For the last several weeks, I’ve put MealMind on the back burner. It’s probably going to remain there for the rest of October, and then it’ll pick back up in November. There are just too many things going on for me to focus on it at the moment.
That said, I’ve been trying to lay out a more specific plan of attack for MealMind when I have the time to get back to it. One of the big issues I’m struggling with is the fact that there are lots of different parts to MealMind, they are all interlinked, and there isn’t a clear place to start. I also haven’t broken it down into bite-sized pieces so I can pick up a small, well-defined task and knock it out.
I think I’m leaning toward working these things next:
- Define a logical recipe & meal schema
- Pick some sample recipes & meals and manually map them to the schema
- Build out the logic code to perform this mapping with a suite a test cases to confirm the mapping is successful
- Start tasking out the process to take this schema and scale it up/down, build a shopping list of ingredients from it, calculate nutrient information, etc
At the end of this process, I’d like to point the system at a specific recipe url and have it give me a schema back along with the nutrient calculations. If I can get this part out of the way, then I can make progress on several other parts of the system, but having a really quick and reliable way to get recipes into the system seems like a good place to start.