Regex to extract this semi formatted data

Question

Alright, I can't quite figure out how to do this.

Given the following text:

Roland AX-1:


/start
Roland's AX-1 strap-on remote MIDI controller has a very impressive 45-note velocity sensitive keyboard, and has switchable velocity curves, goes octave up/down, transpose, split/layering zones, and has fun tempo control for sequencers and more. Roland's AX-1 comes with a built-in GS control for total MIDI control of GM/GS synths. Its "Expression Bar" can control pitch and mod via an almost ribbon-like controller. It's also the newest and most advanced remote controller for your synths or midi modules.
/end

Roland AX-7:

/start
Roland's AX-7 builds on the infamous Roland AX-1 design. You just strap it on and put it to the front of the stage. Offering several controllers, such as: a D-Beam, then you can open the door to amazing live performance. 7-segment LED display, larger patch memory (Around 128 patches with MIDI data backup), and comes with GM2/GS compatibility make it extra easy to use. The 45-note, velocity-sensitive keyboard. 5 realtime controllers including a data entry knob, touch controller knob, opression bar, a hold button, and D-Beam. 128 patches with MIDI data backup. 2 MIDI zones.
/end

I'm trying to use the following:

/^([\w\d \-]*):\s\s\s\s^\/start([^\:]*)\/end$/im

You can see on rubular here:

http://rubular.com/r/BVRRHsnWdp

Thanks for any help. I guess i'm trying to match blocks of text until I hit the next title which always ends with a :$

Answer 1

The problem with your solution is [^\:]* doesn't allow any colons in the paragraph, but there are some. Try this, using a non-greedy match for the paragraphs:

/^([\w\d \-]*):\s+^\/start(.*?)\/end$/im

Answer 2

Take a look at Ruby's "flip-flop" operator. It is very useful for this sort of problem.

http://stackoverflow.com/questions/1111286/when-would-a-ruby-flip-flop-be-useful