Proxy Pac Rb

Build Status Code Climate Coverage Status Gem Version Downloads

proxy_pac_rb is a gem to compress, lint and parse proxy auto-config files. It comes with a cli program, some rack middlewares and can be used from within ruby scripts as well. proxy_pac_rb uses a JavaScript runtime to evaulate a proxy auto-config file the same way a browser does to determine what proxy (if any at all) should a program use to connect to a server. You must install on of the supported JavaScript runtimes: therubyracer or therubyrhino.

Big thanks to sstephenson's execjs for the runtime wrapper code and to samuelkadolph's ruby-pac for the foundation of this gem.

Installing

Add this line to your application's Gemfile:

gem 'proxy_pac_rb'

And then execute:

$ bundle

Or install it yourself as:

$ gem install proxy_pac_rb

Requirements

After installing the proxy_pac_rb gem you must install a JavaScript runtime. Compatible runtimes include:

Usage

Command Line

Find proxy for url

Arguments

  • -p|--proxy-pac FILE: Path to proxy pac file
  • -t|--time YYYY-MM-DD HH:MM:SS: Time to use in proxy.pac
  • -c|--client-ip x.x.x.x: Client-IP to use in proxy.pac
  • -h|--help: Show help

Use

# Download via curl and parse pac
curl -L -o sample.pac https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample.pac
pprb find proxy -c 127.0.0.1 -t "2014-03-09 12:00:00" -p sample.pac -u https://github.com

# Or download via pprb directly and parse pac #2
pprb find proxy -c 127.0.0.1 -t "2014-03-09 12:00:00" -p https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample.pac -u https://github.com

# Or download this example if you are behind a coporate proxy via pprb directly and parse pac #3
pprb find proxy -c 127.0.0.1 -t "2014-03-09 12:00:00" -p https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample.pac -u https://github.com --use-proxy

# =>                url: result
# => https://github.com: DIRECT

Compress proxy.pac-file

You can compress a proxy.pac with pprb to reduce the amount of data transferred to download the proxy.pac.

# Download pac
curl -L -o sample.pac https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample.pac

# Compress pac
pprb compress proxy_pac -p sample.pac

Lint proxy.pac-file

You can lint a proxy.pac with pprb to check a proxy.pac before deploying it.

# Download pac
curl -L -o sample.pac https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample.pac

# Lint pac
pprb lint proxy_pac -p sample.pac

Init new proxy.pac

You can use the following command to start a new proxy.pac:

Plain

pprb init proxy_pac

Add rspec

To test your proxy.pac you can use rspec.

pprb init proxy_pac --test rspec

Add middleman

To build your proxy.pac you can use middleman.

pprb init proxy_pac --build middleman

"rack"-middleware

The middleware which comes with proxy_pac_rb is compliant with the rack-specification - tested via rack/lint and should work with every rack-compliant-server.

Prerequisites

Make sure the content is served with "Content-Type": 'application/x-ns-proxy-autoconfig'. Otherwise the content is ignored by both middlewares.

Warning

The linter-rack-middleware needs to be activated before ANY other middleman-extension, rack-middleware or whatever framework you are using can instantiate the V8-runtime! Only the first time the V8-javascript-engine - aka therubyracer - is instantiated, it is possible to create a binding to ruby code. Every other V8-object created later re-uses this binding.

You might an error like this if you ignore this warning!

error  build/proxy.pac
Unexpected token: name (is) (line: 1, col: 10, pos: 10)

Error
    at new JS_Parse_Error (<eval>:2359:10623)
    at js_error (<eval>:2359:10842)
    at croak (<eval>:2359:19086)
    at token_error (<eval>:2359:19223)
    at unexpected (<eval>:2359:19311)
    at semicolon (<eval>:2359:19784)
    at simple_statement (<eval>:2359:22580)
    at <eval>:2359:20553
    at <eval>:2359:19957
    at <eval>:2359:31968
There were errors during this build

Linter Middleware

require 'proxy_pac_rb/rack/proxy_pac_linter'
use ProxyPacRb::Rack::ProxyPacLinter

Compressor Middleware

require 'proxy_pac_rb/rack/proxy_pac_compressor'
use ProxyPacRb::Rack::ProxyPacCompressor

If you need to change some settings, please pass it a options-parameter. Please see the (uglifier)[https://github.com/lautis/uglifier] for more information about this.

require 'proxy_pac_rb/rack/proxy_pac_compressor'
use ProxyPacRb::Rack::ProxyPacCompressor.new(options: {})

Using "rack"-middleware with "middleman"

If you want to use the rack-middleware with middleman look at the following code snippet captured from the middleman-configuration file config.rb:

  • config.rb:
  # IMPORTANT: Needs to come first before compressor
    require 'proxy_pac_rb/rack/proxy_pac_linter'
    use ProxyPacRb::Rack::ProxyPacLinter

    require 'proxy_pac_rb/rack/proxy_pac_compressor'
    use ProxyPacRb::Rack::ProxyPacCompressor

  # The middleware works on content served with 
  # "Content-Type" 'application/x-ns-proxy-autoconfig'
    page "*.pac", content_type: 'application/x-ns-proxy-autoconfig', layout: false

  # This provides an uncompressed copy of the proxy.pac to make it
  # possible for your support to review it by hand
  Dir.glob(File.join(source_dir, '**', '*.pac*')).each do |f|
    # Path should be relative to source dir
    relative_path = Pathname.new(f).relative_path_from(Pathname.new(source_dir))
    relative_path = relative_path.to_s.gsub(/(?<path>.*\.pac)\..*/, '\k<path>')

    # "text/plain" prevents the middlewares to handle it
    proxy(format('%s.raw', relative_path), relative_path, content_type: 'text/plain', layout: false)
  end

Ruby

Load from website

require 'proxy_pac_rb'

file = ProxyPacRb::Parser.new.parse('https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample.pac')
file.find('https://github.com')        # => "DIRECT"

Load from filesystem

curl -L -o sample.pac https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample.pac

require 'proxy_pac_rb'

file = ProxyPacRb::Parser.new.parse("sample.pac")
file.find('https://github.com')        # => "DIRECT"

Use string

require 'proxy_pac_rb'

file = ProxyPacRb::Parser.new.parse "  function FindProxyForURL(url, host) {\n    return \"DIRECT\";\n  }\n"

file.find('http://localhost') # => "DIRECT"

Use Client IP

require 'proxy_pac_rb'

environment = ProxyPacRb::Environment.new(client_ip: '127.0.0.1')
file = ProxyPacRb::Parser.new(environment).parse('https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample2.pac')
file.find('https://github.com')        # => "PROXY localhost:8080"

environment = ProxyPacRb::Environment.new(client_ip: '127.0.0.2')
file = ProxyPacRb::Parser.new(environment).parse('https://github.com/fedux-org/proxy_pac_rb/raw/master/files/sample2.pac')
file.find('https://github.com')        # => "DIRECT"

Use Date Time

require 'proxy_pac_rb'

string = "function FindProxyForURL(url, host) {\n  if (dateRange(\"JUL\", \"SEP\")) {\n    return \"PROXY localhost:8080\";                                                                                                          \n  } else {\n    return \"DIRECT\";\n  }\n}\n"

environment = ProxyPacRb::Environment.new(time: Time.parse('2014-07-06 12:00'))
file = ProxyPacRb::Parser.new(environment).parse(string)
file.find('http://localhost') # => 'PROXY localhost:8080'

environment = ProxyPacRb::Environment.new(time: Time.parse('2014-03-08 6:00'))
file = ProxyPacRb::Parser.new(environment).parse(string)
file.find('http://localhost') # => 'DIRECT'

Available JavaScript Functions

  • isPlainHostName(host)
  • dnsDomainIs(host, domain)
  • localHostOrDomainIs(host, hostdom)
  • isResolvable(host)
  • isInNet(host, pattern, mask)
  • dnsResolve(host)
  • myIpAddress()
  • dnsDomainLevels(host)
  • shExpMatch(str, shexp)
  • weekdayRange(wd1, wd2, gmt)
  • dateRange(*args)
  • timeRange(*args)
  • alert(msg) (output on stderr by default)

RSpec-integration

proxy_pac_rb comes with helpers and matchers for rspec. To make those helpers and matchers available in your project, add this code snippet in your project:

require 'proxy_pac_rb/rspec'

Helpers

  • proxy_pac:

    This helper makes a proxy.pac available. It requires a source for your proxy.pac given in subject { } - e.g. a file, a string, or a url. It represents a ProxyPacFile.

  • time:

    The time-helper makes 1970-01-01 00:00:00 as time available. Overwrite this helper at will with another time string or a time-object - e.g let(:time) { Time.now }.

  • client_ip:

    The client_ip-helper makes 127.0.0.1 available. Overwrite it with a different ip-address as string or an object which returns an ip-address on #to_s.

  • root_path:

    The root_path-helper is meant for overriding. It is used to find your "proxy.pac"-files. By default its value is Dir.getwd which is set by rspec.

Configuration

You can either configure ProxyPacRb either via a global ProxyPacRb.configure-block:

ProxyPacRb.configure do |config|
  config.use_proxy = true
end

or via RSpec-metadata:

RSpec.describe 'proxy.pac', type: :proxy_pac, use_proxy: true do
end

Examples

To make it easier for you to start, you find some examples below.

Type for specs

It is important that you flag your specs with type: :proxy_pac. Otherwise the helpers are not included and not available in your examples.

RSpec.describe 'proxy.pac', type: :proxy_pac do

Supported sources

String

RSpec.describe 'String', type: :proxy_pac do
  subject do
    "      function FindProxyForURL(url, host) {\n        return \"DIRECT\";\n      }\n    EOS\n  end\nend\n".chomp

Local File

RSpec.describe 'proxy.pac', type: :proxy_pac do
  subject { 'proxy.pac' }
end

URL

RSpec.describe 'http://server/proxy.pac', type: :proxy_pac do
  subject { 'http://server/proxy.pac' }
end

Matchers

Readable

To check if a proxy.pac could be read from filesystem or downloaded via HTTP, check be_readable.

require 'proxy_pac_rb/rspec'

RSpec.describe 'proxy.pac', type: :proxy_pac do
  context 'when proxy pac exist' do
    context 'when is file' do
      subject { 'proxy.pac' }
      it { expect(proxy_pac).to be_readable }
    end

    context 'when is url' do
      subject { 'http://www.example.com/proxy.pac' }
      it { expect(proxy_pac).to be_readable }
    end
  end
end

Equal

If you want to check if a proxy.pac is the same you can use the be_the_same_proxy_pac_file-matcher.

require 'proxy_pac_rb/rspec'

RSpec.describe 'proxy.pac', type: :proxy_pac do
  subject do
    "      function FindProxyForURL(url, host) {\n        return \"DIRECT\";\n      }\n    EOS\n  end\n  context 'Check equality of proxy pac' do\n    context 'when proxy.pac is eq' do\n      it { expect(proxy_pac).to be_the_same_proxy_pac_file 'proxy.pac' }\n    end\n\n    context 'when proxy.pac is not eq' do\n      it { expect(proxy_pac).not_to be_the_same_proxy_pac_file 'proxy.pac' }\n    end\n  end\nend\n".chomp

This is quite handy to compare a local proxy.pac with a remote one - e.g. a deployed one.

require 'proxy_pac_rb/rspec'

RSpec.describe 'proxy.pac', type: :proxy_pac do
  let(:local_proxy_pac) { ProxyPacRb::ProxyPacFile.new(source: file) }

  before(:each) do
    ProxyPacRb::ProxyPacLoader.new.load(local_proxy_pac)
    ProxyPacRb::ProxyPacCompressor.new.compress(local_proxy_pac) if local_proxy_pac.readable?
  end

  context 'when no proxy.pac is given' do
    subject { 'http://server.example.com/proxy.pac' }
    let(:file) { 'source/proxy.pac' }

    it { expect(proxy_pac).to be_the_same_proxy_pac_file(local_proxy_pac) }
  end
end

Valid

You want to check if your proxy.pac is valid. This will find reference errors (e.g. unknown variables) and syntax errors.

require 'proxy_pac_rb/rspec'

RSpec.describe 'proxy.pac', type: :proxy_pac do
  subject do
    "      function FindProxyForURL(url, host) {\n        return \"DIRECT\";\n      }\n    EOS\n  end\n  context 'Check validity of proxy pac' do\n    context 'when proxy.pac is valid' do\n      it { expect(proxy_pac).to be_valid }\n    end\n\n    context 'when proxy.pac is not valid' do\n      subject do\n        <<~EOS.chomp\n          function FindProxyForURL(url, host) {\n            return adsf;\n          }\n        EOS\n      end\n      it { expect(proxy_pac).not_to be_valid }\n    end\n  end\nend\n".chomp

Parse proxy.pac

To make some logic checks use the be_downloaded_via-matcher. This will parse the proxy.pac and returns the proxy to be used for a given url.

require 'proxy_pac_rb/rspec'
RSpec.describe 'proxy.pac', type: :proxy_pac do
  subject do
    "      function FindProxyForURL(url, host) {\n        if (dnsDomainIs(host, 'www1.example.com'')) {\n          return \"PROXY proxy1.example.com:8080\";\n        } else {\n          return \"PROXY proxy2.example.net:8080\";\n        }\n      }\n    EOS\n  end\n  context 'Parse Proxy Pac' do\n    context 'when url is forwarded via' do\n      let(:url) { 'http://www1.example.com' }\n      it { expect(url).to be_downloaded_via 'PROXY proxy.example.com:8080' }\n    end\n\n    context 'when url is not forwarded via' do\n      let(:url) { 'http://www2.example.com' }\n      it { expect(url).not_to be_downloaded_via 'PROXY proxy.example.com:8080' }\n    end\n  end\nend\n".chomp

Developers

After checking out the repo, run script/bootstrap to install dependencies. Then, run script/console for an interactive prompt that will allow you to experiment. To run tests execute script/test.

To install this gem onto your local machine, run bundle exec rake gem:install. To release a new version, update the version number in version.rb, and then run bundle exec rake gem:release to create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Contributing

If you want to contribute: fork, branch & pull request and please see CONTRIBUTING.md.

Running Tests

bundle install
rake test:rspec
rake test:rubyracer
rake test:rubyrhino

If you want to open an issue. Please send a PR with a test describing the bug.