Disaggregate codelist distributions by field

[jpmckinney]

Duncan: I want to understand coverage of documents by documentType and section, e.g. a table like

type	Planning	Tender	Award	Contract	Implementation
tenderNotice	#	#	#	#	#
contractNotice	#	#	#	#	#

Duncan: I want to understand coverage of milestones by type and section, for example by using a table like

type	Planning	Tender	Award	Contract	Implementation
approval	10,278
delivery					2,546
financing					10,278

Duncan: For both of the above, since document type and milestone type are open codelists, we should include any extra codes used by the publisher.

James: the logic can be that all fields using the same codelist form the X axis. We would need some custom logic at the presentation layer to abbreviate the JSON paths to e.g. "Planning"

[jpmckinney]

It was also noted that a shorter table could be included on the overview page:

Duncan: I’d like to know whether documents are being published, for example by providing a count of documents for each section in OCDS (+1: MM, JM, CP, YL)

i.e. instead of a breakdown by documentType code as above, it'd just give the total number of document objects for each stage/section.

[jpmckinney]

Transferred to pelican-frontend, though changes to pelican-backend might also be required.