aisr接入指引

aisr接入指引

包括自动读配置,建立redis队列,mq对象,还有redis队列的异步多进程处理.

1. 导入依赖

pip install aisrfwk==1.0.4 -i https://pypi.org/simple
如果是本地已经引用旧版本,需要使用更新到指定版本
pip3 install --upgrade aisrfwk==1.0.4 -i https://pypi.org/simple

2. 添加配置

根据环境配置.文件名固定server-{env}.conf格式

env即为运行启动命令后的第一个参数. 例如: python mian.py sit

配置文件会根据env加载对应的配置文件,环境支持,dev,uat,prod,默认为local

###====配置文件模板====####

# 系统主题,关系到错误提示.邮件主题等
system.name="地库排布系统"

# 环境 配置项
# env=local
# server 配置项
server.port=15555

# redis 配置项
redis.host=10.220.37.1
redis.port=16379
redis.passwd=***
redis.db=7
# 任务队列key
redis.task_queue_key=task_queue
# 任务状态
redis.task_status_key=task_status
# 任务请求数据
redis.task_data_key=task_data
# 任务过程发送邮件
redis.task_mail_key=task_mail

# mq 配置项
mq.addr=0.0.0.0:9876
mq.producer=aisr-carbarn-ai
mq.topic=aisr-carbarn-park-plan-test

# 图片上传接口
upload.url=https://ipark-dev.bgy.com.cn/background/file-service/file/v1/publicRead/upload


# 邮件主题
mail.subject="算法运行出错"
# 邮件标题
mail.title="智能地库算法排布运行错误"
# 邮件发送人简称
mail.send.user="智能地库"
# 邮件接收人简称
mail.to.user="智能地库管理员"

# 发送邮件配置
mail.host=smtp.countrygarden.com.cn
mail.port=587
mail.ssl=false
mail.user=[email protected]
mail.passwd=***
mail.send.nick=FDDS
# 多个邮件逗号隔开,不配合默认不发送
mail.to.users=[email protected]

3. 自行添加http接口

http使用flask配置

###====启动方法====####

def start():
    """启动web服务"""
    server_host = config.get("server.host", "0.0.0.0")
    server_port = config.get("server.port", 5000)
    log.info("当前配置.端口号:{},ip:{}", server_host, server_port)
    # scheduleTask()
    app.run(host=server_host, port=server_port)

api模板

import sys
import json

from aisrfwk.util import log
from aisrfwk.domain.base_result import Result
from flask import Flask, jsonify, request
from flask_cors import cross_origin
from aisrfwk.util.config import config, service_port
import aisrfwk.util.redis_util as redis

from aisrfwk.util import api

sys.path.append("..")

app = Flask(__name__)


@app.route('/park', methods=['POST'])
@cross_origin()
def park():
    """
    车位排布主方法
    """
    try:
        param = request.get_data()
        json_param = json.loads(param)
        validateParam(json_param)
        res = api.pushQueue({json_param)
        return Result().success(res)
    except Exception as e:
        log.info(type(e))
        log.error("车位排布任务启动失败:" + str(e))
        return Result().error(str(e))


def validateParam(param):
    """根据图纸生成方案参数校验"""
    assert param.get("projectId") is not None, '项目ID参数[projectId]必须传入'
    # assert param.get("params") is not None, '指标参数[params]必须传入'
    # assert param.get("outer") is not None, '地库退线参数[outer]必须传入'
    # assert param.get("building") is not None, '楼栋外轮廓参数[outer]必须传入'
    # assert param.get("column") is not None, '剪力墙参数[column]必须传入'
    # assert param.get("coreBarrel") is not None, '核心筒参数[coreBarrel]必须传入'


def start():
    """启动web服务"""
    server_host = config.get("server.host", "0.0.0.0")
    server_port = service_port if service_port != 0 else config.get("server.port", 5000)
    log.info("当前配置.端口号:{},ip:{}", server_host, server_port)
    # scheduleTask()
    app.run(host=server_host, port=server_port)


if __name__ == '__main__':
    # 本地运行Flask
    start()

4.加入业务逻辑处理类

创建业务逻辑处理类,需要继承BaseAisrTask,并且重写run方法

import sys
import time

from aisrfwk.domain.base_aisr_task import BaseAisrTask
from aisrfwk.util import log, redis_util, rocketmq_util
from aisrfwk.util.exception_util import formatException


class BusinessDemo(BaseAisrTask):

    def run(self, data):
        start_time = time.time()

        task_id = None
        try:
            log.info("业务方法处理逻辑")

            # 过程中处理结果发送到服务端
            rocketmq_util.sendResultToServer()
        except BaseException as e:
            _msg = "任务失败,请检查底图和图层选择后再次重试!"
            _err_msg = "【严重】任务失败,排布过程中出现异常:"
            log.error(_err_msg + "{}", str(e))
            exc_text = _err_msg + "
" + formatException(sys.exc_info())
            redis_util.errorTaskStatus(task_id, exc_text)
            rocketmq_util.sendErrorToServer()
        # 处理结束后发送消息到后端服务
        time.sleep(1)
        rocketmq_util.sendTaskOverToServer()

        end_time = time.time()
        log.info("排布完成,总耗时:[{}]秒", round(end_time - start_time, 2))

5.主方法启动

主方法启动时需要启动队列,并传入业务处理类,然后再启动api

from aisrfwk.aisr_queue import consume_queue as queue

import rest_apis_demo
from business import BusinessDemo

if __name__ == '__main__':
    queue.start(BusinessDemo())
    rest_apis_demo.start()

6.公共API

push消息到队列

from aisrfwk.util import api

# 将参数对象push到消息队列中.
res = api.pushQueue({"projectId": 12,"test": "name"})
# 返回的对象会包含自动生成的id

推送消息到rocketmq

from aisrfwk.util import rocketmq_util

# 推送方案对象到rocketmq
rocketmq_util.sendResultToServer({返回对象})
# 推送错误消息到rocketmq
rocketmq_util.sendErrorToServer()
# 推送任务完结消息到rocketmq
rocketmq_util.sendTaskOverToServer()

详细使用可见上步业务逻辑处理类