Update path in generate_code.py

JDAI-CV · May 31, 2019 · 59646a6 · 59646a6
1 parent b5d6707
commit 59646a6
Showing 1 changed file with 5 additions and 12 deletions.
diff --git a/generate_code.py b/generate_code.py
@@ -5,8 +5,6 @@
 
 str_io = io.StringIO()
 
-optional_namespace = 'std'
-
 
 class Target(Enum):
     ModelBuilder = 1
@@ -39,7 +37,7 @@ def get_param(elem: dict) -> Tuple[str, str]:
     if elem['cpp_type'] == 'str':
         return 'const std::string &', elem['name']
     elif elem['cpp_type'] == 'optional_str':
-        return f'const {optional_namespace}::optional<std::string> &', elem['name']
+        return 'const dnn::optional<std::string> &', elem['name']
     elif elem['cpp_type'] == 'str_list':
         return 'const std::vector<std::string> &', elem['name']
     elif elem['cpp_type'] == 'int32_list':
@@ -177,9 +175,6 @@ def update_code(file: str, label: str) -> None:
 
 
 def generate_onnx_converter():
-    global optional_namespace
-    optional_namespace = 'nonstd'
-
     with open('ops.yml') as f:
         cfg = yaml.load(f)
     infer_cfg(cfg, Target.OnnxConverter)
@@ -250,12 +245,10 @@ def get_input_param(x):
         params = list(map(get_param, ipt_opt))
         params_str = ', '.join(map(lambda param: "{} {}".format(*param), params))
         cogoutl(f"void AddLayer{op['name']}{'' if op['converter'] else 'Impl'}({params_str});")
-    update_code('tools/onnx2daq/OnnxConverter.h', 'OnnxConverter auto generated methods')
+    update_code('include/tools/onnx2daq/OnnxConverter.h', 'OnnxConverter auto generated methods')
 
 
 def generate_model_builder():
-    global optional_namespace
-    optional_namespace = 'std'
     with open('ops.yml') as f:
         cfg = yaml.load(f)
     infer_cfg(cfg, Target.ModelBuilder)
@@ -266,7 +259,7 @@ def generate_model_builder():
         ipt_opt = op['input'] + op['output']
         params = list(map(get_param, ipt_opt))
         if op['support_quant_asymm']:
-            params.append((f'const {optional_namespace}::optional<QuantInfo> &', 'output_quant_info'))
+            params.append(('const dnn::optional<QuantInfo> &', 'output_quant_info'))
         params_str = ', '.join(map(lambda param: "{} {}".format(*param), params))
         cogoutl("ModelBuilder::Index ModelBuilder::Add{}({}) {{".format(op['name'], params_str))
         tensor_input = list(filter(lambda x: x['nnapi_type'] == 'tensor', op['input']))
@@ -310,11 +303,11 @@ def generate_model_builder():
         ipt_opt = op['input'] + op['output']
         params = list(map(get_param, ipt_opt))
         if op['support_quant_asymm']:
-            params.append((f'const {optional_namespace}::optional<QuantInfo> &', 'output_quant_info'))
+            params.append(('const dnn::optional<QuantInfo> &', 'output_quant_info'))
         params_str = ', '.join(map(lambda param: "{} {}".format(*param), params))
         cogoutl("ModelBuilder::Index Add{}({});".format(op['name'], params_str))
         cogoutl('#endif // __ANDROID_API__ >= {}'.format(op['api']))
-    update_code('dnnlibrary/include/ModelBuilder.h', 'ModelBuilder auto generated methods')
+    update_code('include/dnnlibrary/ModelBuilder.h', 'ModelBuilder auto generated methods')
 
 
 def main():