A writeup of this project can be found on Medium.
This simple Python tool takes a list of references refs.txt
(from a journal manuscript, for instance) and scrapes the web for URLs using the RapidAPI Google Search tool. In order to use this script, you will need your own RapidAPI account and an API key for the Google Search tool (see the RapidAPI section below).
Running the script will produce a refs-links.csv
file that will contain 3 columns: ref_full
, title
, link
.
The values for link
will be one of three options:
- A URL for the reference
- The string
'no link found'
if the request was successful, but no link was found - The string
'request failed'
if there was a connection error during the request
Note: This tool was built using python 3.9.1
, but it should work with older versions as well (not 2.x
)
Clone the repository onto your machine, then navigate to that directory.
Create a virtual environment using Pipenv
If you have pip
already installed, you can install Pipenv using $ pip install --user pipenv
.
~/ref-scraper $ pipenv install && pipenv shell
See the refs-example.txt
file for formatting. Each reference should be on its own line, and should include the full title of the text that you'd like to find a link for.
In order to use the ref_scraper.py
script, you will need to create your own API key and edit the file to include it. Create an account with RapidAPI, and then navigate to the Google Search tool. Once there, you should see a button that says "Subscribe to Test". Click on the button, then choose the free option, which will provide 600 requests per month. Once you've done this, you should be redirected to the API page, and in the "Code Snippets" window on the right, there should be a field labelled "x-rapidapi-key"
. Copy the string value (SIGN-UP-FOR-KEY
in the image below) of the key and paste it into ref_scraper.py
for the RAPID_API_KEY
variable.
Note: By default, the script runs silently, and takes quite a bit of time to finish, depending on the size of your refs.txt
file. If you want to see some output, you can uncomment the print
statements in ref_scraper.py
in the get_request_data
function. This will print every title and link that is found while the script runs.
In your terminal with the virutal environment activated, run the script by calling $ python ref_scraper.py
.
When the script is finished, it will save a new file refs-links.csv
and will print the total time in seconds that it took to run.