Issues with data

[strogonoff]

This concerns the main branch. I see there was a push yesterday, so it looks like it’s being actively worked on. In that case, here are some issues that break toolchains that work with that repository:

• UUIDs within files don’t match UUIDs in filenames:
  
• UUIDs between concepts & localized concepts don’t match (concepts must refer to correct UUIDs of localized concepts).
  This may be because concepts are referencing the UUID in filename, whereas some software (e.g., RegistryKit, etc.) references the id from within the file
• Register items (at least localized concepts, I believe) are missing required top-level attribute status, which must be one of: valid, invalid, superseded, or retired. This means they are not recognized as register items by software that validates register contents more strictly

Minor issues—this does not break anything, but will cause unexpected diffs:

• The initial document separator --- will be removed on data updates, because serializers that write register YAML treat each file as a standalone bare document (they do not allow multiple documents within one YAML file and intentionally do not support any directives)

cc @HassanAkbar @ronaldtse

[HassanAkbar]

As @strogonoff mentioned

Register items (at least localized concepts, I believe) are missing required top-level attribute status, which must be one of: valid, invalid, superseded, or retired. This means they are not recognized as register items by software that validates register contents more strictly

@ronaldtse In the 20231214_Multi-Lingual_Glossary–Published__unlocked__withMath sheet, there is no status defined for some terms, what should be the status in that case? e.g concept id = 18 in Danish language does not have a status available.

[ronaldtse]

@HassanAkbar if there are no status defined for those terms, just assume they are valid.

Ping @ReesePlews can we make that assumption? Thanks.

[ReesePlews]

hello @HassanAkbar i guess if the non-english term does not exist the status is "valid" but "unpublished". with the "valid" status coming from the base english term and the "unpublished" coming from no translation from the Danish group being submitted. does this understanding make sense to everyone? @ronaldtse do you think this is an acceptable understanding?

Hassan, please send me a link and i will have a check. we are still working from the staging site, correct?

Question: earlier when you are talking about the UUIDs and such, this does not mean there will be a change in the api (json, rdf, turtle) links, correct?

[strogonoff]

Just for the record, we need to make sure the status attribute value is supported by spec (register or glossarist). I have not heard of "unpublished", the closest I can think of is "invalid".On 13 May 2024, at 10:14 AM, Reese Plews ***@***.***> wrote: hello @HassanAkbar i guess if the non-english term does not exist the status is "valid" but "unpublished". with the "valid" status coming from the base english term and the "unpublished" coming from no translation from the Danish group being submitted. does this understanding make sense to everyone? @ronaldtse do you think this is an acceptable understanding? Hassan, please send me a link and i will have a check. we are still working from the staging site, correct? Question: earlier when you are talking about the UUIDs and such, this does not mean there will be a change in the api (json, rdf, turtle) links, correct? —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: ***@***.***>

[strogonoff]

I think it’s the register data model in question, so valid, invalid, superseded, retired… Unpublished items can exist as part of a proposal, but that’s about it I think.On 13 May 2024, at 7:33 PM, Anton Strogonoff ***@***.***> wrote:Just for the record, we need to make sure the status attribute value is supported by spec (register or glossarist). I have not heard of "unpublished", the closest I can think of is "invalid".On 13 May 2024, at 10:14 AM, Reese Plews ***@***.***> wrote: hello @HassanAkbar i guess if the non-english term does not exist the status is "valid" but "unpublished". with the "valid" status coming from the base english term and the "unpublished" coming from no translation from the Danish group being submitted. does this understanding make sense to everyone? @ronaldtse do you think this is an acceptable understanding? Hassan, please send me a link and i will have a check. we are still working from the staging site, correct? Question: earlier when you are talking about the UUIDs and such, this does not mean there will be a change in the api (json, rdf, turtle) links, correct? —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: ***@***.***>

[HassanAkbar]

Question: earlier when you are talking about the UUIDs and such, this does not mean there will be a change in the api (json, rdf, turtle) links, correct?

@ReesePlews yes that is correct, the api will not be changed just values of the ids will be updated.

i guess if the non-english term does not exist the status is "valid" but "unpublished"

@ReesePlews The issue here is that the translation exists but it does not have a status attribute given in the excel file that is being used to generate the concepts.

By the explanation provided by you and @strogonoff I believe that it should be valid in that case.
Is this correct?

cc: @ronaldtse

[ronaldtse]

Yes @HassanAkbar I believe this is the understanding. Thanks @ReesePlews @strogonoff !