Implement pubid-itu

[ronaldtse]

As stated by @andrew2net :

• https://github.com/relaton/relaton-itu/issues/67#issuecomment-1603673688

Please start with code from this, which is also written using Parsley:

• https://github.com/relaton/relaton-itu/blob/main/lib/relaton_itu/pubid.rb

[mico]

@ronaldtse @andrew2net could you help me find a list of ITU identifiers, as complete as possible but something around 100 different identifiers from different categories and different format will be enough to start building parser.

[mico]

@ronaldtse @andrew2net also any documents related to categories, format would be very helpful.

[ronaldtse]

@mico ITU has 3 bureaus that issue different identifiers:

• ITU-R: all ITU-R document identifiers are here: https://github.com/relaton/relaton-data-itu-r
• ITU-T: ping @andrew2net if we have them
• ITU-D: not sure if we have them

[mico]

@ronaldtse I also found there are document types for each sector (ITU-R, ITU-T, ITU-D) like:

• reports
• questions
• resolutions
• handbooks
• opinions
• terminology
• etc...

But I don't understand how does it reflect in identifier.

[ronaldtse]

Right now we only care about ITU Recommendations. All 3 bureaus publish "Recommendations" (which are "standards").

[mico]

Right now we only care about ITU Recommendations. All 3 bureaus publish "Recommendations" (which are "standards").

@ronaldtse I found "question" type identifiers here https://github.com/relaton/relaton-data-itu-r/blob/master/data/ITU-R_SG07_110-2.yaml So I should skip identifiers like this for now, right?

[ronaldtse]

@mico actually we should parse all the identifiers from relaton-data-itu-r. The thing is the "Question" identifiers may differ per bureau (we have not fully analyzed the patterns).

[andrew2net]

• ITU-T: ping @andrew2net if we have them

here are all cases that I have https://github.com/relaton/relaton-itu/blob/main/spec/relaton_itu/pubid_spec.rb

[ronaldtse]

@mico can you please help document the different patterns used for different identifier types? Thanks!

[mico]

@ronaldtse @andrew2net I'm trying to differentiate "question", "recommendation", "handbook" and other types for identifiers. Right now we have:

• ITU-R SA.364-6 - recommendation
• ITU-R SG01.222-200 - question
• ITU-R 20-200 - handbook

There are other types like "reports" and "opinions" but I didn't find them in https://github.com/relaton/relaton-data-itu-r "reports" also have similar series as "recommendation" so ITU-R SA.364-6 could be valid identifier for "report" as well.

Handbook don't have series in identifier, but have them on website (https://www.itu.int/pub/R-HDB) very similar to question (SG00, SG01, SG02...), but don't appears in identifiers https://github.com/relaton/relaton-data-itu-r/blob/master/data/ITU-R_63-201.yaml

Upd.: for “resolution” type, identifier looks like ITU-R R.9-6, “R” before numbers means “resolution”, this is the only type where the type is clearly indicated.

How should we deal with that? Should we provide identifier type together with identifier to parse?

[ronaldtse]

ITU PubIDs have the following patterns:

• Recommendations. [series].[number] like H.265 or SA.364-6.
• Questions. [study-group].[number] like SG01.222-200. (SG01 means "Study Group 1")
• Other publications. [number] like 20-200.
• Resolutions. It seems that they have the pattern just like Recommendations but the series is "R".

We should definitely separate the identifier classes.