Validity check api

[aglavic]

As pointed out by @andyfaff in #143 it can be hard to validate the Orso header data and find out which attributes are missing/wrong. This adds a check_valid method to each Header derived class that can be used to check a dictionary for correctness and get detailed information on the issues if it is invalid.

The returned class includes a list of any missing or invalid attributes as well as sub-attribute details. with ORSOValidationResult.get_report() the results are summarized so users of the API can quickly fix what is missing.

Examples:

from orsopy.fileio import *
print(
    Person.check_valid({'name': 'Test User', 'affiliacion': 'Test'}).get_report()
)
> Dictionary was invalid for Person!
>    Reasons for classification:
>    User parameters not part of specification
>      affiliacion
>    Required parameters not provided
>      affiliation

print(
    InstrumentSettings.check_valid({
        'incident_angle': {'magnitude': 4.0, 'unit': 'deg'}, 
        'wavelength': {'min': 2.0, 'max': 12.0, 'unit': 12}, 
        'polarization': 'unpolarized'}).get_report()
)
>  Dictionary was invalid for InstrumentSettings!
>    Reasons for classification:
>    Parameters with invalid type or attributes
>      wavelength
>      ORSO parameter wavelength extra information:
>        Dictionary was invalid for ValueRange!
>          Reasons for classification:
>          Parameters with invalid type or attributes

This should work arbitrarily deep, as illustrated by putting the previous result into a DataSource:

Dictionary was invalid for DataSource!
  Reasons for classification:
  Parameters with invalid type or attributes
    measurement
    ORSO parameter measurement extra information:
      Dictionary was invalid for Measurement!
        Reasons for classification:
        Parameters with invalid type or attributes
          instrument_settings
          ORSO parameter instrument_settings extra information:
            Dictionary was invalid for InstrumentSettings!
              Reasons for classification:
              Parameters with invalid type or attributes
                wavelength
                ORSO parameter wavelength extra information:
                  Dictionary was invalid for ValueRange!
                    Reasons for classification:
                    Parameters with invalid type or attributes
                      unit

[andyfaff]

We've had the capability of validating against the schema for a while, via fileio.base._read_header_data. I guess this adds validation to each type of Header?

[aglavic]

Yes, this is the idea and also to have better feedback. There also was no public API for that header validation and it did not seem to work in all cases that I tried using the schema.