feat(GraphQL): Add support for has filter on list of fields.

[minhaj-shakeel]

Fixes GRAPHQL-814.

---

This change is 

[abhimanyusinghgaur]

Reviewed 4 of 58 files at r1.
Reviewable status: 4 of 58 files reviewed, 4 unresolved discussions (waiting on @minhaj-shakeel and @pawanrawal)

---

graphql/e2e/common/mutation.go, line 1077 at r1 (raw file):

0

if we are setting it to 1000 and expecting 0, how can this pass in the test?
seems something fishy with the test.

---

graphql/e2e/common/mutation.go, line 1151 at r1 (raw file):

 0

same here, 1000 expected as 0

---

graphql/e2e/common/query.go, line 697 at r1 (raw file):

	newPost1 := addPostWithNullText(t, newAuthor.ID, newCountry.ID, postExecutor)
	newPost2 := addPostWithNoTags(t, newAuthor.ID, newCountry.ID, postExecutor)

instead of repeating the same code in each of these special cases, can we just write one function

addPostWithVars(t, authorId, countryId string, postVar map[string]interface) *post {
   ...
   addPostParams := &common.GraphQLParams{
      ...
      Vars: map[string]interface{}{"post":postVar}
   }
   ...
}

and use that inside each of these special case functions?

---

graphql/e2e/common/query.go, line 713 at r1 (raw file):

Quoted 6 lines of code…

	var result struct {
		QueryPost []*post
	}
	err := json.Unmarshal([]byte(gqlResponse.Data), &result)
	require.NoError(t, err)
	require.Equal(t, 1, len(result.QueryPost))

instead of making sure that only one post is returned, we should make sure what is exactly returned in that post, because right now, that one post can also be the one that we are not expecting. So, we should do something like:

testutil.CompareJSON(t, `{"queryPost":[{"title": "..."}]}`)

[pawanrawal]

Reviewed 55 of 58 files at r1, 3 of 3 files at r2.
Reviewable status: all files reviewed, 5 unresolved discussions (waiting on @minhaj-shakeel)

---

graphql/resolve/query_rewriter.go, line 1829 at r2 (raw file):

				})
			case []interface{}:
				// has: [comments, text] -> has(comments) AND has(text)

Add 1 e2e test as well for which has takes a list of arguments.

[pawanrawal]

Reviewed 2 of 2 files at r3.
Reviewable status: all files reviewed, 5 unresolved discussions (waiting on @minhaj-shakeel)