Downloading PDF files using Puppeteer

[ags4436]

I am trying to download PDF files using puppeteer, when .pdf URLs are hit the chrome opens the pdf in the tab, instead I want the pdf to be downloaded. This can be achieved through changing the user-preferences of browser and Sites automatically follow this setting when you visit them by default it is Open PDFs in Chrome.

I found that we can use puppeteer-extra for downloading the PDF files. Is that a legit NPM which I could use in my project instead of puppeteer-core ?

Is there any alternatives available from Puppeteer itself ?

[OrKoN]

I believe it should be possible to use CDP's setDownloadBehaviour https://chromedevtools.github.io/devtools-protocol/tot/Browser/#method-setDownloadBehavior but there is no dedicated API in Puppeteer for this yet:

await  page._client.send('Page.setDownloadBehavior', {
	behavior: 'allow',
	downloadPath: tempFolderPath,
})

[ags4436]

This allows us to set the download path while downloading the files. PDFs will open in the tab itself in this case as well.

[OrKoN]

I see. It looks like puppeteer-extra writes those settings into the user profile directory and to me it looks like the approach might be a bit fragile if the structure of the directory changes. But I am not aware of any other way to achieve this. Most likely, it's possible to replicate what puppeteer-extra does and create the profile with the right setting but I think using puppeteer-extra is fine too.

[MrXyfir]

Really surprised puppeteer still doesn't support this. Playwright does.

[nros]

Because of PR 8506 (since release 14.4.0), it's not _client anymore but it is _client(). So use:

await page._client().send('Page.setDownloadBehavior', {
    behavior: 'allow',
    downloadPath: tempFolderPath,
})

See: Page.ts

[Korak-997]

I had to twick a lot till i found this solution 😄
but it works perfectly for me.

And Downloads the pdf automatically which means you do not need to do anything else

// WAITS TILL THE OBJECT IS RENDERED
await page.waitForSelector('SELECTOR FOR PDF OBJECT')
   //SET DOWNLOAD PATH
    await page._client.send("Page.setDownloadBehavior", {
      behavior: "allow",
      downloadPath: PATH,
    })
    
  
    await page.evaluate(() => {
      const pdfObj = document.querySelector("SELECTOR FOR PDF OBJECT")
      const btn = document.createElement("a")
      btn.setAttribute("href", pdfObj.data)
      btn.setAttribute("download","FILE_NAME")
      btn.click()
      })

[stale]

We're marking this issue as unconfirmed because it has not had recent activity and we weren't able to confirm it yet. It will be closed if no further activity occurs within the next 30 days.

[OrKoN]

We will definitely need a better API for this.

[Kle0s]

I had to twick a lot till i found this solution 😄 but it works perfectly for me.

And Downloads the pdf automatically which means you do not need to do anything else

// WAITS TILL THE OBJECT IS RENDERED
await page.waitForSelector('SELECTOR FOR PDF OBJECT')
   //SET DOWNLOAD PATH
    await page._client.send("Page.setDownloadBehavior", {
      behavior: "allow",
      downloadPath: PATH,
    })
    
  
    await page.evaluate(() => {
      const pdfObj = document.querySelector("SELECTOR FOR PDF OBJECT")
      const btn = document.createElement("a")
      btn.setAttribute("href", pdfObj.data)
      btn.setAttribute("download","FILE_NAME")
      btn.click()
      })

This is also not a perfect solution, since you should technically use Browser.setDownloadBehavior but accessing browser._connection does not work anymore as well (I think since release 14.4.0). Is there a way to bypass that and receive the CDPSession used by the browser like there is with the page (Page._client())?