Qiskit / qiskit-ibm-runtime / 27264001633

87%

push

github

Reintroduce support for multiple result decoders (#2899)

<!--
⚠️ The pull request title should be short detailed and understandable
for all.
⚠️ Also, please add a release note file using reno if the change needs
to be
  documented in the release notes.
⚠️ If your pull request fixes an open issue, please link to the issue.

✅ I have added the tests to cover my changes.
✅ I have updated the documentation accordingly.
✅ I have read the CONTRIBUTING document.
-->

### Summary

Revise the methods that accept `result_decoder` to ensure they also
support sequences of decoders (ie. `type[ResultDecoder] |
Sequence[type[ResultDecoder]] | None = None`) consistently, and update
the logic so the decoders are called in chain in that case.

In practice, it re-introduces the ability of using multiple decoders,
with provides for some degree of easy customizability by end users. It
was introduced around #1450 for dealing with the websocket mechanism,
and some remains remained at the interface level although in practice
[it was restricted to one
decoder](https://github.com/Qiskit/qiskit-ibm-runtime/blob/0.47.0/qiskit_ibm_runtime/base_runtime_job.py#L110-L113)
out of two.

Additionally, it avoids passing the decoder explicitly from the
Primitive classes, as `DEFAULT_DECODERS` is used under the hood for
selecting a default (making declaring and passing them from Primitives
unneeded, and saving some import pressure).

### Details and comments

Related #2852 - not marking as fixed as this does not revise existing
decoders.

### AI/LLM disclosure

- [x] I didn't use LLM tooling, or only used it privately.

10 of 10 new or added lines in 3 files covered. (100.0%)

7931 of 9122 relevant lines covered (86.94%)

9.5 hits per line