SlideShare a Scribd company logo

ScalaQuery

•Download as KEY, PDF•
•
T
Trond Marius Øvstetun

The document discusses ScalaQuery, a type-safe database API for Scala. It describes how ScalaQuery can be used to connect to a database, perform queries and updates, and map database tables to case classes through the use of tables. Tables allow defining columns and relations to other tables. Queries can be built declaratively through for-comprehensions and include filtering, joins, and parameters. The results of queries can be lists, single values or options. The document suggests moving to Slick for a more native Scala API.

Related slideshows

Understanding NFT Marketplace Ecosystem.pptx
Understanding NFT Marketplace Ecosystem.pptxUnderstanding NFT Marketplace Ecosystem.pptx
Understanding NFT Marketplace Ecosystem.pptx
 
CI/CD pipelines for CloudHub 2.0 - Wroclaw MuleSoft Meetup #2
CI/CD pipelines for CloudHub 2.0 - Wroclaw MuleSoft Meetup #2CI/CD pipelines for CloudHub 2.0 - Wroclaw MuleSoft Meetup #2
CI/CD pipelines for CloudHub 2.0 - Wroclaw MuleSoft Meetup #2
 
Top Ten UI/UX Design Companies in India
Top Ten UI/UX Design Companies in IndiaTop Ten UI/UX Design Companies in India
Top Ten UI/UX Design Companies in India
 
1 of 54
ScalaQuery flatMap(Oslo) Trond Marius Øvstetun © Mesan AS
Trond Marius Øvstetun • Sjefskonsulentog fagleder i Mesan • Lidenskapelig utvikler, arkitekt, teamleder ++ • Java, Scala, HTML/js • Alltid på jakt etter bedre, enklere måter © Mesan AS
Data • Vi har behov for å lagre data © Mesan AS
Data • Vi har behov for å lagre data • Ikke alle bruker eller kan bruke NoSQL © Mesan AS
Data • Vi har behov for å lagre data • Ikke alle bruker eller kan bruke NoSQL • konservativ it-avdeling? • legacy – eksisterende data • andre behov? © Mesan AS
Data • Vi har behov for å lagre data • Ikke alle bruker eller kan bruke NoSQL • konservativ it-avdeling? • legacy – eksisterende data • andre behov? • av og til er relasjonsmodellen riktig © Mesan AS
SQL • Hvordan jobber vi med RDBMS på best mulig måte? © Mesan AS
SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? © Mesan AS
SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? • ORM? © Mesan AS
SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? • ORM? • Hibernate? © Mesan AS
SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? • ORM? • Hibernate? • Anorm? Mapper? Record? © Mesan AS
ScalaQuery © Mesan AS
ScalaQuery A type-safe database API for Scala © Mesan AS
ScalaQuery A type-safe database API for Scala © Mesan AS
En enkel domenemodell «enumeration» Artist Genre Album name : String Rock 1 biography : Text name : String Pop * Classic mainGenre : Genre release : Date Blues founded : Date rating : Option[Int] Rap split : Option[Date] 1 HipHop Alternative * * * Song name : String Person duration : Int firstname : String trackNumber : Int lastname : String biography : Text © Mesan AS
Koble til en database • Med javax.sql.DataSource def myDs : DataSource = {...} val db = Database.forDataSource(myDs) © Mesan AS
Koble til en database • Alternativt © Mesan AS
Koble til en database • Alternativt Database.forName("jdbc/my_ds") //JNDI Database.forURL("jdbc:h2:test") © Mesan AS
Gjøre noe med databasen val db = {...} db withSession { do_something() } import Database.threadLocalSession © Mesan AS
StaticQuery Håndkodet SQL mot databasen © Mesan AS
Spørringer val q = StaticQuery[Int] + "select count(id) from Artists" val count = q.first() © Mesan AS
Spørringer val q = StaticQuery[Int, (Int, String)] + "select id, name from Artists where id = ?" val (id,name) = q(1001).first val artist : (Int, String) = q(1001).first © Mesan AS
Mapping av resultat case class Artist(id:Int, name:String) val q = StaticQuery.query[Int, (Int, String)]( "select id, name from Artists where id = ?") val qMapped = q.mapResult[Artist]({ case (a, b) => Artist(a, b) }) var a:Artist = qMapped(1001).first() © Mesan AS
Implisitt mapping av case class Artist(id:Int, name:String) implicit val artMapper = GetResult( (r: PositionedResult) => new Artist(r<<, r<<)) val q = StaticQuery.query[Int, Art]( "select id, name from Artists where id = ?") val a:Artist = q(1001).first © Mesan AS
Insert / Update import org.scalaquery.simple.StaticQuery._ val q = query[(String, String), Int]( "insert into Artists(name, biography) "+ "values (?, ?)") val q2 = q("Seigmen", "") q2.execute() © Mesan AS
ScalaQuery © Mesan AS
ScalaQuery På tide å få noe igjen? Hva om ... © Mesan AS
En enkel domenemodell «enumeration» Artist Genre Album name : String Rock 1 biography : Text name : String Pop * Classic mainGenre : Genre release : Date Blues founded : Date rating : Option[Int] Rap split : Option[Date] 1 HipHop Alternative * * * Song name : String Person duration : Int firstname : String trackNumber : Int lastname : String biography : Text © Mesan AS
Mapping til en tabell object Artists extends Table[ (Int, String, String, Genre.Genre, Date, Option[Date]) ]("ARTISTS") { ... } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
Mapping til en tabell object Artists extends Table (...) { def id = column[Int]("ID", O PrimaryKey) def name = column[String]("NAME") def biography = column[String]("BIOGRAPHY") def maingenre = column[Genre.Genre]("MAINGENRE") def founded = column[Date]("FOUNDED") def split = column[Option[Date]]("SPLIT") } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
Mapping til en tabell object Artists extends Table (...) { ... def * = id ~ name ~ biography ~ maingenre ~ founded ~ split def i = name ~ biography ~ maingenre ~ founded ~ split } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
Mapping med case-klasse case class Artist( id:Int,name:String, bio:String, mainGenre:Genre.Genre, founded:Date, split:Option[Date]) object Artists extends Table[Artist]("ARTISTS") { def * = id ~ .. ~ split <> (Artist, Artist.unapply _) } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
Bruke egne typer column[Genre.Genre]("MAINGENRE") object Genre extends Enumeration { type Genre = Value «enumeration» Genre val Rock = Value(1) Rock Pop Classic ... Blues Rap HipHop Alternative val Alternative = Value(7) } implicit val genreMapper = MappedTypeMapper.base[Genre.Genre, Int]( _.id, Genre(_)) © Mesan AS
Bruke egne typer column[Duration]("DURATION") case class Duration(mins:Int, secs:Int) implicit object durationMapper extends MappedTypeMapper[Duration, Int] with BaseTypeMapper[Duration] with NumericTypeMapper { def map(dur: Duration) = dur.mins * 60 + dur.secs def comap(secs: Int) = Duration(secs / 60, secs % 60) } © Mesan AS
Inserts val i: Int = Artists.i.insert( ("Seigmen", "", Genre.Rock, date("1989-12-27"), Some(date("2008-06-22")))) // i is number of rows affected val artists : List[(...)] = ... val i = Artists.i.insertAll(artists :_*) // i is Option[Int] - num affected © Mesan AS
Enkle spørringer def findArtist(id:Int) : Option[(...)] = { val q = Artists.createFinderBy(_.id) q(id).firstOption } © Mesan AS
Enkle spørringer def findArtist(id:Int) : Option[(...)] = { val q = Artists.createFinderBy(_.id) q(id).firstOption } def highRatedAlbums : List[...] = { val q = Albums.filter(_.rating === (Six:Rating)) q.list } def unRatedAlbums : List[...] = { val q = Albums.filter(_.rating isNull) q.list } © Mesan AS
Bruk av spørringer val resList = q.list val resOption = q.firstOption val res = q.first © Mesan AS
Mer generelle spørringer val q1 = Artists.filter(_.maingenre === Genre.Rock) © Mesan AS
Mer generelle spørringer val q1 = Artists.filter(_.maingenre === Genre.Rock) val q = for { a <- Artists if a.maingenre === Genre.Rock } yield a © Mesan AS
Generell spørring val aQuery = for { x1 <- generator x2 <- generator ... xN <- generator guard } yield projection © Mesan AS
Parametrisert val qArtist = for { n <- Parameters[String] a <- Artists if a.name === n } yield a.id ~ a.name © Mesan AS
Parametrisert val qArtist = for { n <- Parameters[String] a <- Artists if a.name === n } yield a.id ~ a.name val qTool = qArtist("Tool") val (id, name) = qTool.first © Mesan AS
Spørringer • Er lazy • Er immutable • Dele og gjenbruke • Byggeklosser © Mesan AS
Spørringer • Er lazy • Er immutable • Dele og gjenbruke • Byggeklosser val q1 = for (a <- Artists) yield a val q2 = for (a <- q1) yield a.id ~ a.name val q3 = q2.take(10).drop(50) © Mesan AS
Update val q = for { a <- Artists if a.id === 1001 } yield a.name q.update("updated") // num affected © Mesan AS
Delete val q = for { a <- Albums if a.id === 1001 } yield a q.delete © Mesan AS
Relasjoner / object Albums extends Table[(...)]("ALBUMS") { def artist_id = column[Int]("ARTIST_ID") def artist = foreignKey("albums_artists_fk", artist_id, Artists)(_.id) def * = id ~ name ~ release ~ rating ~ artist_id } Artist Album name : String 1 biography : Text name : String * mainGenre : Genre release : Date founded : Date rating : Option[Int] split : Option[Date] © Mesan AS
Joins Artist val q = for { Album name : String 1 biography : Text name : String * mainGenre : Genre release : Date al <- Albums founded : Date split : Option[Date] rating : Option[Int] a <- Artists if al.artist_id === a.id } yield a.name ~ al.name val q = for { al <- Albums a <- al.artist } yield a.name ~ al.name val q = for { (al, a) <- Albums innerJoin Artists on (_.artist_id is _.id) } yield a.name ~ al.name © Mesan AS
Veien videre? ScalaQuery => SLICK Scala Language Integrated © Mesan AS
SLICK • Generiskrammeverk for spørringer mot data • Ulike backends • SQL, NoSQL, json, WebServices, XML ... • Nærmere “vanlig” Scala-kode, backend som scala Collection © Mesan AS
Q&A © Mesan AS
Q&A © Mesan AS
takk for meg tmo@mesan.no @ovstetun github.com/ovstetun/scala-persistence © Mesan AS

More Related Content

ScalaQuery

  • 1. ScalaQuery flatMap(Oslo) Trond Marius Øvstetun © Mesan AS
  • 2. Trond Marius Øvstetun • Sjefskonsulentog fagleder i Mesan • Lidenskapelig utvikler, arkitekt, teamleder ++ • Java, Scala, HTML/js • Alltid på jakt etter bedre, enklere måter © Mesan AS
  • 3. Data • Vi har behov for å lagre data © Mesan AS
  • 4. Data • Vi har behov for å lagre data • Ikke alle bruker eller kan bruke NoSQL © Mesan AS
  • 5. Data • Vi har behov for å lagre data • Ikke alle bruker eller kan bruke NoSQL • konservativ it-avdeling? • legacy – eksisterende data • andre behov? © Mesan AS
  • 6. Data • Vi har behov for å lagre data • Ikke alle bruker eller kan bruke NoSQL • konservativ it-avdeling? • legacy – eksisterende data • andre behov? • av og til er relasjonsmodellen riktig © Mesan AS
  • 7. SQL • Hvordan jobber vi med RDBMS på best mulig måte? © Mesan AS
  • 8. SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? © Mesan AS
  • 9. SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? • ORM? © Mesan AS
  • 10. SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? • ORM? • Hibernate? © Mesan AS
  • 11. SQL • Hvordan jobber vi med RDBMS på best mulig måte? • JDBC? • ORM? • Hibernate? • Anorm? Mapper? Record? © Mesan AS
  • 12. ScalaQuery © Mesan AS
  • 13. ScalaQuery A type-safe database API for Scala © Mesan AS
  • 14. ScalaQuery A type-safe database API for Scala © Mesan AS
  • 15. En enkel domenemodell «enumeration» Artist Genre Album name : String Rock 1 biography : Text name : String Pop * Classic mainGenre : Genre release : Date Blues founded : Date rating : Option[Int] Rap split : Option[Date] 1 HipHop Alternative * * * Song name : String Person duration : Int firstname : String trackNumber : Int lastname : String biography : Text © Mesan AS
  • 16. Koble til en database • Med javax.sql.DataSource def myDs : DataSource = {...} val db = Database.forDataSource(myDs) © Mesan AS
  • 17. Koble til en database • Alternativt © Mesan AS
  • 18. Koble til en database • Alternativt Database.forName("jdbc/my_ds") //JNDI Database.forURL("jdbc:h2:test") © Mesan AS
  • 19. Gjøre noe med databasen val db = {...} db withSession { do_something() } import Database.threadLocalSession © Mesan AS
  • 20. StaticQuery Håndkodet SQL mot databasen © Mesan AS
  • 21. Spørringer val q = StaticQuery[Int] + "select count(id) from Artists" val count = q.first() © Mesan AS
  • 22. Spørringer val q = StaticQuery[Int, (Int, String)] + "select id, name from Artists where id = ?" val (id,name) = q(1001).first val artist : (Int, String) = q(1001).first © Mesan AS
  • 23. Mapping av resultat case class Artist(id:Int, name:String) val q = StaticQuery.query[Int, (Int, String)]( "select id, name from Artists where id = ?") val qMapped = q.mapResult[Artist]({ case (a, b) => Artist(a, b) }) var a:Artist = qMapped(1001).first() © Mesan AS
  • 24. Implisitt mapping av case class Artist(id:Int, name:String) implicit val artMapper = GetResult( (r: PositionedResult) => new Artist(r<<, r<<)) val q = StaticQuery.query[Int, Art]( "select id, name from Artists where id = ?") val a:Artist = q(1001).first © Mesan AS
  • 25. Insert / Update import org.scalaquery.simple.StaticQuery._ val q = query[(String, String), Int]( "insert into Artists(name, biography) "+ "values (?, ?)") val q2 = q("Seigmen", "") q2.execute() © Mesan AS
  • 26. ScalaQuery © Mesan AS
  • 27. ScalaQuery På tide å få noe igjen? Hva om ... © Mesan AS
  • 28. En enkel domenemodell «enumeration» Artist Genre Album name : String Rock 1 biography : Text name : String Pop * Classic mainGenre : Genre release : Date Blues founded : Date rating : Option[Int] Rap split : Option[Date] 1 HipHop Alternative * * * Song name : String Person duration : Int firstname : String trackNumber : Int lastname : String biography : Text © Mesan AS
  • 29. Mapping til en tabell object Artists extends Table[ (Int, String, String, Genre.Genre, Date, Option[Date]) ]("ARTISTS") { ... } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
  • 30. Mapping til en tabell object Artists extends Table (...) { def id = column[Int]("ID", O PrimaryKey) def name = column[String]("NAME") def biography = column[String]("BIOGRAPHY") def maingenre = column[Genre.Genre]("MAINGENRE") def founded = column[Date]("FOUNDED") def split = column[Option[Date]]("SPLIT") } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
  • 31. Mapping til en tabell object Artists extends Table (...) { ... def * = id ~ name ~ biography ~ maingenre ~ founded ~ split def i = name ~ biography ~ maingenre ~ founded ~ split } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
  • 32. Mapping med case-klasse case class Artist( id:Int,name:String, bio:String, mainGenre:Genre.Genre, founded:Date, split:Option[Date]) object Artists extends Table[Artist]("ARTISTS") { def * = id ~ .. ~ split <> (Artist, Artist.unapply _) } Artist name : String biography : Text mainGenre : Genre founded : Date © Mesan AS split : Option[Date]
  • 33. Bruke egne typer column[Genre.Genre]("MAINGENRE") object Genre extends Enumeration { type Genre = Value «enumeration» Genre val Rock = Value(1) Rock Pop Classic ... Blues Rap HipHop Alternative val Alternative = Value(7) } implicit val genreMapper = MappedTypeMapper.base[Genre.Genre, Int]( _.id, Genre(_)) © Mesan AS
  • 34. Bruke egne typer column[Duration]("DURATION") case class Duration(mins:Int, secs:Int) implicit object durationMapper extends MappedTypeMapper[Duration, Int] with BaseTypeMapper[Duration] with NumericTypeMapper { def map(dur: Duration) = dur.mins * 60 + dur.secs def comap(secs: Int) = Duration(secs / 60, secs % 60) } © Mesan AS
  • 35. Inserts val i: Int = Artists.i.insert( ("Seigmen", "", Genre.Rock, date("1989-12-27"), Some(date("2008-06-22")))) // i is number of rows affected val artists : List[(...)] = ... val i = Artists.i.insertAll(artists :_*) // i is Option[Int] - num affected © Mesan AS
  • 36. Enkle spørringer def findArtist(id:Int) : Option[(...)] = { val q = Artists.createFinderBy(_.id) q(id).firstOption } © Mesan AS
  • 37. Enkle spørringer def findArtist(id:Int) : Option[(...)] = { val q = Artists.createFinderBy(_.id) q(id).firstOption } def highRatedAlbums : List[...] = { val q = Albums.filter(_.rating === (Six:Rating)) q.list } def unRatedAlbums : List[...] = { val q = Albums.filter(_.rating isNull) q.list } © Mesan AS
  • 38. Bruk av spørringer val resList = q.list val resOption = q.firstOption val res = q.first © Mesan AS
  • 39. Mer generelle spørringer val q1 = Artists.filter(_.maingenre === Genre.Rock) © Mesan AS
  • 40. Mer generelle spørringer val q1 = Artists.filter(_.maingenre === Genre.Rock) val q = for { a <- Artists if a.maingenre === Genre.Rock } yield a © Mesan AS
  • 41. Generell spørring val aQuery = for { x1 <- generator x2 <- generator ... xN <- generator guard } yield projection © Mesan AS
  • 42. Parametrisert val qArtist = for { n <- Parameters[String] a <- Artists if a.name === n } yield a.id ~ a.name © Mesan AS
  • 43. Parametrisert val qArtist = for { n <- Parameters[String] a <- Artists if a.name === n } yield a.id ~ a.name val qTool = qArtist("Tool") val (id, name) = qTool.first © Mesan AS
  • 44. Spørringer • Er lazy • Er immutable • Dele og gjenbruke • Byggeklosser © Mesan AS
  • 45. Spørringer • Er lazy • Er immutable • Dele og gjenbruke • Byggeklosser val q1 = for (a <- Artists) yield a val q2 = for (a <- q1) yield a.id ~ a.name val q3 = q2.take(10).drop(50) © Mesan AS
  • 46. Update val q = for { a <- Artists if a.id === 1001 } yield a.name q.update("updated") // num affected © Mesan AS
  • 47. Delete val q = for { a <- Albums if a.id === 1001 } yield a q.delete © Mesan AS
  • 48. Relasjoner / object Albums extends Table[(...)]("ALBUMS") { def artist_id = column[Int]("ARTIST_ID") def artist = foreignKey("albums_artists_fk", artist_id, Artists)(_.id) def * = id ~ name ~ release ~ rating ~ artist_id } Artist Album name : String 1 biography : Text name : String * mainGenre : Genre release : Date founded : Date rating : Option[Int] split : Option[Date] © Mesan AS
  • 49. Joins Artist val q = for { Album name : String 1 biography : Text name : String * mainGenre : Genre release : Date al <- Albums founded : Date split : Option[Date] rating : Option[Int] a <- Artists if al.artist_id === a.id } yield a.name ~ al.name val q = for { al <- Albums a <- al.artist } yield a.name ~ al.name val q = for { (al, a) <- Albums innerJoin Artists on (_.artist_id is _.id) } yield a.name ~ al.name © Mesan AS
  • 50. Veien videre? ScalaQuery => SLICK Scala Language Integrated © Mesan AS
  • 51. SLICK • Generiskrammeverk for spørringer mot data • Ulike backends • SQL, NoSQL, json, WebServices, XML ... • Nærmere “vanlig” Scala-kode, backend som scala Collection © Mesan AS
  • 54. takk for meg tmo@mesan.no @ovstetun github.com/ovstetun/scala-persistence © Mesan AS

Editor's Notes

  1. \n
  2. \n
  3. \n
  4. \n
  5. \n
  6. \n
  7. \n
  8. \n
  9. \n
  10. \n
  11. \n
  12. \n
  13. \n
  14. \n
  15. Herfra er det nesten bare kode - ROP ut om det er noe!!!\n
  16. Herfra er det nesten bare kode - ROP ut om det er noe!!!\n
  17. Herfra er det nesten bare kode - ROP ut om det er noe!!!\n
  18. &amp;#x201C;min faste&amp;#x201D; modell jeg bruker n&amp;#xE5;r jeg tester nye systemer/teknologier etc\nhar laget ogs&amp;#xE5; i mongo - fungerer str&amp;#xE5;lende der ogs&amp;#xE5;.\n
  19. \n
  20. \n
  21. \n
  22. settes typisk opp med cake/avhengigheter eller hvordan man &amp;#xF8;nsker\nLIFT: i boot\n
  23. \n
  24. \n
  25. \n
  26. \n
  27. \n
  28. \n
  29. hva om jeg forteller ScalaQuery noe om databasen min?\n
  30. \n
  31. starte med &amp;#xE5; definere strukturen p&amp;#xE5; tabellen\n
  32. legge til kolonnenne som finnes\n
  33. definere projeksjoner p&amp;#xE5; kolonnene som finnes\n* M&amp;#xC5; definere STJERNE - denne brukes til og fra mapping som default\n
  34. kan gj&amp;#xF8;res med caseklasser -&gt; her er resten som tidligere\n&lt;&gt; kaller eg til-og-fra - og er laget for akkurat case-klasser\n
  35. ENUMS i scala, kan gj&amp;#xF8;re tilsvarende med CASE OBJECTS\n
  36. mapping til/fra immutable verdiobjekter\n
  37. kaller insert p&amp;#xE5; en projeksjon - med samme struktur som projeksjonen selv!\ngir kompileringsfeil ved feil struktur\ninsert er eager - gj&amp;#xF8;res med en gang mot databasen\n
  38. \n
  39. vanligste operatorer -&gt; trigger databasen!\n
  40. kan ogs&amp;#xE5; skrives som...\n
  41. Generator == tabell (eller en fremmedn&amp;#xF8;kkel)\n
  42. === mappes til = i basen\nlike\nupper, st&amp;#xF8;rreEnn, mindreEnn etc - alt som finnes i databasen\n
  43. ingen av disse g&amp;#xE5;r mot basen f&amp;#xF8;r man faktisk kj&amp;#xF8;rer en list/first\n
  44. sp&amp;#xF8;rring representerer WHERE-delen\nyield-delen det som skal oppdateres - og inputstruktur til .update-kallet\n
  45. sp&amp;#xF8;rring representerer WHERE-delen\n
  46. ForeignKeyQuery\n
  47. \n
  48. \n
  49. kommer for scala 2.10, bruker makroer\n
  50. \n
  51. \n