Refactor.

Modify the way CDefinition is specialized.

test_parallel_vectorize_numpy.py

@@ -1,13 +1,20 @@
+'''
+Test parallel-vectorize with numpy.fromfunc.
+Uses the work load from test_parallel_vectorize.
+'''
+
 from test_parallel_vectorize import *
 
 import numpy as np
 
 def main():
     module = Module.new(__name__)
-    PUfuncDef = ParallelUFuncPosix.specialize(num_thread=2)
-    SPUF = SpecializedParallelUFunc.specialize(
-                                        PUfuncDef, UFuncCore_D_D, Work_D_D)
-    sppufunc = SPUF.define(module)
+
+    spufdef = SpecializedParallelUFunc(ParallelUFuncPosix(num_thread=2),
+                                       UFuncCore_D_D(),
+                                       Work_D_D())
+
+    sppufunc = spufdef(module)
 
     module.verify()
 
@@ -27,6 +34,11 @@ def main():
 
     ptr_t = long # py2 only
 
+    # Becareful that fromfunc does not provide full error checking yet.
+    # If typenum is out-of-bound, we have nasty memory corruptions.
+    # For instance, -1 for typenum will cause segfault.
+    # If elements of type-list (2nd arg) is tuple instead,
+    # there will also memory corruption. (Seems like code rewrite.)
     typenum = np.dtype(np.double).num
     ufunc = np.fromfunc([ptr_t(funcptr)], [[typenum, typenum]], 1, 1, [None])