启动nutch的linux脚本

为了跟踪nutch1.0的如下错误
Generator: 0 records selected for fetching, exiting ..
写了一个nutch启动的linux脚本，功能等同于bin/nutch crawl，可以查看哪一步出问题。
有三个参数，<crawlData>, <urls>, <depth>。
第三个可以省略，默认是3层爬取。

#!/bin/bash
cd /data
Nutch/bin/nutch inject $1/crawldb $2
# echo "Command: nutch inject $1/crawldb $2"
if [ $3 ]; then
  cnt=$3
else
  cnt=3
fi

while [ $cnt -gt 0 ]; do
Nutch/bin/nutch generate $1/crawldb $1/segments
seg=` hadoop dfs -ls $1/segments/ | tail -1 | cut -c57-`
Nutch/bin/nutch fetch $seg
Nutch/bin/nutch updatedb $1/crawldb $seg
let "cnt-=1"
done

Nutch/bin/nutch invertlinks $1/linkdb -dir $1/segments
Nutch/bin/nutch index $1/indexes $1/crawldb $1/linkdb $1/segments/*
Nutch/bin/nutch dedup $1/indexes
Nutch/bin/nutch merge $1/index $1/indexes