7. DDL, Inserting, and Updating
In this chapter we examine operations that modify data in the database, and ways to retrieve the results of these updates.
Setting Up
Again we set up a transactor and pull in YOLO mode, but this time we’re not using the world database.
import doobie.imports._
import scalaz._, Scalaz._
import scalaz.concurrent.Task
import scalaz.stream.Process
val xa = DriverManagerTransactor[Task](
"org.postgresql.Driver", "jdbc:postgresql:world", "postgres", ""
)
import xa.yolo._
Data Definition
It is uncommon to define database structures at runtime, but doobie handles it just fine and treats such operations like any other kind of update. And it happens to be useful here!
Let’s create a new table, which we will use for the examples to follow. This looks a lot like our prior usage of the sql
interpolator, but this time we’re using update
rather than query
. The .run
method gives a ConnectionIO[Int]
that yields the total number of rows modified, and the YOLO-mode .quick
gives a Task[Unit]
that prints out the row count.
val drop: Update0 =
sql"""
DROP TABLE IF EXISTS person
""".update
val create: Update0 =
sql"""
CREATE TABLE person (
id SERIAL,
name VARCHAR NOT NULL UNIQUE,
age SMALLINT
)
""".update
We can compose these and run them together.
scala> (drop.quick *> create.quick).unsafePerformSync
0 row(s) updated
0 row(s) updated
Inserting
Inserting is straightforward and works just as with selects. Here we define a method that constructs an Update0
that inserts a row into the person
table.
def insert1(name: String, age: Option[Short]): Update0 =
sql"insert into person (name, age) values ($name, $age)".update
Let’s insert a few rows.
scala> insert1("Alice", Some(12)).quick.run
<console>:29: warning: method run in class Task is deprecated: use unsafePerformSync
insert1("Alice", Some(12)).quick.run
^
1 row(s) updated
scala> insert1("Bob", None).quick.run
<console>:29: warning: method run in class Task is deprecated: use unsafePerformSync
insert1("Bob", None).quick.run
^
1 row(s) updated
And read them back.
case class Person(id: Long, name: String, age: Option[Short])
scala> sql"select id, name, age from person".query[Person].quick.run
<console>:30: warning: method run in class Task is deprecated: use unsafePerformSync
sql"select id, name, age from person".query[Person].quick.run
^
Person(1,Alice,Some(12))
Person(2,Bob,None)
Updating
Updating follows the same pattern. Here we update Alice’s age.
scala> sql"update person set age = 15 where name = 'Alice'".update.quick.run
<console>:28: warning: method run in class Task is deprecated: use unsafePerformSync
sql"update person set age = 15 where name = 'Alice'".update.quick.run
^
1 row(s) updated
scala> sql"select id, name, age from person".query[Person].quick.run
<console>:30: warning: method run in class Task is deprecated: use unsafePerformSync
sql"select id, name, age from person".query[Person].quick.run
^
Person(2,Bob,None)
Person(1,Alice,Some(15))
Retrieving Results
When we insert we usually want the new row back, so let’s do that. First we’ll do it the hard way, by inserting, getting the last used key via lastVal()
, then selecting the indicated row.
def insert2(name: String, age: Option[Short]): ConnectionIO[Person] =
for {
_ <- sql"insert into person (name, age) values ($name, $age)".update.run
id <- sql"select lastval()".query[Long].unique
p <- sql"select id, name, age from person where id = $id".query[Person].unique
} yield p
scala> insert2("Jimmy", Some(42)).quick.run
<console>:31: warning: method run in class Task is deprecated: use unsafePerformSync
insert2("Jimmy", Some(42)).quick.run
^
Person(3,Jimmy,Some(42))
This is irritating but it is supported by all databases (although the “get the last used id” function will vary by vendor).
Some database (like H2) allow you to return [only] the inserted id, allowing the above operation to be reduced to two statements (see below for an explanation of withUniqueGeneratedKeys
).
def insert2_H2(name: String, age: Option[Short]): ConnectionIO[Person] =
for {
id <- sql"insert into person (name, age) values ($name, $age)".update.withUniqueGeneratedKeys[Int]("id")
p <- sql"select id, name, age from person where id = $id".query[Person].unique
} yield p
scala> insert2_H2("Ramone", Some(42)).quick.run
<console>:31: warning: method run in class Task is deprecated: use unsafePerformSync
insert2_H2("Ramone", Some(42)).quick.run
^
Person(4,Ramone,Some(42))
Other databases (including PostgreSQL) provide a way to do this in one shot by returning multiple specified columns from the inserted row.
def insert3(name: String, age: Option[Short]): ConnectionIO[Person] = {
sql"insert into person (name, age) values ($name, $age)"
.update.withUniqueGeneratedKeys("id", "name", "age")
}
The withUniqueGeneratedKeys
specifies that we expect exactly one row back (otherwise an exception will be thrown), and requires a list of columns to return. This isn’t the most beautiful API but it’s what JDBC gives us. And it does work.
scala> insert3("Elvis", None).quick.run
<console>:31: warning: method run in class Task is deprecated: use unsafePerformSync
insert3("Elvis", None).quick.run
^
Person(5,Elvis,None)
This mechanism also works for updates, for databases that support it. In the case of multiple row updates we omit unique
and get a Process[ConnectionIO, Person]
back.
val up = {
sql"update person set age = age + 1 where age is not null"
.update.withGeneratedKeys[Person]("id", "name", "age")
}
Running this process updates all rows with a non-NULL
age and returns them.
scala> up.quick.run
<console>:31: warning: method run in class Task is deprecated: use unsafePerformSync
up.quick.run
^
Person(1,Alice,Some(16))
Person(3,Jimmy,Some(43))
Person(4,Ramone,Some(43))
scala> up.quick.run // and again!
<console>:31: warning: method run in class Task is deprecated: use unsafePerformSync
up.quick.run // and again!
^
Person(1,Alice,Some(17))
Person(3,Jimmy,Some(44))
Person(4,Ramone,Some(44))
Batch Updates
doobie supports batch updating via the updateMany
and updateManyWithGeneratedKeys
operations on the Update
data type (which we haven’t seen before). An Update0
, which is the type of a sql"..."
expression, represents a parameterized statement where the arguments are known. An Update[A]
is more general, and represents a parameterized statement where the composite argument of type A
is not known.
// Given some values ...
val a = 1; val b = "foo"
// this expression ...
sql"... $a $b ..."
// is syntactic sugar for this one, which is an Update applied to (a, b)
Update[(Int, String)]("... ? ? ...").run((a, b))
By using an Update
directly we can apply many sets of arguments to the same statement, and execute it as a single batch operation.
type PersonInfo = (String, Option[Short])
def insertMany(ps: List[PersonInfo]): ConnectionIO[Int] = {
val sql = "insert into person (name, age) values (?, ?)"
Update[PersonInfo](sql).updateMany(ps)
}
// Some rows to insert
val data = List[PersonInfo](
("Frank", Some(12)),
("Daddy", None))
Running this program yields the number of updated rows.
scala> insertMany(data).quick.run
<console>:31: warning: method run in class Task is deprecated: use unsafePerformSync
insertMany(data).quick.run
^
2
For databases that support it (such as PostgreSQL) we can use updateManyWithGeneratedKeys
to return a stream of updated rows.
def insertMany2(ps: List[PersonInfo]): Process[ConnectionIO, Person] = {
val sql = "insert into person (name, age) values (?, ?)"
Update[PersonInfo](sql).updateManyWithGeneratedKeys[Person]("id", "name", "age")(ps)
}
// Some rows to insert
val data = List[PersonInfo](
("Banjo", Some(39)),
("Skeeter", None),
("Jim-Bob", Some(12)))
Running this program yields the updated instances.
scala> insertMany2(data).quick.run
<console>:33: warning: method run in class Task is deprecated: use unsafePerformSync
insertMany2(data).quick.run
^
Person(8,Banjo,Some(39))
Person(9,Skeeter,None)
Person(10,Jim-Bob,Some(12))