Earlier I wrote about using Boost in Python to process the graphs in my research. Turns out that this is much harder than it looks. 

It’s always a good idea to reuse code and leverage work done by other programmers. No point in reinventing the wheel, right? Sometimes, though, it’s quicker and easier to just stop being lazy and write the code yourself. In this case, I found out that installing an old version of Boost along with a dead bgl-python wrapper is not a very portable solution. 

After struggling with this issue for almost a week, I decided to try implementing the functionality myself. After a couple of simple regular expression and about 30 minutes of my time, I had a working solution. Plus, this code is just as portable as the rest of the project. Wow, all of that Boost stuff was a waste of time.