Update mexico_plataforma_digital_nacional spider

[yolile]

https://www.plataformadigitalnacional.org/contrataciones used to have a link to a Google Drive JSON file. Now they point directly to https://datos.gob.mx/busca/dataset/concentrado-de-contrataciones-abiertas-de-la-apf. However, they are also publishing data from other data source now "Secretaría Ejecutiva del Sistema Estatal Anticorrupción de Aguascalientes".

And now, they have an underlying API, for example:

curl 'https://api.plataformadigitalnacional.org/s6/api/v1/search?supplier_id=SHCP' \ -X 'POST'

To get the list of releases by publisher, so we could update the spider to use that API endpoint instead of the Google Drive file.

The only thing to decide is if we want two spiders, one per publisher (e.g., one for SHCP and another one for Aguascalientes) or just one for Plataforma Digital Nacional as of now.

[jpmckinney]

How does PDN disambiguate the two publications? Is its CompraNet data identical to https://datos.gob.mx/busca/dataset/concentrado-de-contrataciones-abiertas-de-la-apf ?

[yolile]

How does PDN disambiguate the two publications?

In the front end, they have a select where you can select the publisher (or, as they call it, "data provider")

In the API, they have the "supplier_id" query param that accepts SCHP and SESEA_AGS as values.

Is its CompraNet data identical to https://datos.gob.mx/busca/dataset/concentrado-de-contrataciones-abiertas-de-la-apf ?

I don't know, but I can check. Edit, in their FAQ section they say:

¿La PDN va a generar información? No. El objetivo de la PDN es generar interoperabilidad entre los datos que generan los sujetos obligados y es una herramienta de consulta.

So they say they don't generate information, only create interoperability between the different entities, and the platform is only a "query" tool. So I guess we can assume their data from Compranet is identical to what Compranet publishes.

[jpmckinney]

Okay, so let's only add Aguascalientes to Collect and the data registry, since we don't have the upstream source (unless we can find out where it is).

We can add their SHCP data to Collect, but I'm not sure who would want it in favor of the CompraNet data.

[jpmckinney]

I don't know, but I can check

I'd be curious if the ocid is the same, and if so, what ocid is used for Aguascalientes.

[yolile]

I'd be curious if the ocid is the same, and if so, what ocid is used for Aguascalientes.

For Compranet, the prefix is the one registered for SHCP ocds-07smqs, and for Aguascalientes, the one registered for Aguascalientes ocds-ty10ed (we registered that one in the past, so maybe they are sending their data directly to the platform and not publishing it themselves? Aha! from CRM-8104 they said, "we developed the open contracting system for the Plataforma Digital Nacional" so they are indeed only publishing their data through PDN, so new publisher alert? But who? Aguascalientes, I guess?)

[jpmckinney]

Yes, I'd say Aguascalientes. We haven't had a scenario where the publishing platform is independent of the data's author/steward, but I think we want to be tracking the author (despite the field being publisher).

[jpmckinney]

We can maybe add a note to the docs for the CompraNet spider, to indicate that it's also published via PDN. Not sure how useful that fact is for the data registry, but we can also add it to the end of the description.

[yolile]

Ok, so I assume we want one spider for each. And if they add more publishers in the future will add new spiders as we did with Digiwhist, right?

[jpmckinney]

Ok, so I assume we want one spider for each.

I meant the existing CompraNet spider (APF) can just have a note added. We don't need a CompraNet spider for PDN.

Ok, so I assume we want one spider for each. And if they add more publishers in the future will add new spiders as we did with Digiwhist, right?

Yup! Assuming we don't have access to the original source.