2
\$\begingroup\$

I am pretty much a python newbie and was looking at this question on StackOverflow.

Whereas the OP was interested in downloading the .htm |.txt files, I was simply interested in using Beautiful Soup and requests to gather all the links, to those files, into one structure. Ideally, it would have been a single list but I have a list of lists.

The process is:

  1. Start with the landing page and grab all the links in the leftmost column which has the header CIK.

    CIK column sample with single link selected on right


  1. Navigate to each of those links and grab the document button links in the Format column:

    Format column sample with single link selected on right


  1. Navigate to each of those document button links and grab the document links in the Document Format Files section, if the file mask is .htm | .txt.

    Document Format Files sample with single link selected on right


  1. Print the list of links (which is a list of lists)

Other info:

I use CSS selectors throughout to target the links.

What particularly concerns me, given my limited Python experience, is the fact I have a list of lists rather than a single list, whether my re-use of variable names is confusing and potentially bug prone, and I guess whether what I am doing is particularly pythonic and efficient.

In languages I am more familiar with, I might attempt to create a class and provide the class with methods; which in effect are my current functions.


Set-up info:

The version of the jupyter notebook server is: 5.5.0

The server is running on this version of Python:

Python 3.6.5 |Anaconda, Inc.| (default, Mar 29 2018, 13:32:41) [MSC v.1900 64 bit (AMD64)]


Python code:

from bs4 import BeautifulSoup
import requests
from urllib.parse import urljoin

base = 'https://www.sec.gov'
start_urls = ["https://www.sec.gov/cgi-bin/browse-edgar?company=&match=&CIK=&filenum=&State=&Country=&SIC=2834&owner=exclude&Find=Find+Companies&action=getcompany"]

def makeSoup(url):
    res = requests.get(url)
    soup = BeautifulSoup(res.content, "lxml")
    return soup

def getLinks(pattern):
    links = soup.select(pattern)
    return links

def getFullLink(links):
    outputLinks = []
    for link in links:
        outputLinks.append(urljoin(base, link.get("href")))
    return outputLinks

soup = makeSoup(start_urls[0])
links = getLinks("#seriesDiv [href*='&CIK']")
firstLinks = getFullLink(links)

penultimateLinks = []

for link in firstLinks:
    soup = makeSoup(link)
    links = getLinks('[id=documentsbutton]')
    nextLinks = getFullLink(links)
    penultimateLinks.append(nextLinks)

finalLinks = []

for link in penultimateLinks:
    for nextLink in link:
        soup = makeSoup(nextLink)
        links = getLinks("[summary='Document Format Files'] td:nth-of-type(3) [href$='.htm'],td:nth-of-type(3) [href$='.txt']")
        nextLinks = getFullLink(links)
        finalLinks.append(nextLinks)

print(finalLinks)
\$\endgroup\$
1
\$\begingroup\$

Since you never use getLinks without a call to getFullLink afterwards, I would merge these two functions. I would also make it a generator:

def get_links(soup, pattern):
    for link in soup.select(pattern):
        yield urljoin(base, link.get("href"))

Then your main part can become this nested for loop:

if __name__ == "__main__":
    pattern1 = "#seriesDiv [href*='&CIK']"
    pattern2 = "[id=documentsbutton]"
    pattern3 = "[summary='Document Format Files'] td:nth-of-type(3) [href$='.htm'],td:nth-of-type(3) [href$='.txt']"

    final_links = []
    for first_link in get_links(make_soup(start_urls[0]), pattern1):
        for second_link in get_links(make_soup(first_link), pattern2):
            final_links.extend(get_links(make_soup(second_link), pattern3))
    print(final_links)

I also renamed your functions according to Python's official style-guide, PEP8 and added a if __name__ == "__main__": guard.

One way to make this a bit faster (this sounds like it could be a lot of requests) is to use requests.Session to re-use the connection to the server:

session = requests.Session()

def make_soup(url):
    res = session.get(url)
    res.raise_for_status()
    return BeautifulSoup(res.content, "lxml")

Here I also added a guard so that the program stops if any site does not exists (i.e. returns 404 or similar).

\$\endgroup\$
  • 1
    \$\begingroup\$ OMG.... That is sooooooo much better. Thank you for taking the time to review +. I will leave a little longer if you don't mind to see if there is any further feedback. \$\endgroup\$ – QHarr Nov 16 '18 at 7:53

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.