Splitdefinitions is wrong in case loops are not binary
Issue #523
resolved
[I currently have a solution through bootstrapping]
Splitdefinitions works by partitioning all rules of a def into different partitions if they define different symbols. Afterwards, it merges 2 such partitions if there is a loop possible among them.
This misses every loop that goes over more than 2 symbols.
a <- b b <- c c <- a
Comments (5)
-
-
reporter - marked as minor
-
reporter -
assigned issue to
- edited description
-
assigned issue to
-
-
assigned issue to
-
assigned issue to
-
- changed status to resolved
- Log in to comment
Disabled splitting of definitions
see issue
#523→ <<cset fd7fb682f2b9>>