I'm planning to eventually (fairly soon!) publish the source code for this blog, if for no other reason than that I should have some code to point to on github. The one mistake I've made in that regard, however, is that the git repository contains both the scripts to publish the fragments and the text of the fragments themselves. So I have to split those apart, and ideally prune the latter from the git history for the former. (It's not the end of the world if I don't do that pruning, however.)
Admittedly, it seems unlikely that anybody else will ever want to use that source code. And it's not all that representative of the way I program: in particular, there are a lot fewer unit tests than I normally write. I think the breakdown into classes and methods isn't so unrepresentative, at least.