Improved documentation

Added explanations where suggested by Adriaan in his review.

1 files changed, 8 insertions, 3 deletions

diff --git a/src/dotty/tools/dotc/core/Definitions.scala b/src/dotty/tools/dotc/core/Definitions.scala
index 6a318b0fe..5e335e240 100644
--- a/src/dotty/tools/dotc/core/Definitions.scala
+++ b/src/dotty/tools/dotc/core/Definitions.scala
@@ -437,13 +437,18 @@ class Definitions {
    *  to be the type parameters of a higher-kided type). This is a class symbol that
    *  would be generated by the following schema.
    *
-   *      class LambdaXYZ { type v_1 Arg1; ...; type v_N ArgN; type Apply }
+   *      class LambdaXYZ extends Object with P1 with ... with Pn {
+   *        type v_1 $hk$Arg0; ...; type v_N $hk$ArgN;
+   *        type Apply
+   *      }
    *
    *  Here:
    *
-   *   - XYZ is a string of length N with one letter for each variant of a bound symbols,
-   *     using `P` (positive variance), `N` (negative variance), `I` (invariant).
    *   - v_i are the variances of the bound symbols (i.e. +, -, or empty).
+   *   - XYZ is a string of length N with one letter for each variant of a bound symbol,
+   *     using `P` (positive variance), `N` (negative variance), `I` (invariant).
+   *   - for each positive or negative variance v_i there is a parent trait Pj which
+   *     is the same as LambdaXYZ except that it has `I` in i-th position.
    */
   def lambdaTrait(vcs: List[Int]): ClassSymbol = {
     assert(vcs.nonEmpty)