[maximilianruesch]

No description provided.

[hsemenenko]

Let's add a test that the names provided here match the ops provided by the HUGR extension.

[maximilianruesch]

Do you mean that for every op here there is a hugr op name that matches it?

[hsemenenko]

I think it should be the opposite: for every hugr op, there is a cost that matches it.

[hsemenenko]

Scratch that, the dynamic ops should not have costs, it should be just the block primitives.

[maximilianruesch]

I mean we can add such a test (we will have to exclude idle), if you want to stick to the idea.

[hsemenenko]

I'd like to make sure that the hugr ops and costs don't become disjoint. For example, we're missing pauli_exp_all_x_except_z_at and pauli_exp_all_z_except_x_at here. Annoyingly, the naming is not that consistent:

Cost field	Current HUGR extension name
cx_intrablock	cx
cx_interblock	cx_between_blocks
{xx,yy,zz}_phase_intrablock	{xx,yy,zz}_phase
{xx,yy,zz}_phase_interblock	{xx,yy,zz}_phase_between_blocks

[maximilianruesch]

Fair endeavour. My next question would be what naming to adapt though

[hsemenenko]

Imo the names we have for the costs fields are more descriptive and having the "default" cx be the intrablock op has no justification.

[hsemenenko]

Add docstrings to each of these. Depth is related to the block size k.