logo

multimedia-dl

Unnamed repository; edit this file 'description' to name the repository.

generic.pl (1274B)

    

  1. #!/usr/bin/env perl
  2. # Multimedia-DL: Youtube-DL inspired scraper
  3. # Copyright © 2021 Multimedia-DL Authors <https://hacktivis.me/git/multimedia-dl/>
  4. # SPDX-License-Identifier: AGPL-3-only
  5. use strict;
  6. use utf8;
  7. use HTML::TreeBuilder::XPath;
  9. use LWP::UserAgent;
  11. my $ua = LWP::UserAgent->new;
  14. my $webkit_ua = "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.0 Safari/605.1.15";
  16. $ua->agent($webkit_ua . "Multimedia-DL/1.0");
  18. if($#ARGV != 0) {
  19. 	print "usage: multimedia-dl <url>\n";
  20. 	exit 1;
  21. }
  23. my $req = HTTP::Request->new(GET => $ARGV[0]);
  25. my $res = $ua->request($req);
  27. sub scrap_html_response {
  28. 	my ($res) = @_;
  30. 	my $tree = HTML::TreeBuilder::XPath->new_from_content($res->content) or die "HTML parsing failed";
  32. 	my $title = $tree->findvalue('//title');
  33. 	if($title) {
  34. 		print "Title: ", $title, "\n";
  35. 	}
  37. 	foreach($tree->findvalues('//video/@src')) {
  38. 		print $_, "\n";
  39. 	}
  41. 	foreach($tree->findvalues('//video/source/@src')) {
  42. 		print $_, "\n";
  43. 	}
  45. 	# TODO: meta og:video
  46. }
  48. if($res->is_success) {
  49. 	my $content_type = $res->header("Content-Type");
  51. 	if(($content_type == "text/html") or ($content_type == "application/xhtml+xml")) {
  52. 		scrap_html_response($res);
  53. 	}
  54. } else {
  55. 	print "Got ", $res->status_line, " instead of 2xx\n";
  56. }