Implement a capacity-allocated constructor for DAGCircuit in Rust.

[raynelfss]

Summary

Resolves #13002
After #12550 merges, the following commits aim to provide awith_capacity alternative initializer for the DAGCircuit in Rust.

Details and comments

Will be rebased after #12550 merges

These commits add the DAGCircuit::with_capacity constructor to the rust representation of DAGCircuit. This would allow us to allocate an approximation of the necessary space to successfully create a DAGCircuit with certain properties to avoid re-allocating space in memory.

This constructor takes the following arguments:

• num_qubits for the number of qubits in the specific circuit.
• num_clbits for the number of clbits in the circuit.
• num_ops for the number of operations in the circuit (Optional).
• num_vars for the number of variables in the circuit (Optional).

Questions

• What is the best way of estimating the possible number of edges required for num_ops?
  • An operation (or a block) can have as many input/output qargs/cargs as possible
  • I used 3 (2 qubits + 1 classical or 3 qubits) as a baseline. But this is just speculation, if a better approach exists, let me in a comment.
• Best way of estimating the names of operations in a circuit?
  • In a DAGCircuit, there can be 100 op_nodes of just the same operation or 100 different operations occurring just once.
  • I initialized it as num_ops but if a different opinion exists, let me know in a comment.

Blockers

• [DAGCircuit Oxidation] Port DAGCircuit to Rust #12550

[coveralls]

Pull Request Test Coverage Report for Build 10690338962

Warning: This coverage report may be inaccurate.

This pull request's base commit is no longer the HEAD commit of its target branch. This means it includes changes from outside the original pull request, including, potentially, unrelated coverage changes.

• For more information on this, see Tracking coverage changes with pull request builds.
• To avoid this issue with future PRs, see these Recommended CI Configurations.
• For a quick fix, rebase this PR at GitHub. Your next report should be accurate.

Details

• 72 of 72 (100.0%) changed or added relevant lines in 3 files are covered.
• 169 unchanged lines in 13 files lost coverage.
• Overall coverage increased (+0.03%) to 89.171%

---

Files with Coverage Reduction	New Missed Lines	%
qiskit/circuit/commutation_checker.py	1	94.44%
qiskit/synthesis/two_qubit/xx_decompose/decomposer.py	1	95.42%
qiskit/quantum_info/operators/symplectic/random.py	2	95.0%
qiskit/transpiler/passes/optimization/commutative_inverse_cancellation.py	2	96.92%
crates/qasm2/src/lex.rs	4	91.98%
crates/circuit/src/packed_instruction.rs	4	95.19%
crates/accelerate/src/synthesis/linear/utils.rs	8	94.83%
qiskit/circuit/library/data_preparation/pauli_feature_map.py	9	91.0%
crates/circuit/src/dag_circuit.rs	9	88.87%
crates/qasm2/src/parse.rs	12	96.69%

Totals
Change from base Build 10638423704:	0.03%
Covered Lines:	72524
Relevant Lines:	81331

---

💛 - Coveralls

[mtreinish]

#12812 might be useful for you on this. This was me adding basically the same interface to the python space constructor for DAGCircuit. We get more flexibility with it in Rust including for things like the internal caches so I probably will rebase that PR on top of this after this merges (and add a python interface for the alternative constructor).

[qiskit-bot]

One or more of the following people are relevant to this code:

• @Qiskit/terra-core
• @kevinhartman
• @mtreinish

[mtreinish]

Overall this LGTM it's a straightforward addition to the rust api for DAGCircuit. Just one inline comment where I think one argument is in the wrong position.

[mtreinish]

LGTM, thanks for the quick update