This is my solution to a "vacancy test" task.

I'm not sure at all if I have correctly implemented the task, but here is my solution.

Goals of code:

  1. Parse rows of table from a URL and extract some data (make a list of jsons from this data).
  2. Parse rows of table from a URL and download PDF (link is placed inside a row).


import asyncio
import aiofiles
from aiohttp import ClientSession as aiohttp_ClientSession
from bs4 import BeautifulSoup
from pathlib import Path as pathlib_Path
from json import dumps as json_dumps
from loguru import logger
from sys import platform as sys_platform

# # # HARDCODED # # #
MIN_YEAR = 2018
MAX_YEAR = 2020

def is_key_in_dict(dict_key: str, array: list):
    for i, dict_ in enumerate(array):
        if dict_['form_number'] == dict_key:
            return i

async def load_page(url: str, session, ):
    async with session.get(url=url, ) as resp:
        if 200 <= resp.status <= 299:
            return await resp.read()
            logger.error(f'Can not access a page, returned status code: {resp.status_code}')

async def parse_page_forms_json(rows: list) -> list[str]:  # Every page contain set of forms
    page_forms = []
    for row in rows:
            form_number = row.findChild(['LeftCellSpacer', 'a']).string.split(' (', 1)[0]  # split to remove non eng ver
            current_year = int(row.find(name='td', attrs={'class': 'EndCellSpacer'}).string.strip())
            index = is_key_in_dict(dict_key=form_number, array=page_forms)
            if index is None:
                    'form_number': form_number,  # Title in reality
                    'form_title': row.find(name='td', attrs={'class': 'MiddleCellSpacer'}).string.strip(),
                    'min_year': current_year,
                    'max_year': current_year,
            else:  # If exists - modify form
                if page_forms[index]['min_year'] > current_year:
                    page_forms[index]['min_year'] = current_year
                elif page_forms[index]['max_year'] < current_year:
                    page_forms[index]['max_year'] = current_year
        except Exception as e:
            logger.error(f'Error: {e}')
    return [json_dumps(page_form) for page_form in page_forms]  # What to do whit this data?

async def save_page_form_pdf(rows, session):
    for row in rows:
            current_year = int(row.find(name='td', attrs={'class': 'EndCellSpacer'}).string.strip())
            form_number_elem = row.findChild(['LeftCellSpacer', 'a'])  # link and name
            form_number = form_number_elem.string.split(' (', 1)[0]  # split to remove non eng ver
            if MIN_YEAR <= current_year <= MAX_YEAR:
                resp = await load_page(url=form_number_elem.attrs['href'], session=session)
                # See https://docs.python.org/3/library/pathlib.html#pathlib.Path.mkdir
                pathlib_Path(form_number).mkdir(parents=True, exist_ok=True)  # exist_ok - skip FileExistsError
                filename = f"{form_number}/{form_number}_{current_year}.pdf"
                async with aiofiles.open(file=filename, mode='wb') as f:
                    await f.write(resp)
        except Exception as e:
            logger.error(f'Can not save file, error: {e}')

async def main():
    async with aiohttp_ClientSession() as session:
        tasks = []
        pagination = 0
        while 1:
            url = (f'https://apps.irs.gov/app/picklist/list/priorFormPublication.html?'
            page = await load_page(url=url, session=session)
            soup = BeautifulSoup(page, 'html.parser')
            table = soup.find(name='table', attrs={'class': 'picklist-dataTable'})  # Target
            rows = table.find_all(name='tr')[1:]  # [1:] - Just wrong HTML
            if rows:
                tasks.append(await parse_page_forms_json(rows=rows))  # Task 1
                tasks.append(await save_page_form_pdf(rows=rows, session=session))  # Task 2
                pagination += 200
            else:  # Stop pagination
        await asyncio.gather(*tasks)

if __name__ == '__main__':
    # See https://github.com/encode/httpx/issues/914#issuecomment-622586610 (exit code is 0, but error is exists)
    if sys_platform.startswith('win'):
    logger.add('errors.txt', level='ERROR', rotation="30 days", backtrace=True)

The logic of code:

  1. Asynchronously get a page with table rows (tax forms)
  2. Do some actions with forms (blocking and non-blocking)
  3. Repeat P.s. You can skip code of all functions except main, load_page and save_page_form_pdf (they are no matter).


  1. Have I picked a good architecture? Especially I'm curious about the self-implemented if item not in list python pattern because this pattern is good for a simple object inside a list, but I have a list of the dicts and checking by dict key. After some experiments, I didn't find a better realization than presented here.
  2. Is an asynchronous solution good enough? It's my first experience with asynchronous web scraping.
  3. Maybe I need to optimize the check whether the form already exists in the list of forms, and not iterate over the entire list every time, but apply some sorting algorithm or something else?
  4. Do I need to add await keyword in a tasks.append(await save_page_form_pdf(rows=rows, session=session)) line?
  • \$\begingroup\$ Unfortunately this question is off-topic because this site is for reviewing working code. Please take the tour and read up at our help center. When the code works then feel free to edit the post to include it for a review. \$\endgroup\$ Mar 15 at 13:56
  • \$\begingroup\$ @SᴀᴍOnᴇᴌᴀ, Okay, I will fix it in a few hours \$\endgroup\$ Mar 15 at 14:41
  • 1
    \$\begingroup\$ When it's working, please add a good description of what the program is supposed to do. Then the question will be ready for reopening. \$\endgroup\$ Mar 15 at 14:46
  • \$\begingroup\$ @TobySpeight The code is working now. \$\endgroup\$ Mar 17 at 12:21
  • 1
    \$\begingroup\$ You need a little update (still says "code is incorrect currently"). Also, what's the code supposed to do? Can you summarise what's on the other end of that link, so that the question is self-contained? Thanks. \$\endgroup\$ Mar 17 at 12:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.