shell操作文件的几条命令：删除最后一列、删除第一行、diff等-优快云博客

本文介绍了一段用于处理新闻数据的Shell脚本，该脚本可以自动化地从Hadoop文件系统获取最新的新闻数据文件，并筛选出未处理的数据进行进一步处理。通过使用sed、awk、comm等命令，实现了去除文件的首行、提取文件路径的最后一部分等功能。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

删除文件第一行： sed '1d' filename

删除文件最后一列： awk '{print $NF}' filename

awk删除重复行的命令：awk '{if (!seen[$0]++) {print $0;}}' filename

比较文件的两种方法：

1）comm -3 --nocheck-order file1 file2

2) grep -v -f file1 file2 :输出file2中有file1中没有的行

当然还有diff file1 file2

贴一段昨天写的shell脚本~

#!/bin/bash
date_time=`date +'%H_%M_%S'`
yesterday=`date -d"-1 day" +'%Y_%m_%d'`
today=`date +'%Y_%m_%d'`
date_day_time=`date +'%Y_%m_%d_%H_%M_%S'`

mkdir /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/same_similiar_log/$today

# begin to get input files which haven't been deal with
today_input=/home/crawler/petabyte/crawllog/news_data/$today
yesterday_input=/home/crawler/petabyte/crawllog/news_data/$yesterday

/opt/hadoop/program/bin/hadoop fs -ls $yesterday_input/ > /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_get
/opt/hadoop/program/bin/hadoop fs -ls $today_input/ >> /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_get

sed '1d' /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_get > /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_get_without_first_line

awk '{print $NF}' /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_get_without_first_line > /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_input

#comm -3 --nocheck-order /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_input /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/input_done > /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/today_diff

grep -v -f /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/input_done /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_input > /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/today_diff

awk '{print $NF}' /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/today_diff > /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/today_new_input

mv /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/all_input /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/input_done


# begin to compute same_similary_news
inputfile1=""
while read line
do
  inputfile1=$inputfile1,${line}
done < /home/spamdetect/changxiaojia/workspace/finance/same_similar_news_mining/mid_files/input_done
echo $inputfile1