Sequencer pulls L1 -> L2 messages from archiver

[LHerskind]

The sequencer should query the archiver to get L1 to L2 messages to include in its next L2Block.

• The archiver must implement an API that lets sequencer fetch messages for an L2block e.g., getL1ToL2Messages(numberOfMessages)
• The sequencer must use the API to fetch messages to populate its L2block, and insert them into the block.

As the number of messages held by the archiver might be larger than the number of messages that can be put into a single L2block, consider how to prioritise the values:

• Minimum gas -> Prioritise duplicate messages
• Max fee -> Simple prio by fee
• Fifo -> based on time of event.

[LHerskind]

Depends on #520