This project intends to build a search engine for local files that have been lost somewhere within a directory.
The Seach engine works well for small directories with minimal subdirectories. However, because it is all written in python, it is extremely slow for larger directories. I am in the process of rewriting much of the python scripts into Go to take advantage of the speed of go routines.
----------------------------------------------------------------------------------<>
The way this project works is as follows:
1.) Use a sentence embedding model to encode contents of all files within a directory (and its subdirectories) into vector embeddings
2.) Gather a semantic description of the lost file wanting to be found
3.) Encode this semantic description into a vector embedding
4.) Use a simmilarity function to find the file within the directory that most resembles the search
----------------------------------------------------------------------------------<>