Fix parsing of "editorial group"

ronaldtse commented 2 years ago

From: https://github.com/ietf-ribose/relaton-data-3gpp/issues/5

@strogonoff:

editorialgroup seems like some custom extension outside of Relaton core but relevant for some citations?

The dash in the following example makes editorialgroup contents seem ill-parsed, example:

"editorialgroup": [
    {
        "name": "C1",
        "type": "prime"
    },
    {
        "name": "-",
        "type": "other"
    },
    {
        "name": "T2",
        "type": "former"
    }
],

Example items:

https://demo.bibxml.org/api/v1/ref/3gpp/TR_33.879_REL-13_0.1.0/
https://demo.bibxml.org/api/v1/ref/3gpp/TS_29.163_REL-9_9.21.0/
https://demo.bibxml.org/api/v1/ref/3gpp/TS_27.007_REL-15_15.3.0/ (extended “release” value)

This is actually the meeting ID from the database, it is not THE editorial group itself, but it is editorial group information. It shouldn't be broken down, e.g. "C1-T2" is correct without splitting it.

andrew2net commented 2 years ago

@ronaldtse they have already split in DB:

{:Type=>"TR",
 :Number=>"33.879",
 :"For publication"=>"0",
 :Title=>"Study on security enhancements for Mission Critical Push To Talk (MCPTT) over LTE",
 :"former WG"=>nil,
 :"WG prime"=>"S3",
 :"WG other"=>nil,
...

{:Type=>"TS",
 :Number=>"29.163",
 :"For publication"=>"1",
 :Title=>"Interworking between the IP Multimedia (IM) Core Network (CN) subsystem and Circuit Switched (CS) networks",
 :"former WG"=>nil,
 :"WG prime"=>"C3",
 :"WG other"=>nil,
...

{:Type=>"TS",
 :Number=>"27.007",
 :"For publication"=>"1",
 :Title=>"AT command set for User Equipment (UE)",
 :"former WG"=>"T2",
 :"WG prime"=>"C1",
 :"WG other"=>"-",
...

Sope specs look like this:

{:Type=>"TS",
 :Number=>"01.02",
 :"For publication"=>"1",
 :Title=>"General Description of a GSM Public Land Mobile Network (PLMN)",
 :"former WG"=>"SMG01  /  SMG02  /  SMG03",
 :"WG prime"=>"S1",
 :"WG other"=>"GP / CP",
...

Therefore we can't just join them, can we?

ronaldtse commented 2 years ago

@andrew2net sorry, yes you are right. These WG codes are split as "prime" and "other" components in the Specs_GSM+3G table.

It's just so strange the WG other can be just -...

ronaldtse commented 2 years ago

Maybe you should rename them:

  "editorialgroup": [
      {
          "name": "C1",
          "type": "wg-prime"
      },
      {
          "name": "-",
          "type": "wg-other"
      },
      {
          "name": "T2",
          "type": "wg-former"
      }
  ],

ronaldtse commented 2 years ago

For the content of "former WG" being S1 / SMG01 / T2 / SMG04, it should be something like:

"wg-former": [
[
      {
          "name": "S1",
          "type": "wg-prime"
      },
      {
          "name": "SMG01",
          "type": "wg-other"
      }
],
[
      {
          "name": "T2",
          "type": "wg-prime"
      },
      {
          "name": "SMG04",
          "type": "wg-other"
      }
],
]

But many of them don't have both components (e.g. SPAN3 / N4 / SMG03?).

It's too inconsistent, so we might as well not parse "former WG".

ronaldtse commented 2 years ago

Maybe - just means NULL and we just drop that.

andrew2net commented 2 years ago

Yes, I think - should be dropped. Will do it. Concerning nested workgroups "wg-former": [{ "name": "S1", "type": "wg-prime" }] we don't have such a option in the Relaton model.

andrew2net commented 2 years ago

@ronaldtse I've removed workgroups with - names. Can we close this issue?

ronaldtse commented 2 years ago

Yes, thank you.

relaton / relaton-3gpp

Fix parsing of "editorial group" #6