Niklas Saari

Actually error reporting isn't that good. Might be better to give the failed value also in the error message like in other codecs

Related to this https://github.com/librasn/rasn/issues/185, maybe it can be handled in the same time.

I was just finishing it, some small cleaning still to do. I unified all the string types, so that they implemented `StaticPermittedAlphabet`. Also for `Teletex` type while the characters itself...

Invalid restricted string error, for example, should look something like this now: ``` ---- coer::tests::invalid_ia5_oer stdout ---- [src/coer.rs:697:9] e = DecodeError { kind: PermittedAlphabetError { reason: InvalidRestrictedString { source: InvalidIA5String(...

I think it is ready now. Not sure what is the performance since all characters on restricted strings are verified now in every codec.

For nested choices it currently uses conditionally `TagTree::tag_contains` call, which should be enough for that, and reduces the overhead from everywhere else. This happens on derived code rather than on...

Also note - if `&'static [Tag]` is valid and can be used, in many cases tags can also be sorted and we could use `binary_search` instead. On few variants there...

This is still on the list, I just have been a bit busy. Maybe we can do binary search part later.

This seems promising, maybe good for starters? https://github.com/benchmark-action/github-action-benchmark For the most benefits, it might need GitHub pages tho.

Would it be better if we default to `BigInt` internally instead? I think that would work similarly as earlier. It is just slower by default. Instead of type alias it...