| 1 | import threading |
| 2 | from typing import Any, Callable, Dict |
| 3 | |
| 4 | from loguru import logger |
| 5 | |
| 6 | |
| 7 | class TaskQueueFullError(ValueError): |
| 8 | pass |
| 9 | |
| 10 | |
| 11 | class TaskManager: |
| 12 | def __init__(self, max_concurrent_tasks: int, max_queued_tasks: int = 100): |
| 13 | self.max_concurrent_tasks = max_concurrent_tasks |
| 14 | self.max_queued_tasks = max_queued_tasks |
| 15 | self.current_tasks = 0 |
| 16 | self.lock = threading.Lock() |
| 17 | self.queue = self.create_queue() |
| 18 | |
| 19 | def create_queue(self): |
| 20 | raise NotImplementedError() |
| 21 | |
| 22 | def add_task(self, func: Callable, *args: Any, **kwargs: Any): |
| 23 | with self.lock: |
| 24 | if self.current_tasks < self.max_concurrent_tasks: |
| 25 | logger.info( |
| 26 | f"add task: {func.__name__}, current_tasks: {self.current_tasks}" |
| 27 | ) |
| 28 | self.execute_task(func, *args, **kwargs) |
| 29 | else: |
| 30 | queue_size = self.queue_size() |
| 31 | # 并发数已满时才进入排队。队列必须有上限，否则匿名接口可以持续 |
| 32 | # 堆积任务对象和请求参数，最终造成内存耗尽或第三方 API 成本失控。 |
| 33 | if queue_size >= self.max_queued_tasks: |
| 34 | logger.warning( |
| 35 | f"reject task: {func.__name__}, queue_size: {queue_size}, " |
| 36 | f"max_queued_tasks: {self.max_queued_tasks}" |
| 37 | ) |
| 38 | raise TaskQueueFullError("task queue is full, please try again later") |
| 39 | |
| 40 | logger.info( |
| 41 | f"enqueue task: {func.__name__}, current_tasks: {self.current_tasks}, " |
| 42 | f"queue_size: {queue_size}" |
| 43 | ) |
| 44 | self.enqueue({"func": func, "args": args, "kwargs": kwargs}) |
| 45 | |
| 46 | def execute_task(self, func: Callable, *args: Any, **kwargs: Any): |
| 47 | thread = threading.Thread( |
| 48 | target=self.run_task, args=(func, *args), kwargs=kwargs |
| 49 | ) |
| 50 | thread.start() |
| 51 | |
| 52 | def run_task(self, func: Callable, *args: Any, **kwargs: Any): |
| 53 | try: |
| 54 | with self.lock: |
| 55 | self.current_tasks += 1 |
| 56 | func(*args, **kwargs) # call the function here, passing *args and **kwargs. |
| 57 | finally: |
| 58 | self.task_done() |
| 59 | |
| 60 | def check_queue(self): |
| 61 | with self.lock: |
| 62 | if ( |
| 63 | self.current_tasks < self.max_concurrent_tasks |
| 64 | and not self.is_queue_empty() |
| 65 | ): |
| 66 | task_info = self.dequeue() |
| 67 | func = task_info["func"] |
| 68 | args = task_info.get("args", ()) |
| 69 | kwargs = task_info.get("kwargs", {}) |
| 70 | self.execute_task(func, *args, **kwargs) |
| 71 | |
| 72 | def task_done(self): |
| 73 | with self.lock: |
| 74 | self.current_tasks -= 1 |
| 75 | self.check_queue() |
| 76 | |
| 77 | def enqueue(self, task: Dict): |
| 78 | raise NotImplementedError() |
| 79 | |
| 80 | def dequeue(self): |
| 81 | raise NotImplementedError() |
| 82 | |
| 83 | def is_queue_empty(self): |
| 84 | raise NotImplementedError() |
| 85 | |
| 86 | def queue_size(self): |
| 87 | raise NotImplementedError() |
| 88 |