Add Post-Processing Deduplication for Deep GraphQL Queries/mutations

### Describe the feature you'd like to request

Problem
The current amplify-codegen generates GraphQL queries, mutations, and subscriptions with a uniform depth (configurable via --maxDepth), but it doesn’t address redundant data in deeply interrelated schemas.

It may be useful to generate queries with a max depth of 4, but duplicated data from a closely related object may be generated unnecessarily.

Goal
Provide a mechanism to prune any duplicated query branches.

### Describe the solution you'd like

Add a flag --deduplicate

When used with --maxDepth it could be possible to generate queries that can traverse the entire graph with no duplication.

### Describe alternatives you've considered

Write custom queries with varying custom depths for each related object.

### Additional context

_No response_

### Is this something that you'd be interested in working on?

- [x] 👋 I may be able to implement this feature request

### Would this feature include a breaking change?

- [ ] ⚠️ This feature might incur a breaking change