3. Mergeability

CONTEXT

• Data defined in different formats, in different languages, different field names, field value types and measures at hierarchically different levels, is hard to merge and align.

DEFINITION

• Two or more of data records are fully or partially mergeable if their metaheaders define schemas with fully or partially overlapping concepts and their value types (e.g., distance#float64-meters) mapped to their field names.

TEST CONDITIONS

• Suppose you have a data record like this:

  <person><first_name>Neil</first_name><last_name>Armstrong</last_name><properties><age>15</age></properties></person>
  

  and want to merge it with a data record like this:

  {"name": "Hello", "surname": "Kitty", "age_in_days": 12.5}
  

If each data piece has a metaheader, they can be automatically aligned to produce a list of records (or a table) of desired properties (or fields), regardless of different nesting in each source.