enhancement!: CLI overhaul

[jsstevenson]

close #293
close #217
progress on #298

stuff I did:

• reorganize the CLI stuff to more closely match unix tool/click patterns. Instead of a bunch of different entry points, use a common one (gene-normalizer) and then different actions are commands (eg gene-normalizer update).
• For the update command, sources are now arguments, not an option. Since the number of sources can be 0 to n, I think it makes more sense to be used as an argument instead of a param passed to an option (avoids the quoting, spacing etc stuff).
• add a --silent option to suppress feedback to console (false by default) in cases like pipelines that process stuff going to stdout. Addresses Database endpoint usage feedback #217 with a very stringent check (value must be True, can't just be truthy, so no accidental silencing - hence the noqa).
• Some cleanup on CLI descriptions. Also add a CLI autosummary to docs instead of maintaining a separate description -- there's some fun (sort of janky) stuff done to make it look okay in both.
• Refactor a lot of the internal CLI update methods to a public module, gene.etl.update. That way they can easily be used programmatically (this is something that Thera-Py does with the Disease Normalizer, for example).

The updated docs page looks like this: https://gene-normalizer--299.org.readthedocs.build/en/299/managing_data/loading_and_updating_data.html

And this is what the help looks like for the update command:

Usage: gene-normalizer update [OPTIONS] [SOURCES]...

  Update provided normalizer SOURCES in the gene database.

  Valid SOURCES are "HGNC", "NCBI", and "Ensembl" (case is irrelevant).
  SOURCES are optional, but if not provided, either --all or --normalize must
  be used.

  For example, the following command will update NCBI and HGNC source records:

      $ gene-normalizer update HGNC NCBI

  To completely reload all source records and construct normalized concepts,
  use the --all and --normalize options:

      $ gene-normalizer update --all --normalize

  The Gene Normalizer will fetch the latest available data from all sources if
  local data is out-of-date. To suppress this and force usage of local files
  only, use the –use_existing flag:

      $ gene-normalizer update --all --use_existing

Options:
  --all           Update records for all sources.
  --normalize     Create normalized records.
  --db_url TEXT   URL endpoint for the application database. Can either be a
                  URL to a local DynamoDB server (e.g.
                  "http://localhost:8001") or a libpq-compliant PostgreSQL
                  connection description (e.g. "postgresql://postgres:password
                  @localhost:5432/gene_normalizer").
  --aws_instance  Use cloud DynamodDB instance.
  --use_existing  Use most recent local source data instead of fetching latest
                  version
  -s, --silent    Suppress console output.  
  --help          Show this message and exit.

[jsstevenson]

@korikuzma what would you think about shortening the options to --all and --normalized? eg

gene-normalizer update --all and gene-normalizer update --normalized HGNC NCBI ensembl

[korikuzma]

@korikuzma what would you think about shortening the options to --all and --normalized? eg

gene-normalizer update --all and gene-normalizer update --normalized HGNC NCBI ensembl

@jsstevenson to confirm, all would be all sources and concept group creation? I'm not sure about the name normalized for the sources.

[korikuzma]

🚀