build relationships 'ICDO to schema' for 'NULL-XXX' scheme

[dimshitc]

after we done #67
we have to add 'NULL-XXX' to NAACCR Schema

[denyskaduk]

If we look only at site, we see that one 'ICDO Condition' can have multiple schemas. For example: C50.9 - 'Malignant neoplasm of breast of unspecified site' can have 5 schemas:

schema_id	relationship	concept_code	concept_name
breast	Schema to ICDO	NULL-C50.9	Neoplasm defined only by topography: Breast, NOS
myeloma_plasma_cell_disorder	Schema to ICDO	NULL-C50.9	Neoplasm defined only by topography: Breast, NOS
kaposi_sarcoma	Schema to ICDO	NULL-C50.9	Neoplasm defined only by topography: Breast, NOS
heme_retic	Schema to ICDO	NULL-C50.9	Neoplasm defined only by topography: Breast, NOS
lymphoma	Schema to ICDO	NULL-C50.9	Neoplasm defined only by topography: Breast, NOS

We have to choose, what to do in this case. All 'NULL-XXX' conditions have 5 or more schemas. If we leave these relations as is, better don't keep it all. We tend to keep only schema which is site identification, as we do with 'Neoplasm, uncertain whether benign or malignant', for example '8000/1-C50.9':

schema_id	relationship	concept_code	concept_name
breast	Schema to ICDO	8000/1-C50.9	Neoplasm, uncertain whether benign or malignant of breast

Those 'NULL-C50.9' will have such relationship:

schema_id	relationship	concept_code	concept_name
breast	Schema to ICDO	NULL-C50.9	Neoplasm defined only by topography: Breast, NOS

[dimshitc]

@mgurley @cgreich @rimusia
please confirm the logic of schema definition for NULL-XXX concepts.
We link them to organ-specific schema only.
i.e.
NULL-C50.9 -> breast
(Technically NULL can mean Lymphoma as well, but we're omitting this).

[rtmill]

As discussed today on call, this is no longer needed.