Download https://musictheory.pugetsound.edu/mt21c for offline usage

download-musictheory-pugetsound-edu.sh

#!/bin/bash

# Function taken from https://gist.github.com/cdown/1163649
# Install "urlencode" command from the "gridsite-clients" package if it doesn't work
urlencode() {
    # Usage: urlencode <string>

    old_lc_collate=$LC_COLLATE
    LC_COLLATE=C

    local length="${#1}"
    for (( i = 0; i < length; i++ )); do
        local c="${1:$i:1}"
        case $c in
            [a-zA-Z0-9.~_-]) printf '%s' "$c" ;;
            *) printf '%%%02X' "'$c" ;;
        esac
    done

    LC_COLLATE=$old_lc_collate
}


USER_AGENT_STRING='Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:90.0) Gecko/20100101 Firefox/90.0'

# The folder in which wget is going to download the website
WEBSITE_LOCAL_FOLDER_NAME='musictheory.pugetsound.edu/mt21c'


# Recursively download the whole website and convert the links between pages into relative links, pointing to the local files
wget -rpk -l 5 -np -U "$USER_AGENT_STRING" https://musictheory.pugetsound.edu/mt21c/MusicTheory.html


# Extract all JS URLs which the page depends on, except MathJax lib which cannot run locally
js_assets="$(grep -oP '<script src="\K(.*?)(?=\")' "$WEBSITE_LOCAL_FOLDER_NAME/Accidentals.html" | grep -v 'MathJax\.js')"

# Extract all CSS URLs which the page depends on, except Google fonts
css_assets="$(grep -oP '<link href="\K(.*?)(?=\")' "$WEBSITE_LOCAL_FOLDER_NAME/Accidentals.html" | grep -v 'fonts\.google')"


# Download the JS assets using a subshell to prevent changing the directory in the main script
# Constructs download commands which look like this:
# wget -c "https://pretextbook.org/js/lib/jquery.min.js"
(
    mkdir -p "$WEBSITE_LOCAL_FOLDER_NAME/js" &&
    cd "$WEBSITE_LOCAL_FOLDER_NAME/js" &&
    echo "$js_assets" | xargs -I{} echo 'wget -c "{}"' | bash
)


# Download the CSS files using a subshell
# Constructs download commands which look like this:
# wget -c "https://pretextbook.org/css/0.31/pretext.css"
(
    mkdir -p "$WEBSITE_LOCAL_FOLDER_NAME/css" &&
    cd "$WEBSITE_LOCAL_FOLDER_NAME/css" &&
    echo "$css_assets" | xargs -I{} echo 'wget -c "{}"' | bash
)


echo 'Coverting the JS asset URLs into relative file paths...'

for page in "$WEBSITE_LOCAL_FOLDER_NAME"/*.html; do
    for js_url in $(echo -e "$js_assets"); do
        js_file_name="${js_url##*/}"
        js_file_name_encoded="$(urlencode "$js_file_name")"
        sed -i -e "s|$js_url|\./js/$js_file_name_encoded|g" "$page"
    done
done

echo 'Coverting the CSS asset URLs into relative file paths...'

for page in "$WEBSITE_LOCAL_FOLDER_NAME"/*.html; do
    for css_url in $(echo -e "$css_assets"); do
        css_file_name="${css_url##*/}"
        css_file_name_encoded="$(urlencode "$css_file_name")"
        sed -i -e "s|$css_url|\./css/$css_file_name_encoded|g" "$page"
    done
done


echo 'Done!'
echo 'You can now open the MusicTheory.html file to begin studying.'