Unfortunately due to lack of time and passion for Dgraph i'm unable to maintain the project as i'd like to. If any of you would like to be a maintainer of dqlx please comment below

When appending the computed arguments from a select, the slice of arguments was being appended, not the individual elements.

Note that this also resolves the TestExampleTestSuite/TestFundamentals/Applying_filters that has been failing.

Can someone help me understand this?

The following code generates the error result must be a pointer.

var data []map[string]interface{}

client.Query(dqlx.HasFn("name")).
		Filter(
                   dqlx.UID("0x12345"),
		).Select(`
		    name
		`).UnmarshalInto(data).Execute(ctx)

@fenos

Hi, I've not found any information about the @cascade directive in the library's documentation (I assume it's just not supported yet).

I was thinking about the "ugly" workaround:

• Add @filter(@cascade) at the root level - run ToDQL func - replace string @filter(@cascade) with just @cascade - turn back to QueryBuilder (generally I wanted to update only query string, but I didn't find any solution for turning query, args, and error back into QueryBuilder)

Can you give me a solution on how to add @cascade for my query? It's crucial for one of my functions.

Also, it would be nice to be able to modify the query string more directly with function, which would allow adding custom parts to the query at the root level (next to filters etc.). For example, when Dgraph would release another part of the query, needed to be placed at the root level then it would be easy to implement that part without a specific function from the DQLX package (considering that there isn't a function meant to add that part)

Example taken from the official docs:

package main

import (
	"fmt"

	"github.com/fenos/dqlx"
)

func main() {
	variable := dqlx.Variable(dqlx.EqFn("name", "test")).
		As("A").
		EdgeAs("B", "film").
		EdgeAs("C", "film->performance", dqlx.Select(`
			D as genre
		`))
	s, _, _ := variable.ToDQL()
	fmt.Println(s)
}

Current result:

query Rootquery($0:string) { C asvar(func: eq(<name>,$0)) { <film> { <performance> { D AS <genre> } } } }

Fixed with this PR:

query Rootquery($0:string) { A as var(func: eq(<name>,$0)) { B as  { C as  { D as <genre> } } } }

I've encountered an issue where I have an object that satisfies the json.Unmarshaler interface which is not being respected by Response.Unmarshal.

It would likely be a good idea to remove the dependency on a map[string]interface{} and then mapstructure here and instead make use of json.RawMessage. This also solves issues with decoding time.Time values as Time already implements the json.Unmarshaler interface for RFC 3339 formatted strings.

Support cascade directive

	query, variables, err := dql.
		QueryEdge("bladerunner", dql.EqFn("item", "value")).
		Fields(`
			uid
			name
			initial_release_date
			netflix_id
		`).
		Cascade().
		Edge("films", dql.Fields(`
			id
			date
		`), dql.Cascade("date", "id")).
		ToDQL()

Closes #3

data := []map[string]interface{}{
    {
        "uid": "uid(v)",
        "email": "[email protected]",
        "name": "first name"
    },
}

userByEmailQuery := dqlx.Query(dqlx.EqFn("email", "[email protected]")).
    .Select(`
        v as uid
        name
    `)

resp, err := db.Mutation().
    Query(userByEmailQuery).
    Set(data).
    Execute(ctx)

The above example does not work, as userByEmailQuery is of type dqlx.QueryBuilder and Query() needs dqlx.DQLizer

Currently, the response of any mutation returns a map of uids and an empty JSON. This is 100% wrong for the following reasons:

• The uids will be returned for some reason when dgraph won't generate the object because of an error that is never returned.
• The Dgraph database returns also the objects that are being generated (based on the schema), but those objects are not included in the response.

Hey everyone,

First thank you for this amazing repo. My issue is about EdgeAS;

	var output []map[string]interface{}
	_, err := r.Conn.QueryType("User").
		Filter(
			dqlx.UID(uid),
		).
		EdgeAs("total", "friendship", dqlx.Facets(dqlx.And{
			dqlx.GtFn("requested_at", util.BeginningOfDay().Unix()),
			dqlx.LtFn("requested_at", util.EndOfDay().Unix()),
		}), dqlx.Select("uid")).
		EdgeAs("to", "friendship", dqlx.UID(otherUSER), dqlx.Facets(dqlx.And{
			dqlx.GtFn("requested_at", util.BeginningOfWeek().Unix()),
			dqlx.LtFn("requested_at", util.EndOfWeek().Unix()),
		}), dqlx.Select("uid")).
		UnmarshalInto(&output).
		Execute(ctx)

	log.Print(output)

when I run it, I'm getting an error as

"rpc error: code = Unknown desc = line 1 column 35: Expected Left round brackets. Got: lex.Item [9] \"func\" at 1:35",

And the generated DQL is

query Rootquery($0:string, $1:int, $2:int, $3:int, $4:int, $5:string) { to as(func: type(<User>)) @filter(uid($0)) { <hey> @facets((gt(<requested_at>,$1) AND lt(<requested_at>,$2))) { <uid> } <hey> @facets((gt(<requested_at>,$3) AND lt(<requested_at>,$4))) @filter(uid($5)) { <uid> } } }
map[$0:0x57 $1:1662760800 $2:1662847199 $3:1662242400 $4:1662847199 $5:0x56]

anyone can help me ?

Hi, thank you for your library,

I was trying to make an upsert block https://dgraph.io/docs/mutations/upsert-block/#example-of-uid-function and I see in your documentation: https://fenos.github.io/dqlx/docs/mutations/set#logical-upsert the way to do it. But there is a type error.

In your example you add a Query in the mutation like this:

resp, err := db.Mutation().
    Query(userByEmailQuery).
    Set(data).
    Execute(ctx)

but the query method has a dependency on an interface DQLizer which has the method signature: ToDQL() (query string, args []interface{}, err error)

Using the query builder in your example:

userByEmailQuery := dqlx.Query(dqlx.EqFn("email", "[email protected]")).
    .Select(`
        v as uid
        name
    `)

given us an instance of QueryBuilder which has the method ToDQL() (query string, args map[string]string, err error) that is not match to the DQLizer method: