diag-kernel: thread base offset through structure DOP decoding

[lorinciubeica10]

Summary

Replaced push_slice/pop_slice coordinate shifting in map_structure_dop_from_uds with a base offset parameter propagated through all decode functions.
This correctly handles both absolute and relative byte position conventions without corrupting the payload cursor for inner params

Checklist

• I have tested my changes locally
• I have added or updated documentation
• I have linked related issues or discussions
• I have added or updated tests

Related

Fixes #270

Notes for Reviewers

[lorinciubeica10]

Cargo.lock needed because of build_and_test failure:

[theswiftfox]

Cargo.lock needed because of build_and_test failure:

Thanks! :)
I created #284 a minute ago, as I stumbled across the same issue in my PR ^^
Feel free to 👍 or 👎 on the issue if you have any thoughts about it :)

[andreirusu-aumovio]

Thanks @lorinciubeica10 ! I can confirm that this PR solves one of the issues I encountered while extending ODX for DTC Snapshot/Extended data where the response would be wrongly interpreted:
59 04 01 is actually the Positive Service response + Subfuction + First Byte of DTC instead of the actual DTC of 01 E2 40

[theswiftfox]

question: Do you think it would make sense to introduce a struct that wraps the &datatypes::Parameter together with the base_offset?
Something like this:

struct ParamContext<'a> {
    parameter: &'a datatypes::Parameter,
    base_offset: usize,
}

My train of thought being that these two are related to each other, and combining them into a single parameter would avoid the clippy warning at the same time and maybe we have additional context we want to pass with the parameter in the future.

This is more of a discussion point, where I'd like your opinion on it as well. If you prefer it as is that is perfectly fine to me as well.

[lorinciubeica10]

I think it s a good idea since base_offset is always passed alongside param. I will fix it

[lorinciubeica10]

Added several fixes for bugs caused by incorrect or missing data when decoding UDS responses:

• Parameters inside Structure DOPs were decoded using relative byte_position
  instead of absolute position in the full response payload.
• map_dynamic_length_field_from_uds computed the wrong item count when a param
  had no explicit byte_position.
• EnvDataDesc decoding used an incorrect base_offset, causing the discriminator
  lookup to fail and snapshot data to be silently dropped.
• Extended data and snapshot requests always tried both FaultMem and UserMem
  service types; they now select the correct one based on the scope parameter.