Tests for the new Selenium WebDriver addition

Inclusions of `SeleniumBrowserWrapper`, `SeleniumBrowser`, and several required helper functions that are part of the upcoming `ContentCollector` class and the `WebCollectionAgent`.

Provides an optional drop-in replacement for `SimpleTextBrowser` with `SeleniumBrowserWrapper` for use-cases including pages that depend on JavaScript and others that prevent calls from `requests`.  Nearly all compatibility is held through with the exception of page numbering.

    The ContentAgent class is a custom Autogen agent that can be used to collect and store online content from different web pages. It extends the ConversableAgent class and provides additional functionality for managing a list of additional links, storing collected content in local directories, and customizing request headers.
    ContentAgent uses deque to manage a list of additional links for further exploration, with a maximum depth limit set by max_depth parameter. The collected content is stored in the specified storage path (storage_path) using local directories.
    ContentAgent can be customized with request_kwargs and llm_config parameters during instantiation. The default User-Agent header is used for requests, but it can be overridden by providing a new dictionary of headers under request_kwargs.

Very minor updates prior to submitting a PR

small fix in the `fix_missing_protocol` function

Small addition to maintain a dictionary of processed html content, referenced by the source URL (Uniform Resource Locator)

We cover a small sample of websites, asserting expectations against a number of measurements performed on the collected content.  

The assertions include, but are not limited to: 
- the expected variables contain values
- the presence of the expected output files
- that the expected output files are not empty

Further improvements can include:
- evaluation against all choices of WebDriver to confirm functionality 
- evaluation against a larger sample of websites
-

It's noted that `_set_page_content`, `_split_pages`, and `viewport` are likely not yet compatible but seemingly not necessary at this time for the selenium browser wrapper class.

Small updates on imports that have been recently refactored to other locations.  Specifically:
```
from ..agent import Agent
from .. import ConversableAgent, AssistantAgent, UserProxyAgent, GroupChatManager, GroupChat
from ...oai.client import OpenAIWrapper
```

A small change to declaring `self.browser_kwargs` prior to initializing the parent class (ConversableAgent).  This is done to avoid triggering an unexpected argument error for `browser_kwargs`.

fixing the following pre-commit errors:
autogen/agentchat/contrib/content_agent.py:21:1: E402 Module level import not at top of file
autogen/agentchat/contrib/content_agent.py:34:1: E402 Module level import not at top of file
autogen/agentchat/contrib/content_agent.py:65:33: F811 Redefinition of unused `deque` from line 6
autogen/agentchat/contrib/content_agent.py:374:26: F811 Redefinition of unused `filename` from line 7

Fixing the redundant import of selenium webdriver within `SeleniumBrowser`

Small corrections based on pre-commit errors, both resulting in removed code:
content_agent.py:94:9: F821 Undefined name `f`
content_agent.py:371:26: F811 Redefinition of unused `filename` from line 21

pre-commit fixes for:
autogen/browser_utils.py:455: argumnets ==> arguments
autogen/browser_utils.py:486: compatability ==> compatibility

Still a bit new to the unit test framework and had to remove some conditional statements that are covered elsewhere

…thon-package.yml workflow, and additional small fixes to bring the PR into compliance

… Updated to account for refactoring. All updates now stable and done. Inside Dev Docker, all test files Passed, all pre-commit checks Passed.

…e dev docker and the pytest library

…all the emails.

This update should GitHub to use the WebSurfer extras when testing test_web_surfer_selenium.py.

Adding coverage within the Websurfer workflow for this PR: 
- test/agentchat/contrib/test_web_surfer_selenium.py 
- test/agentchat/contrib/test_content_agent.py

Adding `test/agentchat/contrib/test_content_agent.py --skip-openai` under the assumption that all test files must be accounted for or they will rely on the default workflow.  This test requires openAI calls, but still needs to be registered on this file to avoid build errors.

removed duplicate entry for test_web_surfer_selenium.py

Added the missing `pillow` dependency for graphical based web browsing and downstream tasks

Moving the ContentAgent import to be conditional on "not skip_oai" in the hope that it helps avoid the `markdownify` import error during build tests.

…y pre-commit.

…ting web data. Added '_' to internal functions and docstrings to the web_archiver_agent.py file.

change _set_page_content to set_page_content

change _set_page_content to set_page_content

Removing the exception messages related to Selenium

Minor fix to permit testing