typed_python
typed_python copied to clipboard
APrioriInvestments
An llvm-based framework for generating and calling into high-performance native code from Python.
right now if you try to send a very large string (say, 100gb), the deserializer will allocate space for it. We should allow clients to impose strict limits to ensure...
Our serialization library has a similar structure to protobuf - we want each field in a composite datastructure to have a field id that's stable over time as we change...
Right now we construct an Alternative like this ``` A = Alternative("A", OneKind=dict(x=int, y=float), AnotherKind=dict(name=str, hihi=float), memberFunction=lambda self: "this is a member function" ) ``` This can get a little...
Right now, you can mark a Class to be 'Final' which means that you can't subclass it, but in exchange, the compiler can speed up method dispatch (and also inline)...
Python has a behavior where `hash(1.0)` == `hash(1)` == `hash(True)` == 1. This happens so that you can look up an int in a dictionary with a float and vice...
If you write `OneOf(False, True, int)(0)` you are relying on the fact that the Value infrastructure compares 'False' to '0' with 'exact' set to true in order to have the...
if you have two instances of a OneOf(int, float) and you compare them, we check the type first and then we check the instance values. This doesn't match python, where...
Right now we can write ``` def f(x: int, y: int): ... ``` but we have no way of describing the idea that we want 'f' with two arguments that...
Currently, if all concrete alternatives share a number of fields, these have to be repeated explicitly. It would be better to have a mechanism to declare common alternative fields, for...
## Motivation and Context Once merged, this will allow the Sphinx docs in docs/ (and the docstrings) to be built and served at [typed-python.readthedocs.io](http://typed-python.readthedocs.io/). With an appropriate webhook the docs...
Metadata
Owner
Metadata
An llvm-based framework for generating and calling into high-performance native code from Python.