Fix #728 - Support case class with more 22 fields in Scala 3 #729
Conversation
cchantep
force-pushed
the
fix/728_macros
branch
2 times, most recently
from
2d5dedf to
989bddf
Compare
| } | ||
|
|
||
| "from generic Bar" in { | ||
| testWithTuple( | ||
| Bar[Double]("bar1", None, Seq(1.2D, 34.5D)) | ||
| ).mustEqual( | ||
| "scala.Tuple3[scala.Predef.String, scala.Option[scala.Double], scala.collection.immutable.Seq[scala.Double]]/Bar(bar1,None,List(1.2, 34.5))" | ||
| "play.api.libs.json.Bar[scala.Double]/Bar(bar1,None,List(1.2, 34.5))" |
There was a problem hiding this comment.
Case class type is directly usable (and more specific)
| val fieldNme = s"_${i + 1}" | ||
|
|
||
| val resolve: Term => Term = { | ||
| val field = tupleTpeSym.declaredField(fieldNme) |
There was a problem hiding this comment.
Either resolve as field or nullary method
| tt: Type[T], | ||
| ut: Type[U] | ||
| ): Tuple2[TypeRepr, Expr[T] => (Expr[U] => Expr[R]) => Expr[R]] = | ||
| TypeRepr.of(using ut) -> { |
There was a problem hiding this comment.
Directly use the given product type
| "no custom ProductOf" in { | ||
| "Macros.writer[CustomNoProductOf]".mustNot(typeCheck) | ||
| "be handled" when { | ||
| "is declared with more than 22 fields" in { |
cchantep
force-pushed
the
fix/728_macros
branch
from
989bddf to
e0ae719
Compare
cchantep
force-pushed
the
fix/728_macros
branch
from
e0ae719 to
4880314
Compare
| @@ -227,7 +227,8 @@ object Json extends JsonFacade with JsMacros with JsValueMacros { | |||
| JsValueWrapperImpl(w.writes(field)) | |||
|
|
|||
| def obj(fields: (String, JsValueWrapper)*): JsObject = JsObject(fields.map(f => (f._1, unwrap(f._2)))) | |||
| def arr(items: JsValueWrapper*): JsArray = JsArray(items.iterator.map(unwrap).toArray[JsValue]) | |||
|
|
|||
| @@ -143,6 +161,66 @@ case class Foo(bar: String, lorem: Int) | |||
|
|
|||
| case class Bar[T](name: String, opt: Option[T], scores: Seq[Double]) | |||
|
|
|||
| case class BigFat( | |||
There was a problem hiding this comment.
Case class with 26 fields
|
@ihostage is there a reason why this PR doesn't have any comments/feedback? I've run it locally and it seems to be working fine. The tests pass so I wonder why it's stopping us from merging and do a cutover for RC7. |
No special reason. Probably just got out of the radar. There is something wrong with CI, so I will close it and reopen to retrigger it, but I think it's ready for merge. |
Fix #728