... similarly to the change for pdf (#18181), but easier because there are no
baseline images for which we need to provide bug-level backcompat :-)

Drop the FontInfo metadata (which is explicitly optional in the
PostScript spec) to avoid having to figure out the correct encoding
(which can be quite obscure).

Replace the implementation of the_sc command from
7 -1 roll{setcachedevice}{pop pop pop pop pop pop}ifelse to a plain
setcachedevice (as I cannot see any case where the "other" branch is
taken).

Drop the splitting of long commands usingexec (_e) -- this is only
needed for level-1 postscript, which has a small fixed stack size; we
output level-2 postscript (per backend_version) and I guess level-1 is
rarely in use nowadays anyways (probably the feature could be added back
if there's really demand for it, but let's not get ahead of ourselves).

Previously, some composite characters would be output in a "compressed"
form (e.g., accented characters would be recorded as "draw the accent,
then run the charproc for the unaccented character"). This is lost, but
I'd guess outputting .ps.gz is better if compression really matters.

PR Summary

PR Checklist

• Has pytest style unit tests (andpytest passes).
• IsFlake 8 compliant (runflake8 on changed files to check).
• New features are documented, with examples if plot related.
• Documentation is sphinx and numpydoc compliant (the docs shouldbuild without error).
• Conforms to Matplotlib style conventions (installflake8-docstrings andpydocstyle<4 and runflake8 --docstring-convention=all).
• New features have an entry indoc/users/next_whats_new/ (follow instructions in README.rst there).
• API changes documented indoc/api/next_api_changes/ (follow instructions in README.rst there).