Do not throw 409 errors on duplicate statements in POST about ralph

[wilbrdt]

It is indeed a grey area of the LRS specification, and I understand that it can be a pain to work around it.
I think removing the duplicate statements from the batch and continue processing would be a logic thing to do.
I think it would require to implement a deep comparison of statements to cover the following cases:

• If LRS receives a statement with an ID it already has a statement for, and statements don't match: don't do anything, return a 409 Conflict
• if LRS receives a statement with an ID it already has a statement, and statements match: remove statement from batch and continue posting the batch.

This deep comparison would come with a performance cost (at least until pydantic v2 is out 🔜), but would be mitigated by only checking the duplicate ids.

However, I'm not so sure about adding the found duplicates into the response details, as we already return the array of statements that has been posted. What would you suggest?

[bmtcril]

I think the deep comparison would be good once it can be done, though even that is complicated in the spec since some fields should not be considered in the comparison. I forgot that the array of statements was returned, that is totally sufficient!