之前張耀老師的網頁腳本由於51cto升級，課程列表頁面改用javascript失效了
筆者發現視頻課程頁面右邊的列表都爲靜態化後的視頻課程地址，遂將老師初始版本的腳本進行了修改，在對視頻課程頁面使用時結果正常，遂將及修改後的腳本和部分註釋發上來供大家研究
[root@m01 scripts]# cat html_to_table.sh
#!/bin/bash
# oldboy linux training
# 2016-11-13
# 基於老男孩linux21期學員張耀開發腳本
#
EduFile=/tmp/edu.html #process temp file1
EduFile2=/tmp/edu2.html #process 2
Url="$*"
# Check for given parameters
[ $# -eq 0 ] && {
echo "USAGE: /bin/sh $0 http://...."
exit 1
}
# Judge url is ok?
curl -I $Url &>/dev/null
[ $? -ne 0 ] &&{
echo "Bad url,Please check it"
exit 1
}
# Defined get pagenum and CourseId Functions
# Defined get pagenum and CourseId Functions
#function getnum(){
# curl -s $Url>$EduFile
# grep '"pagesGoEnd"' $EduFile &>/dev/null
# if [ $? -eq 0 ]
# then
# num=`sed -rn 's#.*page=([0-9].*)" class="pagesGoEnd".*$#\1#gp' $EduFile`
# else
# num=`sed -rn 's|.*page=([0-9].*)#" class="pagesNum".*$|\1|gp' $EduFile`
# fi
# pagenum=${num:-1}
# CourseId=`echo $Url|awk -F "[-.]" '{print $4}'`
#}
# Defined curl html Functions
#function Curl(){
# getnum
# for i in `seq $pagenum`
# do
# curl "http://edu.51cto.com/index.php?do=course&m=lessions&course_id=$CourseId&page=$i" 1>>$EduFile 2>/dev/null
# done
#}
#分段沒了，原函數保留，視頻頁抓一遍就好
function Curl(){
curl "$*" 1>>$EduFile 2>/dev/null
}
# Defined Create table Functions
function table(){
sum=""
index=1
sed -rn '/lesson/ s#<.*(<a href=")(.*)#\1http://edu.51cto.com\2#gp' $EduFile > $EduFile2
while read line
do
sum=$sum"<tr><th width="40" scope="row">$index</th><td width="520">$line</td>"
((index++))
done <$EduFile2
}
# Defined Create html Functions
function html(){
cat >/tmp/oldboy.html<<-END
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>test</title>
</head>
<body>
<table width="560" border="1">
$sum
</table>
</body>
</html>
END
}
function main(){
Curl
table
html
}
main效果是相似的但是像換頁之類的功能都沒用到，還是有很多值得向前輩學習的地方
tips：cat語句特別是最後end不能有空格，前端縮進最好用tab鍵完成
源腳本 “zhangyao” 博客，請務必保留源出處http://life2death.blog.51cto.com/7550586/1657133
最後歡迎大家訪問我的wp博客>https://www.drscrewdriver.com 豈止博客