Read env vars via os.getenv instead of os.environ

__init__.py

@@ -21,7 +21,7 @@ def detect_cloud_environment():
     
     # Check environment variables
     for indicator in cloud_indicators:
-        if indicator in os.environ:
+        if os.getenv(indicator) is not None:
             print(f"🌩️  Detected cloud environment: {indicator}")
             
             # Check CUDA version - only disable CUDA graph for CUDA 12
@@ -86,7 +86,7 @@ def _auto_install_tensorrt():
     This function installs only the TensorRT packages via pip.
     A marker file prevents repeated install attempts on every ComfyUI startup.
     """
-    disable_auto_install = os.environ.get("DISABLE_TENSORRT_AUTO_INSTALL", "false").lower() == "true"
+    disable_auto_install = os.getenv("DISABLE_TENSORRT_AUTO_INSTALL", "false").lower() == "true"
     if disable_auto_install:
         print("[ComfyUI-RIFE-TensorRT] Auto-installation disabled via DISABLE_TENSORRT_AUTO_INSTALL")
         return True
@@ -153,8 +153,8 @@ def _auto_install_tensorrt():
                 pass
 
         # Try CUDA_PATH
-        if not cuda_version and os.environ.get("CUDA_PATH"):
-            nvcc_path = os.path.join(os.environ["CUDA_PATH"], "bin", "nvcc")
+        if not cuda_version and os.getenv("CUDA_PATH"):
+            nvcc_path = os.path.join(os.getenv("CUDA_PATH"), "bin", "nvcc")
             if os.path.exists(nvcc_path):
                 try:
                     stdout = subprocess.check_output([nvcc_path, "--version"], text=True)
@@ -166,8 +166,8 @@ def _auto_install_tensorrt():
                     pass
 
         # Try CUDA_HOME
-        if not cuda_version and os.environ.get("CUDA_HOME"):
-            nvcc_path = os.path.join(os.environ["CUDA_HOME"], "bin", "nvcc")
+        if not cuda_version and os.getenv("CUDA_HOME"):
+            nvcc_path = os.path.join(os.getenv("CUDA_HOME"), "bin", "nvcc")
             if os.path.exists(nvcc_path):
                 try:
                     stdout = subprocess.check_output([nvcc_path, "--version"], text=True)
@@ -250,11 +250,11 @@ def _setup_cuda_dll_path():
     cuda_root = None
     
     # Check for CUDA_PATH or CUDA_HOME environment variables
-    cuda_root = os.environ.get("CUDA_PATH") or os.environ.get("CUDA_HOME")
+    cuda_root = os.getenv("CUDA_PATH") or os.getenv("CUDA_HOME")
     
     if not cuda_root:
         # Try default Windows install location
-        program_files = os.environ.get("PROGRAMFILES")
+        program_files = os.getenv("PROGRAMFILES")
         if program_files:
             cuda_base = Path(program_files) / "NVIDIA GPU Computing Toolkit" / "CUDA"
             if cuda_base.exists():

install.py

@@ -38,7 +38,7 @@ def detect_cuda_version():
             return version
 
     # Try CUDA_PATH
-    cuda_path = os.environ.get("CUDA_PATH")
+    cuda_path = os.getenv("CUDA_PATH")
     if cuda_path:
         nvcc_path = os.path.join(cuda_path, "bin", "nvcc")
         if os.path.exists(nvcc_path):
@@ -51,7 +51,7 @@ def detect_cuda_version():
                     return version
 
     # Try CUDA_HOME
-    cuda_home = os.environ.get("CUDA_HOME")
+    cuda_home = os.getenv("CUDA_HOME")
     if cuda_home:
         nvcc_path = os.path.join(cuda_home, "bin", "nvcc")
         if os.path.exists(nvcc_path):