Incorrect value of CI in FisherExactTest()

[muimuiz]

It seems that the upper bound of the confidence interval in FisherExactTest() returns an incorrect value (probably when the odds ratio is far from 1).

julia> confint(FisherExactTest(5,192,40,50))
(0.009676931449075, 4.503599627370496e15) <====

julia> confint(FisherExactTest(5,191,40,50))
(0.009727027600800141, 0.09011337150812752)

(@v1.5) pkg> status HypothesisTests
Status `C:\Users\****\.julia\environments\v1.5\Project.toml`
  [09f84164] HypothesisTests v0.10.4

The below is the result in R for the same case above.

R> fisher.test(matrix(c(5,192,40,50),2,2))$conf.int
[1] 0.009675738 0.089635757 <====
attr(,"conf.level")
[1] 0.95

R> fisher.test(matrix(c(5,191,40,50),2,2))$conf.int
[1] 0.009725712 0.090110378
attr(,"conf.level")
[1] 0.95