feat(job_crawler): initialize job crawler service with kafka integration

- Add technical documentation (技术方案.md) with system architecture and design details
- Create FastAPI application structure with modular organization (api, core, models, services, utils)
- Implement job data crawler service with incremental collection from third-party API
- Add Kafka service integration with Docker Compose configuration for message queue
- Create data models for job listings, progress tracking, and API responses
- Implement REST API endpoints for data consumption (/consume, /status) and task management
- Add progress persistence layer using SQLite for tracking collection offsets
- Implement date filtering logic to extract data published within 7 days
- Create API client service for third-party data source integration
- Add configuration management with environment-based settings
- Include Docker support with Dockerfile and docker-compose.yml for containerized deployment
- Add logging configuration and utility functions for date parsing
- Include requirements.txt with all Python dependencies and README documentation
This commit is contained in:
2026-01-15 17:09:43 +08:00
commit ae681575b9
26 changed files with 1898 additions and 0 deletions

View File

@@ -0,0 +1,209 @@
"""多任务增量采集核心逻辑"""
import asyncio
import logging
from typing import Dict, Optional
from concurrent.futures import ThreadPoolExecutor
from app.services.api_client import api_client
from app.services.kafka_service import kafka_service
from app.services.progress_store import progress_store
from app.utils import is_within_days
from app.models import JobData
from app.core.config import settings, TaskConfig
logger = logging.getLogger(__name__)
class TaskCrawler:
"""单个任务采集器"""
def __init__(self, task_config: TaskConfig):
self.task_id = task_config.id
self.task_name = task_config.name or task_config.id
self.batch_size = settings.api.batch_size
self.filter_days = settings.crawler.filter_days
self._running = False
self._total_filtered = 0
self._total_produced = 0
@property
def is_running(self) -> bool:
return self._running
async def start(self, reset: bool = False):
"""开始采集"""
if self._running:
logger.warning(f"[{self.task_name}] 任务已在运行中")
return
self._running = True
self._total_filtered = 0
self._total_produced = 0
logger.info(f"[{self.task_name}] 开始采集任务")
try:
if reset:
progress_store.reset_progress(self.task_id)
current_offset = 0
else:
progress = progress_store.get_progress(self.task_id)
current_offset = progress.current_offset if progress else 0
total = await api_client.get_total_count(self.task_id)
logger.info(f"[{self.task_name}] 数据总数: {total}, 当前偏移: {current_offset}")
if current_offset >= total:
logger.info(f"[{self.task_name}] 数据已全部采集完成")
progress_store.save_progress(self.task_id, current_offset, total, "completed",
self._total_filtered, self._total_produced)
self._running = False
return
while current_offset < total and self._running:
try:
await self._crawl_batch(current_offset)
current_offset += self.batch_size
progress_store.save_progress(self.task_id, current_offset, total, "running",
self._total_filtered, self._total_produced)
progress_pct = min(100, current_offset / total * 100)
logger.info(f"[{self.task_name}] 进度: {progress_pct:.2f}% ({current_offset}/{total})")
await asyncio.sleep(0.5)
except Exception as e:
logger.error(f"[{self.task_name}] 采集批次失败: {e}")
await asyncio.sleep(5)
status = "completed" if current_offset >= total else "stopped"
progress_store.save_progress(self.task_id, current_offset, total, status,
self._total_filtered, self._total_produced)
logger.info(f"[{self.task_name}] 采集任务 {status}")
except Exception as e:
logger.error(f"[{self.task_name}] 采集任务异常: {e}")
progress_store.save_progress(self.task_id, 0, 0, "error",
self._total_filtered, self._total_produced)
finally:
self._running = False
async def _crawl_batch(self, offset: int):
"""采集一批数据"""
result = await api_client.fetch_data(self.task_id, offset, self.batch_size)
data_list = result.get("data", {}).get("data", [])
if not data_list:
return
filtered_jobs = []
for raw in data_list:
aae397 = raw.get("aae397", "")
collect_time = raw.get("Collect_time", "")
if is_within_days(aae397, collect_time, self.filter_days):
job = JobData.from_raw(raw)
job.task_id = self.task_id # 添加任务ID标识
filtered_jobs.append(job)
self._total_filtered += len(filtered_jobs)
if filtered_jobs:
produced = kafka_service.produce_batch(filtered_jobs)
self._total_produced += produced
def stop(self):
"""停止采集"""
logger.info(f"[{self.task_name}] 正在停止采集任务...")
self._running = False
def get_status(self) -> dict:
"""获取采集状态"""
stats = progress_store.get_stats(self.task_id)
if not stats:
return {
"task_id": self.task_id, "task_name": self.task_name,
"total": 0, "current_offset": 0, "progress": "0%",
"status": "idle", "last_update": "",
"filtered_count": 0, "produced_count": 0
}
total = stats.get("total", 0)
current = stats.get("current_offset", 0)
progress = f"{min(100, current / total * 100):.2f}%" if total > 0 else "0%"
return {
"task_id": self.task_id, "task_name": self.task_name,
"total": total, "current_offset": current, "progress": progress,
"status": stats.get("status", "idle"), "last_update": stats.get("last_update", ""),
"filtered_count": stats.get("filtered_count", 0),
"produced_count": stats.get("produced_count", 0)
}
class CrawlerManager:
"""多任务采集管理器"""
def __init__(self):
self._crawlers: Dict[str, TaskCrawler] = {}
self._executor = ThreadPoolExecutor(max_workers=settings.crawler.max_workers)
self._init_crawlers()
def _init_crawlers(self):
"""初始化所有启用的任务采集器"""
for task in settings.get_enabled_tasks():
self._crawlers[task.id] = TaskCrawler(task)
logger.info(f"初始化任务采集器: {task.name} ({task.id})")
def get_crawler(self, task_id: str) -> Optional[TaskCrawler]:
"""获取指定任务的采集器"""
return self._crawlers.get(task_id)
def get_all_crawlers(self) -> Dict[str, TaskCrawler]:
"""获取所有采集器"""
return self._crawlers
async def start_task(self, task_id: str, reset: bool = False) -> bool:
"""启动单个任务"""
crawler = self._crawlers.get(task_id)
if not crawler:
logger.error(f"任务不存在: {task_id}")
return False
if crawler.is_running:
logger.warning(f"任务已在运行: {task_id}")
return False
asyncio.create_task(crawler.start(reset))
return True
async def start_all(self, reset: bool = False):
"""启动所有任务"""
tasks = []
for task_id, crawler in self._crawlers.items():
if not crawler.is_running:
tasks.append(crawler.start(reset))
if tasks:
await asyncio.gather(*tasks)
def stop_task(self, task_id: str) -> bool:
"""停止单个任务"""
crawler = self._crawlers.get(task_id)
if not crawler:
return False
crawler.stop()
return True
def stop_all(self):
"""停止所有任务"""
for crawler in self._crawlers.values():
crawler.stop()
def get_status(self, task_id: str = None) -> dict:
"""获取状态"""
if task_id:
crawler = self._crawlers.get(task_id)
return crawler.get_status() if crawler else {}
# 返回所有任务状态
return {
"tasks": [c.get_status() for c in self._crawlers.values()],
"kafka_lag": kafka_service.get_lag(),
"running_count": sum(1 for c in self._crawlers.values() if c.is_running)
}
@property
def is_any_running(self) -> bool:
"""是否有任务在运行"""
return any(c.is_running for c in self._crawlers.values())
# 全局管理器实例
crawler_manager = CrawlerManager()