[BUG] Identifier numbering is not clear

[chrisdburr]

Issue

The numbering for the element identifiers does not appear to follow a consistent logic.

As you can see in this image, the child elements of S1, which were the first to be added start at P6.

While not shown on this image, P1...P5 also exist and are located under S4.

For some reason, the strategy elements also appear to have moved around (as shown below). Their numbering indicates the order in which they were created, but this is not reflected in their left-to-right ordering.

Desired Behaviour

Sequential numbering of identifiers should be consistent, and ideally based on the order in which elements are added.

[chrisdburr]

Another example. The following strategy elements were added in this order:

• Argument over system safety
• Argument over component safety
• Argument over fair impacts

Initially, the numbering remained consistent until Argument over fair impacts was added. Then, the numbering was automatically revised.

[cptanalatriste]

@chrisdburr just identified the root cause and prepared a PR with the fix: https://github.com/alan-turing-institute/AssurancePlatform/pull/496

I'm deploying the fix to staging right now. Please verify if now it's behaving as expected.

[chrisdburr]

Initial testing on my end seems to have fixed the bug. Please could I request some more thorough testing on this though. @kallewesterling do you have suggestions for a more systematic way of testing this?

[cptanalatriste]

@chrisdburr due to the root cause (a random ordering on rows when performing the query) testing this one manually it's tricky. I can write an automated unit test, so it can flag if the issue reappears.

[chrisdburr]

Thanks, @cptanalatriste. Whatever method you think is best I am happy to go with.

[cptanalatriste]

Please verify in staging.

[chrisdburr]

This seems to be working. However, we will need to update this when we add the feature to refresh identifiers across the case.

Please link to a PR when closing.

[cptanalatriste]

The original fix was part of PR: https://github.com/alan-turing-institute/AssurancePlatform/pull/496/files

The automated test was added in PR: https://github.com/alan-turing-institute/AssurancePlatform/pull/508/files

Closing as agreed with @chrisdburr