在Perl oneliner中成功匹配的URL编码捕获值

Question

I am using the following Perl oneliner in a bash script to capture multiline text between pre tags and add a second iteration of the tag contents wrapped in a different starting and ending string:

new_start_string="NEWSTART"
new_end_string="NEWEND"

perl -i -pe  "BEGIN{undef $/;} s/<pre>(.*?)<\/pre>/<pre>\$1<\/pre>${new_start_string}\$1${new_end_string}/smg" /path/to/file

It works perfectly for the task as specified. I'm aware I'm parsing HTML with regex and that's not ideal, however the formatting of this text file is a known factor, very simple, under local control and done by local specification so it is a rare case in which parsing would be overkill and I've chosen not to do this via parsing.

Using the same oneliner, how can I URL encode the second iteration of the capture value $1 variable? If it's impossible, is there another approach to get this result that is also relatively simple and readable?

Answer 1

For URI escaping you can use URI::Encode . To use a function inside the substitution, see the /e option.

Answer 2

因为我更喜欢在Perl内核中使用模块，所以我最终使用了将e修饰符与CGI :: Util结合使用的oneliner：

perl -MCGI -i -pe  "BEGIN{undef $/;} s/<pre>(.*?)<\/pre>/'<pre>' . \$1 . '<\/pre>' . '$new_start_string' . CGI::escape(\$1) . '$new_end_string'/esmg" /my/path