dump-binaries: Using dynamic tables provided by WMO (#4)

* lazy loading tables

* table f3 improvements

* Implementing Display for F3

* feat: Section 3 display Descriptor child

* refactor: Avoid pub when able to

* feat: Implementing iter() for F3

* wip

* wip

* implement show-data

* width8 and width4 tests

* remove magic

* fix: Fixing authors attribute

* GA to build binaries when BUFR-dump is released

* Adding Tables B

We might do something dynamic later.

* Cleaning redundant print

* Adding more units (incomplete)

* Adding more units

* Adding more units

* Limiting to working table B

Need some work to allow missing tables (B type).

* Adding missing TableD and updating

Note a new field: noteIDs

* Loading missing tables D

* Improving syntax

* Updating codecov-action

* Publish (dry-run) each crate

* Missing Luiz!

* Bumping versions

* Adding some metadata to the crates