clusterType values not listed #1507
Assignees
Labels
area/docsgen
Issues with docs capture or example rendering, historically part of pkg/tfgen
kind/bug
Some behavior is incorrect or out of spec
resolution/fixed
This issue was fixed
Comments
|
Thank you @brucemackenzie, we will look into why these values are missing. |
toriancrane
added
kind/bug
|
Adding a bit of detail. The issue is that these values are present in the TF provider docs, but missing in ours:
It looks like they never made it into the schema: 
So it looks like this is a bug in the doc-parsing code. Transferring to pulumi-terraform-bridge for the team to confirm. |
10 tasks
|
Yep, this is a bug in our docs parsing code. Thanks for reporting it. |
|
I believe this is, if not a duplicate, then at least related to pulumi/pulumi-libvirt#335. |
iwahbe
added
area/docsgen
|
Update: We can in fact fix this by more strictly parsing potential properties as presenting in #1969 will not address situations such as pulumi/pulumi-libvirt#335, since in that case we're seeing only lowercase values. |
guineveresaenger
added a commit
that referenced
this issue
May 13, 2024
This pull request asserts that [according to Terraform best practices](https://www.terraform-best-practices.com/naming) we should not parse backticked values containing uppercase letters as potential schema properties. While it is a convention to use `snake_case` for TF properties, it is a strong one. The AWS provider [requires it](https://github.com/hashicorp/terraform-provider-aws/blob/main/docs/naming.md), and other providers follow this convention strictly as well. The reason I would like to bring this change is because it allows us to detect documentation sections that are not actually nested property types, but rather lists of possible values, as described in #1507 in pulumi-mongodbatlas. Rather than add an ever-elusive regex for "Possible values include" et al, we can rely with reasonable confidence on the fact that TF properties are in general lowercase. See sample comparison diffs below. A [comparison diff for mongodbatlas](https://github.com/pulumi/pulumi-mongodbatlas/compare/missing-settings?expand=1) shows multiple improvements. There are a few regressions but the main one is for [a resource](https://www.pulumi.com/registry/packages/mongodbatlas/api-docs/getthirdpartyintegrations/) whose upstream doc is so unconventional that our representation of it is already flawed. Overall, we gain a lot more information than we're losing - the cost is that in a couple of places we get a nonsensical appendage of a real TF description. This should be fixed with docs overrides or by opening an upstream PR. A [comparison diff for AWS](https://github.com/pulumi/pulumi-aws/compare/missing-settings?expand=1) shows even more improvements, with only `V2modelsBotVersionLocaleSpecification` having a single regression, because the upstream docs are erroneously using camelCase. A [comparison diff for GCP](https://github.com/pulumi/pulumi-gcp/compare/missing-settings?expand=1) shows only improvements. A [comparison diff for Azure](https://github.com/pulumi/pulumi-azure/compare/missing-descriptions?expand=1) shows only improvements. **Note** I refactored the tests slightly because I wanted to be able to run individual tests and leave hints as to what they were testing. Fixes #1507.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
File: themes/default/content/registry/packages/mongodbatlas/api-docs/cluster/_index.md
Looking in the src code I found these values possible for
clusterType:REPLICASET,GEOSHARDED,SHARDEDThe text was updated successfully, but these errors were encountered: