List of traits used in AWS files but not used in code generation

[adam-fowler]

Look into

• smithy.api#retryable: If we receive this error we can retry
• smithy.api#sensitive: Sensitive data, this is not being brought across
• aws.api#clientEndpointDiscoveryId: which member(s) of the operation input should be used to discover an endpoint
• smithy.api#auth: authentication scheme. Currently everything uses sigv4. Will be needed when sigv4a appears

Need to understand how these affect code

• smithy.api#box: not sure how this works with required
• smithy.api#cors: not used at the moment, maybe need to set certain headers
• aws.api#arn: ARN for a resource. Would need to break services into resources for this to work

To investigate

• clientOptional
• addedDefault
• nestedProperties
• property
• notProperty
• aws.api#tagEnabled
• aws.api#taggable
• aws.cloudformation#cfnDefaultValue

Can ignore

• smithy.api#optionalAuth: Set for operations that don't need credentials
• smithy.api#idempotent: We only need the "smithy.api#idempotencyToken" trait I believe
• smithy.api#readonly: indicates an operation is read only
• smithy.api#private: Don't use in other namespaces, probably can ignore
• smithy.api#tags: Tags used to filter shapes on export
• smithy.api#title: Think we can ignore this as we have aws.api#service$sdkId
• smithy.api#mediaType: not sure how useful this is, given it only seems to be applied to members of shape
• smithy.api#references: not sure we need this
• smithy.api#eventPayload: Used by S3.SelectContent probably can ignore
• smithy.api#suppress: Given we don't know what any of the validation events are we can probably ignore
• smithy.api#sparse: Lists/Maps may have null values. Used in one place I can ignore