refactor: allocate smaller tuples for query streams

[jyeshe]

What

• Unify scope deserialization by separating key boundary from the gen or txi scope
• Remove use of max_blob for all /v2/contract/calls and /v2/contract/logs except for data log filtered one
• Remove use of max_blob for name and auction prefixing
• Leverage Erlang term comparison to allocate less memory on query key boundaries
• Adds unit test coverage for /v2/contract/calls and /v2/contract/logs

Why

• Allocate less memory by replacing max_256bit_int with nil
• max_blob is a 256KB

Notes

• Gladly the BEAM doesn't reallocate every time the max_blob
• Analysis of NFT token id with negative values will be done in a separte PR

[jyeshe]

Adding test cases for contract logs and calls ...