gh-155966: Correct handling of tanpi(0.5) (#155966)

As pointed out in the issue, the exceptional half-integer
(0.5, 1.5, etc) inputs of tanpi were not handled properly.

Now, these raise ValueError and this is tested by `test.test_math`
through its list of test cases.

Author: jepler

Lib/test/mathdata/math_testcases.txt

@@ -2247,3 +2247,8 @@ tanpi10275 tanpi -1.6591963470121216 -> 1.829921944286168
 tanpi20001 tanpi inf -> nan invalid
 tanpi20002 tanpi -inf -> nan invalid
 tanpi20003 tanpi nan -> nan
+
+tanpi30050 tanpi 0.5 -> inf invalid
+tanpi30055 tanpi -0.5 -> inf invalid
+tanpi30115 tanpi 1.5 -> inf invalid
+tanpi30118 tanpi -1.5 -> inf invalid

Modules/mathmodule.c

@@ -1343,10 +1343,10 @@ FUNC1D(tan, tan, 0,
 FUNC1(tanh, tanh, 0,
       "tanh($module, x, /)\n--\n\n"
       "Return the hyperbolic tangent of x.")
-FUNC1D(tanpi, m_tanpi, 1,
+FUNC1D(tanpi, m_tanpi, 0,
       "tanpi($module, x, /)\n--\n\n"
       "Return the tangent of x (measured in half-turns).",
-      "expected a finite input, got %s")
+      "expected a finite input not equal to a half-integer, got %s")
 
 /* Precision summation function as msum() by Raymond Hettinger in
    <https://code.activestate.com/recipes/393090-binary-floating-point-summation-accurate-to-full-p/>,