Vulture in GSoC phase 2
Phase 2 is almost over. I think we implemented new features even faster than in phase 1. Rahul kept the pull requests coming and it was a pleasure discussing them with him. Overall, we implemented some interesting new features in Vulture: vulture can now be run with “python -m vulture”, making it easy to be run under different interpreter versions. We also now distinguish between unused classes and functions. The internal “Item” class no longer inherits from “str”, making future changes easier. An important change is that we now choose the shipped whitelists by looking at which modules are actually used in the analyzed code, raising the amount of code we can detect as unused. The biggest new feature added in this phase, however, is that Vulture can now sort unused code by size. This feature builds upon work by @jackric. I’m looking forward to what the final phase will bring!