10:51:44AM +, Gavin Smith wrote: > I had an idea that "locking" could be done in a separate helper program. > The "make" rule would become something like: > > LOCK_ONCE=build_aux/lock-once.sh > > $(srcdir)/main/command_data.c $(srcdir)/main/command_ids.h > $(srcdir)/m

Build - possible bug for "multiple outputs"

2025-03-01 Thread Gavin Smith
I came across a section in the Automake manual describing problems with Makefile rules with multiple outputs. https://www.gnu.org/software/automake/manual/html_node/Multiple-Outputs.html We have such rules in several places for building texi2any. For example, in tta/C/Makefile.am: $(srcdir)/mai