Bash 使用awk格式化和替换时间戳列

我有以下格式的多个列

D,"4/2/2017 2:45:56 PM",ee,"4/2/2017 2:45:56 PM"
D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM"
D,"09/2/2017 6:05:54 AM",ee,"09/2/2017 6:05:54 AM"
D,"5/01/2017 8:29:46 PM",ee,"5/01/2017 8:29:46 PM"
D,"4/2/2017 02:3:26 AM",ee,"4/2/2017 02:3:26 AM"

我想把它们格式化如下

D,"04/02/2017 02:45:56 PM",ee,"04/02/2017 02:45:56 PM"
D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM"
D,"09/02/2017 06:05:54 AM",ee,"09/02/2017 06:05:54 AM"
D,"05/01/2017 08:29:46 PM",ee,"05/01/2017 08:29:46 PM"
D,"04/02/2017 02:03:26 AM",ee,"04/02/2017 02:03:26 AM"

D,"04/02/2017 02:45:56 PM",ee,"04/02/2017 02:45:56 PM"
D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM"
D,"09/02/2017 06:05:54 AM",ee,"09/02/2017 06:05:54 AM"
D,"05/01/2017 08:29:46 PM",ee,"05/01/2017 08:29:46 PM"
D,"04/02/2017 02:03:26 AM",ee,"04/02/2017 02:03:26 AM"

我尝试使用awk-F[，/：]分离列，然后根据长度进行处理

但是当有多个列时，它就变得单调乏味了

请建议awk中是否有任何日期时间或时间戳格式选项，以便我可以快速处理列式格式

我建议使用awk及其printf格式化输出：

awk -F '["/ :]' '{printf "%s\"%.2d/%.2d/%d %.2d:%.2d:%.2d %s\"%s\"%.2d/%.2d/%d %.2d:%.2d:%.2d %s\"\n",$1,$2,$3,$4,$5,$6,$7,$8,$9,$10,$11,$12,$13,$14,$15,$16}' file

输出：

D,"04/02/2017 02:45:56 PM",ee,"04/02/2017 02:45:56 PM" D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM" D,"09/02/2017 06:05:54 AM",ee,"09/02/2017 06:05:54 AM" D,"05/01/2017 08:29:46 PM",ee,"05/01/2017 08:29:46 PM" D,"04/02/2017 02:03:26 AM",ee,"04/02/2017 02:03:26 AM" 使用GNU awk与seps拆分。代码：

输出：

D,"04/02/2017 02:45:56 PM",ee,"04/02/2017 02:45:56 PM" D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM" D,"09/02/2017 06:05:54 AM",ee,"09/02/2017 06:05:54 AM" D,"05/01/2017 08:29:46 PM",ee,"05/01/2017 08:29:46 PM" D,"04/02/2017 02:03:26 AM",ee,"04/02/2017 02:03:26 AM" 您也可以使用sed，将字边界到pad之间的所有单个数字替换为0。但它会改变数据中的任何一个数字，即使它不在日期列中。因此，仅当您希望替换所有单次出现的附加0的数字时才使用它

如果要使更改永久化，请将-i与sed一起使用

它是如何工作的

正则表达式\b\[：digit:][\\b将匹配单词边界之间的单个数字，并用大括号捕获。现在，在sed的“替换”部分中，用第一个匹配的模式对0进行硬编码\1将得到0，并用单个数字填充

正则表达式演示

要了解此正则表达式的工作原理，请参见

工作示例：

---

我更喜欢这种方法，因为它也适用于mawk——基于Debian的发行版中的默认awk。我曾尝试使用gensub，但放弃了，因为它在mawk中不可用。@EdMorton我的版本是Ubuntu16.04附带的1.3.3-17。这与最近的Ubuntu17.04和即将发布的Debian9是同一个版本。在他们的网页中，他们声明：正如前面提到的，mawk被一些包装商忽略了。我真的不明白为什么Debian选择mawk作为默认awk，甚至不想使用最新版本。我可以理解/bin/sh的选择不是/bin/bash，但不是这个。出现错误。。4是Split的无效参数，是GNU awk的，先生。我相信你在评论中提到了mawk，两者都好。

D,"04/02/2017 02:45:56 PM",ee,"04/02/2017 02:45:56 PM"
D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM"
D,"09/02/2017 06:05:54 AM",ee,"09/02/2017 06:05:54 AM"
D,"05/01/2017 08:29:46 PM",ee,"05/01/2017 08:29:46 PM"
D,"04/02/2017 02:03:26 AM",ee,"04/02/2017 02:03:26 AM"

$ cat tst.awk
function fmt(t,    f) {
    split(t,f,/["\/ :]/)
    return sprintf("\"%02d/%02d/%04d %02d:%02d:%02d %s\"",f[2],f[3],f[4],f[5],f[6],f[7],f[8])
}
BEGIN { FS=OFS="," }
{ $2=fmt($2); $4=fmt($4); print }

$ awk -f tst.awk file
D,"04/02/2017 02:45:56 PM",ee,"04/02/2017 02:45:56 PM"
D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM"
D,"09/02/2017 06:05:54 AM",ee,"09/02/2017 06:05:54 AM"
D,"05/01/2017 08:29:46 PM",ee,"05/01/2017 08:29:46 PM"
D,"04/02/2017 02:03:26 AM",ee,"04/02/2017 02:03:26 AM"

sed 's|\b\([[:digit:]]\)\b|0\1|g'

bash-4.2$ cat file1
D,"4/2/2017 2:45:56 PM",ee,"4/2/2017 2:45:56 PM"
D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM"
D,"09/2/2017 6:05:54 AM",ee,"09/2/2017 6:05:54 AM"
D,"5/01/2017 8:29:46 PM",ee,"5/01/2017 8:29:46 PM"
D,"4/2/2017 02:3:26 AM",ee,"4/2/2017 02:3:26 AM"

bash-4.2$ sed -i 's|\b\([[:digit:]]\)\b|0\1|g' file1

bash-4.2$ cat file1
D,"04/02/2017 02:45:56 PM",ee,"04/02/2017 02:45:56 PM"
D,"03/02/2017 03:47:16 PM",ee,"03/02/2017 03:47:16 PM"
D,"09/02/2017 06:05:54 AM",ee,"09/02/2017 06:05:54 AM"
D,"05/01/2017 08:29:46 PM",ee,"05/01/2017 08:29:46 PM"
D,"04/02/2017 02:03:26 AM",ee,"04/02/2017 02:03:26 AM"