From 5164622ba462fe07fc9f2325fccf7f85aecb3ec8 Mon Sep 17 00:00:00 2001
From: Pieter Noordhuis <pietern@fb.com>
Date: Wed, 10 Apr 2019 22:21:45 -0700
Subject: [PATCH] Revert D14878128: [jit] Support attributes when copying
 modules

Differential Revision:
D14878128

Original commit changeset: 7ef5f7b1b16b

fbshipit-source-id: 3818222a897f8c01bc67f550ed0fd3ddecf61015
---
 test/test_jit.py               | 32 --------------------------------
 torch/csrc/jit/script/module.h | 10 ++++------
 2 files changed, 4 insertions(+), 38 deletions(-)

diff --git a/test/test_jit.py b/test/test_jit.py
index 26160ad..787dee4 100644
--- a/test/test_jit.py
+++ b/test/test_jit.py
@@ -3513,38 +3513,6 @@ a")
             for i in range(len(script_funs)):
                 self.assertEqual(test_func(script_funs[i], x, tensor), test_func(funs[i], x, tensor))
 
-    def test_module_copy_with_attributes(self):
-        class Vocabulary(torch.jit.ScriptModule):
-            def __init__(self, vocab_list):
-                super().__init__()
-                self._vocab = torch.jit.Attribute(vocab_list, List[str])
-                self.some_idx = torch.jit.Attribute(2, int)
-                self.idx = torch.jit.Attribute(
-                    {word: i for i, word in enumerate(vocab_list)}, Dict[str, int]
-                )
-
-            @torch.jit.script_method
-            def lookup_indices_1d(self, values):
-                # type: (List[str]) -> List[int]
-                result = torch.jit.annotate(List[int], [])
-                # Direct list iteration not supported
-                for i in range(len(values)):
-                    value = values[i]
-                    result.append(self.idx.get(value, self.some_idx))
-                return result
-
-            @torch.jit.script_method
-            def forward(self, values):
-                # type: (List[List[str]]) -> List[List[int]]
-                result = torch.jit.annotate(List[List[int]], [])
-                # Direct list iteration not supported
-                for i in range(len(values)):
-                    result.append(self.lookup_indices_1d(values[i]))
-                return result
-
-        v = Vocabulary(list('uabcdefg'))
-        v.copy()
-
     def test_tuple_to_opt_list(self):
         @torch.jit.script
         def foo(x):
diff --git a/torch/csrc/jit/script/module.h b/torch/csrc/jit/script/module.h
index d354b2b..13233d9 100644
--- a/torch/csrc/jit/script/module.h
+++ b/torch/csrc/jit/script/module.h
@@ -633,13 +633,11 @@ struct Module {
       parameter_remap[param] = curr->parameter_slot(param.name());
     }
     for (auto& attr : get_attributes()) {
-      if (attr.type()->isSubtypeOf(TensorType::get())) {
-        curr->register_buffer(attr.name(), attr.value().toTensor());
-        parameter_remap[attr] = *curr->find_buffer(attr.name());
-      } else {
-        curr->register_attribute(attr.name(), attr.type(), attr.value());
-        parameter_remap[attr] = *curr->find_attribute(attr.name());
+      if (!attr.type()->isSubtypeOf(TensorType::get())) {
+        continue;
       }
+      curr->register_buffer(attr.name(), attr.value().toTensor());
+      parameter_remap[attr] = *curr->find_buffer(attr.name());
     }
     for (auto& mod : get_modules()) {
       names.push_back(mod->name());
-- 
2.7.4