Read YAML using C++ newSolution in Python / access root level data

[ischoegl]

Changes proposed in this pull request

• Use C++ newSolution to parse YAML strings and file input
• Save root level YAML information and input source in C++
• Ensure that an empty transport manager is created in C++ newSolution
• Replaces previous Python implementation of YAML input
• Implementation expands existing serialization introduced with Enable serialization of Cantera objects #984
• Allow for user-defined top-level YAML fields in YamlWriter (e.g. description)

Also:

• Deprecate Species.fromYaml in favor of Species.from_yaml
• Ensure that resetting transport in Python is registered at the C++ level
• Fix a minor glitch where LaTeX formatting in a figure title crashes example CI runs

If applicable, fill in the issue number this pull request is fixing

Closes #1013, closes #1131, supersedes #881, needed for #692

If applicable, provide an example illustrating new features this pull request is introducing

In [1]: import cantera as ct
   ...: gas = ct.Solution("h2o2.yaml", transport_model=None)
   ...: gas.input_header
   ...:
Out[1]: 
{'cantera-version': '2.5.0',
 'date': 'Wed, 11 Dec 2019 16:59:04 -0500',
 'description': 'Hydrogen-Oxygen submechanism extracted from GRI-Mech 3.0.\nModified from the original to include N2.',
 'generator': 'ck2yaml',
 'input-files': ['h2o2.inp', 'gri30_tran.dat'],
 'units': {'activation-energy': 'cal/mol', 'length': 'cm', 'quantity': 'mol'}}

The transport_model is no longer Transport if not set, but yields

In [2]: gas.transport_model
Out[2]: 'None'

where appropriate safeguards are now also implemented in C++.

Further, it is now possible to specify top-level YAML fields for YamlWriter. For example in extract_submechanism.py

gas2 = ct.Solution(name="gri30-CO-H2-submech",
                   thermo="ideal-gas", kinetics="gas", transport_model="Mix",
                   species=species, reactions=reactions)

# Save the resulting mechanism for later use
gas2.update_user_header({"description": "CO-H2 submechanism extracted from GRI-Mech 3.0"})
gas2.write_yaml("gri30-CO-H2-submech.yaml", header=True)

which produces

description: CO-H2 submechanism extracted from GRI-Mech 3.0
generator: YamlWriter
cantera-version: 2.6.0a3
git-commit: ec800b863
date: Mon Oct 25 03:01:22 2021
...

Checklist

• The pull request includes a clear description of this code change
• Commit messages have short titles and reference relevant issues
• Build passes (scons build & scons test) and unit tests address code coverage
• Style & formatting of contributed code follows contributing guidelines
• The pull request is ready for review

[codecov]

Codecov Report

Merging #1129 (f51bd6d) into main (2d211db) will increase coverage by 0.10%.
The diff coverage is 85.45%.

@@            Coverage Diff             @@
##             main    #1129      +/-   ##
==========================================
+ Coverage   73.52%   73.62%   +0.10%     
==========================================
  Files         365      366       +1     
  Lines       48242    48373     +131     
==========================================
+ Hits        35470    35617     +147     
+ Misses      12772    12756      -16     

Impacted Files	Coverage Δ
include/cantera/base/Solution.h	100.00% <ø> (ø)
include/cantera/base/YamlWriter.h	100.00% <ø> (ø)
src/transport/TransportBase.cpp	31.57% <ø> (+3.00%)	⬆️
include/cantera/transport/TransportBase.h	48.61% <47.22%> (+40.03%)	⬆️
src/transport/TransportFactory.cpp	98.33% <83.33%> (-1.67%)	⬇️
src/base/Solution.cpp	93.06% <93.75%> (-2.28%)	⬇️
include/cantera/kinetics/KineticsFactory.h	100.00% <100.00%> (ø)
include/cantera/thermo/ThermoFactory.h	100.00% <100.00%> (ø)
include/cantera/transport/TransportFactory.h	100.00% <100.00%> (ø)
src/base/YamlWriter.cpp	100.00% <100.00%> (ø)
... and 18 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 2d211db...f51bd6d. Read the comment docs.

[ischoegl]

@speth / @tsikes ... this PR replaces #881 (access to top-level YAML annotations), with a - hopefully - better fate. I improved the linkage of YAML input in Python and the underlying C++ so both use the C++ version of YAML input for Solution. I also replaced the Transport band-aid in Python with something that works universally.

There is currently one failing test (Sundials 5.7 not getting correct thermal_diff_coeffs values), where I'm running out of ideas as I cannot replicate the issue on either of my two machines (may have to create a docker with the exact same environment).

Edit: this docker container can reproduce the failure.

[speth]

Thanks, @ischoegl, this generally looks pretty good to me. I had a couple of suggestions for simplifications, as well as some other minor comments.

[ischoegl]

@speth ... I believe I took care of all the comments. Regarding the transport model, it's possible to replace the C++ "None" with a Python None, but the string version may be more intuitive/descriptive to users.

[ischoegl]

@speth ... thanks for the clarification. I updated as suggested ...

[speth]

Thanks for the updates, @ischoegl, I think this is almost set. I just had a couple of further comments, mainly related to the new factory functions.

[ischoegl]

@speth ... I implemented the one remaining suggestion for newTranport. I believe this should be ready ...

[speth]

Thanks, @ischoegl, this looks good to me.

Regarding one of your last comments, I would agree that the various global factory functions for new Cantera thermo / kinetics / transport / etc. objects could use some housekeeping. At a minimum, there's both the inconsistency whether they return raw pointers, unique_ptr or shared_ptr as well as confusing variability in their names (do we need the Mgr suffix in newKineticsMgr?).