Missing stage_codes in stages.yaml

opoudjis commented 1 year ago

pubid-core-1.8.0/lib/pubid/core/harmonized_stage_code.rb

    14|       @stages = if stage_or_code.is_a?(Array)
    15|                   stage_or_code
    16|                 elsif stage_or_code.is_a?(String) && config.stages["codes_description"].key?(stage_or_code)
    17|                   [stage_or_code]
    18|                   # when stage is stage name
=>  19|                 elsif config.stages["stage_codes"].key?(stage_or_code.to_s)
    20|                   ["#{config.stages["stage_codes"][stage_or_code.to_s]}.#{config.stages["substage_codes"][substage.to_s]}"]

If the pubid flavour gem is initialised without stage_codes in its YAML, as is the case with pubid-iec/stages.yaml, this step crashes.

@mico These gems are not being tested adequately.

Either add stage_codes to pubid-iec, or else make harmonized_stage_code not presuppose the existence of stage_codes: !config.stages("stage_codes", stage_or_code.to_s).nil? The latter is the safest.

mico commented 1 year ago

@opoudjis could you help me with use case where we face this bug?

mico commented 1 year ago

@opoudjis I released pubid-core update (https://github.com/metanorma/pubid-core/releases/tag/v1.8.1) implementing what you was asking here, but it might not help with parsing stages that are not defined in config.stages["codes_description"] and it will raise Pubid::Core::Errors::HarmonizedStageCodeInvalidError for unknown stages like "60.92".

ronaldtse commented 1 year ago

@mico which stage codes do not have harmonize stage descriptions? Thanks.

mico commented 1 year ago

@mico which stage codes do not have harmonize stage descriptions? Thanks.

@ronaldtse from conversation with @opoudjis I learned that "60.92" don't have stage description

ronaldtse commented 1 year ago

I think we need to implement a list that says what stages are not allowed: https://www.iso.org/files/live/sites/isoorg/files/developing_standards/docs/en/stage_codes.pdf

All these empty boxes should not be allowed to be used to create stages:

If someone attempts to use those stages, a HarmonizedStageCodeInvalidError should definitely be thrown.

mico commented 1 year ago

@opoudjis could you comment on "60.92" code you found in IEEE identifiers?

ronaldtse commented 1 year ago

@opoudjis please help clarify -- this code 60.92 is not meant to exist?

opoudjis commented 1 year ago

it was in a sample document header created by @Intelligent2013 as an example, top of https://github.com/metanorma/metanorma-iec/issues/154

Alex, to confirm, that was an example, not a real state, right?

Intelligent2013 commented 1 year ago

@opoudjis when I've worked on the issue https://github.com/metanorma/metanorma-iec/issues/153, just for debug purposes I've get the sample document from mn-samples-iec (https://raw.githubusercontent.com/metanorma/mn-samples-iec/main/sources/iec-rice-en.adoc) with

:docstage: 30
:docsubstage: 92

and update some fields for ISO-IEC-TR-63306-1-2020 (also changed to :docstage: 60), therefore we have:

:docstage: 60
:docsubstage: 92

So, you can ignore it.

opoudjis commented 11 months ago

Just as a reminder:

If you get a request with an illegal stage/substage, and there are no stage codes defined in YAML, you should still not crash as the code currently does, but degrade gracefully. This is a problem with pubid-core, and it will turn up elsewhere.
I need pubid-iec as a source of truth to tell me whether a stage/substage in IEC is legal or not. Will test @mico's code update when I have free time.