Move Option<Box<_>> into ExtraInstructionAttributes.

[kevinhartman]

Summary

The primary motivation here is to encapsulate the hairy logic of managing memory for extra attributes. By moving the Option<Box<_>> into the ExtraInstructionAttributes struct itself, clients are no longer burdened with the responsibility of checking if the Option contains a value, unwrapping it etc.; the semantics of that aren't relevant to clients like DAGCircuit and CircuitData. Instead, the ExtraInstructionAttributes acts as a container for optional attributes, and can easily drop the Box internally if all attributes are cleared.

A give-away that this pattern could be helpful was ExtraInstructionAttributes::new's return type, which was previously Option<Self>.

Details and comments

The size of ExtraInstructionAttributes (now held by value by instructions) should be the same size as the old Option<Box<ExtraInstructionAttributes>>, so this should not have memory implications regarding the size of instructions.

Note that ExtraAttrs is completely private to the circuit_instruction module, i.e. it's only an implementation detail and in no way part of the interface.

[qiskit-bot]

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

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

[coveralls]

Pull Request Test Coverage Report for Build 10837025267

Details

• 107 of 164 (65.24%) changed or added relevant lines in 8 files are covered.
• 21 unchanged lines in 6 files lost coverage.
• Overall coverage decreased (-0.02%) to 88.89%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
crates/circuit/src/packed_instruction.rs	2	3	66.67%
crates/circuit/src/dag_node.rs	5	7	71.43%
crates/circuit/src/circuit_instruction.rs	54	108	50.0%

Files with Coverage Reduction	New Missed Lines	%
crates/circuit/src/dag_circuit.rs	1	88.54%
crates/qasm2/src/expr.rs	1	94.02%
qiskit/transpiler/passes/synthesis/unitary_synthesis.py	2	88.26%
crates/qasm2/src/lex.rs	4	92.23%
crates/qasm2/src/parse.rs	6	97.61%
qiskit/synthesis/two_qubit/xx_decompose/decomposer.py	7	90.91%

Totals
Change from base Build 10834414649:	-0.02%
Covered Lines:	73365
Relevant Lines:	82535

---

💛 - Coveralls

[jakelishman]

Thanks for doing this, it looks like a very sensible idea.

[jakelishman]

Thanks for this and the changes. I'll leave unmerged because I saw elsewhere that Ray was going to look too.

[raynelfss]

This is all sensible to me, my only comment was about the usage of a map-like struct instead of a simple tuple struct but it seems it's already been corrected :)

This should be good as is. Thanks for working on it! 🚀