用 BeforeAndAfter 测试,table 放在后面,table 还在

Test with BeforeAndAfter, table drop in after, table still there

我正尝试在 tdd 中针对我一直在修改的想法进行一些集成测试。

我不是 TDD 的新手,但也不是专家。我昨天才开始在 ScalaTest 中使用它,所以我来了。

这是有效的方法:

class BookTest extends FlatSpec with Matchers with ScalaFutures with BeforeAndAfter
{
  implicit val defaultPatience =
    PatienceConfig(timeout = Span(5, Seconds), interval = Span(500, Millis))

  val db = Database.forURL("jdbc:sqlite:/home/teolha/test.sqlite", driver = "org.sqlite.JDBC")

  private val books = Books.all

  val setup = DBIO.seq(
    (books.schema).create,

    // Insert some dummy data
    books += new Book(-1, 0, "Awesome Title #1"),
    books += new Book(-1, 1, "Gorgeous Sequel"),
    books += new Book(-1, 2, "Mehmeh Prequel")
  )

  db.run(setup)

  "Books" should "be a Sequences of Books" in
  {
    val bookList: Future[Seq[Book]] = db.run(books.result)
    whenReady(bookList)
    {
      result => {
        result shouldBe a [Seq[_]] // Remember type erasure
        result(0) shouldBe a [Book]
      }

    }
  }

  "Books" should "contain the books we inserted" in
  {
    val bookList = db.run(books.result)
    whenReady(bookList)
    {
      result =>
      {
        result should have length 3
        (result(0).title shouldEqual "Awesome Title #1") (after being lowerCased)
        (result(1).title shouldEqual "Gorgeous Sequel") (after being lowerCased)
        (result(2).title shouldEqual "Mehmeh Prequel") (after being lowerCased)
      }
    }
  }
}

现在这个设置有多个问题:

  1. 我更喜欢内存数据库
  2. 我想在每次测试后删除所有 tables 并设置它们 在下一个之前再次向上并在特定测试中进行插入

所以我尝试了这样的事情:

class BookTest extends FlatSpec with Matchers with ScalaFutures with BeforeAndAfter {
  implicit val defaultPatience =
    PatienceConfig(timeout = Span(5, Seconds), interval = Span(500, Millis))

  val db = Database.forURL("jdbc:sqlite:/home/teolha/test.sqlite", driver = "org.sqlite.JDBC")

  private val books = Books.all

  before {
    db.run(setup)
  }

  after {
    db.run(tearDown)
  }

  val setup = DBIO.seq(
    (books.schema).create
  )

  val tearDown = DBIO.seq(
    (books.schema).drop
  )

  "Books" should "be a Sequences of Books" in {
    // Insert some dummy data

    db.run(
      DBIO.seq(
        books += new Book(-1, 0, "Awesome Title #1"),
        books += new Book(-1, 1, "Gorgeous Sequel"),
        books += new Book(-1, 2, "Mehmeh Prequel")
      )
    )

    val bookList: Future[Seq[Book]] = db.run(books.result)
    whenReady(bookList) {
      result => {
        result shouldBe a[Seq[_]] // Remember type erasure
        result(0) shouldBe a[Book]
      }

    }
  }
}

理论上这应该可行,不是吗?好吧,事实并非如此。测试后还有一个table,里面还是填满了数据

如何正确地做到这一点?之后我想把它放在一个内存数据库中(如上所述),但目前为了调试它更有意义。

其实我的做法是对的。唯一的问题是超时时间仍然设置得太低。我认为 5 秒就足够了,但我将其增加到 10 秒,现在一切正常!