The updated web scraping howto
Web scraping is a practice that is basically as old as the web. The desire to extract contents or to machine- generate things from what perhaps was primarily intended to be presented to a browser and to humans pops up all the time.
When I first created the first tool that would later turn into curl back in 1997, it was for the purpose of scraping. When I added more protocols beyond the initial HTTP support it too was to extend its abilities to “scrape” contents for me.
I’ve not (yet!) met Michael Schrenk in person, although I’ve communicated with him back and forth over the years and back in 2007 I got a copy of his book Webbots, Spiders and Screen Scrapers in its 1st edition. Already then I liked it to the extent that I posted this positive little review on the curl-and-php mailing list saying:
this book is a rare exception and previously unmatched to my knowledge in how it covers PHP/CURL. It explains to great details on how to write web clients using PHP/CURL, what pitfalls there are, how to make your code behave well and much more.
Fast-forward to the year 2011. I was contacted by Mike and his publisher at Nostarch, and I was asked to review the book with special regards to protocol facts and curl usage. I didn’t hesitate but gladly accepted as I liked the first edition already and I believe an updated version could be useful to people.
Now, in the early 2012 Mike’s efforts have turned out into a finished second edition of his book. With updated contents and a couple of new chapters, it is refreshed and extended. The web has changed since 2007 and so has this book! I hope that my contributions didn’t only annoy Mike but possibly I helped a little bit to make it even more accurate than the original version. If you find technical or factual errors in this edition, don’t feel shy to tell me (and Mike of course) about them!