Autogenerate semantic convention constants

[srikanthccv]

Description

Fixes #1022

Type of change

Please delete options that are not relevant.

• New feature (non-breaking change which adds functionality)

Does This PR Require a Contrib Repo Change?

• Yes. - Link to PR: Sync with auto generated semantic convention constants opentelemetry-python-contrib#428

Checklist:

• Followed the style guidelines of this project
• Changelogs have been updated

[lzchen]

This is cool. We will need another PR to replace all the constant usages correct? Maybe another issue is needed?

[srikanthccv]

This is cool. We will need another PR to replace all the constant usages correct? Maybe another issue is needed?

Yes here is the PR that does it open-telemetry/opentelemetry-python-contrib#428. Another thing to point out here, there is manual list of resource related attributes here https://github.com/open-telemetry/opentelemetry-python/blob/main/opentelemetry-sdk/src/opentelemetry/sdk/resources/__init__.py#L74; Because removing them would be a breaking left as is and eventually it will be incomplete list. Going forward we can recommend people to use auto generated ones instead.

[codeboten]

Nice, this is great. Any thoughts about putting this in a semconv module to make it easier to differentiate from the rest of the code?

[srikanthccv]

Nice, this is great. Any thoughts about putting this in a semconv module to make it easier to differentiate from the rest of the code?

Sounds good to me. Your suggestion is to move this from attributes to semconv module right? I am open to any kind of alternative placement.

[codeboten]

As discussed in the SIG meeting, let's publish a separate pypi package for all semantic conventions: opentelemetry-semantic-conventions

[lzchen]

Because removing them would be a breaking left as is and eventually it will be incomplete list. Going forward we can recommend people to use auto generated ones instead.

Maybe we can make them derive from the autogenerated ones?

[srikanthccv]

Yes, that's what we decided on doing in the last SIG meeting.

[codeboten]

Thanks for updating this PR, overall it looks good. Blocking on the version number as the spec for this is still not stable.

[codeboten]

i wouldnt expect this package to be marked as stable as the spec is still experimental

[srikanthccv]

Then does it still make sense to use this unstable package in sdk to derive the resource attributes?

[codeboten]

I think that's fine, it just means that the SDK will need to be updated as the conventions package is updated. @lzchen any thoughts?

[lzchen]

Should be fine.

[codeboten]

Thanks for addressing my comments!

[srikanthccv]

@lzchen Please review again. There have been some major changes since you last reviewed.