Qiskit/qiskit

ConsolidateBlocks takes a kak_basis_gate argument that is not exercised by any test in test/python/transpiler/test_consolidate_blocks.py This is a follow up of #14417. The PR #14687 was an attempted to cover it with a single test but it was incomplete. If you base on that PR to fix this issue, dont forget to add WhimsyHippo (@hippowm) as a co-author.

Environment Qiskit version: This probably goes back to the initial release of the Rust drawer in Qiskit 2.4. What is happening? When given an empty circuit, i.e. one without qubits or clbits, TextDrawer::draw panics. How can we reproduce the issue? Add a test function like this to the test module in circuit_drawer.rs: #[test] fn test_empty_circuit() { let circuit = CircuitData::new(None, None, Param::Float(0.5)).unwrap(); let result = draw_circuit(&circuit, false, false, None, false).unwrap(); assert_eq!("global phase: 0.5", result.trim_end()); } What should happen? The test above, or any other call to the drawer (e.g. via the C API) should return an empty string for the circuit section (along with possibly phase information). Any suggestions? No response

What should we add? version 2.0 of pygraphviz now bundles graphviz with it: https://pygraphviz.github.io/documentation/stable/install.html can we get you to use this dependency instead of graphviz directly: https://quantum.cloud.ibm.com/docs/en/api/qiskit/utils#qiskitutilsoptionalshas_graphviz to simplify end user installation requirements for drawing coupling maps?

This issue tracks the remaining gaps between the Rust circuit drawer and the Python text drawer implementation. The goal is to achieve complete feature parity, enabling us to replace the Python text drawer with the Rust implementation for improved interoperability and lower maintenance cost. NOTE: This issue is marked as good first issue so that it appears in searches for that label. However, not all tasks listed below are suitable as good first issue. See below for more details. Point of integration The Rust circuit drawer uses its own internal representation and type system to render a CircuitData object. Therefore no Python-side functionality or class hierarchy is used by it. Given this, it seems that the cleanest and easiest point for replacement would be to call the Rust circuit drawer from this function: qiskit/qiskit/visualization/circuit/circuit_visualization.py Line 402 in cd146f5 def

I have to iterate over all dag nodes for which the operation is either Measure or Reset. Currently in DagCircuit.op_nodes the op argument cannot be a sequence of operations. It will be good to add this option, if it's more performant than two executions of op_nodes, one with op=Measure and the other with op=Reset.