Gff3toddbj

* Add `genbank-to-ddbj` for Genbank → DDBJ conversion
* Add `compare-ddbj`s for comparing two DDBJ annotations for evaluation
* Add a non-SO wild items to the default config for feature and qualifier names
* Update the default config: replace hyphens in qualifier keys with underscores.
* Fix: Respect feature-wise /transl_table value than the globally-set one.
* Fix an error when joined features have .sub_features
* Fix [source] in metadata not used when "source" feature exists in a entry
* Change not to join features directly under "gene"
* Fix the default config for feature and qualifier names around /pseudo
* Fix a bug getting incorrect codon when CompoundLocation has a part < 3bp.
* Fix start codon detection: no longer change /codon_start value even when !=1
* Fix stop codon detection
* consider stop codon does not exists whenever the length
(after subtraction by "phase") is not a multiple of 3.

* Expand the default config for feature and qualifier names
* Now based on Sequence Ontology
* Create bgzip file as a new file, rather than replacing existing FASTA file

* Switch to pysam and samtools to index FASTA (called faidx)
* No longer creates sqlite3 database when reading FASTA
* Minor fixes

* Rename the CLI option from --rename_setting to --config_rename
* Rename the CLI option form --filter_setting to --config_filter
* Update the feature-qualifier config for renaming
* Minor bugfixes

* Add the CLI option --version

* Improve memory usage when reading a large FASTA file
* Add CLI options: --rename_setting and --filter_setting to load custom settings
* Fix locations with inequalities '<' and '>' when start/stop codons are absent
* Fix a critical bug not inserting source properly
* Fix a critical bug missing subfeatures in certain cases
* Fix a critical bug ocassionally getting mRNA's ID wrong.
* Update the translate-feature renaming table
* Update the translate-feature pair filtering
* Add CLI tool: normalize-entry-names for renaming IDs in DDBJ annotation file
* Many bugfixes