[FIXED] Why await inside returned tuple make code slower?

Issue

I have function:

def foo(page_url: str):
    async with ClientSession() as session:
        async with session.get(page_url) as response:
            if response.status == 200:
                page = await response.text()
                return (page_url, page)

And it works correct, but i want to make await inside tuple:

return(page_url, await response.text())

my programm immediately starts working much slower.

Why?

Solution

I can't reproduce a drastic slowdown.

• I have python -m http.server 8254 2>/dev/null running, so there is no network interference here, and no background terminal being noisy either.

import asyncio
import sys

import aiohttp


async def getter_1(sess: aiohttp.ClientSession, page_url: str):
    async with sess.get(page_url) as response:
        if response.status == 200:
            page = await response.text()
            return (page_url, page)


async def getter_2(sess: aiohttp.ClientSession, page_url: str):
    async with sess.get(page_url) as response:
        if response.status == 200:
            return (page_url, await response.text())


def getter_test(getter):
    async def test():
        async with aiohttp.ClientSession() as sess:
            for x in range(3000):
                res = await getter(sess, 'http://127.0.0.1:8254/')
                assert res

    asyncio.run(test())


if sys.argv[1] == '1':
    getter_test(getter_1)
else:
    getter_test(getter_2)

Running this with the hyperfine benchmark tool, I get a result

$ hyperfine 'python3 so73542988.py 1' 'python3 so73542988.py 2'
Benchmark 1: python3 so73542988.py 1
  Time (mean ± σ):      7.547 s ±  1.547 s    [User: 2.228 s, System: 0.298 s]
  Range (min … max):    6.417 s … 11.568 s    10 runs

Benchmark 2: python3 so73542988.py 2
  Time (mean ± σ):      7.863 s ±  0.524 s    [User: 2.217 s, System: 0.297 s]
  Range (min … max):    7.218 s …  8.831 s    10 runs

Summary
  'python3 so73542988.py 1' ran
    1.04 ± 0.22 times faster than 'python3 so73542988.py 2'

so nothing very drastic at all.

Answered By - AKX