pronunciation.sh

pronunciation.sh

#!/bin/sh -e

# This script grabs the pronunciation (either uk or us) from
# <https://dictionary.cambridge.com>.
#
# Author: Ali Bozorgzadeh <https://github.com/reverseila>
#
# USAGE: p word [accent]
# accents: uk (default), us
# example: p aesthetic us
#
# Dependencies
# 1. mpv
# 2. wget
# 3. curl
#
# Note that It stores all the downloaded files in a cache directory.
# To customize the cache path, alter the "_cache_dir" variable.


# =========================================================================== #
#                                Variables (global)                           #
# =========================================================================== #

_word="$(echo $1 | tr A-Z a-z | tr ' ' '-')". # There is a problem with the letter 'p'. [:upper] [:lower:] converts 'p' to 'a'!
_base_url="https://dictionary.cambridge.org"
_cache_dir="${XDG_CACHE_HOME:-$HOME/.cache}/pronunciations"
_default_accent="uk"



# =========================================================================== #
#                                Functions                                    #
# =========================================================================== #

# print usage and exit
_usage() {
    printf "\033[33;1mUSAGE\033[0m: `basename -- $0` word [accent]\n"
    printf "\033[35;1maccent\033[0m: uk, us\n"
    exit 1
}


# get the url and output the two (space-separated) fields of urls
# first field is us accent and the second one is uk version
_get_url() {
    printf "$(curl --silent "$_base_url/us/dictionary/english/$1" | grep -o '/us/media/english/u[ks]_pron/.*\.mp3' | head -2)"
}


# play audio file via mpv
_play() {
    #@ $1: path to file
    mpv --no-config --force-window=no "$1" >/dev/null 2>&1
}



# =========================================================================== #
#                                   Main                                      #
# =========================================================================== #

# If word is not provided, exit
[ $# -lt 1 -a $# -gt 2 ] \
    && {
        printf "\033[31;1mERROR\033[0m: Please provide the word.\n\n"
        _usage
    }


# check for the correctness of accent
if [ $# -eq 2 ]; then
    case  "$2" in
        "us")
            _accent="us"
            _accent_number=1
            ;;
        "uk")
            _accent="uk"
            _accent_number=2
            ;;
        *)
            printf "\033[31;1mERROR\033[0m: The available accents are: uk and us.\n"
            exit 1
    esac
else
    # if no accent is provided as an argument
    _accent="$_default_accent"
    [ $_accent = "uk" ] && _accent_number=2 || _accent_number=1
fi


# name of the file in cache
_file="${_cache_dir}/${_accent}_${_word}.mp3"


# check if you already have it or not
if [ -e "$_file" ]; then
    _play "$_file" && exit 0
fi


# find the urls for us and uk pronunciations
_urls="$(_get_url "$_word")"

# if url cannot be fuond
[ -z "$_urls" ] \
    && printf "Sorry, the pronunciation of \033[0;1m%s\033[0m could not be found.\n" "$_word" \
    && exit 1


# if it doesn't exists, download and play it
# (don't download html files (--fail))
wget -c -q --show-progress -O "$_file" \
    "$_base_url$(echo $_urls | cut -d ' ' -f $_accent_number)"
_play "$_file"