deva 0.9.9

data eval in future

---

The deva lib makes it easy to write streaming data process pipelines,event driven programing,and run async function.

An example of a streaming process and web view

# coding: utf-8
from deva.page import page, Streaming, render_template
from deva import *

# 系统日志监控
s = from_textfile('/var/log/system.log')
s1 = s.sliding_window(5).map(concat('<br>'), name='system.log日志监控')
s.start()


# 实时股票数据

s2 = timer(func=lambda: NB('sample')['df'].sample(
    5).to_html(), start=True, name='实时股票数据', interval=1)

# 系统命令执行
command_s = Stream.from_command()
s3 = command_s.sliding_window(5).map(concat('<br>'), name='系统持续命令')
command_s << 'ping baidu.com'


@page.route('/')
def get():
    streams = [s1, s2, s3]
    return render_template('./web/templates/streams.html', streams=streams)


Streaming().start()
Monitor().start()
Deva.run()

Features

License

Copyright spark, 2018-2020.

Install

pip install deva

or

pip3 install deva

Sample

<b>如果是在jupyter里执行带码，代码尾部不需要添加Deva.run() </b>

bus

<b>如果使用bus跨进程，需要安装redis 5.0</b>

from deva import *

# 每隔一秒写入秒数到bus中
timer(start=True) >> bus
# 打印来自bus到数据
bus >> log
Deva.run()

from deva import *

# bus中的证书进行乘2后打印日志
bus.filter(lambda x: isinstance(x, int)).map(lambda x: x*2) >> log
# bus中来的原始数据全部打印报警
bus >> warn

Deva.run()

Crawler

from deva import *

h = http()
h.map(lambda r: (r.url, r.html.search('<title>{}</title>')[0])) >> log
'http://www.518.is' >> h


s = Stream()
s.rate_limit(1).http(workers=20).map(lambda r: (
    r.url, r.html.search('<title>{}</title>')[0])) >> warn
'http://www.518.is' >> s

Deva.run()

timer

from deva import timer, log, Deva, warn

# 默认每秒执行一次，返回当前秒
timer(start=True) >> log

# 3秒返回一个yahoo，随后启动，结果报警warn
s = timer(func=lambda: 'yahoo', interval=3)
s.start()

s >> warn
# 可用stop方法停止一个定时器
# s.stop()
Deva.run()


# python3 每隔n秒执行.py
# [2020-03-14 10:31:16.847544] INFO: log: 16
# WARNING:root:yahoo
# [2020-03-14 10:31:17.849576] INFO: log: 17
# [2020-03-14 10:31:18.853488] INFO: log: 18
# WARNING:root:yahoo
# [2020-03-14 10:31:19.855116] INFO: log: 19
# [2020-03-14 10:31:20.859602] INFO: log: 20
# [2020-03-14 10:31:21.865973] INFO: log: 21
# WARNING:root:yahoo
# [2020-03-14 10:31:22.868624] INFO: log: 22

scheduler

from deva import *

s = Stream.scheduler()

# 5秒执行一次的任务，返回yahoo到s中

s.add_job(func=lambda: 'yahoo', seconds=5)
# 5秒执行一次的任务，发送yamaha到bus，且返回yamaha到s中

s.add_job(func=lambda: 'yamaha' >> bus, seconds=5)

# 返回open到s中，每天执行一次，启动时间9点25
s.add_job(name='open', func=lambda: 'open', days=1, start_date='2019-04-03 09:25:00')

# 发送关闭到bus，返回值close放到s中，每天执行一次，15点30开始执行


def foo():
    '关闭' >> bus
    return 'close'


s.add_job(name='close', func=foo,
          days=1, start_date='2019-04-03 15:30:00')

# 打印所有任务
s.get_jobs() | pmap(lambda x: x.next_run_time) | ls | print

# 放入s中的所有数据都打印日志
s >> log

bus.map(lambda x: x*2) >> warn

Deva.run()


# $ python3 time_scheduler/scheduler.py

# [datetime.datetime(2020, 3, 14, 18, 6, 17, 830399, tzinfo=<DstTzInfo 'Asia/Shanghai' CST+8:00:00 STD>), datetime.datetime(2020, 3, 14, 18, 6, 17, 830947, tzinfo=<DstTzInfo 'Asia/Shanghai' CST+8:00:00 STD>), datetime.datetime(2020, 3, 15, 9, 25, tzinfo=<DstTzInfo 'Asia/Shanghai' CST+8:00:00 STD>), datetime.datetime(2020, 3, 15, 15, 30, tzinfo=<DstTzInfo 'Asia/Shanghai' CST+8:00:00 STD>)]
# [2020-03-14 10:06:17.835725] INFO: log: yahoo
# [2020-03-14 10:06:17.839594] INFO: log: yamaha
# WARNING:root:yamahayamaha
# [2020-03-14 10:06:22.846482] INFO: log: yahoo
# [2020-03-14 10:06:22.851722] INFO: log: yamaha
# WARNING:root:yamahayamaha
# [2020-03-14 10:06:27.840823] INFO: log: yaho

workers

from deva import bus, log, when, Deva

# 开盘任务
@bus.route(lambda x: x == 'open')
def onopen(x):
    'open' >> log

# 收盘任务
@bus.route(lambda x: x == 'close')
def onclose(x):
    'close' >> log

# 另外一种写法

when('open', source=bus).then(lambda: print(f'开盘啦'))
Deva.run()