shell命令curl抓取网页内容-下载xls文件，


通过curl命令抓取网页内容，关键点如下：

1.curl要模拟浏览器，最好是有个代理，因为很多站点都会搞个反爬虫什么的。

2.我的需求很简单，就是在国家外汇管理局上下载各种货币对美元的汇率。

http://www.safe.gov.cn/wps/portal/sy/tjsj_dmzsl

3.主要命令：curl，grep，awk，xls2txt，msql(LOAD DATA )。

curl：提取url

xls2txt：shell下的xls操作命令


xls2txt-0.14.tar.gz

下载地址：http://wizard.ae.krakow.pl/~jb/xls2txt/


msql(LOAD DATA )：导入到mysql数据库


备注：没有解决的问题就是下载的时候，下载下来的文件名不是转码后的文件名，而是utf8的字符串。
废话少说，贴一个粗糙版本，从抓取到操作xls


[root@DevelopServer http_curl]# cat shellspider.sh 

THE_DATE=`date +"%Y-%m-%d %H:%M:%S"`
echo "[$THE_DATE]:Begin  $1 $2  ......"

base_url="http://www.safe.gov.cn"
list_url="http://www.safe.gov.cn/wps/portal/sy/tjsj_dmzsl"
curl -S  -e "$list_url" -w %{http_code}"\\n"   -A "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:27.0) Gecko/20100101 Firefox/27.0" -H "Host:www.safe.gov.cn"  -D header1.txt -b "JSESSIONID=x3nfVnDP0cGNnNf2d6GPWZ7NnGFVJCJ3pdPLl0pDjjMC31XB0YZ3\!504534437" -c servercookie1.txt   "$list_url"  |grep 'href="/wps/portal/' |grep  hbdm_store |awk -F '"' '{print $4}' > list.lst

echo "http_code=$http_code"
THE_DATE=`date +"%Y-%m-%d %H:%M:%S"`
echo "[$THE_DATE]:get list_url  ......"
sleep 3

index=`ls -lt *.xls|wc -l`
if [ 0 -ne ];then
   index=0
fi
 for i in `cat list.lst `
 do 
  # index=$((${index}+1))
  # xls_file="$index.xls"   

   tmp_url="$base_url$i"
   download_page_url=${tmp_url//\!/\\!}
   #echo "$download_page_url"
   THE_DATE=`date +"%Y-%m-%d %H:%M:%S"`
   echo "[$THE_DATE]:get download_page_url=$download_page_url  ......"

   tmp_url=`curl -S  -e "$list_url" -w %{http_code} -A "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:27.0) Gecko/20100101 Firefox/27.0" -H "Host:www.safe.gov.cn"  -D header2.txt -b "JSESSIONID=x3nfVnDP0cGNnNf2d6GPWZ7NnGFVJCJ3pdPLl0pDjjMC31XB0YZ3\!504534436" -c servercookie2.txt "$download_page_url"|grep urlArr|grep  '/wps/wcm/connect/' |grep '.xls' |awk -F "'" '{print $2}'`
   echo "http_code=$http_code"

   download_url=${tmp_url//\!/\\!}
   download_url="$base_url$download_url"
    #echo "$download_url"
   THE_DATE=`date +"%Y-%m-%d %H:%M:%S"`
   echo "[$THE_DATE]:get download_url=$download_url  ......"

    sleep 1
    curl -S  -e "$list_url" -w %{http_code}  -A "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:27.0) Gecko/20100101 Firefox/27.0" -H "Host:www.safe.gov.cn"  -D header3.txt -b "JSESSIONID=x3nfVnDP0cGNnNf2d6GPWZ7NnGFVJCJ3pdPLl0pDjjMC31XB0YZ3\!504534435" -c servercookie3.txt -O "$download_url" >ok.code
    cat ok.code

    echo "http_code=$http_code"
   THE_DATE=`date +"%Y-%m-%d %H:%M:%S"`
   echo "[$THE_DATE]:end download_url=$download_url  ......"

   sleep 3
 done

  for i in `ls *\%*`
  do
   index=$((${index}+1))
   xls_file="$index.xls"    
      ls -lt $i
      mv -f $i $xls_file
      xls2txt -n 0  $xls_file |head
  done

THE_DATE=`date +"%Y-%m-%d %H:%M:%S"`
echo "[$THE_DATE]:End  $1 $2  ......"

[root@DevelopServer http_curl]#


相关内容