relationships.md

Relationships

Relationship Types

SharpGraph supports all GraphQL relationship patterns:

One-to-Many

type User {
  posts: [Post]  # User has many posts
}

type Post {
  author: User   # Post belongs to one user
}

Many-to-Many

type User {
  friends: [User]  # Users can have many friends
}

Self-Referencing

type Category {
  parent: Category      # Category has one parent
  children: [Category]  # Category has many children
}

Schema Definition

Automatic Relationship Detection:

The schema parser automatically detects relationships and creates foreign key fields:

type Post {
  author: User     # Creates: authorId: ID
  tags: [Tag]      # Creates: tagsIds: [ID]
}

Foreign Key Management

Automatic Foreign Key Generation:

{
  "Post": [
    {
      "id": "post1",
      "title": "Hello World",
      "authorId": "user1",        // Many-to-one
      "tagsIds": ["tag1", "tag2"] // Many-to-many
    }
  ]
}

Query Resolution

Lazy Loading with Relationship Traversal:

{
  posts {
    title
    author {      # Automatically resolved via authorId
      name
      email
    }
    tags {        # Automatically resolved via tagsIds
      name
    }
  }
}

Resolution Process:

1. Query parser detects relationship fields
2. Executor loads primary records
3. Collects foreign key IDs
4. Batch loads related records
5. Projects requested fields

Relationship Resolution Algorithm

Resolution Algorithm:

1. Parse query - identify relationship fields
2. Load primary records - execute base query
3. Collect foreign keys - extract IDs from primary records
4. Batch load related - single query for all related records
5. Map relationships - connect records via foreign keys
6. Project fields - return only requested fields

Optimization Strategies:

• Foreign key batching reduces N+1 queries
• Index usage for efficient lookups
• Field projection minimizes data transfer
• Lazy loading prevents over-fetching