def isAlreadyExistsWeb(companyName, currentTime, expired, shortUrl, subTaskId, url, isExistsWebId):
subTask = TaskInfo.getOne(TaskInfo.id == subTaskId)
taskResult = subTask.taskResultId
bigTask = taskResult.taskId
bigTaskId = bigTask.taskId
# 获取当前网站上一次的更新时间
websiteResult = Website.getOne(Website.webId == isExistsWebId)
webUpdateTime = websiteResult.updateDate
if webUpdateTime is None:
diffDay = expired + 1
else:
diffDay = (currentTime - webUpdateTime).days
if diffDay > expired:
print '时间已经过期'
# 网站信息过期,需要重新抓取并检测
fetchWebsite(companyName, bigTaskId, subTaskId, url)
else:
# 判断是否有当前网站信息是否存在亮照编号
if websiteResult.licID != '':
#logger.debug('已经亮照:', shortUrl)
q = TaskInfo.update(state='2').where(TaskInfo.id == subTaskId)
q.execute()
else:
#logger.debug('未亮照:', shortUrl)
q = TaskInfo.update(state='3').where(TaskInfo.id == subTaskId)
q.execute()
def judgeLzResult(com, web, shortUrl, subTask):
# 更新网站信息和公司信息
impCompanyInfo(com)
# impWebsite(web)
# nWeb = Website.getOne(Website.domain == web.domain)
# judgeWeb = Website.getOne(Website.domain ** str.format("%{}%", shortUrl))
# 如果查询网址与抓取亮照后的网址不匹配
if shortUrl != '':
shortUrl = shortUrl.replace('http://', '').replace(' ', '')
print 'shortUrl:', shortUrl
print "web.domain:", web.domain
com = Company.getOne(Company.coname == com.coname)
if shortUrl != web.domain:
# 当已经存在跳转关系记录时,不再操作
existsJumpWeb = Website.getOne((Website.domain == web.domain))
if existsJumpWeb is None:
dt = format(datetime.datetime.now(), '%Y-%m-%d %H:%M:%S')
web.updateDate = dt
web.regID = com
web.save(force_insert=True)
# 更新网站跳转地址
q = Website.update(jump=web).where(Website.webId == subTask.webId.webId)
q.execute()
else:
impWebsite(web)
# 更新网站跳转地址
q = Website.update(jump=existsJumpWeb).where(Website.webId == subTask.webId.webId)
q.execute()
else:
#更新网站信息
impWebsite(web)
count = Website.select().where(
(Website.licID != '') & (Website.regID == com) & (Website.domain == shortUrl)
).count()
if count == 0:
onlyCname = Website.select().where(
(Website.licID != '') & (Website.regID == com)).count()
onlyDomain = Website.select().where(
(Website.licID != '') & (Website.domain == shortUrl)).count()
if onlyCname > 0:
q = TaskInfo.update(state='9').where(TaskInfo.id == subTask.id)
q.execute()
elif onlyDomain > 0:
q = TaskInfo.update(state='8').where(TaskInfo.id == subTask.id)
q.execute()
else:
q = TaskInfo.update(state='4').where(TaskInfo.id == subTask.id)
q.execute()
else:
q = TaskInfo.update(state='2').where(TaskInfo.id == subTask.id)
q.execute()
def intervalDelayTask(taskResultId):
# 获取celery中当前已经正在进行的任务数
nowCount = TaskInfo.select().order_by(TaskInfo.id).where((TaskInfo.state == '6'))
if nowCount >= 0:
# 当celery中正在做的任务数量少于指定的数量时，向celery添加需要执行的任务
if nowCount <= config.celeryMaxCount:
taskCount = TaskInfo.select().where(
(TaskInfo.state == '1') & (TaskInfo.taskResultId == taskResultId)).count()
print 'taskCount:', taskCount
if taskCount == 0:
singalCheck(taskResultId)
# 查询该任务设置的延迟时间开启下一次需要检查的任务
interval = Task.getOne(Task.taskId == (
TaskResult.select(TaskResult.taskId).where(TaskResult.taskResultId == taskResultId))).intervalDay
print 'interval:', interval
if interval != "":
# 生成需要轮巡的新主任务结果记录
taskResult = TaskResult()
lastTaskResult = TaskResult.getOne(TaskResult.taskResultId == taskResultId)
taskResult.taskId = lastTaskResult.taskId
taskResult.state = '1'
taskResult.save()
#将上一次的任务结果编号所对应的webId指定给新的任务结果
psql_db.transaction()
try:
query = (TaskInfo
.insert_from(
fields=[TaskInfo.webId],
query=TaskInfo.select(TaskInfo.webId).where(TaskInfo.taskResultId == lastTaskResult)))
query.execute()
q = TaskInfo.update(taskResultId=taskResult).where(TaskInfo.taskResultId.is_null())
q.execute()
except Exception, e:
print e
psql_db.rollback()
# 获取当前时间
ctime = datetime.datetime.now()
delay_time = int(interval)
stime = ctime + datetime.timedelta(seconds=delay_time)
scheduler.add_job(intervalDelayTask, "date", next_run_time=stime, args=[taskResult.taskResultId],
jobstore="default", id=taskResult.taskResultId)
else:
tasks = TaskInfo.select().order_by(TaskInfo.id).paginate(0, config.sendCeleryCount).where(
(TaskInfo.taskResultId == taskResultId) & (TaskInfo.state == '1'))
for subTask in tasks:
subtaskId = subTask.id
fetchCycle.apply_async((subtaskId,), queue="celery")
# 更新taskinfo状态为已发送
q = TaskInfo.update(state='6').where(TaskInfo.id == subtaskId)
q.execute()
def checkLz(subtaskId):
try:
# 根据任务结果编号获取本次需要检查的任务记录
subTask = TaskInfo.getOne(TaskInfo.id == subtaskId)
checkTaskResult = subTask.taskResultId
bigTaskId = checkTaskResult.taskId.taskId
if subTask is not None:
companyName = subTask.webId.regID.coname
url = subTask.webId.domain
if (validateUrl(url)):
# 从任务表中获取是否已经进行检查过了(根据网站有没有更新时间进行判断)
isExistsTask = TaskInfo.getOne(
TaskInfo.webId == Website.getOne((Website.domain == url) & (Website.updateDate.is_null(False))))
subId = subTask.id
# 数据库中无此task对应的网站记录
if isExistsTask is None:
# 抓取检测
fetchWebsite(companyName, bigTaskId,subId, url)
else:
isExistsWebId = isExistsTask.webId.webId
# 如果当前网站更新时间小于过期时间,说明不用重新进行抓取并检查
expired = Configs.getOne(Configs.type == 'update').expired
# 获取当前时间
currentTime = datetime.datetime.now()
isAlreadyExistsWeb(companyName, currentTime, expired, url, subId,
url,
isExistsWebId)
except Exception:
q = TaskInfo.update(state='-1')
q.execute()
def noAccess(dirPath, filePath, mainTask, shortUrl, subTask):
dt = format(datetime.datetime.now(), '%Y-%m-%d %H:%M:%S')
qw = Website.update(updateDate=dt).where(Website.webId == subTask.webId.webId)
qw.execute()
q = TaskInfo.update(state='7').where(TaskInfo.id == subTask.id)
q.execute()
# 删除此首页文件
if os.path.exists(filePath):
os.remove(filePath)
# 删除本次任务目录
if os.path.exists(dirPath):
os.rmdir(dirPath)
def executeOnceTaskInfo(taskResultId):
# 获取celery中当前已经正在进行的任务数
nowCount = TaskInfo.select().where((TaskInfo.state == '6')).count()
# 当celery中正在做的任务数量少于指定的数量时，向celery添加需要执行的任务
if nowCount <= config.celeryMaxCount:
taskCount = TaskInfo.select().where(
(TaskInfo.state == '1') & (TaskInfo.taskResultId == taskResultId)).count()
print 'taskCount:', taskCount
if taskCount == 0:
singalCheck(taskResultId)
else:
tasks = TaskInfo.select().order_by(TaskInfo.id).paginate(0, config.sendCeleryCount).where(
(TaskInfo.taskResultId == taskResultId) & (TaskInfo.state == '1'))
for subTask in tasks:
subtaskId = subTask.id
checkLz.apply_async((subtaskId,), queue="celery")
# 更新taskinfo状态为已发送
q = TaskInfo.update(state='6').where(TaskInfo.id == subtaskId)
q.execute()
def fetchLzPage(isLzUrl, lzPath, shortUrl, subTask):
# 根据亮照的完整url进行抓取
status = downloadByPath(isLzUrl, lzPath)
if not os.path.exists(lzPath):
print "lzpath:",lzPath
#logger.debug('亮照页面无法访问:', isLzUrl)
#dt = format(datetime.datetime.now(), '%Y-%m-%d %H:%M:%S')
# qw = Website.update(updateDate=dt).where(Website.webId == subTask.webId.webId)
# qw.execute()
q = TaskInfo.update(state='5', remark=isLzUrl).where(TaskInfo.id == subTask.id)
q.execute()
else:
try:
f = open(lzPath, 'r')
parseData = parserCompanyAndWeb(f.read())
com = buildCompany(parseData['company'])
tempBuildWeb = buildWebsite(parseData['web'])
judgeLzResult(com, tempBuildWeb, shortUrl, subTask)
except Exception, e:
print e
def judgeNoLz(companyName, shortUrl, subTask):
dt = format(datetime.datetime.now(), '%Y-%m-%d %H:%M:%S')
qw = Website.update(updateDate=dt).where(Website.webId == subTask.webId.webId)
qw.execute()
q = TaskInfo.update(state='3').where(TaskInfo.id == subTask.id)
q.execute()
