This code consists of two main things:
- A Python script which fetches the day's newspaper content from the Open Platform and saves each article as a separate HTML file locally.
The Python script should be run at least once per day, with the Guardian/Observer's daily content rolling over roughly around midnight, UK time. The list of content continues to change, so it's worth having the script run every hour or so for a while (but not all day, or you'll hit daily API limits). Running the script subsequent times in a day will update/replace any existing files.
Only one day's worth of content is saved at a time, with older days being deleted (the API only allows for you to keep content for 24 hours). Each day's content is stored in a dated directory within the archive/ directory.
Install the required modules using pip by doing:
$ pip install -r requirements.txt
There are various known issues/enhancements/proposals listed on the project's Issues page.
The site should work fine in most desktop web browsers, and not too bad on iPhone/iPad/iPod Touch. I've never used it on an Android device.
Get an API key for the Guardian Open Platform.
Install the python modules.
daily-paper/public/directory readable from the web. You can move it elsewhere without breaking the Python script, although I guess that will confuse Git. Or you could symlink it to somewhere appropriate.
- Your API key.
- The path to the directory where you want the dated folders for saved HTML
files to be kept. This should be called
archiveand be within the project's
publicdirectory, at the same level as
- If the
verbosesetting is left to
1, you will see a little debugging output when the script runs, including which articles have incomplete data.
daily-paper/scripts/scraper.pyscript. This should leave you with a dated directory in the archive directory you specified in the config file. It should contain an HTML file for every article from today's paper.
public/in your web browser. You should be able to read today's paper. If you're not using an actual web server, you should be able to load the
public/index.htmlfile itself in your browser and still use the site.