SI-7841 More robust unspecialization of names

Names like `T1$sp`, which can arise from `AnyRef` specialization,
were leading to negative length Names if they ever passed through
`unspecializedName` or `splitSpecializedName`. This code path was
touched when printing the tree of a certain AnyRef specialized
classes after specialization, such as `AbstractPartialFunction`
(which had such specialization until a few commits ago.)

This commit handles that case correctly, and generally hardens
against unexpected names, which could pop up from third party
classes.

The documentation for `splitSpecializedName` transposed the
class and method specializations. The things you discover
when you turn examples in documentation in to test cases!

In addition, we now require non-negative length and offset in
`newTermName`

1 files changed, 2 insertions, 0 deletions

diff --git a/src/reflect/scala/reflect/internal/Names.scala b/src/reflect/scala/reflect/internal/Names.scala
index ed248d6e1e..f998d95349 100644
--- a/src/reflect/scala/reflect/internal/Names.scala
+++ b/src/reflect/scala/reflect/internal/Names.scala
@@ -90,6 +90,8 @@ trait Names extends api.Names {
    */
   final def newTermName(cs: Array[Char], offset: Int, len: Int, cachedString: String): TermName = {
     def body = {
+      require(offset >= 0, "offset must be non-negative, got " + offset)
+      require(len >= 0, "length must be non-negative, got " + len)
       val h = hashValue(cs, offset, len) & HASH_MASK
       var n = termHashtable(h)
       while ((n ne null) && (n.length != len || !equals(n.start, cs, offset, len)))