A script I made to index .txt files in a directory and sub directorys so I can use
grep to search them. I chose Go because 1. I wanted to learn it and 2. indexing multiple files could be parallelized and Go seemed like a good choice.
How it works
A main thread scans the first level directory and makes a channel for goroutines to dump
word: filename maps to, then the main thread collects the maps together and writes an index line by line. Words in the index are lowercased and sanitized of any non-alphanumeric start and end characters.
The index file can then be
greped for the word you’re looking for.
Stuff I’d like to do
- add support for different text document types (pdf, namely, probably docx, odt?)
- add more terminal options/configuration (thread count, recursion depth, whether the output is sorted)
- add a way to better search the end result and display it better to the user
- reduce memory consumption while indexing, it all dumps into a huge map right now
- use some naive map-reduce to create the index faster/more distributed