如何正确使用FirstOrCreate

I have the following simple struct

type Profile struct {
    gorm.Model

    Email     string    `json:"email" sql:"not null;unique"`
    LastLogin time.Time `json:"lastlogin"`
}

I'm trying to make a insert if it doesn't exist via

  db.Con.Debug().Where(db.Profile{Email: "user@domain.com"}).Assign(db.Profile{LastLogin: time.Now()}).FirstOrCreate(&profile)

I get the following in my logs

(/Users/mzupan/go/src/gitlab.com/org/app/pkg/auth/login.go:182)
[2018-09-24 13:35:58]  [4.56ms]  SELECT * FROM "profiles"  WHERE "profiles"."deleted_at" IS NULL AND (("profiles"."email" = 'user@domain.com')) ORDER BY "profiles"."id" ASC LIMIT 1
[0 rows affected or returned ]

(/Users/mzupan/go/src/gitlab.com/org/app/pkg/auth/login.go:182)
[2018-09-24 13:35:58]  [1.77ms]  UPDATE "profiles" SET "last_login" = '2018-09-24 13:35:58', "updated_at" = '2018-09-24 13:35:58'  WHERE "profiles"."deleted_at" IS NULL AND (("profiles"."email" = 'user@domain.com'))
[0 rows affected or returned ]

So it's trying to do the select/update even though 0 rows are found in the select. Seems to me I'm doing the right thing.

I think you forget to create the table db.CreateTable(&Profile{})

Here is a working example:

package main

import (
    "time"

    "github.com/jinzhu/gorm"
    _ "github.com/jinzhu/gorm/dialects/sqlite"
)

type Profile struct {
    gorm.Model

    Email     string    `json:"email" sql:"not null;unique"`
    LastLogin time.Time `json:"lastlogin"`
}

func main() {
    db, err := gorm.Open("sqlite3", "test.db")
    if err != nil {
        panic("failed to connect database")
    }
    defer db.Close()
    // Create the table...
    // Read the doc at: http://doc.gorm.io/database.html#migration -> sec: Create Table
    db.CreateTable(&Profile{})

    var profile Profile
    db.Debug().Where(Profile{Email: "user@domain.com"}).Assign(Profile{LastLogin: time.Now()}).FirstOrCreate(&profile)
}

Out put:

[2018-09-25 09:47:35]  [0.18ms]  INSERT INTO "profiles" ("created_at","updated_at","deleted_at","email","last_login") VALUES ('2018-09-25 09:47:35','2018-09-25 09:47:35',NULL,'user@domain.com','2018-09-25 09:47:35')
[1 rows affected or returned ]

Hope this help :)