python-detector-worker/siwatsystem/pympta.py

import os
import json
import logging
import torch
import cv2
import requests
import zipfile
import shutil
import traceback
from ultralytics import YOLO
from urllib.parse import urlparse

# Create a logger specifically for this module
logger = logging.getLogger("detector_worker.pympta")

def load_pipeline_node(node_config: dict, mpta_dir: str) -> dict:
    # Recursively load a model node from configuration.
    model_path = os.path.join(mpta_dir, node_config["modelFile"])
    if not os.path.exists(model_path):
        logger.error(f"Model file {model_path} not found. Current directory: {os.getcwd()}")
        logger.error(f"Directory content: {os.listdir(os.path.dirname(model_path))}")
        raise FileNotFoundError(f"Model file {model_path} not found.")
    logger.info(f"Loading model for node {node_config['modelId']} from {model_path}")
    model = YOLO(model_path)
    if torch.cuda.is_available():
        logger.info(f"CUDA available. Moving model {node_config['modelId']} to GPU")
        model.to("cuda")
    else:
        logger.info(f"CUDA not available. Using CPU for model {node_config['modelId']}")
    node = {
        "modelId": node_config["modelId"],
        "modelFile": node_config["modelFile"],
        "triggerClasses": node_config.get("triggerClasses", []),
        "crop": node_config.get("crop", False),
        "minConfidence": node_config.get("minConfidence", None),
        "model": model,
        "branches": []
    }
    logger.debug(f"Configured node {node_config['modelId']} with trigger classes: {node['triggerClasses']}")
    for child in node_config.get("branches", []):
        logger.debug(f"Loading branch for parent node {node_config['modelId']}")
        node["branches"].append(load_pipeline_node(child, mpta_dir))
    return node

def load_pipeline_from_zip(zip_source: str, target_dir: str) -> dict:
    logger.info(f"Attempting to load pipeline from {zip_source} to {target_dir}")
    os.makedirs(target_dir, exist_ok=True)
    zip_path = os.path.join(target_dir, "pipeline.mpta")
    
    # Parse the source; only local files are supported here.
    parsed = urlparse(zip_source)
    if parsed.scheme in ("", "file"):
        local_path = parsed.path if parsed.scheme == "file" else zip_source
        logger.debug(f"Checking if local file exists: {local_path}")
        if os.path.exists(local_path):
            try:
                shutil.copy(local_path, zip_path)
                logger.info(f"Copied local .mpta file from {local_path} to {zip_path}")
            except Exception as e:
                logger.error(f"Failed to copy local .mpta file from {local_path}: {str(e)}", exc_info=True)
                return None
        else:
            logger.error(f"Local file {local_path} does not exist. Current directory: {os.getcwd()}")
            # List all subdirectories of models directory to help debugging
            if os.path.exists("models"):
                logger.error(f"Content of models directory: {os.listdir('models')}")
                for root, dirs, files in os.walk("models"):
                    logger.error(f"Directory {root} contains subdirs: {dirs} and files: {files}")
            else:
                logger.error("The models directory doesn't exist")
            return None
    else:
        logger.error(f"HTTP download functionality has been moved. Use a local file path here. Received: {zip_source}")
        return None

    try:
        if not os.path.exists(zip_path):
            logger.error(f"Zip file not found at expected location: {zip_path}")
            return None
            
        logger.debug(f"Extracting .mpta file from {zip_path} to {target_dir}")
        # Extract contents and track the directories created
        extracted_dirs = []
        with zipfile.ZipFile(zip_path, "r") as zip_ref:
            file_list = zip_ref.namelist()
            logger.debug(f"Files in .mpta archive: {file_list}")
            
            # Extract and track the top-level directories
            for file_path in file_list:
                parts = file_path.split('/')
                if len(parts) > 1:
                    top_dir = parts[0]
                    if top_dir and top_dir not in extracted_dirs:
                        extracted_dirs.append(top_dir)
            
            # Now extract the files
            zip_ref.extractall(target_dir)
            
        logger.info(f"Successfully extracted .mpta file to {target_dir}")
        logger.debug(f"Extracted directories: {extracted_dirs}")
        
        # Check what was actually created after extraction
        actual_dirs = [d for d in os.listdir(target_dir) if os.path.isdir(os.path.join(target_dir, d))]
        logger.debug(f"Actual directories created: {actual_dirs}")
    except zipfile.BadZipFile as e:
        logger.error(f"Bad zip file {zip_path}: {str(e)}", exc_info=True)
        return None
    except Exception as e:
        logger.error(f"Failed to extract .mpta file {zip_path}: {str(e)}", exc_info=True)
        return None
    finally:
        if os.path.exists(zip_path):
            os.remove(zip_path)
            logger.debug(f"Removed temporary zip file: {zip_path}")

    # Use the first extracted directory if it exists, otherwise use the expected name
    pipeline_name = os.path.basename(zip_source)
    pipeline_name = os.path.splitext(pipeline_name)[0]
    
    # Find the directory with pipeline.json
    mpta_dir = None
    # First try the expected directory name
    expected_dir = os.path.join(target_dir, pipeline_name)
    if os.path.exists(expected_dir) and os.path.exists(os.path.join(expected_dir, "pipeline.json")):
        mpta_dir = expected_dir
        logger.debug(f"Found pipeline.json in the expected directory: {mpta_dir}")
    else:
        # Look through all subdirectories for pipeline.json
        for subdir in actual_dirs:
            potential_dir = os.path.join(target_dir, subdir)
            if os.path.exists(os.path.join(potential_dir, "pipeline.json")):
                mpta_dir = potential_dir
                logger.info(f"Found pipeline.json in directory: {mpta_dir} (different from expected: {expected_dir})")
                break
    
    if not mpta_dir:
        logger.error(f"Could not find pipeline.json in any extracted directory. Directory content: {os.listdir(target_dir)}")
        return None
        
    pipeline_json_path = os.path.join(mpta_dir, "pipeline.json")
    if not os.path.exists(pipeline_json_path):
        logger.error(f"pipeline.json not found in the .mpta file. Files in directory: {os.listdir(mpta_dir)}")
        return None

    try:
        with open(pipeline_json_path, "r") as f:
            pipeline_config = json.load(f)
        logger.info(f"Successfully loaded pipeline configuration from {pipeline_json_path}")
        logger.debug(f"Pipeline config: {json.dumps(pipeline_config, indent=2)}")
        return load_pipeline_node(pipeline_config["pipeline"], mpta_dir)
    except json.JSONDecodeError as e:
        logger.error(f"Error parsing pipeline.json: {str(e)}", exc_info=True)
        return None
    except KeyError as e:
        logger.error(f"Missing key in pipeline.json: {str(e)}", exc_info=True)
        return None
    except Exception as e:
        logger.error(f"Error loading pipeline.json: {str(e)}", exc_info=True)
        return None

def run_pipeline(frame, node: dict, return_bbox: bool = False, is_last_stage: bool = True):
    """
    Processes the frame with the given pipeline node. When return_bbox is True,
    the function returns a tuple (detection, bbox) where bbox is (x1,y1,x2,y2)
    for drawing. Otherwise, returns only the detection.
    
    The is_last_stage parameter controls whether this node is considered the last
    in the pipeline chain. Only the last stage will return detection results.
    """
    try:
        # Check model type and use appropriate method
        model_task = getattr(node["model"], "task", None)
        
        if model_task == "classify":
            # Classification models need to use predict() instead of track()
            logger.debug(f"Running classification model: {node.get('modelId')}")
            results = node["model"].predict(frame, stream=False)
            detection = None
            best_box = None
            
            # Process classification results
            for r in results:
                probs = r.probs
                if probs is not None and len(probs) > 0:
                    # Get the most confident class
                    class_id = int(probs.top1)
                    conf = float(probs.top1conf)
                    detection = {
                        "class": node["model"].names[class_id],
                        "confidence": conf,
                        "id": None  # Classification doesn't have tracking IDs
                    }
                    logger.debug(f"Classification detection: {detection}")
                else:
                    logger.debug(f"Empty classification results for model {node.get('modelId')}")
            
            # Classification doesn't produce bounding boxes
            bbox = None
            
        else:
            # Detection/segmentation models use tracking
            logger.debug(f"Running detection/tracking model: {node.get('modelId')}")
            results = node["model"].track(frame, stream=False, persist=True)
            detection = None
            best_box = None
            max_conf = -1

            # Log raw detection count
            detection_count = 0
            for r in results:
                if hasattr(r.boxes, 'cpu') and len(r.boxes.cpu()) > 0:
                    detection_count += len(r.boxes.cpu())
            
            if detection_count == 0:
                logger.debug(f"Empty detection results (no objects found) for model {node.get('modelId')}")
            else:
                logger.debug(f"Detection model {node.get('modelId')} found {detection_count} objects")

            for r in results:
                for box in r.boxes:
                    box_cpu = box.cpu()
                    conf = float(box_cpu.conf[0])
                    if conf > max_conf and hasattr(box, "id") and box.id is not None:
                        max_conf = conf
                        detection = {
                            "class": node["model"].names[int(box_cpu.cls[0])],
                            "confidence": conf,
                            "id": box.id.item()
                        }
                        best_box = box_cpu
            
            if detection:
                logger.debug(f"Best detection: {detection}")
            else:
                logger.debug(f"No valid detection with tracking ID for model {node.get('modelId')}")

            bbox = None
            # Calculate bbox if best_box exists
            if detection and best_box is not None:
                coords = best_box.xyxy[0]
                x1, y1, x2, y2 = map(int, coords)
                h, w = frame.shape[:2]
                x1, y1 = max(0, x1), max(0, y1)
                x2, y2 = min(w, x2), min(h, y2)
                if x2 > x1 and y2 > y1:
                    bbox = (x1, y1, x2, y2)
                    logger.debug(f"Detection bounding box: {bbox}")
                    if node.get("crop", False):
                        frame = frame[y1:y2, x1:x2]
                        logger.debug(f"Cropped frame to {frame.shape}")

        # Check if we should process branches
        if detection is not None:
            for branch in node["branches"]:
                if detection["class"] in branch.get("triggerClasses", []):
                    min_conf = branch.get("minConfidence")
                    if min_conf is not None and detection["confidence"] < min_conf:
                        logger.debug(f"Confidence {detection['confidence']} below threshold {min_conf} for branch {branch['modelId']}.")
                        break
                    
                    # If we have branches, this is not the last stage
                    branch_result = run_pipeline(frame, branch, return_bbox, is_last_stage=True)
                    
                    # This node is no longer the last stage, so its results shouldn't be returned
                    is_last_stage = False
                    
                    if branch_result is not None:
                        if return_bbox:
                            return branch_result
                        return branch_result
                    break
            
            # Return this node's detection only if it's considered the last stage
            if is_last_stage:
                if return_bbox:
                    return detection, bbox
                return detection
            
        # No detection or not the last stage
        if return_bbox:
            return None, None
        return None
    except Exception as e:
        logger.error(f"Error running pipeline on node {node.get('modelId')}: {e}")
        if return_bbox:
            return None, None
        return None