Test: libcrmcommon: Fix another iso8601 test for alpine linux.

Alpine uses musl by default, which handles bare percent signs in
strftime() differently.  On that platform, a bare percent sign (or an
unknown specifier) results in an empty string.  We need to check for
that and handle it in our test cases.

Author: clumens

lib/common/tests/iso8601/pcmk__time_format_hr_test.c

@@ -85,24 +85,43 @@ no_specifiers(void **state)
      * as "%%".
      */
     assert_hr_format("this has a literal % in it",
+#ifdef HAVE_STRFTIME_EMPTY_SPEC_FAILS
+                     NULL, NULL, 0);
+#else
                      "this has a literal % in it",
                      // *BSD strftime() strips single %
                      "this has a literal  in it", 0);
+#endif
+
     assert_hr_format("this has a literal %01 in it",
+#ifdef HAVE_STRFTIME_EMPTY_SPEC_FAILS
+                     NULL, NULL, 0);
+#else
                      "this has a literal %01 in it",
                      // *BSD strftime() strips %0
                      "this has a literal 1 in it", 0);
+#endif
+
     assert_hr_format("%2 this starts and ends with %",
+#ifdef HAVE_STRFTIME_EMPTY_SPEC_FAILS
+                     NULL, NULL, 0);
+#else
                      "%2 this starts and ends with %",
                      // *BSD strftime() strips % in front of nonzero number
                      "2 this starts and ends with %", 0);
+#endif
 
     /* strftime() treats % with a number (and no specifier) as a literal string
      * to be formatted with a field width (undocumented and probably a bug ...)
      */
-    assert_hr_format("this ends with %10", "this ends with        %10",
+    assert_hr_format("this ends with %10",
+#ifdef HAVE_STRFTIME_EMPTY_SPEC_FAILS
+                     NULL, NULL, 0);
+#else
+                     "this ends with        %10",
                      // *BSD strftime() strips % in front of nonzero number
                      "this ends with 10", 0);
+#endif
 }
 
 static void