allow mbld and fmc to auto assign number of sets as per WCA

[pranav-027]

Autoassigning till the number of sets mentioned on WCA wasn't implemented for FMC and MBLD.

This will avoid the repetitive sanity check failure of [Consistency of Results & Rounds Data] Inconsistent scramble counts in Scrambles table and rounds table

I have also explained the bug in the email thread "Bug in scramble matcher."

[dunkOnIT]

LGTM and makes sense - @gregorbg you happy if we merge?

[gregorbg]

I need some more context on this slice. Without any real-world context, I cannot immediately see what it does or why it doesn't hurt existing events that aren't (usually) split into attempts.

[pranav-027]

Hello @gregorbg,

The scramble matcher has an auto-assign function that automatically assigns scrambles to their respective rounds. Delegates typically generate more scrambles than the actual number of sets required, and previously, the auto-assign function would include all scrambles. This slice was added in this #139 PR to match the scrambles up to the number of scrambles mentioned on WCA.

This PR did not handle the logic for FMC and MBLD, which are split into multiple attempts. This PR extends the same logic to those events.

The issue is that delegates assume the auto-matching works consistently across all events. As a result, they often submit the wrong number of scrambles for MBLD, thinking that editing the number of sets on the WCA website is enough. And scramble matcher will take care of the rest. This mismatch gets ignored because of a results validation issue and only gets caught during monthly sanity checks. Addressing this within the Scramble Matcher itself will prevent such errors from occurring in the first place.

CC: @thewca/wrt

[pranav-027]

ping @gregorbg

[gregorbg]

LGTM, thanks for your detailed explanation! By any chance, is this related to the per-attempt work in #16?

Anyhow, I'd like you to take a look at that other PR too, because you clearly have the deeper understanding of auto-matching