在任何父解析器中使用子 graphql 解析器 - 解析器链接

Question

. Answers to this question are eligible for a +50 reputation bounty. BiasInput wants to to this question.

Let's say I have an author with different books.
I can either fetch a book alone, or fetch an author with its books.

type Query {
  author(authorId: ID!): Author
  book(bookId: ID!): Book
}

type Author {
  books: [Book!]!
}

type Book {
  title: String!
}

The book can already resolve by himself:

// book-resolver.js
{
  // single source of truth for book resolving
  Book: {
    title: (_, {bookId}) => fetchBookTitleByBookId(bookId)
  }
}

How can I reuse the book resolver in the Author's resolver? I would like to avoid rewriting it:

// author-resolver.js
{
  Author: {
    books: (_, {authorId}) => {
      const bookIds = fetchBookIdsByAuthorId(authorId)
    
      return {
        // here I'm duplicating logic. Can I avoid it?
        books: bookIds.map(bookId => ({
          title: fetchBookTitleByBookId(bookId)
        })
      }
    }
  }
}

I think that creating an utility function (eg fetchBook(id) ) and using it everywhere wouldn't match the "paradigm" of resolver chaining.

Answer 1

Generally, it is a bad idea to query so many items singularly, but if you must do so, you can use Apollo resolver chains

Keep in mind thad by doing so, you will call fetchBookTitleByBookId(bookId) n times (where n is the number of books) + fetchBookIdsByAuthorId(authorId) once. See this answer for more info https://stackoverflow.com/a/60832838/10909317