My first experience with Regular Expressions was writing whats called a web scraper to fetch weather data from Weather.com. Back then, this would involve a ton of IF statements and a large array of data. Very tedious and time consuming, but it was very educational. Today, Ruby has a couple convenient libraries built for this sort of task. Even a way to mimic a browser so you can get to pages normally reserved for site members.
After some careful research I settled on Nokogiri because its the library of choice with Mechanize, which I use to grab data from web sites and scrape the data from them easily. Before you can really use this library on OSX Leopard, you have to jump through a few hoops.
Mechanize may give you an error message like so:
WARNING: Nokogiri was built against LibXML version 2.7.3, but has dynamically loaded 2.6.16 HI. You’re using libxml2 version 2.6.16 which is over 4 years old and has plenty of bugs. We suggest that for maximum HTML/XML parsing pleasure, you upgrade your version of libxml2 and re-install nokogiri. If you like using libxml2 version 2.6.16, but don’t like this warning, please define the constant I_KNOW_I_AM_USING_AN_OLD_AND_BUGGY_VERSION_OF_LIBXML2 before requring nokogiri.
This is easy to fix, so here’s how to get rolling in Leopard:
Install the latest libxml2 and libxslt
Make a source directory somewhere and do the following commands to grab the latest versions of libxml2 and libxslt:
Once done, unpack both like so:
Now change into the libxml directory and build it:
Now jump back to the libxslt directory and do the same:
If you want, go ahead and delete the sources to save space. Now you’re ready for the final step.
Now installing Nokogiri is super easy once we tell it where to link.
Once you run that you are good to go.