1) Fixing a problem with unique identifier generation (Issue OpenCyphal#88)
2) Generating C++ data structures (Issue OpenCyphal#91)

Still TODO for Issue OpenCyphal#91 is serialization for types.

Disabling the Jinja cache results in the unique name generator state disappearing from the global Jinja context:

    >       return _UniqueNameGenerator.ensure_generator_in_globals(env.globals)('c', adj_base_token, '_', '_')
    E       TypeError: 'NoneType' object is not callable
    .tox/py35-test/lib/python3.5/site-packages/nunavut/lang/c.py:527: TypeError

This issue affects all languages. The following test modification indicates that the state gets replaced with None:

    ------------------------- src/nunavut/lang/__init__.py -------------------------
    index bd6524c..135a29a 100644
    @@ -226,7 +226,7 @@ class _UniqueNameGenerator:
        def ensure_generator_in_globals(cls, environment_globals: typing.Dict[str, typing.Any]) -> '_UniqueNameGenerator':
            from .. import TypeLocalGlobalKey

    -        if TypeLocalGlobalKey not in environment_globals:
    +        if environment_globals.get(TypeLocalGlobalKey) is None:
                environment_globals[TypeLocalGlobalKey] = cls()
            return typing.cast('_UniqueNameGenerator', environment_globals[TypeLocalGlobalKey])

Am stuck.

Progress on #91

1) Fixing a problem with unique identifier generation (Issue #88)
2) Generating C++ data structures (Issue #91)

Still TODO for Issue #91 is serialization for types.

* Combining additional #88 test with #88 fix and #91

Disabling the Jinja cache results in the unique name generator state disappearing from the global Jinja context:

    >       return _UniqueNameGenerator.ensure_generator_in_globals(env.globals)('c', adj_base_token, '_', '_')
    E       TypeError: 'NoneType' object is not callable
    .tox/py35-test/lib/python3.5/site-packages/nunavut/lang/c.py:527: TypeError

This issue affects all languages. The following test modification indicates that the state gets replaced with None:

    ------------------------- src/nunavut/lang/__init__.py -------------------------
    index bd6524c..135a29a 100644
    @@ -226,7 +226,7 @@ class _UniqueNameGenerator:
        def ensure_generator_in_globals(cls, environment_globals: typing.Dict[str, typing.Any]) -> '_UniqueNameGenerator':
            from .. import TypeLocalGlobalKey

    -        if TypeLocalGlobalKey not in environment_globals:
    +        if environment_globals.get(TypeLocalGlobalKey) is None:
                environment_globals[TypeLocalGlobalKey] = cls()
            return typing.cast('_UniqueNameGenerator', environment_globals[TypeLocalGlobalKey])

Am stuck.

Progress on #91

1) Fixing a problem with unique identifier generation (Issue #88)
2) Generating C++ data structures (Issue #91)

Still TODO for Issue #91 is serialization for types.

* A new way to generate unique ids in a template

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

* Issue #94 and progress on Issue #91

Major refactor of language context to drive filter options from
configuration. The ability to modify things like stropping and
encoding were fundamentally broken before this change.

This work was done in support of adding named types as a concept to
languages. This will allow generation of serialization code that is more
configurable and will avoid fixing decisions like the use of size_t into
templates instead moving these decisions into user-overridable
configuration.

* updates per PR review comments

1. refactor to support embedding support code in nunavut distribution.
2. refactor to support optional inclusion of serialization support and
routines. Retaining ability to generate POD types.
3. Simplification of some CMake stuff to deduplicate what TOX does for
us.
4. Unification of test fixtures between Sybil and our regular unit
tests.

1. refactor to support embedding support code in nunavut distribution.
2. refactor to support optional inclusion of serialization support and
routines. Retaining ability to generate POD types.
3. Simplification of some CMake stuff to deduplicate what TOX does for
us.
4. Unification of test fixtures between Sybil and our regular unit
tests.

1. refactor to support embedding support code in nunavut distribution.
2. refactor to support optional inclusion of serialization support and
routines. Retaining ability to generate POD types.
3. Simplification of some CMake stuff to deduplicate what TOX does for
us.
4. Unification of test fixtures between Sybil and our regular unit
tests.

1. refactor to support embedding support code in nunavut distribution.
2. refactor to support optional inclusion of serialization support and
routines. Retaining ability to generate POD types.
3. Simplification of some CMake stuff to deduplicate what TOX does for
us.
4. Unification of test fixtures between Sybil and our regular unit
tests.

* A lot of stuff in this one:

1. Fixed support generator to properly use specified file and line
   post-processors.
2. General progress on the variable length array.
3. Added "no-cov" to cmake build so it doesn't get in the way during
   development. This was a constant problem on OSX.

Still left is to get all the noexcept guarentees to be correct and to
figure out how to verify them.

* adding array type override

* version bump

* disabling coverage for now

* Fixing GCC 9 and earlier syntax

* parallelizing python builds

* increasing host parallelism

* fixing concurrency issues with native build