mongrel 0.3.8 → 0.3.9

data/README

@@ -11,11 +11,27 @@ scream without too many portability issues.
 
 == Status
 
-The 0.3.6 release supports Ruby On Rails much better than previously, and also
-sports the beginning of a command and plugin infrastructure.  There is now a more
-complete CGIWrapper that handles most of the CGI usage, but still doesn't do the
-MIME decoding or file upload/send (it leaves that to CGI).  Finally, there's a
-great mongrel_rails_service script for running under Win32 as a service.
+Mongrel 0.3.9 now supports a fancy RubyGems based plugin system called GemPlugin.
+It uses the basic machinery of RubyGems to implement dynamically loaded plugins
+based on dependencies.  Writing a plugin is pretty easy, but right now it's
+not as well documented as it should be.  There is a simple example plugin 
+for adding a status command to your mongrel.  Just do:
+
+ > gem install mongrel_status
+
+And you'll then get a new status command.  Then just do:
+
+ > cd myrailsapp
+ > mongrel_rails start -d
+ > mongrel_rails status
+
+And it'll print out the PID your Rails app is running under.
+
+The GemPlugin project is a sub-project of Mongrel, but it's licensed under
+the Ruby license and is usable outside Mongrel
+
+
+== Quick Start
 
 After you've installed (either with gem install mongrel or via source) you should
 have the mongrel_rails command available in your PATH.  Then you just do the following:
@@ -58,7 +74,11 @@ and do:
      -r c:\my\path\to\myapp -p 4000 -e production
  $ mongrel_rails_service start -n myapp
 
-Now hit the port and poof, works.  *Stopping the service is a little problematic right now.*
+Now hit the port and poof, works.  
+
+Stopping a service is simple:
+
+ $ mongrel_rails_service stop -n myapp
 
 If you run into an app that's not running right, my suggestion is to run it with
 the regular mongrel_rails runner:
@@ -68,6 +88,10 @@ the regular mongrel_rails runner:
 
 Since that will spit out error messages and stuff to the console.  *Use CTRL-Pause/Break to stop.*
 
+Best thing about the win32 support is that you can simply use the Windows Services
+in Control Panel->Admin Tools to work with it.  You can also install the same
+Rails app as different installs.  For example I've got myapp_dev, and myapp_prod and
+just start/stop which one I want to work with.
 
 == Install
 
@@ -131,24 +155,6 @@ create a matching database connection for each processor thread.  More on
 this in future releases.
 
 
-== The Future
-
-With the core of Mongrel completed I'm now turning to the next set of features
-to make Mongrel useful for hosting web applications in a heavily utilized
-production environment.  Right now I'm looking at:
-
-* An idea I've had for an insane caching handler which could speed up quite a
-few deployments.
-
-Overall though the goal of Mongrel is to be just enough HTTP to serve a Ruby
-web application that sits behind a more complete web server.  Everything 
-in the next will focus on actually hosting the major web frameworks for Ruby:
-
-* Camping -- because it's already done (thanks Why).
-* Ruby on Rails -- that's where my bread is buttered right now.
-* Nitro -- Nitro folks have already hooked this up and started using it.  Nice.
-* ????? -- Others people might be interested in.
-
 == Contact
 
 E-mail zedshaw at zedshaw.com and I'll help.  Comments about the API are welcome.

data/Rakefile

@@ -10,7 +10,7 @@ include FileUtils
 setup_tests
 setup_clean ["ext/http11/*.{bundle,so,obj,pdb,lib,def,exp}", "ext/http11/Makefile", "pkg", "lib/*.bundle", "*.gem", "doc/site/output", ".config"]
 
-setup_rdoc ['README', 'LICENSE', 'COPYING', 'lib/*.rb', 'doc/**/*.rdoc', 'ext/http11/http11.c']
+setup_rdoc ['README', 'LICENSE', 'COPYING', 'lib/**/*.rb', 'doc/**/*.rdoc', 'ext/http11/http11.c']
 
 desc "Does a full compile, test run"
 task :default => [:compile, :test]
@@ -26,11 +26,12 @@ end
 task :site do
   sh %{pushd doc/site; webgen; scp -r output/* #{ENV['SSH_USER']}@rubyforge.org:/var/www/gforge-projects/mongrel/; popd }
   sh %{ scp -r doc/rdoc/* #{ENV['SSH_USER']}@rubyforge.org:/var/www/gforge-projects/mongrel/rdoc/ }
+  sh %{ cd projects/gem_plugin; rake site }
 end
 
 setup_extension("http11", "http11")
 
-version="0.3.8"
+version="0.3.9"
 summary = "A small fast HTTP library and server that runs Rails, Camping, and Nitro apps."
 test_file = "test/test_ws.rb"
 author="Zed A. Shaw"
@@ -39,6 +40,7 @@ scripts=['mongrel_rails']
 
 setup_gem(name, version,  author, summary, scripts, test_file) do |spec|
   spec.add_dependency('daemons', '>= 0.4.2')
+  spec.add_dependency('gem_plugin', ">= 0.1")
 end
 
 desc "Build a binary gem for Win32"
@@ -48,9 +50,17 @@ scripts_win32 = scripts + ['mongrel_rails_service']
 task :package_win32 do
   setup_win32_gem(name, version,  version, summary, scripts_win32, test_file) do |spec|
     spec.add_dependency('win32-service', '>= 0.5.0')
+    spec.add_dependency('gem_plugin', ">= 0.1")
     spec.files << 'ext/http11/http11.so'
     spec.extensions = []
     spec.platform = Gem::Platform::WIN32
   end
 end
 
+task :gem_plugin_project do 
+  sh %{cd projects/gem_plugin; rake gem_test; }
+end
+
+task :gem_test => [:gem_plugin_project, :package] do
+  sh %{sudo gem install pkg/mongrel-#{version}}
+end  

data/bin/mongrel_rails

@@ -1,8 +1,10 @@
 require 'rubygems'
 require 'mongrel/rails'
+require 'yaml'
 
 
-class Start < Mongrel::Plugin "/commands"
+
+class Start < GemPlugin::Plugin "/commands"
   include Mongrel::Command::Base
 
   def configure
@@ -18,7 +20,6 @@ class Start < Mongrel::Plugin "/commands"
       ['-m', '--mime PATH', "A YAML file that lists additional MIME types", :@mime_map, nil],
       ['-c', '--chdir PATH', "Change to dir before starting (will be expanded)", :@cwd, Dir.pwd],
       ['-r', '--root PATH', "Set the document root (default 'public')", :@docroot, "public"],
-      ['-L', '--load PATH', "Loads plugins from the given directory", :@load_from, nil],
     ]
   end
   
@@ -34,8 +35,6 @@ class Start < Mongrel::Plugin "/commands"
     valid_dir? @docroot, "Path to docroot not valid: #@docroot"
     valid_exists? @mime_map, "MIME mapping file does not exist: #@mime_map" if @mime_map
 
-    valid_dir? @load_from, "Plugin directory path does not exist" if @load_from
-
     return @valid
   end
 
@@ -87,6 +86,8 @@ class Start < Mongrel::Plugin "/commands"
 
     server = Mongrel::HttpServer.new(@address, @port, @num_procs.to_i, @timeout.to_i)
     server.register("/", rails)
+
+    # start mongrel processing thread
     server.run
 
     # signal trapping just applies to posix systems
@@ -109,7 +110,10 @@ class Start < Mongrel::Plugin "/commands"
         @restart = true
       }
     end
-    
+
+    # hook up any rails specific plugins
+    GemPlugin::Manager.instance.load "mongrel" => GemPlugin::INCLUDE
+
     begin
       STDERR.puts "Server ready."
       server.acceptor.join
@@ -123,7 +127,6 @@ class Start < Mongrel::Plugin "/commands"
   end
 
   def run
-    Mongrel::PluginManager.instance.load(@load_from) if @load_from
     daemonize
     rails = configure_rails
     start_mongrel(rails)
@@ -144,7 +147,7 @@ def send_signal(signal, pid_file)
 end
 
 
-class Stop < Mongrel::Plugin "/commands"
+class Stop < GemPlugin::Plugin "/commands"
   include Mongrel::Command::Base
 
   def configure 
@@ -179,7 +182,7 @@ end
 
 
 
-class Restart < Mongrel::Plugin "/commands"
+class Restart < GemPlugin::Plugin "/commands"
   include Mongrel::Command::Base
 
   def configure 
@@ -212,5 +215,6 @@ class Restart < Mongrel::Plugin "/commands"
   end
 end
 
+GemPlugin::Manager.instance.load "mongrel" => GemPlugin::INCLUDE, "rails" => GemPlugin::EXCLUDE
 
 Mongrel::Command::Registry.instance.run ARGV

data/doc/rdoc/classes/Mongrel.html

@@ -58,6 +58,18 @@
                 <a href="../files/lib/mongrel_rb.html">
                 lib/mongrel.rb
                 </a>
+        <br />
+                <a href="../files/lib/mongrel/cgi_rb.html">
+                lib/mongrel/cgi.rb
+                </a>
+        <br />
+                <a href="../files/lib/mongrel/command_rb.html">
+                lib/mongrel/command.rb
+                </a>
+        <br />
+                <a href="../files/lib/mongrel/handlers_rb.html">
+                lib/mongrel/handlers.rb
+                </a>
         <br />
                 <a href="../files/ext/http11/http11_c.html">
                 ext/http11/http11.c
@@ -100,8 +112,13 @@ to service web application requests fast as possible.
     <div id="class-list">
       <h3 class="section-bar">Classes and Modules</h3>
 
-      Module <a href="Mongrel/Const.html" class="link">Mongrel::Const</a><br />
+      Module <a href="Mongrel/Command.html" class="link">Mongrel::Command</a><br />
+Module <a href="Mongrel/Const.html" class="link">Mongrel::Const</a><br />
+Class <a href="Mongrel/CGIWrapper.html" class="link">Mongrel::CGIWrapper</a><br />
+Class <a href="Mongrel/DirHandler.html" class="link">Mongrel::DirHandler</a><br />
+Class <a href="Mongrel/Error404Handler.html" class="link">Mongrel::Error404Handler</a><br />
 Class <a href="Mongrel/HeaderOut.html" class="link">Mongrel::HeaderOut</a><br />
+Class <a href="Mongrel/HttpHandler.html" class="link">Mongrel::HttpHandler</a><br />
 Class <a href="Mongrel/HttpParser.html" class="link">Mongrel::HttpParser</a><br />
 Class <a href="Mongrel/HttpRequest.html" class="link">Mongrel::HttpRequest</a><br />
 Class <a href="Mongrel/HttpResponse.html" class="link">Mongrel::HttpResponse</a><br />

data/doc/rdoc/classes/Mongrel/CGIWrapper.html

@@ -0,0 +1,381 @@
+<?xml version="1.0" encoding="iso-8859-1"?>
+<!DOCTYPE html 
+     PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
+     "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+<head>
+  <title>Class: Mongrel::CGIWrapper</title>
+  <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
+  <meta http-equiv="Content-Script-Type" content="text/javascript" />
+  <link rel="stylesheet" href="../.././rdoc-style.css" type="text/css" media="screen" />
+  <script type="text/javascript">
+  // <![CDATA[
+
+  function popupCode( url ) {
+    window.open(url, "Code", "resizable=yes,scrollbars=yes,toolbar=no,status=no,height=150,width=400")
+  }
+
+  function toggleCode( id ) {
+    if ( document.getElementById )
+      elem = document.getElementById( id );
+    else if ( document.all )
+      elem = eval( "document.all." + id );
+    else
+      return false;
+
+    elemStyle = elem.style;
+    
+    if ( elemStyle.display != "block" ) {
+      elemStyle.display = "block"
+    } else {
+      elemStyle.display = "none"
+    }
+
+    return true;
+  }
+  
+  // Make codeblocks hidden by default
+  document.writeln( "<style type=\"text/css\">div.method-source-code { display: none }</style>" )
+  
+  // ]]>
+  </script>
+
+</head>
+<body>
+
+
+
+    <div id="classHeader">
+        <table class="header-table">
+        <tr class="top-aligned-row">
+          <td><strong>Class</strong></td>
+          <td class="class-name-in-header">Mongrel::CGIWrapper</td>
+        </tr>
+        <tr class="top-aligned-row">
+            <td><strong>In:</strong></td>
+            <td>
+                <a href="../../files/lib/mongrel/cgi_rb.html">
+                lib/mongrel/cgi.rb
+                </a>
+        <br />
+            </td>
+        </tr>
+
+        <tr class="top-aligned-row">
+            <td><strong>Parent:</strong></td>
+            <td>
+                ::CGI
+            </td>
+        </tr>
+        </table>
+    </div>
+  <!-- banner header -->
+
+  <div id="bodyContent">
+
+
+
+  <div id="contextContent">
+
+    <div id="description">
+      <p>
+The beginning of a complete wrapper around <a
+href="../Mongrel.html">Mongrel</a>&#8217;s internal HTTP processing system
+but maintaining the original Ruby CGI module. Use this only as a crutch to
+get existing CGI based systems working. It should handle everything, but
+please notify me if you see special warnings. This work is still very alpha
+so I need testers to help work out the various corner cases.
+</p>
+<p>
+The CGIWrapper.handler attribute is normally not set and is available for
+frameworks that need to get back to the handler. Rails uses this to give
+people access to the RailsHandler#files (<a
+href="DirHandler.html">DirHandler</a> really) so they can look-up paths and
+do other things withthe files managed there.
+</p>
+<p>
+In Rails you can get the real file for a request with:
+</p>
+<pre>
+ path = @request.cgi.handler.files.can_serve(@request['PATH_INFO'])
+</pre>
+<p>
+Which is ugly but does the job. Feel free to write a Rails helper for that.
+Refer to <a href="DirHandler.html#M000024">DirHandler#can_serve</a> for
+more information on this.
+</p>
+
+    </div>
+
+
+   </div>
+
+    <div id="method-list">
+      <h3 class="section-bar">Methods</h3>
+
+      <div class="name-list">
+      <a href="#M000056">args</a>&nbsp;&nbsp;
+      <a href="#M000057">env_table</a>&nbsp;&nbsp;
+      <a href="#M000052">header</a>&nbsp;&nbsp;
+      <a href="#M000051">new</a>&nbsp;&nbsp;
+      <a href="#M000054">out</a>&nbsp;&nbsp;
+      <a href="#M000053">send_cookies</a>&nbsp;&nbsp;
+      <a href="#M000055">status</a>&nbsp;&nbsp;
+      <a href="#M000058">stdinput</a>&nbsp;&nbsp;
+      <a href="#M000059">stdoutput</a>&nbsp;&nbsp;
+      </div>
+    </div>
+
+  </div>
+
+
+    <!-- if includes -->
+
+    <div id="section">
+
+
+    <div id="constants-list">
+      <h3 class="section-bar">Constants</h3>
+
+      <div class="name-list">
+        <table summary="Constants">
+        <tr class="top-aligned-row context-row">
+          <td class="context-item-name">REMOVED_KEYS</td>
+          <td>=</td>
+          <td class="context-item-value">[ &quot;nph&quot;,&quot;status&quot;,&quot;server&quot;,&quot;connection&quot;,&quot;type&quot;,                      &quot;charset&quot;,&quot;length&quot;,&quot;language&quot;,&quot;expires&quot;]</td>
+          <td width="3em">&nbsp;</td>
+          <td class="context-item-desc">
+these are stripped out of any keys passed to <a
+href="CGIWrapper.html#M000052">CGIWrapper.header</a> function
+
+</td>
+        </tr>
+        </table>
+      </div>
+    </div>
+
+
+
+    <div id="attribute-list">
+      <h3 class="section-bar">Attributes</h3>
+
+      <div class="name-list">
+        <table>
+        <tr class="top-aligned-row context-row">
+          <td class="context-item-name">handler</td>
+          <td class="context-item-value">&nbsp;[W]&nbsp;</td>
+          <td class="context-item-desc"></td>
+        </tr>
+        <tr class="top-aligned-row context-row">
+          <td class="context-item-name">handler</td>
+          <td class="context-item-value">&nbsp;[R]&nbsp;</td>
+          <td class="context-item-desc"></td>
+        </tr>
+        <tr class="top-aligned-row context-row">
+          <td class="context-item-name">options</td>
+          <td class="context-item-value">&nbsp;[R]&nbsp;</td>
+          <td class="context-item-desc"></td>
+        </tr>
+        </table>
+      </div>
+    </div>
+      
+
+
+    <!-- if method_list -->
+    <div id="methods">
+      <h3 class="section-bar">Public Class methods</h3>
+
+      <div id="method-M000051" class="method-detail">
+        <a name="M000051"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000051.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000051.html');return false;">
+          <span class="method-name">new</span><span class="method-args">(request, response, *args)</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+Takes an <a href="HttpRequest.html">HttpRequest</a> and <a
+href="HttpResponse.html">HttpResponse</a> object, plus any additional
+arguments normally passed to CGI. These are used internally to create a
+wrapper around the real CGI while maintaining <a
+href="../Mongrel.html">Mongrel</a>&#8217;s view of the world.
+</p>
+        </div>
+      </div>
+
+      <h3 class="section-bar">Public Instance methods</h3>
+
+      <div id="method-M000056" class="method-detail">
+        <a name="M000056"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000056.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000056.html');return false;">
+          <span class="method-name">args</span><span class="method-args">()</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+Used to wrap the normal args variable used inside CGI.
+</p>
+        </div>
+      </div>
+
+      <div id="method-M000057" class="method-detail">
+        <a name="M000057"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000057.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000057.html');return false;">
+          <span class="method-name">env_table</span><span class="method-args">()</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+Used to wrap the normal <a href="CGIWrapper.html#M000057">env_table</a>
+variable used inside CGI.
+</p>
+        </div>
+      </div>
+
+      <div id="method-M000052" class="method-detail">
+        <a name="M000052"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000052.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000052.html');return false;">
+          <span class="method-name">header</span><span class="method-args">(options = &quot;text/html&quot;)</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+The header is typically called to send back the header. In our case we
+collect it into a hash for later usage.
+</p>
+<p>
+nph &#8212; Mostly ignored. It&#8216;ll output the date. connection &#8212;
+Completely ignored. Why is CGI doing this? length &#8212; Ignored since <a
+href="../Mongrel.html">Mongrel</a> figures this out from what you write to
+output.
+</p>
+        </div>
+      </div>
+
+      <div id="method-M000054" class="method-detail">
+        <a name="M000054"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000054.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000054.html');return false;">
+          <span class="method-name">out</span><span class="method-args">(options = &quot;text/html&quot;) {||| &quot;&quot;)| ...}</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+The dumb thing is people can call header or this or both and in any order.
+So, we just reuse header and then finalize the <a
+href="HttpResponse.html">HttpResponse</a> the right way. Status is taken
+from the various options and converted to what <a
+href="../Mongrel.html">Mongrel</a> needs via the <a
+href="CGIWrapper.html#M000055">CGIWrapper.status</a> function.
+</p>
+        </div>
+      </div>
+
+      <div id="method-M000053" class="method-detail">
+        <a name="M000053"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000053.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000053.html');return false;">
+          <span class="method-name">send_cookies</span><span class="method-args">(to)</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+Takes any &#8216;cookie&#8217; setting and sends it over the <a
+href="../Mongrel.html">Mongrel</a> header, then removes the setting from
+the options. If cookie is an Array or Hash then it sends those on with
+.to_s, otherwise it just calls .to_s on it and hopefully your
+&quot;cookie&quot; can write itself correctly.
+</p>
+        </div>
+      </div>
+
+      <div id="method-M000055" class="method-detail">
+        <a name="M000055"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000055.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000055.html');return false;">
+          <span class="method-name">status</span><span class="method-args">()</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+Computes the status once, but lazily so that people who call header twice
+don&#8217;t get penalized. Because CGI insists on including the options
+status message in the status we have to do a bit of parsing.
+</p>
+        </div>
+      </div>
+
+      <div id="method-M000058" class="method-detail">
+        <a name="M000058"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000058.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000058.html');return false;">
+          <span class="method-name">stdinput</span><span class="method-args">()</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+Used to wrap the normal stdinput variable used inside CGI.
+</p>
+        </div>
+      </div>
+
+      <div id="method-M000059" class="method-detail">
+        <a name="M000059"></a>
+
+        <div class="method-heading">
+          <a href="CGIWrapper.src/M000059.html" target="Code" class="method-signature"
+            onclick="popupCode('CGIWrapper.src/M000059.html');return false;">
+          <span class="method-name">stdoutput</span><span class="method-args">()</span>
+          </a>
+        </div>
+      
+        <div class="method-description">
+          <p>
+The stdoutput should be completely bypassed but we&#8217;ll drop a warning
+just in case
+</p>
+        </div>
+      </div>
+
+
+    </div>
+
+
+  </div>
+
+
+<div id="validator-badges">
+  <p><small><a href="http://validator.w3.org/check/referer">[Validate]</a></small></p>
+</div>
+
+</body>
+</html>