Handle statically linked case in macOS as well

It turned out macOS can have statically linked Python when it's
installed via conda:
#150 (comment)

So it seems `linked_libpython` (which calls `libdl`) is the only way
to reliably detect if the Python executable is statically linked or
not.  Since cd2e408 implements it for
Windows as well, we can now rely on `linked_libpython` everywhere
which simplifies core.py.

julia/core.py

@@ -39,7 +39,7 @@
 # this is python 3.3 specific
 from types import ModuleType, FunctionType
 
-from .find_libpython import find_libpython, normalize_path
+from .find_libpython import find_libpython, linked_libpython, normalize_path
 
 #-----------------------------------------------------------------------------
 # Classes and funtions
@@ -260,11 +260,7 @@ def isafunction(julia, julia_name, mod_name=""):
 
 def determine_if_statically_linked():
     """Determines if this python executable is statically linked"""
-    # Windows and OS X are generally always dynamically linked
-    if not sys.platform.startswith('linux'):
-        return False
-    lddoutput = subprocess.check_output(["ldd",sys.executable])
-    return not (b"libpython" in lddoutput)
+    return linked_libpython() is None
 
 
 JuliaInfo = namedtuple(
@@ -356,18 +352,6 @@ def is_compatible_exe(jlinfo, _debug=lambda *_: None):
         _debug("libpython cannot be read from PyCall/deps/deps.jl")
         return False
 
-    if determine_if_statically_linked():
-        _debug(sys.executable, "is statically linked.")
-        return False
-
-    # Note that the following check is OK since statically linked case
-    # is already excluded.
-    if is_same_path(jlinfo.pyprogramname, sys.executable):
-        # In macOS and Windows, find_libpython does not work as good
-        # as in Linux.  We add this shortcut so that PyJulia can work
-        # in those environments.
-        return True
-
     py_libpython = find_libpython()
     jl_libpython = normalize_path(jlinfo.libpython)
     _debug("py_libpython =", py_libpython)

test/test_find_libpython.py

@@ -1,5 +1,7 @@
+import subprocess
+import sys
+
 from julia.find_libpython import finding_libpython, linked_libpython
-from julia.core import determine_if_statically_linked
 
 try:
     unicode
@@ -14,6 +16,22 @@ def test_finding_libpython_yield_type():
 # let's just check returned type of finding_libpython.
 
 
+def determine_if_statically_linked():
+    """Determines if this python executable is statically linked"""
+    if not sys.platform.startswith('linux'):
+        # Assuming that Windows and OS X are generally always
+        # dynamically linked.  Note that this is not the case in
+        # Python installed via conda:
+        # https://github.com/JuliaPy/pyjulia/issues/150#issuecomment-432912833
+        # However, since we do not use conda in our CI, this function
+        # is OK to use in tests.
+        return False
+    lddoutput = subprocess.check_output(["ldd", sys.executable])
+    return not (b"libpython" in lddoutput)
+
+
 def test_linked_libpython():
+    # TODO: Special-case conda (check `sys.version`).  See the above
+    # comments in `determine_if_statically_linked.
     if not determine_if_statically_linked():
         assert linked_libpython() is not None