Jenkins kill重复任务

脚本的功能是查看当前所有task，如果发现同一个url触发的task，就会abort之前的task，只保留最新的。
流程:

1. 在job_list填写需要kill的job路径,如果是多层,可以只写入顶层,脚本会递归查询,如填写devops,他会
   把子目录查询出来
2. 通过jenkins的api查询build的来源,是同一个来源。保存最新的时间。当后面的job时间比他早,干掉后面的job
   添加crontab定时任务：*/5 * * * * python3 stop_build.py

import json
import os
import re
import time
import logging

import jenkins
import requests
from lxml import etree
from configparser import ConfigParser

logger = logging.getLogger()

logging.basicConfig(
    format='%(levelname)s:%(asctime)s : %(lineno)d :%(message)s',
    datefmt='%Y/%m/%d %H:%M:%S',
    filename='earth_job_kill.log',
    level=logging.WARNING)


class JenkinsStopTask(object):

    def __init__(self, user, passwd, job_list=[]):
        self.host = "https://ci-earth.hobot.cc:8443"
        self.user = user
        self.password = passwd
        self.server = None
        self.init_server()
        self.job_list = job_list

    def init_server(self):

        self.server = jenkins.Jenkins(self.host,
                                      username=self.user,
                                      password=self.password)

    def get_running_build(self, job):
        tmp_url = ''
        if isinstance(job, str):
            job = job.split('/')
        flag = False
        for jj in job:
            if jj == 'view':
                tmp_url += f'/{jj}'
                flag = True
            elif flag:
                tmp_url += f'/{jj}'
                flag = False
            else:
                tmp_url += f'/job/{jj}'
    
        url = f'{self.host}{tmp_url}'
        api_url = f'{self.host}{tmp_url}/api/json'
    
        response = self.server.jenkins_open(requests.Request('GET', api_url))
        response_json = json.loads(response)
    
        if response_json['_class'] == 'org.jenkinsci.plugins.workflow.job.WorkflowJob':
            response = self.server.jenkins_open(requests.Request('GET', url))
            element = etree.HTML(response)
            self.parse_job(element, response_json['fullName'])
        else:
            # If the job is not a WorkflowJob, it may contain nested jobs
            for sub_job in response_json['jobs']:
                if sub_job['_class'] == 'org.jenkinsci.plugins.workflow.job.WorkflowJob':
                    sub_job_response = self.server.jenkins_open(requests.Request('GET', sub_job['15dd']))
                    sub_job_element = etree.HTML(sub_job_response)
                    sub_full_name = response_json['fullName'] + '/' + sub_job['name']  # Use 'name' instead of splitting URL
                    #print(sub_full_name)
                    self.parse_job(sub_job_element, sub_full_name)
                    # Recursively call get_running_build for nested jobs
                    #self.get_running_build(sub_full_name)  # Pass sub job name for recursion
                else:
                    sub_full_name = response_json['fullName'] + '/' + sub_job['name']  # Use 'name' instead of splitting URL
                    self.get_running_build(sub_full_name)

    def parse_job(self, element, job):

        task_list = element.xpath('//*[@id="buildHistory"]/div[2]/table/tr/td')
        running_map = dict()
        for task in task_list:
            res = task.xpath('div[1]/div/a/span/svg/@tooltip')
            if res and 'In progress' in res[0]:
                try:
                    task_name = None
                    start_time = None
                    # merge requests
                    if task.xpath('div[4]/a/text()'):
                        merge_text = task.xpath('div[4]/a/text()')[0]
                        if merge_text and 'GitLab Merge Request' in merge_text:
                            task_name_array = task.xpath('./div[4]/a/@href')
                            if task_name_array:
                                task_name = task_name_array[0]
                    # ph 重复任务
                    if task.xpath('div[3]/div[1]/a/span/text()'):
                        ph_text = task.xpath('div[3]/div[1]/a/span/text()')[0]
                        task_name_array = re.search(r'D(\d+)',
                                                    ph_text)
                        if task_name_array:
                            task_name = task_name_array.group(1)
                    # 不是这两项，不处理
                    if not task_name:
                        continue
                    index = None
                    start_time_array = task.xpath('./div[2]/@time')
                    if start_time_array:
                        start_time = start_time_array[0]
                    index_array = task.xpath('./div[1]/a/text()')
                    if index_array:
                        index_str = index_array[0]
                        index = re.search(r'(\d+)', index_str).group(1)
                    if (task_name and start_time and index) is None:
                        continue
                    if running_map.get(task_name,
                                       None) is None:
                        running_map[task_name] = {
                            'start_time': start_time,
                            'index': index
                        }
                    elif running_map[task_name]['start_time'] > start_time:
                        self.stop_task(job, index)
                    else:
                        last_task = running_map[task_name]
                        running_map[task_name] = {
                            'start_time': start_time,
                            'index': index
                        }
                        self.stop_task(job, last_task[index])

                except Exception as err:
                    logger.error(err)
        #logging.warning(str(running_map))

    def stop_task(self, job, index):
        if isinstance(job, list):
            job = '/'.join(job)
        print(f"{job} --- {index}")
        logging.warning(f"{job} --- {index}")
        self.server.stop_build(job, index)

    def handle(self):
       # for _ in range(2):
            for job in self.job_list:
                self.get_running_build(job)
            #time.sleep(20)


if __name__ == '__main__':
    cp = ConfigParser()
    config_path = os.path.dirname(os.path.abspath(__file__))
    cfg = os.path.join(config_path, 'config.cfg')
    cp.read(cfg)
    username = cp.get('jenkins', 'username')
    password = cp.get('jenkins', 'password')
    job_list = [
        #'test/test',
    ]
    jen = JenkinsStopTask(username, password, job_list)
    jen.handle()

config.cfg格式
[jenkins]
username =
password =