47 lines
1.0 KiB
YAML
47 lines
1.0 KiB
YAML
# 招聘数据采集服务配置文件
|
||
|
||
# 应用配置
|
||
app:
|
||
name: job-crawler
|
||
version: 1.0.0
|
||
debug: false
|
||
|
||
# 八爪鱼API配置
|
||
api:
|
||
base_url: https://openapi.bazhuayu.com
|
||
username: "13051331101"
|
||
password: "abc19910515"
|
||
batch_size: 100
|
||
# 多任务配置
|
||
tasks:
|
||
- id: "00f3b445-d8ec-44e8-88b2-4b971a228b1e"
|
||
name: "青岛招聘数据"
|
||
enabled: true
|
||
- id: "task-id-2"
|
||
name: "任务2"
|
||
enabled: false
|
||
- id: "task-id-3"
|
||
name: "任务3"
|
||
enabled: false
|
||
|
||
# RabbitMQ配置
|
||
rabbitmq:
|
||
host: rabbitmq
|
||
port: 5672
|
||
username: guest
|
||
password: guest
|
||
queue: job_data
|
||
message_ttl: 604800000 # 消息过期时间：7天(毫秒)
|
||
|
||
# 采集配置
|
||
crawler:
|
||
interval: 300 # 采集间隔(秒)
|
||
filter_days: 7 # 过滤天数
|
||
max_workers: 5 # 最大并行任务数
|
||
max_expired_batches: 3 # 连续过期批次阈值（首次采集时生效）
|
||
auto_start: true # 容器启动时自动开始采集
|
||
|
||
# 数据库配置
|
||
database:
|
||
path: data/crawl_progress.db
|