Previously, symbols like int and int1 could have the same unique name
generated for them, depending on the fresh number used e.g. n=12: int ->
int12; n=2: int1 -> int12. This can break things further up that expect
unique names to be, well, unique. Adding a separator between the
variable and the fresh number prevents this type of edge case from
occurring.
Previously, symbols like int and int1 could have the same unique name generated for them, depending on the fresh number used e.g. n=12: int -> int12; n=2: int1 -> int12. This can break things further up that expect unique names to be, well, unique. Adding a separator between the variable and the fresh number prevents this type of edge case from occurring.
Also see the GitHub discussion for further info: https://github.com/camfort/fortran-src/issues/190