38 fix types in the ast

[pablolh]

Changes

• Implement ControlledSemantic to be able to express CNOT as a
  controlled X, for instance
• Separate validation of qubit ranges from type-checking in a separate
  ANTLR tree visitor
• snake_case instead of camelCase for variable names
• Add docstrings to parsing classes
• The GateLibrary superclass is used to factorize code for gateset-aware
  classes (e.g. TypeChecker)
• Semantic classes are removed, absorbed into Gate subclasses
• Move ALL ANTLR parsing classes to parsing folder - this has already
  been done, but only partially
• Gates are specified directly with Python functions and a decorator
• Lay the basis of handling of anonymous gates
• Writer turned into impl pattern
• Rename AST -> IR
• Make test_testinterpreter a proper unit test, before it was testing
  parsing too
• Visitor pattern for the SquirrelIR class
• Use those visitors in e.g. writer and test_interpreter
• Various code fixes
• Use inspect to know Python function signatures
  This makes the gate semantic API nice to use from Python.

[elenbaasc]

I have been through the code, but mostly to check if there are no weird things or things going wrong. I haven't been able to check/optimize the functionalities. It's quite a big refactor and a really nice one as well! I would still need some time to fully understand how everything links together, but our meeting this morning helped a lot.

The test coverage is now 82%, it would be good if we could increase this (perhaps this could be an issue of its own though).

[pablolh]

Thanks Chris for the review, will look asap

[pablolh]

I've tried increasing the test coverage a bit, by looking at what was not covered
Most of the non-covered code is now generated code from ANTLR... Not sure if I can do much about it, the test_parsing.py is already pretty complete

[elenbaasc]

I've tried increasing the test coverage a bit, by looking at what was not covered Most of the non-covered code is now generated code from ANTLR... Not sure if I can do much about it, the test_parsing.py is already pretty complete

Yes, I see, good point. We could exclude the generated files by adding them to the omit list of the coverage tool in the pyproject.toml:

[tool.coverage.run]
branch = true
omit = ["tests/*", "**/.tox/**","opensquirrel/parsing/antlr/*"]