Currently, it’s not possible to sort GSoC participating organizations by the programming languages they use in their code. This results in students spending a lot of time going through each organization’s page and manually sorting through them.
This article introduces a way for students to write their own Python script using the BeautifulSoup4 library. Using this script the students can find the organization that uses the language they desire to contribute in.
You will learn the following through this article:
- How to use Requests library to send HTTPS requests to webpages
- How to use BeautifulSoup4 library in python to parse HTML code
- Output data in the form of a spreadsheet (eg. MS Excel) using OpenPyXL
Installation
The above module does not come pre-installed with Python. To install them type the below command in the terminal.
pip install requests pip install beautifulsoup4 pip install openpyxl
Note: Only beginner level knowledge of Python 3 is required for following this article. For more information, refer to Python Programming Language
Getting Started
Step 1: Import the required libraries
import requests, bs4, openpyxl |
Step 2: Create a response object using Requests. We will be using the Archive page as our source
# Replace "YEAR" by the year you # want to get data from. Eg. "2018" # Creating a response object # from the given url res = requests.get(url) # We'll be using the Archive page # of GSoC's website as our source. # Checking the url's status res.raise_for_status() |
Step 3: Create a BeautifulSoup object
From the Archive page’s source code:
< li class = "organization-card__container" layout flex-xs = "100" flex-sm = "50" flex = "33" aria-label = "AerospaceResearch.net" > ... ... < div class = "organization-card__footer md-padding" > < h4 class = "organization-card__name font-black-54" >AerospaceResearch.net</ h4 > </ div > ... ... </ li > |
We can see that the Orgs’s name is in a H4
tag with class name “organization-card__name font-black-54
” .
Using BS4, we can search for this particular tag in the HTML code and store the text in a list.
# Specify the language you # want to search for language = 'python' # BS4 object to store the # html text We use res.text # to get the html code in text format soup = bs4.BeautifulSoup(res.text, 'html.parser' ) # Selecting the specific tag # with class name orgElem = soup.select( 'h4[class ="organization-card__name font-black-54"]' ) # Similarly finding the links # for each org's gsoc page orgLink = soup.find_all( "a" , class_ = "organization-card__link" ) languageCheck = [ 'no' ] * len (orgElem) orgURL = [ 'none' ] * len (orgElem) |
Step 4: Opening each Orgs’s GSoC page and finding the languages used
item = 0 # Loop to go through each organisation for link in orgLink: # Gets the anchor tag's hyperlink presentLink = link.get( 'href' ) print (item) print (url2) orgURL[item] = url2 res2 = requests.get(url2) res2.raise_for_status() soup2 = bs4.BeautifulSoup(res2.text, 'html.parser' ) tech = soup2.find_all( "li" , class_ = "organization__tag organization__tag--technology" ) # Finding if the org uses # the specified language for name in tech: if language in name.getText(): languageCheck[item] = 'yes' item = item + 1 |
Step 5: Writing the list to a spreadsheet
Using the openpyxl
library, we first a create a workbook. In this workbook we open a sheet using wb[‘Sheet’], where we will actually write the data. Using the cell().value
function, we can directly write values to each cell. Finally we save the workbook using save()
function.
wb = openpyxl.Workbook() sheet = wb[ 'Sheet' ] for i in range ( 0 , len (orgElem)): sheet.cell(row = i + 1 , column = 1 ).value = orgElem[i].getText() sheet.cell(row = i + 1 , column = 2 ).value = languageCheck[i] sheet.cell(row = i + 1 , column = 3 ).value = orgURL[i] wb.save( 'gsocOrgsList.xlsx' ) |
Note: The spreadsheet will be stored in the same directory as the Python file
Troubleshooting
Due to repeated requests to the website, the server may block your IP address after repeated attempts. Using a VPN will solve this issue.
If the problem still persists, add the following to your code:
from fake_useragent import UserAgent ua = UserAgent() header = { "User-Agent" : ua.random } |