SCRIPT: parser comments from YouTube.com

Example of automation scripts (PHP, C#, Java Script sand Phyton) and tutorials.
Post Reply
Ivan
Posts: 1
Joined: Mon Nov 09, 2020 12:00 pm

SCRIPT: parser comments from YouTube.com

Post by Ivan » Mon Nov 09, 2020 1:06 pm

Program version: Human Emulator Studio 7.0.54.

Browser: Chromium.

Logics:
1. Follow the link to the video on Youtube;
2. After navigate to the video, new comments are loaded, until the limit is reached, and the script parses the basic data (author's nickname, comment date, comment text, number of likes);
3. All data is added to a separate file result.csv;

The script consists of:
The executable file parser_comments_youtube.php. Plug-in commands are written in the tools folder in the functions.php file.
The file result.scv records information with the results.
The log folder contains the logs of the script. There is a separate log file for each day.

Download script:
parser_comments_youtube_EN.rar
(3.98 KiB) Downloaded 741 times
Script settings:
//link to video
$url = "youtube.com/watch?v=moOxq_8l_34";
$comments_limit = 20;

Script code:
//follow the link to the video
$browser->navigate($url);
sleep(4);

//scroll through the scroll until the required number of comments is loaded
while(true)
{
// get the given objects
$obj = $element->get_all_by_tag("YTD-COMMENT-THREAD-RENDERER");
// count
$cur_num_obj = count($obj->elements);
echo "comments: $cur_num_obj <br>";
	if($cur_num_obj < $comments_limit)
	{
		$browser->set_vertical_scroll_pos($browser->get_page_height() + 100);
		sleep(3);
	}
	else
	{
		break;
	}
}
$result = "";
foreach($obj->elements as $item)
{
	//go down into nested tags until we find the one we need
	$it = $item->get_child_by_attribute("class", "style-scope ytd-comment-renderer", true, true);
	$text = $it->get_inner_text();
	$a = explode("\n", $text);
	$name = "Name:" . $a[0];
	$date = "Date:" . $a[1];
	//number of likes
	$cnt = count($a);
	$votes = "Number of likes:" . $a[$cnt-2];
	//collect comments
	$comment = "Comment:" . $a[2];
	for($j = 3; $j < $cnt-2; $j++)
	{
		if($a[$j] == "Read more")
		{
			break;
		}
		$comment .= "\n" . $a[$j];
	}
	$result = $result . $name . "\n" . $date . "\n" . $comment . "\n" . $votes . "\n" . "\n";
}
file_put_contents("result.csv", $result);

Example of a file with the result of the script:
Name:QueenYoko OfKei
Date:1 день назад
Comment:You know Trump was imagining the golf balls were Biden’s face.
Number of likes:2,5 тыс.

Name:Zach Coates
Date:10 часов назад
Comment:“How you get fired on your day off???”
Number of likes:2 тыс.

Name:Nite & Day
Date:11 часов назад
Comment:Ended how it started. On a Golf Course of course. 🤣😂
Number of likes:1,3 тыс.

Post Reply