TODO
====

* Add a web interface.

* Add a REST interface.

* Add tests.

* Search (with restriction on ranks).

* Convert taxid to/from scientific name and output results as a table.

* Use file as input.

* Use stdin as input.

* Add support for `accession2taxid`.

* Add info about the database (number of nodes, species, time of last update, size in Mb).

* Remove nodes cloning in the code.

* Add a benchmark (because why not).

* Add a documentation, to clean up the README.