#4864 limit(atan2(0, cos(x)), x, %pi) gives ind instead of %pi

[David Scherfgen]

Spotted by Claude:

(%i1) limit(atan2(0, cos(x)), x, %pi);
(%o1) ind

This is incorrect. atan2(0, negative) is %pi. simplim%atan2 misses the case where the first argument is identical to zero, not an expression that approaches zero (with unknown direction).

[David Scherfgen]

Fixed by commit [1a175a].