如何将复杂的 sed 命令转换为 PowerShell 脚本 [英] How to convert complex sed commands to a PowerShell script

问题描述

我有一个 Windows 批处理文件，它使用 sed 来操作文件.操作包括简单的替换，如

I have a Windows batch file the uses sed to manipulate files. The manipulation includes simple substitutions like

sed -e s/PATTERN/OTHERS/g                 infile >outfile
sed -e "/^.*COPY_START/,/^.*COPY_END/ d"  infile >outfile

虽然第一个可以替换为

Get-Content infile | %{ $_ -replace 'PATTERN', 'OTHERS' } | Set-Content outfile

我没有找到如何选择由行标记 COPY_START 和 COPY_END 删除的一堆行.

I didn't found how to select a bunch of lines that shall be removed by the line markers COPY_START and COPY_END.

如何用 PowerShell 替换这些 sed 命令.

How can these sed commands be replaced with PowerShell.

关于杰夫的评论，我试图用这段代码捕捉多行问题:

Regarding Jeff's comment I tried to catch the multi-line problem with this code:

Get-Content infile -replace '(.*)COPY_START\n.*COPY_END(.*)','$1$2'

但是 \n 不满足搜索条件.为了隔离我使用的多行"标准

But the \n does not fulfill the search criteria. To isolate the "multi-line" criteria I used

Get-Content infile -match 'COPY_START\n.*COPY_END'

使用这个文件

before
do COPY_START
skip
until COPY_END
after

但匹配总是$False.我希望表达式 \n.* 应该与 COPY_START 之后的行结束以及直到 COPY_END 出现的所有行匹配.

But the match is always $False. I expected that the expression \n.* should match with the line end after COPY_START and all lines until COPY_END appears.

为什么与正则表达式不匹配，我该如何解决?

Why doesn't match the regular expression and how can I fix it?

推荐答案

sed -e "/^.*COPY_START/,/^.*COPY_END/d" infile >outfile 命令找到COPY_START 和 COPY_END 行之间的所有子字符串，并将它们从 infile 中删除，并将输出保存到 outfile.

The sed -e "/^.*COPY_START/,/^.*COPY_END/ d" infile >outfile command finds all substrings between lines having COPY_START and COPY_END and removes them from the infile and saves the output to outfile.

要使其在 PS 中工作，您需要使用 -Raw 将文件读入单个变量，并使用像

To make it work in PS, you need to read the file into a single variables with -Raw and use a regex like

(?m)^.*COPY_START(?s:.*?)COPY_END.*$\n?

请参阅正则表达式演示.

详情

• (?m) - 一个多行内联修饰符，^ 现在将匹配一行的开头
• ^ - 行开始
• .* - LF 以外的任意 0+ 个字符，尽可能多
• COPY_START - 一个子字符串
• (?s:.*?) - 内联修饰符组(其中 s 启用 RegexOptions.Singleline 选项，也称为DOTALL，当 . 匹配任何字符) 匹配任何 0+ 个字符时，尽可能少到第一个
• COPY_END.*$ - COPY_END 然后是尽可能多的除换行符以外的任何 0+ 个字符
• \n? - 和一个可选的换行符.

• (?m) - a MULTILINE inline modifier, ^ will match the start of a line now
• ^ - line start
• .* - any 0+ chars other than LF, as many as possible
• COPY_START - a substring
• (?s:.*?) - an inline modifier group (where s enables the RegexOptions.Singleline option, also known as DOTALL, when . matches any char) that matches any 0+ chars, as few as possible up to the first
• COPY_END.*$ - COPY_END and then any 0+ chars other than a newline as many as possible
• \n? - and an optional newline.

这篇关于如何将复杂的 sed 命令转换为 PowerShell 脚本的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！