nGraph integration with TensorFlow

* Added nGraph bridge as a third_party to be built with TensorFlow based on user selection.
* Added a limited set of C++ unit tests to verify the correctness of the computation

1 files changed, 16 insertions, 0 deletions

diff --git a/third_party/ngraph/build_defs.bzl b/third_party/ngraph/build_defs.bzl
new file mode 100644
index 0000000000..2c9027a6b8
--- /dev/null
+++ b/third_party/ngraph/build_defs.bzl
@@ -0,0 +1,16 @@
+def clean_dep(dep):
+  return str(Label(dep))
+
+def if_ngraph(a):
+    """Shorthand for select()'ing on whether we're building with nGraph support.
+
+    Returns a select statement which evaluates to if_true if we're building
+    with nGraph.  Otherwise, the select statement evaluates to default.
+
+    """
+    ret_val = select({
+        clean_dep("//tensorflow:with_ngraph_support"): a,
+        "//conditions:default": []
+    })
+
+    return ret_val