Skip to main content
This example demonstrates how to handle file downloads using Zendriver’s expect_download() context manager. You can intercept downloads, access file metadata, and save files with custom names.

​
What it does

The script:
  1. Opens a file upload/conversion website (Yandex OCR translator)
  2. Uploads an image file
  3. Triggers a download by clicking a button
  4. Intercepts the download
  5. Decodes the file data and saves it with the suggested filename

​
Complete code

import asyncio
import base64
import os

import zendriver as zd


async def main() -> None:
    path_image = r"Your Image Path"
    out_dir = r"."
    async with await zd.start() as browser:
        page = browser.main_tab
        await page.get("https://translate.yandex.com/en/ocr")
        await (await page.select('input[type="file"]')).send_file(path_image)
        await asyncio.sleep(3)

        async with page.expect_download() as download_ex:
            await (await page.select("#downloadButton")).mouse_click()
            download = await download_ex.value
        print(download.suggested_filename)
        with open(os.path.join(out_dir, download.suggested_filename), "wb") as fw:
            bytes_file = base64.b64decode(download.url.split(",", 1)[-1])
            fw.write(bytes_file)


if __name__ == "__main__":
    asyncio.run(main())

​
Key techniques

​
Context manager pattern

Use async with to handle the browser lifecycle: 
async with await zd.start() as browser:
    # Browser automatically closes after this block
    page = browser.main_tab

​
Download interception

Wrap the download-triggering action in expect_download(): 
async with page.expect_download() as download_ex:
    await (await page.select("#downloadButton")).mouse_click()
    download = await download_ex.value

​
Download metadata

Access information about the downloaded file: 
print(download.suggested_filename)  # Original filename
print(download.url)  # Data URL or download URL

​
Data URL decoding

For data URLs (base64-encoded content), decode and save: 
bytes_file = base64.b64decode(download.url.split(",", 1)[-1])
with open(filename, "wb") as fw:
    fw.write(bytes_file)

​
Mouse click

Use mouse_click() for more realistic clicking behavior: 
await element.mouse_click()
This simulates actual mouse movements and is more reliable for certain elements than .click().

​
Running the example

  1. Update the path_image variable with your actual image path
  2. Save the code to a file (e.g., download_file.py)
  3. Run it with Python:
python download_file.py
  1. The downloaded file will be saved in the current directory

​
Use cases

  • Automated downloads: Download reports, exports, or generated files
  • File validation: Verify downloaded content matches expectations
  • Custom file handling: Process or rename files based on metadata
  • Batch downloads: Loop through multiple downloads in a workflow
  • Testing: Verify download functionality works correctly

​
Important notes

​
Data URLs vs regular downloads

Some downloads return data URLs (base64-encoded content in the URL itself), while others initiate regular file downloads. This example handles data URLs. For regular file downloads, you may need to configure download behavior differently.

​
Timing considerations

The example uses asyncio.sleep(3) to wait for file processing. In production, consider using more robust waiting strategies: 
# Wait for specific elements to appear
await page.wait_for(selector=".upload-complete")

# Or use Zendriver's built-in waiting
await page.select(".processing-done")

​
File paths

Use raw strings (r"path") or forward slashes for file paths to avoid issues with backslashes on Windows: 
path_image = r"C:\Users\Name\image.png"  # Raw string
# or
path_image = "C:/Users/Name/image.png"  # Forward slashes

Build docs developers (and LLMs) love

Get started for freeTalk to us