如何用Perl从Ask.com特定标签中精准提取域名URL

The issue with your current code is that it processes every <p> tag in the HTML, which includes unrelated content like "Videos" or "Change Settings". To fix this, we need to filter <p> tags to only those with the class PartialSearchResults-item-url, and optionally validate that the content is a valid domain URL.

Here's the corrected code:

use HTML::TokeParser;

my $p = HTML::TokeParser->new(\$rrs);

# Open output file once (more efficient than opening/closing in loop)
open(my $out_fh, '>>', 'askurls.txt') or die "Could not open file: $!";

while (my $tag_info = $p->get_tag("p")) {
    # Extract attributes from the <p> tag
    my %tag_attrs = %{$tag_info->[1]};
    
    # Check if the tag has the target class
    if (exists $tag_attrs{class} && $tag_attrs{class} eq 'PartialSearchResults-item-url') {
        # Get trimmed text inside the <p> tag
        my $url = $p->get_trimmed_text("/p");
        
        # Optional: Validate the content is a valid domain (avoids any stray text)
        if ($url =~ /^[a-z0-9.-]+\.[a-z]{2,}$/i) {
            print "$url\n";
            print $out_fh "$url\n";
        }
    }
}

close($out_fh);

Key Changes Explained: #

• Filter by Tag Class: We now check if the <p> tag has the exact class PartialSearchResults-item-url before extracting its text. This eliminates all unrelated <p> tags.
• Efficient File Handling: Instead of opening and closing the output file inside the loop (which is slow for large datasets), we open it once before processing and close it afterward.
• Optional URL Validation: The regex /^[a-z0-9.-]+\.[a-z]{2,}$/i ensures we only keep valid domain URLs (like maps.google.com) and skip any unexpected text that might slip through.
• Scalar Variable for URL: get_trimmed_text("/p") returns a single string, so using a scalar ($url) is cleaner than an array.

This code will now only extract the domain URLs from the specific <p> tags you're targeting, ignoring all other content.

内容的提问来源于stack exchange，提问作者Haroon Ahmad