http_parser.rb 0.5.3 → 0.6.0.beta.1

data/ext/ruby_http_parser/ruby_http_parser.c

@@ -17,9 +17,6 @@ typedef struct ParserWrapper {
   ryah_http_parser parser;
 
   VALUE request_url;
-  VALUE request_path;
-  VALUE query_string;
-  VALUE fragment;
 
   VALUE headers;
 
@@ -49,9 +46,6 @@ void ParserWrapper_init(ParserWrapper *wrapper) {
   wrapper->parser.http_minor = 0;
 
   wrapper->request_url = Qnil;
-  wrapper->request_path = Qnil;
-  wrapper->query_string = Qnil;
-  wrapper->fragment = Qnil;
 
   wrapper->upgrade_data = Qnil;
 
@@ -66,9 +60,6 @@ void ParserWrapper_mark(void *data) {
   if(data) {
     ParserWrapper *wrapper = (ParserWrapper *) data;
     rb_gc_mark_maybe(wrapper->request_url);
-    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);
@@ -111,9 +102,6 @@ int on_message_begin(ryah_http_parser *parser) {
   GET_WRAPPER(wrapper, parser);
 
   wrapper->request_url = rb_str_new2("");
-  wrapper->request_path = rb_str_new2("");
-  wrapper->query_string = rb_str_new2("");
-  wrapper->fragment = rb_str_new2("");
   wrapper->headers = rb_hash_new();
   wrapper->upgrade_data = rb_str_new2("");
 
@@ -139,24 +127,6 @@ int on_url(ryah_http_parser *parser, const char *at, size_t length) {
   return 0;
 }
 
-int on_path(ryah_http_parser *parser, const char *at, size_t length) {
-  GET_WRAPPER(wrapper, parser);
-  rb_str_cat(wrapper->request_path, at, length);
-  return 0;
-}
-
-int on_query_string(ryah_http_parser *parser, const char *at, size_t length) {
-  GET_WRAPPER(wrapper, parser);
-  rb_str_cat(wrapper->query_string, at, length);
-  return 0;
-}
-
-int on_fragment(ryah_http_parser *parser, const char *at, size_t length) {
-  GET_WRAPPER(wrapper, parser);
-  rb_str_cat(wrapper->fragment, at, length);
-  return 0;
-}
-
 int on_header_field(ryah_http_parser *parser, const char *at, size_t length) {
   GET_WRAPPER(wrapper, parser);
 
@@ -278,10 +248,7 @@ int on_message_complete(ryah_http_parser *parser) {
 
 static ryah_http_parser_settings settings = {
   .on_message_begin = on_message_begin,
-  .on_path = on_path,
-  .on_query_string = on_query_string,
   .on_url = on_url,
-  .on_fragment = on_fragment,
   .on_header_field = on_header_field,
   .on_header_value = on_header_value,
   .on_headers_complete = on_headers_complete,
@@ -318,6 +285,10 @@ VALUE ResponseParser_alloc(VALUE klass) {
   return Parser_alloc_by_type(klass, HTTP_RESPONSE);
 }
 
+VALUE Parser_strict_p(VALUE klass) {
+  return HTTP_PARSER_STRICT == 1 ? Qtrue : Qfalse;
+}
+
 VALUE Parser_initialize(int argc, VALUE *argv, VALUE self) {
   ParserWrapper *wrapper = NULL;
   DATA_GET(self, ParserWrapper, wrapper);
@@ -349,11 +320,14 @@ VALUE Parser_execute(VALUE self, VALUE data) {
   size_t nparsed = ryah_http_parser_execute(&wrapper->parser, &settings, ptr, len);
 
   if (wrapper->parser.upgrade) {
-    rb_str_cat(wrapper->upgrade_data, ptr + nparsed + 1, len - nparsed - 1);
+    if (RTEST(wrapper->stopped))
+      nparsed += 1;
+
+    rb_str_cat(wrapper->upgrade_data, ptr + nparsed, len - nparsed);
 
   } else if (nparsed != (size_t)len) {
     if (!RTEST(wrapper->stopped) && !RTEST(wrapper->completed))
-      rb_raise(eParserError, "Could not parse data entirely");
+      rb_raise(eParserError, "Could not parse data entirely (%zu != %zu)", nparsed, len);
     else
       nparsed += 1; // error states fail on the current character
   }
@@ -465,9 +439,6 @@ VALUE Parser_status_code(VALUE self) {
   }
 
 DEFINE_GETTER(request_url);
-DEFINE_GETTER(request_path);
-DEFINE_GETTER(query_string);
-DEFINE_GETTER(fragment);
 DEFINE_GETTER(headers);
 DEFINE_GETTER(upgrade_data);
 DEFINE_GETTER(header_value_type);
@@ -515,6 +486,7 @@ void Init_ruby_http_parser() {
   rb_define_alloc_func(cRequestParser, RequestParser_alloc);
   rb_define_alloc_func(cResponseParser, ResponseParser_alloc);
 
+  rb_define_singleton_method(cParser, "strict?", Parser_strict_p, 0);
   rb_define_method(cParser, "initialize", Parser_initialize, -1);
 
   rb_define_method(cParser, "on_message_begin=", Parser_set_on_message_begin, 1);
@@ -534,9 +506,6 @@ void Init_ruby_http_parser() {
   rb_define_method(cParser, "status_code", Parser_status_code, 0);
 
   rb_define_method(cParser, "request_url", Parser_request_url, 0);
-  rb_define_method(cParser, "request_path", Parser_request_path, 0);
-  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);

data/ext/ruby_http_parser/vendor/http-parser-java/AUTHORS

@@ -0,0 +1,32 @@
+# Authors ordered by first contribution.
+Ryan Dahl <ry@tinyclouds.org>
+Jeremy Hinegardner <jeremy@hinegardner.org>
+Sergey Shepelev <temotor@gmail.com>
+Joe Damato <ice799@gmail.com>
+tomika <tomika_nospam@freemail.hu>
+Phoenix Sol <phoenix@burninglabs.com>
+Cliff Frey <cliff@meraki.com>
+Ewen Cheslack-Postava <ewencp@cs.stanford.edu>
+Santiago Gala <sgala@apache.org>
+Tim Becker <tim.becker@syngenio.de>
+Jeff Terrace <jterrace@gmail.com>
+Ben Noordhuis <info@bnoordhuis.nl>
+Nathan Rajlich <nathan@tootallnate.net>
+Mark Nottingham <mnot@mnot.net>
+Aman Gupta <aman@tmm1.net>
+Tim Becker <tim.becker@kuriositaet.de>
+Sean Cunningham <sean.cunningham@mandiant.com>
+Peter Griess <pg@std.in>
+Salman Haq <salman.haq@asti-usa.com>
+Cliff Frey <clifffrey@gmail.com>
+Jon Kolb <jon@b0g.us>
+Fouad Mardini <f.mardini@gmail.com>
+Paul Querna <pquerna@apache.org>
+Felix Geisendörfer <felix@debuggable.com>
+koichik <koichik@improvement.jp>
+Andre Caron <andre.l.caron@gmail.com>
+Ivo Raisr <ivosh@ivosh.net>
+James McLaughlin <jamie@lacewing-project.org>
+David Gwynne <loki@animata.net>
+LE ROUX Thomas <thomas@procheo.fr>
+Randy Rizun <rrizun@ortivawireless.com>

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

@@ -23,7 +23,11 @@ IN THE SOFTWARE.
 This code mainly based on code with the following license:
 
 
-Copyright Joyent, Inc. and other Node contributors. All rights reserved.
+http_parser.c is based on src/http/ngx_http_parse.c from NGINX copyright
+Igor Sysoev.
+
+Additional changes are licensed under the same terms as NGINX and
+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-java/README.md

@@ -24,7 +24,7 @@ The parser extracts the following information from HTTP messages:
   * Response status code
   * Transfer-Encoding
   * HTTP version
-  * Request path, query string, fragment
+  * Request URL
   * Message body
 
 Building
@@ -49,3 +49,135 @@ Usage
   help or have suggestions, feel free to contact me at
   (tim.becker@kuriositaet.de).
 
+
+One `http_parser` object is used per TCP connection. Initialize the struct
+using `http_parser_init()` and set the callbacks. That might look something
+like this for a request parser:
+
+    http_parser_settings settings;
+    settings.on_path = my_path_callback;
+    settings.on_header_field = my_header_field_callback;
+    /* ... */
+
+    http_parser *parser = malloc(sizeof(http_parser));
+    http_parser_init(parser, HTTP_REQUEST);
+    parser->data = my_socket;
+
+When data is received on the socket execute the parser and check for errors.
+
+    size_t len = 80*1024, nparsed;
+    char buf[len];
+    ssize_t recved;
+
+    recved = recv(fd, buf, len, 0);
+
+    if (recved < 0) {
+      /* Handle error. */
+    }
+
+    /* Start up / continue the parser.
+     * Note we pass recved==0 to signal that EOF has been recieved.
+     */
+    nparsed = http_parser_execute(parser, &settings, buf, recved);
+
+    if (parser->upgrade) {
+      /* handle new protocol */
+    } else if (nparsed != recved) {
+      /* Handle error. Usually just close the connection. */
+    }
+
+HTTP needs to know where the end of the stream is. For example, sometimes
+servers send responses without Content-Length and expect the client to
+consume input (for the body) until EOF. To tell http_parser about EOF, give
+`0` as the forth parameter to `http_parser_execute()`. Callbacks and errors
+can still be encountered during an EOF, so one must still be prepared
+to receive them.
+
+Scalar valued message information such as `status_code`, `method`, and the
+HTTP version are stored in the parser structure. This data is only
+temporally stored in `http_parser` and gets reset on each new message. If
+this information is needed later, copy it out of the structure during the
+`headers_complete` callback.
+
+The parser decodes the transfer-encoding for both requests and responses
+transparently. That is, a chunked encoding is decoded before being sent to
+the on_body callback.
+
+
+The Special Problem of Upgrade
+------------------------------
+
+HTTP supports upgrading the connection to a different protocol. An
+increasingly common example of this is the Web Socket protocol which sends
+a request like
+
+        GET /demo HTTP/1.1
+        Upgrade: WebSocket
+        Connection: Upgrade
+        Host: example.com
+        Origin: http://example.com
+        WebSocket-Protocol: sample
+
+followed by non-HTTP data.
+
+(See http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-75 for more
+information the Web Socket protocol.)
+
+To support this, the parser will treat this as a normal HTTP message without a
+body. Issuing both on_headers_complete and on_message_complete callbacks. However
+http_parser_execute() will stop parsing at the end of the headers and return.
+
+The user is expected to check if `parser->upgrade` has been set to 1 after
+`http_parser_execute()` returns. Non-HTTP data begins at the buffer supplied
+offset by the return value of `http_parser_execute()`.
+
+
+Callbacks
+---------
+
+During the `http_parser_execute()` call, the callbacks set in
+`http_parser_settings` will be executed. The parser maintains state and
+never looks behind, so buffering the data is not necessary. If you need to
+save certain data for later usage, you can do that from the callbacks.
+
+There are two types of callbacks:
+
+* notification `typedef int (*http_cb) (http_parser*);`
+    Callbacks: on_message_begin, on_headers_complete, on_message_complete.
+* data `typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);`
+    Callbacks: (requests only) on_uri,
+               (common) on_header_field, on_header_value, on_body;
+
+Callbacks must return 0 on success. Returning a non-zero value indicates
+error to the parser, making it exit immediately.
+
+In case you parse HTTP message in chunks (i.e. `read()` request line
+from socket, parse, read half headers, parse, etc) your data callbacks
+may be called more than once. Http-parser guarantees that data pointer is only
+valid for the lifetime of callback. You can also `read()` into a heap allocated
+buffer to avoid copying memory around if this fits your application.
+
+Reading headers may be a tricky task if you read/parse headers partially.
+Basically, you need to remember whether last header callback was field or value
+and apply following logic:
+
+    (on_header_field and on_header_value shortened to on_h_*)
+     ------------------------ ------------ --------------------------------------------
+    | State (prev. callback) | Callback   | Description/action                         |
+     ------------------------ ------------ --------------------------------------------
+    | nothing (first call)   | on_h_field | Allocate new buffer and copy callback data |
+    |                        |            | into it                                    |
+     ------------------------ ------------ --------------------------------------------
+    | value                  | on_h_field | New header started.                        |
+    |                        |            | Copy current name,value buffers to headers |
+    |                        |            | list and allocate new buffer for new name  |
+     ------------------------ ------------ --------------------------------------------
+    | field                  | on_h_field | Previous name continues. Reallocate name   |
+    |                        |            | buffer and append callback data to it      |
+     ------------------------ ------------ --------------------------------------------
+    | field                  | on_h_value | Value for current header started. Allocate |
+    |                        |            | new buffer and copy callback data to it    |
+     ------------------------ ------------ --------------------------------------------
+    | value                  | on_h_value | Value continues. Reallocate value buffer   |
+    |                        |            | and append callback data to it             |
+     ------------------------ ------------ --------------------------------------------

data/ext/ruby_http_parser/vendor/http-parser-java/TODO

@@ -1,4 +1,10 @@
+decide how to handle errs per default:
+  - ry: "set state to dead", return `read`
+  - current: call on_error w/ details, if no on_error handler set,
+    throw Exception, else call on_error and behave like orig...
+
 some tests from test.c left to port
+  (scan ...)
 documentation
 
 hi level callback interface 

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

@@ -1,4 +1,7 @@
-/* Copyright Joyent, Inc. and other Node contributors. All rights reserved.
+/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev
+ *
+ * Additional changes are licensed under the same terms as NGINX and
+ * 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
@@ -18,48 +21,94 @@
  * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
  * IN THE SOFTWARE.
  */
-#include <http_parser.h>
+#include "http_parser.h"
 #include <assert.h>
 #include <stddef.h>
+#include <ctype.h>
+#include <stdlib.h>
+#include <string.h>
+#include <limits.h>
 
+#ifndef ULLONG_MAX
+# define ULLONG_MAX ((uint64_t) -1) /* 2^64-1 */
+#endif
 
 #ifndef MIN
 # define MIN(a,b) ((a) < (b) ? (a) : (b))
 #endif
 
 
-#define CALLBACK2(FOR)                                               \
+#if HTTP_PARSER_DEBUG
+#define SET_ERRNO(e)                                                 \
 do {                                                                 \
-  if (settings->on_##FOR) {                                          \
-    if (0 != settings->on_##FOR(parser)) return (p - data);          \
-  }                                                                  \
+  parser->http_errno = (e);                                          \
+  parser->error_lineno = __LINE__;                                   \
 } while (0)
+#else
+#define SET_ERRNO(e)                                                 \
+do {                                                                 \
+  parser->http_errno = (e);                                          \
+} while(0)
+#endif
 
 
-#define MARK(FOR)                                                    \
+/* Run the notify callback FOR, returning ER if it fails */
+#define CALLBACK_NOTIFY_(FOR, ER)                                    \
 do {                                                                 \
-  FOR##_mark = p;                                                    \
+  assert(HTTP_PARSER_ERRNO(parser) == HPE_OK);                       \
+                                                                     \
+  if (settings->on_##FOR) {                                          \
+    if (0 != settings->on_##FOR(parser)) {                           \
+      SET_ERRNO(HPE_CB_##FOR);                                       \
+    }                                                                \
+                                                                     \
+    /* We either errored above or got paused; get out */             \
+    if (HTTP_PARSER_ERRNO(parser) != HPE_OK) {                       \
+      return (ER);                                                   \
+    }                                                                \
+  }                                                                  \
 } while (0)
 
-#define CALLBACK_NOCLEAR(FOR)                                        \
+/* Run the notify callback FOR and consume the current byte */
+#define CALLBACK_NOTIFY(FOR)            CALLBACK_NOTIFY_(FOR, p - data + 1)
+
+/* Run the notify callback FOR and don't consume the current byte */
+#define CALLBACK_NOTIFY_NOADVANCE(FOR)  CALLBACK_NOTIFY_(FOR, p - data)
+
+/* Run data callback FOR with LEN bytes, returning ER if it fails */
+#define CALLBACK_DATA_(FOR, LEN, ER)                                 \
 do {                                                                 \
+  assert(HTTP_PARSER_ERRNO(parser) == HPE_OK);                       \
+                                                                     \
   if (FOR##_mark) {                                                  \
     if (settings->on_##FOR) {                                        \
-      if (0 != settings->on_##FOR(parser,                            \
-                                 FOR##_mark,                         \
-                                 p - FOR##_mark))                    \
-      {                                                              \
-        return (p - data);                                           \
+      if (0 != settings->on_##FOR(parser, FOR##_mark, (LEN))) {      \
+        SET_ERRNO(HPE_CB_##FOR);                                     \
+      }                                                              \
+                                                                     \
+      /* We either errored above or got paused; get out */           \
+      if (HTTP_PARSER_ERRNO(parser) != HPE_OK) {                     \
+        return (ER);                                                 \
       }                                                              \
     }                                                                \
+    FOR##_mark = NULL;                                               \
   }                                                                  \
 } while (0)
+  
+/* Run the data callback FOR and consume the current byte */
+#define CALLBACK_DATA(FOR)                                           \
+    CALLBACK_DATA_(FOR, p - FOR##_mark, p - data + 1)
 
+/* Run the data callback FOR and don't consume the current byte */
+#define CALLBACK_DATA_NOADVANCE(FOR)                                 \
+    CALLBACK_DATA_(FOR, p - FOR##_mark, p - data)
 
-#define CALLBACK(FOR)                                                \
+/* Set the mark FOR; non-destructive if mark is already set */
+#define MARK(FOR)                                                    \
 do {                                                                 \
-  CALLBACK_NOCLEAR(FOR);                                             \
-  FOR##_mark = NULL;                                                 \
+  if (!FOR##_mark) {                                                 \
+    FOR##_mark = p;                                                  \
+  }                                                                  \
 } while (0)
 
 
@@ -97,6 +146,8 @@ static const char *method_strings[] =
   , "NOTIFY"
   , "SUBSCRIBE"
   , "UNSUBSCRIBE"
+  , "PATCH"
+  , "PURGE"
   };
 
 
@@ -117,9 +168,9 @@ static const char tokens[256] = {
 /*  24 can   25 em    26 sub   27 esc   28 fs    29 gs    30 rs    31 us  */
         0,       0,       0,       0,       0,       0,       0,       0,
 /*  32 sp    33  !    34  "    35  #    36  $    37  %    38  &    39  '  */
-       ' ',      '!',     '"',     '#',     '$',     '%',     '&',    '\'',
+        0,      '!',      0,      '#',     '$',     '%',     '&',    '\'',
 /*  40  (    41  )    42  *    43  +    44  ,    45  -    46  .    47  /  */
-        0,       0,      '*',     '+',      0,      '-',     '.',     '/',
+        0,       0,      '*',     '+',      0,      '-',     '.',      0,
 /*  48  0    49  1    50  2    51  3    52  4    53  5    54  6    55  7  */
        '0',     '1',     '2',     '3',     '4',     '5',     '6',     '7',
 /*  56  8    57  9    58  :    59  ;    60  <    61  =    62  >    63  ?  */
@@ -139,7 +190,7 @@ static const char tokens[256] = {
 /* 112  p   113  q   114  r   115  s   116  t   117  u   118  v   119  w  */
        'p',     'q',     'r',     's',     't',     'u',     'v',     'w',
 /* 120  x   121  y   122  z   123  {   124  |   125  }   126  ~   127 del */
-       'x',     'y',     'z',      0,      '|',     '}',     '~',       0 };
+       'x',     'y',     'z',      0,      '|',      0,      '~',       0 };
 
 
 static const int8_t unhex[256] =
@@ -186,28 +237,7 @@ 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,
-
-/* 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 };
+        1,       1,       1,       1,       1,       1,       1,       0, };
 
 
 enum state
@@ -236,7 +266,12 @@ enum state
   , s_req_schema
   , s_req_schema_slash
   , s_req_schema_slash_slash
+  , s_req_host_start
+  , s_req_host_v6_start
+  , s_req_host_v6
+  , s_req_host_v6_end
   , s_req_host
+  , s_req_port_start
   , s_req_port
   , s_req_path
   , s_req_query_string_start
@@ -258,6 +293,7 @@ enum state
   , s_header_field
   , s_header_value_start
   , s_header_value
+  , s_header_value_lws
 
   , s_header_almost_done
 
@@ -265,9 +301,11 @@ enum state
   , 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
+  , s_headers_done
+
+  /* Important: 's_headers_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.
    */
@@ -278,10 +316,12 @@ enum state
 
   , s_body_identity
   , s_body_identity_eof
+
+  , s_message_done
   };
 
 
-#define PARSING_HEADER(state) (state <= s_headers_almost_done)
+#define PARSING_HEADER(state) (state <= s_headers_done)
 
 
 enum header_states
@@ -311,27 +351,39 @@ enum header_states
   };
 
 
-enum flags
-  { F_CHUNKED               = 1 << 0
-  , F_CONNECTION_KEEP_ALIVE = 1 << 1
-  , F_CONNECTION_CLOSE      = 1 << 2
-  , F_TRAILING              = 1 << 3
-  , F_UPGRADE               = 1 << 4
-  , F_SKIPBODY              = 1 << 5
-  };
-
+/* Macros for character classes; depends on strict-mode  */
+#define CR                  '\r'
+#define LF                  '\n'
+#define LOWER(c)            (unsigned char)(c | 0x20)
+#define IS_ALPHA(c)         (LOWER(c) >= 'a' && LOWER(c) <= 'z')
+#define IS_NUM(c)           ((c) >= '0' && (c) <= '9')
+#define IS_ALPHANUM(c)      (IS_ALPHA(c) || IS_NUM(c))
+#define IS_HEX(c)           (IS_NUM(c) || (LOWER(c) >= 'a' && LOWER(c) <= 'f'))
 
-#define CR '\r'
-#define LF '\n'
-#define LOWER(c) (unsigned char)(c | 0x20)
-#define TOKEN(c) tokens[(unsigned char)c]
+#if HTTP_PARSER_STRICT
+#define TOKEN(c)            (tokens[(unsigned char)c])
+#define IS_URL_CHAR(c)      (normal_url_char[(unsigned char) (c)])
+#define IS_HOST_CHAR(c)     (IS_ALPHANUM(c) || (c) == '.' || (c) == '-')
+#else
+#define TOKEN(c)            ((c == ' ') ? ' ' : tokens[(unsigned char)c])
+#define IS_URL_CHAR(c)                                                         \
+  (normal_url_char[(unsigned char) (c)] || ((c) & 0x80))
+#define IS_HOST_CHAR(c)                                                        \
+  (IS_ALPHANUM(c) || (c) == '.' || (c) == '-' || (c) == '_')
+#endif
 
 
 #define start_state (parser->type == HTTP_REQUEST ? s_start_req : s_start_res)
 
 
 #if HTTP_PARSER_STRICT
-# define STRICT_CHECK(cond) if (cond) goto error
+# define STRICT_CHECK(cond)                                          \
+do {                                                                 \
+  if (cond) {                                                        \
+    SET_ERRNO(HPE_STRICT);                                           \
+    goto error;                                                      \
+  }                                                                  \
+} while (0)
 # define NEW_MESSAGE() (http_should_keep_alive(parser) ? start_state : s_dead)
 #else
 # define STRICT_CHECK(cond)
@@ -339,24 +391,228 @@ enum flags
 #endif
 
 
+/* Map errno values to strings for human-readable output */
+#define HTTP_STRERROR_GEN(n, s) { "HPE_" #n, s },
+static struct {
+  const char *name;
+  const char *description;
+} http_strerror_tab[] = {
+  HTTP_ERRNO_MAP(HTTP_STRERROR_GEN)
+};
+#undef HTTP_STRERROR_GEN
+
+int http_message_needs_eof(http_parser *parser);
+
+/* Our URL parser.
+ *
+ * This is designed to be shared by http_parser_execute() for URL validation,
+ * hence it has a state transition + byte-for-byte interface. In addition, it
+ * is meant to be embedded in http_parser_parse_url(), which does the dirty
+ * work of turning state transitions URL components for its API.
+ *
+ * This function should only be invoked with non-space characters. It is
+ * assumed that the caller cares about (and can detect) the transition between
+ * URL and non-URL states by looking for these.
+ */
+static enum state
+parse_url_char(enum state s, const char ch)
+{
+  assert(!isspace(ch));
+
+  switch (s) {
+    case s_req_spaces_before_url:
+      /* Proxied requests are followed by scheme of an absolute URI (alpha).
+       * All methods except CONNECT are followed by '/' or '*'.
+       */
+
+      if (ch == '/' || ch == '*') {
+        return s_req_path;
+      }
+
+      if (IS_ALPHA(ch)) {
+        return s_req_schema;
+      }
+
+      break;
+
+    case s_req_schema:
+      if (IS_ALPHA(ch)) {
+        return s;
+      }
+
+      if (ch == ':') {
+        return s_req_schema_slash;
+      }
+
+      break;
+
+    case s_req_schema_slash:
+      if (ch == '/') {
+        return s_req_schema_slash_slash;
+      }
+
+      break;
+
+    case s_req_schema_slash_slash:
+      if (ch == '/') {
+        return s_req_host_start;
+      }
+
+      break;
+
+    case s_req_host_start:
+      if (ch == '[') {
+        return s_req_host_v6_start;
+      }
+
+      if (IS_HOST_CHAR(ch)) {
+        return s_req_host;
+      }
+
+      break;
+
+    case s_req_host:
+      if (IS_HOST_CHAR(ch)) {
+        return s_req_host;
+      }
+
+      /* FALLTHROUGH */
+    case s_req_host_v6_end:
+      switch (ch) {
+        case ':':
+          return s_req_port_start;
+
+        case '/':
+          return s_req_path;
+
+        case '?':
+          return s_req_query_string_start;
+      }
+
+      break;
+
+    case s_req_host_v6:
+      if (ch == ']') {
+        return s_req_host_v6_end;
+      }
+
+      /* FALLTHROUGH */
+    case s_req_host_v6_start:
+      if (IS_HEX(ch) || ch == ':') {
+        return s_req_host_v6;
+      }
+      break;
+
+    case s_req_port:
+      switch (ch) {
+        case '/':
+          return s_req_path;
+
+        case '?':
+          return s_req_query_string_start;
+      }
+
+      /* FALLTHROUGH */
+    case s_req_port_start:
+      if (IS_NUM(ch)) {
+        return s_req_port;
+      }
+
+      break;
+
+    case s_req_path:
+      if (IS_URL_CHAR(ch)) {
+        return s;
+      }
+
+      switch (ch) {
+        case '?':
+          return s_req_query_string_start;
+
+        case '#':
+          return s_req_fragment_start;
+      }
+
+      break;
+
+    case s_req_query_string_start:
+    case s_req_query_string:
+      if (IS_URL_CHAR(ch)) {
+        return s_req_query_string;
+      }
+
+      switch (ch) {
+        case '?':
+          /* allow extra '?' in query string */
+          return s_req_query_string;
+
+        case '#':
+          return s_req_fragment_start;
+      }
+
+      break;
+
+    case s_req_fragment_start:
+      if (IS_URL_CHAR(ch)) {
+        return s_req_fragment;
+      }
+
+      switch (ch) {
+        case '?':
+          return s_req_fragment;
+
+        case '#':
+          return s;
+      }
+
+      break;
+
+    case s_req_fragment:
+      if (IS_URL_CHAR(ch)) {
+        return s;
+      }
+
+      switch (ch) {
+        case '?':
+        case '#':
+          return s;
+      }
+
+      break;
+
+    default:
+      break;
+  }
+
+  /* We should never fall out of the switch above unless there's an error */
+  return s_dead;
+}
+
 size_t http_parser_execute (http_parser *parser,
                             const http_parser_settings *settings,
                             const char *data,
                             size_t len)
 {
   char c, ch;
-  const char *p = data, *pe;
-  int64_t to_read;
+  int8_t unhex_val;
+  const char *p = data;
+  const char *header_field_mark = 0;
+  const char *header_value_mark = 0;
+  const char *url_mark = 0;
+  const char *body_mark = 0;
 
-  enum state state = (enum state) parser->state;
-  enum header_states header_state = (enum header_states) parser->header_state;
-  uint64_t index = parser->index;
-  uint64_t nread = parser->nread;
+  /* We're in an error state. Don't bother doing anything. */
+  if (HTTP_PARSER_ERRNO(parser) != HPE_OK) {
+    return 0;
+  }
 
   if (len == 0) {
-    switch (state) {
+    switch (parser->state) {
       case s_body_identity_eof:
-        CALLBACK2(message_complete);
+        /* Use of CALLBACK_NOTIFY() here would erroneously return 1 byte read if
+         * we got paused.
+         */
+        CALLBACK_NOTIFY_NOADVANCE(message_complete);
         return 0;
 
       case s_dead:
@@ -366,52 +622,59 @@ size_t http_parser_execute (http_parser *parser,
         return 0;
 
       default:
-        return 1; // error
+        SET_ERRNO(HPE_INVALID_EOF_STATE);
+        return 1;
     }
   }
 
-  /* technically we could combine all of these (except for url_mark) into one
-     variable, saving stack space, but it seems more clear to have them
-     separated. */
-  const char *header_field_mark = 0;
-  const char *header_value_mark = 0;
-  const char *fragment_mark = 0;
-  const char *query_string_mark = 0;
-  const char *path_mark = 0;
-  const char *url_mark = 0;
 
-  if (state == s_header_field)
+  if (parser->state == s_header_field)
     header_field_mark = data;
-  if (state == s_header_value)
+  if (parser->state == s_header_value)
     header_value_mark = data;
-  if (state == s_req_fragment)
-    fragment_mark = data;
-  if (state == s_req_query_string)
-    query_string_mark = data;
-  if (state == s_req_path)
-    path_mark = data;
-  if (state == s_req_path || state == s_req_schema || state == s_req_schema_slash
-      || state == s_req_schema_slash_slash || state == s_req_port
-      || state == s_req_query_string_start || state == s_req_query_string
-      || state == s_req_host
-      || state == s_req_fragment_start || state == s_req_fragment)
+  switch (parser->state) {
+  case s_req_path:
+  case s_req_schema:
+  case s_req_schema_slash:
+  case s_req_schema_slash_slash:
+  case s_req_host_start:
+  case s_req_host_v6_start:
+  case s_req_host_v6:
+  case s_req_host_v6_end:
+  case s_req_host:
+  case s_req_port_start:
+  case s_req_port:
+  case s_req_query_string_start:
+  case s_req_query_string:
+  case s_req_fragment_start:
+  case s_req_fragment:
     url_mark = data;
+    break;
+  }
 
-  for (p=data, pe=data+len; p != pe; p++) {
+  for (p=data; p != data + len; p++) {
     ch = *p;
 
-    if (PARSING_HEADER(state)) {
-      ++nread;
+    if (PARSING_HEADER(parser->state)) {
+      ++parser->nread;
       /* Buffer overflow attack */
-      if (nread > HTTP_MAX_HEADER_SIZE) goto error;
+      if (parser->nread > HTTP_MAX_HEADER_SIZE) {
+        SET_ERRNO(HPE_HEADER_OVERFLOW);
+        goto error;
+      }
     }
 
-    switch (state) {
+    reexecute_byte:
+    switch (parser->state) {
 
       case s_dead:
         /* this state is used after a 'Connection: close' message
          * the parser will error out if it reads another message
          */
+        if (ch == CR || ch == LF)
+          break;
+
+        SET_ERRNO(HPE_CLOSED_CONNECTION);
         goto error;
 
       case s_start_req_or_res:
@@ -419,42 +682,46 @@ size_t http_parser_execute (http_parser *parser,
         if (ch == CR || ch == LF)
           break;
         parser->flags = 0;
-        parser->content_length = -1;
+        parser->content_length = ULLONG_MAX;
 
-        CALLBACK2(message_begin);
+        if (ch == 'H') {
+          parser->state = s_res_or_resp_H;
 
-        if (ch == 'H')
-          state = s_res_or_resp_H;
-        else {
+          CALLBACK_NOTIFY(message_begin);
+        } else {
           parser->type = HTTP_REQUEST;
-          goto start_req_method_assign;
+          parser->state = s_start_req;
+          goto reexecute_byte;
         }
+
         break;
       }
 
       case s_res_or_resp_H:
         if (ch == 'T') {
           parser->type = HTTP_RESPONSE;
-          state = s_res_HT;
+          parser->state = s_res_HT;
         } else {
-          if (ch != 'E') goto error;
+          if (ch != 'E') {
+            SET_ERRNO(HPE_INVALID_CONSTANT);
+            goto error;
+          }
+
           parser->type = HTTP_REQUEST;
           parser->method = HTTP_HEAD;
-          index = 2;
-          state = s_req_method;
+          parser->index = 2;
+          parser->state = s_req_method;
         }
         break;
 
       case s_start_res:
       {
         parser->flags = 0;
-        parser->content_length = -1;
-
-        CALLBACK2(message_begin);
+        parser->content_length = ULLONG_MAX;
 
         switch (ch) {
           case 'H':
-            state = s_res_H;
+            parser->state = s_res_H;
             break;
 
           case CR:
@@ -462,105 +729,133 @@ size_t http_parser_execute (http_parser *parser,
             break;
 
           default:
+            SET_ERRNO(HPE_INVALID_CONSTANT);
             goto error;
         }
+
+        CALLBACK_NOTIFY(message_begin);
         break;
       }
 
       case s_res_H:
         STRICT_CHECK(ch != 'T');
-        state = s_res_HT;
+        parser->state = s_res_HT;
         break;
 
       case s_res_HT:
         STRICT_CHECK(ch != 'T');
-        state = s_res_HTT;
+        parser->state = s_res_HTT;
         break;
 
       case s_res_HTT:
         STRICT_CHECK(ch != 'P');
-        state = s_res_HTTP;
+        parser->state = s_res_HTTP;
         break;
 
       case s_res_HTTP:
         STRICT_CHECK(ch != '/');
-        state = s_res_first_http_major;
+        parser->state = s_res_first_http_major;
         break;
 
       case s_res_first_http_major:
-        if (ch < '1' || ch > '9') goto error;
+        if (ch < '0' || ch > '9') {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         parser->http_major = ch - '0';
-        state = s_res_http_major;
+        parser->state = s_res_http_major;
         break;
 
       /* major HTTP version or dot */
       case s_res_http_major:
       {
         if (ch == '.') {
-          state = s_res_first_http_minor;
+          parser->state = s_res_first_http_minor;
           break;
         }
 
-        if (ch < '0' || ch > '9') goto error;
+        if (!IS_NUM(ch)) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
 
         parser->http_major *= 10;
         parser->http_major += ch - '0';
 
-        if (parser->http_major > 999) goto error;
+        if (parser->http_major > 999) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         break;
       }
 
       /* first digit of minor HTTP version */
       case s_res_first_http_minor:
-        if (ch < '0' || ch > '9') goto error;
+        if (!IS_NUM(ch)) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         parser->http_minor = ch - '0';
-        state = s_res_http_minor;
+        parser->state = s_res_http_minor;
         break;
 
       /* minor HTTP version or end of request line */
       case s_res_http_minor:
       {
         if (ch == ' ') {
-          state = s_res_first_status_code;
+          parser->state = s_res_first_status_code;
           break;
         }
 
-        if (ch < '0' || ch > '9') goto error;
+        if (!IS_NUM(ch)) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
 
         parser->http_minor *= 10;
         parser->http_minor += ch - '0';
 
-        if (parser->http_minor > 999) goto error;
+        if (parser->http_minor > 999) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         break;
       }
 
       case s_res_first_status_code:
       {
-        if (ch < '0' || ch > '9') {
+        if (!IS_NUM(ch)) {
           if (ch == ' ') {
             break;
           }
+
+          SET_ERRNO(HPE_INVALID_STATUS);
           goto error;
         }
         parser->status_code = ch - '0';
-        state = s_res_status_code;
+        parser->state = s_res_status_code;
         break;
       }
 
       case s_res_status_code:
       {
-        if (ch < '0' || ch > '9') {
+        if (!IS_NUM(ch)) {
           switch (ch) {
             case ' ':
-              state = s_res_status;
+              parser->state = s_res_status;
               break;
             case CR:
-              state = s_res_line_almost_done;
+              parser->state = s_res_line_almost_done;
               break;
             case LF:
-              state = s_header_field_start;
+              parser->state = s_header_field_start;
               break;
             default:
+              SET_ERRNO(HPE_INVALID_STATUS);
               goto error;
           }
           break;
@@ -569,7 +864,11 @@ size_t http_parser_execute (http_parser *parser,
         parser->status_code *= 10;
         parser->status_code += ch - '0';
 
-        if (parser->status_code > 999) goto error;
+        if (parser->status_code > 999) {
+          SET_ERRNO(HPE_INVALID_STATUS);
+          goto error;
+        }
+
         break;
       }
 
@@ -577,19 +876,19 @@ size_t http_parser_execute (http_parser *parser,
         /* the human readable status. e.g. "NOT FOUND"
          * we are not humans so just ignore this */
         if (ch == CR) {
-          state = s_res_line_almost_done;
+          parser->state = s_res_line_almost_done;
           break;
         }
 
         if (ch == LF) {
-          state = s_header_field_start;
+          parser->state = s_header_field_start;
           break;
         }
         break;
 
       case s_res_line_almost_done:
         STRICT_CHECK(ch != LF);
-        state = s_header_field_start;
+        parser->state = s_header_field_start;
         break;
 
       case s_start_req:
@@ -597,15 +896,15 @@ size_t http_parser_execute (http_parser *parser,
         if (ch == CR || ch == LF)
           break;
         parser->flags = 0;
-        parser->content_length = -1;
+        parser->content_length = ULLONG_MAX;
 
-        CALLBACK2(message_begin);
-
-        if (ch < 'A' || 'Z' < ch) goto error;
+        if (!IS_ALPHA(ch)) {
+          SET_ERRNO(HPE_INVALID_METHOD);
+          goto error;
+        }
 
-      start_req_method_assign:
         parser->method = (enum http_method) 0;
-        index = 1;
+        parser->index = 1;
         switch (ch) {
           case 'C': parser->method = HTTP_CONNECT; /* or COPY, CHECKOUT */ break;
           case 'D': parser->method = HTTP_DELETE; break;
@@ -615,342 +914,157 @@ size_t http_parser_execute (http_parser *parser,
           case 'M': parser->method = HTTP_MKCOL; /* or MOVE, MKACTIVITY, MERGE, M-SEARCH */ break;
           case 'N': parser->method = HTTP_NOTIFY; break;
           case 'O': parser->method = HTTP_OPTIONS; break;
-          case 'P': parser->method = HTTP_POST; /* or PROPFIND or PROPPATCH or PUT */ break;
+          case 'P': parser->method = HTTP_POST;
+            /* or PROPFIND|PROPPATCH|PUT|PATCH|PURGE */
+            break;
           case 'R': parser->method = HTTP_REPORT; break;
           case 'S': parser->method = HTTP_SUBSCRIBE; break;
           case 'T': parser->method = HTTP_TRACE; break;
           case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE */ break;
-          default: goto error;
+          default:
+            SET_ERRNO(HPE_INVALID_METHOD);
+            goto error;
         }
-        state = s_req_method;
+        parser->state = s_req_method;
+
+        CALLBACK_NOTIFY(message_begin);
+
         break;
       }
 
       case s_req_method:
       {
-        if (ch == '\0')
+        const char *matcher;
+        if (ch == '\0') {
+          SET_ERRNO(HPE_INVALID_METHOD);
           goto error;
+        }
 
-        const char *matcher = method_strings[parser->method];
-        if (ch == ' ' && matcher[index] == '\0') {
-          state = s_req_spaces_before_url;
-        } else if (ch == matcher[index]) {
+        matcher = method_strings[parser->method];
+        if (ch == ' ' && matcher[parser->index] == '\0') {
+          parser->state = s_req_spaces_before_url;
+        } else if (ch == matcher[parser->index]) {
           ; /* nada */
         } else if (parser->method == HTTP_CONNECT) {
-          if (index == 1 && ch == 'H') {
+          if (parser->index == 1 && ch == 'H') {
             parser->method = HTTP_CHECKOUT;
-          } else if (index == 2  && ch == 'P') {
+          } else if (parser->index == 2  && ch == 'P') {
             parser->method = HTTP_COPY;
+          } else {
+            goto error;
           }
         } else if (parser->method == HTTP_MKCOL) {
-          if (index == 1 && ch == 'O') {
+          if (parser->index == 1 && ch == 'O') {
             parser->method = HTTP_MOVE;
-          } else if (index == 1 && ch == 'E') {
+          } else if (parser->index == 1 && ch == 'E') {
             parser->method = HTTP_MERGE;
-          } else if (index == 1 && ch == '-') {
+          } else if (parser->index == 1 && ch == '-') {
             parser->method = HTTP_MSEARCH;
-          } else if (index == 2 && ch == 'A') {
+          } else if (parser->index == 2 && ch == 'A') {
             parser->method = HTTP_MKACTIVITY;
+          } else {
+            goto error;
           }
-        } else if (index == 1 && parser->method == HTTP_POST && ch == 'R') {
-          parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */
-        } else if (index == 1 && parser->method == HTTP_POST && ch == 'U') {
-          parser->method = HTTP_PUT;
-        } else if (index == 2 && parser->method == HTTP_UNLOCK && ch == 'S') {
-          parser->method = HTTP_UNSUBSCRIBE;
-        } else if (index == 4 && parser->method == HTTP_PROPFIND && ch == 'P') {
+        } else if (parser->index == 1 && parser->method == HTTP_POST) {
+          if (ch == 'R') {
+            parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */
+          } else if (ch == 'U') {
+            parser->method = HTTP_PUT; /* or HTTP_PURGE */
+          } else if (ch == 'A') {
+            parser->method = HTTP_PATCH;
+          } else {
+            goto error;
+          }
+        } else if (parser->index == 2) {
+          if (parser->method == HTTP_PUT) {
+            if (ch == 'R') parser->method = HTTP_PURGE;
+          } else if (parser->method == HTTP_UNLOCK) {
+            if (ch == 'S') parser->method = HTTP_UNSUBSCRIBE;
+          }
+        } else if (parser->index == 4 && parser->method == HTTP_PROPFIND && ch == 'P') {
           parser->method = HTTP_PROPPATCH;
         } else {
+          SET_ERRNO(HPE_INVALID_METHOD);
           goto error;
         }
 
-        ++index;
+        ++parser->index;
         break;
       }
+
       case s_req_spaces_before_url:
       {
         if (ch == ' ') break;
 
-        if (ch == '/' || ch == '*') {
-          MARK(url);
-          MARK(path);
-          state = s_req_path;
-          break;
+        MARK(url);
+        if (parser->method == HTTP_CONNECT) {
+          parser->state = s_req_host_start;
         }
 
-        c = LOWER(ch);
-
-        if (c >= 'a' && c <= 'z') {
-          MARK(url);
-          state = s_req_schema;
-          break;
+        parser->state = parse_url_char((enum state)parser->state, ch);
+        if (parser->state == s_dead) {
+          SET_ERRNO(HPE_INVALID_URL);
+          goto error;
         }
 
-        goto error;
+        break;
       }
 
       case s_req_schema:
-      {
-        c = LOWER(ch);
-
-        if (c >= 'a' && c <= 'z') break;
-
-        if (ch == ':') {
-          state = s_req_schema_slash;
-          break;
-        } else if (ch == '.') {
-          state = s_req_host;
-          break;
-        } else if ('0' <= ch && ch <= '9') {
-          state = s_req_host;
-          break;
-        }
-
-        goto error;
-      }
-
       case s_req_schema_slash:
-        STRICT_CHECK(ch != '/');
-        state = s_req_schema_slash_slash;
-        break;
-
       case s_req_schema_slash_slash:
-        STRICT_CHECK(ch != '/');
-        state = s_req_host;
-        break;
-
-      case s_req_host:
-      {
-        c = LOWER(ch);
-        if (c >= 'a' && c <= 'z') break;
-        if ((ch >= '0' && ch <= '9') || ch == '.' || ch == '-') break;
-        switch (ch) {
-          case ':':
-            state = s_req_port;
-            break;
-          case '/':
-            MARK(path);
-            state = s_req_path;
-            break;
-          case ' ':
-            /* The request line looks like:
-             *   "GET http://foo.bar.com HTTP/1.1"
-             * That is, there is no path.
-             */
-            CALLBACK(url);
-            state = s_req_http_start;
-            break;
-          case '?':
-            state = s_req_query_string_start;
-            break;
-          default:
-            goto error;
-        }
-        break;
-      }
-
-      case s_req_port:
+      case s_req_host_start:
+      case s_req_host_v6_start:
+      case s_req_host_v6:
+      case s_req_port_start:
       {
-        if (ch >= '0' && ch <= '9') break;
         switch (ch) {
-          case '/':
-            MARK(path);
-            state = s_req_path;
-            break;
+          /* No whitespace allowed here */
           case ' ':
-            /* The request line looks like:
-             *   "GET http://foo.bar.com:1234 HTTP/1.1"
-             * That is, there is no path.
-             */
-            CALLBACK(url);
-            state = s_req_http_start;
-            break;
-          case '?':
-            state = s_req_query_string_start;
-            break;
-          default:
-            goto error;
-        }
-        break;
-      }
-
-      case s_req_path:
-      {
-        if (normal_url_char[(unsigned char)ch]) break;
-
-        switch (ch) {
-          case ' ':
-            CALLBACK(url);
-            CALLBACK(path);
-            state = s_req_http_start;
-            break;
           case CR:
-            CALLBACK(url);
-            CALLBACK(path);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_req_line_almost_done;
-            break;
           case LF:
-            CALLBACK(url);
-            CALLBACK(path);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_header_field_start;
-            break;
-          case '?':
-            CALLBACK(path);
-            state = s_req_query_string_start;
-            break;
-          case '#':
-            CALLBACK(path);
-            state = s_req_fragment_start;
-            break;
-          default:
+            SET_ERRNO(HPE_INVALID_URL);
             goto error;
-        }
-        break;
-      }
-
-      case s_req_query_string_start:
-      {
-        if (normal_url_char[(unsigned char)ch]) {
-          MARK(query_string);
-          state = s_req_query_string;
-          break;
-        }
-
-        switch (ch) {
-          case '?':
-            break; /* XXX ignore extra '?' ... is this right? */
-          case ' ':
-            CALLBACK(url);
-            state = s_req_http_start;
-            break;
-          case CR:
-            CALLBACK(url);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_req_line_almost_done;
-            break;
-          case LF:
-            CALLBACK(url);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_header_field_start;
-            break;
-          case '#':
-            state = s_req_fragment_start;
-            break;
           default:
-            goto error;
+            parser->state = parse_url_char((enum state)parser->state, ch);
+            if (parser->state == s_dead) {
+              SET_ERRNO(HPE_INVALID_URL);
+              goto error;
+            }
         }
-        break;
-      }
 
-      case s_req_query_string:
-      {
-        if (normal_url_char[(unsigned char)ch]) break;
-
-        switch (ch) {
-          case '?':
-            /* allow extra '?' in query string */
-            break;
-          case ' ':
-            CALLBACK(url);
-            CALLBACK(query_string);
-            state = s_req_http_start;
-            break;
-          case CR:
-            CALLBACK(url);
-            CALLBACK(query_string);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_req_line_almost_done;
-            break;
-          case LF:
-            CALLBACK(url);
-            CALLBACK(query_string);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_header_field_start;
-            break;
-          case '#':
-            CALLBACK(query_string);
-            state = s_req_fragment_start;
-            break;
-          default:
-            goto error;
-        }
         break;
       }
 
+      case s_req_host:
+      case s_req_host_v6_end:
+      case s_req_port:
+      case s_req_path:
+      case s_req_query_string_start:
+      case s_req_query_string:
       case s_req_fragment_start:
-      {
-        if (normal_url_char[(unsigned char)ch]) {
-          MARK(fragment);
-          state = s_req_fragment;
-          break;
-        }
-
-        switch (ch) {
-          case ' ':
-            CALLBACK(url);
-            state = s_req_http_start;
-            break;
-          case CR:
-            CALLBACK(url);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_req_line_almost_done;
-            break;
-          case LF:
-            CALLBACK(url);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_header_field_start;
-            break;
-          case '?':
-            MARK(fragment);
-            state = s_req_fragment;
-            break;
-          case '#':
-            break;
-          default:
-            goto error;
-        }
-        break;
-      }
-
       case s_req_fragment:
       {
-        if (normal_url_char[(unsigned char)ch]) break;
-
         switch (ch) {
           case ' ':
-            CALLBACK(url);
-            CALLBACK(fragment);
-            state = s_req_http_start;
+            parser->state = s_req_http_start;
+            CALLBACK_DATA(url);
             break;
           case CR:
-            CALLBACK(url);
-            CALLBACK(fragment);
-            parser->http_major = 0;
-            parser->http_minor = 9;
-            state = s_req_line_almost_done;
-            break;
           case LF:
-            CALLBACK(url);
-            CALLBACK(fragment);
             parser->http_major = 0;
             parser->http_minor = 9;
-            state = s_header_field_start;
-            break;
-          case '?':
-          case '#':
+            parser->state = (ch == CR) ?
+              s_req_line_almost_done :
+              s_header_field_start;
+            CALLBACK_DATA(url);
             break;
           default:
-            goto error;
+            parser->state = parse_url_char((enum state)parser->state, ch);
+            if (parser->state == s_dead) {
+              SET_ERRNO(HPE_INVALID_URL);
+              goto error;
+            }
         }
         break;
       }
@@ -958,140 +1072,170 @@ size_t http_parser_execute (http_parser *parser,
       case s_req_http_start:
         switch (ch) {
           case 'H':
-            state = s_req_http_H;
+            parser->state = s_req_http_H;
             break;
           case ' ':
             break;
           default:
+            SET_ERRNO(HPE_INVALID_CONSTANT);
             goto error;
         }
         break;
 
       case s_req_http_H:
         STRICT_CHECK(ch != 'T');
-        state = s_req_http_HT;
+        parser->state = s_req_http_HT;
         break;
 
       case s_req_http_HT:
         STRICT_CHECK(ch != 'T');
-        state = s_req_http_HTT;
+        parser->state = s_req_http_HTT;
         break;
 
       case s_req_http_HTT:
         STRICT_CHECK(ch != 'P');
-        state = s_req_http_HTTP;
+        parser->state = s_req_http_HTTP;
         break;
 
       case s_req_http_HTTP:
         STRICT_CHECK(ch != '/');
-        state = s_req_first_http_major;
+        parser->state = s_req_first_http_major;
         break;
 
       /* first digit of major HTTP version */
       case s_req_first_http_major:
-        if (ch < '1' || ch > '9') goto error;
+        if (ch < '1' || ch > '9') {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         parser->http_major = ch - '0';
-        state = s_req_http_major;
+        parser->state = s_req_http_major;
         break;
 
       /* major HTTP version or dot */
       case s_req_http_major:
       {
         if (ch == '.') {
-          state = s_req_first_http_minor;
+          parser->state = s_req_first_http_minor;
           break;
         }
 
-        if (ch < '0' || ch > '9') goto error;
+        if (!IS_NUM(ch)) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
 
         parser->http_major *= 10;
         parser->http_major += ch - '0';
 
-        if (parser->http_major > 999) goto error;
+        if (parser->http_major > 999) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         break;
       }
 
       /* first digit of minor HTTP version */
       case s_req_first_http_minor:
-        if (ch < '0' || ch > '9') goto error;
+        if (!IS_NUM(ch)) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         parser->http_minor = ch - '0';
-        state = s_req_http_minor;
+        parser->state = s_req_http_minor;
         break;
 
       /* minor HTTP version or end of request line */
       case s_req_http_minor:
       {
         if (ch == CR) {
-          state = s_req_line_almost_done;
+          parser->state = s_req_line_almost_done;
           break;
         }
 
         if (ch == LF) {
-          state = s_header_field_start;
+          parser->state = s_header_field_start;
           break;
         }
 
         /* XXX allow spaces after digit? */
 
-        if (ch < '0' || ch > '9') goto error;
+        if (!IS_NUM(ch)) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
 
         parser->http_minor *= 10;
         parser->http_minor += ch - '0';
 
-        if (parser->http_minor > 999) goto error;
+        if (parser->http_minor > 999) {
+          SET_ERRNO(HPE_INVALID_VERSION);
+          goto error;
+        }
+
         break;
       }
 
       /* end of request line */
       case s_req_line_almost_done:
       {
-        if (ch != LF) goto error;
-        state = s_header_field_start;
+        if (ch != LF) {
+          SET_ERRNO(HPE_LF_EXPECTED);
+          goto error;
+        }
+
+        parser->state = s_header_field_start;
         break;
       }
 
       case s_header_field_start:
       {
         if (ch == CR) {
-          state = s_headers_almost_done;
+          parser->state = s_headers_almost_done;
           break;
         }
 
         if (ch == LF) {
           /* they might be just sending \n instead of \r\n so this would be
            * the second \n to denote the end of headers*/
-          state = s_headers_almost_done;
-          goto headers_almost_done;
+          parser->state = s_headers_almost_done;
+          goto reexecute_byte;
         }
 
         c = TOKEN(ch);
 
-        if (!c) goto error;
+        if (!c) {
+          SET_ERRNO(HPE_INVALID_HEADER_TOKEN);
+          goto error;
+        }
 
         MARK(header_field);
 
-        index = 0;
-        state = s_header_field;
+        parser->index = 0;
+        parser->state = s_header_field;
 
         switch (c) {
           case 'c':
-            header_state = h_C;
+            parser->header_state = h_C;
             break;
 
           case 'p':
-            header_state = h_matching_proxy_connection;
+            parser->header_state = h_matching_proxy_connection;
             break;
 
           case 't':
-            header_state = h_matching_transfer_encoding;
+            parser->header_state = h_matching_transfer_encoding;
             break;
 
           case 'u':
-            header_state = h_matching_upgrade;
+            parser->header_state = h_matching_upgrade;
             break;
 
           default:
-            header_state = h_general;
+            parser->header_state = h_general;
             break;
         }
         break;
@@ -1102,31 +1246,31 @@ size_t http_parser_execute (http_parser *parser,
         c = TOKEN(ch);
 
         if (c) {
-          switch (header_state) {
+          switch (parser->header_state) {
             case h_general:
               break;
 
             case h_C:
-              index++;
-              header_state = (c == 'o' ? h_CO : h_general);
+              parser->index++;
+              parser->header_state = (c == 'o' ? h_CO : h_general);
               break;
 
             case h_CO:
-              index++;
-              header_state = (c == 'n' ? h_CON : h_general);
+              parser->index++;
+              parser->header_state = (c == 'n' ? h_CON : h_general);
               break;
 
             case h_CON:
-              index++;
+              parser->index++;
               switch (c) {
                 case 'n':
-                  header_state = h_matching_connection;
+                  parser->header_state = h_matching_connection;
                   break;
                 case 't':
-                  header_state = h_matching_content_length;
+                  parser->header_state = h_matching_content_length;
                   break;
                 default:
-                  header_state = h_general;
+                  parser->header_state = h_general;
                   break;
               }
               break;
@@ -1134,60 +1278,60 @@ size_t http_parser_execute (http_parser *parser,
             /* connection */
 
             case h_matching_connection:
-              index++;
-              if (index > sizeof(CONNECTION)-1
-                  || c != CONNECTION[index]) {
-                header_state = h_general;
-              } else if (index == sizeof(CONNECTION)-2) {
-                header_state = h_connection;
+              parser->index++;
+              if (parser->index > sizeof(CONNECTION)-1
+                  || c != CONNECTION[parser->index]) {
+                parser->header_state = h_general;
+              } else if (parser->index == sizeof(CONNECTION)-2) {
+                parser->header_state = h_connection;
               }
               break;
 
             /* proxy-connection */
 
             case h_matching_proxy_connection:
-              index++;
-              if (index > sizeof(PROXY_CONNECTION)-1
-                  || c != PROXY_CONNECTION[index]) {
-                header_state = h_general;
-              } else if (index == sizeof(PROXY_CONNECTION)-2) {
-                header_state = h_connection;
+              parser->index++;
+              if (parser->index > sizeof(PROXY_CONNECTION)-1
+                  || c != PROXY_CONNECTION[parser->index]) {
+                parser->header_state = h_general;
+              } else if (parser->index == sizeof(PROXY_CONNECTION)-2) {
+                parser->header_state = h_connection;
               }
               break;
 
             /* content-length */
 
             case h_matching_content_length:
-              index++;
-              if (index > sizeof(CONTENT_LENGTH)-1
-                  || c != CONTENT_LENGTH[index]) {
-                header_state = h_general;
-              } else if (index == sizeof(CONTENT_LENGTH)-2) {
-                header_state = h_content_length;
+              parser->index++;
+              if (parser->index > sizeof(CONTENT_LENGTH)-1
+                  || c != CONTENT_LENGTH[parser->index]) {
+                parser->header_state = h_general;
+              } else if (parser->index == sizeof(CONTENT_LENGTH)-2) {
+                parser->header_state = h_content_length;
               }
               break;
 
             /* transfer-encoding */
 
             case h_matching_transfer_encoding:
-              index++;
-              if (index > sizeof(TRANSFER_ENCODING)-1
-                  || c != TRANSFER_ENCODING[index]) {
-                header_state = h_general;
-              } else if (index == sizeof(TRANSFER_ENCODING)-2) {
-                header_state = h_transfer_encoding;
+              parser->index++;
+              if (parser->index > sizeof(TRANSFER_ENCODING)-1
+                  || c != TRANSFER_ENCODING[parser->index]) {
+                parser->header_state = h_general;
+              } else if (parser->index == sizeof(TRANSFER_ENCODING)-2) {
+                parser->header_state = h_transfer_encoding;
               }
               break;
 
             /* upgrade */
 
             case h_matching_upgrade:
-              index++;
-              if (index > sizeof(UPGRADE)-1
-                  || c != UPGRADE[index]) {
-                header_state = h_general;
-              } else if (index == sizeof(UPGRADE)-2) {
-                header_state = h_upgrade;
+              parser->index++;
+              if (parser->index > sizeof(UPGRADE)-1
+                  || c != UPGRADE[parser->index]) {
+                parser->header_state = h_general;
+              } else if (parser->index == sizeof(UPGRADE)-2) {
+                parser->header_state = h_upgrade;
               }
               break;
 
@@ -1195,7 +1339,7 @@ size_t http_parser_execute (http_parser *parser,
             case h_content_length:
             case h_transfer_encoding:
             case h_upgrade:
-              if (ch != ' ') header_state = h_general;
+              if (ch != ' ') parser->header_state = h_general;
               break;
 
             default:
@@ -1206,84 +1350,89 @@ size_t http_parser_execute (http_parser *parser,
         }
 
         if (ch == ':') {
-          CALLBACK(header_field);
-          state = s_header_value_start;
+          parser->state = s_header_value_start;
+          CALLBACK_DATA(header_field);
           break;
         }
 
         if (ch == CR) {
-          state = s_header_almost_done;
-          CALLBACK(header_field);
+          parser->state = s_header_almost_done;
+          CALLBACK_DATA(header_field);
           break;
         }
 
         if (ch == LF) {
-          CALLBACK(header_field);
-          state = s_header_field_start;
+          parser->state = s_header_field_start;
+          CALLBACK_DATA(header_field);
           break;
         }
 
+        SET_ERRNO(HPE_INVALID_HEADER_TOKEN);
         goto error;
       }
 
       case s_header_value_start:
       {
-        if (ch == ' ') break;
+        if (ch == ' ' || ch == '\t') break;
 
         MARK(header_value);
 
-        state = s_header_value;
-        index = 0;
-
-        c = LOWER(ch);
+        parser->state = s_header_value;
+        parser->index = 0;
 
         if (ch == CR) {
-          CALLBACK(header_value);
-          header_state = h_general;
-          state = s_header_almost_done;
+          parser->header_state = h_general;
+          parser->state = s_header_almost_done;
+          CALLBACK_DATA(header_value);
           break;
         }
 
         if (ch == LF) {
-          CALLBACK(header_value);
-          state = s_header_field_start;
+          parser->state = s_header_field_start;
+          CALLBACK_DATA(header_value);
           break;
         }
 
-        switch (header_state) {
+        c = LOWER(ch);
+
+        switch (parser->header_state) {
           case h_upgrade:
             parser->flags |= F_UPGRADE;
-            header_state = h_general;
+            parser->header_state = h_general;
             break;
 
           case h_transfer_encoding:
             /* looking for 'Transfer-Encoding: chunked' */
             if ('c' == c) {
-              header_state = h_matching_transfer_encoding_chunked;
+              parser->header_state = h_matching_transfer_encoding_chunked;
             } else {
-              header_state = h_general;
+              parser->header_state = h_general;
             }
             break;
 
           case h_content_length:
-            if (ch < '0' || ch > '9') goto error;
+            if (!IS_NUM(ch)) {
+              SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
+              goto error;
+            }
+
             parser->content_length = ch - '0';
             break;
 
           case h_connection:
             /* looking for 'Connection: keep-alive' */
             if (c == 'k') {
-              header_state = h_matching_connection_keep_alive;
+              parser->header_state = h_matching_connection_keep_alive;
             /* looking for 'Connection: close' */
             } else if (c == 'c') {
-              header_state = h_matching_connection_close;
+              parser->header_state = h_matching_connection_close;
             } else {
-              header_state = h_general;
+              parser->header_state = h_general;
             }
             break;
 
           default:
-            header_state = h_general;
+            parser->header_state = h_general;
             break;
         }
         break;
@@ -1291,20 +1440,22 @@ size_t http_parser_execute (http_parser *parser,
 
       case s_header_value:
       {
-        c = LOWER(ch);
 
         if (ch == CR) {
-          CALLBACK(header_value);
-          state = s_header_almost_done;
+          parser->state = s_header_almost_done;
+          CALLBACK_DATA(header_value);
           break;
         }
 
         if (ch == LF) {
-          CALLBACK(header_value);
-          goto header_almost_done;
+          parser->state = s_header_almost_done;
+          CALLBACK_DATA_NOADVANCE(header_value);
+          goto reexecute_byte;
         }
 
-        switch (header_state) {
+        c = LOWER(ch);
+
+        switch (parser->header_state) {
           case h_general:
             break;
 
@@ -1314,66 +1465,83 @@ size_t http_parser_execute (http_parser *parser,
             break;
 
           case h_content_length:
+          {
+            uint64_t t;
+
             if (ch == ' ') break;
-            if (ch < '0' || ch > '9') goto error;
-            parser->content_length *= 10;
-            parser->content_length += ch - '0';
+
+            if (!IS_NUM(ch)) {
+              SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
+              goto error;
+            }
+
+            t = parser->content_length;
+            t *= 10;
+            t += ch - '0';
+
+            /* Overflow? */
+            if (t < parser->content_length || t == ULLONG_MAX) {
+              SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
+              goto error;
+            }
+
+            parser->content_length = t;
             break;
+          }
 
           /* Transfer-Encoding: chunked */
           case h_matching_transfer_encoding_chunked:
-            index++;
-            if (index > sizeof(CHUNKED)-1
-                || c != CHUNKED[index]) {
-              header_state = h_general;
-            } else if (index == sizeof(CHUNKED)-2) {
-              header_state = h_transfer_encoding_chunked;
+            parser->index++;
+            if (parser->index > sizeof(CHUNKED)-1
+                || c != CHUNKED[parser->index]) {
+              parser->header_state = h_general;
+            } else if (parser->index == sizeof(CHUNKED)-2) {
+              parser->header_state = h_transfer_encoding_chunked;
             }
             break;
 
           /* looking for 'Connection: keep-alive' */
           case h_matching_connection_keep_alive:
-            index++;
-            if (index > sizeof(KEEP_ALIVE)-1
-                || c != KEEP_ALIVE[index]) {
-              header_state = h_general;
-            } else if (index == sizeof(KEEP_ALIVE)-2) {
-              header_state = h_connection_keep_alive;
+            parser->index++;
+            if (parser->index > sizeof(KEEP_ALIVE)-1
+                || c != KEEP_ALIVE[parser->index]) {
+              parser->header_state = h_general;
+            } else if (parser->index == sizeof(KEEP_ALIVE)-2) {
+              parser->header_state = h_connection_keep_alive;
             }
             break;
 
           /* looking for 'Connection: close' */
           case h_matching_connection_close:
-            index++;
-            if (index > sizeof(CLOSE)-1 || c != CLOSE[index]) {
-              header_state = h_general;
-            } else if (index == sizeof(CLOSE)-2) {
-              header_state = h_connection_close;
+            parser->index++;
+            if (parser->index > sizeof(CLOSE)-1 || c != CLOSE[parser->index]) {
+              parser->header_state = h_general;
+            } else if (parser->index == sizeof(CLOSE)-2) {
+              parser->header_state = h_connection_close;
             }
             break;
 
           case h_transfer_encoding_chunked:
           case h_connection_keep_alive:
           case h_connection_close:
-            if (ch != ' ') header_state = h_general;
+            if (ch != ' ') parser->header_state = h_general;
             break;
 
           default:
-            state = s_header_value;
-            header_state = h_general;
+            parser->state = s_header_value;
+            parser->header_state = h_general;
             break;
         }
         break;
       }
 
       case s_header_almost_done:
-      header_almost_done:
       {
         STRICT_CHECK(ch != LF);
 
-        state = s_header_field_start;
+        parser->state = s_header_value_lws;
 
-        switch (header_state) {
+        switch (parser->header_state) {
           case h_connection_keep_alive:
             parser->flags |= F_CONNECTION_KEEP_ALIVE;
             break;
@@ -1386,32 +1554,47 @@ size_t http_parser_execute (http_parser *parser,
           default:
             break;
         }
+
+        break;
+      }
+
+      case s_header_value_lws:
+      {
+        if (ch == ' ' || ch == '\t')
+          parser->state = s_header_value_start;
+        else
+        {
+          parser->state = s_header_field_start;
+          goto reexecute_byte;
+        }
         break;
       }
 
       case s_headers_almost_done:
-      headers_almost_done:
       {
         STRICT_CHECK(ch != LF);
 
         if (parser->flags & F_TRAILING) {
           /* End of a chunked request */
-          CALLBACK2(message_complete);
-          state = NEW_MESSAGE();
+          parser->state = NEW_MESSAGE();
+          CALLBACK_NOTIFY(message_complete);
           break;
         }
 
-        nread = 0;
+        parser->state = s_headers_done;
 
-        if (parser->flags & F_UPGRADE || parser->method == HTTP_CONNECT) {
-          parser->upgrade = 1;
-        }
+        /* Set this here so that on_headers_complete() callbacks can see it */
+        parser->upgrade =
+          (parser->flags & F_UPGRADE || parser->method == HTTP_CONNECT);
 
         /* Here we call the headers_complete callback. This is somewhat
          * different than other callbacks because if the user returns 1, we
          * will interpret that as saying that this message has no body. This
          * is needed for the annoying case of recieving a response to a HEAD
          * request.
+         *
+         * We'd like to use CALLBACK_NOTIFY_NOADVANCE() here but we cannot, so
+         * we have to simulate it by handling a change in errno below.
          */
         if (settings->on_headers_complete) {
           switch (settings->on_headers_complete(parser)) {
@@ -1423,39 +1606,54 @@ size_t http_parser_execute (http_parser *parser,
               break;
 
             default:
-              parser->state = state;
+              SET_ERRNO(HPE_CB_headers_complete);
               return p - data; /* Error */
           }
         }
 
+        if (HTTP_PARSER_ERRNO(parser) != HPE_OK) {
+          return p - data;
+        }
+
+        goto reexecute_byte;
+      }
+
+      case s_headers_done:
+      {
+        STRICT_CHECK(ch != LF);
+
+        parser->nread = 0;
+
         /* Exit, the rest of the connect is in a different protocol. */
         if (parser->upgrade) {
-          CALLBACK2(message_complete);
-          return (p - data);
+          parser->state = NEW_MESSAGE();
+          CALLBACK_NOTIFY(message_complete);
+          return (p - data) + 1;
         }
 
         if (parser->flags & F_SKIPBODY) {
-          CALLBACK2(message_complete);
-          state = NEW_MESSAGE();
+          parser->state = NEW_MESSAGE();
+          CALLBACK_NOTIFY(message_complete);
         } else if (parser->flags & F_CHUNKED) {
           /* chunked encoding - ignore Content-Length header */
-          state = s_chunk_size_start;
+          parser->state = s_chunk_size_start;
         } else {
           if (parser->content_length == 0) {
             /* Content-Length header given but zero: Content-Length: 0\r\n */
-            CALLBACK2(message_complete);
-            state = NEW_MESSAGE();
-          } else if (parser->content_length > 0) {
+            parser->state = NEW_MESSAGE();
+            CALLBACK_NOTIFY(message_complete);
+          } else if (parser->content_length != ULLONG_MAX) {
             /* Content-Length header given and non-zero */
-            state = s_body_identity;
+            parser->state = s_body_identity;
           } else {
-            if (parser->type == HTTP_REQUEST || http_should_keep_alive(parser)) {
+            if (parser->type == HTTP_REQUEST ||
+                !http_message_needs_eof(parser)) {
               /* Assume content-length 0 - read the next */
-              CALLBACK2(message_complete);
-              state = NEW_MESSAGE();
+              parser->state = NEW_MESSAGE();
+              CALLBACK_NOTIFY(message_complete);
             } else {
               /* Read body until EOF */
-              state = s_body_identity_eof;
+              parser->state = s_body_identity_eof;
             }
           }
         }
@@ -1464,60 +1662,103 @@ size_t http_parser_execute (http_parser *parser,
       }
 
       case s_body_identity:
-        to_read = MIN(pe - p, (int64_t)parser->content_length);
-        if (to_read > 0) {
-          if (settings->on_body) settings->on_body(parser, p, to_read);
-          p += to_read - 1;
-          parser->content_length -= to_read;
-          if (parser->content_length == 0) {
-            CALLBACK2(message_complete);
-            state = NEW_MESSAGE();
-          }
+      {
+        uint64_t to_read = MIN(parser->content_length,
+                               (uint64_t) ((data + len) - p));
+
+        assert(parser->content_length != 0
+            && parser->content_length != ULLONG_MAX);
+
+        /* The difference between advancing content_length and p is because
+         * the latter will automaticaly advance on the next loop iteration.
+         * Further, if content_length ends up at 0, we want to see the last
+         * byte again for our message complete callback.
+         */
+        MARK(body);
+        parser->content_length -= to_read;
+        p += to_read - 1;
+
+        if (parser->content_length == 0) {
+          parser->state = s_message_done;
+
+          /* Mimic CALLBACK_DATA_NOADVANCE() but with one extra byte.
+           *
+           * The alternative to doing this is to wait for the next byte to
+           * trigger the data callback, just as in every other case. The
+           * problem with this is that this makes it difficult for the test
+           * harness to distinguish between complete-on-EOF and
+           * complete-on-length. It's not clear that this distinction is
+           * important for applications, but let's keep it for now.
+           */
+          CALLBACK_DATA_(body, p - body_mark + 1, p - data);
+          goto reexecute_byte;
         }
+
         break;
+      }
 
       /* read until EOF */
       case s_body_identity_eof:
-        to_read = pe - p;
-        if (to_read > 0) {
-          if (settings->on_body) settings->on_body(parser, p, to_read);
-          p += to_read - 1;
-        }
+        MARK(body);
+        p = data + len - 1;
+
+        break;
+
+      case s_message_done:
+        parser->state = NEW_MESSAGE();
+        CALLBACK_NOTIFY(message_complete);
         break;
 
       case s_chunk_size_start:
       {
-        assert(nread == 1);
+        assert(parser->nread == 1);
         assert(parser->flags & F_CHUNKED);
 
-        c = unhex[(unsigned char)ch];
-        if (c == -1) goto error;
-        parser->content_length = c;
-        state = s_chunk_size;
+        unhex_val = unhex[(unsigned char)ch];
+        if (unhex_val == -1) {
+          SET_ERRNO(HPE_INVALID_CHUNK_SIZE);
+          goto error;
+        }
+
+        parser->content_length = unhex_val;
+        parser->state = s_chunk_size;
         break;
       }
 
       case s_chunk_size:
       {
+        uint64_t t;
+
         assert(parser->flags & F_CHUNKED);
 
         if (ch == CR) {
-          state = s_chunk_size_almost_done;
+          parser->state = s_chunk_size_almost_done;
           break;
         }
 
-        c = unhex[(unsigned char)ch];
+        unhex_val = unhex[(unsigned char)ch];
 
-        if (c == -1) {
+        if (unhex_val == -1) {
           if (ch == ';' || ch == ' ') {
-            state = s_chunk_parameters;
+            parser->state = s_chunk_parameters;
             break;
           }
+
+          SET_ERRNO(HPE_INVALID_CHUNK_SIZE);
+          goto error;
+        }
+
+        t = parser->content_length;
+        t *= 16;
+        t += unhex_val;
+
+        /* Overflow? */
+        if (t < parser->content_length || t == ULLONG_MAX) {
+          SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
           goto error;
         }
 
-        parser->content_length *= 16;
-        parser->content_length += c;
+        parser->content_length = t;
         break;
       }
 
@@ -1526,7 +1767,7 @@ size_t http_parser_execute (http_parser *parser,
         assert(parser->flags & F_CHUNKED);
         /* just ignore this shit. TODO check for overflow */
         if (ch == CR) {
-          state = s_chunk_size_almost_done;
+          parser->state = s_chunk_size_almost_done;
           break;
         }
         break;
@@ -1537,74 +1778,117 @@ size_t http_parser_execute (http_parser *parser,
         assert(parser->flags & F_CHUNKED);
         STRICT_CHECK(ch != LF);
 
-        nread = 0;
+        parser->nread = 0;
 
         if (parser->content_length == 0) {
           parser->flags |= F_TRAILING;
-          state = s_header_field_start;
+          parser->state = s_header_field_start;
         } else {
-          state = s_chunk_data;
+          parser->state = s_chunk_data;
         }
         break;
       }
 
       case s_chunk_data:
       {
-        assert(parser->flags & F_CHUNKED);
+        uint64_t to_read = MIN(parser->content_length,
+                               (uint64_t) ((data + len) - p));
 
-        to_read = MIN(pe - p, (int64_t)(parser->content_length));
+        assert(parser->flags & F_CHUNKED);
+        assert(parser->content_length != 0
+            && parser->content_length != ULLONG_MAX);
 
-        if (to_read > 0) {
-          if (settings->on_body) settings->on_body(parser, p, to_read);
-          p += to_read - 1;
-        }
+        /* See the explanation in s_body_identity for why the content
+         * length and data pointers are managed this way.
+         */
+        MARK(body);
+        parser->content_length -= to_read;
+        p += to_read - 1;
 
-        if (to_read == parser->content_length) {
-          state = s_chunk_data_almost_done;
+        if (parser->content_length == 0) {
+          parser->state = s_chunk_data_almost_done;
         }
 
-        parser->content_length -= to_read;
         break;
       }
 
       case s_chunk_data_almost_done:
         assert(parser->flags & F_CHUNKED);
+        assert(parser->content_length == 0);
         STRICT_CHECK(ch != CR);
-        state = s_chunk_data_done;
+        parser->state = s_chunk_data_done;
+        CALLBACK_DATA(body);
         break;
 
       case s_chunk_data_done:
         assert(parser->flags & F_CHUNKED);
         STRICT_CHECK(ch != LF);
-        state = s_chunk_size_start;
+        parser->nread = 0;
+        parser->state = s_chunk_size_start;
         break;
 
       default:
         assert(0 && "unhandled state");
+        SET_ERRNO(HPE_INVALID_INTERNAL_STATE);
         goto error;
     }
   }
 
-  CALLBACK_NOCLEAR(header_field);
-  CALLBACK_NOCLEAR(header_value);
-  CALLBACK_NOCLEAR(fragment);
-  CALLBACK_NOCLEAR(query_string);
-  CALLBACK_NOCLEAR(path);
-  CALLBACK_NOCLEAR(url);
+  /* Run callbacks for any marks that we have leftover after we ran our of
+   * bytes. There should be at most one of these set, so it's OK to invoke
+   * them in series (unset marks will not result in callbacks).
+   *
+   * We use the NOADVANCE() variety of callbacks here because 'p' has already
+   * overflowed 'data' and this allows us to correct for the off-by-one that
+   * we'd otherwise have (since CALLBACK_DATA() is meant to be run with a 'p'
+   * value that's in-bounds).
+   */
 
-  parser->state = state;
-  parser->header_state = header_state;
-  parser->index = index;
-  parser->nread = nread;
+  assert(((header_field_mark ? 1 : 0) +
+          (header_value_mark ? 1 : 0) +
+          (url_mark ? 1 : 0)  +
+          (body_mark ? 1 : 0)) <= 1);
+
+  CALLBACK_DATA_NOADVANCE(header_field);
+  CALLBACK_DATA_NOADVANCE(header_value);
+  CALLBACK_DATA_NOADVANCE(url);
+  CALLBACK_DATA_NOADVANCE(body);
 
   return len;
 
 error:
-  parser->state = s_dead;
+  if (HTTP_PARSER_ERRNO(parser) == HPE_OK) {
+    SET_ERRNO(HPE_UNKNOWN);
+  }
+
   return (p - data);
 }
 
 
+/* Does the parser need to see an EOF to find the end of the message? */
+int
+http_message_needs_eof (http_parser *parser)
+{
+  if (parser->type == HTTP_REQUEST) {
+    return 0;
+  }
+
+  /* See RFC 2616 section 4.4 */
+  if (parser->status_code / 100 == 1 || /* 1xx e.g. Continue */
+      parser->status_code == 204 ||     /* No Content */
+      parser->status_code == 304 ||     /* Not Modified */
+      parser->flags & F_SKIPBODY) {     /* response to a HEAD request */
+    return 0;
+  }
+
+  if ((parser->flags & F_CHUNKED) || parser->content_length != ULLONG_MAX) {
+    return 0;
+  }
+
+  return 1;
+}
+
+
 int
 http_should_keep_alive (http_parser *parser)
 {
@@ -1612,17 +1896,15 @@ http_should_keep_alive (http_parser *parser)
     /* HTTP/1.1 */
     if (parser->flags & F_CONNECTION_CLOSE) {
       return 0;
-    } else {
-      return 1;
     }
   } else {
     /* HTTP/1.0 or earlier */
-    if (parser->flags & F_CONNECTION_KEEP_ALIVE) {
-      return 1;
-    } else {
+    if (!(parser->flags & F_CONNECTION_KEEP_ALIVE)) {
       return 0;
     }
   }
+
+  return !http_message_needs_eof(parser);
 }
 
 
@@ -1635,10 +1917,142 @@ const char * http_method_str (enum http_method m)
 void
 http_parser_init (http_parser *parser, enum http_parser_type t)
 {
+  void *data = parser->data; /* preserve application data */
+  memset(parser, 0, sizeof(*parser));
+  parser->data = data;
   parser->type = t;
   parser->state = (t == HTTP_REQUEST ? s_start_req : (t == HTTP_RESPONSE ? s_start_res : s_start_req_or_res));
-  parser->nread = 0;
-  parser->upgrade = 0;
-  parser->flags = 0;
-  parser->method = 0;
+  parser->http_errno = HPE_OK;
+}
+
+const char *
+http_errno_name(enum http_errno err) {
+  assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0])));
+  return http_strerror_tab[err].name;
+}
+
+const char *
+http_errno_description(enum http_errno err) {
+  assert(err < (sizeof(http_strerror_tab)/sizeof(http_strerror_tab[0])));
+  return http_strerror_tab[err].description;
+}
+
+int
+http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
+                      struct http_parser_url *u)
+{
+  enum state s;
+  const char *p;
+  enum http_parser_url_fields uf, old_uf;
+
+  u->port = u->field_set = 0;
+  s = is_connect ? s_req_host_start : s_req_spaces_before_url;
+  uf = old_uf = UF_MAX;
+
+  for (p = buf; p < buf + buflen; p++) {
+    s = parse_url_char(s, *p);
+
+    /* Figure out the next field that we're operating on */
+    switch (s) {
+      case s_dead:
+        return 1;
+
+      /* Skip delimeters */
+      case s_req_schema_slash:
+      case s_req_schema_slash_slash:
+      case s_req_host_start:
+      case s_req_host_v6_start:
+      case s_req_host_v6_end:
+      case s_req_port_start:
+      case s_req_query_string_start:
+      case s_req_fragment_start:
+        continue;
+
+      case s_req_schema:
+        uf = UF_SCHEMA;
+        break;
+
+      case s_req_host:
+      case s_req_host_v6:
+        uf = UF_HOST;
+        break;
+
+      case s_req_port:
+        uf = UF_PORT;
+        break;
+
+      case s_req_path:
+        uf = UF_PATH;
+        break;
+
+      case s_req_query_string:
+        uf = UF_QUERY;
+        break;
+
+      case s_req_fragment:
+        uf = UF_FRAGMENT;
+        break;
+
+      default:
+        assert(!"Unexpected state");
+        return 1;
+    }
+
+    /* Nothing's changed; soldier on */
+    if (uf == old_uf) {
+      u->field_data[uf].len++;
+      continue;
+    }
+
+    u->field_data[uf].off = p - buf;
+    u->field_data[uf].len = 1;
+
+    u->field_set |= (1 << uf);
+    old_uf = uf;
+  }
+
+  /* CONNECT requests can only contain "hostname:port" */
+  if (is_connect && u->field_set != ((1 << UF_HOST)|(1 << UF_PORT))) {
+    return 1;
+  }
+
+  /* Make sure we don't end somewhere unexpected */
+  switch (s) {
+  case s_req_host_v6_start:
+  case s_req_host_v6:
+  case s_req_host_v6_end:
+  case s_req_host:
+  case s_req_port_start:
+    return 1;
+  default:
+    break;
+  }
+
+  if (u->field_set & (1 << UF_PORT)) {
+    /* Don't bother with endp; we've already validated the string */
+    unsigned long v = strtoul(buf + u->field_data[UF_PORT].off, NULL, 10);
+
+    /* Ports have a max value of 2^16 */
+    if (v > 0xffff) {
+      return 1;
+    }
+
+    u->port = (uint16_t) v;
+  }
+
+  return 0;
+}
+
+void
+http_parser_pause(http_parser *parser, int paused) {
+  /* Users should only be pausing/unpausing a parser that is not in an error
+   * state. In non-debug builds, there's not much that we can do about this
+   * other than ignore it.
+   */
+  if (HTTP_PARSER_ERRNO(parser) == HPE_OK ||
+      HTTP_PARSER_ERRNO(parser) == HPE_PAUSED) {
+    SET_ERRNO((paused) ? HPE_PAUSED : HPE_OK);
+  } else {
+    assert(0 && "Attempting to pause parser in error state");
+  }
 }