How to use GetHashCode method of PuppeteerSharp.PageAccessibility.SerializedAXNode class

I would use WaitForSelectorAsync instead:

var frameElement = await page.WaitForSelectorAsync("#google_ads_frame5");
var frame = await frameElement.ContentFrameAsync();
var frameContent = await frame.GetContentAsync();

To use XPath based and not CSS selection use page.XPathAsync(xpath-expression) e.g.

var tds = page.XPathAsync(".//td[contains(., 'ABC-DEF-GHI')]");
await tds[0].ClickAsync();

Based on Line 43 of the LanchOptions.cs you can specify arguments to the launch process.

Now connecting this with the native JavaScript (using the --load-extension flag) the output should look similar to the following:

var browser = await Puppeteer.LaunchAsync(new LaunchOptions
{
    Args = new string[1] { "--load-extension=/Path/To/Extension/Folder" }
});

Note: that I put path to folder, not the .zip as you have in your sample.. if you want to load multiple extensions then just separate the paths using a comma.

Here's a full list of possible arguments that can be passed in to Chromium.

Based on Line 43 of the LanchOptions.cs you can specify arguments to the launch process.

Now connecting this with the native JavaScript (using the --load-extension flag) the output should look similar to the following:

var browser = await Puppeteer.LaunchAsync(new LaunchOptions
{
    Args = new string[1] { "--load-extension=/Path/To/Extension/Folder" }
});

Note: that I put path to folder, not the .zip as you have in your sample.. if you want to load multiple extensions then just separate the paths using a comma.

Here's a full list of possible arguments that can be passed in to Chromium.

The issue was using Action instead of Func<Page, Task>.

Action always returns a void. Since the function Action was calling was a async method but was not sending the method's output it causes the function that called Action to exit which released the unmanged resources. The Func<Page, Task> retains the output so it can be awaited so the main thread is blocked till completion preventing the unmanaged resources from being release early.

async static void Main()
{
   Func<FileStream,Task> func = (FileStream fs) => Foo(fs); //Input to function is FileStream and Output of Function is Task
   Action<FileStream> action = (FileStream fs) => Foo(fs); //Equivalent to Func<FileStream , void>
   using var fs = new FileStream();
   //action(fs); //This can fail if the main thread is finished before it
   await func(fs); //This will succeed since the main thread is blocked till completion

}
async static Task Foo(FileStream fs) => await fs.SomeLongAsyncMethod();

That's the right way, for example:

var six = await page.EvaluateFunctionAsync<int>("async () => await Promise.resolve(6)");

"TimeoutError: Timed out after 30000 ms while trying to connect to the browser! Only Chrome at revision r1056772 is guaranteed to work." might be an explanation, as of 12/2022. Tried on Chrome Portable and got this error, very similar to yours.

This is how you can use System.Net.WebSockets.Managed on Puppeteer-Sharp

var browserOptions = new LaunchOptions
{
    Headless = true,
    WebSocketFactory = (uri, socketOptions, cancellationToken)
        => System.Net.WebSockets.SystemClientWebSocket.ConnectAsync(uri, cancellationToken)
};
using (var browser = await Puppeteer.LaunchAsync(browserOptions))
using (var page = await browser.NewPageAsync())
{

}

There are few ways to implement what you need (considering a C# console application).

Maybe the easiest one is to use tools that interact with an instance of a browser, i.e. Selenium (used for unit tests). So:

1. Install Selenium.WebDriver nuget package
2. Install a browser where your application will run (let's suppose chrome)
3. Download the browser driver (chromedriver)

4. Write something like:

   IWebDriver driver = null;
   try
   {
       ChromeOptions options = new ChromeOptions();
       options.AddArguments("--incognito");
   
       driver = new ChromeDriver(options);
       driver.Manage().Timeouts().ImplicitWait = TimeSpan.FromSeconds(5);
       driver.Url = "https://www.wedj.com/dj-photo-video.nsf/firstdance.html";
   
       var musicTable = driver.FindElement(By.Id("musicTable"));
       // interact with driver to get data from the page.
    }
    finally
    {
       if (driver != null)
          driver.Dispose();
    }
   

Otherwise, you need to investigate a little bit more how the webpage works. As far as I can see, the page loads a javascript, https://www.wedj.com/dj-photo-video.nsf/musiclist.js, that it is responsible to load the list of music from server. This js script basically load data from following url: https://www.wedj.com/gbmusic.nsf/musicList?open&wedj=1&list=category_firstdance&count=100 (you can open it also in a browser). Excluding "(" and ")", the result is a json you can parse (maybe using newtonsoft.json package):

{
  "more": "yes",
  "title": "<h1>Most Requested Wedding First Dance Songs<\/h...",
  "event": "<table class='musicTable g6-table-all g6-small' id='musicTable' borde..."
}

The event property contains the data you need (you can use HtmlAgilityPack nuget package to parse it).

Pro Selenium:

1. easy to interact with
2. the behavior is the same of what you see by the browser

Cons Selenium:

1. you need chrome or another browser installed
2. the browser is running when you interact with it
3. the browser download the full page (images, html, js, css...)

Pro manual:

1. you load only what you need
2. no dependencies to external programs (i.e. browsers)

Cons manual:

1. you need to understand how html/js works
2. you need to manually parse the json/html

In this specific case, I prefer the second options.