[Bug]: In-consistant generated name for anonymous model caused breaking between m4 and tcgc

[MaryGao]

Describe the bug

This happened in common types but I believe this breaking would be a general issue for anonymous model when we migrate swagger to typespec. For case with below and here is the difference.

• name given by M4 is AccessRulePropertiesSubscriptionsItem and the pattern is ${Modelname}${Propertyname}Item
• name from tcgc is AccessRulePropertiesSubscription and the pattern is ${Modelname}${Singular(Propertyname)}

The common types in typespec is here.

model AccessRuleProperties {
  // ...

  /** Subscriptions for inbound rules */
  subscriptions?: {
    /** The fully qualified Azure resource ID of the subscription e.g. ('/subscriptions/00000000-0000-0000-0000-000000000000') */
    id?: Azure.Core.armResourceIdentifier;
  }[];
}

And relevant swagger definition is here.

"subscriptions": {
          "description": "Subscriptions for inbound rules",
          "type": "array",
          "items": {
            "description": "Subscription identifiers",
            "type": "object",
            "properties": {
              "id": {
                "description": "The fully qualified Azure resource ID of the subscription e.g. ('/subscriptions/00000000-0000-0000-0000-000000000000') ",
                "type": "string",
                "format": "arm-id"
              }
            }
          }
        },

Possible solutions

• Allow to add client name for anonymous model(Cannot augment properties in nested anonymous model microsoft/typespec#4818)
• Discourage the anonymous model in typespec and always give an name for that model([linter]discourage to use anonymous model and always give a model name #1727)

Reproduction

NA

Checklist

• Follow our Code of Conduct
• Check that this issue is about the Azure libraries for typespec. For bug in the typespec language or core libraries file it in the TypeSpec repo
• Check that there isn't already an issue that request the same bug to avoid creating a duplicate.
• The provided reproduction is a minimal reproducible example of the bug.

[MaryGao]

@allenjzhang I'd like to clarify from typespec ARM side, do we have any preference on how to avoid this breaking? Since anonymous model is not recommanded to use, do we plan to not use any anonymous models in common types?

[markcowl]

Moved to TCGC - if anaonymous models are used, changes in naming are always possible.

[MaryGao]

@markcowl Since anaonymous models are not a good practice do we have a plan in common types to not using it?