Metadata normalizing tool

The purpose of this tool is to extract a defined set of parameter from raw metadata. It is meant primarily for use with geo-spatial datasets, but can be extend to process any kind of data.

Principle

Input: raw metadata attributes in the form of a dictionary

Output: a dictionary in which normalized parameter names are associated with values found in the raw metadata.

The actual work of extracting attribute values from the raw metadata is done by normalizers.

Each normalizer is a class able to deal with a particular type of metadata. In the case of geo-spatial datasets, a normalizer is typically able to deal with the metadata format of a particular data provider.

Usage

Although normalizers can be used directly, the easiest way to normalize metadata is to use a MetadataHandler. A metadata handler is initialized using a base normalizer class.

When trying to normalize metadata (using the handler's get_parameters() method), the handler tries all normalizers which inherit from this base class. The first normalizer which is able to deal with the metadata is used.

To determine if a normalizer is able to deal with a dictionary of raw metadata, the handler calls its check() method.

Example to normalize data for use in django-geo-spaas:

import metanorm.handlers as handlers
import metanorm.normalizers as normalizers

metadata_to_normalize = {
  'foo': 'bar,
  'baz': 'qux'
}

m = handlers.MetadataHandler(normalizers.geospaas.GeoSPaaSMetadataNormalizer)
normalized_metadata = m.get_parameters(metadata_to_normalize)