Fixes for fuzzing tests reported on Gitter #4401
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Untyped trees may appear in programs that already had some errors reported.
We might provoke failures while checking whether all parts have a denotation. Also, if this is the case, we should not propagate the bad type.
|
This was quite enjoyable, trying to fix fuzzing problems in real time, as they are reported. But I won't be able to spend more time on this. So, reviewers, if you notice something, it would be good if you could fix it yourself and push on top of this PR. |
|
This is best reviewed commit by commit. |
Should not be attempted if argument does not have a value type.
Test case is this one. I don't want to include it in the rgeression tests though
because it looks too noisy (i.e. we would likely get spurious failures for tiny
changes in the future).
class i0 {
type Ordering[i1, i2] = (=> i3[i1])
def i4[i2, i3](i5: i1): i4[Any] // error
i8 += i14 {
type i4[+i9] = i5
def i7: Int = i6(5)
class i9[i10 <: i16 <: i4[_], i12](val i14: i11[_], i4: i11) { // error
def i16[i14 >: i1]: i0 = ???
implicit def i13[i4](i5: (=> i6)): i1[i6] // error
type i13[i7] = i13[i7]
def i8[+i1[+i2] <: i0[i1]]](i9: => ii5[i7]): i7[i5] = sys.error(); // error // error
def i10(i11: i4[i10.i4])(implicit i8: i0[i6#i0]) = i4 * i7 i5 i4! this match {
case i14(i9) => 3 }
}
class i10 {
type i13 = i10.i8
val i11 =
var i21 =
case class Tuple1(i3: i1, i4: i4, i6)
}
(new i10(i13))
}
}
package i1.i0
object i12 {
def i9(i10: i1)(implicit i9: i8[Boolean]) = implicitly[i9](i10: Any, i4, i9) {}
case class i9[i14[i2]](i16: i11[i8[i2]]) extends i10[i4, i14] {
def this(i11: i12) = {
val i5 = new i5[Int, Int]
}
}
smarter
reviewed
Apr 28, 2018
| @@ -1004,7 +1004,7 @@ object tpd extends Trees.Instance[Type] with TypedTreeInfo { | |||
| def sameTypes(trees: List[tpd.Tree], trees1: List[tpd.Tree]): Boolean = { | |||
| if (trees.isEmpty) trees.isEmpty | |||
| else if (trees1.isEmpty) trees.isEmpty | |||
| else (trees.head.tpe eq trees1.head.tpe) && sameTypes(trees.tail, trees1.tail) | |||
| else (trees.head.typeOpt eq trees1.head.typeOpt) && sameTypes(trees.tail, trees1.tail) | |||
There was a problem hiding this comment.
I think I have a better fix for that one: #4402
There was a problem hiding this comment.
I see you have already started doing this, so please go ahead. Otherwise I could have also approved yours and done the rest myself.
I assumed the existing tests were already sufficient to avoid confusion before
we try a substitution of one for the other, but was wrong in the case where
both sides are riddled with errors.
Test case (too long/fuzzy to be included in regression tests):
case class i0[i1, i4](i4: i1)
object i4 {
implicit def i4[i1: i0](i4: i1): i1
def i4[i4[_], i4[_], i4[i4]](i4: i1[i4]): i4[i4] = null
}
class i1 extends i4 {
case _ =>
}
val i1 = new i0 {
type i1[i4] = i4 with i0[i4]
trait i1 extends i4 {
type i1[i4] = i1 type i4 = i4#i4 def i4[i4 >: i4](i4: => i4): i4.i4[_ <: i4] = i4
i4(i1, i4: i4)
def unapply(i2: i4) = sys.error("")
}
trait i4
trait throw {
val i3: i4 = i4(5, i1: i1,
i4: i4)
def i3: i4[Int] = i1(null: Any) i3(0, i1())
println(i3 <= i4)
}
object i1 {
implicit def i1(i4: i4): Int
def i1[@specialized i4 <: i4[i4]](i1: i4): i4[Int]
def i1[@specialized i4, i4 <: i1](i4: i4): i4[i1[i4]] = new i4[i1] {
val i1 = (new i2.i1
}
}
This reverts commit f61596b. A better fix is implemented in the next commit
|
Some existing neg tests give more errors now that we check arguments of erroneous functions anyway. So these error counts need to be updated. Otherwise the last two commits LGTM. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.