Avoid LaTeX scanner errors without a TeX tool

[Boulea7]

A Command on a LaTeX source fails with an undefined LATEXSUFFIXES error when the environment uses tools=[] and AllowSubstExceptions(). Skip the default LaTeX scanner when that variable is absent. Configured suffixes, custom scanners, and errors in configured substitution values retain their existing behavior.

Related #4511. This addresses the reported LaTeX error while preserving the current scanner registration model.

The regression covers .tex, .ltx, and .latex copies, unchanged builds, source updates, and strict command substitution errors. Unit tests cover absent, empty, configured, and custom suffixes and invalid variable references.

Validation on Python 3.11 / Linux:

python runtest.py \
  SCons/Scanner/LaTeXTests.py SCons/Scanner/ScannerTests.py \
  SCons/Tool/ToolTests.py SCons/SubstTests.py \
  test/Subst/AllowSubstExceptions.py test/Scanner \
  test/TEX/LATEXCOMSTR.py test/TEX/PDFLATEXCOMSTR.py \
  test/TEX/TEXCOMSTR.py test/TEX/PDFTEXCOMSTR.py

All 26 selected test files passed, with no failures or skips. The issue's original three-line copy reproduction also succeeds. The full suite and real TeX executables were not run; the four TeX checks above use the repository's simulated commands.

The added LaTeX regression test also fails on the base and passes on this head with Python 3.12/Linux. An earlier Windows AppVeyor run failed in test/Interactive/configure.py; that test passes on Linux at both commits, but the Windows failure's cause remains unconfirmed. Native Windows validation is still outstanding.

Contributor Checklist:

• New and updated tests cover the changed functionality.
• CHANGES.txt and RELEASE.txt are updated; README.rst was checked for relevant guidance.
• The release notes document the correction. No public API documentation change is needed.

This change and its tests were developed with OpenAI Codex.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[bdbaddog]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

Who requested this?

[Boulea7]

Who requested this?

Maybe Github's default setting...

[bdbaddog]

Who requested this?

Maybe Github's default setting...

Nope. Never happened on any SCons PR's before.
Do you have it set up in your account?

[Boulea7]

Looks like default setting for new users of Copilot (just paid) , I just closed it , NVM

[bdbaddog]

I think this is the wrong fix.
The latex scanners should not be registered if no tool which uses them has been loaded.

@mwichmann thoughts?

[Boulea7]

I'm checking the registration path when no LaTeX tool is loaded, including how a change there would affect other environments.

[Boulea7]

In local tests, removing the global registrations preserves nested input tracking for DVI/PDF builds from TeX. It does change env.Command('copied.tex', 'source.tex', ...) with tex loaded: changing an \input dependency no longer rebuilds the target. Should we preserve that implicit Command() scanning, or require an explicit source_scanner there?

[bdbaddog]

In local tests, removing the global registrations preserves nested input tracking for DVI/PDF builds from TeX. It does change env.Command('copied.tex', 'source.tex', ...) with tex loaded: changing an \input dependency no longer rebuilds the target. Should we preserve that implicit Command() scanning, or require an explicit source_scanner there?

With tool tex loaded it's not using the scanner?

[Boulea7]

It does with tools=['tex'] on the base commit: changing an \input dependency rebuilds the Command target. I meant my unpushed experiment that removes the global registrations. Generic Command relies on that global scanner mapping; the TeX builders retain their own explicit scanners. Removing the mapping caused the difference I reported, so I haven't applied that change to this PR.