Implement extractors for retrieving dataset information

[ricardogsilva]

Extracting information about datasets shall be done via dedicated extractors. An extractor needs to be able to retrieve the relevant details in order to allow creation of new DB items. This means being able to instantiate new objects of types:

• schemas.surveyrelatedrecords.RecordAssetCreate
• schemas.surveyrelatedrecords.SurveyRelatedRecordCreate
• schemas.surveymissions.SurveyMissionCreate

These will then be used to feed into the DB functions for creating new items, such as:

• db.commands.surveymissions.create_survey_mission()
• etc.

Extractors need to be configured in a somewhat dynamic way that takes into account the properties of the related records, which are mostly:

• DatasetCategory
• DomainType
• WorkflowStage

These are bootstrapped in

cliapp.bootstrapdata

The extractor needs to know about media types too. The below diagram depicts this combination:

This setup of extractors shall be configurable by means of an external file, for example a JSON array, as can be seen in the image above

[lcalisto]

@ricardogsilva to support gridded CF NetCDF extraction I'll probably need the gdal:ubuntu-full base image instead of gdal:ubuntu-small (the small ships without the netCDF driver). What do you think? I can try ubuntu-full, measure the image size increase, and if we later find a lighter alternative we switch.

[ricardogsilva]

Let's tackle simpler formats first and try to establish working patterns for extraction and generation of previews before dealing with things like this. The idea is to get something up and running with the already existing base first, and then improve upon it