download the bunch of nltk data instead of the nltk.downloader module

nltk_utils.sh

#!/bin/bash
# refer to https://www.nltk.org/data.html for more information;
# this gist belongs to the "manual installation" section.

function nltk_fetch {
    curl https://www.nltk.org/nltk_data/ | ggrep -Po 'url=[^ ]*' | awk -F '"' '/githubuser/ {print $2}' | awk -F 'packages/' '{print $2" "$0}' > copora-name-url.txt;
}

function nltk_download_item {
    mkdir -p nltk_data/$(dirname $1);
    echo download $1 from $2 ... ;
    wget $2 -O nltk_data/$1;
};

# download each url
function nltk_download_zip {
    while IFS=' ' read -r filename url
    do
       download $filename $url
    done <copora-name-url.txt
}

# unzip each files
function nltk_unpack_zips {
    while IFS=' ' read -r filename url
    do
        path=$(dirname $filename)
        unzip -d nltk_data/$path nltk_data/$filename
    done <copora-name-url.txt
}

function nltk_remove_zips {
    find nltk_data -name '*.zip' | xargs rm 
}