mongrel 0.2.2 → 0.3

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

@@ -86,10 +86,10 @@
       <h3 class="section-bar">Methods</h3>
 
       <div class="name-list">
-      <a href="#M000019">new</a>&nbsp;&nbsp;
-      <a href="#M000020">register</a>&nbsp;&nbsp;
-      <a href="#M000022">resolve</a>&nbsp;&nbsp;
-      <a href="#M000021">unregister</a>&nbsp;&nbsp;
+      <a href="#M000021">new</a>&nbsp;&nbsp;
+      <a href="#M000022">register</a>&nbsp;&nbsp;
+      <a href="#M000024">resolve</a>&nbsp;&nbsp;
+      <a href="#M000023">unregister</a>&nbsp;&nbsp;
       </div>
     </div>
 
@@ -111,12 +111,12 @@
     <div id="methods">
       <h3 class="section-bar">Public Class methods</h3>
 
-      <div id="method-M000019" class="method-detail">
-        <a name="M000019"></a>
+      <div id="method-M000021" class="method-detail">
+        <a name="M000021"></a>
 
         <div class="method-heading">
-          <a href="URIClassifier.src/M000019.html" target="Code" class="method-signature"
-            onclick="popupCode('URIClassifier.src/M000019.html');return false;">
+          <a href="URIClassifier.src/M000021.html" target="Code" class="method-signature"
+            onclick="popupCode('URIClassifier.src/M000021.html');return false;">
           <span class="method-name">URIClassifier.new &rarr; URIClassifier<br />
 </span>
           </a>
@@ -143,12 +143,12 @@ portion.
 
       <h3 class="section-bar">Public Instance methods</h3>
 
-      <div id="method-M000020" class="method-detail">
-        <a name="M000020"></a>
+      <div id="method-M000022" class="method-detail">
+        <a name="M000022"></a>
 
         <div class="method-heading">
-          <a href="URIClassifier.src/M000020.html" target="Code" class="method-signature"
-            onclick="popupCode('URIClassifier.src/M000020.html');return false;">
+          <a href="URIClassifier.src/M000022.html" target="Code" class="method-signature"
+            onclick="popupCode('URIClassifier.src/M000022.html');return false;">
           <span class="method-name">uc.register("/someuri", SampleHandler.new) &rarr; nil<br />
 </span>
           </a>
@@ -173,15 +173,17 @@ to get right with a Hash.
         </div>
       </div>
 
-      <div id="method-M000022" class="method-detail">
-        <a name="M000022"></a>
+      <div id="method-M000024" class="method-detail">
+        <a name="M000024"></a>
 
         <div class="method-heading">
-          <a href="URIClassifier.src/M000022.html" target="Code" class="method-signature"
-            onclick="popupCode('URIClassifier.src/M000022.html');return false;">
+          <a href="URIClassifier.src/M000024.html" target="Code" class="method-signature"
+            onclick="popupCode('URIClassifier.src/M000024.html');return false;">
           <span class="method-name">uc.resolve("/someuri") &rarr; "/someuri", "", handler<br />
 uc.resolve("/someuri/pathinfo") &rarr; "/someuri", "/pathinfo", handler<br />
 uc.resolve("/notfound/orhere") &rarr; nil, nil, nil<br />
+uc.resolve("/") &rarr; "/", "/", handler  # if uc.register("/", handler)<br />
+uc.resolve("/path/from/root") &rarr; "/", "/path/from/root", handler  # if uc.register("/", handler) <br />
 </span>
           </a>
         </div>
@@ -209,18 +211,26 @@ chars, anything really. It also means that it&#8217;s very efficient to do
 this only taking as long as the URI has characters.
 </p>
 <p>
+A slight modification to the CGI 1.2 standard is given for handlers
+registered to &quot;/&quot;. CGI expects all CGI scripts to be at some
+script path, so it doesn&#8217;t really say anything about a script that
+handles the root. To make this work, the resolver will detect that the
+requested handler is at &quot;/&quot;, and return that for script_name, and
+then simply return the full URI back as path_info.
+</p>
+<p>
 It expects strings with no embedded &#8217;\0&#8217; characters.
-Don&#8216;t try other string-line stuff yet.
+Don&#8216;t try other string-like stuff yet.
 </p>
         </div>
       </div>
 
-      <div id="method-M000021" class="method-detail">
-        <a name="M000021"></a>
+      <div id="method-M000023" class="method-detail">
+        <a name="M000023"></a>
 
         <div class="method-heading">
-          <a href="URIClassifier.src/M000021.html" target="Code" class="method-signature"
-            onclick="popupCode('URIClassifier.src/M000021.html');return false;">
+          <a href="URIClassifier.src/M000023.html" target="Code" class="method-signature"
+            onclick="popupCode('URIClassifier.src/M000023.html');return false;">
           <span class="method-name">uc.unregister("/someuri")<br />
 </span>
           </a>

data/doc/rdoc/classes/Mongrel/URIClassifier.src/M000021.html

@@ -5,32 +5,35 @@
 
 <html>
 <head>
-  <title>unregister (Mongrel::URIClassifier)</title>
+  <title>new (Mongrel::URIClassifier)</title>
   <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
   <link rel="stylesheet" href="../../.././rdoc-style.css" type="text/css" media="screen" />
 </head>
 <body class="standalone-code">
   <pre>/**
  * call-seq:
- *    uc.unregister(&quot;/someuri&quot;)
+ *    URIClassifier.new -&gt; URIClassifier
+ *
+ * Initializes a new URIClassifier object that you can use to associate URI sequences
+ * with objects.  You can actually use it with any string sequence and any objects,
+ * but it's mostly used with URIs.
+ *
+ * It uses TST from http://www.octavian.org/cs/software.html to build an ternary search
+ * trie to hold all of the URIs.  It uses this to do an initial search for the a URI
+ * prefix, and then to break the URI into SCRIPT_NAME and PATH_INFO portions.  It actually
+ * will do two searches most of the time in order to find the right handler for the
+ * registered prefix portion.
  *
- * Yep, just removes this uri and it's handler from the trie.
  */
-VALUE URIClassifier_unregister(VALUE self, VALUE uri)
+VALUE URIClassifier_init(VALUE self)
 {
-  void *handler = NULL;
-  struct tst *tst = NULL;
-  DATA_GET(self, struct tst, tst);
-
-  handler = tst_delete((unsigned char *)StringValueCStr(uri), tst);
+  VALUE hash;
 
-  if(handler) {
-    rb_hash_delete(rb_ivar_get(self, id_handler_map), uri);
+  // we create an internal hash to protect stuff from the GC
+  hash = rb_hash_new();
+  rb_ivar_set(self, id_handler_map, hash);
 
-    return (VALUE)handler;
-  } else {
-    return Qnil;
-  }
+  return self;
 }</pre>
 </body>
 </html>

data/doc/rdoc/classes/Mongrel/URIClassifier.src/M000022.html

@@ -5,64 +5,47 @@
 
 <html>
 <head>
-  <title>resolve (Mongrel::URIClassifier)</title>
+  <title>register (Mongrel::URIClassifier)</title>
   <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
   <link rel="stylesheet" href="../../.././rdoc-style.css" type="text/css" media="screen" />
 </head>
 <body class="standalone-code">
   <pre>/**
  * call-seq:
- *    uc.resolve(&quot;/someuri&quot;) -&gt; &quot;/someuri&quot;, &quot;&quot;, handler
- *    uc.resolve(&quot;/someuri/pathinfo&quot;) -&gt; &quot;/someuri&quot;, &quot;/pathinfo&quot;, handler
- *    uc.resolve(&quot;/notfound/orhere&quot;) -&gt; nil, nil, nil
+ *    uc.register(&quot;/someuri&quot;, SampleHandler.new) -&gt; nil
  *
- * Attempts to resolve either the whole URI or at the longest prefix, returning
- * the prefix (as script_info), path (as path_info), and registered handler
- * (usually an HttpHandler).  If it doesn't find a handler registered at the longest
- * match then it returns nil,nil,nil.
+ * Registers the SampleHandler (one for all requests) with the &quot;/someuri&quot;.
+ * When URIClassifier::resolve is called with &quot;/someuri&quot; it'll return
+ * SampleHandler immediately.  When called with &quot;/someuri/iwant&quot; it'll also
+ * return SomeHandler immediatly, with no additional searches, but it will
+ * return path info with &quot;/iwant&quot;.
  *
- * Because the resolver uses a trie you are able to register a handler at *any* character
- * in the URI and it will be handled as long as it's the longest prefix.  So, if you 
- * registered handler #1 at &quot;/something/lik&quot;, and #2 at &quot;/something/like/that&quot;, then a
- * a search for &quot;/something/like&quot; would give you #1.  A search for &quot;/something/like/that/too&quot;
- * would give you #2.
- * 
- * This is very powerful since it means you can also attach handlers to parts of the ; 
- * (semi-colon) separated path params, any part of the path, use off chars, anything really.
- * It also means that it's very efficient to do this only taking as long as the URI has
- * characters.
- *
- * It expects strings with no embedded '\0' characters.  Don't try other string-line stuff yet.
+ * You actually can reuse this class to register nearly anything and 
+ * quickly resolve it.  This could be used for caching, fast mapping, etc.
+ * The downside is it uses much more memory than a Hash, but it can be
+ * a lot faster.  It's main advantage is that it works on prefixes, which
+ * is damn hard to get right with a Hash.
  */
-VALUE URIClassifier_resolve(VALUE self, VALUE uri)
+VALUE URIClassifier_register(VALUE self, VALUE uri, VALUE handler)
 {
-  void *handler = NULL;
-  int pref_len = 0;
+  int rc = 0;
+  void *ptr = NULL;
   struct tst *tst = NULL;
-  VALUE result;
-  unsigned char *uri_str = NULL;
-  unsigned char *script_name_str = NULL;
-
   DATA_GET(self, struct tst, tst);
-  uri_str = (unsigned char *)StringValueCStr(uri);
-
-  handler = tst_search(uri_str, tst, &amp;pref_len);
 
-  // setup for multiple return values
-  result = rb_ary_new();
+  rc = tst_insert((unsigned char *)StringValueCStr(uri), (void *)handler , tst, 0, &amp;ptr);
 
-  if(handler) {
-    rb_ary_push(result, rb_str_substr (uri, 0, pref_len));
-    rb_ary_push(result, rb_str_substr(uri, pref_len, RSTRING(uri)-&gt;len));
-    rb_ary_push(result, (VALUE)handler);
-  } else {
-    // not found so push back nothing
-    rb_ary_push(result, Qnil);
-    rb_ary_push(result, Qnil);
-    rb_ary_push(result, Qnil);
+  if(rc == TST_DUPLICATE_KEY) {
+    rb_raise(rb_eStandardError, &quot;Handler already registered with that name&quot;);
+  } else if(rc == TST_ERROR) {
+    rb_raise(rb_eStandardError, &quot;Memory error registering handler&quot;);
+  } else if(rc == TST_NULL_KEY) {
+    rb_raise(rb_eStandardError, &quot;URI was empty&quot;);
   }
+  
+  rb_hash_aset(rb_ivar_get(self, id_handler_map), uri, handler);
 
-  return result;
+  return Qnil;
 }</pre>
 </body>
 </html>

data/doc/rdoc/classes/Mongrel/URIClassifier.src/M000023.html

@@ -0,0 +1,36 @@
+<?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>
+<head>
+  <title>unregister (Mongrel::URIClassifier)</title>
+  <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
+  <link rel="stylesheet" href="../../.././rdoc-style.css" type="text/css" media="screen" />
+</head>
+<body class="standalone-code">
+  <pre>/**
+ * call-seq:
+ *    uc.unregister(&quot;/someuri&quot;)
+ *
+ * Yep, just removes this uri and it's handler from the trie.
+ */
+VALUE URIClassifier_unregister(VALUE self, VALUE uri)
+{
+  void *handler = NULL;
+  struct tst *tst = NULL;
+  DATA_GET(self, struct tst, tst);
+
+  handler = tst_delete((unsigned char *)StringValueCStr(uri), tst);
+
+  if(handler) {
+    rb_hash_delete(rb_ivar_get(self, id_handler_map), uri);
+
+    return (VALUE)handler;
+  } else {
+    return Qnil;
+  }
+}</pre>
+</body>
+</html>

data/doc/rdoc/classes/Mongrel/URIClassifier.src/M000024.html

@@ -0,0 +1,84 @@
+<?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>
+<head>
+  <title>resolve (Mongrel::URIClassifier)</title>
+  <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1" />
+  <link rel="stylesheet" href="../../.././rdoc-style.css" type="text/css" media="screen" />
+</head>
+<body class="standalone-code">
+  <pre>/**
+ * call-seq:
+ *    uc.resolve(&quot;/someuri&quot;) -&gt; &quot;/someuri&quot;, &quot;&quot;, handler
+ *    uc.resolve(&quot;/someuri/pathinfo&quot;) -&gt; &quot;/someuri&quot;, &quot;/pathinfo&quot;, handler
+ *    uc.resolve(&quot;/notfound/orhere&quot;) -&gt; nil, nil, nil
+ *    uc.resolve(&quot;/&quot;) -&gt; &quot;/&quot;, &quot;/&quot;, handler  # if uc.register(&quot;/&quot;, handler)
+ *    uc.resolve(&quot;/path/from/root&quot;) -&gt; &quot;/&quot;, &quot;/path/from/root&quot;, handler  # if uc.register(&quot;/&quot;, handler) 
+ * 
+ * Attempts to resolve either the whole URI or at the longest prefix, returning
+ * the prefix (as script_info), path (as path_info), and registered handler
+ * (usually an HttpHandler).  If it doesn't find a handler registered at the longest
+ * match then it returns nil,nil,nil.
+ *
+ * Because the resolver uses a trie you are able to register a handler at *any* character
+ * in the URI and it will be handled as long as it's the longest prefix.  So, if you 
+ * registered handler #1 at &quot;/something/lik&quot;, and #2 at &quot;/something/like/that&quot;, then a
+ * a search for &quot;/something/like&quot; would give you #1.  A search for &quot;/something/like/that/too&quot;
+ * would give you #2.
+ * 
+ * This is very powerful since it means you can also attach handlers to parts of the ; 
+ * (semi-colon) separated path params, any part of the path, use off chars, anything really.
+ * It also means that it's very efficient to do this only taking as long as the URI has
+ * characters.
+ *
+ * A slight modification to the CGI 1.2 standard is given for handlers registered to &quot;/&quot;.
+ * CGI expects all CGI scripts to be at some script path, so it doesn't really say anything
+ * about a script that handles the root.  To make this work, the resolver will detect that
+ * the requested handler is at &quot;/&quot;, and return that for script_name, and then simply return
+ * the full URI back as path_info.
+ *
+ * It expects strings with no embedded '\0' characters.  Don't try other string-like stuff yet.
+ */
+VALUE URIClassifier_resolve(VALUE self, VALUE uri)
+{
+  void *handler = NULL;
+  int pref_len = 0;
+  struct tst *tst = NULL;
+  VALUE result;
+  unsigned char *uri_str = NULL;
+  unsigned char *script_name_str = NULL;
+
+  DATA_GET(self, struct tst, tst);
+  uri_str = (unsigned char *)StringValueCStr(uri);
+
+  handler = tst_search(uri_str, tst, &amp;pref_len);
+
+  // setup for multiple return values
+  result = rb_ary_new();
+
+  if(handler) {
+    rb_ary_push(result, rb_str_substr (uri, 0, pref_len));
+    // compensate for a script_name=&quot;/&quot; where we need to add the &quot;/&quot; to path_info to keep it consistent
+    if(pref_len == 1 &amp;&amp; uri_str[0] == '/') {
+      // matches the root URI so we have to use the whole URI as the path_info
+      rb_ary_push(result, uri);
+    } else {
+      // matches a script so process like normal
+      rb_ary_push(result, rb_str_substr(uri, pref_len, RSTRING(uri)-&gt;len));
+    }
+      
+    rb_ary_push(result, (VALUE)handler);
+  } else {
+    // not found so push back nothing
+    rb_ary_push(result, Qnil);
+    rb_ary_push(result, Qnil);
+    rb_ary_push(result, Qnil);
+  }
+
+  return result;
+}</pre>
+</body>
+</html>

data/doc/rdoc/created.rid

@@ -1 +1 @@
-Fri Feb 03 01:14:07 EST 2006
+Fri Feb 10 02:50:17 EST 2006

data/doc/rdoc/files/README.html

@@ -56,7 +56,7 @@
     </tr>
     <tr class="top-aligned-row">
       <td><strong>Last Update:</strong></td>
-      <td>Fri Feb 03 01:13:51 EST 2006</td>
+      <td>Fri Feb 10 02:18:14 EST 2006</td>
     </tr>
     </table>
   </div>
@@ -84,38 +84,41 @@ portability issues.
 </p>
 <h2>Status</h2>
 <p>
-The 0.2.2 release of <a href="../classes/Mongrel.html">Mongrel</a> features
-an HTTP core server that is the fastest possible thing I could get without
-using something other than Ruby. It features a few bug fixes, but mostly
-just a change to the <a
-href="../classes/Mongrel/HttpResponse.html">Mongrel::HttpResponse</a> class
-to make it more feature complete. The remaining development will be spent
-getting <a href="../classes/Mongrel.html">Mongrel</a> to work with other
-frameworks, adding additional needed features, and improving the
-concurrency and speed.
+The 0.3 release is the first official release to start supporting Ruby on
+Rails and to have a more complete DirHandler for serving directories of
+files. This release is actually closer to a full functioning web server
+than the previous releases.
 </p>
 <p>
-The current release has samples from &quot;why the lucky stiff&quot; for
-his Camping framework in the examples directory. Camping is a small micro
-framework (<a
-href="http://rubyforge.org/projects/camping">rubyforge.org/projects/camping</a>)
-which should work with <a href="../classes/Mongrel.html">Mongrel</a> if you
-use the subversion source for Camping.
+The Rails support is pretty rough right now, but check out the
+bin/mongrel_rails file, which should be installed into your PATH if you use
+a gem. You should be able to do the following to run your Rails
+applications:
 </p>
+<pre>
+  &gt; cd myrailsapp
+  &gt; mongrel_rails 0.0.0.0 3000
+</pre>
 <p>
-This is also the first release onto the new <a
-href="../classes/Mongrel.html">Mongrel</a> RubyForge project page found at
-<a
-href="http://rubyforge.org/projects/mongrel">rubyforge.org/projects/mongrel</a>/
-thanks to Tom Copland. I&#8217;ll be looking to automate management of
-this, but feel free to use rubyforge to post feature requests, bugs, and
-join the mailing list.
+And then hit <a href="http://localhost:3000">localhost:3000</a>/ to see
+your app. One thing is that if you have a public/index.html file then
+you&#8217;ll get that served instead of your Rails application.
 </p>
 <p>
-Finally, it now supports all CGI parameters that don&#8217;t cause a
-performance hit, and it has a <a
-href="../classes/Mongrel/DirHandler.html">Mongrel::DirHandler</a> which can
-serve files out of a directory and do (optional) directory listings.
+People with the daemons gem installed will see that mongrel_rails will go
+into the background. You can kill it with:
+</p>
+<pre>
+  &gt; kill -TERM `cat log/mongrel-3000.pid`
+</pre>
+<p>
+Where &quot;3000&quot; is whatever port you told it to listen on when you
+ran it.
+</p>
+<p>
+The file serving is still a little rough and the redirects might not work
+well, but try it out and tell me about any weird errors. File uploads will
+definitely have some problems.
 </p>
 <h2>Install</h2>
 <p>
@@ -203,12 +206,6 @@ applications in a heavily utilized production environment. Right now
 I&#8217;m looking at:
 </p>
 <ul>
-<li>Fast static file handling with directory listings.
-
-</li>
-<li>More testing on more platforms.
-
-</li>
 <li>An idea I&#8217;ve had for an insane caching handler which could speed up
 quite a
 
@@ -217,14 +214,6 @@ quite a
 <p>
 few deployments.
 </p>
-<ul>
-<li>General little things most web servers need.
-
-</li>
-<li>A nice management system or interface for controlling mongrel servers.
-
-</li>
-</ul>
 <p>
 Overall though the goal of <a href="../classes/Mongrel.html">Mongrel</a> is
 to be just enough HTTP to serve a Ruby web application that sits behind a
@@ -238,7 +227,8 @@ hosting the major web frameworks for Ruby:
 <li>Ruby on Rails &#8212; that&#8217;s where my bread is buttered right now.
 
 </li>
-<li>Nitro &#8212; George is a nice guy, and Nitro is thread safe. Might be fun.
+<li>Nitro &#8212; Nitro folks have already hooked this up and started using it.
+Nice.
 
 </li>
 <li>????? &#8212; Others people might be interested in.