gh-76187: Document the c typecode for multiprocessing.Array.

[tomasr8]

I linked to the ctypes types table and added this sentence:

In addition, the c typecode is an alias for :class:ctypes.c_char.

The rest of the diff is just wrapping.

• Issue: multiprocessing.Array 'c' code is not documented #76187

---

📚 Documentation preview 📚: https://cpython-previews--132504.org.readthedocs.build/

[gpshead]

Could you also mention that the w typecode of array is not supported? (glancing at the code, it is missing from Lib/multiprocessing/sharedctypes.py)

[tomasr8]

Could you also mention that the w typecode of array is not supported? (glancing at the code, it is missing from Lib/multiprocessing/sharedctypes.py)

Will do. Do you think it's worth to improve the error message for invalid typecodes as well? TypeError: this type has no size is a bit cryptic imo. Compare with the array.array error:

>>> array('x')
...
ValueError: bad typecode (must be b, B, u, h, H, i, I, l, L, q, Q, f or d)

(And I just noticed that the error message doesn't include 'w' 😆 )

[gpshead]

Yeah, improving the error message and/or even adding 'w' support are all also viable options. I expect this part of the code was inherited from the original upstream project multiprocessing came from and likely hasn't been meaningfully touched since? It's a dark corner.

[tomasr8]

Updated to mention that 'w' is not supported.

improving the error message and/or even adding 'w' support are all also viable options.

I'll send a patch to improve the error message and align it with that of array.array. I'm not that familiar with this module so I'm not sure how useful it'd be to add support for 'w'. I'll check how many changes that would require.