The Lucene search library has something like this. Data is stored /retrieved via codecs. Most write binary data, but one codec actually writes out all data as text. It is, of course, too slow for production use, but it is great for figuring out what is going on under the covers.
This is a really neat idea! I wonder how close you could get the prose to actually explaining the codes purpose as opposed to just saying "this is a for loop"?
The purpose (intent) wouldn’t be explained, but it could be useful for tools which have oddball syntax and are only used sometimes (infrequent enough for you to forget).
