vision-agent 0.2.238__tar.gz → 0.2.239__tar.gz

{vision_agent-0.2.238 → vision_agent-0.2.239}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vision-agent
-Version: 0.2.238
+Version: 0.2.239
 Summary: Toolset for Vision Agent
 Author: Landing AI
 Author-email: dev@landing.ai

{vision_agent-0.2.238 → vision_agent-0.2.239}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "vision-agent"
-version = "0.2.238"
+version = "0.2.239"
 description = "Toolset for Vision Agent"
 authors = ["Landing AI <dev@landing.ai>"]
 readme = "README.md"

{vision_agent-0.2.238 → vision_agent-0.2.239}/vision_agent/tools/tools.py

@@ -2804,7 +2804,7 @@ def save_video(
     else:
         Path(output_video_path).parent.mkdir(parents=True, exist_ok=True)
 
-    output_video_path = video_writer(frames, fps, output_video_path)
+    output_video_path = video_writer(frames, fps, filename=output_video_path)
     _save_video_to_result(output_video_path)
     return output_video_path
 

{vision_agent-0.2.238 → vision_agent-0.2.239}/vision_agent/utils/video.py

@@ -1,5 +1,5 @@
-import base64
 import logging
+import os
 import tempfile
 from functools import lru_cache
 from typing import List, Optional, Tuple
@@ -15,37 +15,6 @@ _DEFAULT_VIDEO_FPS = 24
 _DEFAULT_INPUT_FPS = 1.0
 
 
-def play_video(video_base64: str) -> None:
-    """Play a video file"""
-    video_data = base64.b64decode(video_base64)
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as temp_video:
-        temp_video.write(video_data)
-        temp_video_path = temp_video.name
-
-        cap = cv2.VideoCapture(temp_video_path)
-        if not cap.isOpened():
-            _LOGGER.error("Error: Could not open video.")
-            return
-
-        # Display the first frame and wait for any key press to start the video
-        ret, frame = cap.read()
-        if ret:
-            cv2.imshow("Video Player", frame)
-            _LOGGER.info(f"Press any key to start playing the video: {temp_video_path}")
-            cv2.waitKey(0)  # Wait for any key press
-
-        while cap.isOpened():
-            ret, frame = cap.read()
-            if not ret:
-                break
-            cv2.imshow("Video Player", frame)
-            # Press 'q' to exit the video
-            if cv2.waitKey(200) & 0xFF == ord("q"):
-                break
-        cap.release()
-        cv2.destroyAllWindows()
-
-
 def _resize_frame(frame: np.ndarray) -> np.ndarray:
     height, width = frame.shape[:2]
     new_width = width - (width % 2)
@@ -57,12 +26,15 @@ def video_writer(
     frames: List[np.ndarray],
     fps: float = _DEFAULT_INPUT_FPS,
     filename: Optional[str] = None,
+    file_ext: str = ".mp4",
 ) -> str:
+    tempf = None
     if isinstance(fps, str):
         # fps could be a string when it's passed in from a web endpoint deployment
         fps = float(fps)
     if filename is None:
-        filename = tempfile.NamedTemporaryFile(delete=False, suffix=".mp4").name
+        tempf = tempfile.NamedTemporaryFile(delete=False, suffix=file_ext)
+        filename = tempf.name
     container = av.open(filename, mode="w")
     stream = container.add_stream("h264", rate=fps)
     height, width = frames[0].shape[:2]
@@ -82,6 +54,9 @@ def video_writer(
     for packet in stream.encode():
         container.mux(packet)
     container.close()
+    # for windows nee to manually close tempfile, cannot use with NamedTemporaryFile(delete=True)
+    if tempf is not None:
+        tempf.close()
     return filename
 
 
@@ -98,11 +73,11 @@ def frames_to_bytes(
     if isinstance(fps, str):
         # fps could be a string when it's passed in from a web endpoint deployment
         fps = float(fps)
-    with tempfile.NamedTemporaryFile(delete=True, suffix=file_ext) as temp_file:
-        video_writer(frames, fps, temp_file.name)
-
-        with open(temp_file.name, "rb") as f:
-            buffer_bytes = f.read()
+    filename = video_writer(frames, fps, file_ext=file_ext)
+    # TODO: look into memory-mapped files to avoid reading the entire file into memory
+    with open(filename, "rb") as f:
+        buffer_bytes = f.read()
+    os.unlink(filename)
     return buffer_bytes