Advanced Query

Smart Select Fields

GORM에서는, Select 메소드를 사용하여 특정 필드를 효율적으로 선택할 수 있습니다. 이는 특히 대규모 모델을 다룰 때 모델의 일부 필드만 필요한 경우에 API 응답에서 특히 유용합니다.

type User struct {
  ID     uint
  Name   string
  Age    int
  Gender string
  // 수백 개의 필드가 있다고 가정
}

type APIUser struct {
  ID   uint
  Name string
}

// 쿼리를 실행할 때 `id`, `name` 필드를 자동으로 select 합니다.
db.Model(&User{}).Limit(10).Find(&APIUser{})
// SELECT `id`, `name` FROM `users` LIMIT 10

NOTE QueryFields 모드에서는 모든 모델 필드가 그들의 이름에 따라 선택됩니다.

db, err := gorm.Open(sqlite.Open("gorm.db"), &gorm.Config{
  QueryFields: true,
})

// QueryFields가 true로 설정되어 있을 때 기본 동작
db.Find(&user)
// SQL: SELECT `users`.`name`, `users`.`age`, ... FROM `users`

// QueryFields와 함께 세션 모드를 사용할 때
db.Session(&gorm.Session{QueryFields: true}).Find(&user)
// SQL: SELECT `users`.`name`, `users`.`age`, ... FROM `users`

Locking

GORM 은 다른 종류의 lock을 지원합니다.

// Basic FOR UPDATE lock
db.Clauses(clause.Locking{Strength: "UPDATE"}).Find(&users)
// SQL: SELECT * FROM `users` FOR UPDATE

위의 명령문은 트랜잭션이 진행되는 동안 선택된 행을 잠그게 됩니다. 이는 행을 업데이트하기 전에 다른 트랜잭션이 해당 행을 수정하지 못하도록 원한다면 사용될 수 있습니다.

강도(Strength)는 SHARE로 설정될 수 있으며, 이는 다른 트랜잭션이 잠긴 행을 읽을 수 있지만 수정할 수는 없도록 행을 잠근다는 의미입니다.

db.Clauses(clause.Locking{
  Strength: "SHARE",
  Table: clause.Table{Name: clause.CurrentTable},
}).Find(&users)
// SQL: SELECT * FROM `users` FOR SHARE OF `users`

테이블(Table) 옵션은 특정 테이블을 잠그는데 사용될 수 있습니다. 여러 테이블을 조인하고, 그 중 하나만 잠그려고 할 때 유용할 수 있습니다.

옵션(Options) 에는 NOWAIT을 사용할 수 있습니다. 이는 lock을 획득하려고 시도하고 lock이 사용될 수 없는 경우 즉시 오류가 발생합니다. 다른 트랜잭션이 lock을 풀 때까지 기다리는 것을 방지합니다.

db.Clauses(clause.Locking{
  Strength: "UPDATE",
  Options: "NOWAIT",
}).Find(&users)
// SQL: SELECT * FROM `users` FOR UPDATE NOWAIT

다른 옵션으로는 SKIP LOCKED을 사용할 수 있습니다. 이는 다른 트랜잭션에 의해 이미 잠겨 있는 행을 건너뛰게 됩니다. 이는 다른 트랜잭션에 의해 현재 잠겨 있지 않은 행을 처리하고자 하는 높은 동시성 상황에서 유용합니다.

잠금 전략들에 대한 더 상세한 정보를 원한다면 Raw SQL and SQL Builder를 참고해주세요.

SubQuery

Subqueries are a powerful feature in SQL, allowing nested queries. GORM can generate subqueries automatically when using a *gorm.DB object as a parameter.

// Simple subquery
db.Where("amount > (?)", db.Table("orders").Select("AVG(amount)")).Find(&orders)
// SQL: SELECT * FROM "orders" WHERE amount > (SELECT AVG(amount) FROM "orders");

// Nested subquery
subQuery := db.Select("AVG(age)").Where("name LIKE ?", "name%").Table("users")
db.Select("AVG(age) as avgage").Group("name").Having("AVG(age) > (?)", subQuery).Find(&results)
// SQL: SELECT AVG(age) as avgage FROM `users` GROUP BY `name` HAVING AVG(age) > (SELECT AVG(age) FROM `users` WHERE name LIKE "name%")

From SubQuery

GORM allows the use of subqueries in the FROM clause, enabling complex queries and data organization.

// Using subquery in FROM clause
db.Table("(?) as u", db.Model(&User{}).Select("name", "age")).Where("age = ?", 18).Find(&User{})
// SQL: SELECT * FROM (SELECT `name`,`age` FROM `users`) as u WHERE `age` = 18

// Combining multiple subqueries in FROM clause
subQuery1 := db.Model(&User{}).Select("name")
subQuery2 := db.Model(&Pet{}).Select("name")
db.Table("(?) as u, (?) as p", subQuery1, subQuery2).Find(&User{})
// SQL: SELECT * FROM (SELECT `name` FROM `users`) as u, (SELECT `name` FROM `pets`) as p

Group Conditions (그룹화 조건)

Group Conditions in GORM provide a more readable and maintainable way to write complex SQL queries involving multiple conditions.

// Complex SQL query using Group Conditions
db.Where(
  db.Where("pizza = ?", "pepperoni").Where(db.Where("size = ?", "small").Or("size = ?", "medium")),
).Or(
  db.Where("pizza = ?", "hawaiian").Where("size = ?", "xlarge"),
).Find(&Pizza{})
// SQL: SELECT * FROM `pizzas` WHERE (pizza = "pepperoni" AND (size = "small" OR size = "medium")) OR (pizza = "hawaiian" AND size = "xlarge")

IN with multiple columns

GORM supports the IN clause with multiple columns, allowing you to filter data based on multiple field values in a single query.

// Using IN with multiple columns
db.Where("(name, age, role) IN ?", [][]interface{}{{"jinzhu", 18, "admin"}, {"jinzhu2", 19, "user"}}).Find(&users)
// SQL: SELECT * FROM users WHERE (name, age, role) IN (("jinzhu", 18, "admin"), ("jinzhu 2", 19, "user"));

Named Argument

GORM enhances the readability and maintainability of SQL queries by supporting named arguments. This feature allows for clearer and more organized query construction, especially in complex queries with multiple parameters. Named arguments can be utilized using either sql.NamedArg or map[string]interface{}{}, providing flexibility in how you structure your queries.

// Example using sql.NamedArg for named arguments
db.Where("name1 = @name OR name2 = @name", sql.Named("name", "jinzhu")).Find(&user)
// SQL: SELECT * FROM `users` WHERE name1 = "jinzhu" OR name2 = "jinzhu"

// Example using a map for named arguments
db.Where("name1 = @name OR name2 = @name", map[string]interface{}{"name": "jinzhu"}).First(&user)
// SQL: SELECT * FROM `users` WHERE name1 = "jinzhu" OR name2 = "jinzhu" ORDER BY `users`.`id` LIMIT 1

For more examples and details, see Raw SQL and SQL Builder

Find To Map

GORM provides flexibility in querying data by allowing results to be scanned into a map[string]interface{} or []map[string]interface{}, which can be useful for dynamic data structures.

When using Find To Map, it’s crucial to include Model or Table in your query to explicitly specify the table name. This ensures that GORM understands which table to query against.

// Scanning the first result into a map with Model
result := map[string]interface{}{}
db.Model(&User{}).First(&result, "id = ?", 1)
// SQL: SELECT * FROM `users` WHERE id = 1 LIMIT 1

// Scanning multiple results into a slice of maps with Table
var results []map[string]interface{}
db.Table("users").Find(&results)
// SQL: SELECT * FROM `users`

FirstOrInit

GORM’s FirstOrInit method is utilized to fetch the first record that matches given conditions, or initialize a new instance if no matching record is found. This method is compatible with both struct and map conditions and allows additional flexibility with the Attrs and Assign methods.

// If no User with the name "non_existing" is found, initialize a new User
var user User
db.FirstOrInit(&user, User{Name: "non_existing"})
// user -> User{Name: "non_existing"} if not found

// Retrieving a user named "jinzhu"
db.Where(User{Name: "jinzhu"}).FirstOrInit(&user)
// user -> User{ID: 111, Name: "Jinzhu", Age: 18} if found

// Using a map to specify the search condition
db.FirstOrInit(&user, map[string]interface{}{"name": "jinzhu"})
// user -> User{ID: 111, Name: "Jinzhu", Age: 18} if found

Using Attrs for Initialization

When no record is found, you can use Attrs to initialize a struct with additional attributes. These attributes are included in the new struct but are not used in the SQL query.

// If no User is found, initialize with given conditions and additional attributes
db.Where(User{Name: "non_existing"}).Attrs(User{Age: 20}).FirstOrInit(&user)
// SQL: SELECT * FROM USERS WHERE name = 'non_existing' ORDER BY id LIMIT 1;
// user -> User{Name: "non_existing", Age: 20} if not found

// If a User named "Jinzhu" is found, `Attrs` are ignored
db.Where(User{Name: "Jinzhu"}).Attrs(User{Age: 20}).FirstOrInit(&user)
// SQL: SELECT * FROM USERS WHERE name = 'Jinzhu' ORDER BY id LIMIT 1;
// user -> User{ID: 111, Name: "Jinzhu", Age: 18} if found

Using Assign for Attributes

The Assign method allows you to set attributes on the struct regardless of whether the record is found or not. These attributes are set on the struct but are not used to build the SQL query and the final data won’t be saved into the database.

// Initialize with given conditions and Assign attributes, regardless of record existence
db.Where(User{Name: "non_existing"}).Assign(User{Age: 20}).FirstOrInit(&user)
// user -> User{Name: "non_existing", Age: 20} if not found

// If a User named "Jinzhu" is found, update the struct with Assign attributes
db.Where(User{Name: "Jinzhu"}).Assign(User{Age: 20}).FirstOrInit(&user)
// SQL: SELECT * FROM USERS WHERE name = 'Jinzhu' ORDER BY id LIMIT 1;
// user -> User{ID: 111, Name: "Jinzhu", Age: 20} if found

FirstOrInit, along with Attrs and Assign, provides a powerful and flexible way to ensure a record exists and is initialized or updated with specific attributes in a single step.

FirstOrCreate

FirstOrCreate in GORM is used to fetch the first record that matches given conditions or create a new one if no matching record is found. This method is effective with both struct and map conditions. The RowsAffected property is useful to determine the number of records created or updated.

// Create a new record if not found
result := db.FirstOrCreate(&user, User{Name: "non_existing"})
// SQL: INSERT INTO "users" (name) VALUES ("non_existing");
// user -> User{ID: 112, Name: "non_existing"}
// result.RowsAffected // => 1 (record created)

// If the user is found, no new record is created
result = db.Where(User{Name: "jinzhu"}).FirstOrCreate(&user)
// user -> User{ID: 111, Name: "jinzhu", Age: 18}
// result.RowsAffected // => 0 (no record created)

Using Attrs with FirstOrCreate

Attrs can be used to specify additional attributes for the new record if it is not found. These attributes are used for creation but not in the initial search query.

// Create a new record with additional attributes if not found
db.Where(User{Name: "non_existing"}).Attrs(User{Age: 20}).FirstOrCreate(&user)
// SQL: SELECT * FROM users WHERE name = 'non_existing';
// SQL: INSERT INTO "users" (name, age) VALUES ("non_existing", 20);
// user -> User{ID: 112, Name: "non_existing", Age: 20}

// If the user is found, `Attrs` are ignored
db.Where(User{Name: "jinzhu"}).Attrs(User{Age: 20}).FirstOrCreate(&user)
// SQL: SELECT * FROM users WHERE name = 'jinzhu';
// user -> User{ID: 111, Name: "jinzhu", Age: 18}

Using Assign with FirstOrCreate

The Assign method sets attributes on the record regardless of whether it is found or not, and these attributes are saved back to the database.

// Initialize and save new record with `Assign` attributes if not found
db.Where(User{Name: "non_existing"}).Assign(User{Age: 20}).FirstOrCreate(&user)
// SQL: SELECT * FROM users WHERE name = 'non_existing';
// SQL: INSERT INTO "users" (name, age) VALUES ("non_existing", 20);
// user -> User{ID: 112, Name: "non_existing", Age: 20}

// Update found record with `Assign` attributes
db.Where(User{Name: "jinzhu"}).Assign(User{Age: 20}).FirstOrCreate(&user)
// SQL: SELECT * FROM users WHERE name = 'jinzhu';
// SQL: UPDATE users SET age=20 WHERE id = 111;
// user -> User{ID: 111, Name: "Jinzhu", Age: 20}

Optimizer/Index Hints

GORM includes support for optimizer and index hints, allowing you to influence the query optimizer’s execution plan. This can be particularly useful in optimizing query performance or when dealing with complex queries.

Optimizer hints are directives that suggest how a database’s query optimizer should execute a query. GORM facilitates the use of optimizer hints through the gorm.io/hints package.

import "gorm.io/hints"

// Using an optimizer hint to set a maximum execution time
db.Clauses(hints.New("MAX_EXECUTION_TIME(10000)")).Find(&User{})
// SQL: SELECT * /*+ MAX_EXECUTION_TIME(10000) */ FROM `users`

Index Hints

Index hints provide guidance to the database about which indexes to use. They can be beneficial if the query planner is not selecting the most efficient indexes for a query.

import "gorm.io/hints"

// Suggesting the use of a specific index
db.Clauses(hints.UseIndex("idx_user_name")).Find(&User{})
// SQL: SELECT * FROM `users` USE INDEX (`idx_user_name`)

// Forcing the use of certain indexes for a JOIN operation
db.Clauses(hints.ForceIndex("idx_user_name", "idx_user_id").ForJoin()).Find(&User{})
// SQL: SELECT * FROM `users` FORCE INDEX FOR JOIN (`idx_user_name`,`idx_user_id`)

These hints can significantly impact query performance and behavior, especially in large databases or complex data models. For more detailed information and additional examples, refer to Optimizer Hints/Index/Comment in the GORM documentation.

Iteration

GORM supports the iteration over query results using the Rows method. This feature is particularly useful when you need to process large datasets or perform operations on each record individually.

You can iterate through rows returned by a query, scanning each row into a struct. This method provides granular control over how each record is handled.

rows, err := db.Model(&User{}).Where("name = ?", "jinzhu").Rows()
defer rows.Close()

for rows.Next() {
  var user User
  // ScanRows scans a row into a struct
  db.ScanRows(rows, &user)

  // Perform operations on each user
}

This approach is ideal for complex data processing that cannot be easily achieved with standard query methods.

FindInBatches

FindInBatches allows querying and processing records in batches. This is especially useful for handling large datasets efficiently, reducing memory usage and improving performance.

With FindInBatches, GORM processes records in specified batch sizes. Inside the batch processing function, you can apply operations to each batch of records.

// Processing records in batches of 100
result := db.Where("processed = ?", false).FindInBatches(&results, 100, func(tx *gorm.DB, batch int) error {
  for _, result := range results {
    // Operations on each record in the batch
  }

  // Save changes to the records in the current batch
  tx.Save(&results)

  // tx.RowsAffected provides the count of records in the current batch
  // The variable 'batch' indicates the current batch number

  // Returning an error will stop further batch processing
  return nil
})

// result.Error contains any errors encountered during batch processing
// result.RowsAffected provides the count of all processed records across batches

FindInBatches is an effective tool for processing large volumes of data in manageable chunks, optimizing resource usage and performance.

Query Hooks

GORM offers the ability to use hooks, such as AfterFind, which are triggered during the lifecycle of a query. These hooks allow for custom logic to be executed at specific points, such as after a record has been retrieved from the database.

This hook is useful for post-query data manipulation or default value settings. For more detailed information and additional hook types, refer to Hooks in the GORM documentation.

func (u *User) AfterFind(tx *gorm.DB) (err error) {
  // Custom logic after finding a user
  if u.Role == "" {
    u.Role = "user" // Set default role if not specified
  }
  return
}

// Usage of AfterFind hook happens automatically when a User is queried

Pluck

The Pluck method in GORM is used to query a single column from the database and scan the result into a slice. This method is ideal for when you need to retrieve specific fields from a model.

If you need to query more than one column, you can use Select with Scan or Find instead.

// Retrieving ages of all users
var ages []int64
db.Model(&User{}).Pluck("age", &ages)

// Retrieving names of all users
var names []string
db.Model(&User{}).Pluck("name", &names)

// Retrieving names from a different table
db.Table("deleted_users").Pluck("name", &names)

// Using Distinct with Pluck
db.Model(&User{}).Distinct().Pluck("Name", &names)
// SQL: SELECT DISTINCT `name` FROM `users`

// Querying multiple columns
db.Select("name", "age").Scan(&users)
db.Select("name", "age").Find(&users)

Scopes

Scopes in GORM are a powerful feature that allows you to define commonly-used query conditions as reusable methods. These scopes can be easily referenced in your queries, making your code more modular and readable.

Defining Scopes

Scopes are defined as functions that modify and return a gorm.DB instance. You can define a variety of conditions as scopes based on your application’s requirements.

// Scope for filtering records where amount is greater than 1000
func AmountGreaterThan1000(db *gorm.DB) *gorm.DB {
  return db.Where("amount > ?", 1000)
}

// Scope for orders paid with a credit card
func PaidWithCreditCard(db *gorm.DB) *gorm.DB {
  return db.Where("pay_mode_sign = ?", "C")
}

// Scope for orders paid with cash on delivery (COD)
func PaidWithCod(db *gorm.DB) *gorm.DB {
  return db.Where("pay_mode_sign = ?", "COD")
}

// Scope for filtering orders by status
func OrderStatus(status []string) func(db *gorm.DB) *gorm.DB {
  return func(db *gorm.DB) *gorm.DB {
    return db.Where("status IN (?)", status)
  }
}

Applying Scopes in Queries

You can apply one or more scopes to a query by using the Scopes method. This allows you to chain multiple conditions dynamically.

// Applying scopes to find all credit card orders with an amount greater than 1000
db.Scopes(AmountGreaterThan1000, PaidWithCreditCard).Find(&orders)

// Applying scopes to find all COD orders with an amount greater than 1000
db.Scopes(AmountGreaterThan1000, PaidWithCod).Find(&orders)

// Applying scopes to find all orders with specific statuses and an amount greater than 1000
db.Scopes(AmountGreaterThan1000, OrderStatus([]string{"paid", "shipped"})).Find(&orders)

Scopes are a clean and efficient way to encapsulate common query logic, enhancing the maintainability and readability of your code. For more detailed examples and usage, refer to Scopes in the GORM documentation.

Count

The Count method in GORM is used to retrieve the number of records that match a given query. It’s a useful feature for understanding the size of a dataset, particularly in scenarios involving conditional queries or data analysis.

Getting the Count of Matched Records

You can use Count to determine the number of records that meet specific criteria in your queries.

var count int64

// Counting users with specific names
db.Model(&User{}).Where("name = ?", "jinzhu").Or("name = ?", "jinzhu 2").Count(&count)
// SQL: SELECT count(1) FROM users WHERE name = 'jinzhu' OR name = 'jinzhu 2'

// Counting users with a single name condition
db.Model(&User{}).Where("name = ?", "jinzhu").Count(&count)
// SQL: SELECT count(1) FROM users WHERE name = 'jinzhu'

// Counting records in a different table
db.Table("deleted_users").Count(&count)
// SQL: SELECT count(1) FROM deleted_users

Count with Distinct and Group

GORM also allows counting distinct values and grouping results.

// Counting distinct names
db.Model(&User{}).Distinct("name").Count(&count)
// SQL: SELECT COUNT(DISTINCT(`name`)) FROM `users`

// Counting distinct values with a custom select
db.Table("deleted_users").Select("count(distinct(name))").Count(&count)
// SQL: SELECT count(distinct(name)) FROM deleted_users

// Counting grouped records
users := []User{
  {Name: "name1"},
  {Name: "name2"},
  {Name: "name3"},
  {Name: "name3"},
}

db.Model(&User{}).Group("name").Count(&count)
// Count after grouping by name
// count => 3