backtesting.lib combing barssince and cross?

Currently in the documentation, it says barssince can be used like this:

def barssince
(
condition, default=inf)
Return the number of bars since condition sequence was last True, or if never, return default.

>>> barssince(self.data.Close > self.data.Open)
3

I have tried to use 2 versions of this
One:

if barrsince(self.data.EMA_9 > self.data.EMA_50) > 1

This works but the algo will re-enter a long position even if the bars crossed many bars ago which is something I don't want.

if barrsince(self.data.EMA_9 > self.data.EMA_50) < 5 

Something I'm trying to mitigate the issue from the first example.

barssince(cross(self.data.EMA_9 > self.data.EMA_50)) > 1

This code doesn't work because I get the following error:

barssince(condition, default)
def barssince(condition: Sequence[bool], default=np.inf) -> int:
          """
          Return the number of bars since `condition` sequence was last `True`,
          or if never, return `default`.
   (...)
             3
     [     """
--->     return next(compress(range(len(condition)), reversed(condition)), default)

TypeError: object of type 'numpy.bool_' has no len()

Is it possible to combine barrsince with cross?

Start in backtesting.lib with the barssince and cross entry points, then reproduce the reported numpy.bool_ TypeError using the examples in the issue. Check existing tests or documentation for these helpers and define a testable outcome for combining them. Done means the supported behavior is covered and the usage is no longer ambiguous.