Tags

[hotfix] fix encoded text byte count

- Add text data type
    - only UTF-8 encoding is supported for now

- [tech] cleaner data module code
    - separate data type implementations
    - separate data type unit tests
- Now only one constructor for the UnsignedInteger data type
    - always have to specify min and max value (which are already typed as u64, so no mistake can be made here concerning the capacity)
    - no more error needed for max=0 => already handled with min=max or max>min

- Add floating point number data type, with two possible precisions
    - single precision : like Rust's f32
    - double precision : like Rust's f64

- Tables and columns
    - tables : have columns, max column count is 64
    - columns : have name and data type and optionality (columns are either optionally or mandatorily valued in rows)
- Boolean and unsigned integer data types
    - unsigned integers : have min value and max value, max capacity is u64 : 8 bytes
- Encoding row to save
    - columns presence is encoded in a 64-bit leading section called "optionality flags"
    - columns values are encoded into bytes, byte count depending on the data type
- Decoding saved row
    - reverting the encoding to fetch the values that have been saved (lossless)

- Misnaming fixed