By slightly modifying one of the examples scripts from PEAR’s HTTP_Client package, you have yourself quite a useful link checker script.
Here’s the code:
To get things to work:
- install the required classes using the PEAR package manager (if you’re not using this yet slap yourself on the wrist now):
$ pear install HTTP_Request
$ pear install HTPP_Client
If you get an error about missing dependencies, obviously install the dependencies first.
- change the $url variable to one of your sites you want to check (line 81)
- load the script in a browser
The concept is simple:
- Use HTTP_Request to return the html of a given webpage
- use the extractlinks() function (slightly modified from Snoopy code) to get an array of links from the page
- use HTTP_Client and a link checker listener (Observer pattern) to iterate through all the links and return the appropriate response code
- colour code pass/fail responses and update the browser with flush()