sed / grep-获取两个字符串之间的文本（html）

Question

I am trying to extract "pagename" from the following:

<a class="timetable work" href="http://www.test.com/pagename?tag=meta376">Test</a>

I tried to get it to work using "sed" but it only says invalid command code.

What line of code would you guys suggest to get the pagename? By the way: This is not a single line but there is more content on the same line - but that should not make a difference as it should just matter what is between the limiters, right?

Thanks in advance for helping me out!

Answer 1

As you commented, if you want to extract "<a class="timetable work" href="test.com/"; and "?tag=meta376">Test</a>" you can use the following regex:

<a class="timetable.*?<\/a>

Working demo

If you want to grab the content just surround the regex with capturing groups:

(<a class="timetable.*?<\/a>)

The match is:

MATCH 1
1.  [9-80]  `<a class="timetable work" href="test.com/"; and "?tag=meta376">Test</a>`

Answer 2

我认为这是您想要的：

sed 's_^.*<a [^<>]* href="https*://[^/]*/\([^"?]*\).*$_\1_'

Answer 3

Giving you exactly what you asked for using exactly the delimiters you told us to use:

$ sed -n 's|.*<a class="timetable work" href="http://www\.test\.com/\(.*\)?tag=meta376">Test</a>|\1|p' file
pagename

Answer 4

I know it may be tempting to handle this using a regular expression but here's an alternative.

You are trying to parse some HTML, so use an HTML parser. Here's an example in Perl:

use strict;
use warnings;
use feature qw(say);

use HTML::TokeParser::Simple;
use URI::URL;

my $filename = 'file.html'; 
my $parser = HTML::TokeParser::Simple->new($filename);

while (my $anchor = $parser->get_tag('a')) {
    next unless defined(my $class = $anchor->get_attr('class'));
    next unless $class =~ /\btimetable\b/ and $class =~ /\bwork\b/;    
    my $url = url $anchor->get_attr('href');    
    say substr($url->path, 1);
}

Parse the HTML using HTML::TokeParser::Simple . loop through the <a> tags, skipping any that don't have the correct classes defined. For the ones that do, use URI::URL to parse the url and extract the "path" component (which in your case, would be "/pagename"). As you didn't want the leading slash, I used substr to remove the first character.

Output:

pagename

I know it's much longer than a single regex but it's also a lot more robust and will continue to work even when the format of your HTML changes slightly in the future. HTML parsers exist for a reason :)

Answer 5

I would use awk for this:

awk -F"[/?]" '/timetable work/ {print $4}'file
pagename

It search for a line containing timetable work , then print fourth field using \\ or ? as separator.