Fix display of registerless bits in all drawers and change display of one-bit registers

[enavarro51]

Summary

Fixes #6732
Fixes #5784

Details and comments

This PR fixes an issue where displaying circuits containing registerless bits would cause a failure of the 'mpl' and 'latex' circuit drawers. It also changes the display format in the 'text' drawer.

In addition, one-bit registers are now displayed in all drawers without a subscript 0. This standardizes the text formatting for 'latex' and 'mpl' drawers. All register names are now in italics, whether one-bit or more. Only the cregbundled registers are displayed in normal formatting.

In approaching this change, it became clear that the code for choosing a register name/number format was significantly different in the 3 drawers. Prior to this there were 4 if clauses which were done in 3 different orders, and this PR added 2 additional if clauses. It seemed prudent to standardize this selection process across the 3 drawers.

There is now a new function in utils.py called get_bit_label which returns the text that will be displayed for the registers and bits for all 3 drawers. This greatly simplified this code section in the 3 drawers and will make maintenance of register and bit names easier in the future.

There were quite a few reference image changes, but were mostly minor font or spacing changes in bit labels. Reviewers should still check these out for possible unacceptable changes.

qrx = QuantumRegister(2, "qrx")
qry = QuantumRegister(1, "qry")
crx = ClassicalRegister(2, "crx")
circuit = QuantumCircuit(qrx, [Qubit(), Qubit()], qry, [Clbit(), Clbit()], crx)
circuit.draw(cregbundle=True)

       
qrx_0: 
       
qrx_1: 
       
    2: 
       
    3: 
       
  qry: 
       
    0: 
       
    1: 
       
crx: 2/

[javabster]

This is looking good, just a few minor comments. Moving generic code out into the utils file makes the drawer files much more readable, would love to see more enhancements like this in future!

[javabster]

Could you add to the docstring what each of the arguments are please?

[enavarro51]

Done in 8a7c24a.

[javabster]

This is a v minor comment but I think it would be better to split this into 2 bullet points (one for each issue that was fixed), see code snippet example here in the contributing guidelines

[enavarro51]

Done in 8a7c24a.

[javabster]

This big nested if statement is kind of ugly but I understand it's partly because of messiness from before. At some point I think this should be refactored, but perhaps that is beyond the scope of this PR

[enavarro51]

Couldn't help myself. Refactored in 8a7c24a. Let me know what you think.

[javabster]

LGTM 🚀