Expose the proxy bypass option

[grahamhar]

In some scenarios there will be a mixture of requests some will need a proxy some won't. This adds and exposes the proxy bypass option.

I have only tested this in chrome on linux, can you help by suggesting:

• What tests need to be added.
• Documentation to be added.
• Updates for other browsers.

[mdmintz]

There are some things that should be fixed first:

You need to flake8 all your code and make sure there are no issues there. I see some long lines that would fail right away.

You left debug statements in your code:
logging.exception('Big issue from GRAHAM')

You mixed args by inserting proxy_bypass_list in the middle of grouped args. It should go after proxy_pass.

    proxy_string,
    proxy_bypass_list,
    proxy_auth,
    proxy_user,
    proxy_pass,

[grahamhar]

I think I addressed all those points @mdmintz

[mdmintz]

@grahamhar According to https://www.chromium.org/developers/design-documents/network-settings , it's a semi-colon separated list, not a comma-separated one. Have you tried an example with multiple domains bypassed?

--proxy-bypass-list=(<trailing_domain>|<ip-address>)[:<port>][;...]

This tells chrome to bypass any specified proxy for the given semi-colon-separated list of hosts. This flag must be used (or rather, only has an effect) in tandem with --proxy-server.
Note that trailing-domain matching doesn't require "." separators so "*google.com" will match "igoogle.com" for example.

For example,
  --proxy-server="foopy:8080" --proxy-bypass-list="*.google.com;*foo.com;127.0.0.1:8080"
will use the proxy server "foopy" on port 8080 for all hosts except those pointing to *.google.com, those pointing to *foo.com and those pointing to localhost on port 8080. 
igoogle.com requests would still be proxied. ifoo.com requests would not be proxied since *foo, not *.foo was specified. 

[mdmintz]

@grahamhar It's mostly for the Python-comment description, but I'd check for a semicolon-separated list, as a comma could appear in valid urls such as data:,Hello.

[grahamhar]

@grahamhar It's mostly for the Python-comment description, but I'd check for a semicolon-separated list, as a comma could appear in valid urls such as data:,Hello.

I think the bypass list only covers domains and IP address not URL so commas are not actually valid. I'll update the comment to reflect it should be a semi colon separated list.

[mdmintz]

@grahamhar I added more comments. Looks like there are still a few places where proxy_bypass_list was not in the correct location in browser_launcher.py.

[grahamhar]

@grahamhar I added more comments. Looks like there are still a few places where proxy_bypass_list was not in the correct location in browser_launcher.py.

Sorry, not sure how I missed them. Updated now.

[mdmintz]

@grahamhar On your line 943 of browser_launcher.py, it looks like you're calling chrome_options = _set_chrome_options( without the proxy_bypass_list as a variable, which breaks my internal tests that include selenium grid tests. Those tests are private because they include private authentication codes to private proxy servers, but I'm happy to walk through all issues I encounter like this.

[grahamhar]

@mdmintz I think there will possibly be quite a few gaps like this. I only managed to test using the pycharm fixture. Let me know any more changes and I'll try to update.

[mdmintz]

@grahamhar Sounds good. You'll need to add that into get_remote_driver( to make that new line also work.

[mdmintz]

@grahamhar Your current line 1017 of browser_launcher.py is missing the proxy_bypass_list in firefox_options = _set_firefox_options(.

[mdmintz]

@grahamhar Merged! Nice work. It'll be part of the next release.

[mdmintz]

@grahamhar It has been released as part of https://github.com/seleniumbase/SeleniumBase/releases/tag/v2.2.8