fix(app): update specific analysis keys passed through CommandText with terminal run record

[ncdiehl11]

closes RQA-2645
closes RQA-2647

Overview

We now grab commands for terminal runs from the run record rather than the most recent completed analysis to give a more accurate representation of the commands actually executed during a run. However, this introduced some unintended downstream bugs in CommandText and its utilities for hashing commands by labware, pipette, and module IDs— the run commands' IDs no longer match those contained in the analysis's labware, pipette, and module values.

To address this, here, I overwrite those properties in the robot-side analysis that is passed to CommandText and all of its children/utilities so that the command param IDs can be properly hashed.

Test Plan

• run a protocol, like flex_MOAM_218.py linked here
• observe that the running protocol command text is formatted properly
• allow the run to finish or cancel the run
• observe that the terminal protocol commands go only up to the last command actually executed, and the text formatted properly, including all labware display names and locations

Changelog

• overwrite labware, modules, and pipettes data on analysis if run is terminal and run record data is not null

Review requests

@shlokamin per working on previous fix

Risk assessment

medium

[jerader]

should this be targeting edge or the release branch?

[shlokamin]

thank you!

[ncdiehl11]

should this be targeting edge or the release branch?

release, thank you!