之前张耀老师的网页脚本由于51cto升级，课程列表页面改用javascript失效了
笔者发现视频课程页面右边的列表都为静态化后的视频课程地址，遂将老师初始版本的脚本进行了修改，在对视频课程页面使用时结果正常，遂将及修改后的脚本和部分注释发上来供大家研究
[root@m01 scripts]# cat html_to_table.sh
#!/bin/bash
# oldboy linux training
# 2016-11-13
# 基于老男孩linux21期学员张耀开发脚本
#
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
#}
#分段没了，原函数保留，视频页抓一遍就好
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换页之类的功能都没用到，还是有很多值得向前辈学习的地方
tips：cat语句特别是最后end不能有空格，前端缩进最好用tab键完成
源脚本 “zhangyao” 博客，请务必保留源出处http://life2death.blog.51cto.com/7550586/1657133
最后欢迎大家访问我的wp博客>https://www.drscrewdriver.com 岂止博客