Create scripts/gethimport.py

[lithp]

What was wrong?

Nothing was wrong, per say, but it wasn't previously possible to get a full-syncing trinity node this far:

How was it fixed?

I wrote a script with a few helpful subcommands which can be used to create a trinity database, bootstrapping off an already-synced geth node.

To-Do

• Why is the build failing while building snappy? (needed for decoding ancient blocks)
• Appease lint and such
• Break into multiple PRs
• Remove the code which iterates over the state trie? It's too slow to be practical except for very early blocks.
• Support simulating a full sync, starting from genesis, by also importing the genesis state when creating a trinity database. This depends on moving the EIP1085 code over from trinity -> py-evm, so it probably doesn't belong in this PR.
• Figure out how to test any of this without committing large example geth databases to the repo.
• Clean up commit history
• Add entry to the release notes
• Find a cute animal pic
• Add a command to compact the database. If the database is not compacted it will compact itself,
  freezing your process for an uncomfortably long amount of time.
• Document how to use this script to build a trinity database to reproduce the above screenshot
• Remove import_block_flame.svg
• Apply some bug fixes I made on my server, but not locally.
• Rebase onto Improve header chain extension performance #1891
• Make sure this is modular enough that a wrapper script in trinity could call it

Cute Animal Picture

A hermit crab swapping out shells: