Update GHA TPL cache; resolve new test failures

[jsiirola]

Fixes # .

Summary/Motivation:

This updates the GHA TPL caches. Sort of. We are updating GAMS to the almost-current version (52.5) because the current version (53.1) causes segfaults (at least under pytest). The updated versions led to some test errors that this PR also resolves:

• Rework the BARON license test to rely on the LST ("times") file and not parsing the log (in the most recent BARON version the log message changed).
  • This also takes the opportunity to significantly simplify the logic around parsing and caching the BARON version / licensing information.
  • This also adds some "defensive" exceptions to catch incomplete results parsing if the Baron API ever expands/
• Update Xpress "locally optimal" test so Xpress doesn't recognize it as a conic problem it "knows how to solve"
• Update GAMS on GHA:
  • Stop installing gams distribution on OSX (self-extracting zip is no longer distributed, and I couldn't see how to (easily) install the package on GHA.
  • Update the gams Python installation for recent package layout
  • Add fallback on gamspy from PyPI if we either don't recognize the layout or didn't install the package

While working on things, some other things were cleaned up:

• Rework/simplify the download-extensions command
  • simplify (and document!) function control flow.
  • move retry parameters from module constants to command-line arguments.
• GHA drivers
  • standardize on use of the curlrc for passing standard options to curl
  • remove creation of local variables that led to obfuscation of the driver logic
  • update test-branches python versions to get better coverage without adding tests
• cplex: IBMDecisionOptimization released new versions of cplex and docplex that are both a bit buggy
  • cplex 22.1.2.1 errors when computing the IIS on Python 3.13 & 3.14. It looks like the file they write to is getting closed/corrupted before the routine exits. Either way, it results in some (fatal) error is coming out of the compiled library. I don't see how to fix or workaround, so we are just going to ignore that release.
  • docplex 2.32.257 is missing required dependencies for Python<=3.12. This highlighted that our tests for whether docplex is available were fragile / incomplete. This PR updates those tests to be more robust.
• Quit attempting to verify links to SourceForge: it appears that SF is categorically rejecting connections from GHA.

NOTE: the intent is to let this PR build and generate the new caches, then we will update the PR to unpin GAMS (but not trigger the caches to be rebuilt).

Changes proposed in this PR:

• (see above)

Legal Acknowledgement

By contributing to this software project, I have read the contribution guide and agree to the following terms and conditions for my contribution:

1. I agree my contributions are submitted under the BSD license.
2. I represent I am authorized to make the contributions and grant the license. If my employer has rights to intellectual property that includes these contributions, I represent that I have received permission to make contributions and grant the required license on behalf of that employer.

[codecov]

Codecov Report

❌ Patch coverage is 90.16393% with 6 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.68%. Comparing base (ae10c13) to head (1fe200b).
⚠️ Report is 61 commits behind head on main.

Files with missing lines	Patch %	Lines
pyomo/solvers/plugins/solvers/BARON.py	83.78%	6 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3865      +/-   ##
==========================================
+ Coverage   89.65%   89.68%   +0.02%     
==========================================
  Files         908      908              
  Lines      106757   106735      -22     
==========================================
+ Hits        95715    95726      +11     
+ Misses      11042    11009      -33     

Flag	Coverage Δ
builders	29.07% <65.57%> (-0.02%)	⬇️
default	85.96% <86.88%> (?)
expensive	35.51% <65.57%> (?)
linux	87.13% <88.52%> (-2.01%)	⬇️
linux_other	87.13% <88.52%> (+0.29%)	⬆️
oldsolvers	28.00% <27.86%> (+<0.01%)	⬆️
osx	82.04% <88.52%> (-1.01%)	⬇️
win	85.57% <88.52%> (+0.29%)	⬆️
win_other	85.57% <88.52%> (+0.29%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[blnicho]

I have a couple (hopefully easy) questions

[blnicho]

We no longer need idaes-lib on Windows?

[mrmundt]

I think the better question here is, "Did we ever actually need it in the first place?", and I think the answer is "no."

[jsiirola]

I think there was a time when we did need it (we were distributing some "standard libraries" that are always present on *nix). I think at some point we started compiling things more statically, which cut out the need for the extra libraries (or we changed how things were packaged so that the libraries we needed for the solvers were bundled with the solvers). Either way, I don't think there is anything in idaes-lib that we need for Pyomo testing.