Support querying for blocks and transactions

[0x009922]

In Iroha Block Explorer, we need to show up to a user the list of latest blocks and transactions. Also it would be useful to query for some specific block/transaction by its id, height and hash relatively.

Thus, it would be nice to extend Query API with the following:

• FindAllBlocks, that returns a list of blocks. With pagination, of course. Ordered by height desc.

  There are several life stages of a block:

  1. pending
  2. chained
  3. valid
  4. committed

  AFAIK in explorer we only need pending and committed blocks, and we need to distinguish them from each other.

• FindBlockByHeight, that returns a single block if it has been found. I think we need to know the block status here as well.

• FindAllTransactions, that returns a list of latest transactions, ordered by timestamp desc. The most debatable query.

  I am not sure that this query is a good idea, because actually transactions are always contained within blocks, and its enough to query for blocks only to get transactions as well. But if we want to show the list of transactions decoupled from their blocks, how to paginate this data independently from blocks? Maybe we should drop the idea of "transactions list" decoupled from blocks.

• FindTransactionByHash, that returns a single transaction if it has been found. This is not as debatable as FindAllTransactions is, but I think it might be really hard to query for it in an efficient way. Am I right?

[appetrosyan]

• FindTransactionByHash is there since GaroRobe/issue1485 #1517.
• FindAllTransactions has some overlap with FindAllBlocks. Still, because of the difference in pagination and how it would be a nightmare to implement FindAllTransactions given only a paginated FindAllBlocks I think we should implement both.
• FindBlockByHeight is redundant. Pagination has a lower and upper bound. FindAllBlocks can be used to get blocks of height n to n+1 and it would have the same effect.

[appetrosyan]

@QuentinI If you're looking for something new to implement, seems like a good place to start

[s8sato]

Does block stream not work for it?

[QuentinI]

Regarding FindAllBlocks: structures describing blocks are in defined in iroha_core, which depends on iroha_data_model where the queries need to be defined.
As far as I understand, iroha_data_model doesn't concern itself with blocks at all at the moment. Do we actually want to change that, or is /block/stream API endpoint @s8sato mentioned sufficient here? My take would be that /block/stream should suffice if we need this information only for Block Explorer.

[appetrosyan]

@0x009922 You're probably the best person to explain what's missing in the block streaming API.