Agustín Borgna
Agustín Borgna
The schema on [hugr's side](https://github.com/CQCL/hugr/blob/main/specification/schema/hugr_schema_v1.json) has been manually updated, and is currently out of sync with the pydantic definition here. Hugr now has a python package that defines the serialization...
Replaces the local `RawHugr`. https://pypi.org/project/hugr/ The `quantinuum_hugr` python package version `0.1.0a1` implements the serialization format `v1`, so it may require non-trivial changes here (and an update to the unpublished `0.3.0`...
Currently, `serialize` removes dummy nodes and adds state edges before doing the serialisation. https://github.com/CQCL/guppylang/blob/89edc2d724586d20438395b2c9f141821753dabe/guppylang/hugr_builder/hugr.py#L783 This is unexpected behaviour. It should probably copy the Hugr before doing any mutation.
This function fails to compile ```python main = GuppyModule("main") @guppy(main) def const_false(x: int) -> bool: """Return False.""" return False ``` with error ```bash Guppy compilation failed. Error in file test_files/even_odd.py:26...
`test_depth_hyp` runs depth optimisation on a set of circuits, with a 30ms deadline. The test may take longer in some cases, which causes CI to fail. See https://github.com/CQCL/tket2/actions/runs/9189665381/job/25272188461 We could...
Move the bindings to a `tket2._tket2` submodule and re-define all structures from python, backed by their binded counterparts. See [Quantomatic/quizx](https://github.com/Quantomatic/quizx/tree/master/pybindings) for an example. This gives us more flexibility on the...
As mentioned in https://github.com/CQCL-DEV/tket2/pull/100#discussion_r1331693153_ , - Most uses of `Units` only care about the `CircuitUnit` - We have some add-hoc maps that add the `Wire`s using the corresponding `CircuitLabeller`. We...
Add non-borrowing alternative to `Command` that pre-computes the data instead of keeping a reference to the circuit. This would be useful for simplifying #74 .
Loading a new ECC from `nam_6_3.rwr` takes a couple seconds. Is there any way to speed up this? The answer may very well be _no_, since that compiled rewriter contains...