asyncio.pdfasyncio in python simple introduction f

文件名称: asyncio.pdf

所属分类: Python

开发工具:

文件大小: 186kb

下载次数: 0

上传时间: 2019-07-03

提供者: da***

下载 (186kb)

不能下载？报告错误

详细说明：asyncio in python simple introduction for asyncio.Contents 1 Chapter 1: First steps with asyncio 1. 1 Why use asencio 1. 2 Getting Started 1. 3 Hello world 1. 4 Hello clock 1.5 Http client example 1.6 asyncio perfornance 34556 1.7 Learn asyncio if you come from Twisted 1. 8 Getting Help · 8 2 Chapter 2: Advanced topics 2.1 TCP echo client and server 2.2 Threads 2.3 Subprocess 2.4 Producer/consumer .12 2.5 Asyncio Debug Mode ...14 3 Chapter 3: Larger examples 15 3.1 Web Scraping 15 4 Indices and tables 29 4.1 Glossary 29 ee aso 6 Contributing 33 6.1 Asyncio documentation 33 6.2 Notes to writers 33 6.3 Ideas 6.4 How to install Sphinx 6.5 How to build the documentation 34 6.6 See also...,,,,,,,,.,,,,,,, 34 CHAPTER Chapter 1: First steps with asyncio 1.1 Why use asyncio? 1.1.1 Why asynchronous programming? asyncio is a library to write asynchronous applications. It is the most efficient way to implement a network server having to handle many concurrent users 1.1.2 But gevent and eventlet just work Or Why should i bother with all these extra annoying async and await keywords In short, asyncio adopted a radically different solution for race conditions Parallel computing using threads is hard because of race conditions. Gevent and eventlet have a similar issue using green(lightweight)threads Code written with asyncio is less error-prone: by just looking at the code, it is possible to identify which parts of the code are under our controls and where the event loop takes over the control fow and is able to run other tasks when our task is waiting for something. gevent and eventlet are designed to hide the asynchronous programming. For non-expert, and sometimes even for experts, it is really hard to guess where the event loop is allowed to suspend the task and run other tasks in background It is even worse. A modification in a third party library can change the behaviour of our code, introduce a new point where the task is suspended For an example, see the Ca(shlche Coherent) Money"section of the Unyielding article(by Glyph, February, 2014) Asyncio Documentation Documentation, Release 0.0 1.2 Getting Started 1.2.1 Python 3.5(or higher) only This documentation is written for Python 3.5 to avail of the new a sync and await keywords IfyouhavePython3.5installedyouonlyneedtoinstallaiohttp: pipinstall-uaiohttp If you don t have python 3. 5 installed yet, you have several options to install it All platforms with conda Download and install Miniconda for our platform Create a new Python 3.5 environment(named aio35, use a different if you like highlight: bash conda create -n aio35 python=3.5 Activate it. Linux and os x highlight:: bash S source activate aio35 Windows highlight:: bash S source activate aio35 Installaiohttp highlight:: bash S(aio35) pip install aiohtti Platform specific Windows: The easiest way to use Python 3.5 would be to use a package manager such as conda. See the installation instructions above Mac os X: Install Homebrew and then type brew install python Linux: Ubuntu 16.04+ and Arch linux ship with Python 3.5 included. If you don t have Python 3.5+on your computer, you can compile it or use Python 1.2.2 Create a virtual environment to run examples If you dont use conda(see above), create a virtual environment python -m venv venv Chapter 1. Chapter 1: First steps with asyncio Asyncio Documentation documentation Release 0.0 Note: Depending on your platform, the Python 3 interpreter could be invoked by python instead. This is the case for conda on Windows for example Installaiohttpinthevirtualenvironment /venv/bin/python-mpipinstall-uaiohttp 1.3 Hello World This is a series of examples showing the basics of how to write coroutines and schedule them in the asyncio event 1.3.1 Simple coroutine This example uses the asyncio. AbstractEventLoop run_until_ complete() method to schedule a sim- ple function that will wait one second, print hello and then finish Because it is launched with asyncio. AbstractEventLoop. run_until_complete(), the event loop itself will terminate once the coroutine is completed import asencio async def say (what, when) await asyncio. sleep(when) print(what) 1。op= asyrc1o.qet_even=_1。p() loop. run_until_complete(say(hello world, 1) loop. clcs( 1.3.2 Creating tasks This second example shows how you can schedule multiple coroutines in the event loop, and then run the event loop Notice that this example will print second_ hello before first _ hello, as the first task scheduled waits longer that the second one before printing Also note that this example will never terminate, as the loop is asked to run forever import asyncio async def say(what, when) await asyncio sleep(when) print(what) loop asencio. get_ even=_loop() loop. create task(say('first hello',2) loop. create task(say (' second hello, 1)) (continues on next page 1.3. Hello world 3 Asyncio Documentation documentation release 0.0 (continued from previous page) loop.run_forever( loop. close() 1.3.3 Stopping the loop This third example adds another task that will stop the event loop before all scheduled tasks could execute, which results in a warning import asyncio async def say (what, when): await asyncio sleep(when) print (what) async def stop_after(loop, when) await asyncio. sleep(when) loop. stop() loop= asyncio get even- loop( oop create_task(say( first hello,2)) loop. create-task(say(' second hello', 1)) loop. create task(say( third hello, 4) loop. crcatetask(stop _after(loop, 3)) f loop. close() Warning: highlight:: none Task was destroyed but it is pending! task: wait_ for=> 1. 4 Hello clock Example illustrating how to schedule two coroutines to run concurrently. They run for ten minutes, during which the first coroutine is scheduled to run every second, while the second is scheduled to run every minute The function asyncio gather is used to schedule both coroutines at once Import async1o async def print_ second( imprint seconds m while true f。 r 1 n range(50): print (ir's) await asyncio sleep(1) (continues on next page) Chapter 1. Chapter 1: First steps with asyncio Asyncio Documentation documentation Release 0.0 (continued from previous page) async def print_every minute(): for i in range(1r 10) await asyncio sleep(60) print(i, 'minuLe' loop= asyncio get_even=_loop( loop. run_untiIcomplete( asyncio. gather(prin-_ second(), p≌in= every minute() loop. close() 1.5 Http client example Http ClienT example import asyncio importaiohttp async def fetch page(session, url): withaiohttp.Timeou=(l0 async with session. get (url)as response: assert response status ==200 return await response. read() loop asyncio, get_even=_loop() withaichttp.Clientsession(loop-loop)assession conCerT n unTil lele fetch_page(session,'http://python.org')) print (content) oop.C⊥cse() Formoreinformationseeaiohttpdocumentation 1.6 asyncio performance Random notes about tuning asyncio for performance. Performance means two different terms which might be incom- patible Number of concurrent requests per second Request latency in seconds: min/average/max time to complete a request 1.6.1 Architecture: Worker processes Because of its GIL, CPython is basically only able to use 1 CPU. To increase the number of concurrent requests, one solution is to spawn multiple worker processes. See for example Unicorn API-Hour 1.5. Http client example Asyncio Documentation documentation release 0.0 1.6.2 Stream limits limit parameter of Stream Reader/open_connection set write buffer limits( low/high water mark on writing for transports iohttpusesset_writer_buffer_limits(0)forbackpressuresupportandimplementedtheirownbuffering see aio-libs/aiohttp#136 Some thoughts on asynchronous API design in a post-asyncawait world (November, 2016) by Nathaniel J Smith 1.6.3 TCP NODELAY Since Python 3.6, asyncio now sets the TCP_NODELAY option on newly created sockets: disable the Nagle algorithm for send coalescing. Disable segment buffering so data can be sent out to peer as quickly as possible, so this is typically used to improve network utilisation See Nagle's algorithm 1.6.4 TCP QUICKACK This option is not used by asyncio by default. The TCP_QUICKACK option can be used to send out acknow ledgements as early as possible than delayed under some protocol level exchanging, and it's not stable/permanent, subsequent TCP transactions(which may happen under the hood) can disregard this option depending on actual protocol level processing or any actual disagreements between user setting and stack behaviour 1.6.5 Tune the linux kernel Linux TCP sysctl /proc/sys/net/ipv4/tcp_mem /proc/sys/net/core/rmem_default and/proc/sys/net/core/rmem_max: The default and maximum amount for the receive socket memorv /proc/sys/net/core/wmem_de fault and/proc/sys/net/ core/wmem_max: The default and maximum amount for the send socket memor /proc/sys/net/core/optmem_max: The maximum amount of option memory buffers net. ipv4. tcp no metrics_save net core netdev_max_ backlog: Set maximum number of packets, queued on the INPUT side, when the interface receives packets faster than kernel can process them 1.7 Learn asyncio if you come from Twisted The twisted project is probably one of the oldest libraries that supports asynchronous programming in Python. It has been used by many programmers to develop a variety of applications. It supports many network protocols and can be used for many different types of network prograMming. In fact, asyncio was heavily inspired by twisted. The Chapter 1. Chapter 1: First steps with asyncio

(系统自动生成,下载前可以参看下载内容)