disable pdf viewer and download to skyvern download folder

[LawyZheng]

Important

PDF files are now downloaded to a specified Skyvern directory, with browser preferences updated to disable in-browser viewing.

• Behavior:
  • PDF files are now downloaded to the Skyvern download folder instead of being viewed in the browser.
  • initialize_download_dir() in browser_factory.py sets the download directory based on workflow and task IDs.
  • update_chromium_browser_preferences() in browser_factory.py modifies Chromium preferences to set download directories.
• Files:
  • Adds chromium_preferences.json to define default browser preferences, including external PDF handling.
• Functions:
  • Modifies _create_headless_chromium() and _create_headful_chromium() in browser_factory.py to use the new download directory and preferences.

^{This description was created by for c5ff494. It will automatically update as commits are pushed.}

[ellipsis-dev]

❌ Changes requested. Reviewed everything up to c5ff494 in 2 minutes and 0 seconds

More details

• Looked at 115 lines of code in 2 files
• Skipped 0 files when reviewing.
• Skipped posting 2 drafted comments based on config settings.

1. skyvern/webeye/browser_factory.py:284

• Draft comment:
  Ensure that temporary directories created for user_data_dir and downloads_path are cleaned up after use to prevent disk space issues.
• Reason this comment was not posted:
  Comment did not seem useful.

2. skyvern/webeye/chromium_preferences.json:1

• Draft comment:
  Consider adding validation for the JSON structure to ensure it is well-formed and handle potential parsing errors.
• Reason this comment was not posted:
  Decided after close inspection that this draft comment was likely wrong and/or not actionable:
  The JSON file is likely to be read by a JSON parser, which inherently checks for well-formedness. Adding additional validation might be redundant unless there is a specific reason to suspect the file could be malformed outside of typical JSON parsing. The comment does not provide specific evidence or context that suggests additional validation is needed beyond what a JSON parser would provide.
  I might be underestimating the potential for non-syntax-related issues, such as logical errors in the JSON structure that a parser wouldn't catch. However, the comment does not specify any such issues.
  Without specific evidence or context indicating a need for additional validation beyond syntax checking, the comment seems unnecessary. JSON parsers are reliable for syntax validation.
  The comment should be deleted as it suggests redundant validation that is typically handled by JSON parsers. There is no strong evidence provided that additional validation is necessary.

Workflow ID: wflow_GvstE97marA12tLv

---

Want Ellipsis to fix these issues? Tag @ellipsis-dev in a comment. You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

Consider adding error handling for file I/O operations when reading and writing the preferences file to prevent unhandled exceptions.