http_parser.rb 0.5.1 → 0.5.3

data/.gitmodules

@@ -1,6 +1,6 @@
 [submodule "ext/ruby_http_parser/vendor/http-parser"]
 	path = ext/ruby_http_parser/vendor/http-parser
-	url = git://github.com/ry/http-parser.git
+	url = git://github.com/joyent/http-parser.git
 [submodule "ext/ruby_http_parser/vendor/http-parser-java"]
 	path = ext/ruby_http_parser/vendor/http-parser-java
 	url = git://github.com/a2800276/http-parser.java.git

data/Gemfile.lock

@@ -1,16 +1,15 @@
 PATH
   remote: .
   specs:
-    http_parser.rb (0.5.1)
+    http_parser.rb (0.5.2)
 
 GEM
   remote: http://rubygems.org/
   specs:
     diff-lcs (1.1.2)
     json (1.5.1)
-    json (1.5.1-java)
-    rake (0.8.7)
-    rake-compiler (0.7.5)
+    rake (0.9.2)
+    rake-compiler (0.7.9)
       rake
     rspec (2.4.0)
       rspec-core (~> 2.4.0)
@@ -20,13 +19,14 @@ GEM
     rspec-expectations (2.4.0)
       diff-lcs (~> 1.1.2)
     rspec-mocks (2.4.0)
+    yajl-ruby (0.8.2)
 
 PLATFORMS
-  java
   ruby
 
 DEPENDENCIES
   http_parser.rb!
   json (>= 1.4.6)
-  rake-compiler (>= 0.7.5)
+  rake-compiler (>= 0.7.9)
   rspec (>= 2.0.1)
+  yajl-ruby (>= 0.8.1)

data/LICENSE-MIT

@@ -0,0 +1,20 @@
+Copyright 2009,2010 Marc-André Cournoyer <macournoyer@gmail.com>
+Copyright 2010,2011 Aman Gupta <aman@tmm1.net>
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to
+deal in the Software without restriction, including without limitation the
+rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+sell copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+IN THE SOFTWARE.

data/README.md

@@ -3,7 +3,7 @@
 A simple callback-based HTTP request/response parser for writing http
 servers, clients and proxies.
 
-This gem is built on top of [ry/http-parser](http://github.com/ry/http-parser) and its java port [a2800276/http-parser.java](http://github.com/a2800276/http-parser.java).
+This gem is built on top of [joyent/http-parser](http://github.com/joyent/http-parser) and its java port [a2800276/http-parser.java](http://github.com/a2800276/http-parser.java).
 
 ## Supported Platforms
 

data/ext/ruby_http_parser/org/ruby_http_parser/RubyHttpParser.java

@@ -1,6 +1,7 @@
 package org.ruby_http_parser;
 
 import org.jruby.Ruby;
+import org.jruby.RubyArray;
 import org.jruby.RubyClass;
 import org.jruby.RubyHash;
 import org.jruby.RubyModule;
@@ -59,6 +60,9 @@ public class RubyHttpParser extends RubyObject {
   private IRubyObject queryString;
   private IRubyObject fragment;
 
+  private IRubyObject header_value_type;
+  private IRubyObject upgradeData;
+
   private IRubyObject callback_object;
 
   private String _current_header;
@@ -77,6 +81,8 @@ public class RubyHttpParser extends RubyObject {
 
     this.callback_object = null;
 
+    this.header_value_type = runtime.getModule("HTTP").getClass("Parser").getInstanceVariable("@default_header_value_type");
+
     initSettings();
     init();
   }
@@ -130,25 +136,45 @@ public class RubyHttpParser extends RubyObject {
         byte[] data = fetchBytes(buf, pos, len);
         ThreadContext context = headers.getRuntime().getCurrentContext();
         IRubyObject key, val;
+        int new_field = 0;
 
         if (_current_header != null) {
+          new_field = 1;
           _last_header = _current_header;
           _current_header = null;
+        }
 
-          key = (IRubyObject)runtime.newString(_last_header);
+        key = (IRubyObject)runtime.newString(_last_header);
+        val = headers.op_aref(context, key);
+
+        if (new_field == 1) {
+          if (val.isNil()) {
+            if (header_value_type == runtime.newSymbol("arrays")) {
+              headers.op_aset(context, key, RubyArray.newArrayLight(runtime, runtime.newString("")));
+            } else {
+              headers.op_aset(context, key, runtime.newString(""));
+            }
+          } else {
+            if (header_value_type == runtime.newSymbol("mixed")) {
+              if (val instanceof RubyString) {
+                headers.op_aset(context, key, RubyArray.newArrayLight(runtime, val, runtime.newString("")));
+              } else {
+                ((RubyArray)val).add(runtime.newString(""));
+              }
+            } else if (header_value_type == runtime.newSymbol("arrays")) {
+              ((RubyArray)val).add(runtime.newString(""));
+            } else {
+              ((RubyString)val).cat(", ".getBytes());
+            }
+          }
           val = headers.op_aref(context, key);
+        }
 
-          if (!val.isNil())
-            ((RubyString)val).cat(", ".getBytes());
-        } else {
-          key = (IRubyObject)runtime.newString(_last_header);
-          val = headers.op_aref(context, key);
+        if (val instanceof RubyArray) {
+          val = ((RubyArray)val).entry(-1);
         }
 
-        if (val.isNil())
-          headers.op_aset(context, key, runtime.newString(new String(data)));
-        else
-          ((RubyString)val).cat(data);
+        ((RubyString)val).cat(data);
 
         return 0;
       }
@@ -163,6 +189,8 @@ public class RubyHttpParser extends RubyObject {
         queryString = runtime.newString("");
         fragment = runtime.newString("");
 
+        upgradeData = runtime.newString("");
+
         IRubyObject ret = runtime.getNil();
 
         if (callback_object != null) {
@@ -256,6 +284,8 @@ public class RubyHttpParser extends RubyObject {
     this.requestPath = runtime.getNil();
     this.queryString = runtime.getNil();
     this.fragment = runtime.getNil();
+
+    this.upgradeData = runtime.getNil();
   }
 
   @JRubyMethod(name = "initialize")
@@ -269,6 +299,12 @@ public class RubyHttpParser extends RubyObject {
     return initialize();
   }
 
+  @JRubyMethod(name = "initialize")
+  public IRubyObject initialize(IRubyObject arg, IRubyObject arg2) {
+    header_value_type = arg2;
+    return initialize(arg);
+  }
+
   @JRubyMethod(name = "on_message_begin=")
   public IRubyObject set_on_message_begin(IRubyObject cb) {
     on_message_begin = cb;
@@ -308,7 +344,9 @@ public class RubyHttpParser extends RubyObject {
     }
 
     if (parser.getUpgrade()) {
-      // upgrade request
+      byte[] upData = fetchBytes(buf, buf.position(), buf.limit() - buf.position());
+      ((RubyString)upgradeData).concat(runtime.newString(new String(upData)));
+
     } else if (buf.hasRemaining()) {
       if (!stopped)
         throw new RaiseException(runtime, eParserError, "Could not parse data entirely", true);
@@ -394,6 +432,25 @@ public class RubyHttpParser extends RubyObject {
     return fragment == null ? runtime.getNil() : fragment;
   }
 
+  @JRubyMethod(name = "header_value_type")
+  public IRubyObject getHeaderValueType() {
+    return header_value_type == null ? runtime.getNil() : header_value_type;
+  }
+
+  @JRubyMethod(name = "header_value_type=")
+  public IRubyObject set_header_value_type(IRubyObject val) {
+    if (val != runtime.newSymbol("mixed") && val != runtime.newSymbol("arrays") && val != runtime.newSymbol("strings")) {
+      throw runtime.newArgumentError("Invalid header value type");
+    }
+    header_value_type = val;
+    return val;
+  }
+
+  @JRubyMethod(name = "upgrade_data")
+  public IRubyObject upgradeData() {
+    return upgradeData == null ? runtime.getNil() : upgradeData;
+  }
+
   @JRubyMethod(name = "reset!")
   public IRubyObject reset() {
     init();

data/ext/ruby_http_parser/ruby_http_parser.c

@@ -23,6 +23,8 @@ typedef struct ParserWrapper {
 
   VALUE headers;
 
+  VALUE upgrade_data;
+
   VALUE on_message_begin;
   VALUE on_headers_complete;
   VALUE on_body;
@@ -32,6 +34,8 @@ typedef struct ParserWrapper {
   VALUE stopped;
   VALUE completed;
 
+  VALUE header_value_type;
+
   VALUE last_field_name;
   VALUE curr_field_name;
 
@@ -49,6 +53,8 @@ void ParserWrapper_init(ParserWrapper *wrapper) {
   wrapper->query_string = Qnil;
   wrapper->fragment = Qnil;
 
+  wrapper->upgrade_data = Qnil;
+
   wrapper->headers = Qnil;
   wrapper->completed = Qfalse;
 
@@ -63,6 +69,7 @@ void ParserWrapper_mark(void *data) {
     rb_gc_mark_maybe(wrapper->request_path);
     rb_gc_mark_maybe(wrapper->query_string);
     rb_gc_mark_maybe(wrapper->fragment);
+    rb_gc_mark_maybe(wrapper->upgrade_data);
     rb_gc_mark_maybe(wrapper->headers);
     rb_gc_mark_maybe(wrapper->on_message_begin);
     rb_gc_mark_maybe(wrapper->on_headers_complete);
@@ -93,6 +100,10 @@ static ID Ion_body;
 static ID Ion_message_complete;
 
 static VALUE Sstop;
+static VALUE Sreset;
+static VALUE Sarrays;
+static VALUE Sstrings;
+static VALUE Smixed;
 
 /** Callbacks **/
 
@@ -104,6 +115,7 @@ int on_message_begin(ryah_http_parser *parser) {
   wrapper->query_string = rb_str_new2("");
   wrapper->fragment = rb_str_new2("");
   wrapper->headers = rb_hash_new();
+  wrapper->upgrade_data = rb_str_new2("");
 
   VALUE ret = Qnil;
 
@@ -161,17 +173,45 @@ int on_header_field(ryah_http_parser *parser, const char *at, size_t length) {
 int on_header_value(ryah_http_parser *parser, const char *at, size_t length) {
   GET_WRAPPER(wrapper, parser);
 
+  int new_field = 0;
+  VALUE current_value;
+
   if (wrapper->last_field_name == Qnil) {
+    new_field = 1;
     wrapper->last_field_name = wrapper->curr_field_name;
     wrapper->curr_field_name = Qnil;
+  }
 
-    VALUE val = rb_hash_aref(wrapper->headers, wrapper->last_field_name);
-    if (val != Qnil) {
-      rb_str_cat(val, ", ", 2);
+  current_value = rb_hash_aref(wrapper->headers, wrapper->last_field_name);
+
+  if (new_field == 1) {
+    if (current_value == Qnil) {
+      if (wrapper->header_value_type == Sarrays) {
+        rb_hash_aset(wrapper->headers, wrapper->last_field_name, rb_ary_new3(1, rb_str_new2("")));
+      } else {
+        rb_hash_aset(wrapper->headers, wrapper->last_field_name, rb_str_new2(""));
+      }
+    } else {
+      if (wrapper->header_value_type == Smixed) {
+        if (TYPE(current_value) == T_STRING) {
+          rb_hash_aset(wrapper->headers, wrapper->last_field_name, rb_ary_new3(2, current_value, rb_str_new2("")));
+        } else {
+          rb_ary_push(current_value, rb_str_new2(""));
+        }
+      } else if (wrapper->header_value_type == Sarrays) {
+        rb_ary_push(current_value, rb_str_new2(""));
+      } else {
+        rb_str_cat(current_value, ", ", 2);
+      }
     }
+    current_value = rb_hash_aref(wrapper->headers, wrapper->last_field_name);
   }
 
-  HASH_CAT(wrapper->headers, wrapper->last_field_name, at, length);
+  if (TYPE(current_value) == T_ARRAY) {
+    current_value = rb_ary_entry(current_value, -1);
+  }
+
+  rb_str_cat(current_value, at, length);
 
   return 0;
 }
@@ -190,6 +230,8 @@ int on_headers_complete(ryah_http_parser *parser) {
   if (ret == Sstop) {
     wrapper->stopped = Qtrue;
     return -1;
+  } else if (ret == Sreset){
+    return 1;
   } else {
     return 0;
   }
@@ -280,8 +322,16 @@ VALUE Parser_initialize(int argc, VALUE *argv, VALUE self) {
   ParserWrapper *wrapper = NULL;
   DATA_GET(self, ParserWrapper, wrapper);
 
-  if (argc == 1)
+  wrapper->header_value_type = rb_iv_get(CLASS_OF(self), "@default_header_value_type");
+
+  if (argc == 1) {
+    wrapper->callback_object = argv[0];
+  }
+
+  if (argc == 2) {
     wrapper->callback_object = argv[0];
+    wrapper->header_value_type = argv[1];
+  }
 
   return self;
 }
@@ -299,8 +349,9 @@ VALUE Parser_execute(VALUE self, VALUE data) {
   size_t nparsed = ryah_http_parser_execute(&wrapper->parser, &settings, ptr, len);
 
   if (wrapper->parser.upgrade) {
-    // upgrade request
-  } else if (nparsed != len) {
+    rb_str_cat(wrapper->upgrade_data, ptr + nparsed + 1, len - nparsed - 1);
+
+  } else if (nparsed != (size_t)len) {
     if (!RTEST(wrapper->stopped) && !RTEST(wrapper->completed))
       rb_raise(eParserError, "Could not parse data entirely");
     else
@@ -353,7 +404,7 @@ VALUE Parser_upgrade_p(VALUE self) {
   ParserWrapper *wrapper = NULL;
   DATA_GET(self, ParserWrapper, wrapper);
 
-  return wrapper->parser.upgrade == 1 ? Qtrue : Qfalse;
+  return wrapper->parser.upgrade ? Qtrue : Qfalse;
 }
 
 VALUE Parser_http_version(VALUE self) {
@@ -418,6 +469,19 @@ DEFINE_GETTER(request_path);
 DEFINE_GETTER(query_string);
 DEFINE_GETTER(fragment);
 DEFINE_GETTER(headers);
+DEFINE_GETTER(upgrade_data);
+DEFINE_GETTER(header_value_type);
+
+VALUE Parser_set_header_value_type(VALUE self, VALUE val) {
+  if (val != Sarrays && val != Sstrings && val != Smixed) {
+    rb_raise(rb_eArgError, "Invalid header value type");
+  }
+
+  ParserWrapper *wrapper = NULL;
+  DATA_GET(self, ParserWrapper, wrapper);
+  wrapper->header_value_type = val;
+  return wrapper->header_value_type;
+}
 
 VALUE Parser_reset(VALUE self) {
   ParserWrapper *wrapper = NULL;
@@ -441,6 +505,11 @@ void Init_ruby_http_parser() {
   Ion_body = rb_intern("on_body");
   Ion_message_complete = rb_intern("on_message_complete");
   Sstop = ID2SYM(rb_intern("stop"));
+  Sreset = ID2SYM(rb_intern("reset"));
+
+  Sarrays = ID2SYM(rb_intern("arrays"));
+  Sstrings = ID2SYM(rb_intern("strings"));
+  Smixed = ID2SYM(rb_intern("mixed"));
 
   rb_define_alloc_func(cParser, Parser_alloc);
   rb_define_alloc_func(cRequestParser, RequestParser_alloc);
@@ -469,6 +538,9 @@ void Init_ruby_http_parser() {
   rb_define_method(cParser, "query_string", Parser_query_string, 0);
   rb_define_method(cParser, "fragment", Parser_fragment, 0);
   rb_define_method(cParser, "headers", Parser_headers, 0);
+  rb_define_method(cParser, "upgrade_data", Parser_upgrade_data, 0);
+  rb_define_method(cParser, "header_value_type", Parser_header_value_type, 0);
+  rb_define_method(cParser, "header_value_type=", Parser_set_header_value_type, 1);
 
   rb_define_method(cParser, "reset!", Parser_reset, 0);
 }

data/ext/ruby_http_parser/vendor/http-parser/LICENSE-MIT

@@ -1,4 +1,4 @@
-Copyright 2009,2010 Ryan Dahl <ry@tinyclouds.org>
+Copyright Joyent, Inc. and other Node contributors. All rights reserved.
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to

data/ext/ruby_http_parser/vendor/http-parser/http_parser.c

@@ -1,4 +1,4 @@
-/* Copyright 2009,2010 Ryan Dahl <ry@tinyclouds.org>
+/* Copyright Joyent, Inc. and other Node contributors. All rights reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a copy
  * of this software and associated documentation files (the "Software"), to
@@ -186,7 +186,28 @@ static const uint8_t normal_url_char[256] = {
 /* 112  p   113  q   114  r   115  s   116  t   117  u   118  v   119  w  */
         1,       1,       1,       1,       1,       1,       1,       1,
 /* 120  x   121  y   122  z   123  {   124  |   125  }   126  ~   127 del */
-        1,       1,       1,       1,       1,       1,       1,       0 };
+        1,       1,       1,       1,       1,       1,       1,       0,
+
+/* Remainder of non-ASCII range are accepted as-is to support implicitly UTF-8
+   encoded paths. This is out of spec, but clients generate this and most other
+   HTTP servers support it. We should, too. */
+
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1,
+        1,       1,       1,       1,       1,       1,       1,       1 };
 
 
 enum state
@@ -240,15 +261,17 @@ enum state
 
   , s_header_almost_done
 
+  , s_chunk_size_start
+  , s_chunk_size
+  , s_chunk_parameters
+  , s_chunk_size_almost_done
+  
   , s_headers_almost_done
   /* Important: 's_headers_almost_done' must be the last 'header' state. All
    * states beyond this must be 'body' states. It is used for overflow
    * checking. See the PARSING_HEADER() macro.
    */
-  , s_chunk_size_start
-  , s_chunk_size
-  , s_chunk_size_almost_done
-  , s_chunk_parameters
+
   , s_chunk_data
   , s_chunk_data_almost_done
   , s_chunk_data_done
@@ -258,7 +281,7 @@ enum state
   };
 
 
-#define PARSING_HEADER(state) (state <= s_headers_almost_done && 0 == (parser->flags & F_TRAILING))
+#define PARSING_HEADER(state) (state <= s_headers_almost_done)
 
 
 enum header_states
@@ -331,10 +354,20 @@ size_t http_parser_execute (http_parser *parser,
   uint64_t nread = parser->nread;
 
   if (len == 0) {
-    if (state == s_body_identity_eof) {
-      CALLBACK2(message_complete);
+    switch (state) {
+      case s_body_identity_eof:
+        CALLBACK2(message_complete);
+        return 0;
+
+      case s_dead:
+      case s_start_req_or_res:
+      case s_start_res:
+      case s_start_req:
+        return 0;
+
+      default:
+        return 1; // error
     }
-    return 0;
   }
 
   /* technically we could combine all of these (except for url_mark) into one
@@ -707,6 +740,9 @@ size_t http_parser_execute (http_parser *parser,
             CALLBACK(url);
             state = s_req_http_start;
             break;
+          case '?':
+            state = s_req_query_string_start;
+            break;
           default:
             goto error;
         }
@@ -729,6 +765,9 @@ size_t http_parser_execute (http_parser *parser,
             CALLBACK(url);
             state = s_req_http_start;
             break;
+          case '?':
+            state = s_req_query_string_start;
+            break;
           default:
             goto error;
         }
@@ -1448,6 +1487,7 @@ size_t http_parser_execute (http_parser *parser,
 
       case s_chunk_size_start:
       {
+        assert(nread == 1);
         assert(parser->flags & F_CHUNKED);
 
         c = unhex[(unsigned char)ch];
@@ -1497,6 +1537,8 @@ size_t http_parser_execute (http_parser *parser,
         assert(parser->flags & F_CHUNKED);
         STRICT_CHECK(ch != LF);
 
+        nread = 0;
+
         if (parser->content_length == 0) {
           parser->flags |= F_TRAILING;
           state = s_header_field_start;

data/ext/ruby_http_parser/vendor/http-parser/http_parser.h

@@ -1,4 +1,4 @@
-/* Copyright 2009,2010 Ryan Dahl <ry@tinyclouds.org>
+/* Copyright Joyent, Inc. and other Node contributors. All rights reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a copy
  * of this software and associated documentation files (the "Software"), to
@@ -24,6 +24,8 @@
 extern "C" {
 #endif
 
+#define HTTP_PARSER_VERSION_MAJOR 1
+#define HTTP_PARSER_VERSION_MINOR 0
 
 #include <sys/types.h>
 #if defined(_WIN32) && !defined(__MINGW32__)

data/ext/ruby_http_parser/vendor/http-parser/test.c

@@ -1,4 +1,4 @@
-/* Copyright 2009,2010 Ryan Dahl <ry@tinyclouds.org>
+/* Copyright Joyent, Inc. and other Node contributors. All rights reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a copy
  * of this software and associated documentation files (the "Software"), to
@@ -557,7 +557,7 @@ const struct message requests[] =
   ,.body= ""
   }
 
-#define MSEARCH_REQ 19
+#define MSEARCH_REQ 20
 , {.name= "m-search request"
   ,.type= HTTP_REQUEST
   ,.raw= "M-SEARCH * HTTP/1.1\r\n"
@@ -582,6 +582,84 @@ const struct message requests[] =
   ,.body= ""
   }
 
+#define UTF8_PATH_REQ 21
+, {.name= "utf-8 path request"
+  ,.type= HTTP_REQUEST
+  ,.raw= "GET /δ¶/δt/pope?q=1#narf HTTP/1.1\r\n"
+         "Host: github.com\r\n"
+         "\r\n"
+  ,.should_keep_alive= TRUE
+  ,.message_complete_on_eof= FALSE
+  ,.http_major= 1
+  ,.http_minor= 1
+  ,.method= HTTP_GET
+  ,.query_string= "q=1"
+  ,.fragment= "narf"
+  ,.request_path= "/δ¶/δt/pope"
+  ,.request_url= "/δ¶/δt/pope?q=1#narf"
+  ,.num_headers= 1
+  ,.headers= { {"Host", "github.com" }
+             }
+  ,.body= ""
+  }
+
+#define QUERY_TERMINATED_HOST 22
+, {.name= "host terminated by a query string"
+  ,.type= HTTP_REQUEST
+  ,.raw= "GET http://hypnotoad.org?hail=all HTTP/1.1\r\n"
+         "\r\n"
+  ,.should_keep_alive= TRUE
+  ,.message_complete_on_eof= FALSE
+  ,.http_major= 1
+  ,.http_minor= 1
+  ,.method= HTTP_GET
+  ,.query_string= "hail=all"
+  ,.fragment= ""
+  ,.request_path= ""
+  ,.request_url= "http://hypnotoad.org?hail=all"
+  ,.num_headers= 0
+  ,.headers= { }
+  ,.body= ""
+  }
+
+#define QUERY_TERMINATED_HOSTPORT 23
+, {.name= "host:port terminated by a query string"
+  ,.type= HTTP_REQUEST
+  ,.raw= "GET http://hypnotoad.org:1234?hail=all HTTP/1.1\r\n"
+         "\r\n"
+  ,.should_keep_alive= TRUE
+  ,.message_complete_on_eof= FALSE
+  ,.http_major= 1
+  ,.http_minor= 1
+  ,.method= HTTP_GET
+  ,.query_string= "hail=all"
+  ,.fragment= ""
+  ,.request_path= ""
+  ,.request_url= "http://hypnotoad.org:1234?hail=all"
+  ,.num_headers= 0
+  ,.headers= { }
+  ,.body= ""
+  }
+
+#define SPACE_TERMINATED_HOSTPORT 24
+, {.name= "host:port terminated by a space"
+  ,.type= HTTP_REQUEST
+  ,.raw= "GET http://hypnotoad.org:1234 HTTP/1.1\r\n"
+         "\r\n"
+  ,.should_keep_alive= TRUE
+  ,.message_complete_on_eof= FALSE
+  ,.http_major= 1
+  ,.http_minor= 1
+  ,.method= HTTP_GET
+  ,.query_string= ""
+  ,.fragment= ""
+  ,.request_path= ""
+  ,.request_url= "http://hypnotoad.org:1234"
+  ,.num_headers= 0
+  ,.headers= { }
+  ,.body= ""
+  }
+
 , {.name= NULL } /* sentinel */
 };
 
@@ -1735,14 +1813,22 @@ main (void)
 
   /// REQUESTS
 
-
   test_simple("hello world", 0);
   test_simple("GET / HTP/1.1\r\n\r\n", 0);
 
+
   test_simple("ASDF / HTTP/1.1\r\n\r\n", 0);
   test_simple("PROPPATCHA / HTTP/1.1\r\n\r\n", 0);
   test_simple("GETA / HTTP/1.1\r\n\r\n", 0);
 
+  // Well-formed but incomplete
+  test_simple("GET / HTTP/1.1\r\n"
+              "Content-Type: text/plain\r\n"
+              "Content-Length: 6\r\n"
+              "\r\n"
+              "fooba",
+              0);
+
   static const char *all_methods[] = {
     "DELETE",
     "GET",