get_mp3.sh

#!/bin/bash
# 下载 txts 中记录的歌曲 id 对应的歌曲文件

get_clip() {
    echo $1 | while read ind tag id tag_folder; do 
        echo $ind", "$tag", "$id", "$tag_folder; 
        url="https://www.xiami.com/song/"$id
        date;
        AUDIO_PATH=${tag_folder}"/"${id}".%(ext)s"
        if [ ! -f ${tag_folder}"/"${id}".mp3" ]; then
            youtube-dl --quiet --extract-audio ${url} -o "${AUDIO_PATH}";
        fi
    done
}
export -f get_clip;

PARA_NUM="32"           # 爬取并行进程数
MUSIC_NUM="2000"        # 每个歌曲标签下载歌曲数
TYPE="YZ语种"
AUDIO_FOLDER="mp3s/"$TYPE;
TXT_FOLDER="txts/"$TYPE;
for TXT_NAME in `ls ${TXT_FOLDER}`; do
    TXT_PATH=${TXT_FOLDER}/${TXT_NAME};
    TAG_NAME=${TXT_NAME%.txt};
    TAG_FOLDER=${AUDIO_FOLDER}"/"${TAG_NAME};
    mkdir -p $TAG_FOLDER;
    cat -n ${TXT_PATH} | head -${MUSIC_NUM} | while read line; do echo ${line}" "${TAG_FOLDER}; done | parallel -j ${PARA_NUM} -k get_clip
done