SQLBoiler is a tool to generate a Go ORM tailored to your database schema.

It is a "database-first" ORM as opposed to "code-first" (like gorm/gorp). That means you must first create your database schema. Please use something like goose, sql-migrate or some other migration tool to manage this part of the database's life-cycle.

v3 has been released, please upgrade when possible, v2 is on life support only now.

While attempting to migrate a legacy Rails database, we realized how much ActiveRecord benefitted us in terms of development velocity. Coming over to the Go database/sql package after using ActiveRecord feels extremely repetitive, super long-winded and down-right boring. Being Go veterans we knew the state of ORMs was shaky, and after a quick review we found what our fears confirmed. Most packages out there are code-first, reflect-based and have a very weak story around relationships between models. So with that we set out with these goals:

• Work with existing databases: Don't be the tool to define the schema, that's better left to other tools.
• ActiveRecord-like productivity: Eliminate all sql boilerplate, have relationships as a first-class concept.
• Go-like feel: Work with normal structs, call functions, no hyper-magical struct tags, small interfaces.
• Go-like performance: Benchmark and optimize the hot-paths, perform like hand-rolled sql.DB code.

We believe with SQLBoiler and our database-first code-generation approach we've been able to successfully meet all of these goals. On top of that SQLBoiler also confers the following benefits:

• The models package is type safe. This means no chance of random panics due to passing in the wrong type. No need for interface{}.
• Our types closely correlate to your database column types. This is expanded by our extended null package which supports nearly all Go data types.
• A system that is easy to debug. Your ORM is tailored to your schema, the code paths should be easy to trace since it's not all buried in reflect.
• Auto-completion provides work-flow efficiency gains.

• SQLBoiler
  • Why another ORM
  • About SQL Boiler
    • Features
    • Supported Databases
    • A Small Taste
  • Requirements & Pro Tips
    • Requirements
    • Pro Tips
  • Getting started
    • Videos
    • Download
    • Configuration
    • Initial Generation
    • Regeneration
    • Controlling Generation
      • Aliases
      • Types
      • Imports
      • Templates
    • Extending Generated Models
  • Diagnosing Problems
  • Features & Examples
    • Automatic CreatedAt/UpdatedAt
      • Overriding Automatic Timestamps
    • Query Building
    • Query Mod System
    • Function Variations
    • Finishers
    • Raw Query
    • Binding
    • Relationships
    • Hooks
    • Transactions
    • Debug Logging
    • Select
    • Find
    • Insert
    • Update
    • Delete
    • Upsert
    • Reload
    • Exists
    • Enums
    • Constants
  • FAQ
    • Won't compiling models for a huge database be very slow?
    • Missing imports for generated package
• Benchmarks

• Full model generation
• Extremely fast code generation
• High performance through generation & intelligent caching
• Uses boil.Executor (simple interface, sql.DB, sqlx.DB etc. compatible)
• Uses context.Context
• Easy workflow (models can always be regenerated, full auto-complete)
• Strongly typed querying (usually no converting or binding to pointers)
• Hooks (Before/After Create/Select/Update/Delete/Upsert)
• Automatic CreatedAt/UpdatedAt
• Table and column whitelist/blacklist
• Relationships/Associations
• Eager loading (recursive)
• Custom struct tags
• Transactions
• Raw SQL fallback
• Compatibility tests (Run against your own DB schema)
• Debug logging
• Basic multiple schema support (no cross-schema support)
• 1d arrays, json, hstore & more
• Enum types
• Out of band driver support

Note: SQLBoiler supports out of band driver support so you can make your own

We are seeking contributors for other database engines.

For a comprehensive list of available operations and examples please see Features & Examples.

import (
  // Import this so we don't have to use qm.Limit etc.
  . "github.com/volatiletech/sqlboiler/queries/qm"
)

// Open handle to database like normal
db, err := sql.Open("postgres", "dbname=fun user=abc")
if err != nil {
  return err
}

// If you don't want to pass in db to all generated methods
// you can use boil.SetDB to set it globally, and then use
// the G variant methods like so (--add-global-variants to enable)
boil.SetDB(db)
users, err := models.Users().AllG(ctx)

// Query all users
users, err := models.Users().All(ctx, db)

// Panic-able if you like to code that way (--add-panic-variants to enable)
users := models.Users().AllP(db)

// More complex query
users, err := models.Users(Where("age > ?", 30), Limit(5), Offset(6)).All(db, ctx)

// Ultra complex query
users, err := models.Users(
  Select("id", "name"),
  InnerJoin("credit_cards c on c.user_id = users.id"),
  Where("age > ?", 30),
  AndIn("c.kind in ?", "visa", "mastercard"),
  Or("email like ?", `%aol.com%`),
  GroupBy("id", "name"),
  Having("count(c.id) > ?", 2),
  Limit(5),
  Offset(6),
).All(ctx, db)

// Use any "boil.Executor" implementation (*sql.DB, *sql.Tx, data-dog mock db)
// for any query.
tx, err := db.BeginTx(ctx, nil)
if err != nil {
  return err
}
users, err := models.Users().All(ctx, tx)

// Relationships
user, err := models.Users().One(ctx, db)
if err != nil {
  return err
}
movies, err := user.FavoriteMovies().All(ctx, db)

// Eager loading
users, err := models.Users(Load("FavoriteMovies")).All(ctx, db)
if err != nil {
  return err
}
fmt.Println(len(users.R.FavoriteMovies))

• Go 1.6 minimum, and Go 1.7 for compatibility tests.
• Table names and column names should use snake_case format.
  • We require snake_case table names and column names. This is a recommended default in Postgres, and we agree that it's good form, so we're enforcing this format for all drivers for the time being.
• Join tables should use a composite primary key.
  • For join tables to be used transparently for relationships your join table must have a composite primary key that encompasses both foreign table foreign keys and no other columns in the table. For example, on a join table named user_videos you should have: primary key(user_id, video_id), with both user_id and video_id being foreign key columns to the users and videos tables respectively and there are no other columns on this table.
• For MySQL if using the github.com/go-sql-driver/mysql driver, please activate time.Time parsing when making your MySQL database connection. SQLBoiler uses time.Time and null.Time to represent time in it's models and without this enabled any models with DATE/DATETIME columns will not work.

• It's highly recommended to use transactions where sqlboiler will be doing multiple database calls (relationship setops with insertions for example) for both performance and data integrity.
• Foreign key column names should end with _id.
  • Foreign key column names in the format x_id will generate clearer method names. It is advisable to use this naming convention whenever it makes sense for your database schema.
• If you never plan on using the hooks functionality you can disable generation of this feature using the --no-hooks flag. This will save you some binary size.

If you like learning via a video medium, sqlboiler has a number of screencasts available.

SQLBoiler: Getting Started

SQLBoiler: What's New in v3

SQLBoiler: Advanced Queries and Relationships

Old (v2): SQLBoiler Screencast #1: How to get started

go get -u -t github.com/volatiletech/sqlboiler

# Also install the driver of your choice, there exists pqsl, mysql, mssql
# These are separate binaries.
go get github.com/volatiletech/sqlboiler/drivers/sqlboiler-psql

Create a configuration file. Because the project uses viper, TOML, JSON and YAML are all supported. Environment variables are also able to be used.

The configuration file should be named sqlboiler.toml for TOML, sqlboiler.json for JSON and sqlboiler.yaml or sqlboiler.yml for YAML and is searched for in the following directories in this order:

• ./
• $XDG_CONFIG_HOME/sqlboiler/
• $HOME/.config/sqlboiler/

We will assume TOML for the rest of the documentation.

We require you pass in your psql and mysql database configuration via the configuration file rather than env vars. There is no command line argument support for database configuration. Values given under the postgres and mysql block are passed directly to the psql and mysql drivers. Here is a rundown of all the different values that can go in that section:

Example of whitelist/blacklist:

[psql]
# Removes migrations table, and the name column from the addresses table
# from being generated. Foreign keys that reference tables or columns that
# are no longer generated because of whitelists or blacklists may cause problems.
blacklist = ["migrations", "addresses.name"]

You can also pass in these top level configuration values if you would prefer not to pass them through the command line or environment variables:

Example:

[psql]
  dbname = "dbname"
  host   = "localhost"
  port   = 5432
  user   = "dbusername"
  pass   = "dbpassword"
  schema = "myschema"
  blacklist = ["migrations", "other"]

[mysql]
  dbname  = "dbname"
  host    = "localhost"
  port    = 3306
  user    = "dbusername"
  pass    = "dbpassword"
  sslmode = "false"

[mssql]
  dbname  = "dbname"
  host    = "localhost"
  port    = 1433
  user    = "dbusername"
  pass    = "dbpassword"
  sslmode = "disable"
  schema  = "notdbo"

After creating a configuration file that points at the database we want to generate models for, we can invoke the sqlboiler command line utility.

SQL Boiler generates a Go ORM from template files, tailored to your database schema.
Complete documentation is available at https://github.com/volatiletech/sqlboiler

Usage:
  sqlboiler [flags] <driver>

Examples:
sqlboiler psql

Flags:
      --add-global-variants        Enable generation for global variants
      --add-panic-variants         Enable generation for panic variants
  -c, --config string              Filename of config file to override default lookup
  -d, --debug                      Debug mode prints stack traces on error
  -h, --help                       help for sqlboiler
      --no-auto-timestamps         Disable automatic timestamps for created_at/updated_at
      --no-context                 Disable context.Context usage in the generated code
      --no-hooks                   Disable hooks feature for your models
      --no-rows-affected           Disable rows affected in the generated API
      --no-tests                   Disable generated go test files
  -o, --output string              The name of the folder to output to (default "models")
  -p, --pkgname string             The name you wish to assign to your generated package (default "models")
      --struct-tag-casing string   Decides the casing for go structure tag names. camel or snake (default snake) (default "snake")
  -t, --tag strings                Struct tags to be included on your models in addition to json, yaml, toml
      --templates strings          A templates directory, overrides the bindata'd template folders in sqlboiler
      --version                    Print the version
      --wipe                       Delete the output folder (rm -rf) before generation to ensure sanity

Follow the steps below to do some basic model generation. Once you've generated your models, you can run the compatibility tests which will exercise the entirety of the generated code. This way you can ensure that your database is compatible with SQLBoiler. If you find there are some failing tests, please check the Diagnosing Problems section.

# Generate our models and exclude the migrations table
# When passing 'psql' here, it looks for a binary called
# 'sqlboiler-psql' in your CWD and PATH. You can also pass
# an absolute path to a driver if you desire.
sqlboiler psql

# Run the generated tests
go test ./models

Note: No mysqldump or pg_dump equivalent for Microsoft SQL Server, so generated tests must be supplemented by tables_schema.sql with CREATE TABLE ... queries

You can use go generate for SQLBoiler if you want to to make it easy to run the command for your application:

//go:generate sqlboiler --flags-go-here psql

It's important to not modify anything in the output folder, which brings us to the next topic: regeneration.

When regenerating the models it's recommended that you completely delete the generated directory in a build script or use the --wipe flag in SQLBoiler. The reasons for this are that sqlboiler doesn't try to diff your files in any smart way, it simply writes the files it's going to write whether they're there or not and doesn't delete any files that were added by you or previous runs of SQLBoiler. In the best case this can cause compilation errors, in the worst case this may leave extraneous and unusable code that was generated against tables that are no longer in the database.

The bottom line is that this tool should always produce the same result from the same source. And the intention is to always regenerate from a pure state. The only reason the --wipe flag isn't defaulted to on is because we don't like programs that rm -rf things on the filesystem without being asked to.

The templates get executed in a specific way each time. There's a variety of configuration options on the command line/config file that can control what features are turned on or off.

In addition to the command line flags there are a few features that are only available via the config file and can use some explanation.

In sqlboiler, names are automatically generated for you. If you name your database nice things you will likely have nice names in the end. However in the case where your names in your database are bad AND unchangeable, or sqlboiler's inference doesn't understand the names you do have (even though they are good and correct) you can use aliases to change the name of your tables, columns and relationships in the generated Go code.

Note: It is not required to provide all parts of all names. Anything left out will be inferred as it was in the past.

# Although team_names works fine without configuration, we use it here for illustrative purposes
[aliases.tables.team_names]
up_plural     = "TeamNames"
up_singular   = "TeamName"
down_plural   = "teamNames"
down_singular = "teamName"

  # Columns can also be aliased.
  [aliases.tables.team_names.columns]
  team_name = "OurTeamName"

When creating aliases for relationships, it's important to know how the concept of local/foreign map to what is used for generation. First off, everything is renamed using the foreign key as a unique identifier (namespaced to the table). If you don't know your foreign key names, it's likely you have not been naming them manually and it's possible they change suddenly for whatever reason. If you're going to rename relationships it's recommended that you use manually named foreign keys for stability. Therefore to rename a relationship based on a foreign key on the videos table you would use the key [aliases.tables.videos.relationships.fk_name].

In terms of how to understand what local and foreign are in the context of renaming relationships, local simply means "the side with the foreign key". For example in a users <-> videos relationship where we have a author_id on the video that refers to the id of the users table, the foreign key is on the videos table itself, so the local key refers to how we refer to the videos side of the relationship. In the example below we're naming the local side (how we refer to the videos) AuthoredVideos. The table foreign to the foreign key is the users table and we want to refer to that side of the relationship as the Author.

[aliases.tables.videos.relationships.videos_author_id_fkey]
# The local side would originally be inferred as AuthorVideos, which
# is probably good enough to not want to mess around with this feature, avoid it where possible.
local   = "AuthoredVideos"
# Even if left unspecified, the foreign side would have been inferred correctly
# due to the proper naming of the foreign key column.
foreign = "Author"

In a many-to-many relationship it's a bit more complicated. In an example where videos <-> tags with a join table in the middle. Imagine if the join table didn't exist, and instead both of the id columns in the join table were slapped on to the tables themselves. You'd have videos.tag_id and tags.video_id. Using a similar method to the above (the side with the foreign key) we can rename the relationships. To change Videos.Tags to Videos.Rags we can use the example below.

Keep in mind that naming ONE side of the many-to-many relationship is sufficient as the other side will be automatically mirrored, though you can specify both if you so choose.

[aliases.tables.video_tags.relationships.fk_video_id]
local   = "Rags"
foreign = "Videos"

There is an alternative syntax available for those who are challenged by the key syntax of toml or challenged by viper lowercasing all of your keys. Instead of using a regular table in toml, use an array of tables, and add a name field to each object. The only one that changes past that is columns, which now has to have a new field called alias.

[[aliases.tables]]
name          = "team_names"
up_plural     = "TeamNames"
up_singular   = "TeamName"
down_plural   = "teamNames"
down_singular = "teamName"

  [[aliases.tables.columns]]
  name  = "team_name"
  alias = "OurTeamName"

  [[aliases.tables.video_tags.relationships]]
  name    = "fk_video_id"
  local   = "Rags"
  foreign = "Videos"

There exists the ability to override types that the driver has inferred. The way to accomplish this is through the config file.

[[types]]
  # The match is a drivers.Column struct, and matches on almost all fields.
  # Notable exception for the unique bool. Matches are done
  # with "logical and" meaning it must match all specified matchers. Boolean values
  # are only checked if all the string specifiers match first, and they
  # must always match.
  # Not shown here: db_type is the database type and a very useful matcher
  [types.match]
    type = "null.String"
    nullable = true

  # The replace is what we replace the strings with. You cannot modify any
  # boolean values in here. But we could change the Go type (the most useful thing)
  # or the DBType or FullDBType etc. if for some reason we needed to.
  [types.replace]
    type = "mynull.String"

  # These imports specified here overwrite the definition of the type's "based_on_type"
  # list. The type entry that is replaced is the replaced type's "type" field.
  # In the above example it would add an entry for mynull.String, if we did not
  # change the type in our replacement, it would overwrite the null.String entry.
  [types.imports]
    third_party = ['"github.com/me/mynull"']

Imports are overridable by the user. This can be used in conjunction with replacing the templates for extreme cases. Typically this should be avoided.

Note that specifying any section of the imports completely overwrites that section. It's also true that the driver can still specify imports and those will be merged in to what is provided here.

[imports.all]
  standard = ['"context"']
  third_party = ['"github.com/my/package"']

# Changes imports for the boil_queries file
[imports.singleton."boil_queries"]
  standard = ['"context"']
  third_party = ['"github.com/my/package"']

# Same syntax as all
[imports.test]

# Same syntax as singleton
[imports.test_singleton]

# Changes imports when a model contains null.Int32
[imports.based_on_type.string]
  standard = ['"context"']
  third_party = ['"github.com/my/package"']

When defining maps it's possible to use an alternative syntax since viper automatically lowercases all configuration keys (same as aliases).

[[imports.singleton]]
  name = "boil_queries"
  third_party = ['"github.com/my/package"']

[[imports.based_on_type]]
  name = "null.Int64"
  third_party = ['"github.com/my/int64"']

In advanced scenarios it may be desirable to generate additional files that are not go code. You can accomplish this by using the --templates flag to specify all the directories you wish to generate code for. With this flag you specify root directories, that is top-level container directories.

If root directories have a _test suffix in the name, this folder is considered a folder full of templates for testing only and will be ommitted when --no-tests is specified and its templates will be generated into files with a _test suffix.

Each root directory is recursively walked. Each template found will be merged into table_name.ext where ext is defined by the shared extension of the templates. The directory structure is preserved with the exception of singletons.

For files that should not be generated for each model, you can use a singleton directory inside the directory where the singleton file should be generated. This will make sure that the file is only generated once.

Here's an example:

templates/
├── 00_struct.go.tpl               # Merged into output_dir/table_name.go
├── 00_struct.js.tpl               # Merged into output_dir/table_name.js
├── singleton
│   └── boil_queries.go.tpl        # Rendered as output_dir/boil_queries.go
└── js
    ├── jsmodel.js.tpl             # Merged into output_dir/js/table_name.js
    └── singleton
        └── jssingle.js.tpl        # Merged into output_dir/js/jssingle.js

The output files of which would be:

output_dir/
├── boil_queries.go
├── table_name.go
├── table_name.js
└── js
    ├── table_name.js
    └── jssingle.js

Note: Because the --templates flag overrides the internal bindata of sqlboiler, if you still wish to generate the default templates it's recommended that you include the path to sqlboiler's templates as well.

templates = [
  "/path/to/sqlboiler/templates",
  "/path/to/sqlboiler/templates_test"
  "/path/to/your_project/more_templates"
]

There will probably come a time when you want to extend the generated models with some kinds of helper functions. A general guideline is to put your extension functions into a separate package so that your functions aren't accidentally deleted when regenerating. Past that there are 3 main ways to extend the models, the first way is the most desirable:

Method 1: Simple Functions

// Package modext is for SQLBoiler helper methods
package modext

// UserFirstTimeSetup is an extension of the user model.
func UserFirstTimeSetup(ctx context.Context, db *sql.DB, u *models.User) error { ... }

Code organization is accomplished by using multiple files, and everything is passed as a parameter so these kinds of methods are very easy to test.

Calling code is also very straightforward:

user, err := Users().One(ctx, db)
// elided error check

err = modext.UserFirstTimeSetup(ctx, db, user)
// elided error check

Method 2: Empty struct methods

The above is the best way to code extensions for SQLBoiler, however there may be times when the number of methods grows too large and code completion is not as helpful anymore. In these cases you may consider structuring the code like this:

// Package modext is for SQLBoiler helper methods
package modext

type users struct {}

var Users = users{}

// FirstTimeSetup is an extension of the user model.
func (users) FirstTimeSetup(ctx context.Context, db *sql.DB, u *models.User) error { ... }

Calling code then looks a little bit different:

user, err := Users().One(ctx, db)
// elided error check

err = modext.Users.FirstTimeSetup(ctx, db, user)
// elided error check

This is almost identical to the method above, but gives slight amounts more organization at virtually no cost at runtime. It is however not as desirable as the first method since it does have some runtime cost and doesn't offer that much benefit over it.

Method 3: Embedding

This pattern is not for the faint of heart, what it provides in benefits it more than makes up for in downsides. It's possible to embed the SQLBoiler structs inside your own to enhance them. However it's subject to easy breakages and a dependency on these additional objects. It can also introduce inconsistencies as some objects may have no extended functionality and therefore have no reason to be embedded so you either have to have a struct for each generated struct even if it's empty, or have inconsistencies, some places where you use the enhanced model, and some where you do not.

user, err := Users().One(ctx, db)
// elided error check

enhUser := modext.User{user}
err = ehnUser.FirstTimeSetup(ctx, db)
// elided error check

I don't recommend this pattern, but included it so that people know it's an option and also know the problems with it.

The most common causes of problems and panics are:

• Forgetting to exclude tables you do not want included in your generation, like migration tables.
• Tables without a primary key. All tables require one.
• Forgetting to put foreign key constraints on your columns that reference other tables.
• The compatibility tests require privileges to create a database for testing purposes, ensure the user supplied in your sqlboiler.toml config has adequate privileges.
• A nil or closed database handle. Ensure your passed in boil.Executor is not nil.
  • If you decide to use the G variant of functions instead, make sure you've initialized your global database handle using boil.SetDB().
• Naming collisions, if the code fails to compile because there are naming collisions, look at the aliasing feature.
• A field not being inserted (usually a default true boolean), boil.Infer looks at the zero value of your Go type (it doesn't care what the default value in the database is) to determine if it should insert your field or not. In the case of a default true boolean value, when you want to set it to false; you set that in the struct but that's the zero value for the bool field in Go so sqlboiler assumes you do not want to insert that field and you want the default value from the database. Use a whitelist/greylist to add that field to the list of fields to insert.

For errors with other causes, it may be simple to debug yourself by looking at the generated code. Setting boil.DebugMode to true can help with this. You can change the output using boil.DebugWriter (defaults to os.Stdout).

If you're still stuck and/or you think you've found a bug, feel free to leave an issue and we'll do our best to help you.

Most examples in this section will be demonstrated using the following Postgres schema, structs and variables:

CREATE TABLE pilots (
  id integer NOT NULL,
  name text NOT NULL
);

ALTER TABLE pilots ADD CONSTRAINT pilot_pkey PRIMARY KEY (id);

CREATE TABLE jets (
  id integer NOT NULL,
  pilot_id integer NOT NULL,
  age integer NOT NULL,
  name text NOT NULL,
  color text NOT NULL
);

ALTER TABLE jets ADD CONSTRAINT jet_pkey PRIMARY KEY (id);
ALTER TABLE jets ADD CONSTRAINT jet_pilots_fkey FOREIGN KEY (pilot_id) REFERENCES pilots(id);

CREATE TABLE languages (
  id integer NOT NULL,
  language text NOT NULL
);

ALTER TABLE languages ADD CONSTRAINT language_pkey PRIMARY KEY (id);

-- Join table
CREATE TABLE pilot_languages (
  pilot_id integer NOT NULL,
  language_id integer NOT NULL
);

-- Composite primary key
ALTER TABLE pilot_languages ADD CONSTRAINT pilot_language_pkey PRIMARY KEY (pilot_id, language_id);
ALTER TABLE pilot_languages ADD CONSTRAINT pilot_language_pilots_fkey FOREIGN KEY (pilot_id) REFERENCES pilots(id);
ALTER TABLE pilot_languages ADD CONSTRAINT pilot_language_languages_fkey FOREIGN KEY (language_id) REFERENCES languages(id);

The generated model structs for this schema look like the following. Note that we've included the relationship structs as well so you can see how it all pieces together:

type Pilot struct {
  ID   int    `boil:"id" json:"id" toml:"id" yaml:"id"`
  Name string `boil:"name" json:"name" toml:"name" yaml:"name"`

  R *pilotR `boil:"-" json:"-" toml:"-" yaml:"-"`
  L pilotR  `boil:"-" json:"-" toml:"-" yaml:"-"`
}

type pilotR struct {
  Licenses  LicenseSlice
  Languages LanguageSlice
  Jets      JetSlice
}

type Jet struct {
  ID      int    `boil:"id" json:"id" toml:"id" yaml:"id"`
  PilotID int    `boil:"pilot_id" json:"pilot_id" toml:"pilot_id" yaml:"pilot_id"`
  Age     int    `boil:"age" json:"age" toml:"age" yaml:"age"`
  Name    string `boil:"name" json:"name" toml:"name" yaml:"name"`
  Color   string `boil:"color" json:"color" toml:"color" yaml:"color"`

  R *jetR `boil:"-" json:"-" toml:"-" yaml:"-"`
  L jetR  `boil:"-" json:"-" toml:"-" yaml:"-"`
}

type jetR struct {
  Pilot *Pilot
}

type Language struct {
  ID       int    `boil:"id" json:"id" toml:"id" yaml:"id"`
  Language string `boil:"language" json:"language" toml:"language" yaml:"language"`

  R *languageR `boil:"-" json:"-" toml:"-" yaml:"-"`
  L languageR  `boil:"-" json:"-" toml:"-" yaml:"-"`
}

type languageR struct {
  Pilots PilotSlice
}

// Open handle to database like normal
db, err := sql.Open("postgres", "dbname=fun user=abc")
if err != nil {
  return err
}

If your generated SQLBoiler models package can find columns with the names created_at or updated_at it will automatically set them to time.Now() in your database, and update your object appropriately. To disable this feature use --no-auto-timestamps.

Note: You can set the timezone for this feature by calling boil.SetLocation()

• Insert
  • Timestamps for both updated_at and created_at that are zero values will be set automatically.
  • To set the timestamp to null, set Valid to false and Time to a non-zero value. This is somewhat of a work around until we can devise a better solution in a later version.
• Update
  • The updated_at column will always be set to time.Now(). If you need to override this value you will need to fall back to another method in the meantime: queries.Raw(), overriding updated_at in all of your objects using a hook, or create your own wrapper.
• Upsert
  • created_at will be set automatically if it is a zero value, otherwise your supplied value will be used. To set created_at to null, set Valid to false and Time to a non-zero value.
  • The updated_at column will always be set to time.Now().

We generate "Starter" methods for you. These methods are named as the plural versions of your model, for example: models.Jets(). Starter methods are used to build queries using our Query Mod System. They take a slice of Query Mods as parameters, and end with a call to a Finisher method.

Here are a few examples:

// SELECT COUNT(*) FROM pilots;
count, err := models.Pilots().Count(ctx, db)

// SELECT * FROM "pilots" LIMIT 5;
pilots, err := models.Pilots(qm.Limit(5)).All(ctx, db)

// DELETE FROM "pilots" WHERE "id"=$1;
err := models.Pilots(qm.Where("id=?", 1)).DeleteAll(ctx, db)

In the event that you would like to build a query and specify the table yourself, you can do so using models.NewQuery():

// Select all rows from the pilots table by using the From query mod.
err := models.NewQuery(db, From("pilots")).All(ctx, db)

As you can see, Query Mods allow you to modify your queries, and Finishers allow you to execute the final action.

We also generate query building helper methods for your relationships as well. Take a look at our Relationships Query Building section for some additional query building information.

The query mod system allows you to modify queries created with Starter methods when performing query building. Here is a list of all of your generated query mods using examples:

// Dot import so we can access query mods directly instead of prefixing with "qm."
import . "github.com/volatiletech/sqlboiler/queries/qm"

// Use a raw query against a generated struct (Pilot in this example)
// If this query mod exists in your call, it will override the others.
// "?" placeholders are not supported here, use "$1, $2" etc.
SQL("select * from pilots where id=$1", 10)
models.Pilots(SQL("select * from pilots where id=$1", 10)).All()

Select("id", "name") // Select specific columns.
From("pilots as p") // Specify the FROM table manually, can be useful for doing complex queries.

// WHERE clause building
Where("name=?", "John")
And("age=?", 24)
Or("height=?", 183)

// WHERE IN clause building
WhereIn("name, age in ?", "John", 24, "Tim", 33) // Generates: WHERE ("name","age") IN (($1,$2),($3,$4))
AndIn("weight in ?", 84)
OrIn("height in ?", 183, 177, 204)

InnerJoin("pilots p on jets.pilot_id=?", 10)

GroupBy("name")
OrderBy("age, height")

Having("count(jets) > 2")

Limit(15)
Offset(5)

// Explicit locking
For("update nowait")

// Eager Loading -- Load takes the relationship name, ie the struct field name of the
// Relationship struct field you want to load. Optionally also takes query mods to filter on that query.
Load("Languages", Where(...)) // If it's a ToOne relationship it's in singular form, ToMany is plural.

Note: We don't force you to break queries apart like this if you don't want to, the following is also valid and supported by query mods that take a clause:

Where("(name=? OR age=?) AND height=?", "John", 24, 183)

Functions can have variations generated for them by using the flags --add-global-variants and --add-panic-variants. Once you've used these flags or set the appropriate values in your configuration file extra method overloads will be generated. We've used the Delete method to demonstrate:

// Set the global db handle for G method variants.
boil.SetDB(db)

pilot, _ := models.FindPilot(ctx, db, 1)

err := pilot.Delete(ctx, db) // Regular variant, takes a db handle (boil.Executor interface).
pilot.DeleteP(ctx, db)       // Panic variant, takes a db handle and panics on error.
err := pilot.DeleteG(ctx)    // Global variant, uses the globally set db handle (boil.SetDB()).
pilot.DeleteGP(ctx)          // Global&Panic variant, combines the global db handle and panic on error.

db.Begin()                   // Normal sql package way of creating a transaction
boil.BeginTx(ctx, nil)       // Uses the global database handle set by boil.SetDB() (doesn't require flag)

Note that it's slightly different for query building.

Here are a list of all of the finishers that can be used in combination with Query Building.

Finishers all have P (panic) method variations. To specify your db handle use the G or regular variation of the Starter method.

// These are called like the following:
models.Pilots().All(ctx, db)

One() // Retrieve one row as object (same as LIMIT(1))
All() // Retrieve all rows as objects (same as SELECT * FROM)
Count() // Number of rows (same as COUNT(*))
UpdateAll(models.M{"name": "John", "age": 23}) // Update all rows matching the built query.
DeleteAll() // Delete all rows matching the built query.
Exists() // Returns a bool indicating whether the row(s) for the built query exists.
Bind(&myObj) // Bind the results of a query to your own struct object.
Exec() // Execute an SQL query that does not require any rows returned.
QueryRow() // Execute an SQL query expected to return only a single row.
Query() // Execute an SQL query expected to return multiple rows.

We provide queries.Raw() for executing raw queries. Generally you will want to use Bind() with this, like the following:

err := queries.Raw(db, "select * from pilots where id=$1", 5).Bind(ctx, db, &obj)

You can use your own structs or a generated struct as a parameter to Bind. Bind supports both a single object for single row queries and a slice of objects for multiple row queries.

queries.Raw() also has a method that can execute a query without binding to an object, if required.

You also have models.NewQuery() at your disposal if you would still like to use Query Building in combination with your own custom, non-generated model.

For a comprehensive ruleset for Bind() you can refer to our godoc.

The Bind() Finisher allows the results of a query built with the Raw SQL method or the Query Builder methods to be bound to your generated struct objects, or your own custom struct objects.

This can be useful for complex queries, queries that only require a small subset of data and have no need for the rest of the object variables, or custom join struct objects like the following:

// Custom struct using two generated structs
type PilotAndJet struct {
  models.Pilot `boil:",bind"`
  models.Jet   `boil:",bind"`
}

var paj PilotAndJet
// Use a raw query
err := queries.Raw(db, `
  select pilots.id as "pilots.id", pilots.name as "pilots.name",
  jets.id as "jets.id", jets.pilot_id as "jets.pilot_id",
  jets.age as "jets.age", jets.name as "jets.name", jets.color as "jets.color"
  from pilots inner join jets on jets.pilot_id=?`, 23,
).Bind(&paj)

// Use query building
err := models.NewQuery(db,
  Select("pilots.id", "pilots.name", "jets.id", "jets.pilot_id", "jets.age", "jets.name", "jets.color"),
  From("pilots"),
  InnerJoin("jets on jets.pilot_id = pilots.id"),
).Bind(&paj)

// Custom struct for selecting a subset of data
type JetInfo struct {
  AgeSum int `boil:"age_sum"`
  Count int `boil:"juicy_count"`
}

var info JetInfo

// Use query building
err := models.NewQuery(db, Select("sum(age) as age_sum", "count(*) as juicy_count", From("jets"))).Bind(&info)

// Use a raw query
err := queries.Raw(db, `select sum(age) as "age_sum", count(*) as "juicy_count" from jets`).Bind(&info)

We support the following struct tag modes for Bind() control:

type CoolObject struct {
  // Don't specify a name, Bind will TitleCase the column
  // name, and try to match against this.
  Frog int

  // Specify an alternative name for the column, it will
  // be titlecased for matching, can be whatever you like.
  Cat int  `boil:"kitten"`

  // Ignore this struct field, do not attempt to bind it.
  Pig int  `boil:"-"`

  // Instead of binding to this as a regular struct field
  // (like other sql-able structs eg. time.Time)
  // Recursively search inside the Dog struct for field names from the query.
  Dog      `boil:",bind"`

  // Same as the above, except specify a different table name
  Mouse    `boil:"rodent,bind"`

  // Ignore this struct field, do not attempt to bind it.
  Bird     `boil:"-"`
}

Helper methods will be generated for every to one and to many relationship structure you have defined in your database by using foreign keys.

We attach these helpers directly to your model struct, for example:

jet, _ := models.FindJet(ctx, db, 1)

// "to one" relationship helper method.
// This will retrieve the pilot for the jet.
pilot, err := jet.Pilot().One(ctx, db)

// "to many" relationship helper method.
// This will retrieve all languages for the pilot.
languages, err := pilot.Languages().All(ctx, db)

If your relationship involves a join table SQLBoiler will figure it out for you transparently.

It is important to note that you should use Eager Loading if you plan on loading large collections of rows, to avoid N+1 performance problems.

For example, take the following:

// Avoid this loop query pattern, it is slow.
jets, _ := models.Jets().All(ctx, db)
pilots := make([]models.Pilot, len(jets))
for i := 0; i < len(jets); i++ {
  pilots[i] = jets[i].Pilot().OneP(ctx, db)
}

// Instead, use Eager Loading!
jets, _ := models.Jets(Load("Pilot")).All(ctx, db)

Eager loading can be combined with other query mods, and it can also eager load recursively.

// Example of a nested load.
// Each jet will have its pilot loaded, and each pilot will have its languages loaded.
jets, _ := models.Jets(Load("Pilot.Languages")).All(ctx, db)
// Note that each level of a nested Load call will be loaded. No need to call Load() multiple times.

// A larger example. In the below scenario, Pets will only be queried one time, despite
// showing up twice because they're the same query (the user's pets)
users, _ := models.Users(
  Load("Pets.Vets"),
  // the query mods passed in below only affect the query for Toys
  // to use query mods against Pets itself, you must declare it separately
  Load("Pets.Toys", Where("toys.deleted = ?", isDeleted)),
  Load("Property"),
  Where("age > ?", 23),
).All(ctx, db)

We provide the following methods for managing relationships on objects:

To One

• SetX(): Set the foreign key to point to something else: jet.SetPilot(...)
• RemoveX(): Null out the foreign key, effectively removing the relationship between these two objects: jet.RemovePilot(...)

To Many

• AddX(): Add more relationships to the existing set of related Xs: pilot.AddLanguages(...)
• SetX(): Remove all existing relationships, and replace them with the provided set: pilot.SetLanguages(...)
• RemoveX(): Remove all provided relationships: pilot.RemoveLanguages(...)

Important: Remember to use transactions around these set helpers for performance and data integrity. SQLBoiler does not do this automatically due to it's transparent API which allows you to batch any number of calls in a transaction without spawning subtransactions you don't know about or are not supported by your database.

To One code examples:

  jet, _ := models.FindJet(ctx, db, 1)
  pilot, _ := models.FindPilot(ctx, db, 1)

  // Set the pilot to an existing pilot
  err := jet.SetPilot(ctx, db, false, &pilot)

  pilot = models.Pilot{
    Name: "Erlich",
  }

  // Insert the pilot into the database and assign it to a jet
  err := jet.SetPilot(ctx, db, true, &pilot)

  // Remove a relationship. This method only exists for foreign keys that can be NULL.
  err := jet.RemovePilot(ctx, db, &pilot)

To Many code examples:

  pilots, _ := models.Pilots().All(ctx, db)
  languages, _ := models.Languages().All(ctx, db)

  // Set a group of language relationships
  err := pilots.SetLanguages(db, false, &languages)

  languages := []*models.Language{
    {Language: "Strayan"},
    {Language: "Yupik"},
    {Language: "Pawnee"},
  }

  // Insert new a group of languages and assign them to a pilot
  err := pilots.SetLanguages(ctx, db, true, languages...)

  // Add another language relationship to the existing set of relationships
  err := pilots.AddLanguages(ctx, db, false, &someOtherLanguage)

  anotherLanguage := models.Language{Language: "Archi"}

  // Insert and then add another language relationship
  err := pilots.AddLanguages(ctx, db, true, &anotherLanguage)

  // Remove a group of relationships
  err := pilots.RemoveLanguages(ctx, db, languages...)

Before and After hooks are available for most operations. If you don't need them you can shrink the size of the generated code by disabling them with the --no-hooks flag.

Every generated package that includes hooks has the following HookPoints defined:

const (
  BeforeInsertHook HookPoint = iota + 1
  BeforeUpdateHook
  BeforeDeleteHook
  BeforeUpsertHook
  AfterInsertHook
  AfterSelectHook
  AfterUpdateHook
  AfterDeleteHook
  AfterUpsertHook
)

To register a hook for your model you will need to create the hook function, and attach it with the AddModelHook method. Here is an example of a before insert hook:

// Define my hook function
func myHook(ctx context.Context, exec boil.ContextExecutor, p *Pilot) error {
  // Do stuff
  return nil
}

// Register my before insert hook for pilots
models.AddPilotHook(boil.BeforeInsertHook, myHook)

Your ModelHook will always be defined as func(context.Context, boil.ContextExecutor, *Model) error if context is not turned off.

The boil.Executor and boil.ContextExecutor interface powers all of SQLBoiler. This means anything that conforms to the three Exec/Query/QueryRow methods (and their context-aware variants) can be used to execute queries. sql.DB, sql.Tx as well as other libraries (sqlx) conform to this interface, and therefore any of these things may be used as an executor for any query in the system. This makes using transactions very simple:

tx, err := db.BeginTx(ctx, nil)
if err != nil {
  return err
}

users, _ := models.Pilots().All(ctx, tx)
users.DeleteAll(ctx, tx)

// Rollback or commit
tx.Commit()
tx.Rollback()

It's also worth noting that there's a way to take advantage of boil.SetDB() by using the boil.BeginTx() function. This opens a transaction using the globally stored database.

Debug logging will print your generated SQL statement and the arguments it is using. Debug logging can be toggled on globally by setting the following global variable to true:

boil.DebugMode = true

// Optionally set the writer as well. Defaults to os.Stdout
fh, _ := os.Open("debug.txt")
boil.DebugWriter = fh

Note: Debug output is messy at the moment. This is something we would like addressed.

Select is done through Query Building and Find. Here's a short example:

// Select one pilot
pilot, err := models.Pilots(qm.Where("name=?", "Tim")).One(ctx, db)

// Select specific columns of many jets
jets, err := models.Jets(qm.Select("age", "name")).All(ctx, db)

Find is used to find a single row by primary key:

// Retrieve pilot with all columns filled
pilot, err := models.FindPilot(ctx, db, 1)

// Retrieve a subset of column values
jet, err := models.FindJet(ctx, db, 1, "name", "color")

The main thing to be aware of with Insert is how the columns argument operates. You can supply one of the following column lists: boil.Infer, boil.Whitelist, boil.Blacklist, or boil.Greylist.

These lists control what fields are inserted into the database, and what values are returned to your struct from the database (default, auto incrementing, trigger-based columns are candidates for this). Your struct will have those values after the insert is complete.

When you use inference sqlboiler looks at your Go struct field values and if the field value is the zero value it will not insert that field, instead it will get the value from the database. Keep in mind sqlboiler cannot read or understand your default values set in the database, so the Go zero value is what's important here (this can be especially troubling for default true bool fields). Use a whitelist or greylist in cases where you want to insert a Go zero value.

See the documentation for boil.Columns.InsertColumnSet for more details.

var p1 models.Pilot
p1.Name = "Larry"
err := p1.Insert(ctx, db, boil.Infer()) // Insert the first pilot with name "Larry"
// p1 now has an ID field set to 1

var p2 models.Pilot
p2.Name "Boris"
err := p2.Insert(ctx, db, boil.Infer()) // Insert the second pilot with name "Boris"
// p2 now has an ID field set to 2

var p3 models.Pilot
p3.ID = 25
p3.Name = "Rupert"
err := p3.Insert(ctx, db, boil.Infer()) // Insert the third pilot with a specific ID
// The id for this row was inserted as 25 in the database.

var p4 models.Pilot
p4.ID = 0
p4.Name = "Nigel"
err := p4.Insert(ctx, db, boil.Whitelist("id", "name")) // Insert the fourth pilot with a zero value ID
// The id for this row was inserted as 0 in the database.
// Note: We had to use the whitelist for this, otherwise
// SQLBoiler would presume you wanted to auto-increment

Update can be performed on a single object, a slice of objects or as a Finisher for a collection of rows.

Update on a single object optionally takes a whitelist. The purpose of the whitelist is to specify which columns in your object should be updated in the database.

Like Insert, this method also takes a Columns type, but the behavior is slighty different. Although the descriptions below look similar the full documentation reveals the differences. Note that all inference is based on the Go types zero value and not the database default value, read the Insert documentation above for more details.

See the documentation for boil.Columns.UpdateColumnSet for more details.

// Find a pilot and update his name
pilot, _ := models.FindPilot(ctx, db, 1)
pilot.Name = "Neo"
rowsAff, err := pilot.Update(ctx, db)

// Update a slice of pilots to have the name "Smith"
pilots, _ := models.Pilots().All(ctx, db)
rowsAff, err := pilots.UpdateAll(ctx, db, models.M{"name": "Smith"})

// Update all pilots in the database to to have the name "Smith"
rowsAff, err := models.Pilots().UpdateAll(ctx, db, models.M{"name": "Smith"})

Delete a single object, a slice of objects or specific objects through Query Building.

pilot, _ := models.FindPilot(db, 1)
// Delete the pilot from the database
rowsAff, err := pilot.Delete(ctx, db)

// Delete all pilots from the database
rowsAff, err := models.Pilots().DeleteAll(ctx, db)

// Delete a slice of pilots from the database
pilots, _ := models.Pilots().All(ctx, db)
rowsAff, err := pilots.DeleteAll(ctx, db)

Upsert allows you to perform an insert that optionally performs an update when a conflict is found against existing row values.

The updateColumns and insertColumns operates in the same fashion that it does for Update and Insert.

If an insert is performed, your object will be updated with any missing default values from the database, such as auto-incrementing column values.

var p1 models.Pilot
p1.ID = 5
p1.Name = "Gaben"

// INSERT INTO pilots ("id", "name") VALUES($1, $2)
// ON CONFLICT DO NOTHING
err := p1.Upsert(ctx, db, false, nil, boil.Infer())

// INSERT INTO pilots ("id", "name") VALUES ($1, $2)
// ON CONFLICT ("id") DO UPDATE SET "name" = EXCLUDED."name"
err := p1.Upsert(ctx, db, true, []string{"id"}, boil.Whitelist("name"), boil.Infer())

// Set p1.ID to a zero value. We will have to use the whitelist now.
p1.ID = 0
p1.Name = "Hogan"

// INSERT INTO pilots ("id", "name") VALUES ($1, $2)
// ON CONFLICT ("id") DO UPDATE SET "name" = EXCLUDED."name"
err := p1.Upsert(ctx, db, true, []string{"id"}, boil.Whitelist("name"), boil.Whitelist("id", "name"))

The updateOnConflict argument allows you to specify whether you would like Postgres to perform a DO NOTHING on conflict, opposed to a DO UPDATE. For MySQL, this param will not be generated.

The conflictColumns argument allows you to specify the ON CONFLICT columns for Postgres. For MySQL, this param will not be generated.

Note: Passing a different set of column values to the update component is not currently supported.

Note: Upsert is now not guaranteed to be provided by SQLBoiler and it's now up to each driver individually to support it since it's a bit outside of the reach of the sql standard.

In the event that your objects get out of sync with the database for whatever reason, you can use Reload and ReloadAll to reload the objects using the primary key values attached to the objects.

pilot, _ := models.FindPilot(ctx, db, 1)

// > Object becomes out of sync for some reason, perhaps async processing

// Refresh the object with the latest data from the db
err := pilot.Reload(ctx, db)

// Reload all objects in a slice
pilots, _ := models.Pilots().All(ctx, db)
err := pilots.ReloadAll(ctx, db)

Note: Reload and ReloadAll are not recursive, if you need your relationships reloaded you will need to call the Reload methods on those yourself.

jet, err := models.FindJet(ctx, db, 1)

// Check if the pilot assigned to this jet exists.
exists, err := jet.Pilot(ctx, db).Exists()

// Check if the pilot with ID 5 exists
exists, err := models.Pilots(ctx, db, Where("id=?", 5)).Exists()

If your MySQL or Postgres tables use enums we will generate constants that hold their values that you can use in your queries. For example:

CREATE TYPE workday AS ENUM('monday', 'tuesday', 'wednesday', 'thursday', 'friday');

CREATE TABLE event_one (
  id     serial PRIMARY KEY NOT NULL,
  name   VARCHAR(255),
  day    workday NOT NULL
);

An enum type defined like the above, being used by a table, will generate the following enums:

const (
  WorkdayMonday    = "monday"
  WorkdayTuesday   = "tuesday"
  WorkdayWednesday = "wednesday"
  WorkdayThursday  = "thursday"
  WorkdayFriday    = "friday"
)

For Postgres we use enum type name + title cased value to generate the const variable name. For MySQL we use table name + column name + title cased value to generate the const variable name.

Note: If your enum holds a value we cannot parse correctly due, to non-alphabet characters for example, it may not be generated. In this event, you will receive errors in your generated tests because the value randomizer in the test suite does not know how to generate valid enum values. You will still be able to use your generated library, and it will still work as expected, but the only way to get the tests to pass in this event is to either use a parsable enum value or use a regular column instead of an enum.

The models package will also contain some structs that contain all of the table and column names harvested from the database at generation time. Eager loading constants are also generated mainly to avoid hardcoding and possible runtime issues.

For table names they're generated under models.TableNames:

// Generated code from models package
var TableNames = struct {
	Messages  string
	Purchases string
}{
	Messages:  "messages",
	Purchases: "purchases",
}

// Usage example:
fmt.Println(models.TableNames.Messages)

For column names they're generated under models.{Model}Columns:

// Generated code from models package
var MessageColumns = struct {
	ID         string
	PurchaseID string
}{
	ID:         "id",
	PurchaseID: "purchase_id",
}

// Usage example:
fmt.Println(models.MessageColumns.ID)

For eager loading relationships ther're generated under models.{Model}Rels:

// Generated code from models package
var MessageRels = struct {
	Purchase string
}{
	Purchase: "Purchase",
}

// Usage example:
fmt.Println(models.MessageRels.Purchase)

No, because Go's toolchain - unlike traditional toolchains - makes the compiler do most of the work instead of the linker. This means that when the first go install is done it can take a little bit of time because there is a lot of code that is generated. However, because of this work balance between the compiler and linker in Go, linking to that code afterwards in the subsequent compiles is extremely fast.

The generated models might import a couple of packages that are not on your system already, so cd into your generated models directory and type go get -u -t to fetch them. You will only need to run this command once, not per generation.

If your database uses multiple schemas you should generate a new package for each of your schemas. Note that this only applies to databases that use real, SQL standard schemas (like PostgreSQL), not fake schemas (like MySQL).

Only "escaped format" is supported for types.BytesArray. This means that your byte slice needs to have a format of "\x00" (4 bytes per byte) opposed to "\x00" (1 byte per byte). This is to maintain compatibility with all Postgres drivers. Example:

x := types.BytesArray{0: []byte("\\x68\\x69")}

Please note that multi-dimensional Postgres ARRAY types are not supported at this time.

You must use a DSN flag in MySQL connections, see: Requirements

The homepage for the SQLBoiler Golang ORM generator is located at: https://github.com/volatiletech/sqlboiler

If you'd like to run the benchmarks yourself check out our boilbench repo.

go test -bench . -benchmem

Test machine:

OS:  Ubuntu 16.04
CPU: Intel(R) Core(TM) i7-4771 CPU @ 3.50GHz
Mem: 16GB
Go:  go version go1.8.1 linux/amd64

The graphs below have many runs like this as input to calculate errors. Here is a sample run:

BenchmarkGORMSelectAll/gorm-8         20000   66500 ns/op   28998 B/op    455 allocs/op
BenchmarkGORPSelectAll/gorp-8         50000   31305 ns/op    9141 B/op    318 allocs/op
BenchmarkXORMSelectAll/xorm-8         20000   66074 ns/op   16317 B/op    417 allocs/op
BenchmarkKallaxSelectAll/kallax-8    100000   18278 ns/op    7428 B/op    145 allocs/op
BenchmarkBoilSelectAll/boil-8        100000   12759 ns/op    3145 B/op     67 allocs/op

BenchmarkGORMSelectSubset/gorm-8      20000    69469 ns/op   30008 B/op   462 allocs/op
BenchmarkGORPSelectSubset/gorp-8      50000    31102 ns/op    9141 B/op   318 allocs/op
BenchmarkXORMSelectSubset/xorm-8      20000    64151 ns/op   15933 B/op   414 allocs/op
BenchmarkKallaxSelectSubset/kallax-8 100000    16996 ns/op    6499 B/op   132 allocs/op
BenchmarkBoilSelectSubset/boil-8     100000    13579 ns/op    3281 B/op    71 allocs/op

BenchmarkGORMSelectComplex/gorm-8     20000    76284 ns/op   34566 B/op   521 allocs/op
BenchmarkGORPSelectComplex/gorp-8     50000    31886 ns/op    9501 B/op   328 allocs/op
BenchmarkXORMSelectComplex/xorm-8     20000    68430 ns/op   17694 B/op   464 allocs/op
BenchmarkKallaxSelectComplex/kallax-8 50000    26095 ns/op   10293 B/op   212 allocs/op
BenchmarkBoilSelectComplex/boil-8    100000    16403 ns/op    4205 B/op   102 allocs/op

BenchmarkGORMDelete/gorm-8           200000    10356 ns/op    5059 B/op    98 allocs/op
BenchmarkGORPDelete/gorp-8          1000000     1335 ns/op     352 B/op    13 allocs/op
BenchmarkXORMDelete/xorm-8           200000    10796 ns/op    4146 B/op   122 allocs/op
BenchmarkKallaxDelete/kallax-8       300000     5141 ns/op    2241 B/op    48 allocs/op
BenchmarkBoilDelete/boil-8          2000000      796 ns/op     168 B/op     8 allocs/op

BenchmarkGORMInsert/gorm-8           100000    15238 ns/op    8278 B/op   150 allocs/op
BenchmarkGORPInsert/gorp-8           300000     4648 ns/op    1616 B/op    38 allocs/op
BenchmarkXORMInsert/xorm-8           100000    12600 ns/op    6092 B/op   138 allocs/op
BenchmarkKallaxInsert/kallax-8       100000    15115 ns/op    6003 B/op   126 allocs/op
BenchmarkBoilInsert/boil-8          1000000     2249 ns/op     984 B/op    23 allocs/op

BenchmarkGORMUpdate/gorm-8           100000    18609 ns/op    9389 B/op   174 allocs/op
BenchmarkGORPUpdate/gorp-8           500000     3180 ns/op    1536 B/op    35 allocs/op
BenchmarkXORMUpdate/xorm-8           100000    13149 ns/op    5098 B/op   149 allocs/op
BenchmarkKallaxUpdate/kallax-8       100000    22880 ns/op   11366 B/op   219 allocs/op
BenchmarkBoilUpdate/boil-8          1000000     1810 ns/op     936 B/op    18 allocs/op

BenchmarkGORMRawBind/gorm-8           20000    65821 ns/op   30502 B/op   444 allocs/op
BenchmarkGORPRawBind/gorp-8           50000    31300 ns/op    9141 B/op   318 allocs/op
BenchmarkXORMRawBind/xorm-8           20000    62024 ns/op   15588 B/op   403 allocs/op
BenchmarkKallaxRawBind/kallax-8      200000     7843 ns/op    4380 B/op    46 allocs/op
BenchmarkSQLXRawBind/sqlx-8          100000    13056 ns/op    4572 B/op    55 allocs/op
BenchmarkBoilRawBind/boil-8          200000    11519 ns/op    4638 B/op    55 allocs/op