feat: add Puppeteer integration for web downloads
This introduces a new Puppeteer-based mechanism for downloading web content. It provides a flexible way to manipulate pages during downloads, enhancing the ability to handle dynamic content and improve the overall download process.
This commit is contained in:
@@ -0,0 +1,37 @@
|
||||
|
||||
using HtmlAgilityPack;
|
||||
using Microsoft.Playwright;
|
||||
|
||||
namespace Beam.Puppeteer {
|
||||
public class PuppetUnitPageDownloader<T> : UnitDownloader<T> {
|
||||
public AsyncManipulator PuppetManipulator { get; }
|
||||
|
||||
public PuppetUnitPageDownloader(HtmlWeb web, AsyncManipulator puppetManipulator, AsyncTransformer<HtmlDocument, T> asyncHtmlTransformer, AsyncDownloadFailurePredicate<HtmlDocument>[] asyncDownloadFailurePredicates)
|
||||
: base(web, asyncHtmlTransformer, asyncDownloadFailurePredicates) {
|
||||
PuppetManipulator = puppetManipulator;
|
||||
}
|
||||
|
||||
protected override async Task<(bool, T?)> TryDownloadWithNoRetries(string link, CancellationToken ct) {
|
||||
var page = await PuppetContext.Browser.Value.NewPageAsync();
|
||||
try {
|
||||
await page.GotoAsync(link);
|
||||
await PuppetManipulator(page);
|
||||
var content = await page.ContentAsync();
|
||||
await page.CloseAsync();
|
||||
|
||||
HtmlDocument doc = new();
|
||||
doc.LoadHtml(content);
|
||||
var transformed = await Transformer(doc);
|
||||
if (FailurePredicates is null || !(await IsFailure(doc)))
|
||||
return (true, transformed);
|
||||
return (false, default);
|
||||
} catch (Exception) {
|
||||
return (false, default);
|
||||
} finally {
|
||||
if (!page.IsClosed)
|
||||
await page.CloseAsync();
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
}
|
||||
Reference in New Issue
Block a user