FrequencyDictionaries

This repo contains frequency dictionaries that are text files with one word per line.

It is composed of two folders:

• freq_dicts_dirty : in this folder there the dictionaries contains words that are not in a "standard" dictionary.
• freq_dicts_clean : in this folder the dictionaries has been cleaned (and completed) to contain only words that are in a "standard" dictionary.

freq_dicts_dirty

The files in this folder were obtained from LuminosoInsight/wordfreq project. The corresponding dictionaries were transformed to .txt files with one word per line (the more frequents come first) by keeping only the words longer than 2 characters.

The transformation was done using jakm/msgpack-cli tool to convert the .msgpack files to .json files, and then using sed and grep they are transformed to .txt files with one word per line.

freq_dicts_clean

The files in this folder were obtained from the files in the freq_dicts_dirty folder by removing all words that are not in the corresponding dictionary of titoBouzout/Dictionaries.

• every short_xx.txt file was named with the same name
• every long_xx.txt file was renamed to medium_xx.txt
• the files long_xx.txt are obtained from medium_xx.txt (or short_xx.txt) by completing them (at the end in alphabetical order) by all words that are in the "standard" dictionary but not present in the "frequency" dictionary.