Bash 具有特定格式的Shell排序

Bash 具有特定格式的Shell排序,bash,shell,sorting,awk,Bash,Shell,Sorting,Awk,我的数据库是以这种格式构建的,我需要按第6列对其进行排序 10027|Chen|Ning|female|1982-12-08|2010-02-22T17:59:59.221+0000|1.2.9.86|Firefox 10995116908|Chen|Wei|female|1985-08-02|2010-05-2420:52:26.582+0000|27.98.244.108|Firefox (注意第6列中的T) 到目前为止,我已经尝试通过sort-M和特别是sort-k6m-t“|”file

我的数据库是以这种格式构建的,我需要按第6列对其进行排序

10027|Chen|Ning|female|1982-12-08|2010-02-22T17:59:59.221+0000|1.2.9.86|Firefox
10995116908|Chen|Wei|female|1985-08-02|2010-05-2420:52:26.582+0000|27.98.244.108|Firefox
(注意第6列中的T)

到目前为止,我已经尝试通过
sort-M
和特别是
sort-k6m-t“|”file.dat“
sort-k6-M-t”|“
等方式进行排序

此文件的所需排序输出

933|Perera|Mahinda|male|1989-12-03|2010-03-17T13:32:10.447+0000|192.248.2.123|Firefox
1129|Lepland|Carmen|female|1984-02-18|2010-02-28T04:39:58.781+0000|81.25.252.111|Internet Explorer
4194|Do|Hα» ChΓ­|male|1988-10-14|2010-03-17T22:46:17.657+0000|103.10.89.118|Internet Explorer
8333|Wang|Chen|female|1980-02-02|2010-03-15T10:21:43.365+0000|1.4.16.148|Internet Explorer
8698|Liu|Chen|female|1982-05-29|2010-02-21T08:44:41.479+0000|14.103.81.196|Firefox
一定是这个

8698|Liu|Chen|female|1982-05-29|2010-02-21T08:44:41.479+0000|14.103.81.196|Firefox 
1129|Lepland|Carmen|female|1984-02-18|2010-02-28T04:39:58.781+0000|81.25.252.111|Internet Explorer
8333|Wang|Chen|female|1980-02-02|2010-03-15T10:21:43.365+0000|1.4.16.148|Internet Explorer
933|Perera|Mahinda|male|1989-12-03|2010-03-17T13:32:10.447+0000|192.248.2.123|Firefox
4194|Do|Hα» ChΓ­|male|1988-10-14|2010-03-17T22:46:17.657+0000|103.10.89.118|Internet Explorer

最终,我看不出这项任务有什么特别之处——只是简单的排序:

sort -k6,6 -t "|" file.dat
输出:

8698|Liu|Chen|female|1982-05-29|2010-02-21T08:44:41.479+0000|14.103.81.196|Firefox
1129|Lepland|Carmen|female|1984-02-18|2010-02-28T04:39:58.781+0000|81.25.252.111|Internet Explorer
8333|Wang|Chen|female|1980-02-02|2010-03-15T10:21:43.365+0000|1.4.16.148|Internet Explorer
933|Perera|Mahinda|male|1989-12-03|2010-03-17T13:32:10.447+0000|192.248.2.123|Firefox
4194|Do|Hα» ChΓ­|male|1988-10-14|2010-03-17T22:46:17.657+0000|103.10.89.118|Internet Explorer

添加了两行额外数据,使搜索示例更易于查看:

933|Perera|Mahinda|male|1989-12-03|2010-03-17T13:32:10.447+0000|192.248.2.123|Firefox
1129|Lepland|Carmen|female|1984-02-18|2010-02-28T04:39:58.781+0000|81.25.252.111|Internet Explorer
4194|Do|H? Ch?|male|1988-10-14|2010-03-17T22:46:17.657+0000|103.10.89.118|Internet Explorer
8333|Wang|Chen|female|1980-02-02|2010-03-15T10:21:43.365+0000|1.4.16.148|Internet Explorer
8698|Liu|Chen|female|1982-05-29|2010-02-21T08:44:41.479+0000|14.103.81.196|Firefox
4567|Kim|Lisa|female|1982-05-29|2009-02-21T08:44:41.479+0000|14.103.81.196|Firefox
1234|Axe|John|male|1982-05-29|2012-02-21T08:44:41.479+0000|14.103.81.196|Firefox
我将使用以下输入参数定义一个
bash
脚本[
search.sh
]

search.sh [--born_after <dateA>] [--born_before <dateB>] -f <dbfile>

`--born_after  <dateA>` : [optional] search for data records with field6 >= this search parameter; [format=YYYY-MM-DDTHH:MM:SS.sss+HHMM] [default=0000-00-00T00:00:00.000+0000]
`--born_before <dateB>` : [optional] search for data records with field6 <= this search parameter; [format=YYYY-MM-DDTHH:MM:SS.sss+HHMM] [default=9999-99-99T99:99:99.999+9999]
`-f <dbfile>`           : [required] data file to search

请将该示例输入的所需输出添加到问题中。演示如何将给定日期a指定给另一给定日期adateB@RomanPerekhrest可能日期A 2010-02-15T09:33:33.400+0000,因为我们也没有指定该日期B 2010-03-16T20:20:20.300+0000@Cyrus是的,我的错,我做了如果有日期
2010-01-01T09:33:33.400+0000
?根据您指定的日期记帐
2010-02-15T09:33:33.400+0000
谢谢,尽管我认为需要-M选项。
-M
用于月排序,而不是整个日期您是一个救命恩人,尽管它对小数据库非常有效,但当我处理大文件时,它并不像它应该的那样工作。我是否应该检查更复杂的东西,如编译器或shell实例/版本?您必须在注释“它不能正常工作”上展开;另外,如果您通常搜索(相对)少量的行,那么如果您先执行
awk
工作,然后将输出传输到
sort
(最终结果是您将花费更少的资源来排序较小的数据集),效率应该会更高一些。还有一些事情需要考虑。。。如果您总是按字段6排序此文件。。。考虑将其排序一次,并将结果保存到新文件,然后对新的/排序的文件运行搜索(即,消除重复排序原始文件的开销)。
$ cat search.sh
#!/bin/bash

# set default search dates, clear the dbfile variable:

dateA="0000-00-00T00:00:00.000+0000"
dateB="9999-99-99T99:99:99.999+9999"

unset dbfile

# simulate getopts so we can parse for long and short option names

while [ $# -gt 0 ]
do
        case $1 in
                --born-after)   dateA=$2                                   ; shift  ;;
                --born-before)  dateB=$2                                   ; shift  ;;
                -f)             dbfile=$2                                  ; shift  ;;
                *)              echo "Unexpected argument '$1'. Aborting." ; exit 1 ;;
        esac

        shift
done

# if we didn't get receive/parse a `-f <dbfile>` option then abort:

[[ "${dbfile}" = '' ]] && echo "Missing a dbfile. Aborting." && exit 1

# start by sorting dbfile using RomanPerekhrest's solution; then pipe results to
# an awk script to handle the 'search'

sort -k6,6 -t "|" ${dbfile} | awk -F"|" -v dateA="${dateA}" -v dateB="${dateB}" '$6>=dateA && $6<=dateB'
# no search dates provided (ie, use defaults; display entire file (sorted))
$ search.sh -f file.dat
4567|Kim|Lisa|female|1982-05-29|2009-02-21T08:44:41.479+0000|14.103.81.196|Firefox
8698|Liu|Chen|female|1982-05-29|2010-02-21T08:44:41.479+0000|14.103.81.196|Firefox
1129|Lepland|Carmen|female|1984-02-18|2010-02-28T04:39:58.781+0000|81.25.252.111|Internet Explorer
8333|Wang|Chen|female|1980-02-02|2010-03-15T10:21:43.365+0000|1.4.16.148|Internet Explorer
933|Perera|Mahinda|male|1989-12-03|2010-03-17T13:32:10.447+0000|192.248.2.123|Firefox
4194|Do|H? Ch?|male|1988-10-14|2010-03-17T22:46:17.657+0000|103.10.89.118|Internet Explorer
1234|Axe|John|male|1982-05-29|2012-02-21T08:44:41.479+0000|14.103.81.196|Firefox

# only print records (sorted) with field6 >= 2009-10-01
$ search.sh --born-after '2009-10-01T00:00:00.000+0000' -f file.dat
8698|Liu|Chen|female|1982-05-29|2010-02-21T08:44:41.479+0000|14.103.81.196|Firefox
1129|Lepland|Carmen|female|1984-02-18|2010-02-28T04:39:58.781+0000|81.25.252.111|Internet Explorer
8333|Wang|Chen|female|1980-02-02|2010-03-15T10:21:43.365+0000|1.4.16.148|Internet Explorer
933|Perera|Mahinda|male|1989-12-03|2010-03-17T13:32:10.447+0000|192.248.2.123|Firefox
4194|Do|H? Ch?|male|1988-10-14|2010-03-17T22:46:17.657+0000|103.10.89.118|Internet Explorer
1234|Axe|John|male|1982-05-29|2012-02-21T08:44:41.479+0000|14.103.81.196|Firefox

# only print records (sorted) with field6 between 2009-10-01 and 2011-05-05
$ search.sh --born-after '2009-10-01T00:00:00.000+0000' --born-before '2011-05-05T23:59:59.999+9999' -f file.dat
8698|Liu|Chen|female|1982-05-29|2010-02-21T08:44:41.479+0000|14.103.81.196|Firefox
1129|Lepland|Carmen|female|1984-02-18|2010-02-28T04:39:58.781+0000|81.25.252.111|Internet Explorer
8333|Wang|Chen|female|1980-02-02|2010-03-15T10:21:43.365+0000|1.4.16.148|Internet Explorer
933|Perera|Mahinda|male|1989-12-03|2010-03-17T13:32:10.447+0000|192.248.2.123|Firefox
4194|Do|H? Ch?|male|1988-10-14|2010-03-17T22:46:17.657+0000|103.10.89.118|Internet Explorer