The current implementation of dictionary is simple and can no doubt be optimized.
One optimization would be: when the raw file is loaded, process into a data structure, and save the serialized data structure to a file. Then load that file the next time. (that file can be excluded from scm, so that updates to the raw file will be picked up and applied)
I'm sure there's other optimizations. Let's worry about it later.
The current implementation of
dictionaryis simple and can no doubt be optimized.One optimization would be: when the raw file is loaded, process into a data structure, and save the serialized data structure to a file. Then load that file the next time. (that file can be excluded from scm, so that updates to the raw file will be picked up and applied)
I'm sure there's other optimizations. Let's worry about it later.