source: draft-ietf-httpbis/latest/p1-messaging.xml @ 1773

Last change on this file since 1773 was 1773, checked in by julian.reschke@…, 10 years ago

Remove ABNF diagnostics for known to be non-referenced productions (header field names)

  • Property svn:eol-style set to native
  • Property svn:mime-type set to text/xml
File size: 251.2 KB
1<?xml version="1.0" encoding="utf-8"?>
2<?xml-stylesheet type='text/xsl' href='../myxml2rfc.xslt'?>
3<!DOCTYPE rfc [
4  <!ENTITY MAY "<bcp14 xmlns=''>MAY</bcp14>">
5  <!ENTITY MUST "<bcp14 xmlns=''>MUST</bcp14>">
6  <!ENTITY MUST-NOT "<bcp14 xmlns=''>MUST NOT</bcp14>">
7  <!ENTITY OPTIONAL "<bcp14 xmlns=''>OPTIONAL</bcp14>">
8  <!ENTITY RECOMMENDED "<bcp14 xmlns=''>RECOMMENDED</bcp14>">
9  <!ENTITY REQUIRED "<bcp14 xmlns=''>REQUIRED</bcp14>">
10  <!ENTITY SHALL "<bcp14 xmlns=''>SHALL</bcp14>">
11  <!ENTITY SHALL-NOT "<bcp14 xmlns=''>SHALL NOT</bcp14>">
12  <!ENTITY SHOULD "<bcp14 xmlns=''>SHOULD</bcp14>">
13  <!ENTITY SHOULD-NOT "<bcp14 xmlns=''>SHOULD NOT</bcp14>">
14  <!ENTITY ID-VERSION "latest">
15  <!ENTITY ID-MONTH "July">
16  <!ENTITY ID-YEAR "2012">
17  <!ENTITY mdash "&#8212;">
18  <!ENTITY Note "<x:h xmlns:x=''>Note:</x:h>">
19  <!ENTITY caching-overview       "<xref target='Part6' x:rel='#caching.overview' xmlns:x=''/>">
20  <!ENTITY cache-incomplete       "<xref target='Part6' x:rel='#response.cacheability' xmlns:x=''/>">
21  <!ENTITY payload                "<xref target='Part2' xmlns:x=''/>">
22  <!ENTITY media-types            "<xref target='Part2' x:rel='#media.types' xmlns:x=''/>">
23  <!ENTITY content-codings        "<xref target='Part2' x:rel='#content.codings' xmlns:x=''/>">
24  <!ENTITY CONNECT                "<xref target='Part2' x:rel='#CONNECT' xmlns:x=''/>">
25  <!ENTITY content.negotiation    "<xref target='Part2' x:rel='#content.negotiation' xmlns:x=''/>">
26  <!ENTITY diff-mime              "<xref target='Part2' x:rel='#differences.between.http.and.mime' xmlns:x=''/>">
27  <!ENTITY representation         "<xref target='Part2' x:rel='#representation' xmlns:x=''/>">
28  <!ENTITY header-allow           "<xref target='Part2' x:rel='#header.allow' xmlns:x=''/>">
29  <!ENTITY header-cache-control   "<xref target='Part6' x:rel='#header.cache-control' xmlns:x=''/>">
30  <!ENTITY header-content-encoding    "<xref target='Part2' x:rel='#header.content-encoding' xmlns:x=''/>">
31  <!ENTITY header-content-location    "<xref target='Part2' x:rel='#header.content-location' xmlns:x=''/>">
32  <!ENTITY header-content-range   "<xref target='Part5' x:rel='#header.content-range' xmlns:x=''/>">
33  <!ENTITY header-content-type    "<xref target='Part2' x:rel='#header.content-type' xmlns:x=''/>">
34  <!ENTITY header-date            "<xref target='Part2' x:rel='' xmlns:x=''/>">
35  <!ENTITY header-etag            "<xref target='Part4' x:rel='#header.etag' xmlns:x=''/>">
36  <!ENTITY header-expect          "<xref target='Part2' x:rel='#header.expect' xmlns:x=''/>">
37  <!ENTITY header-expires         "<xref target='Part6' x:rel='#header.expires' xmlns:x=''/>">
38  <!ENTITY header-last-modified   "<xref target='Part4' x:rel='#header.last-modified' xmlns:x=''/>">
39  <!ENTITY header-mime-version    "<xref target='Part2' x:rel='#mime-version' xmlns:x=''/>">
40  <!ENTITY header-pragma          "<xref target='Part6' x:rel='#header.pragma' xmlns:x=''/>">
41  <!ENTITY header-proxy-authenticate  "<xref target='Part7' x:rel='#header.proxy-authenticate' xmlns:x=''/>">
42  <!ENTITY header-proxy-authorization "<xref target='Part7' x:rel='#header.proxy-authorization' xmlns:x=''/>">
43  <!ENTITY header-server          "<xref target='Part2' x:rel='#header.server' xmlns:x=''/>">
44  <!ENTITY header-warning         "<xref target='Part6' x:rel='#header.warning' xmlns:x=''/>">
45  <!ENTITY idempotent-methods     "<xref target='Part2' x:rel='#idempotent.methods' xmlns:x=''/>">
46  <!ENTITY methods                "<xref target='Part2' x:rel='#methods' xmlns:x=''/>">
47  <!ENTITY OPTIONS                "<xref target='Part2' x:rel='#OPTIONS' xmlns:x=''/>">
48  <!ENTITY status-codes           "<xref target='Part2' x:rel='' xmlns:x=''/>">
49  <!ENTITY status-100             "<xref target='Part2' x:rel='#status.100' xmlns:x=''/>">
50  <!ENTITY status-1xx             "<xref target='Part2' x:rel='#status.1xx' xmlns:x=''/>">
51  <!ENTITY status-203             "<xref target='Part2' x:rel='#status.203' xmlns:x=''/>">
52  <!ENTITY status-3xx             "<xref target='Part2' x:rel='#status.3xx' xmlns:x=''/>">
53  <!ENTITY status-304             "<xref target='Part4' x:rel='#status.304' xmlns:x=''/>">
54  <!ENTITY status-4xx             "<xref target='Part2' x:rel='#status.4xx' xmlns:x=''/>">
55  <!ENTITY status-414             "<xref target='Part2' x:rel='#status.414' xmlns:x=''/>">
56  <!ENTITY cons-new-header-fields "<xref target='Part2' x:rel='#considerations.for.creating.header.fields' xmlns:x=''/>">
58<?rfc toc="yes" ?>
59<?rfc symrefs="yes" ?>
60<?rfc sortrefs="yes" ?>
61<?rfc compact="yes"?>
62<?rfc subcompact="no" ?>
63<?rfc linkmailto="no" ?>
64<?rfc editing="no" ?>
65<?rfc comments="yes"?>
66<?rfc inline="yes"?>
67<?rfc rfcedstyle="yes"?>
68<?rfc-ext allow-markup-in-artwork="yes" ?>
69<?rfc-ext include-references-in-index="yes" ?>
70<rfc obsoletes="2145,2616" updates="2817" category="std" x:maturity-level="proposed"
71     ipr="pre5378Trust200902" docName="draft-ietf-httpbis-p1-messaging-&ID-VERSION;"
72     xmlns:x=''>
73<x:link rel="next" basename="p2-semantics"/>
74<x:feedback template="{docname},%20%22{section}%22&amp;body=&lt;{ref}&gt;:"/>
77  <title abbrev="HTTP/1.1, Part 1">HTTP/1.1, part 1: URIs, Connections, and Message Parsing</title>
79  <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
80    <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
81    <address>
82      <postal>
83        <street>345 Park Ave</street>
84        <city>San Jose</city>
85        <region>CA</region>
86        <code>95110</code>
87        <country>USA</country>
88      </postal>
89      <email></email>
90      <uri></uri>
91    </address>
92  </author>
94  <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
95    <organization abbrev="W3C">World Wide Web Consortium</organization>
96    <address>
97      <postal>
98        <street>W3C / ERCIM</street>
99        <street>2004, rte des Lucioles</street>
100        <city>Sophia-Antipolis</city>
101        <region>AM</region>
102        <code>06902</code>
103        <country>France</country>
104      </postal>
105      <email></email>
106      <uri></uri>
107    </address>
108  </author>
110  <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
111    <organization abbrev="greenbytes">greenbytes GmbH</organization>
112    <address>
113      <postal>
114        <street>Hafenweg 16</street>
115        <city>Muenster</city><region>NW</region><code>48155</code>
116        <country>Germany</country>
117      </postal>
118      <email></email>
119      <uri></uri>
120    </address>
121  </author>
123  <date month="&ID-MONTH;" year="&ID-YEAR;"/>
124  <workgroup>HTTPbis Working Group</workgroup>
128   The Hypertext Transfer Protocol (HTTP) is an application-level protocol for
129   distributed, collaborative, hypertext information systems. HTTP has been in
130   use by the World Wide Web global information initiative since 1990. This
131   document is Part 1 of the seven-part specification that defines the protocol
132   referred to as "HTTP/1.1" and, taken together, obsoletes
133   <xref target="RFC2616" x:fmt="none">RFC 2616</xref> and moves it to historic
134   status, along with its predecessor <xref target="RFC2068" x:fmt="none">RFC
135   2068</xref>.
138   Part 1 provides an overview of HTTP and its associated terminology, defines
139   the "http" and "https" Uniform Resource Identifier (URI) schemes, defines
140   the generic message syntax and parsing requirements for HTTP message frames,
141   and describes general security concerns for implementations.
144   This part also obsoletes RFCs <xref target="RFC2145" x:fmt="none">2145</xref>
145   (on HTTP version numbers) and <xref target="RFC2817" x:fmt="none">2817</xref>
146   (on using CONNECT for TLS upgrades) and moves them to historic status.
150<note title="Editorial Note (To be removed by RFC Editor)">
151  <t>
152    Discussion of this draft takes place on the HTTPBIS working group
153    mailing list (, which is archived at
154    <eref target=""/>.
155  </t>
156  <t>
157    The current issues list is at
158    <eref target=""/> and related
159    documents (including fancy diffs) can be found at
160    <eref target=""/>.
161  </t>
162  <t>
163    The changes in this draft are summarized in <xref target="changes.since.19"/>.
164  </t>
168<section title="Introduction" anchor="introduction">
170   The Hypertext Transfer Protocol (HTTP) is an application-level
171   request/response protocol that uses extensible semantics and MIME-like
172   message payloads for flexible interaction with network-based hypertext
173   information systems. HTTP relies upon the Uniform Resource Identifier (URI)
174   standard <xref target="RFC3986"/> to indicate the target resource
175   (<xref target="target-resource"/>) and relationships between resources.
176   Messages are passed in a format similar to that used by Internet mail
177   <xref target="RFC5322"/> and the Multipurpose Internet Mail Extensions
178   (MIME) <xref target="RFC2045"/> (see &diff-mime; for the differences
179   between HTTP and MIME messages).
182   HTTP is a generic interface protocol for information systems. It is
183   designed to hide the details of how a service is implemented by presenting
184   a uniform interface to clients that is independent of the types of
185   resources provided. Likewise, servers do not need to be aware of each
186   client's purpose: an HTTP request can be considered in isolation rather
187   than being associated with a specific type of client or a predetermined
188   sequence of application steps. The result is a protocol that can be used
189   effectively in many different contexts and for which implementations can
190   evolve independently over time.
193   HTTP is also designed for use as an intermediation protocol for translating
194   communication to and from non-HTTP information systems.
195   HTTP proxies and gateways can provide access to alternative information
196   services by translating their diverse protocols into a hypertext
197   format that can be viewed and manipulated by clients in the same way
198   as HTTP services.
201   One consequence of HTTP flexibility is that the protocol cannot be
202   defined in terms of what occurs behind the interface. Instead, we
203   are limited to defining the syntax of communication, the intent
204   of received communication, and the expected behavior of recipients.
205   If the communication is considered in isolation, then successful
206   actions ought to be reflected in corresponding changes to the
207   observable interface provided by servers. However, since multiple
208   clients might act in parallel and perhaps at cross-purposes, we
209   cannot require that such changes be observable beyond the scope
210   of a single response.
213   This document is Part 1 of the seven-part specification of HTTP,
214   defining the protocol referred to as "HTTP/1.1", obsoleting
215   <xref target="RFC2616"/> and <xref target="RFC2145"/>.
216   Part 1 describes the architectural elements that are used or
217   referred to in HTTP, defines the "http" and "https" URI schemes,
218   describes overall network operation and connection management,
219   and defines HTTP message framing and forwarding requirements.
220   Our goal is to define all of the mechanisms necessary for HTTP message
221   handling that are independent of message semantics, thereby defining the
222   complete set of requirements for message parsers and
223   message-forwarding intermediaries.
226<section title="Requirement Notation" anchor="intro.requirements">
228   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
229   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
230   document are to be interpreted as described in <xref target="RFC2119"/>.
234<section title="Syntax Notation" anchor="notation">
235<iref primary="true" item="Grammar" subitem="ALPHA"/>
236<iref primary="true" item="Grammar" subitem="CR"/>
237<iref primary="true" item="Grammar" subitem="CRLF"/>
238<iref primary="true" item="Grammar" subitem="CTL"/>
239<iref primary="true" item="Grammar" subitem="DIGIT"/>
240<iref primary="true" item="Grammar" subitem="DQUOTE"/>
241<iref primary="true" item="Grammar" subitem="HEXDIG"/>
242<iref primary="true" item="Grammar" subitem="HTAB"/>
243<iref primary="true" item="Grammar" subitem="LF"/>
244<iref primary="true" item="Grammar" subitem="OCTET"/>
245<iref primary="true" item="Grammar" subitem="SP"/>
246<iref primary="true" item="Grammar" subitem="VCHAR"/>
248   This specification uses the Augmented Backus-Naur Form (ABNF) notation
249   of <xref target="RFC5234"/> with the list rule extension defined in
250   <xref target="abnf.extension"/>.  <xref target="collected.abnf"/> shows
251   the collected ABNF with the list rule expanded.
253<t anchor="core.rules">
254  <x:anchor-alias value="ALPHA"/>
255  <x:anchor-alias value="CTL"/>
256  <x:anchor-alias value="CR"/>
257  <x:anchor-alias value="CRLF"/>
258  <x:anchor-alias value="DIGIT"/>
259  <x:anchor-alias value="DQUOTE"/>
260  <x:anchor-alias value="HEXDIG"/>
261  <x:anchor-alias value="HTAB"/>
262  <x:anchor-alias value="LF"/>
263  <x:anchor-alias value="OCTET"/>
264  <x:anchor-alias value="SP"/>
265  <x:anchor-alias value="VCHAR"/>
266   The following core rules are included by
267   reference, as defined in <xref target="RFC5234" x:fmt="," x:sec="B.1"/>:
268   ALPHA (letters), CR (carriage return), CRLF (CR LF), CTL (controls),
269   DIGIT (decimal 0-9), DQUOTE (double quote),
270   HEXDIG (hexadecimal 0-9/A-F/a-f), HTAB (horizontal tab), LF (line feed),
271   OCTET (any 8-bit sequence of data), SP (space), and
272   VCHAR (any visible <xref target="USASCII"/> character).
275   As a convention, ABNF rule names prefixed with "obs-" denote
276   "obsolete" grammar rules that appear for historical reasons.
281<section title="Architecture" anchor="architecture">
283   HTTP was created for the World Wide Web architecture
284   and has evolved over time to support the scalability needs of a worldwide
285   hypertext system. Much of that architecture is reflected in the terminology
286   and syntax productions used to define HTTP.
289<section title="Client/Server Messaging" anchor="operation">
290<iref primary="true" item="client"/>
291<iref primary="true" item="server"/>
292<iref primary="true" item="connection"/>
294   HTTP is a stateless request/response protocol that operates by exchanging
295   <x:dfn>messages</x:dfn> (<xref target="http.message"/>) across a reliable
296   transport or session-layer
297   "<x:dfn>connection</x:dfn>". An HTTP "<x:dfn>client</x:dfn>" is a
298   program that establishes a connection to a server for the purpose of
299   sending one or more HTTP requests.  An HTTP "<x:dfn>server</x:dfn>" is a
300   program that accepts connections in order to service HTTP requests by
301   sending HTTP responses.
303<iref primary="true" item="user agent"/>
304<iref primary="true" item="origin server"/>
305<iref primary="true" item="browser"/>
306<iref primary="true" item="spider"/>
307<iref primary="true" item="sender"/>
308<iref primary="true" item="recipient"/>
310   The terms client and server refer only to the roles that
311   these programs perform for a particular connection.  The same program
312   might act as a client on some connections and a server on others.  We use
313   the term "<x:dfn>user agent</x:dfn>" to refer to the program that initiates a request,
314   such as a WWW browser, editor, or spider (web-traversing robot), and
315   the term "<x:dfn>origin server</x:dfn>" to refer to the program that can originate
316   authoritative responses to a request.  For general requirements, we use
317   the term "<x:dfn>sender</x:dfn>" to refer to whichever component sent a given message
318   and the term "<x:dfn>recipient</x:dfn>" to refer to any component that receives the
319   message.
322   Most HTTP communication consists of a retrieval request (GET) for
323   a representation of some resource identified by a URI.  In the
324   simplest case, this might be accomplished via a single bidirectional
325   connection (===) between the user agent (UA) and the origin server (O).
327<figure><artwork type="drawing">
328         request   &gt;
329    <x:highlight>UA</x:highlight> ======================================= <x:highlight>O</x:highlight>
330                                &lt;   response
332<iref primary="true" item="message"/>
333<iref primary="true" item="request"/>
334<iref primary="true" item="response"/>
336   A client sends an HTTP request to a server in the form of a <x:dfn>request</x:dfn>
337   message, beginning with a request-line that includes a method, URI, and
338   protocol version (<xref target="request.line"/>),
339   followed by MIME-like header fields containing
340   request modifiers, client information, and representation metadata
341   (<xref target="header.fields"/>),
342   an empty line to indicate the end of the header section, and finally
343   a message body containing the payload body (if any,
344   <xref target="message.body"/>).
347   A server responds to a client's request by sending one or more HTTP
348   <x:dfn>response</x:dfn>
349   messages, each beginning with a status line that
350   includes the protocol version, a success or error code, and textual
351   reason phrase (<xref target="status.line"/>),
352   possibly followed by MIME-like header fields containing server
353   information, resource metadata, and representation metadata
354   (<xref target="header.fields"/>),
355   an empty line to indicate the end of the header section, and finally
356   a message body containing the payload body (if any,
357   <xref target="message.body"/>).
360   The following example illustrates a typical message exchange for a
361   GET request on the URI "":
364client request:
365</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
366GET /hello.txt HTTP/1.1
367User-Agent: curl/7.16.3 libcurl/7.16.3 OpenSSL/0.9.7l zlib/1.2.3
369Accept: */*
373server response:
374</preamble><artwork type="message/http; msgtype=&#34;response&#34;" x:indent-with="  ">
375HTTP/1.1 200 OK
376Date: Mon, 27 Jul 2009 12:28:53 GMT
377Server: Apache
378Last-Modified: Wed, 22 Jul 2009 19:15:56 GMT
379ETag: "34aa387-d-1568eb00"
380Accept-Ranges: bytes
381Content-Length: <x:length-of target="exbody"/>
382Vary: Accept-Encoding
383Content-Type: text/plain
385<x:span anchor="exbody">Hello World!
389<section title="Implementation Diversity" anchor="implementation-diversity">
391   When considering the design of HTTP, it is easy to fall into a trap of
392   thinking that all user agents are general-purpose browsers and all origin
393   servers are large public websites. That is not the case in practice.
394   Common HTTP user agents include household appliances, stereos, scales,
395   software/firmware updaters, command-line programs, mobile apps,
396   and communication devices in a multitude of shapes and sizes.  Likewise,
397   common HTTP origin servers include home automation units, configurable
398   networking components, office machines, autonomous robots, news feeds,
399   traffic cameras, ad selectors, and video delivery platforms.
402   The term "user agent" does not imply that there is a human user directly
403   interacting with the software agent at the time of a request. In many
404   cases, a user agent is installed or configured to run in the background
405   and save its results for later inspection (or save only a subset of those
406   results that might be interesting or erroneous). Spiders, for example, are
407   typically given a start URI and configured to follow certain behavior while
408   crawling the Web as a hypertext graph.
411   The implementation diversity of HTTP means that we cannot assume the
412   user agent can make interactive suggestions to a user or provide adequate
413   warning for security or privacy options.  In the few cases where this
414   specification requires reporting of errors to the user, it is acceptable
415   for such reporting to only be visible in an error console or log file.
416   Likewise, requirements that an automated action be confirmed by the user
417   before proceeding can me met via advance configuration choices,
418   run-time options, or simply not proceeding with the unsafe action.
422<section title="Connections and Transport Independence" anchor="transport-independence">
424   HTTP messaging is independent of the underlying transport or
425   session-layer connection protocol(s).  HTTP only presumes a reliable
426   transport with in-order delivery of requests and the corresponding
427   in-order delivery of responses.  The mapping of HTTP request and
428   response structures onto the data units of the underlying transport
429   protocol is outside the scope of this specification.
432   The specific connection protocols to be used for an interaction
433   are determined by client configuration and the target URI
434   (<xref target="target-resource"/>).
435   For example, the "http" URI scheme
436   (<xref target="http.uri"/>) indicates a default connection of TCP
437   over IP, with a default TCP port of 80, but the client might be
438   configured to use a proxy via some other connection port or protocol
439   instead of using the defaults.
442   A connection might be used for multiple HTTP request/response exchanges,
443   as defined in <xref target="persistent.connections"/>.
447<section title="Intermediaries" anchor="intermediaries">
448<iref primary="true" item="intermediary"/>
450   HTTP enables the use of intermediaries to satisfy requests through
451   a chain of connections.  There are three common forms of HTTP
452   <x:dfn>intermediary</x:dfn>: proxy, gateway, and tunnel.  In some cases,
453   a single intermediary might act as an origin server, proxy, gateway,
454   or tunnel, switching behavior based on the nature of each request.
456<figure><artwork type="drawing">
457         &gt;             &gt;             &gt;             &gt;
458    <x:highlight>UA</x:highlight> =========== <x:highlight>A</x:highlight> =========== <x:highlight>B</x:highlight> =========== <x:highlight>C</x:highlight> =========== <x:highlight>O</x:highlight>
459               &lt;             &lt;             &lt;             &lt;
462   The figure above shows three intermediaries (A, B, and C) between the
463   user agent and origin server. A request or response message that
464   travels the whole chain will pass through four separate connections.
465   Some HTTP communication options
466   might apply only to the connection with the nearest, non-tunnel
467   neighbor, only to the end-points of the chain, or to all connections
468   along the chain. Although the diagram is linear, each participant might
469   be engaged in multiple, simultaneous communications. For example, B
470   might be receiving requests from many clients other than A, and/or
471   forwarding requests to servers other than C, at the same time that it
472   is handling A's request.
475<iref primary="true" item="upstream"/><iref primary="true" item="downstream"/>
476<iref primary="true" item="inbound"/><iref primary="true" item="outbound"/>
477   We use the terms "<x:dfn>upstream</x:dfn>" and "<x:dfn>downstream</x:dfn>"
478   to describe various requirements in relation to the directional flow of a
479   message: all messages flow from upstream to downstream.
480   Likewise, we use the terms inbound and outbound to refer to
481   directions in relation to the request path:
482   "<x:dfn>inbound</x:dfn>" means toward the origin server and
483   "<x:dfn>outbound</x:dfn>" means toward the user agent.
485<t><iref primary="true" item="proxy"/>
486   A "<x:dfn>proxy</x:dfn>" is a message forwarding agent that is selected by the
487   client, usually via local configuration rules, to receive requests
488   for some type(s) of absolute URI and attempt to satisfy those
489   requests via translation through the HTTP interface.  Some translations
490   are minimal, such as for proxy requests for "http" URIs, whereas
491   other requests might require translation to and from entirely different
492   application-layer protocols. Proxies are often used to group an
493   organization's HTTP requests through a common intermediary for the
494   sake of security, annotation services, or shared caching.
497<iref primary="true" item="transforming proxy"/>
498<iref primary="true" item="non-transforming proxy"/>
499   An HTTP-to-HTTP proxy is called a "<x:dfn>transforming proxy</x:dfn>" if it is designed
500   or configured to modify request or response messages in a semantically
501   meaningful way (i.e., modifications, beyond those required by normal
502   HTTP processing, that change the message in a way that would be
503   significant to the original sender or potentially significant to
504   downstream recipients).  For example, a transforming proxy might be
505   acting as a shared annotation server (modifying responses to include
506   references to a local annotation database), a malware filter, a
507   format transcoder, or an intranet-to-Internet privacy filter.  Such
508   transformations are presumed to be desired by the client (or client
509   organization) that selected the proxy and are beyond the scope of
510   this specification.  However, when a proxy is not intended to transform
511   a given message, we use the term "<x:dfn>non-transforming proxy</x:dfn>" to target
512   requirements that preserve HTTP message semantics. See &status-203; and
513   &header-warning; for status and warning codes related to transformations.
515<t><iref primary="true" item="gateway"/><iref primary="true" item="reverse proxy"/>
516<iref primary="true" item="accelerator"/>
517   A "<x:dfn>gateway</x:dfn>" (a.k.a., "<x:dfn>reverse proxy</x:dfn>")
518   is a receiving agent that acts
519   as a layer above some other server(s) and translates the received
520   requests to the underlying server's protocol.  Gateways are often
521   used to encapsulate legacy or untrusted information services, to
522   improve server performance through "<x:dfn>accelerator</x:dfn>" caching, and to
523   enable partitioning or load-balancing of HTTP services across
524   multiple machines.
527   A gateway behaves as an origin server on its outbound connection and
528   as a user agent on its inbound connection.
529   All HTTP requirements applicable to an origin server
530   also apply to the outbound communication of a gateway.
531   A gateway communicates with inbound servers using any protocol that
532   it desires, including private extensions to HTTP that are outside
533   the scope of this specification.  However, an HTTP-to-HTTP gateway
534   that wishes to interoperate with third-party HTTP servers &MUST;
535   conform to HTTP user agent requirements on the gateway's inbound
536   connection and &MUST; implement the <x:ref>Connection</x:ref>
537   (<xref target="header.connection"/>) and <x:ref>Via</x:ref>
538   (<xref target="header.via"/>) header fields for both connections.
540<t><iref primary="true" item="tunnel"/>
541   A "<x:dfn>tunnel</x:dfn>" acts as a blind relay between two connections
542   without changing the messages. Once active, a tunnel is not
543   considered a party to the HTTP communication, though the tunnel might
544   have been initiated by an HTTP request. A tunnel ceases to exist when
545   both ends of the relayed connection are closed. Tunnels are used to
546   extend a virtual connection through an intermediary, such as when
547   transport-layer security is used to establish private communication
548   through a shared firewall proxy.
550<t><iref primary="true" item="interception proxy"/>
551<iref primary="true" item="transparent proxy"/>
552<iref primary="true" item="captive portal"/>
553   The above categories for intermediary only consider those acting as
554   participants in the HTTP communication.  There are also intermediaries
555   that can act on lower layers of the network protocol stack, filtering or
556   redirecting HTTP traffic without the knowledge or permission of message
557   senders. Network intermediaries often introduce security flaws or
558   interoperability problems by violating HTTP semantics.  For example, an
559   "<x:dfn>interception proxy</x:dfn>" <xref target="RFC3040"/> (also commonly
560   known as a "<x:dfn>transparent proxy</x:dfn>" <xref target="RFC1919"/> or
561   "<x:dfn>captive portal</x:dfn>")
562   differs from an HTTP proxy because it is not selected by the client.
563   Instead, an interception proxy filters or redirects outgoing TCP port 80
564   packets (and occasionally other common port traffic).
565   Interception proxies are commonly found on public network access points,
566   as a means of enforcing account subscription prior to allowing use of
567   non-local Internet services, and within corporate firewalls to enforce
568   network usage policies.
569   They are indistinguishable from a man-in-the-middle attack.
572   HTTP is defined as a stateless protocol, meaning that each request message
573   can be understood in isolation.  Many implementations depend on HTTP's
574   stateless design in order to reuse proxied connections or dynamically
575   load balance requests across multiple servers.  Hence, servers &MUST-NOT;
576   assume that two requests on the same connection are from the same user
577   agent unless the connection is secured and specific to that agent.
578   Some non-standard HTTP extensions (e.g., <xref target="RFC4559"/>) have
579   been known to violate this requirement, resulting in security and
580   interoperability problems.
584<section title="Caches" anchor="caches">
585<iref primary="true" item="cache"/>
587   A "<x:dfn>cache</x:dfn>" is a local store of previous response messages and the
588   subsystem that controls its message storage, retrieval, and deletion.
589   A cache stores cacheable responses in order to reduce the response
590   time and network bandwidth consumption on future, equivalent
591   requests. Any client or server &MAY; employ a cache, though a cache
592   cannot be used by a server while it is acting as a tunnel.
595   The effect of a cache is that the request/response chain is shortened
596   if one of the participants along the chain has a cached response
597   applicable to that request. The following illustrates the resulting
598   chain if B has a cached copy of an earlier response from O (via C)
599   for a request which has not been cached by UA or A.
601<figure><artwork type="drawing">
602            &gt;             &gt;
603       <x:highlight>UA</x:highlight> =========== <x:highlight>A</x:highlight> =========== <x:highlight>B</x:highlight> - - - - - - <x:highlight>C</x:highlight> - - - - - - <x:highlight>O</x:highlight>
604                  &lt;             &lt;
606<t><iref primary="true" item="cacheable"/>
607   A response is "<x:dfn>cacheable</x:dfn>" if a cache is allowed to store a copy of
608   the response message for use in answering subsequent requests.
609   Even when a response is cacheable, there might be additional
610   constraints placed by the client or by the origin server on when
611   that cached response can be used for a particular request. HTTP
612   requirements for cache behavior and cacheable responses are
613   defined in &caching-overview;. 
616   There are a wide variety of architectures and configurations
617   of caches and proxies deployed across the World Wide Web and
618   inside large organizations. These systems include national hierarchies
619   of proxy caches to save transoceanic bandwidth, systems that
620   broadcast or multicast cache entries, organizations that distribute
621   subsets of cached data via optical media, and so on.
625<section title="Conformance and Error Handling" anchor="intro.conformance.and.error.handling">
627   This specification targets conformance criteria according to the role of
628   a participant in HTTP communication.  Hence, HTTP requirements are placed
629   on senders, recipients, clients, servers, user agents, intermediaries,
630   origin servers, proxies, gateways, or caches, depending on what behavior
631   is being constrained by the requirement.
634   The verb "generate" is used instead of "send" where a requirement
635   differentiates between creating a protocol element and merely forwarding a
636   received element downstream.
639   An implementation is considered conformant if it complies with all of the
640   requirements associated with the roles it partakes in HTTP. Note that
641   SHOULD-level requirements are relevant here, unless one of the documented
642   exceptions is applicable.
645   In addition to the prose requirements placed upon them, senders &MUST-NOT;
646   generate protocol elements that do not match the grammar defined by the
647   ABNF rules for those protocol elements that are applicable to the sender's
648   role. If a received protocol element is processed, the recipient &MUST; be
649   able to parse any value that would match the ABNF rules for that protocol
650   element, excluding only those rules not applicable to the recipient's role.
653   Unless noted otherwise, a recipient &MAY; attempt to recover a usable
654   protocol element from an invalid construct.  HTTP does not define
655   specific error handling mechanisms except when they have a direct impact
656   on security, since different applications of the protocol require
657   different error handling strategies.  For example, a Web browser might
658   wish to transparently recover from a response where the
659   <x:ref>Location</x:ref> header field doesn't parse according to the ABNF,
660   whereas a systems control client might consider any form of error recovery
661   to be dangerous.
665<section title="Protocol Versioning" anchor="http.version">
666  <x:anchor-alias value="HTTP-version"/>
667  <x:anchor-alias value="HTTP-name"/>
669   HTTP uses a "&lt;major&gt;.&lt;minor&gt;" numbering scheme to indicate
670   versions of the protocol. This specification defines version "1.1".
671   The protocol version as a whole indicates the sender's conformance
672   with the set of requirements laid out in that version's corresponding
673   specification of HTTP.
676   The version of an HTTP message is indicated by an HTTP-version field
677   in the first line of the message. HTTP-version is case-sensitive.
679<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-version"/><iref primary="true" item="Grammar" subitem="HTTP-name"/>
680  <x:ref>HTTP-version</x:ref>  = <x:ref>HTTP-name</x:ref> "/" <x:ref>DIGIT</x:ref> "." <x:ref>DIGIT</x:ref>
681  <x:ref>HTTP-name</x:ref>     = <x:abnf-char-sequence>"HTTP"</x:abnf-char-sequence> ; "HTTP", case-sensitive
684   The HTTP version number consists of two decimal digits separated by a "."
685   (period or decimal point).  The first digit ("major version") indicates the
686   HTTP messaging syntax, whereas the second digit ("minor version") indicates
687   the highest minor version to which the sender is
688   conformant and able to understand for future communication.  The minor
689   version advertises the sender's communication capabilities even when the
690   sender is only using a backwards-compatible subset of the protocol,
691   thereby letting the recipient know that more advanced features can
692   be used in response (by servers) or in future requests (by clients).
695   When an HTTP/1.1 message is sent to an HTTP/1.0 recipient
696   <xref target="RFC1945"/> or a recipient whose version is unknown,
697   the HTTP/1.1 message is constructed such that it can be interpreted
698   as a valid HTTP/1.0 message if all of the newer features are ignored.
699   This specification places recipient-version requirements on some
700   new features so that a conformant sender will only use compatible
701   features until it has determined, through configuration or the
702   receipt of a message, that the recipient supports HTTP/1.1.
705   The interpretation of a header field does not change between minor
706   versions of the same major HTTP version, though the default
707   behavior of a recipient in the absence of such a field can change.
708   Unless specified otherwise, header fields defined in HTTP/1.1 are
709   defined for all versions of HTTP/1.x.  In particular, the <x:ref>Host</x:ref>
710   and <x:ref>Connection</x:ref> header fields ought to be implemented by all
711   HTTP/1.x implementations whether or not they advertise conformance with
712   HTTP/1.1.
715   New header fields can be defined such that, when they are
716   understood by a recipient, they might override or enhance the
717   interpretation of previously defined header fields.  When an
718   implementation receives an unrecognized header field, the recipient
719   &MUST; ignore that header field for local processing regardless of
720   the message's HTTP version.  An unrecognized header field received
721   by a proxy &MUST; be forwarded downstream unless the header field's
722   field-name is listed in the message's <x:ref>Connection</x:ref> header field
723   (see <xref target="header.connection"/>).
724   These requirements allow HTTP's functionality to be enhanced without
725   requiring prior update of deployed intermediaries.
728   Intermediaries that process HTTP messages (i.e., all intermediaries
729   other than those acting as tunnels) &MUST; send their own HTTP-version
730   in forwarded messages.  In other words, they &MUST-NOT; blindly
731   forward the first line of an HTTP message without ensuring that the
732   protocol version in that message matches a version to which that
733   intermediary is conformant for both the receiving and
734   sending of messages.  Forwarding an HTTP message without rewriting
735   the HTTP-version might result in communication errors when downstream
736   recipients use the message sender's version to determine what features
737   are safe to use for later communication with that sender.
740   An HTTP client &SHOULD; send a request version equal to the highest
741   version to which the client is conformant and
742   whose major version is no higher than the highest version supported
743   by the server, if this is known.  An HTTP client &MUST-NOT; send a
744   version to which it is not conformant.
747   An HTTP client &MAY; send a lower request version if it is known that
748   the server incorrectly implements the HTTP specification, but only
749   after the client has attempted at least one normal request and determined
750   from the response status or header fields (e.g., <x:ref>Server</x:ref>) that
751   the server improperly handles higher request versions.
754   An HTTP server &SHOULD; send a response version equal to the highest
755   version to which the server is conformant and
756   whose major version is less than or equal to the one received in the
757   request.  An HTTP server &MUST-NOT; send a version to which it is not
758   conformant.  A server &MAY; send a <x:ref>505 (HTTP Version Not
759   Supported)</x:ref> response if it cannot send a response using the
760   major version used in the client's request.
763   An HTTP server &MAY; send an HTTP/1.0 response to an HTTP/1.0 request
764   if it is known or suspected that the client incorrectly implements the
765   HTTP specification and is incapable of correctly processing later
766   version responses, such as when a client fails to parse the version
767   number correctly or when an intermediary is known to blindly forward
768   the HTTP-version even when it doesn't conform to the given minor
769   version of the protocol. Such protocol downgrades &SHOULD-NOT; be
770   performed unless triggered by specific client attributes, such as when
771   one or more of the request header fields (e.g., <x:ref>User-Agent</x:ref>)
772   uniquely match the values sent by a client known to be in error.
775   The intention of HTTP's versioning design is that the major number
776   will only be incremented if an incompatible message syntax is
777   introduced, and that the minor number will only be incremented when
778   changes made to the protocol have the effect of adding to the message
779   semantics or implying additional capabilities of the sender.  However,
780   the minor version was not incremented for the changes introduced between
781   <xref target="RFC2068"/> and <xref target="RFC2616"/>, and this revision
782   is specifically avoiding any such changes to the protocol.
786<section title="Uniform Resource Identifiers" anchor="uri">
787<iref primary="true" item="resource"/>
789   Uniform Resource Identifiers (URIs) <xref target="RFC3986"/> are used
790   throughout HTTP as the means for identifying resources. URI references
791   are used to target requests, indicate redirects, and define relationships.
792   HTTP does not limit what a resource might be; it merely defines an interface
793   that can be used to interact with a resource via HTTP. More information on
794   the scope of URIs and resources can be found in <xref target="RFC3986"/>.
796  <x:anchor-alias value="URI-reference"/>
797  <x:anchor-alias value="absolute-URI"/>
798  <x:anchor-alias value="relative-part"/>
799  <x:anchor-alias value="authority"/>
800  <x:anchor-alias value="path-abempty"/>
801  <x:anchor-alias value="path-absolute"/>
802  <x:anchor-alias value="port"/>
803  <x:anchor-alias value="query"/>
804  <x:anchor-alias value="uri-host"/>
805  <x:anchor-alias value="partial-URI"/>
807   This specification adopts the definitions of "URI-reference",
808   "absolute-URI", "relative-part", "port", "host",
809   "path-abempty", "path-absolute", "query", and "authority" from the
810   URI generic syntax <xref target="RFC3986"/>.
811   In addition, we define a partial-URI rule for protocol elements
812   that allow a relative URI but not a fragment.
814<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="URI-reference"/><iref primary="true" item="Grammar" subitem="absolute-URI"/><iref primary="true" item="Grammar" subitem="authority"/><iref primary="true" item="Grammar" subitem="path-absolute"/><iref primary="true" item="Grammar" subitem="port"/><iref primary="true" item="Grammar" subitem="query"/><iref primary="true" item="Grammar" subitem="uri-host"/>
815  <x:ref>URI-reference</x:ref> = &lt;URI-reference, defined in <xref target="RFC3986" x:fmt="," x:sec="4.1"/>&gt;
816  <x:ref>absolute-URI</x:ref>  = &lt;absolute-URI, defined in <xref target="RFC3986" x:fmt="," x:sec="4.3"/>&gt;
817  <x:ref>relative-part</x:ref> = &lt;relative-part, defined in <xref target="RFC3986" x:fmt="," x:sec="4.2"/>&gt;
818  <x:ref>authority</x:ref>     = &lt;authority, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2"/>&gt;
819  <x:ref>path-abempty</x:ref>  = &lt;path-abempty, defined in <xref target="RFC3986" x:fmt="," x:sec="3.3"/>&gt;
820  <x:ref>path-absolute</x:ref> = &lt;path-absolute, defined in <xref target="RFC3986" x:fmt="," x:sec="3.3"/>&gt;
821  <x:ref>port</x:ref>          = &lt;port, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2.3"/>&gt;
822  <x:ref>query</x:ref>         = &lt;query, defined in <xref target="RFC3986" x:fmt="," x:sec="3.4"/>&gt;
823  <x:ref>uri-host</x:ref>      = &lt;host, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2.2"/>&gt;
825  <x:ref>partial-URI</x:ref>   = relative-part [ "?" query ]
828   Each protocol element in HTTP that allows a URI reference will indicate
829   in its ABNF production whether the element allows any form of reference
830   (URI-reference), only a URI in absolute form (absolute-URI), only the
831   path and optional query components, or some combination of the above.
832   Unless otherwise indicated, URI references are parsed
833   relative to the effective request URI
834   (<xref target="effective.request.uri"/>).
837<section title="http URI scheme" anchor="http.uri">
838  <x:anchor-alias value="http-URI"/>
839  <iref item="http URI scheme" primary="true"/>
840  <iref item="URI scheme" subitem="http" primary="true"/>
842   The "http" URI scheme is hereby defined for the purpose of minting
843   identifiers according to their association with the hierarchical
844   namespace governed by a potential HTTP origin server listening for
845   TCP connections on a given port.
847<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="http-URI"/>
848  <x:ref>http-URI</x:ref> = "http:" "//" <x:ref>authority</x:ref> <x:ref>path-abempty</x:ref> [ "?" <x:ref>query</x:ref> ]
851   The HTTP origin server is identified by the generic syntax's
852   <x:ref>authority</x:ref> component, which includes a host identifier
853   and optional TCP port (<xref target="RFC3986" x:fmt="," x:sec="3.2.2"/>).
854   The remainder of the URI, consisting of both the hierarchical path
855   component and optional query component, serves as an identifier for
856   a potential resource within that origin server's name space.
859   If the host identifier is provided as an IP literal or IPv4 address,
860   then the origin server is any listener on the indicated TCP port at
861   that IP address. If host is a registered name, then that name is
862   considered an indirect identifier and the recipient might use a name
863   resolution service, such as DNS, to find the address of a listener
864   for that host.
865   The host &MUST-NOT; be empty; if an "http" URI is received with an
866   empty host, then it &MUST; be rejected as invalid.
867   If the port subcomponent is empty or not given, then TCP port 80 is
868   assumed (the default reserved port for WWW services).
871   Regardless of the form of host identifier, access to that host is not
872   implied by the mere presence of its name or address. The host might or might
873   not exist and, even when it does exist, might or might not be running an
874   HTTP server or listening to the indicated port. The "http" URI scheme
875   makes use of the delegated nature of Internet names and addresses to
876   establish a naming authority (whatever entity has the ability to place
877   an HTTP server at that Internet name or address) and allows that
878   authority to determine which names are valid and how they might be used.
881   When an "http" URI is used within a context that calls for access to the
882   indicated resource, a client &MAY; attempt access by resolving
883   the host to an IP address, establishing a TCP connection to that address
884   on the indicated port, and sending an HTTP request message
885   (<xref target="http.message"/>) containing the URI's identifying data
886   (<xref target="message.routing"/>) to the server.
887   If the server responds to that request with a non-interim HTTP response
888   message, as described in &status-codes;, then that response
889   is considered an authoritative answer to the client's request.
892   Although HTTP is independent of the transport protocol, the "http"
893   scheme is specific to TCP-based services because the name delegation
894   process depends on TCP for establishing authority.
895   An HTTP service based on some other underlying connection protocol
896   would presumably be identified using a different URI scheme, just as
897   the "https" scheme (below) is used for servers that require an SSL/TLS
898   transport layer on a connection. Other protocols might also be used to
899   provide access to "http" identified resources &mdash; it is only the
900   authoritative interface used for mapping the namespace that is
901   specific to TCP.
904   The URI generic syntax for authority also includes a deprecated
905   userinfo subcomponent (<xref target="RFC3986" x:fmt="," x:sec="3.2.1"/>)
906   for including user authentication information in the URI.  Some
907   implementations make use of the userinfo component for internal
908   configuration of authentication information, such as within command
909   invocation options, configuration files, or bookmark lists, even
910   though such usage might expose a user identifier or password.
911   Senders &MUST-NOT; include a userinfo subcomponent (and its "@"
912   delimiter) when transmitting an "http" URI in a message.  Recipients
913   of HTTP messages that contain a URI reference &SHOULD; parse for the
914   existence of userinfo and treat its presence as an error, likely
915   indicating that the deprecated subcomponent is being used to obscure
916   the authority for the sake of phishing attacks.
920<section title="https URI scheme" anchor="https.uri">
921   <x:anchor-alias value="https-URI"/>
922   <iref item="https URI scheme"/>
923   <iref item="URI scheme" subitem="https"/>
925   The "https" URI scheme is hereby defined for the purpose of minting
926   identifiers according to their association with the hierarchical
927   namespace governed by a potential HTTP origin server listening for
928   SSL/TLS-secured connections on a given TCP port.
931   All of the requirements listed above for the "http" scheme are also
932   requirements for the "https" scheme, except that a default TCP port
933   of 443 is assumed if the port subcomponent is empty or not given,
934   and the TCP connection &MUST; be secured for privacy through the
935   use of strong encryption prior to sending the first HTTP request.
937<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="https-URI"/>
938  <x:ref>https-URI</x:ref> = "https:" "//" <x:ref>authority</x:ref> <x:ref>path-abempty</x:ref> [ "?" <x:ref>query</x:ref> ]
941   Unlike the "http" scheme, responses to "https" identified requests
942   are never "public" and thus &MUST-NOT; be reused for shared caching.
943   They can, however, be reused in a private cache if the message is
944   cacheable by default in HTTP or specifically indicated as such by
945   the Cache-Control header field (&header-cache-control;).
948   Resources made available via the "https" scheme have no shared
949   identity with the "http" scheme even if their resource identifiers
950   indicate the same authority (the same host listening to the same
951   TCP port).  They are distinct name spaces and are considered to be
952   distinct origin servers.  However, an extension to HTTP that is
953   defined to apply to entire host domains, such as the Cookie protocol
954   <xref target="RFC6265"/>, can allow information
955   set by one service to impact communication with other services
956   within a matching group of host domains.
959   The process for authoritative access to an "https" identified
960   resource is defined in <xref target="RFC2818"/>.
964<section title="http and https URI Normalization and Comparison" anchor="uri.comparison">
966   Since the "http" and "https" schemes conform to the URI generic syntax,
967   such URIs are normalized and compared according to the algorithm defined
968   in <xref target="RFC3986" x:fmt="," x:sec="6"/>, using the defaults
969   described above for each scheme.
972   If the port is equal to the default port for a scheme, the normal
973   form is to elide the port subcomponent. Likewise, an empty path
974   component is equivalent to an absolute path of "/", so the normal
975   form is to provide a path of "/" instead. The scheme and host
976   are case-insensitive and normally provided in lowercase; all
977   other components are compared in a case-sensitive manner.
978   Characters other than those in the "reserved" set are equivalent
979   to their percent-encoded octets (see <xref target="RFC3986"
980   x:fmt="," x:sec="2.1"/>): the normal form is to not encode them.
983   For example, the following three URIs are equivalent:
985<figure><artwork type="example">
994<section title="Message Format" anchor="http.message">
995<x:anchor-alias value="generic-message"/>
996<x:anchor-alias value="message.types"/>
997<x:anchor-alias value="HTTP-message"/>
998<x:anchor-alias value="start-line"/>
999<iref item="header section"/>
1000<iref item="headers"/>
1001<iref item="header field"/>
1003   All HTTP/1.1 messages consist of a start-line followed by a sequence of
1004   octets in a format similar to the Internet Message Format
1005   <xref target="RFC5322"/>: zero or more header fields (collectively
1006   referred to as the "headers" or the "header section"), an empty line
1007   indicating the end of the header section, and an optional message body.
1009<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-message"/>
1010  <x:ref>HTTP-message</x:ref>   = <x:ref>start-line</x:ref>
1011                   *( <x:ref>header-field</x:ref> <x:ref>CRLF</x:ref> )
1012                   <x:ref>CRLF</x:ref>
1013                   [ <x:ref>message-body</x:ref> ]
1016   The normal procedure for parsing an HTTP message is to read the
1017   start-line into a structure, read each header field into a hash
1018   table by field name until the empty line, and then use the parsed
1019   data to determine if a message body is expected.  If a message body
1020   has been indicated, then it is read as a stream until an amount
1021   of octets equal to the message body length is read or the connection
1022   is closed.
1025   Recipients &MUST; parse an HTTP message as a sequence of octets in an
1026   encoding that is a superset of US-ASCII <xref target="USASCII"/>.
1027   Parsing an HTTP message as a stream of Unicode characters, without regard
1028   for the specific encoding, creates security vulnerabilities due to the
1029   varying ways that string processing libraries handle invalid multibyte
1030   character sequences that contain the octet LF (%x0A).  String-based
1031   parsers can only be safely used within protocol elements after the element
1032   has been extracted from the message, such as within a header field-value
1033   after message parsing has delineated the individual fields.
1036   An HTTP message can be parsed as a stream for incremental processing or
1037   forwarding downstream.  However, recipients cannot rely on incremental
1038   delivery of partial messages, since some implementations will buffer or
1039   delay message forwarding for the sake of network efficiency, security
1040   checks, or payload transformations.
1043<section title="Start Line" anchor="start.line">
1044  <x:anchor-alias value="Start-Line"/>
1046   An HTTP message can either be a request from client to server or a
1047   response from server to client.  Syntactically, the two types of message
1048   differ only in the start-line, which is either a request-line (for requests)
1049   or a status-line (for responses), and in the algorithm for determining
1050   the length of the message body (<xref target="message.body"/>).
1051   In theory, a client could receive requests and a server could receive
1052   responses, distinguishing them by their different start-line formats,
1053   but in practice servers are implemented to only expect a request
1054   (a response is interpreted as an unknown or invalid request method)
1055   and clients are implemented to only expect a response.
1057<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="start-line"/>
1058  <x:ref>start-line</x:ref>     = <x:ref>request-line</x:ref> / <x:ref>status-line</x:ref>
1061   Implementations &MUST-NOT; send whitespace between the start-line and
1062   the first header field. The presence of such whitespace in a request
1063   might be an attempt to trick a server into ignoring that field or
1064   processing the line after it as a new request, either of which might
1065   result in a security vulnerability if other implementations within
1066   the request chain interpret the same message differently.
1067   Likewise, the presence of such whitespace in a response might be
1068   ignored by some clients or cause others to cease parsing.
1071<section title="Request Line" anchor="request.line">
1072  <x:anchor-alias value="Request"/>
1073  <x:anchor-alias value="request-line"/>
1075   A request-line begins with a method token, followed by a single
1076   space (SP), the request-target, another single space (SP), the
1077   protocol version, and ending with CRLF.
1079<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="request-line"/>
1080  <x:ref>request-line</x:ref>   = <x:ref>method</x:ref> <x:ref>SP</x:ref> <x:ref>request-target</x:ref> <x:ref>SP</x:ref> <x:ref>HTTP-version</x:ref> <x:ref>CRLF</x:ref>
1083   A server &MUST; be able to parse any received message that begins
1084   with a request-line and matches the ABNF rule for HTTP-message.
1086<iref primary="true" item="method"/>
1087<t anchor="method">
1088   The method token indicates the request method to be performed on the
1089   target resource. The request method is case-sensitive.
1091<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="method"/>
1092  <x:ref>method</x:ref>         = <x:ref>token</x:ref>
1095   The methods defined by this specification can be found in
1096   &methods;, along with information regarding the HTTP method registry
1097   and considerations for defining new methods.
1099<iref item="request-target"/>
1101   The request-target identifies the target resource upon which to apply
1102   the request, as defined in <xref target="request-target"/>.
1105   No whitespace is allowed inside the method, request-target, and
1106   protocol version.  Hence, recipients typically parse the request-line
1107   into its component parts by splitting on the SP characters.
1110   Unfortunately, some user agents fail to properly encode hypertext
1111   references that have embedded whitespace, sending the characters
1112   directly instead of properly percent-encoding the disallowed characters.
1113   Recipients of an invalid request-line &SHOULD; respond with either a
1114   <x:ref>400 (Bad Request)</x:ref> error or a <x:ref>301 (Moved Permanently)</x:ref>
1115   redirect with the request-target properly encoded.  Recipients &SHOULD-NOT;
1116   attempt to autocorrect and then process the request without a redirect,
1117   since the invalid request-line might be deliberately crafted to bypass
1118   security filters along the request chain.
1121   HTTP does not place a pre-defined limit on the length of a request-line.
1122   A server that receives a method longer than any that it implements
1123   &SHOULD; respond with either a <x:ref>405 (Method Not Allowed)</x:ref>, if it is an origin
1124   server, or a <x:ref>501 (Not Implemented)</x:ref> status code.
1125   A server &MUST; be prepared to receive URIs of unbounded length and
1126   respond with the <x:ref>414 (URI Too Long)</x:ref> status code if the received
1127   request-target would be longer than the server wishes to handle
1128   (see &status-414;).
1131   Various ad-hoc limitations on request-line length are found in practice.
1132   It is &RECOMMENDED; that all HTTP senders and recipients support, at a
1133   minimum, request-line lengths of up to 8000 octets.
1137<section title="Status Line" anchor="status.line">
1138  <x:anchor-alias value="response"/>
1139  <x:anchor-alias value="status-line"/>
1140  <x:anchor-alias value="status-code"/>
1141  <x:anchor-alias value="reason-phrase"/>
1143   The first line of a response message is the status-line, consisting
1144   of the protocol version, a space (SP), the status code, another space,
1145   a possibly-empty textual phrase describing the status code, and
1146   ending with CRLF.
1148<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="status-line"/>
1149  <x:ref>status-line</x:ref> = <x:ref>HTTP-version</x:ref> <x:ref>SP</x:ref> <x:ref>status-code</x:ref> <x:ref>SP</x:ref> <x:ref>reason-phrase</x:ref> <x:ref>CRLF</x:ref>
1152   A client &MUST; be able to parse any received message that begins
1153   with a status-line and matches the ABNF rule for HTTP-message.
1156   The status-code element is a 3-digit integer code describing the
1157   result of the server's attempt to understand and satisfy the client's
1158   corresponding request. The rest of the response message is to be
1159   interpreted in light of the semantics defined for that status code.
1160   See &status-codes; for information about the semantics of status codes,
1161   including the classes of status code (indicated by the first digit),
1162   the status codes defined by this specification, considerations for the
1163   definition of new status codes, and the IANA registry.
1165<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="status-code"/>
1166  <x:ref>status-code</x:ref>    = 3<x:ref>DIGIT</x:ref>
1169   The reason-phrase element exists for the sole purpose of providing a
1170   textual description associated with the numeric status code, mostly
1171   out of deference to earlier Internet application protocols that were more
1172   frequently used with interactive text clients. A client &SHOULD; ignore
1173   the reason-phrase content.
1175<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="reason-phrase"/>
1176  <x:ref>reason-phrase</x:ref>  = *( <x:ref>HTAB</x:ref> / <x:ref>SP</x:ref> / <x:ref>VCHAR</x:ref> / <x:ref>obs-text</x:ref> )
1181<section title="Header Fields" anchor="header.fields">
1182  <x:anchor-alias value="header-field"/>
1183  <x:anchor-alias value="field-content"/>
1184  <x:anchor-alias value="field-name"/>
1185  <x:anchor-alias value="field-value"/>
1186  <x:anchor-alias value="obs-fold"/>
1188   Each HTTP header field consists of a case-insensitive field name
1189   followed by a colon (":"), optional whitespace, and the field value.
1191<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="header-field"/><iref primary="true" item="Grammar" subitem="field-name"/><iref primary="true" item="Grammar" subitem="field-value"/><iref primary="true" item="Grammar" subitem="field-content"/><iref primary="true" item="Grammar" subitem="obs-fold"/>
1192  <x:ref>header-field</x:ref>   = <x:ref>field-name</x:ref> ":" <x:ref>OWS</x:ref> <x:ref>field-value</x:ref> <x:ref>BWS</x:ref>
1193  <x:ref>field-name</x:ref>     = <x:ref>token</x:ref>
1194  <x:ref>field-value</x:ref>    = *( <x:ref>field-content</x:ref> / <x:ref>obs-fold</x:ref> )
1195  <x:ref>field-content</x:ref>  = *( <x:ref>HTAB</x:ref> / <x:ref>SP</x:ref> / <x:ref>VCHAR</x:ref> / <x:ref>obs-text</x:ref> )
1196  <x:ref>obs-fold</x:ref>       = <x:ref>CRLF</x:ref> ( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1197                 ; obsolete line folding
1198                 ; see <xref target="field.parsing"/>
1201   The field-name token labels the corresponding field-value as having the
1202   semantics defined by that header field.  For example, the <x:ref>Date</x:ref>
1203   header field is defined in &header-date; as containing the origination
1204   timestamp for the message in which it appears.
1207   HTTP header fields are fully extensible: there is no limit on the
1208   introduction of new field names, each presumably defining new semantics,
1209   or on the number of header fields used in a given message.  Existing
1210   fields are defined in each part of this specification and in many other
1211   specifications outside the standards process.
1212   New header fields can be introduced without changing the protocol version
1213   if their defined semantics allow them to be safely ignored by recipients
1214   that do not recognize them.
1217   New HTTP header fields &SHOULD; be registered with IANA according
1218   to the procedures in &cons-new-header-fields;.
1219   Unrecognized header fields &MUST; be forwarded by a proxy unless the
1220   field-name is listed in the <x:ref>Connection</x:ref> header field
1221   (<xref target="header.connection"/>) or the proxy is specifically
1222   configured to block or otherwise transform such fields.
1223   Unrecognized header fields &SHOULD; be ignored by other recipients.
1226   The order in which header fields with differing field names are
1227   received is not significant. However, it is "good practice" to send
1228   header fields that contain control data first, such as <x:ref>Host</x:ref>
1229   on requests and <x:ref>Date</x:ref> on responses, so that implementations
1230   can decide when not to handle a message as early as possible.  A server
1231   &MUST; wait until the entire header section is received before interpreting
1232   a request message, since later header fields might include conditionals,
1233   authentication credentials, or deliberately misleading duplicate
1234   header fields that would impact request processing.
1237   Multiple header fields with the same field name &MUST-NOT; be
1238   sent in a message unless the entire field value for that
1239   header field is defined as a comma-separated list [i.e., #(values)].
1240   Multiple header fields with the same field name can be combined into
1241   one "field-name: field-value" pair, without changing the semantics of the
1242   message, by appending each subsequent field value to the combined
1243   field value in order, separated by a comma. The order in which
1244   header fields with the same field name are received is therefore
1245   significant to the interpretation of the combined field value;
1246   a proxy &MUST-NOT; change the order of these field values when
1247   forwarding a message.
1250  <t>
1251   &Note; The "Set-Cookie" header field as implemented in
1252   practice can occur multiple times, but does not use the list syntax, and
1253   thus cannot be combined into a single line (<xref target="RFC6265"/>). (See Appendix A.2.3 of <xref target="Kri2001"/>
1254   for details.) Also note that the Set-Cookie2 header field specified in
1255   <xref target="RFC2965"/> does not share this problem.
1256  </t>
1259<section title="Whitespace" anchor="whitespace">
1260<t anchor="rule.LWS">
1261   This specification uses three rules to denote the use of linear
1262   whitespace: OWS (optional whitespace), RWS (required whitespace), and
1263   BWS ("bad" whitespace).
1265<t anchor="rule.OWS">
1266   The OWS rule is used where zero or more linear whitespace octets might
1267   appear. OWS &SHOULD; either not be produced or be produced as a single
1268   SP. Multiple OWS octets that occur within field-content &SHOULD; either
1269   be replaced with a single SP or transformed to all SP octets (each
1270   octet other than SP replaced with SP) before interpreting the field value
1271   or forwarding the message downstream.
1273<t anchor="rule.RWS">
1274   RWS is used when at least one linear whitespace octet is required to
1275   separate field tokens. RWS &SHOULD; be produced as a single SP.
1276   Multiple RWS octets that occur within field-content &SHOULD; either
1277   be replaced with a single SP or transformed to all SP octets before
1278   interpreting the field value or forwarding the message downstream.
1280<t anchor="rule.BWS">
1281   BWS is used where the grammar allows optional whitespace for historical
1282   reasons but senders &SHOULD-NOT; produce it in messages. HTTP/1.1
1283   recipients &MUST; accept such bad optional whitespace and remove it before
1284   interpreting the field value or forwarding the message downstream.
1286<t anchor="rule.whitespace">
1287  <x:anchor-alias value="BWS"/>
1288  <x:anchor-alias value="OWS"/>
1289  <x:anchor-alias value="RWS"/>
1291<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="OWS"/><iref primary="true" item="Grammar" subitem="RWS"/><iref primary="true" item="Grammar" subitem="BWS"/>
1292  <x:ref>OWS</x:ref>            = *( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1293                 ; "optional" whitespace
1294  <x:ref>RWS</x:ref>            = 1*( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1295                 ; "required" whitespace
1296  <x:ref>BWS</x:ref>            = <x:ref>OWS</x:ref>
1297                 ; "bad" whitespace
1301<section title="Field Parsing" anchor="field.parsing">
1303   No whitespace is allowed between the header field-name and colon.
1304   In the past, differences in the handling of such whitespace have led to
1305   security vulnerabilities in request routing and response handling.
1306   Any received request message that contains whitespace between a header
1307   field-name and colon &MUST; be rejected with a response code of 400
1308   (Bad Request).  A proxy &MUST; remove any such whitespace from a response
1309   message before forwarding the message downstream.
1312   A field value &MAY; be preceded by optional whitespace (OWS); a single SP is
1313   preferred. The field value does not include any leading or trailing white
1314   space: OWS occurring before the first non-whitespace octet of the
1315   field value or after the last non-whitespace octet of the field value
1316   is ignored and &SHOULD; be removed before further processing (as this does
1317   not change the meaning of the header field).
1320   Historically, HTTP header field values could be extended over multiple
1321   lines by preceding each extra line with at least one space or horizontal
1322   tab (obs-fold). This specification deprecates such line
1323   folding except within the message/http media type
1324   (<xref target=""/>).
1325   HTTP senders &MUST-NOT; produce messages that include line folding
1326   (i.e., that contain any field-value that matches the obs-fold rule) unless
1327   the message is intended for packaging within the message/http media type.
1328   HTTP recipients &SHOULD; accept line folding and replace any embedded
1329   obs-fold whitespace with either a single SP or a matching number of SP
1330   octets (to avoid buffer copying) prior to interpreting the field value or
1331   forwarding the message downstream.
1334   Historically, HTTP has allowed field content with text in the ISO-8859-1
1335   <xref target="ISO-8859-1"/> character encoding and supported other
1336   character sets only through use of <xref target="RFC2047"/> encoding.
1337   In practice, most HTTP header field values use only a subset of the
1338   US-ASCII character encoding <xref target="USASCII"/>. Newly defined
1339   header fields &SHOULD; limit their field values to US-ASCII octets.
1340   Recipients &SHOULD; treat other (obs-text) octets in field content as
1341   opaque data.
1345<section title="Field Length" anchor="field.length">
1347   HTTP does not place a pre-defined limit on the length of header fields,
1348   either in isolation or as a set. A server &MUST; be prepared to receive
1349   request header fields of unbounded length and respond with a <x:ref>4xx
1350   (Client Error)</x:ref> status code if the received header field(s) would be
1351   longer than the server wishes to handle.
1354   A client that receives response header fields that are longer than it wishes
1355   to handle can only treat it as a server error.
1358   Various ad-hoc limitations on header field length are found in practice. It
1359   is &RECOMMENDED; that all HTTP senders and recipients support messages whose
1360   combined header fields have 4000 or more octets.
1364<section title="Field value components" anchor="field.components">
1365<t anchor="rule.token.separators">
1366  <x:anchor-alias value="tchar"/>
1367  <x:anchor-alias value="token"/>
1368  <x:anchor-alias value="special"/>
1369  <x:anchor-alias value="word"/>
1370   Many HTTP/1.1 header field values consist of words (token or quoted-string)
1371   separated by whitespace or special characters. These special characters
1372   &MUST; be in a quoted string to be used within a parameter value (as defined
1373   in <xref target="transfer.codings"/>).
1375<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="word"/><iref primary="true" item="Grammar" subitem="token"/><iref primary="true" item="Grammar" subitem="tchar"/><iref primary="true" item="Grammar" subitem="special"/>
1376  <x:ref>word</x:ref>           = <x:ref>token</x:ref> / <x:ref>quoted-string</x:ref>
1378  <x:ref>token</x:ref>          = 1*<x:ref>tchar</x:ref>
1380  IMPORTANT: when editing "tchar" make sure that "special" is updated accordingly!!!
1381 -->
1382  <x:ref>tchar</x:ref>          = "!" / "#" / "$" / "%" / "&amp;" / "'" / "*"
1383                 / "+" / "-" / "." / "^" / "_" / "`" / "|" / "~"
1384                 / <x:ref>DIGIT</x:ref> / <x:ref>ALPHA</x:ref>
1385                 ; any <x:ref>VCHAR</x:ref>, except <x:ref>special</x:ref>
1387  <x:ref>special</x:ref>        = "(" / ")" / "&lt;" / ">" / "@" / ","
1388                 / ";" / ":" / "\" / DQUOTE / "/" / "["
1389                 / "]" / "?" / "=" / "{" / "}"
1391<t anchor="rule.quoted-string">
1392  <x:anchor-alias value="quoted-string"/>
1393  <x:anchor-alias value="qdtext"/>
1394  <x:anchor-alias value="obs-text"/>
1395   A string of text is parsed as a single word if it is quoted using
1396   double-quote marks.
1398<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-string"/><iref primary="true" item="Grammar" subitem="qdtext"/><iref primary="true" item="Grammar" subitem="obs-text"/>
1399  <x:ref>quoted-string</x:ref>  = <x:ref>DQUOTE</x:ref> *( <x:ref>qdtext</x:ref> / <x:ref>quoted-pair</x:ref> ) <x:ref>DQUOTE</x:ref>
1400  <x:ref>qdtext</x:ref>         = <x:ref>OWS</x:ref> / %x21 / %x23-5B / %x5D-7E / <x:ref>obs-text</x:ref>
1401  <x:ref>obs-text</x:ref>       = %x80-FF
1403<t anchor="rule.quoted-pair">
1404  <x:anchor-alias value="quoted-pair"/>
1405   The backslash octet ("\") can be used as a single-octet
1406   quoting mechanism within quoted-string constructs:
1408<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-pair"/>
1409  <x:ref>quoted-pair</x:ref>    = "\" ( <x:ref>HTAB</x:ref> / <x:ref>SP</x:ref> / <x:ref>VCHAR</x:ref> / <x:ref>obs-text</x:ref> )
1412   Recipients that process the value of the quoted-string &MUST; handle a
1413   quoted-pair as if it were replaced by the octet following the backslash.
1416   Senders &SHOULD-NOT; escape octets in quoted-strings that do not require
1417   escaping (i.e., other than DQUOTE and the backslash octet).
1419<t anchor="rule.comment">
1420  <x:anchor-alias value="comment"/>
1421  <x:anchor-alias value="ctext"/>
1422   Comments can be included in some HTTP header fields by surrounding
1423   the comment text with parentheses. Comments are only allowed in
1424   fields containing "comment" as part of their field value definition.
1426<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="comment"/><iref primary="true" item="Grammar" subitem="ctext"/>
1427  <x:ref>comment</x:ref>        = "(" *( <x:ref>ctext</x:ref> / <x:ref>quoted-cpair</x:ref> / <x:ref>comment</x:ref> ) ")"
1428  <x:ref>ctext</x:ref>          = <x:ref>OWS</x:ref> / %x21-27 / %x2A-5B / %x5D-7E / <x:ref>obs-text</x:ref>
1430<t anchor="rule.quoted-cpair">
1431  <x:anchor-alias value="quoted-cpair"/>
1432   The backslash octet ("\") can be used as a single-octet
1433   quoting mechanism within comment constructs:
1435<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-cpair"/>
1436  <x:ref>quoted-cpair</x:ref>   = "\" ( <x:ref>HTAB</x:ref> / <x:ref>SP</x:ref> / <x:ref>VCHAR</x:ref> / <x:ref>obs-text</x:ref> )
1439   Senders &SHOULD-NOT; escape octets in comments that do not require escaping
1440   (i.e., other than the backslash octet "\" and the parentheses "(" and ")").
1444<section title="ABNF list extension: #rule" anchor="abnf.extension">
1446  A #rule extension to the ABNF rules of <xref target="RFC5234"/> is used to
1447  improve readability in the definitions of some header field values.
1450  A construct "#" is defined, similar to "*", for defining comma-delimited
1451  lists of elements. The full form is "&lt;n&gt;#&lt;m&gt;element" indicating
1452  at least &lt;n&gt; and at most &lt;m&gt; elements, each separated by a single
1453  comma (",") and optional whitespace (OWS).   
1456  Thus,
1457</preamble><artwork type="example">
1458  1#element =&gt; element *( OWS "," OWS element )
1461  and:
1462</preamble><artwork type="example">
1463  #element =&gt; [ 1#element ]
1466  and for n &gt;= 1 and m &gt; 1:
1467</preamble><artwork type="example">
1468  &lt;n&gt;#&lt;m&gt;element =&gt; element &lt;n-1&gt;*&lt;m-1&gt;( OWS "," OWS element )
1471  For compatibility with legacy list rules, recipients &SHOULD; accept empty
1472  list elements. In other words, consumers would follow the list productions:
1474<figure><artwork type="example">
1475  #element =&gt; [ ( "," / element ) *( OWS "," [ OWS element ] ) ]
1477  1#element =&gt; *( "," OWS ) element *( OWS "," [ OWS element ] )
1480  Note that empty elements do not contribute to the count of elements present,
1481  though.
1484  For example, given these ABNF productions:
1486<figure><artwork type="example">
1487  example-list      = 1#example-list-elmt
1488  example-list-elmt = token ; see <xref target="field.components"/>
1491  Then these are valid values for example-list (not including the double
1492  quotes, which are present for delimitation only):
1494<figure><artwork type="example">
1495  "foo,bar"
1496  "foo ,bar,"
1497  "foo , ,bar,charlie   "
1500  But these values would be invalid, as at least one non-empty element is
1501  required:
1503<figure><artwork type="example">
1504  ""
1505  ","
1506  ",   ,"
1509  <xref target="collected.abnf"/> shows the collected ABNF, with the list rules
1510  expanded as explained above.
1515<section title="Message Body" anchor="message.body">
1516  <x:anchor-alias value="message-body"/>
1518   The message body (if any) of an HTTP message is used to carry the
1519   payload body of that request or response.  The message body is
1520   identical to the payload body unless a transfer coding has been
1521   applied, as described in <xref target="header.transfer-encoding"/>.
1523<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="message-body"/>
1524  <x:ref>message-body</x:ref> = *OCTET
1527   The rules for when a message body is allowed in a message differ for
1528   requests and responses.
1531   The presence of a message body in a request is signaled by a
1532   a <x:ref>Content-Length</x:ref> or <x:ref>Transfer-Encoding</x:ref> header
1533   field. Request message framing is independent of method semantics,
1534   even if the method does not define any use for a message body.
1537   The presence of a message body in a response depends on both
1538   the request method to which it is responding and the response
1539   status code (<xref target="status.line"/>).
1540   Responses to the HEAD request method never include a message body
1541   because the associated response header fields (e.g.,
1542   <x:ref>Transfer-Encoding</x:ref>, <x:ref>Content-Length</x:ref>, etc.) only
1543   indicate what their values would have been if the request method had been
1544   GET. <x:ref>2xx (Successful)</x:ref> responses to CONNECT switch to tunnel
1545   mode instead of having a message body.
1546   All <x:ref>1xx (Informational)</x:ref>, <x:ref>204 (No Content)</x:ref>, and
1547   <x:ref>304 (Not Modified)</x:ref> responses &MUST-NOT; include a message body.
1548   All other responses do include a message body, although the body
1549   &MAY; be of zero length.
1552<section title="Transfer-Encoding" anchor="header.transfer-encoding">
1553  <iref primary="true" item="Transfer-Encoding header field" x:for-anchor=""/>
1554  <iref primary="true" item="Header Fields" subitem="Transfer-Encoding" x:for-anchor=""/>
1555  <x:anchor-alias value="Transfer-Encoding"/>
1557   When one or more transfer codings are applied to a payload body in order
1558   to form the message body, a Transfer-Encoding header field &MUST; be sent
1559   in the message and &MUST; contain the list of corresponding
1560   transfer-coding names in the same order that they were applied.
1561   Transfer codings are defined in <xref target="transfer.codings"/>.
1563<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Transfer-Encoding"/>
1564  <x:ref>Transfer-Encoding</x:ref> = 1#<x:ref>transfer-coding</x:ref>
1567   Transfer-Encoding is analogous to the Content-Transfer-Encoding field of
1568   MIME, which was designed to enable safe transport of binary data over a
1569   7-bit transport service (<xref target="RFC2045" x:fmt="," x:sec="6"/>).
1570   However, safe transport has a different focus for an 8bit-clean transfer
1571   protocol. In HTTP's case, Transfer-Encoding is primarily intended to
1572   accurately delimit a dynamically generated payload and to distinguish
1573   payload encodings that are only applied for transport efficiency or
1574   security from those that are characteristics of the target resource.
1577   The "chunked" transfer-coding (<xref target="chunked.encoding"/>)
1578   &MUST; be implemented by all HTTP/1.1 recipients because it plays a
1579   crucial role in delimiting messages when the payload body size is not
1580   known in advance.
1581   When the "chunked" transfer-coding is used, it &MUST; be the last
1582   transfer-coding applied to form the message body and &MUST-NOT;
1583   be applied more than once in a message body.
1584   If any transfer-coding is applied to a request payload body,
1585   the final transfer-coding applied &MUST; be "chunked".
1586   If any transfer-coding is applied to a response payload body, then either
1587   the final transfer-coding applied &MUST; be "chunked" or
1588   the message &MUST; be terminated by closing the connection.
1591   For example,
1592</preamble><artwork type="example">
1593  Transfer-Encoding: gzip, chunked
1595   indicates that the payload body has been compressed using the gzip
1596   coding and then chunked using the chunked coding while forming the
1597   message body.
1600   If more than one Transfer-Encoding header field is present in a message,
1601   the multiple field-values &MUST; be combined into one field-value,
1602   according to the algorithm defined in <xref target="header.fields"/>,
1603   before determining the message body length.
1606   Unlike <x:ref>Content-Encoding</x:ref> (&content-codings;),
1607   Transfer-Encoding is a property of the message, not of the payload, and thus
1608   &MAY; be added or removed by any implementation along the request/response
1609   chain. Additional information about the encoding parameters &MAY; be
1610   provided by other header fields not defined by this specification.
1613   Transfer-Encoding &MAY; be sent in a response to a HEAD request or in a
1614   <x:ref>304 (Not Modified)</x:ref> response (&status-304;) to a GET request,
1615   neither of which includes a message body,
1616   to indicate that the origin server would have applied a transfer coding
1617   to the message body if the request had been an unconditional GET.
1618   This indication is not required, however, because any recipient on
1619   the response chain (including the origin server) can remove transfer
1620   codings when they are not needed.
1623   Transfer-Encoding was added in HTTP/1.1.  It is generally assumed that
1624   implementations advertising only HTTP/1.0 support will not understand
1625   how to process a transfer-encoded payload.
1626   A client &MUST-NOT; send a request containing Transfer-Encoding unless it
1627   knows the server will handle HTTP/1.1 (or later) requests; such knowledge
1628   might be in the form of specific user configuration or by remembering the
1629   version of a prior received response.
1630   A server &MUST-NOT; send a response containing Transfer-Encoding unless
1631   the corresponding request indicates HTTP/1.1 (or later).
1634   A server that receives a request message with a transfer-coding it does
1635   not understand &SHOULD; respond with <x:ref>501 (Not Implemented)</x:ref> and then
1636   close the connection.
1640<section title="Content-Length" anchor="header.content-length">
1641  <iref primary="true" item="Content-Length header field" x:for-anchor=""/>
1642  <iref primary="true" item="Header Fields" subitem="Content-Length" x:for-anchor=""/>
1643  <x:anchor-alias value="Content-Length"/>
1645   When a message does not have a <x:ref>Transfer-Encoding</x:ref> header field
1646   and the payload body length can be determined prior to being transferred, a
1647   Content-Length header field &SHOULD; be sent to indicate the length of the
1648   payload body that is either present as the message body, for requests
1649   and non-HEAD responses other than <x:ref>304 (Not Modified)</x:ref>, or
1650   would have been present had the request been an unconditional GET.  The
1651   length is expressed as a decimal number of octets.
1653<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Content-Length"/>
1654  <x:ref>Content-Length</x:ref> = 1*<x:ref>DIGIT</x:ref>
1657   An example is
1659<figure><artwork type="example">
1660  Content-Length: 3495
1663   In the case of a response to a HEAD request, Content-Length indicates
1664   the size of the payload body (without any potential transfer-coding)
1665   that would have been sent had the request been a GET.
1666   In the case of a <x:ref>304 (Not Modified)</x:ref> response (&status-304;)
1667   to a GET request, Content-Length indicates the size of the payload body (without
1668   any potential transfer-coding) that would have been sent in a <x:ref>200 (OK)</x:ref>
1669   response.
1672   Any Content-Length field value greater than or equal to zero is valid.
1673   Since there is no predefined limit to the length of an HTTP payload,
1674   recipients &SHOULD; anticipate potentially large decimal numerals and
1675   prevent parsing errors due to integer conversion overflows
1676   (<xref target="attack.protocol.element.size.overflows"/>).
1679   If a message is received that has multiple Content-Length header fields
1680   with field-values consisting of the same decimal value, or a single
1681   Content-Length header field with a field value containing a list of
1682   identical decimal values (e.g., "Content-Length: 42, 42"), indicating that
1683   duplicate Content-Length header fields have been generated or combined by an
1684   upstream message processor, then the recipient &MUST; either reject the
1685   message as invalid or replace the duplicated field-values with a single
1686   valid Content-Length field containing that decimal value prior to
1687   determining the message body length.
1690  <t>
1691   &Note; HTTP's use of Content-Length for message framing differs
1692   significantly from the same field's use in MIME, where it is an optional
1693   field used only within the "message/external-body" media-type.
1694  </t>
1698<section title="Message Body Length" anchor="message.body.length">
1700   The length of a message body is determined by one of the following
1701   (in order of precedence):
1704  <list style="numbers">
1705    <x:lt><t>
1706     Any response to a HEAD request and any response with a
1707     <x:ref>1xx (Informational)</x:ref>, <x:ref>204 (No Content)</x:ref>, or
1708     <x:ref>304 (Not Modified)</x:ref> status code is always
1709     terminated by the first empty line after the header fields, regardless of
1710     the header fields present in the message, and thus cannot contain a
1711     message body.
1712    </t></x:lt>
1713    <x:lt><t>
1714     Any <x:ref>2xx (Successful)</x:ref> response to a CONNECT request implies that the
1715     connection will become a tunnel immediately after the empty line that
1716     concludes the header fields.  A client &MUST; ignore any
1717     <x:ref>Content-Length</x:ref> or <x:ref>Transfer-Encoding</x:ref> header
1718     fields received in such a message.
1719    </t></x:lt>
1720    <x:lt><t>
1721     If a <x:ref>Transfer-Encoding</x:ref> header field is present
1722     and the "chunked" transfer-coding (<xref target="chunked.encoding"/>)
1723     is the final encoding, the message body length is determined by reading
1724     and decoding the chunked data until the transfer-coding indicates the
1725     data is complete.
1726    </t>
1727    <t>
1728     If a <x:ref>Transfer-Encoding</x:ref> header field is present in a
1729     response and the "chunked" transfer-coding is not the final encoding, the
1730     message body length is determined by reading the connection until it is
1731     closed by the server.
1732     If a Transfer-Encoding header field is present in a request and the
1733     "chunked" transfer-coding is not the final encoding, the message body
1734     length cannot be determined reliably; the server &MUST; respond with
1735     the <x:ref>400 (Bad Request)</x:ref> status code and then close the connection.
1736    </t>
1737    <t>
1738     If a message is received with both a <x:ref>Transfer-Encoding</x:ref>
1739     and a <x:ref>Content-Length</x:ref> header field, the
1740     Transfer-Encoding overrides the Content-Length.
1741     Such a message might indicate an attempt to perform request or response
1742     smuggling (bypass of security-related checks on message routing or content)
1743     and thus ought to be handled as an error.  The provided Content-Length &MUST;
1744     be removed, prior to forwarding the message downstream, or replaced with
1745     the real message body length after the transfer-coding is decoded.
1746    </t></x:lt>
1747    <x:lt><t>
1748     If a message is received without <x:ref>Transfer-Encoding</x:ref> and with
1749     either multiple <x:ref>Content-Length</x:ref> header fields having
1750     differing field-values or a single Content-Length header field having an
1751     invalid value, then the message framing is invalid and &MUST; be treated
1752     as an error to prevent request or response smuggling.
1753     If this is a request message, the server &MUST; respond with
1754     a <x:ref>400 (Bad Request)</x:ref> status code and then close the connection.
1755     If this is a response message received by a proxy, the proxy
1756     &MUST; discard the received response, send a <x:ref>502 (Bad Gateway)</x:ref>
1757     status code as its downstream response, and then close the connection.
1758     If this is a response message received by a user-agent, it &MUST; be
1759     treated as an error by discarding the message and closing the connection.
1760    </t></x:lt>
1761    <x:lt><t>
1762     If a valid <x:ref>Content-Length</x:ref> header field is present without
1763     <x:ref>Transfer-Encoding</x:ref>, its decimal value defines the
1764     message body length in octets.  If the actual number of octets sent in
1765     the message is less than the indicated Content-Length, the recipient
1766     &MUST; consider the message to be incomplete and treat the connection
1767     as no longer usable.
1768     If the actual number of octets sent in the message is more than the indicated
1769     Content-Length, the recipient &MUST; only process the message body up to the
1770     field value's number of octets; the remainder of the message &MUST; either
1771     be discarded or treated as the next message in a pipeline.  For the sake of
1772     robustness, a user-agent &MAY; attempt to detect and correct such an error
1773     in message framing if it is parsing the response to the last request on
1774     a connection and the connection has been closed by the server.
1775    </t></x:lt>
1776    <x:lt><t>
1777     If this is a request message and none of the above are true, then the
1778     message body length is zero (no message body is present).
1779    </t></x:lt>
1780    <x:lt><t>
1781     Otherwise, this is a response message without a declared message body
1782     length, so the message body length is determined by the number of octets
1783     received prior to the server closing the connection.
1784    </t></x:lt>
1785  </list>
1788   Since there is no way to distinguish a successfully completed,
1789   close-delimited message from a partially-received message interrupted
1790   by network failure, implementations &SHOULD; use encoding or
1791   length-delimited messages whenever possible.  The close-delimiting
1792   feature exists primarily for backwards compatibility with HTTP/1.0.
1795   A server &MAY; reject a request that contains a message body but
1796   not a <x:ref>Content-Length</x:ref> by responding with
1797   <x:ref>411 (Length Required)</x:ref>.
1800   Unless a transfer-coding other than "chunked" has been applied,
1801   a client that sends a request containing a message body &SHOULD;
1802   use a valid <x:ref>Content-Length</x:ref> header field if the message body
1803   length is known in advance, rather than the "chunked" encoding, since some
1804   existing services respond to "chunked" with a <x:ref>411 (Length Required)</x:ref>
1805   status code even though they understand the chunked encoding.  This
1806   is typically because such services are implemented via a gateway that
1807   requires a content-length in advance of being called and the server
1808   is unable or unwilling to buffer the entire request before processing.
1811   A client that sends a request containing a message body &MUST; include a
1812   valid <x:ref>Content-Length</x:ref> header field if it does not know the
1813   server will handle HTTP/1.1 (or later) requests; such knowledge can be in
1814   the form of specific user configuration or by remembering the version of a
1815   prior received response.
1820<section anchor="incomplete.messages" title="Handling Incomplete Messages">
1822   Request messages that are prematurely terminated, possibly due to a
1823   cancelled connection or a server-imposed time-out exception, &MUST;
1824   result in closure of the connection; sending an HTTP/1.1 error response
1825   prior to closing the connection is &OPTIONAL;.
1828   Response messages that are prematurely terminated, usually by closure
1829   of the connection prior to receiving the expected number of octets or by
1830   failure to decode a transfer-encoded message body, &MUST; be recorded
1831   as incomplete.  A response that terminates in the middle of the header
1832   block (before the empty line is received) cannot be assumed to convey the
1833   full semantics of the response and &MUST; be treated as an error.
1836   A message body that uses the chunked transfer encoding is
1837   incomplete if the zero-sized chunk that terminates the encoding has not
1838   been received.  A message that uses a valid <x:ref>Content-Length</x:ref> is
1839   incomplete if the size of the message body received (in octets) is less than
1840   the value given by Content-Length.  A response that has neither chunked
1841   transfer encoding nor Content-Length is terminated by closure of the
1842   connection, and thus is considered complete regardless of the number of
1843   message body octets received, provided that the header block was received
1844   intact.
1847   A user agent &MUST-NOT; render an incomplete response message body as if
1848   it were complete (i.e., some indication needs to be given to the user that an
1849   error occurred).  Cache requirements for incomplete responses are defined
1850   in &cache-incomplete;.
1853   A server &MUST; read the entire request message body or close
1854   the connection after sending its response, since otherwise the
1855   remaining data on a persistent connection would be misinterpreted
1856   as the next request.  Likewise,
1857   a client &MUST; read the entire response message body if it intends
1858   to reuse the same connection for a subsequent request.  Pipelining
1859   multiple requests on a connection is described in <xref target="pipelining"/>.
1863<section title="Message Parsing Robustness" anchor="message.robustness">
1865   Older HTTP/1.0 client implementations might send an extra CRLF
1866   after a POST request as a lame workaround for some early server
1867   applications that failed to read message body content that was
1868   not terminated by a line-ending. An HTTP/1.1 client &MUST-NOT;
1869   preface or follow a request with an extra CRLF.  If terminating
1870   the request message body with a line-ending is desired, then the
1871   client &MUST; include the terminating CRLF octets as part of the
1872   message body length.
1875   In the interest of robustness, servers &SHOULD; ignore at least one
1876   empty line received where a request-line is expected. In other words, if
1877   the server is reading the protocol stream at the beginning of a
1878   message and receives a CRLF first, it &SHOULD; ignore the CRLF.
1879   Likewise, although the line terminator for the start-line and header
1880   fields is the sequence CRLF, we recommend that recipients recognize a
1881   single LF as a line terminator and ignore any CR.
1884   When a server listening only for HTTP request messages, or processing
1885   what appears from the start-line to be an HTTP request message,
1886   receives a sequence of octets that does not match the HTTP-message
1887   grammar aside from the robustness exceptions listed above, the
1888   server &MUST; respond with an HTTP/1.1 <x:ref>400 (Bad Request)</x:ref> response. 
1893<section title="Transfer Codings" anchor="transfer.codings">
1894  <x:anchor-alias value="transfer-coding"/>
1895  <x:anchor-alias value="transfer-extension"/>
1897   Transfer-coding values are used to indicate an encoding
1898   transformation that has been, can be, or might need to be applied to a
1899   payload body in order to ensure "safe transport" through the network.
1900   This differs from a content coding in that the transfer-coding is a
1901   property of the message rather than a property of the representation
1902   that is being transferred.
1904<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="transfer-coding"/><iref primary="true" item="Grammar" subitem="transfer-extension"/>
1905  <x:ref>transfer-coding</x:ref>    = "chunked" ; <xref target="chunked.encoding"/>
1906                     / "compress" ; <xref target="compress.coding"/>
1907                     / "deflate" ; <xref target="deflate.coding"/>
1908                     / "gzip" ; <xref target="gzip.coding"/>
1909                     / <x:ref>transfer-extension</x:ref>
1910  <x:ref>transfer-extension</x:ref> = <x:ref>token</x:ref> *( <x:ref>OWS</x:ref> ";" <x:ref>OWS</x:ref> <x:ref>transfer-parameter</x:ref> )
1912<t anchor="rule.parameter">
1913  <x:anchor-alias value="attribute"/>
1914  <x:anchor-alias value="transfer-parameter"/>
1915  <x:anchor-alias value="value"/>
1916   Parameters are in the form of attribute/value pairs.
1918<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="transfer-parameter"/><iref primary="true" item="Grammar" subitem="attribute"/><iref primary="true" item="Grammar" subitem="value"/><iref primary="true" item="Grammar" subitem="date2"/><iref primary="true" item="Grammar" subitem="date3"/>
1919  <x:ref>transfer-parameter</x:ref> = <x:ref>attribute</x:ref> <x:ref>BWS</x:ref> "=" <x:ref>BWS</x:ref> <x:ref>value</x:ref>
1920  <x:ref>attribute</x:ref>          = <x:ref>token</x:ref>
1921  <x:ref>value</x:ref>              = <x:ref>word</x:ref>
1924   All transfer-coding values are case-insensitive.
1925   The HTTP Transfer Coding registry is defined in
1926   <xref target="transfer.coding.registry"/>.
1927   HTTP/1.1 uses transfer-coding values in the <x:ref>TE</x:ref> header field
1928   (<xref target="header.te"/>) and in the <x:ref>Transfer-Encoding</x:ref>
1929   header field (<xref target="header.transfer-encoding"/>).
1932<section title="Chunked Transfer Coding" anchor="chunked.encoding">
1933  <iref item="chunked (Coding Format)"/>
1934  <iref item="Coding Format" subitem="chunked"/>
1935  <x:anchor-alias value="chunk"/>
1936  <x:anchor-alias value="chunked-body"/>
1937  <x:anchor-alias value="chunk-data"/>
1938  <x:anchor-alias value="chunk-ext"/>
1939  <x:anchor-alias value="chunk-ext-name"/>
1940  <x:anchor-alias value="chunk-ext-val"/>
1941  <x:anchor-alias value="chunk-size"/>
1942  <x:anchor-alias value="last-chunk"/>
1943  <x:anchor-alias value="trailer-part"/>
1944  <x:anchor-alias value="quoted-str-nf"/>
1945  <x:anchor-alias value="qdtext-nf"/>
1947   The chunked encoding modifies the body of a message in order to
1948   transfer it as a series of chunks, each with its own size indicator,
1949   followed by an &OPTIONAL; trailer containing header fields. This
1950   allows dynamically produced content to be transferred along with the
1951   information necessary for the recipient to verify that it has
1952   received the full message.
1954<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="chunked-body"/><iref primary="true" item="Grammar" subitem="chunk"/><iref primary="true" item="Grammar" subitem="chunk-size"/><iref primary="true" item="Grammar" subitem="last-chunk"/><iref primary="true" item="Grammar" subitem="chunk-ext"/><iref primary="true" item="Grammar" subitem="chunk-ext-name"/><iref primary="true" item="Grammar" subitem="chunk-ext-val"/><iref primary="true" item="Grammar" subitem="chunk-data"/><iref primary="true" item="Grammar" subitem="trailer-part"/><iref primary="true" item="Grammar" subitem="quoted-str-nf"/><iref primary="true" item="Grammar" subitem="qdtext-nf"/>
1955  <x:ref>chunked-body</x:ref>   = *<x:ref>chunk</x:ref>
1956                   <x:ref>last-chunk</x:ref>
1957                   <x:ref>trailer-part</x:ref>
1958                   <x:ref>CRLF</x:ref>
1960  <x:ref>chunk</x:ref>          = <x:ref>chunk-size</x:ref> [ <x:ref>chunk-ext</x:ref> ] <x:ref>CRLF</x:ref>
1961                   <x:ref>chunk-data</x:ref> <x:ref>CRLF</x:ref>
1962  <x:ref>chunk-size</x:ref>     = 1*<x:ref>HEXDIG</x:ref>
1963  <x:ref>last-chunk</x:ref>     = 1*("0") [ <x:ref>chunk-ext</x:ref> ] <x:ref>CRLF</x:ref>
1965  <x:ref>chunk-ext</x:ref>      = *( ";" <x:ref>chunk-ext-name</x:ref> [ "=" <x:ref>chunk-ext-val</x:ref> ] )
1966  <x:ref>chunk-ext-name</x:ref> = <x:ref>token</x:ref>
1967  <x:ref>chunk-ext-val</x:ref>  = <x:ref>token</x:ref> / <x:ref>quoted-str-nf</x:ref>
1968  <x:ref>chunk-data</x:ref>     = 1*<x:ref>OCTET</x:ref> ; a sequence of chunk-size octets
1969  <x:ref>trailer-part</x:ref>   = *( <x:ref>header-field</x:ref> <x:ref>CRLF</x:ref> )
1971  <x:ref>quoted-str-nf</x:ref>  = <x:ref>DQUOTE</x:ref> *( <x:ref>qdtext-nf</x:ref> / <x:ref>quoted-pair</x:ref> ) <x:ref>DQUOTE</x:ref>
1972                 ; like <x:ref>quoted-string</x:ref>, but disallowing line folding
1973  <x:ref>qdtext-nf</x:ref>      = <x:ref>HTAB</x:ref> / <x:ref>SP</x:ref> / %x21 / %x23-5B / %x5D-7E / <x:ref>obs-text</x:ref>
1976   The chunk-size field is a string of hex digits indicating the size of
1977   the chunk-data in octets. The chunked encoding is ended by any chunk whose size is
1978   zero, followed by the trailer, which is terminated by an empty line.
1981   The trailer allows the sender to include additional HTTP header
1982   fields at the end of the message. The <x:ref>Trailer</x:ref> header field
1983   can be used to indicate which header fields are included in a trailer (see
1984   <xref target="header.trailer"/>).
1987   A server using chunked transfer-coding in a response &MUST-NOT; use the
1988   trailer for any header fields unless at least one of the following is
1989   true:
1990  <list style="numbers">
1991    <t>the request included a <x:ref>TE</x:ref> header field that indicates
1992    "trailers" is acceptable in the transfer-coding of the response, as
1993    described in <xref target="header.te"/>; or,</t>
1995    <t>the trailer fields consist entirely of optional metadata, and the
1996    recipient could use the message (in a manner acceptable to the server where
1997    the field originated) without receiving it. In other words, the server that
1998    generated the header field (often but not always the origin server) is
1999    willing to accept the possibility that the trailer fields might be silently
2000    discarded along the path to the client.</t>
2001  </list>
2004   This requirement prevents an interoperability failure when the
2005   message is being received by an HTTP/1.1 (or later) proxy and
2006   forwarded to an HTTP/1.0 recipient. It avoids a situation where
2007   conformance with the protocol would have necessitated a possibly
2008   infinite buffer on the proxy.
2011   A process for decoding the "chunked" transfer-coding
2012   can be represented in pseudo-code as:
2014<figure><artwork type="code">
2015  length := 0
2016  read chunk-size, chunk-ext (if any) and CRLF
2017  while (chunk-size &gt; 0) {
2018     read chunk-data and CRLF
2019     append chunk-data to decoded-body
2020     length := length + chunk-size
2021     read chunk-size and CRLF
2022  }
2023  read header-field
2024  while (header-field not empty) {
2025     append header-field to existing header fields
2026     read header-field
2027  }
2028  Content-Length := length
2029  Remove "chunked" from Transfer-Encoding
2032   All HTTP/1.1 applications &MUST; be able to receive and decode the
2033   "chunked" transfer-coding and &MUST; ignore chunk-ext extensions
2034   they do not understand.
2037   Use of chunk-ext extensions by senders is deprecated; they &SHOULD-NOT; be
2038   sent and definition of new chunk-extensions is discouraged.
2042<section title="Compression Codings" anchor="compression.codings">
2044   The codings defined below can be used to compress the payload of a
2045   message.
2048   &Note; Use of program names for the identification of encoding formats
2049   is not desirable and is discouraged for future encodings. Their
2050   use here is representative of historical practice, not good
2051   design.
2054   &Note; For compatibility with previous implementations of HTTP,
2055   applications &SHOULD; consider "x-gzip" and "x-compress" to be
2056   equivalent to "gzip" and "compress" respectively.
2059<section title="Compress Coding" anchor="compress.coding">
2060<iref item="compress (Coding Format)"/>
2061<iref item="Coding Format" subitem="compress"/>
2063   The "compress" format is produced by the common UNIX file compression
2064   program "compress". This format is an adaptive Lempel-Ziv-Welch
2065   coding (LZW).
2069<section title="Deflate Coding" anchor="deflate.coding">
2070<iref item="deflate (Coding Format)"/>
2071<iref item="Coding Format" subitem="deflate"/>
2073   The "deflate" format is defined as the "deflate" compression mechanism
2074   (described in <xref target="RFC1951"/>) used inside the "zlib"
2075   data format (<xref target="RFC1950"/>).
2078  <t>
2079    &Note; Some incorrect implementations send the "deflate"
2080    compressed data without the zlib wrapper.
2081   </t>
2085<section title="Gzip Coding" anchor="gzip.coding">
2086<iref item="gzip (Coding Format)"/>
2087<iref item="Coding Format" subitem="gzip"/>
2089   The "gzip" format is produced by the file compression program
2090   "gzip" (GNU zip), as described in <xref target="RFC1952"/>. This format is a
2091   Lempel-Ziv coding (LZ77) with a 32 bit CRC.
2097<section title="TE" anchor="header.te">
2098  <iref primary="true" item="TE header field" x:for-anchor=""/>
2099  <iref primary="true" item="Header Fields" subitem="TE" x:for-anchor=""/>
2100  <x:anchor-alias value="TE"/>
2101  <x:anchor-alias value="t-codings"/>
2102  <x:anchor-alias value="te-params"/>
2103  <x:anchor-alias value="te-ext"/>
2105   The "TE" header field indicates what extension transfer-codings
2106   the client is willing to accept in the response, and whether or not it is
2107   willing to accept trailer fields in a chunked transfer-coding.
2110   Its value consists of the keyword "trailers" and/or a comma-separated
2111   list of extension transfer-coding names with optional accept
2112   parameters (as described in <xref target="transfer.codings"/>).
2114<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="TE"/><iref primary="true" item="Grammar" subitem="t-codings"/><iref primary="true" item="Grammar" subitem="te-params"/><iref primary="true" item="Grammar" subitem="te-ext"/>
2115  <x:ref>TE</x:ref>        = #<x:ref>t-codings</x:ref>
2116  <x:ref>t-codings</x:ref> = "trailers" / ( <x:ref>transfer-extension</x:ref> [ <x:ref>te-params</x:ref> ] )
2117  <x:ref>te-params</x:ref> = <x:ref>OWS</x:ref> ";" <x:ref>OWS</x:ref> "q=" <x:ref>qvalue</x:ref> *( <x:ref>te-ext</x:ref> )
2118  <x:ref>te-ext</x:ref>    = <x:ref>OWS</x:ref> ";" <x:ref>OWS</x:ref> <x:ref>token</x:ref> [ "=" <x:ref>word</x:ref> ]
2121   The presence of the keyword "trailers" indicates that the client is
2122   willing to accept trailer fields in a chunked transfer-coding, as
2123   defined in <xref target="chunked.encoding"/>. This keyword is reserved for use with
2124   transfer-coding values even though it does not itself represent a
2125   transfer-coding.
2128   Examples of its use are:
2130<figure><artwork type="example">
2131  TE: deflate
2132  TE:
2133  TE: trailers, deflate;q=0.5
2136   The TE header field only applies to the immediate connection.
2137   Therefore, the keyword &MUST; be supplied within a <x:ref>Connection</x:ref>
2138   header field (<xref target="header.connection"/>) whenever TE is present in
2139   an HTTP/1.1 message.
2142   A server tests whether a transfer-coding is acceptable, according to
2143   a TE field, using these rules:
2144  <list style="numbers">
2145    <x:lt>
2146      <t>The "chunked" transfer-coding is always acceptable. If the
2147         keyword "trailers" is listed, the client indicates that it is
2148         willing to accept trailer fields in the chunked response on
2149         behalf of itself and any downstream clients. The implication is
2150         that, if given, the client is stating that either all
2151         downstream clients are willing to accept trailer fields in the
2152         forwarded response, or that it will attempt to buffer the
2153         response on behalf of downstream recipients.
2154      </t><t>
2155         &Note; HTTP/1.1 does not define any means to limit the size of a
2156         chunked response such that a client can be assured of buffering
2157         the entire response.</t>
2158    </x:lt>
2159    <x:lt>
2160      <t>If the transfer-coding being tested is one of the transfer-codings
2161         listed in the TE field, then it is acceptable unless it
2162         is accompanied by a qvalue of 0. (As defined in <xref target="quality.values"/>, a
2163         qvalue of 0 means "not acceptable".)</t>
2164    </x:lt>
2165    <x:lt>
2166      <t>If multiple transfer-codings are acceptable, then the
2167         acceptable transfer-coding with the highest non-zero qvalue is
2168         preferred.  The "chunked" transfer-coding always has a qvalue
2169         of 1.</t>
2170    </x:lt>
2171  </list>
2174   If the TE field-value is empty or if no TE field is present, the only
2175   acceptable transfer-coding is "chunked". A message with no transfer-coding is
2176   always acceptable.
2179<section title="Quality Values" anchor="quality.values">
2180  <x:anchor-alias value="qvalue"/>
2182   Both transfer codings (<x:ref>TE</x:ref> request header field,
2183   <xref target="header.te"/>) and content negotiation (&content.negotiation;)
2184   use short "floating point" numbers to indicate the relative importance
2185   ("weight") of various negotiable parameters.  A weight is normalized to a
2186   real number in the range 0 through 1, where 0 is the minimum and 1 the
2187   maximum value. If a parameter has a quality value of 0, then content with
2188   this parameter is "not acceptable" for the client. HTTP/1.1
2189   applications &MUST-NOT; generate more than three digits after the
2190   decimal point. User configuration of these values &SHOULD; also be
2191   limited in this fashion.
2193<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="qvalue"/>
2194  <x:ref>qvalue</x:ref>         = ( "0" [ "." 0*3<x:ref>DIGIT</x:ref> ] )
2195                 / ( "1" [ "." 0*3("0") ] )
2198  <t>
2199     &Note; "Quality values" is a misnomer, since these values merely represent
2200     relative degradation in desired quality.
2201  </t>
2206<section title="Trailer" anchor="header.trailer">
2207  <iref primary="true" item="Trailer header field" x:for-anchor=""/>
2208  <iref primary="true" item="Header Fields" subitem="Trailer" x:for-anchor=""/>
2209  <x:anchor-alias value="Trailer"/>
2211   The "Trailer" header field indicates that the given set of
2212   header fields is present in the trailer of a message encoded with
2213   chunked transfer-coding.
2215<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Trailer"/>
2216  <x:ref>Trailer</x:ref> = 1#<x:ref>field-name</x:ref>
2219   An HTTP/1.1 message &SHOULD; include a Trailer header field in a
2220   message using chunked transfer-coding with a non-empty trailer. Doing
2221   so allows the recipient to know which header fields to expect in the
2222   trailer.
2225   If no Trailer header field is present, the trailer &SHOULD-NOT;  include
2226   any header fields. See <xref target="chunked.encoding"/> for restrictions on the use of
2227   trailer fields in a "chunked" transfer-coding.
2230   Message header fields listed in the Trailer header field &MUST-NOT;
2231   include the following header fields:
2232  <list style="symbols">
2233    <t><x:ref>Transfer-Encoding</x:ref></t>
2234    <t><x:ref>Content-Length</x:ref></t>
2235    <t><x:ref>Trailer</x:ref></t>
2236  </list>
2241<section title="Message Routing" anchor="message.routing">
2243   HTTP request message routing is determined by each client based on the
2244   target resource, the client's proxy configuration, and
2245   establishment or reuse of an inbound connection.  The corresponding
2246   response routing follows the same connection chain back to the client.
2249<section title="Identifying a Target Resource" anchor="target-resource">
2250  <iref primary="true" item="target resource"/>
2251  <iref primary="true" item="target URI"/>
2253   HTTP is used in a wide variety of applications, ranging from
2254   general-purpose computers to home appliances.  In some cases,
2255   communication options are hard-coded in a client's configuration.
2256   However, most HTTP clients rely on the same resource identification
2257   mechanism and configuration techniques as general-purpose Web browsers.
2260   HTTP communication is initiated by a user agent for some purpose.
2261   The purpose is a combination of request semantics, which are defined in
2262   <xref target="Part2"/>, and a target resource upon which to apply those
2263   semantics.  A URI reference (<xref target="uri"/>) is typically used as
2264   an identifier for the "target resource", which a user agent would resolve
2265   to its absolute form in order to obtain the "target URI".  The target URI
2266   excludes the reference's fragment identifier component, if any,
2267   since fragment identifiers are reserved for client-side processing
2268   (<xref target="RFC3986" x:fmt="," x:sec="3.5"/>).
2271   HTTP intermediaries obtain the request semantics and target URI
2272   from the request-line of an incoming request message.
2276<section title="Connecting Inbound" anchor="connecting.inbound">
2278   Once the target URI is determined, a client needs to decide whether
2279   a network request is necessary to accomplish the desired semantics and,
2280   if so, where that request is to be directed.
2283   If the client has a response cache and the request semantics can be
2284   satisfied by a cache (<xref target="Part6"/>), then the request is
2285   usually directed to the cache first.
2288   If the request is not satisfied by a cache, then a typical client will
2289   check its configuration to determine whether a proxy is to be used to
2290   satisfy the request.  Proxy configuration is implementation-dependent,
2291   but is often based on URI prefix matching, selective authority matching,
2292   or both, and the proxy itself is usually identified by an "http" or
2293   "https" URI.  If a proxy is applicable, the client connects inbound by
2294   establishing (or reusing) a connection to that proxy.
2297   If no proxy is applicable, a typical client will invoke a handler routine,
2298   usually specific to the target URI's scheme, to connect directly
2299   to an authority for the target resource.  How that is accomplished is
2300   dependent on the target URI scheme and defined by its associated
2301   specification, similar to how this specification defines origin server
2302   access for resolution of the "http" (<xref target="http.uri"/>) and
2303   "https" (<xref target="https.uri"/>) schemes.
2307<section title="Request Target" anchor="request-target">
2309   Once an inbound connection is obtained
2310   (<xref target=""/>),
2311   the client sends an HTTP request message (<xref target="http.message"/>)
2312   with a request-target derived from the target URI.
2313   There are four distinct formats for the request-target, depending on both
2314   the method being requested and whether the request is to a proxy.
2316<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="request-target"/><iref primary="true" item="Grammar" subitem="origin-form"/><iref primary="true" item="Grammar" subitem="absolute-form"/><iref primary="true" item="Grammar" subitem="authority-form"/><iref primary="true" item="Grammar" subitem="asterisk-form"/>
2317  <x:ref>request-target</x:ref> = <x:ref>origin-form</x:ref>
2318                 / <x:ref>absolute-form</x:ref>
2319                 / <x:ref>authority-form</x:ref>
2320                 / <x:ref>asterisk-form</x:ref>
2322  <x:ref>origin-form</x:ref>    = <x:ref>path-absolute</x:ref> [ "?" <x:ref>query</x:ref> ]
2323  <x:ref>absolute-form</x:ref>  = <x:ref>absolute-URI</x:ref>
2324  <x:ref>authority-form</x:ref> = <x:ref>authority</x:ref>
2325  <x:ref>asterisk-form</x:ref>  = "*"
2327<t anchor="origin-form"><iref item="origin-form (of request-target)"/>
2328   The most common form of request-target is the origin-form.
2329   When making a request directly to an origin server, other than a CONNECT
2330   or server-wide OPTIONS request (as detailed below),
2331   a client &MUST; send only the absolute path and query components of
2332   the target URI as the request-target.
2333   If the target URI's path component is empty, then the client &MUST; send
2334   "/" as the path within the origin-form of request-target.
2335   A <x:ref>Host</x:ref> header field is also sent, as defined in
2336   <xref target=""/>, containing the target URI's
2337   authority component (excluding any userinfo).
2340   For example, a client wishing to retrieve a representation of the resource
2341   identified as
2343<figure><artwork x:indent-with="  " type="example">
2347   directly from the origin server would open (or reuse) a TCP connection
2348   to port 80 of the host "" and send the lines:
2350<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2351GET /where?q=now HTTP/1.1
2355   followed by the remainder of the request message.
2357<t anchor="absolute-form"><iref item="absolute-form (of request-target)"/>
2358   When making a request to a proxy, other than a CONNECT or server-wide
2359   OPTIONS request (as detailed below), a client &MUST; send the target URI
2360   in absolute-form as the request-target.
2361   The proxy is requested to either service that request from a valid cache,
2362   if possible, or make the same request on the client's behalf to either
2363   the next inbound proxy server or directly to the origin server indicated
2364   by the request-target.  Requirements on such "forwarding" of messages are
2365   defined in <xref target="intermediary.forwarding"/>.
2368   An example absolute-form of request-line would be:
2370<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2371GET HTTP/1.1
2374   To allow for transition to the absolute-form for all requests in some
2375   future version of HTTP, HTTP/1.1 servers &MUST; accept the absolute-form
2376   in requests, even though HTTP/1.1 clients will only send them in requests
2377   to proxies.
2379<t anchor="authority-form"><iref item="authority-form (of request-target)"/>
2380   The authority-form of request-target is only used for CONNECT requests
2381   (&CONNECT;).  When making a CONNECT request to establish a tunnel through
2382   one or more proxies, a client &MUST; send only the target URI's
2383   authority component (excluding any userinfo) as the request-target.
2384   For example,
2386<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2389<t anchor="asterisk-form"><iref item="asterisk-form (of request-target)"/>
2390   The asterisk-form of request-target is only used for a server-wide
2391   OPTIONS request (&OPTIONS;).  When a client wishes to request OPTIONS
2392   for the server as a whole, as opposed to a specific named resource of
2393   that server, the client &MUST; send only "*" (%x2A) as the request-target.
2394   For example,
2396<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2397OPTIONS * HTTP/1.1
2400   If a proxy receives an OPTIONS request with an absolute-form of
2401   request-target in which the URI has an empty path and no query component,
2402   then the last proxy on the request chain &MUST; send a request-target
2403   of "*" when it forwards the request to the indicated origin server.
2406   For example, the request
2407</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2411  would be forwarded by the final proxy as
2412</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2413OPTIONS * HTTP/1.1
2417   after connecting to port 8001 of host "".
2422<section title="Host" anchor="">
2423  <iref primary="true" item="Host header field" x:for-anchor=""/>
2424  <iref primary="true" item="Header Fields" subitem="Host" x:for-anchor=""/>
2425  <x:anchor-alias value="Host"/>
2427   The "Host" header field in a request provides the host and port
2428   information from the target URI, enabling the origin
2429   server to distinguish among resources while servicing requests
2430   for multiple host names on a single IP address.  Since the Host
2431   field-value is critical information for handling a request, it
2432   &SHOULD; be sent as the first header field following the request-line.
2434<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Host"/>
2435  <x:ref>Host</x:ref> = <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ; <xref target="http.uri"/>
2438   A client &MUST; send a Host header field in all HTTP/1.1 request
2439   messages.  If the target URI includes an authority component, then
2440   the Host field-value &MUST; be identical to that authority component
2441   after excluding any userinfo (<xref target="http.uri"/>).
2442   If the authority component is missing or undefined for the target URI,
2443   then the Host header field &MUST; be sent with an empty field-value.
2446   For example, a GET request to the origin server for
2447   &lt;; would begin with:
2449<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2450GET /pub/WWW/ HTTP/1.1
2454   The Host header field &MUST; be sent in an HTTP/1.1 request even
2455   if the request-target is in the absolute-form, since this
2456   allows the Host information to be forwarded through ancient HTTP/1.0
2457   proxies that might not have implemented Host.
2460   When an HTTP/1.1 proxy receives a request with an absolute-form of
2461   request-target, the proxy &MUST; ignore the received
2462   Host header field (if any) and instead replace it with the host
2463   information of the request-target.  If the proxy forwards the request,
2464   it &MUST; generate a new Host field-value based on the received
2465   request-target rather than forward the received Host field-value.
2468   Since the Host header field acts as an application-level routing
2469   mechanism, it is a frequent target for malware seeking to poison
2470   a shared cache or redirect a request to an unintended server.
2471   An interception proxy is particularly vulnerable if it relies on
2472   the Host field-value for redirecting requests to internal
2473   servers, or for use as a cache key in a shared cache, without
2474   first verifying that the intercepted connection is targeting a
2475   valid IP address for that host.
2478   A server &MUST; respond with a <x:ref>400 (Bad Request)</x:ref> status code
2479   to any HTTP/1.1 request message that lacks a Host header field and
2480   to any request message that contains more than one Host header field
2481   or a Host header field with an invalid field-value.
2485<section title="Effective Request URI" anchor="effective.request.uri">
2486  <iref primary="true" item="effective request URI"/>
2488   A server that receives an HTTP request message &MUST; reconstruct
2489   the user agent's original target URI, based on the pieces of information
2490   learned from the request-target, <x:ref>Host</x:ref> header field, and
2491   connection context, in order to identify the intended target resource and
2492   properly service the request. The URI derived from this reconstruction
2493   process is referred to as the "effective request URI".
2496   For a user agent, the effective request URI is the target URI.
2499   If the request-target is in absolute-form, then the effective request URI
2500   is the same as the request-target.  Otherwise, the effective request URI
2501   is constructed as follows.
2504   If the request is received over an SSL/TLS-secured TCP connection,
2505   then the effective request URI's scheme is "https"; otherwise, the
2506   scheme is "http".
2509   If the request-target is in authority-form, then the effective
2510   request URI's authority component is the same as the request-target.
2511   Otherwise, if a <x:ref>Host</x:ref> header field is supplied with a
2512   non-empty field-value, then the authority component is the same as the
2513   Host field-value. Otherwise, the authority component is the concatenation of
2514   the default host name configured for the server, a colon (":"), and the
2515   connection's incoming TCP port number in decimal form.
2518   If the request-target is in authority-form or asterisk-form, then the
2519   effective request URI's combined path and query component is empty.
2520   Otherwise, the combined path and query component is the same as the
2521   request-target.
2524   The components of the effective request URI, once determined as above,
2525   can be combined into absolute-URI form by concatenating the scheme,
2526   "://", authority, and combined path and query component.
2530   Example 1: the following message received over an insecure TCP connection
2532<artwork type="example" x:indent-with="  ">
2533GET /pub/WWW/TheProject.html HTTP/1.1
2539  has an effective request URI of
2541<artwork type="example" x:indent-with="  ">
2547   Example 2: the following message received over an SSL/TLS-secured TCP
2548   connection
2550<artwork type="example" x:indent-with="  ">
2551OPTIONS * HTTP/1.1
2557  has an effective request URI of
2559<artwork type="example" x:indent-with="  ">
2564   An origin server that does not allow resources to differ by requested
2565   host &MAY; ignore the <x:ref>Host</x:ref> field-value and instead replace it
2566   with a configured server name when constructing the effective request URI.
2569   Recipients of an HTTP/1.0 request that lacks a <x:ref>Host</x:ref> header
2570   field &MAY; attempt to use heuristics (e.g., examination of the URI path for
2571   something unique to a particular host) in order to guess the
2572   effective request URI's authority component.
2576<section title="Intermediary Forwarding" anchor="intermediary.forwarding">
2578   As described in <xref target="intermediaries"/>, intermediaries can serve
2579   a variety of roles in the processing of HTTP requests and responses.
2580   Some intermediaries are used to improve performance or availability.
2581   Others are used for access control or to filter content.
2582   Since an HTTP stream has characteristics similar to a pipe-and-filter
2583   architecture, there are no inherent limits to the extent an intermediary
2584   can enhance (or interfere) with either direction of the stream.
2587   In order to avoid request loops, a proxy that forwards requests to other
2588   proxies &MUST; be able to recognize and exclude all of its own server
2589   names, including any aliases, local variations, or literal IP addresses.
2592   If a proxy receives a request-target with a host name that is not a
2593   fully qualified domain name, it &MAY; add its domain to the host name
2594   it received when forwarding the request.  A proxy &MUST-NOT; change the
2595   host name if it is a fully qualified domain name.
2598   A non-transforming proxy &MUST-NOT; rewrite the "path-absolute" and "query"
2599   parts of the received request-target when forwarding it to the next inbound
2600   server, except as noted above to replace an empty path with "/" or "*".
2603   Intermediaries that forward a message &MUST; implement the
2604   <x:ref>Connection</x:ref> header field as specified in
2605   <xref target="header.connection"/>.
2608<section title="End-to-end and Hop-by-hop Header Fields" anchor="end-to-end.and.hop-by-hop.header-fields">
2610  <cref anchor="TODO-end-to-end" source="jre">
2611    Restored from <eref target=""/>.
2612    See also <eref target=""/>.
2613  </cref>
2616   For the purpose of defining the behavior of caches and non-caching
2617   proxies, we divide HTTP header fields into two categories:
2618  <list style="symbols">
2619      <t>End-to-end header fields, which are  transmitted to the ultimate
2620        recipient of a request or response. End-to-end header fields in
2621        responses &MUST; be stored as part of a cache entry and &MUST; be
2622        transmitted in any response formed from a cache entry.</t>
2624      <t>Hop-by-hop header fields, which are meaningful only for a single
2625        transport-level connection, and are not stored by caches or
2626        forwarded by proxies.</t>
2627  </list>
2630   The following HTTP/1.1 header fields are hop-by-hop header fields:
2631  <list style="symbols">
2632      <t><x:ref>Connection</x:ref></t>
2633      <t>Keep-Alive (<xref target="RFC2068" x:fmt="of" x:sec=""/>)</t>
2634      <t><x:ref>Proxy-Authenticate</x:ref> (&header-proxy-authenticate;)</t>
2635      <t><x:ref>Proxy-Authorization</x:ref> (&header-proxy-authorization;)</t>
2636      <t><x:ref>TE</x:ref></t>
2637      <t><x:ref>Trailer</x:ref></t>
2638      <t><x:ref>Transfer-Encoding</x:ref></t>
2639      <t><x:ref>Upgrade</x:ref></t>
2640  </list>
2643   All other header fields defined by HTTP/1.1 are end-to-end header fields.
2646   Other hop-by-hop header fields &MUST; be listed in a
2647   <x:ref>Connection</x:ref> header field (<xref target="header.connection"/>).
2651<section title="Non-modifiable Header Fields" anchor="non-modifiable.header-fields">
2653  <cref anchor="TODO-non-mod-headers" source="jre">
2654    Restored from <eref target=""/>.
2655    See also <eref target=""/>.
2656  </cref>
2659   Some features of HTTP/1.1, such as Digest Authentication, depend on the
2660   value of certain end-to-end header fields. A non-transforming proxy &SHOULD-NOT;
2661   modify an end-to-end header field unless the definition of that header field requires
2662   or specifically allows that.
2665   A non-transforming proxy &MUST-NOT; modify any of the following fields in a
2666   request or response, and it &MUST-NOT; add any of these fields if not
2667   already present:
2668  <list style="symbols">
2669    <t><x:ref>Allow</x:ref> (&header-allow;)</t>
2670    <t><x:ref>Content-Location</x:ref> (&header-content-location;)</t>
2671    <t>Content-MD5 (<xref target="RFC2616" x:fmt="of" x:sec="14.15"/>)</t>
2672    <t><x:ref>ETag</x:ref> (&header-etag;)</t>
2673    <t><x:ref>Last-Modified</x:ref> (&header-last-modified;)</t>
2674    <t><x:ref>Server</x:ref> (&header-server;)</t>
2675  </list>
2678   A non-transforming proxy &MUST-NOT; modify any of the following fields in a
2679   response:
2680  <list style="symbols">
2681    <t><x:ref>Expires</x:ref> (&header-expires;)</t>
2682  </list>
2685   but it &MAY; add any of these fields if not already present. If an
2686   <x:ref>Expires</x:ref> header field is added, it &MUST; be given a
2687   field value identical to that of the <x:ref>Date</x:ref> header field in
2688   that response.
2691   A proxy &MUST-NOT; modify or add any of the following fields in a
2692   message that contains the no-transform cache-control directive, or in
2693   any request:
2694  <list style="symbols">
2695    <t><x:ref>Content-Encoding</x:ref> (&header-content-encoding;)</t>
2696    <t><x:ref>Content-Range</x:ref> (&header-content-range;)</t>
2697    <t><x:ref>Content-Type</x:ref> (&header-content-type;)</t>
2698  </list>
2701   A transforming proxy &MAY; modify or add these fields to a message
2702   that does not include no-transform, but if it does so, it &MUST; add a
2703   Warning 214 (Transformation applied) if one does not already appear
2704   in the message (see &header-warning;).
2707  <t>
2708    <x:h>Warning:</x:h> Unnecessary modification of end-to-end header fields might
2709    cause authentication failures if stronger authentication
2710    mechanisms are introduced in later versions of HTTP. Such
2711    authentication mechanisms &MAY; rely on the values of header fields
2712    not listed here.
2713  </t>
2716   A non-transforming proxy &MUST; preserve the message payload (&payload;),
2717   though it &MAY; change the message body through application or removal
2718   of a transfer-coding (<xref target="transfer.codings"/>).
2724<section title="Associating a Response to a Request" anchor="">
2726   HTTP does not include a request identifier for associating a given
2727   request message with its corresponding one or more response messages.
2728   Hence, it relies on the order of response arrival to correspond exactly
2729   to the order in which requests are made on the same connection.
2730   More than one response message per request only occurs when one or more
2731   informational responses (<x:ref>1xx</x:ref>, see &status-1xx;) precede a final response
2732   to the same request.
2735   A client that uses persistent connections and sends more than one request
2736   per connection &MUST; maintain a list of outstanding requests in the
2737   order sent on that connection and &MUST; associate each received response
2738   message to the highest ordered request that has not yet received a final
2739   (non-<x:ref>1xx</x:ref>) response.
2744<section title="Connection Management" anchor="">
2746<section title="Connection" anchor="header.connection">
2747  <iref primary="true" item="Connection header field" x:for-anchor=""/>
2748  <iref primary="true" item="Header Fields" subitem="Connection" x:for-anchor=""/>
2749  <x:anchor-alias value="Connection"/>
2750  <x:anchor-alias value="connection-option"/>
2752   The "Connection" header field allows the sender to specify
2753   options that are desired only for that particular connection.
2754   Such connection options &MUST; be removed or replaced before the
2755   message can be forwarded downstream by a proxy or gateway.
2756   This mechanism also allows the sender to indicate which HTTP
2757   header fields used in the message are only intended for the
2758   immediate recipient ("hop-by-hop"), as opposed to all recipients
2759   on the chain ("end-to-end"), enabling the message to be
2760   self-descriptive and allowing future connection-specific extensions
2761   to be deployed in HTTP without fear that they will be blindly
2762   forwarded by previously deployed intermediaries.
2765   The Connection header field's value has the following grammar:
2767<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Connection"/><iref primary="true" item="Grammar" subitem="connection-option"/>
2768  <x:ref>Connection</x:ref>        = 1#<x:ref>connection-option</x:ref>
2769  <x:ref>connection-option</x:ref> = <x:ref>token</x:ref>
2772   Connection options are compared case-insensitively.
2775   A proxy or gateway &MUST; parse a received Connection
2776   header field before a message is forwarded and, for each
2777   connection-option in this field, remove any header field(s) from
2778   the message with the same name as the connection-option, and then
2779   remove the Connection header field itself or replace it with the
2780   sender's own connection options for the forwarded message.
2783   A sender &MUST-NOT; include field-names in the Connection header
2784   field-value for fields that are defined as expressing constraints
2785   for all recipients in the request or response chain, such as the
2786   Cache-Control header field (&header-cache-control;).
2789   The connection options do not have to correspond to a header field
2790   present in the message, since a connection-specific header field
2791   might not be needed if there are no parameters associated with that
2792   connection option.  Recipients that trigger certain connection
2793   behavior based on the presence of connection options &MUST; do so
2794   based on the presence of the connection-option rather than only the
2795   presence of the optional header field.  In other words, if the
2796   connection option is received as a header field but not indicated
2797   within the Connection field-value, then the recipient &MUST; ignore
2798   the connection-specific header field because it has likely been
2799   forwarded by an intermediary that is only partially conformant.
2802   When defining new connection options, specifications ought to
2803   carefully consider existing deployed header fields and ensure
2804   that the new connection option does not share the same name as
2805   an unrelated header field that might already be deployed.
2806   Defining a new connection option essentially reserves that potential
2807   field-name for carrying additional information related to the
2808   connection option, since it would be unwise for senders to use
2809   that field-name for anything else.
2812   HTTP/1.1 defines the "close" connection option for the sender to
2813   signal that the connection will be closed after completion of the
2814   response. For example,
2816<figure><artwork type="example">
2817  Connection: close
2820   in either the request or the response header fields indicates that
2821   the connection &SHOULD-NOT;  be considered "persistent" (<xref target="persistent.connections"/>)
2822   after the current request/response is complete.
2825   An HTTP/1.1 client that does not support persistent connections &MUST;
2826   include the "close" connection option in every request message.
2829   An HTTP/1.1 server that does not support persistent connections &MUST;
2830   include the "close" connection option in every response message that
2831   does not have a <x:ref>1xx (Informational)</x:ref> status code.
2835<section title="Via" anchor="header.via">
2836  <iref primary="true" item="Via header field" x:for-anchor=""/>
2837  <iref primary="true" item="Header Fields" subitem="Via" x:for-anchor=""/>
2838  <x:anchor-alias value="pseudonym"/>
2839  <x:anchor-alias value="received-by"/>
2840  <x:anchor-alias value="received-protocol"/>
2841  <x:anchor-alias value="Via"/>
2843   The "Via" header field &MUST; be sent by a proxy or gateway to
2844   indicate the intermediate protocols and recipients between the user
2845   agent and the server on requests, and between the origin server and
2846   the client on responses. It is analogous to the "Received" field
2847   used by email systems (<xref target="RFC5322" x:fmt="of" x:sec="3.6.7"/>)
2848   and is intended to be used for tracking message forwards,
2849   avoiding request loops, and identifying the protocol capabilities of
2850   all senders along the request/response chain.
2852<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Via"/><iref primary="true" item="Grammar" subitem="received-protocol"/><iref primary="true" item="Grammar" subitem="protocol-name"/><iref primary="true" item="Grammar" subitem="protocol-version"/><iref primary="true" item="Grammar" subitem="received-by"/><iref primary="true" item="Grammar" subitem="pseudonym"/>
2853  <x:ref>Via</x:ref>               = 1#( <x:ref>received-protocol</x:ref> <x:ref>RWS</x:ref> <x:ref>received-by</x:ref>
2854                          [ <x:ref>RWS</x:ref> <x:ref>comment</x:ref> ] )
2855  <x:ref>received-protocol</x:ref> = [ <x:ref>protocol-name</x:ref> "/" ] <x:ref>protocol-version</x:ref>
2856  <x:ref>received-by</x:ref>       = ( <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ) / <x:ref>pseudonym</x:ref>
2857  <x:ref>pseudonym</x:ref>         = <x:ref>token</x:ref>
2860   The received-protocol indicates the protocol version of the message
2861   received by the server or client along each segment of the
2862   request/response chain. The received-protocol version is appended to
2863   the Via field value when the message is forwarded so that information
2864   about the protocol capabilities of upstream applications remains
2865   visible to all recipients.
2868   The protocol-name is excluded if and only if it would be "HTTP". The
2869   received-by field is normally the host and optional port number of a
2870   recipient server or client that subsequently forwarded the message.
2871   However, if the real host is considered to be sensitive information,
2872   it &MAY; be replaced by a pseudonym. If the port is not given, it &MAY;
2873   be assumed to be the default port of the received-protocol.
2876   Multiple Via field values represent each proxy or gateway that has
2877   forwarded the message. Each recipient &MUST; append its information
2878   such that the end result is ordered according to the sequence of
2879   forwarding applications.
2882   Comments &MAY; be used in the Via header field to identify the software
2883   of each recipient, analogous to the <x:ref>User-Agent</x:ref> and
2884   <x:ref>Server</x:ref> header fields. However, all comments in the Via field
2885   are optional and &MAY; be removed by any recipient prior to forwarding the
2886   message.
2889   For example, a request message could be sent from an HTTP/1.0 user
2890   agent to an internal proxy code-named "fred", which uses HTTP/1.1 to
2891   forward the request to a public proxy at, which completes
2892   the request by forwarding it to the origin server at
2893   The request received by would then have the following
2894   Via header field:
2896<figure><artwork type="example">
2897  Via: 1.0 fred, 1.1 (Apache/1.1)
2900   A proxy or gateway used as a portal through a network firewall
2901   &SHOULD-NOT; forward the names and ports of hosts within the firewall
2902   region unless it is explicitly enabled to do so. If not enabled, the
2903   received-by host of any host behind the firewall &SHOULD; be replaced
2904   by an appropriate pseudonym for that host.
2907   For organizations that have strong privacy requirements for hiding
2908   internal structures, a proxy or gateway &MAY; combine an ordered
2909   subsequence of Via header field entries with identical received-protocol
2910   values into a single such entry. For example,
2912<figure><artwork type="example">
2913  Via: 1.0 ricky, 1.1 ethel, 1.1 fred, 1.0 lucy
2916  could be collapsed to
2918<figure><artwork type="example">
2919  Via: 1.0 ricky, 1.1 mertz, 1.0 lucy
2922   Senders &SHOULD-NOT; combine multiple entries unless they are all
2923   under the same organizational control and the hosts have already been
2924   replaced by pseudonyms. Senders &MUST-NOT; combine entries which
2925   have different received-protocol values.
2929<section title="Persistent Connections" anchor="persistent.connections">
2931<section title="Purpose" anchor="persistent.purpose">
2933   Prior to persistent connections, a separate TCP connection was
2934   established for each request, increasing the load on HTTP servers
2935   and causing congestion on the Internet. The use of inline images and
2936   other associated data often requires a client to make multiple
2937   requests of the same server in a short amount of time. Analysis of
2938   these performance problems and results from a prototype
2939   implementation are available <xref target="Pad1995"/> <xref target="Spe"/>. Implementation experience and
2940   measurements of actual HTTP/1.1 implementations show good
2941   results <xref target="Nie1997"/>. Alternatives have also been explored, for example,
2942   T/TCP <xref target="Tou1998"/>.
2945   Persistent HTTP connections have a number of advantages:
2946  <list style="symbols">
2947      <t>
2948        By opening and closing fewer TCP connections, CPU time is saved
2949        in routers and hosts (clients, servers, proxies, gateways,
2950        tunnels, or caches), and memory used for TCP protocol control
2951        blocks can be saved in hosts.
2952      </t>
2953      <t>
2954        HTTP requests and responses can be pipelined on a connection.
2955        Pipelining allows a client to make multiple requests without
2956        waiting for each response, allowing a single TCP connection to
2957        be used much more efficiently, with much lower elapsed time.
2958      </t>
2959      <t>
2960        Network congestion is reduced by reducing the number of packets
2961        caused by TCP opens, and by allowing TCP sufficient time to
2962        determine the congestion state of the network.
2963      </t>
2964      <t>
2965        Latency on subsequent requests is reduced since there is no time
2966        spent in TCP's connection opening handshake.
2967      </t>
2968      <t>
2969        HTTP can evolve more gracefully, since errors can be reported
2970        without the penalty of closing the TCP connection. Clients using
2971        future versions of HTTP might optimistically try a new feature,
2972        but if communicating with an older server, retry with old
2973        semantics after an error is reported.
2974      </t>
2975    </list>
2978   HTTP implementations &SHOULD; implement persistent connections.
2982<section title="Overall Operation" anchor="persistent.overall">
2984   A significant difference between HTTP/1.1 and earlier versions of
2985   HTTP is that persistent connections are the default behavior of any
2986   HTTP connection. That is, unless otherwise indicated, the client
2987   &SHOULD; assume that the server will maintain a persistent connection,
2988   even after error responses from the server.
2991   Persistent connections provide a mechanism by which a client and a
2992   server can signal the close of a TCP connection. This signaling takes
2993   place using the <x:ref>Connection</x:ref> header field
2994   (<xref target="header.connection"/>). Once a close has been signaled, the
2995   client &MUST-NOT; send any more requests on that
2996   connection.
2999<section title="Negotiation" anchor="persistent.negotiation">
3001   An HTTP/1.1 server &MAY; assume that a HTTP/1.1 client intends to
3002   maintain a persistent connection unless a <x:ref>Connection</x:ref> header
3003   field including the connection option "close" was sent in the request. If
3004   the server chooses to close the connection immediately after sending the
3005   response, it &SHOULD; send a Connection header field including the
3006   connection option "close".
3009   An HTTP/1.1 client &MAY; expect a connection to remain open, but would
3010   decide to keep it open based on whether the response from a server
3011   contains a <x:ref>Connection</x:ref> header field with the connection option
3012   "close". In case the client does not want to maintain a connection for more
3013   than that request, it &SHOULD; send a Connection header field including the
3014   connection option "close".
3017   If either the client or the server sends the "close" option in the
3018   <x:ref>Connection</x:ref> header field, that request becomes the last one
3019   for the connection.
3022   Clients and servers &SHOULD-NOT;  assume that a persistent connection is
3023   maintained for HTTP versions less than 1.1 unless it is explicitly
3024   signaled. See <xref target="compatibility.with.http.1.0.persistent.connections"/> for more information on backward
3025   compatibility with HTTP/1.0 clients.
3028   Each persistent connection applies to only one transport link.
3031   A proxy server &MUST-NOT; establish a HTTP/1.1 persistent connection
3032   with an HTTP/1.0 client (but see <xref x:sec="19.7.1" x:fmt="of" target="RFC2068"/>
3033   for information and discussion of the problems with the Keep-Alive header field
3034   implemented by many HTTP/1.0 clients).
3037   In order to remain persistent, all messages on the connection &MUST;
3038   have a self-defined message length (i.e., one not defined by closure
3039   of the connection), as described in <xref target="message.body"/>.
3043<section title="Pipelining" anchor="pipelining">
3045   A client that supports persistent connections &MAY; "pipeline" its
3046   requests (i.e., send multiple requests without waiting for each
3047   response). A server &MUST; send its responses to those requests in the
3048   same order that the requests were received.
3051   Clients which assume persistent connections and pipeline immediately
3052   after connection establishment &SHOULD; be prepared to retry their
3053   connection if the first pipelined attempt fails. If a client does
3054   such a retry, it &MUST-NOT; pipeline before it knows the connection is
3055   persistent. Clients &MUST; also be prepared to resend their requests if
3056   the server closes the connection before sending all of the
3057   corresponding responses.
3060   Clients &SHOULD-NOT; pipeline requests using non-idempotent request methods or
3061   non-idempotent sequences of request methods (see &idempotent-methods;). Otherwise, a
3062   premature termination of the transport connection could lead to
3063   indeterminate results. A client wishing to send a non-idempotent
3064   request &SHOULD; wait to send that request until it has received the
3065   response status line for the previous request.
3070<section title="Practical Considerations" anchor="persistent.practical">
3072   Servers will usually have some time-out value beyond which they will
3073   no longer maintain an inactive connection. Proxy servers might make
3074   this a higher value since it is likely that the client will be making
3075   more connections through the same server. The use of persistent
3076   connections places no requirements on the length (or existence) of
3077   this time-out for either the client or the server.
3080   When a client or server wishes to time-out it &SHOULD; issue a graceful
3081   close on the transport connection. Clients and servers &SHOULD; both
3082   constantly watch for the other side of the transport close, and
3083   respond to it as appropriate. If a client or server does not detect
3084   the other side's close promptly it could cause unnecessary resource
3085   drain on the network.
3088   A client, server, or proxy &MAY; close the transport connection at any
3089   time. For example, a client might have started to send a new request
3090   at the same time that the server has decided to close the "idle"
3091   connection. From the server's point of view, the connection is being
3092   closed while it was idle, but from the client's point of view, a
3093   request is in progress.
3096   Clients (including proxies) &SHOULD; limit the number of simultaneous
3097   connections that they maintain to a given server (including proxies).
3100   Previous revisions of HTTP gave a specific number of connections as a
3101   ceiling, but this was found to be impractical for many applications. As a
3102   result, this specification does not mandate a particular maximum number of
3103   connections, but instead encourages clients to be conservative when opening
3104   multiple connections.
3107   In particular, while using multiple connections avoids the "head-of-line
3108   blocking" problem (whereby a request that takes significant server-side
3109   processing and/or has a large payload can block subsequent requests on the
3110   same connection), each connection used consumes server resources (sometimes
3111   significantly), and furthermore using multiple connections can cause
3112   undesirable side effects in congested networks.
3115   Note that servers might reject traffic that they deem abusive, including an
3116   excessive number of connections from a client.
3120<section title="Retrying Requests" anchor="persistent.retrying.requests">
3122   Senders can close the transport connection at any time. Therefore,
3123   clients, servers, and proxies &MUST; be able to recover
3124   from asynchronous close events. Client software &MAY; reopen the
3125   transport connection and retransmit the aborted sequence of requests
3126   without user interaction so long as the request sequence is
3127   idempotent (see &idempotent-methods;). Non-idempotent request methods or sequences
3128   &MUST-NOT; be automatically retried, although user agents &MAY; offer a
3129   human operator the choice of retrying the request(s). Confirmation by
3130   user-agent software with semantic understanding of the application
3131   &MAY; substitute for user confirmation. The automatic retry &SHOULD-NOT;
3132   be repeated if the second sequence of requests fails.
3137<section title="Message Transmission Requirements" anchor="message.transmission.requirements">
3139<section title="Persistent Connections and Flow Control" anchor="persistent.flow">
3141   HTTP/1.1 servers &SHOULD; maintain persistent connections and use TCP's
3142   flow control mechanisms to resolve temporary overloads, rather than
3143   terminating connections with the expectation that clients will retry.
3144   The latter technique can exacerbate network congestion.
3148<section title="Monitoring Connections for Error Status Messages" anchor="persistent.monitor">
3150   An HTTP/1.1 (or later) client sending a message body &SHOULD; monitor
3151   the network connection for an error status code while it is transmitting
3152   the request. If the client sees an error status code, it &SHOULD;
3153   immediately cease transmitting the body. If the body is being sent
3154   using a "chunked" encoding (<xref target="transfer.codings"/>), a zero length chunk and
3155   empty trailer &MAY; be used to prematurely mark the end of the message.
3156   If the body was preceded by a Content-Length header field, the client &MUST;
3157   close the connection.
3161<section title="Use of the 100 (Continue) Status" anchor="use.of.the.100.status">
3163   The purpose of the <x:ref>100 (Continue)</x:ref> status code (see &status-100;)
3164   is to allow a client that is sending a request message with a request body
3165   to determine if the origin server is willing to accept the request
3166   (based on the request header fields) before the client sends the request
3167   body. In some cases, it might either be inappropriate or highly
3168   inefficient for the client to send the body if the server will reject
3169   the message without looking at the body.
3172   Requirements for HTTP/1.1 clients:
3173  <list style="symbols">
3174    <t>
3175        If a client will wait for a <x:ref>100 (Continue)</x:ref> response before
3176        sending the request body, it &MUST; send an <x:ref>Expect</x:ref> header
3177        field (&header-expect;) with the "100-continue" expectation.
3178    </t>
3179    <t>
3180        A client &MUST-NOT; send an <x:ref>Expect</x:ref> header field with
3181        the "100-continue" expectation if it does not intend to send a request
3182        body.
3183    </t>
3184  </list>
3187   Because of the presence of older implementations, the protocol allows
3188   ambiguous situations in which a client might send "Expect: 100-continue"
3189   without receiving either a <x:ref>417 (Expectation Failed)</x:ref>
3190   or a <x:ref>100 (Continue)</x:ref> status code. Therefore, when a client sends this
3191   header field to an origin server (possibly via a proxy) from which it
3192   has never seen a <x:ref>100 (Continue)</x:ref> status code, the client &SHOULD-NOT; 
3193   wait for an indefinite period before sending the request body.
3196   Requirements for HTTP/1.1 origin servers:
3197  <list style="symbols">
3198    <t> Upon receiving a request which includes an <x:ref>Expect</x:ref> header
3199        field with the "100-continue" expectation, an origin server &MUST;
3200        either respond with <x:ref>100 (Continue)</x:ref> status code and continue to read
3201        from the input stream, or respond with a final status code. The
3202        origin server &MUST-NOT; wait for the request body before sending
3203        the <x:ref>100 (Continue)</x:ref> response. If it responds with a final status
3204        code, it &MAY; close the transport connection or it &MAY; continue
3205        to read and discard the rest of the request.  It &MUST-NOT;
3206        perform the request method if it returns a final status code.
3207    </t>
3208    <t> An origin server &SHOULD-NOT;  send a <x:ref>100 (Continue)</x:ref> response if
3209        the request message does not include an <x:ref>Expect</x:ref> header
3210        field with the "100-continue" expectation, and &MUST-NOT; send a
3211        <x:ref>100 (Continue)</x:ref> response if such a request comes from an HTTP/1.0
3212        (or earlier) client. There is an exception to this rule: for
3213        compatibility with <xref target="RFC2068"/>, a server &MAY; send a <x:ref>100 (Continue)</x:ref>
3214        status code in response to an HTTP/1.1 PUT or POST request that does
3215        not include an Expect header field with the "100-continue"
3216        expectation. This exception, the purpose of which is
3217        to minimize any client processing delays associated with an
3218        undeclared wait for <x:ref>100 (Continue)</x:ref> status code, applies only to
3219        HTTP/1.1 requests, and not to requests with any other HTTP-version
3220        value.
3221    </t>
3222    <t> An origin server &MAY; omit a <x:ref>100 (Continue)</x:ref> response if it has
3223        already received some or all of the request body for the
3224        corresponding request.
3225    </t>
3226    <t> An origin server that sends a <x:ref>100 (Continue)</x:ref> response &MUST;
3227        ultimately send a final status code, once the request body is
3228        received and processed, unless it terminates the transport
3229        connection prematurely.
3230    </t>
3231    <t> If an origin server receives a request that does not include an
3232        <x:ref>Expect</x:ref> header field with the "100-continue" expectation,
3233        the request includes a request body, and the server responds
3234        with a final status code before reading the entire request body
3235        from the transport connection, then the server &SHOULD-NOT;  close
3236        the transport connection until it has read the entire request,
3237        or until the client closes the connection. Otherwise, the client
3238        might not reliably receive the response message. However, this
3239        requirement ought not be construed as preventing a server from
3240        defending itself against denial-of-service attacks, or from
3241        badly broken client implementations.
3242      </t>
3243    </list>
3246   Requirements for HTTP/1.1 proxies:
3247  <list style="symbols">
3248    <t> If a proxy receives a request that includes an <x:ref>Expect</x:ref>
3249        header field with the "100-continue" expectation, and the proxy
3250        either knows that the next-hop server complies with HTTP/1.1 or
3251        higher, or does not know the HTTP version of the next-hop
3252        server, it &MUST; forward the request, including the Expect header
3253        field.
3254    </t>
3255    <t> If the proxy knows that the version of the next-hop server is
3256        HTTP/1.0 or lower, it &MUST-NOT; forward the request, and it &MUST;
3257        respond with a <x:ref>417 (Expectation Failed)</x:ref> status code.
3258    </t>
3259    <t> Proxies &SHOULD; maintain a record of the HTTP version
3260        numbers received from recently-referenced next-hop servers.
3261    </t>
3262    <t> A proxy &MUST-NOT; forward a <x:ref>100 (Continue)</x:ref> response if the
3263        request message was received from an HTTP/1.0 (or earlier)
3264        client and did not include an <x:ref>Expect</x:ref> header field with
3265        the "100-continue" expectation. This requirement overrides the
3266        general rule for forwarding of <x:ref>1xx</x:ref> responses (see &status-1xx;).
3267    </t>
3268  </list>
3272<section title="Closing Connections on Error" anchor="closing.connections.on.error">
3274   If the client is sending data, a server implementation using TCP
3275   &SHOULD; be careful to ensure that the client acknowledges receipt of
3276   the packet(s) containing the response, before the server closes the
3277   input connection. If the client continues sending data to the server
3278   after the close, the server's TCP stack will send a reset packet to
3279   the client, which might erase the client's unacknowledged input buffers
3280   before they can be read and interpreted by the HTTP application.
3286<section title="Upgrade" anchor="header.upgrade">
3287  <iref primary="true" item="Upgrade header field" x:for-anchor=""/>
3288  <iref primary="true" item="Header Fields" subitem="Upgrade" x:for-anchor=""/>
3289  <x:anchor-alias value="Upgrade"/>
3290  <x:anchor-alias value="protocol"/>
3291  <x:anchor-alias value="protocol-name"/>
3292  <x:anchor-alias value="protocol-version"/>
3294   The "Upgrade" header field allows the client to specify what
3295   additional communication protocols it would like to use, if the server
3296   chooses to switch protocols. Servers can use it to indicate what protocols
3297   they are willing to switch to.
3299<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Upgrade"/>
3300  <x:ref>Upgrade</x:ref>          = 1#<x:ref>protocol</x:ref>
3302  <x:ref>protocol</x:ref>         = <x:ref>protocol-name</x:ref> ["/" <x:ref>protocol-version</x:ref>]
3303  <x:ref>protocol-name</x:ref>    = <x:ref>token</x:ref>
3304  <x:ref>protocol-version</x:ref> = <x:ref>token</x:ref>
3307   For example,
3309<figure><artwork type="example">
3310  Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11
3313   The Upgrade header field is intended to provide a simple mechanism
3314   for transitioning from HTTP/1.1 to some other, incompatible protocol. It
3315   does so by allowing the client to advertise its desire to use another
3316   protocol, such as a later version of HTTP with a higher major version
3317   number, even though the current request has been made using HTTP/1.1.
3318   This eases the difficult transition between incompatible protocols by
3319   allowing the client to initiate a request in the more commonly
3320   supported protocol while indicating to the server that it would like
3321   to use a "better" protocol if available (where "better" is determined
3322   by the server, possibly according to the nature of the request method
3323   or target resource).
3326   The Upgrade header field only applies to switching application-layer
3327   protocols upon the existing transport-layer connection. Upgrade
3328   cannot be used to insist on a protocol change; its acceptance and use
3329   by the server is optional. The capabilities and nature of the
3330   application-layer communication after the protocol change is entirely
3331   dependent upon the new protocol chosen, although the first action
3332   after changing the protocol &MUST; be a response to the initial HTTP
3333   request containing the Upgrade header field.
3336   The Upgrade header field only applies to the immediate connection.
3337   Therefore, the upgrade keyword &MUST; be supplied within a
3338   <x:ref>Connection</x:ref> header field (<xref target="header.connection"/>)
3339   whenever Upgrade is present in an HTTP/1.1 message.
3342   The Upgrade header field cannot be used to indicate a switch to a
3343   protocol on a different connection. For that purpose, it is more
3344   appropriate to use a <x:ref>3xx (Redirection)</x:ref> response (&status-3xx;).
3347   Servers &MUST; include the "Upgrade" header field in <x:ref>101 (Switching
3348   Protocols)</x:ref> responses to indicate which protocol(s) are being switched to,
3349   and &MUST; include it in <x:ref>426 (Upgrade Required)</x:ref> responses to indicate
3350   acceptable protocols to upgrade to. Servers &MAY; include it in any other
3351   response to indicate that they are willing to upgrade to one of the
3352   specified protocols.
3355   This specification only defines the protocol name "HTTP" for use by
3356   the family of Hypertext Transfer Protocols, as defined by the HTTP
3357   version rules of <xref target="http.version"/> and future updates to this
3358   specification. Additional tokens can be registered with IANA using the
3359   registration procedure defined in <xref target="upgrade.token.registry"/>.
3365<section title="IANA Considerations" anchor="IANA.considerations">
3367<section title="Header Field Registration" anchor="header.field.registration">
3369   HTTP header fields are registered within the Message Header Field Registry
3370   <xref target="RFC3864"/> maintained by IANA at
3371   <eref target=""/>.
3374   This document defines the following HTTP header fields, so their
3375   associated registry entries shall be updated according to the permanent
3376   registrations below:
3378<?BEGININC p1-messaging.iana-headers ?>
3379<!--AUTOGENERATED FROM extract-header-defs.xslt, do not edit manually-->
3380<texttable align="left" suppress-title="true" anchor="iana.header.registration.table">
3381   <ttcol>Header Field Name</ttcol>
3382   <ttcol>Protocol</ttcol>
3383   <ttcol>Status</ttcol>
3384   <ttcol>Reference</ttcol>
3386   <c>Connection</c>
3387   <c>http</c>
3388   <c>standard</c>
3389   <c>
3390      <xref target="header.connection"/>
3391   </c>
3392   <c>Content-Length</c>
3393   <c>http</c>
3394   <c>standard</c>
3395   <c>
3396      <xref target="header.content-length"/>
3397   </c>
3398   <c>Host</c>
3399   <c>http</c>
3400   <c>standard</c>
3401   <c>
3402      <xref target=""/>
3403   </c>
3404   <c>TE</c>
3405   <c>http</c>
3406   <c>standard</c>
3407   <c>
3408      <xref target="header.te"/>
3409   </c>
3410   <c>Trailer</c>
3411   <c>http</c>
3412   <c>standard</c>
3413   <c>
3414      <xref target="header.trailer"/>
3415   </c>
3416   <c>Transfer-Encoding</c>
3417   <c>http</c>
3418   <c>standard</c>
3419   <c>
3420      <xref target="header.transfer-encoding"/>
3421   </c>
3422   <c>Upgrade</c>
3423   <c>http</c>
3424   <c>standard</c>
3425   <c>
3426      <xref target="header.upgrade"/>
3427   </c>
3428   <c>Via</c>
3429   <c>http</c>
3430   <c>standard</c>
3431   <c>
3432      <xref target="header.via"/>
3433   </c>
3436<?ENDINC p1-messaging.iana-headers ?>
3438   Furthermore, the header field-name "Close" shall be registered as
3439   "reserved", since using that name as an HTTP header field might
3440   conflict with the "close" connection option of the "<x:ref>Connection</x:ref>"
3441   header field (<xref target="header.connection"/>).
3443<texttable align="left" suppress-title="true">
3444   <ttcol>Header Field Name</ttcol>
3445   <ttcol>Protocol</ttcol>
3446   <ttcol>Status</ttcol>
3447   <ttcol>Reference</ttcol>
3449   <c>Close</c>
3450   <c>http</c>
3451   <c>reserved</c>
3452   <c>
3453      <xref target="header.field.registration"/>
3454   </c>
3457   The change controller is: "IETF ( - Internet Engineering Task Force".
3461<section title="URI Scheme Registration" anchor="uri.scheme.registration">
3463   IANA maintains the registry of URI Schemes <xref target="RFC4395"/> at
3464   <eref target=""/>.
3467   This document defines the following URI schemes, so their
3468   associated registry entries shall be updated according to the permanent
3469   registrations below:
3471<texttable align="left" suppress-title="true">
3472   <ttcol>URI Scheme</ttcol>
3473   <ttcol>Description</ttcol>
3474   <ttcol>Reference</ttcol>
3476   <c>http</c>
3477   <c>Hypertext Transfer Protocol</c>
3478   <c><xref target="http.uri"/></c>
3480   <c>https</c>
3481   <c>Hypertext Transfer Protocol Secure</c>
3482   <c><xref target="https.uri"/></c>
3486<section title="Internet Media Type Registrations" anchor="">
3488   This document serves as the specification for the Internet media types
3489   "message/http" and "application/http". The following is to be registered with
3490   IANA (see <xref target="RFC4288"/>).
3492<section title="Internet Media Type message/http" anchor="">
3493<iref item="Media Type" subitem="message/http" primary="true"/>
3494<iref item="message/http Media Type" primary="true"/>
3496   The message/http type can be used to enclose a single HTTP request or
3497   response message, provided that it obeys the MIME restrictions for all
3498   "message" types regarding line length and encodings.
3501  <list style="hanging" x:indent="12em">
3502    <t hangText="Type name:">
3503      message
3504    </t>
3505    <t hangText="Subtype name:">
3506      http
3507    </t>
3508    <t hangText="Required parameters:">
3509      none
3510    </t>
3511    <t hangText="Optional parameters:">
3512      version, msgtype
3513      <list style="hanging">
3514        <t hangText="version:">
3515          The HTTP-version number of the enclosed message
3516          (e.g., "1.1"). If not present, the version can be
3517          determined from the first line of the body.
3518        </t>
3519        <t hangText="msgtype:">
3520          The message type &mdash; "request" or "response". If not
3521          present, the type can be determined from the first
3522          line of the body.
3523        </t>
3524      </list>
3525    </t>
3526    <t hangText="Encoding considerations:">
3527      only "7bit", "8bit", or "binary" are permitted
3528    </t>
3529    <t hangText="Security considerations:">
3530      none
3531    </t>
3532    <t hangText="Interoperability considerations:">
3533      none
3534    </t>
3535    <t hangText="Published specification:">
3536      This specification (see <xref target=""/>).
3537    </t>
3538    <t hangText="Applications that use this media type:">
3539    </t>
3540    <t hangText="Additional information:">
3541      <list style="hanging">
3542        <t hangText="Magic number(s):">none</t>
3543        <t hangText="File extension(s):">none</t>
3544        <t hangText="Macintosh file type code(s):">none</t>
3545      </list>
3546    </t>
3547    <t hangText="Person and email address to contact for further information:">
3548      See Authors Section.
3549    </t>
3550    <t hangText="Intended usage:">
3551      COMMON
3552    </t>
3553    <t hangText="Restrictions on usage:">
3554      none
3555    </t>
3556    <t hangText="Author/Change controller:">
3557      IESG
3558    </t>
3559  </list>
3562<section title="Internet Media Type application/http" anchor="">
3563<iref item="Media Type" subitem="application/http" primary="true"/>
3564<iref item="application/http Media Type" primary="true"/>
3566   The application/http type can be used to enclose a pipeline of one or more
3567   HTTP request or response messages (not intermixed).
3570  <list style="hanging" x:indent="12em">
3571    <t hangText="Type name:">
3572      application
3573    </t>
3574    <t hangText="Subtype name:">
3575      http
3576    </t>
3577    <t hangText="Required parameters:">
3578      none
3579    </t>
3580    <t hangText="Optional parameters:">
3581      version, msgtype
3582      <list style="hanging">
3583        <t hangText="version:">
3584          The HTTP-version number of the enclosed messages
3585          (e.g., "1.1"). If not present, the version can be
3586          determined from the first line of the body.
3587        </t>
3588        <t hangText="msgtype:">
3589          The message type &mdash; "request" or "response". If not
3590          present, the type can be determined from the first
3591          line of the body.
3592        </t>
3593      </list>
3594    </t>
3595    <t hangText="Encoding considerations:">
3596      HTTP messages enclosed by this type
3597      are in "binary" format; use of an appropriate
3598      Content-Transfer-Encoding is required when
3599      transmitted via E-mail.
3600    </t>
3601    <t hangText="Security considerations:">
3602      none
3603    </t>
3604    <t hangText="Interoperability considerations:">
3605      none
3606    </t>
3607    <t hangText="Published specification:">
3608      This specification (see <xref target=""/>).
3609    </t>
3610    <t hangText="Applications that use this media type:">
3611    </t>
3612    <t hangText="Additional information:">
3613      <list style="hanging">
3614        <t hangText="Magic number(s):">none</t>
3615        <t hangText="File extension(s):">none</t>
3616        <t hangText="Macintosh file type code(s):">none</t>
3617      </list>
3618    </t>
3619    <t hangText="Person and email address to contact for further information:">
3620      See Authors Section.
3621    </t>
3622    <t hangText="Intended usage:">
3623      COMMON
3624    </t>
3625    <t hangText="Restrictions on usage:">
3626      none
3627    </t>
3628    <t hangText="Author/Change controller:">
3629      IESG
3630    </t>
3631  </list>
3636<section title="Transfer Coding Registry" anchor="transfer.coding.registry">
3638   The HTTP Transfer Coding Registry defines the name space for transfer
3639   coding names.
3642   Registrations &MUST; include the following fields:
3643   <list style="symbols">
3644     <t>Name</t>
3645     <t>Description</t>
3646     <t>Pointer to specification text</t>
3647   </list>
3650   Names of transfer codings &MUST-NOT; overlap with names of content codings
3651   (&content-codings;) unless the encoding transformation is identical, as it
3652   is the case for the compression codings defined in
3653   <xref target="compression.codings"/>.
3656   Values to be added to this name space require IETF Review (see
3657   <xref target="RFC5226" x:fmt="of" x:sec="4.1"/>), and &MUST;
3658   conform to the purpose of transfer coding defined in this section.
3661   The registry itself is maintained at
3662   <eref target=""/>.
3666<section title="Transfer Coding Registrations" anchor="transfer.coding.registration">
3668   The HTTP Transfer Coding Registry shall be updated with the registrations
3669   below:
3671<texttable align="left" suppress-title="true" anchor="iana.transfer.coding.registration.table">
3672   <ttcol>Name</ttcol>
3673   <ttcol>Description</ttcol>
3674   <ttcol>Reference</ttcol>
3675   <c>chunked</c>
3676   <c>Transfer in a series of chunks</c>
3677   <c>
3678      <xref target="chunked.encoding"/>
3679   </c>
3680   <c>compress</c>
3681   <c>UNIX "compress" program method</c>
3682   <c>
3683      <xref target="compress.coding"/>
3684   </c>
3685   <c>deflate</c>
3686   <c>"deflate" compression mechanism (<xref target="RFC1951"/>) used inside
3687   the "zlib" data format (<xref target="RFC1950"/>)
3688   </c>
3689   <c>
3690      <xref target="deflate.coding"/>
3691   </c>
3692   <c>gzip</c>
3693   <c>Same as GNU zip <xref target="RFC1952"/></c>
3694   <c>
3695      <xref target="gzip.coding"/>
3696   </c>
3700<section title="Upgrade Token Registry" anchor="upgrade.token.registry">
3702   The HTTP Upgrade Token Registry defines the name space for protocol-name
3703   tokens used to identify protocols in the <x:ref>Upgrade</x:ref> header
3704   field. Each registered protocol name is associated with contact information
3705   and an optional set of specifications that details how the connection
3706   will be processed after it has been upgraded.
3709   Registrations happen on a "First Come First Served" basis (see
3710   <xref target="RFC5226" x:sec="4.1" x:fmt="of"/>) and are subject to the
3711   following rules:
3712  <list style="numbers">
3713    <t>A protocol-name token, once registered, stays registered forever.</t>
3714    <t>The registration &MUST; name a responsible party for the
3715       registration.</t>
3716    <t>The registration &MUST; name a point of contact.</t>
3717    <t>The registration &MAY; name a set of specifications associated with
3718       that token. Such specifications need not be publicly available.</t>
3719    <t>The registration &SHOULD; name a set of expected "protocol-version"
3720       tokens associated with that token at the time of registration.</t>
3721    <t>The responsible party &MAY; change the registration at any time.
3722       The IANA will keep a record of all such changes, and make them
3723       available upon request.</t>
3724    <t>The IESG &MAY; reassign responsibility for a protocol token.
3725       This will normally only be used in the case when a
3726       responsible party cannot be contacted.</t>
3727  </list>
3730   This registration procedure for HTTP Upgrade Tokens replaces that
3731   previously defined in <xref target="RFC2817" x:fmt="of" x:sec="7.2"/>.
3735<section title="Upgrade Token Registration" anchor="upgrade.token.registration">
3737   The HTTP Upgrade Token Registry shall be updated with the registration
3738   below:
3740<texttable align="left" suppress-title="true">
3741   <ttcol>Value</ttcol>
3742   <ttcol>Description</ttcol>
3743   <ttcol>Expected Version Tokens</ttcol>
3744   <ttcol>Reference</ttcol>
3746   <c>HTTP</c>
3747   <c>Hypertext Transfer Protocol</c>
3748   <c>any DIGIT.DIGIT (e.g, "2.0")</c>
3749   <c><xref target="http.version"/></c>
3752   The responsible party is: "IETF ( - Internet Engineering Task Force".
3758<section title="Security Considerations" anchor="security.considerations">
3760   This section is meant to inform application developers, information
3761   providers, and users of the security limitations in HTTP/1.1 as
3762   described by this document. The discussion does not include
3763   definitive solutions to the problems revealed, though it does make
3764   some suggestions for reducing security risks.
3767<section title="Personal Information" anchor="personal.information">
3769   HTTP clients are often privy to large amounts of personal information
3770   (e.g., the user's name, location, mail address, passwords, encryption
3771   keys, etc.), and &SHOULD; be very careful to prevent unintentional
3772   leakage of this information.
3773   We very strongly recommend that a convenient interface be provided
3774   for the user to control dissemination of such information, and that
3775   designers and implementors be particularly careful in this area.
3776   History shows that errors in this area often create serious security
3777   and/or privacy problems and generate highly adverse publicity for the
3778   implementor's company.
3782<section title="Abuse of Server Log Information" anchor="abuse.of.server.log.information">
3784   A server is in the position to save personal data about a user's
3785   requests which might identify their reading patterns or subjects of
3786   interest.  In particular, log information gathered at an intermediary
3787   often contains a history of user agent interaction, across a multitude
3788   of sites, that can be traced to individual users.
3791   HTTP log information is confidential in nature; its handling is often
3792   constrained by laws and regulations.  Log information needs to be securely
3793   stored and appropriate guidelines followed for its analysis.
3794   Anonymization of personal information within individual entries helps,
3795   but is generally not sufficient to prevent real log traces from being
3796   re-identified based on correlation with other access characteristics.
3797   As such, access traces that are keyed to a specific client should not
3798   be published even if the key is pseudonymous.
3801   To minimize the risk of theft or accidental publication, log information
3802   should be purged of personally identifiable information, including
3803   user identifiers, IP addresses, and user-provided query parameters,
3804   as soon as that information is no longer necessary to support operational
3805   needs for security, auditing, or fraud control.
3809<section title="Attacks Based On File and Path Names" anchor="attack.pathname">
3811   Implementations of HTTP origin servers &SHOULD; be careful to restrict
3812   the documents returned by HTTP requests to be only those that were
3813   intended by the server administrators. If an HTTP server translates
3814   HTTP URIs directly into file system calls, the server &MUST; take
3815   special care not to serve files that were not intended to be
3816   delivered to HTTP clients. For example, UNIX, Microsoft Windows, and
3817   other operating systems use ".." as a path component to indicate a
3818   directory level above the current one. On such a system, an HTTP
3819   server &MUST; disallow any such construct in the request-target if it
3820   would otherwise allow access to a resource outside those intended to
3821   be accessible via the HTTP server. Similarly, files intended for
3822   reference only internally to the server (such as access control
3823   files, configuration files, and script code) &MUST; be protected from
3824   inappropriate retrieval, since they might contain sensitive
3825   information. Experience has shown that minor bugs in such HTTP server
3826   implementations have turned into security risks.
3830<section title="DNS-related Attacks" anchor="dns.related.attacks">
3832   HTTP clients rely heavily on the Domain Name Service (DNS), and are thus
3833   generally prone to security attacks based on the deliberate misassociation
3834   of IP addresses and DNS names not protected by DNSSec. Clients need to be
3835   cautious in assuming the validity of an IP number/DNS name association unless
3836   the response is protected by DNSSec (<xref target="RFC4033"/>).
3840<section title="Intermediaries and Caching" anchor="attack.intermediaries">
3842   By their very nature, HTTP intermediaries are men-in-the-middle, and
3843   represent an opportunity for man-in-the-middle attacks. Compromise of
3844   the systems on which the intermediaries run can result in serious security
3845   and privacy problems. Intermediaries have access to security-related
3846   information, personal information about individual users and
3847   organizations, and proprietary information belonging to users and
3848   content providers. A compromised intermediary, or an intermediary
3849   implemented or configured without regard to security and privacy
3850   considerations, might be used in the commission of a wide range of
3851   potential attacks.
3854   Intermediaries that contain a shared cache are especially vulnerable
3855   to cache poisoning attacks.
3858   Implementors need to consider the privacy and security
3859   implications of their design and coding decisions, and of the
3860   configuration options they provide to operators (especially the
3861   default configuration).
3864   Users need to be aware that intermediaries are no more trustworthy than
3865   the people who run them; HTTP itself cannot solve this problem.
3868   The judicious use of cryptography, when appropriate, might suffice to
3869   protect against a broad range of security and privacy attacks. Such
3870   cryptography is beyond the scope of the HTTP/1.1 specification.
3874<section title="Protocol Element Size Overflows" anchor="attack.protocol.element.size.overflows">
3876   Because HTTP uses mostly textual, character-delimited fields, attackers can
3877   overflow buffers in implementations, and/or perform a Denial of Service
3878   against implementations that accept fields with unlimited lengths.
3881   To promote interoperability, this specification makes specific
3882   recommendations for minimum size limits on request-line
3883   (<xref target="request.line"/>)
3884   and blocks of header fields (<xref target="header.fields"/>). These are
3885   minimum recommendations, chosen to be supportable even by implementations
3886   with limited resources; it is expected that most implementations will
3887   choose substantially higher limits.
3890   This specification also provides a way for servers to reject messages that
3891   have request-targets that are too long (&status-414;) or request entities
3892   that are too large (&status-4xx;).
3895   Other fields (including but not limited to request methods, response status
3896   phrases, header field-names, and body chunks) &SHOULD; be limited by
3897   implementations carefully, so as to not impede interoperability.
3902<section title="Acknowledgments" anchor="acks">
3904   This edition of HTTP builds on the many contributions that went into
3905   <xref target="RFC1945" format="none">RFC 1945</xref>,
3906   <xref target="RFC2068" format="none">RFC 2068</xref>,
3907   <xref target="RFC2145" format="none">RFC 2145</xref>, and
3908   <xref target="RFC2616" format="none">RFC 2616</xref>, including
3909   substantial contributions made by the previous authors, editors, and
3910   working group chairs: Tim Berners-Lee, Ari Luotonen, Roy T. Fielding,
3911   Henrik Frystyk Nielsen, Jim Gettys, Jeffrey C. Mogul, Larry Masinter,
3912   Paul J. Leach, and Mark Nottingham.
3913   See <xref target="RFC2616" x:fmt="of" x:sec="16"/> for additional
3914   acknowledgements from prior revisions.
3917   Since 1999, the following contributors have helped improve the HTTP
3918   specification by reporting bugs, asking smart questions, drafting or
3919   reviewing text, and evaluating open issues:
3921<?BEGININC acks ?>
3922<t>Adam Barth,
3923Adam Roach,
3924Addison Phillips,
3925Adrian Chadd,
3926Adrien W. de Croy,
3927Alan Ford,
3928Alan Ruttenberg,
3929Albert Lunde,
3930Alek Storm,
3931Alex Rousskov,
3932Alexandre Morgaut,
3933Alexey Melnikov,
3934Alisha Smith,
3935Amichai Rothman,
3936Amit Klein,
3937Amos Jeffries,
3938Andreas Maier,
3939Andreas Petersson,
3940Anne van Kesteren,
3941Anthony Bryan,
3942Asbjorn Ulsberg,
3943Balachander Krishnamurthy,
3944Barry Leiba,
3945Ben Laurie,
3946Benjamin Niven-Jenkins,
3947Bil Corry,
3948Bill Burke,
3949Bjoern Hoehrmann,
3950Bob Scheifler,
3951Boris Zbarsky,
3952Brett Slatkin,
3953Brian Kell,
3954Brian McBarron,
3955Brian Pane,
3956Brian Smith,
3957Bryce Nesbitt,
3958Cameron Heavon-Jones,
3959Carl Kugler,
3960Carsten Bormann,
3961Charles Fry,
3962Chris Newman,
3963Cyrus Daboo,
3964Dale Robert Anderson,
3965Dan Winship,
3966Daniel Stenberg,
3967Dave Cridland,
3968Dave Crocker,
3969Dave Kristol,
3970David Booth,
3971David Singer,
3972David W. Morris,
3973Diwakar Shetty,
3974Dmitry Kurochkin,
3975Drummond Reed,
3976Duane Wessels,
3977Edward Lee,
3978Eliot Lear,
3979Eran Hammer-Lahav,
3980Eric D. Williams,
3981Eric J. Bowman,
3982Eric Lawrence,
3983Eric Rescorla,
3984Erik Aronesty,
3985Florian Weimer,
3986Frank Ellermann,
3987Fred Bohle,
3988Geoffrey Sneddon,
3989Gervase Markham,
3990Greg Wilkins,
3991Harald Tveit Alvestrand,
3992Harry Halpin,
3993Helge Hess,
3994Henrik Nordstrom,
3995Henry S. Thompson,
3996Henry Story,
3997Herbert van de Sompel,
3998Howard Melman,
3999Hugo Haas,
4000Ian Hickson,
4001Ingo Struck,
4002J. Ross Nicoll,
4003James H. Manger,
4004James Lacey,
4005James M. Snell,
4006Jamie Lokier,
4007Jan Algermissen,
4008Jeff Hodges (who came up with the term 'effective Request-URI'),
4009Jeff Walden,
4010Jim Luther,
4011Joe D. Williams,
4012Joe Gregorio,
4013Joe Orton,
4014John C. Klensin,
4015John C. Mallery,
4016John Cowan,
4017John Kemp,
4018John Panzer,
4019John Schneider,
4020John Stracke,
4021John Sullivan,
4022Jonas Sicking,
4023Jonathan Billington,
4024Jonathan Moore,
4025Jonathan Rees,
4026Jonathan Silvera,
4027Jordi Ros,
4028Joris Dobbelsteen,
4029Josh Cohen,
4030Julien Pierre,
4031Jungshik Shin,
4032Justin Chapweske,
4033Justin Erenkrantz,
4034Justin James,
4035Kalvinder Singh,
4036Karl Dubost,
4037Keith Hoffman,
4038Keith Moore,
4039Koen Holtman,
4040Konstantin Voronkov,
4041Kris Zyp,
4042Lisa Dusseault,
4043Maciej Stachowiak,
4044Marc Schneider,
4045Marc Slemko,
4046Mark Baker,
4047Mark Pauley,
4048Mark Watson,
4049Markus Isomaki,
4050Markus Lanthaler,
4051Martin J. Duerst,
4052Martin Musatov,
4053Martin Nilsson,
4054Martin Thomson,
4055Matt Lynch,
4056Matthew Cox,
4057Max Clark,
4058Michael Burrows,
4059Michael Hausenblas,
4060Mike Amundsen,
4061Mike Belshe,
4062Mike Kelly,
4063Mike Schinkel,
4064Miles Sabin,
4065Murray S. Kucherawy,
4066Mykyta Yevstifeyev,
4067Nathan Rixham,
4068Nicholas Shanks,
4069Nico Williams,
4070Nicolas Alvarez,
4071Nicolas Mailhot,
4072Noah Slater,
4073Pablo Castro,
4074Pat Hayes,
4075Patrick R. McManus,
4076Paul E. Jones,
4077Paul Hoffman,
4078Paul Marquess,
4079Peter Lepeska,
4080Peter Saint-Andre,
4081Peter Watkins,
4082Phil Archer,
4083Phillip Hallam-Baker,
4084Poul-Henning Kamp,
4085Preethi Natarajan,
4086Ray Polk,
4087Reto Bachmann-Gmuer,
4088Richard Cyganiak,
4089Robert Brewer,
4090Robert Collins,
4091Robert O'Callahan,
4092Robert Olofsson,
4093Robert Sayre,
4094Robert Siemer,
4095Robert de Wilde,
4096Roberto Javier Godoy,
4097Roberto Peon,
4098Ronny Widjaja,
4099S. Mike Dierken,
4100Salvatore Loreto,
4101Sam Johnston,
4102Sam Ruby,
4103Scott Lawrence (who maintained the original issues list),
4104Sean B. Palmer,
4105Shane McCarron,
4106Stefan Eissing,
4107Stefan Tilkov,
4108Stefanos Harhalakis,
4109Stephane Bortzmeyer,
4110Stephen Farrell,
4111Stephen Ludin,
4112Stuart Williams,
4113Subbu Allamaraju,
4114Sylvain Hellegouarch,
4115Tapan Divekar,
4116Ted Hardie,
4117Thomas Broyer,
4118Thomas Nordin,
4119Thomas Roessler,
4120Tim Bray,
4121Tim Morgan,
4122Tim Olsen,
4123Tom Zhou,
4124Travis Snoozy,
4125Tyler Close,
4126Vincent Murphy,
4127Wenbo Zhu,
4128Werner Baumann,
4129Wilbur Streett,
4130Wilfredo Sanchez Vega,
4131William A. Rowe Jr.,
4132William Chan,
4133Willy Tarreau,
4134Xiaoshu Wang,
4135Yaron Goland,
4136Yngve Nysaeter Pettersen,
4137Yoav Nir,
4138Yogesh Bang,
4139Yutaka Oiwa,
4140Zed A. Shaw, and
4141Zhong Yu.
4143<?ENDINC acks ?>
4149<references title="Normative References">
4151<reference anchor="ISO-8859-1">
4152  <front>
4153    <title>
4154     Information technology -- 8-bit single-byte coded graphic character sets -- Part 1: Latin alphabet No. 1
4155    </title>
4156    <author>
4157      <organization>International Organization for Standardization</organization>
4158    </author>
4159    <date year="1998"/>
4160  </front>
4161  <seriesInfo name="ISO/IEC" value="8859-1:1998"/>
4164<reference anchor="Part2">
4165  <front>
4166    <title>HTTP/1.1, part 2: Message Semantics, Payload and Content Negotiation</title>
4167    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4168      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4169      <address><email></email></address>
4170    </author>
4171    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
4172      <organization abbrev="W3C">World Wide Web Consortium</organization>
4173      <address><email></email></address>
4174    </author>
4175    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4176      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4177      <address><email></email></address>
4178    </author>
4179    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4180  </front>
4181  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p2-semantics-&ID-VERSION;"/>
4182  <x:source href="p2-semantics.xml" basename="p2-semantics">
4183    <x:defines>1xx (Informational)</x:defines>
4184    <x:defines>1xx</x:defines>
4185    <x:defines>100 (Continue)</x:defines>
4186    <x:defines>101 (Switching Protocols)</x:defines>
4187    <x:defines>2xx (Successful)</x:defines>
4188    <x:defines>2xx</x:defines>
4189    <x:defines>200 (OK)</x:defines>
4190    <x:defines>204 (No Content)</x:defines>
4191    <x:defines>3xx (Redirection)</x:defines>
4192    <x:defines>3xx</x:defines>
4193    <x:defines>301 (Moved Permanently)</x:defines>
4194    <x:defines>4xx (Client Error)</x:defines>
4195    <x:defines>4xx</x:defines>
4196    <x:defines>400 (Bad Request)</x:defines>
4197    <x:defines>405 (Method Not Allowed)</x:defines>
4198    <x:defines>411 (Length Required)</x:defines>
4199    <x:defines>414 (URI Too Long)</x:defines>
4200    <x:defines>417 (Expectation Failed)</x:defines>
4201    <x:defines>426 (Upgrade Required)</x:defines>
4202    <x:defines>501 (Not Implemented)</x:defines>
4203    <x:defines>502 (Bad Gateway)</x:defines>
4204    <x:defines>505 (HTTP Version Not Supported)</x:defines>
4205    <x:defines>Allow</x:defines>
4206    <x:defines>Content-Encoding</x:defines>
4207    <x:defines>Content-Location</x:defines>
4208    <x:defines>Content-Type</x:defines>
4209    <x:defines>Date</x:defines>
4210    <x:defines>Expect</x:defines>
4211    <x:defines>Location</x:defines>
4212    <x:defines>Server</x:defines>
4213    <x:defines>User-Agent</x:defines>
4214  </x:source>
4217<reference anchor="Part4">
4218  <front>
4219    <title>HTTP/1.1, part 4: Conditional Requests</title>
4220    <author fullname="Roy T. Fielding" initials="R." role="editor" surname="Fielding">
4221      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4222      <address><email></email></address>
4223    </author>
4224    <author fullname="Yves Lafon" initials="Y." role="editor" surname="Lafon">
4225      <organization abbrev="W3C">World Wide Web Consortium</organization>
4226      <address><email></email></address>
4227    </author>
4228    <author fullname="Julian F. Reschke" initials="J. F." role="editor" surname="Reschke">
4229      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4230      <address><email></email></address>
4231    </author>
4232    <date month="&ID-MONTH;" year="&ID-YEAR;" />
4233  </front>
4234  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p4-conditional-&ID-VERSION;" />
4235  <x:source basename="p4-conditional" href="p4-conditional.xml">
4236    <x:defines>304 (Not Modified)</x:defines>
4237    <x:defines>ETag</x:defines>
4238    <x:defines>Last-Modified</x:defines>
4239  </x:source>
4242<reference anchor="Part5">
4243  <front>
4244    <title>HTTP/1.1, part 5: Range Requests and Partial Responses</title>
4245    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4246      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4247      <address><email></email></address>
4248    </author>
4249    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
4250      <organization abbrev="W3C">World Wide Web Consortium</organization>
4251      <address><email></email></address>
4252    </author>
4253    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4254      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4255      <address><email></email></address>
4256    </author>
4257    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4258  </front>
4259  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p5-range-&ID-VERSION;"/>
4260  <x:source href="p5-range.xml" basename="p5-range">
4261    <x:defines>Content-Range</x:defines>
4262  </x:source>
4265<reference anchor="Part6">
4266  <front>
4267    <title>HTTP/1.1, part 6: Caching</title>
4268    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4269      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4270      <address><email></email></address>
4271    </author>
4272    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
4273      <organization abbrev="W3C">World Wide Web Consortium</organization>
4274      <address><email></email></address>
4275    </author>
4276    <author initials="M." surname="Nottingham" fullname="Mark Nottingham" role="editor">
4277      <organization>Rackspace</organization>
4278      <address><email></email></address>
4279    </author>
4280    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4281      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4282      <address><email></email></address>
4283    </author>
4284    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4285  </front>
4286  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p6-cache-&ID-VERSION;"/>
4287  <x:source href="p6-cache.xml" basename="p6-cache">
4288    <x:defines>Expires</x:defines>
4289  </x:source>
4292<reference anchor="Part7">
4293  <front>
4294    <title>HTTP/1.1, part 7: Authentication</title>
4295    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4296      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4297      <address><email></email></address>
4298    </author>
4299    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
4300      <organization abbrev="W3C">World Wide Web Consortium</organization>
4301      <address><email></email></address>
4302    </author>
4303    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4304      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4305      <address><email></email></address>
4306    </author>
4307    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4308  </front>
4309  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p7-auth-&ID-VERSION;"/>
4310  <x:source href="p7-auth.xml" basename="p7-auth">
4311    <x:defines>Proxy-Authenticate</x:defines>
4312    <x:defines>Proxy-Authorization</x:defines>
4313  </x:source>
4316<reference anchor="RFC5234">
4317  <front>
4318    <title abbrev="ABNF for Syntax Specifications">Augmented BNF for Syntax Specifications: ABNF</title>
4319    <author initials="D." surname="Crocker" fullname="Dave Crocker" role="editor">
4320      <organization>Brandenburg InternetWorking</organization>
4321      <address>
4322        <email></email>
4323      </address> 
4324    </author>
4325    <author initials="P." surname="Overell" fullname="Paul Overell">
4326      <organization>THUS plc.</organization>
4327      <address>
4328        <email></email>
4329      </address>
4330    </author>
4331    <date month="January" year="2008"/>
4332  </front>
4333  <seriesInfo name="STD" value="68"/>
4334  <seriesInfo name="RFC" value="5234"/>
4337<reference anchor="RFC2119">
4338  <front>
4339    <title>Key words for use in RFCs to Indicate Requirement Levels</title>
4340    <author initials="S." surname="Bradner" fullname="Scott Bradner">
4341      <organization>Harvard University</organization>
4342      <address><email></email></address>
4343    </author>
4344    <date month="March" year="1997"/>
4345  </front>
4346  <seriesInfo name="BCP" value="14"/>
4347  <seriesInfo name="RFC" value="2119"/>
4350<reference anchor="RFC3986">
4351 <front>
4352  <title abbrev='URI Generic Syntax'>Uniform Resource Identifier (URI): Generic Syntax</title>
4353  <author initials='T.' surname='Berners-Lee' fullname='Tim Berners-Lee'>
4354    <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
4355    <address>
4356       <email></email>
4357       <uri></uri>
4358    </address>
4359  </author>
4360  <author initials='R.' surname='Fielding' fullname='Roy T. Fielding'>
4361    <organization abbrev="Day Software">Day Software</organization>
4362    <address>
4363      <email></email>
4364      <uri></uri>
4365    </address>
4366  </author>
4367  <author initials='L.' surname='Masinter' fullname='Larry Masinter'>
4368    <organization abbrev="Adobe Systems">Adobe Systems Incorporated</organization>
4369    <address>
4370      <email></email>
4371      <uri></uri>
4372    </address>
4373  </author>
4374  <date month='January' year='2005'></date>
4375 </front>
4376 <seriesInfo name="STD" value="66"/>
4377 <seriesInfo name="RFC" value="3986"/>
4380<reference anchor="USASCII">
4381  <front>
4382    <title>Coded Character Set -- 7-bit American Standard Code for Information Interchange</title>
4383    <author>
4384      <organization>American National Standards Institute</organization>
4385    </author>
4386    <date year="1986"/>
4387  </front>
4388  <seriesInfo name="ANSI" value="X3.4"/>
4391<reference anchor="RFC1950">
4392  <front>
4393    <title>ZLIB Compressed Data Format Specification version 3.3</title>
4394    <author initials="L.P." surname="Deutsch" fullname="L. Peter Deutsch">
4395      <organization>Aladdin Enterprises</organization>
4396      <address><email></email></address>
4397    </author>
4398    <author initials="J-L." surname="Gailly" fullname="Jean-Loup Gailly"/>
4399    <date month="May" year="1996"/>
4400  </front>
4401  <seriesInfo name="RFC" value="1950"/>
4402  <!--<annotation>
4403    RFC 1950 is an Informational RFC, thus it might be less stable than
4404    this specification. On the other hand, this downward reference was
4405    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4406    therefore it is unlikely to cause problems in practice. See also
4407    <xref target="BCP97"/>.
4408  </annotation>-->
4411<reference anchor="RFC1951">
4412  <front>
4413    <title>DEFLATE Compressed Data Format Specification version 1.3</title>
4414    <author initials="P." surname="Deutsch" fullname="L. Peter Deutsch">
4415      <organization>Aladdin Enterprises</organization>
4416      <address><email></email></address>
4417    </author>
4418    <date month="May" year="1996"/>
4419  </front>
4420  <seriesInfo name="RFC" value="1951"/>
4421  <!--<annotation>
4422    RFC 1951 is an Informational RFC, thus it might be less stable than
4423    this specification. On the other hand, this downward reference was
4424    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4425    therefore it is unlikely to cause problems in practice. See also
4426    <xref target="BCP97"/>.
4427  </annotation>-->
4430<reference anchor="RFC1952">
4431  <front>
4432    <title>GZIP file format specification version 4.3</title>
4433    <author initials="P." surname="Deutsch" fullname="L. Peter Deutsch">
4434      <organization>Aladdin Enterprises</organization>
4435      <address><email></email></address>
4436    </author>
4437    <author initials="J-L." surname="Gailly" fullname="Jean-Loup Gailly">
4438      <address><email></email></address>
4439    </author>
4440    <author initials="M." surname="Adler" fullname="Mark Adler">
4441      <address><email></email></address>
4442    </author>
4443    <author initials="L.P." surname="Deutsch" fullname="L. Peter Deutsch">
4444      <address><email></email></address>
4445    </author>
4446    <author initials="G." surname="Randers-Pehrson" fullname="Glenn Randers-Pehrson">
4447      <address><email></email></address>
4448    </author>
4449    <date month="May" year="1996"/>
4450  </front>
4451  <seriesInfo name="RFC" value="1952"/>
4452  <!--<annotation>
4453    RFC 1952 is an Informational RFC, thus it might be less stable than
4454    this specification. On the other hand, this downward reference was
4455    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4456    therefore it is unlikely to cause problems in practice. See also
4457    <xref target="BCP97"/>.
4458  </annotation>-->
4463<references title="Informative References">
4465<reference anchor="Nie1997" target="">
4466  <front>
4467    <title>Network Performance Effects of HTTP/1.1, CSS1, and PNG</title>
4468    <author initials="H." surname="Frystyk" fullname="Henrik Frystyk Nielsen"/>
4469    <author initials="J." surname="Gettys" fullname="J. Gettys"/>
4470    <author initials="E." surname="Prud'hommeaux" fullname="E. Prud'hommeaux"/>
4471    <author initials="H." surname="Lie" fullname="H. Lie"/>
4472    <author initials="C." surname="Lilley" fullname="C. Lilley"/>
4473    <date year="1997" month="September"/>
4474  </front>
4475  <seriesInfo name="ACM" value="Proceedings of the ACM SIGCOMM '97 conference on Applications, technologies, architectures, and protocols for computer communication SIGCOMM '97"/>
4478<reference anchor="Pad1995" target="">
4479  <front>
4480    <title>Improving HTTP Latency</title>
4481    <author initials="V.N." surname="Padmanabhan" fullname="Venkata N. Padmanabhan"/>
4482    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul"/>
4483    <date year="1995" month="December"/>
4484  </front>
4485  <seriesInfo name="Computer Networks and ISDN Systems" value="v. 28, pp. 25-35"/>
4488<reference anchor='RFC1919'>
4489  <front>
4490    <title>Classical versus Transparent IP Proxies</title>
4491    <author initials='M.' surname='Chatel' fullname='Marc Chatel'>
4492      <address><email></email></address>
4493    </author>
4494    <date year='1996' month='March' />
4495  </front>
4496  <seriesInfo name='RFC' value='1919' />
4499<reference anchor="RFC1945">
4500  <front>
4501    <title abbrev="HTTP/1.0">Hypertext Transfer Protocol -- HTTP/1.0</title>
4502    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
4503      <organization>MIT, Laboratory for Computer Science</organization>
4504      <address><email></email></address>
4505    </author>
4506    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
4507      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
4508      <address><email></email></address>
4509    </author>
4510    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4511      <organization>W3 Consortium, MIT Laboratory for Computer Science</organization>
4512      <address><email></email></address>
4513    </author>
4514    <date month="May" year="1996"/>
4515  </front>
4516  <seriesInfo name="RFC" value="1945"/>
4519<reference anchor="RFC2045">
4520  <front>
4521    <title abbrev="Internet Message Bodies">Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies</title>
4522    <author initials="N." surname="Freed" fullname="Ned Freed">
4523      <organization>Innosoft International, Inc.</organization>
4524      <address><email></email></address>
4525    </author>
4526    <author initials="N.S." surname="Borenstein" fullname="Nathaniel S. Borenstein">
4527      <organization>First Virtual Holdings</organization>
4528      <address><email></email></address>
4529    </author>
4530    <date month="November" year="1996"/>
4531  </front>
4532  <seriesInfo name="RFC" value="2045"/>
4535<reference anchor="RFC2047">
4536  <front>
4537    <title abbrev="Message Header Extensions">MIME (Multipurpose Internet Mail Extensions) Part Three: Message Header Extensions for Non-ASCII Text</title>
4538    <author initials="K." surname="Moore" fullname="Keith Moore">
4539      <organization>University of Tennessee</organization>
4540      <address><email></email></address>
4541    </author>
4542    <date month="November" year="1996"/>
4543  </front>
4544  <seriesInfo name="RFC" value="2047"/>
4547<reference anchor="RFC2068">
4548  <front>
4549    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
4550    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
4551      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
4552      <address><email></email></address>
4553    </author>
4554    <author initials="J." surname="Gettys" fullname="Jim Gettys">
4555      <organization>MIT Laboratory for Computer Science</organization>
4556      <address><email></email></address>
4557    </author>
4558    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
4559      <organization>Digital Equipment Corporation, Western Research Laboratory</organization>
4560      <address><email></email></address>
4561    </author>
4562    <author initials="H." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4563      <organization>MIT Laboratory for Computer Science</organization>
4564      <address><email></email></address>
4565    </author>
4566    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
4567      <organization>MIT Laboratory for Computer Science</organization>
4568      <address><email></email></address>
4569    </author>
4570    <date month="January" year="1997"/>
4571  </front>
4572  <seriesInfo name="RFC" value="2068"/>
4575<reference anchor="RFC2145">
4576  <front>
4577    <title abbrev="HTTP Version Numbers">Use and Interpretation of HTTP Version Numbers</title>
4578    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
4579      <organization>Western Research Laboratory</organization>
4580      <address><email></email></address>
4581    </author>
4582    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
4583      <organization>Department of Information and Computer Science</organization>
4584      <address><email></email></address>
4585    </author>
4586    <author initials="J." surname="Gettys" fullname="Jim Gettys">
4587      <organization>MIT Laboratory for Computer Science</organization>
4588      <address><email></email></address>
4589    </author>
4590    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4591      <organization>W3 Consortium</organization>
4592      <address><email></email></address>
4593    </author>
4594    <date month="May" year="1997"/>
4595  </front>
4596  <seriesInfo name="RFC" value="2145"/>
4599<reference anchor="RFC2616">
4600  <front>
4601    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
4602    <author initials="R." surname="Fielding" fullname="R. Fielding">
4603      <organization>University of California, Irvine</organization>
4604      <address><email></email></address>
4605    </author>
4606    <author initials="J." surname="Gettys" fullname="J. Gettys">
4607      <organization>W3C</organization>
4608      <address><email></email></address>
4609    </author>
4610    <author initials="J." surname="Mogul" fullname="J. Mogul">
4611      <organization>Compaq Computer Corporation</organization>
4612      <address><email></email></address>
4613    </author>
4614    <author initials="H." surname="Frystyk" fullname="H. Frystyk">
4615      <organization>MIT Laboratory for Computer Science</organization>
4616      <address><email></email></address>
4617    </author>
4618    <author initials="L." surname="Masinter" fullname="L. Masinter">
4619      <organization>Xerox Corporation</organization>
4620      <address><email></email></address>
4621    </author>
4622    <author initials="P." surname="Leach" fullname="P. Leach">
4623      <organization>Microsoft Corporation</organization>
4624      <address><email></email></address>
4625    </author>
4626    <author initials="T." surname="Berners-Lee" fullname="T. Berners-Lee">
4627      <organization>W3C</organization>
4628      <address><email></email></address>
4629    </author>
4630    <date month="June" year="1999"/>
4631  </front>
4632  <seriesInfo name="RFC" value="2616"/>
4635<reference anchor='RFC2817'>
4636  <front>
4637    <title>Upgrading to TLS Within HTTP/1.1</title>
4638    <author initials='R.' surname='Khare' fullname='R. Khare'>
4639      <organization>4K Associates / UC Irvine</organization>
4640      <address><email></email></address>
4641    </author>
4642    <author initials='S.' surname='Lawrence' fullname='S. Lawrence'>
4643      <organization>Agranat Systems, Inc.</organization>
4644      <address><email></email></address>
4645    </author>
4646    <date year='2000' month='May' />
4647  </front>
4648  <seriesInfo name='RFC' value='2817' />
4651<reference anchor='RFC2818'>
4652  <front>
4653    <title>HTTP Over TLS</title>
4654    <author initials='E.' surname='Rescorla' fullname='Eric Rescorla'>
4655      <organization>RTFM, Inc.</organization>
4656      <address><email></email></address>
4657    </author>
4658    <date year='2000' month='May' />
4659  </front>
4660  <seriesInfo name='RFC' value='2818' />
4663<reference anchor='RFC2965'>
4664  <front>
4665    <title>HTTP State Management Mechanism</title>
4666    <author initials='D. M.' surname='Kristol' fullname='David M. Kristol'>
4667      <organization>Bell Laboratories, Lucent Technologies</organization>
4668      <address><email></email></address>
4669    </author>
4670    <author initials='L.' surname='Montulli' fullname='Lou Montulli'>
4671      <organization>, Inc.</organization>
4672      <address><email></email></address>
4673    </author>
4674    <date year='2000' month='October' />
4675  </front>
4676  <seriesInfo name='RFC' value='2965' />
4679<reference anchor='RFC3040'>
4680  <front>
4681    <title>Internet Web Replication and Caching Taxonomy</title>
4682    <author initials='I.' surname='Cooper' fullname='I. Cooper'>
4683      <organization>Equinix, Inc.</organization>
4684    </author>
4685    <author initials='I.' surname='Melve' fullname='I. Melve'>
4686      <organization>UNINETT</organization>
4687    </author>
4688    <author initials='G.' surname='Tomlinson' fullname='G. Tomlinson'>
4689      <organization>CacheFlow Inc.</organization>
4690    </author>
4691    <date year='2001' month='January' />
4692  </front>
4693  <seriesInfo name='RFC' value='3040' />
4696<reference anchor='RFC3864'>
4697  <front>
4698    <title>Registration Procedures for Message Header Fields</title>
4699    <author initials='G.' surname='Klyne' fullname='G. Klyne'>
4700      <organization>Nine by Nine</organization>
4701      <address><email></email></address>
4702    </author>
4703    <author initials='M.' surname='Nottingham' fullname='M. Nottingham'>
4704      <organization>BEA Systems</organization>
4705      <address><email></email></address>
4706    </author>
4707    <author initials='J.' surname='Mogul' fullname='J. Mogul'>
4708      <organization>HP Labs</organization>
4709      <address><email></email></address>
4710    </author>
4711    <date year='2004' month='September' />
4712  </front>
4713  <seriesInfo name='BCP' value='90' />
4714  <seriesInfo name='RFC' value='3864' />
4717<reference anchor='RFC4033'>
4718  <front>
4719    <title>DNS Security Introduction and Requirements</title>
4720    <author initials='R.' surname='Arends' fullname='R. Arends'/>
4721    <author initials='R.' surname='Austein' fullname='R. Austein'/>
4722    <author initials='M.' surname='Larson' fullname='M. Larson'/>
4723    <author initials='D.' surname='Massey' fullname='D. Massey'/>
4724    <author initials='S.' surname='Rose' fullname='S. Rose'/>
4725    <date year='2005' month='March' />
4726  </front>
4727  <seriesInfo name='RFC' value='4033' />
4730<reference anchor="RFC4288">
4731  <front>
4732    <title>Media Type Specifications and Registration Procedures</title>
4733    <author initials="N." surname="Freed" fullname="N. Freed">
4734      <organization>Sun Microsystems</organization>
4735      <address>
4736        <email></email>
4737      </address>
4738    </author>
4739    <author initials="J." surname="Klensin" fullname="J. Klensin">
4740      <address>
4741        <email></email>
4742      </address>
4743    </author>
4744    <date year="2005" month="December"/>
4745  </front>
4746  <seriesInfo name="BCP" value="13"/>
4747  <seriesInfo name="RFC" value="4288"/>
4750<reference anchor='RFC4395'>
4751  <front>
4752    <title>Guidelines and Registration Procedures for New URI Schemes</title>
4753    <author initials='T.' surname='Hansen' fullname='T. Hansen'>
4754      <organization>AT&amp;T Laboratories</organization>
4755      <address>
4756        <email></email>
4757      </address>
4758    </author>
4759    <author initials='T.' surname='Hardie' fullname='T. Hardie'>
4760      <organization>Qualcomm, Inc.</organization>
4761      <address>
4762        <email></email>
4763      </address>
4764    </author>
4765    <author initials='L.' surname='Masinter' fullname='L. Masinter'>
4766      <organization>Adobe Systems</organization>
4767      <address>
4768        <email></email>
4769      </address>
4770    </author>
4771    <date year='2006' month='February' />
4772  </front>
4773  <seriesInfo name='BCP' value='115' />
4774  <seriesInfo name='RFC' value='4395' />
4777<reference anchor='RFC4559'>
4778  <front>
4779    <title>SPNEGO-based Kerberos and NTLM HTTP Authentication in Microsoft Windows</title>
4780    <author initials='K.' surname='Jaganathan' fullname='K. Jaganathan'/>
4781    <author initials='L.' surname='Zhu' fullname='L. Zhu'/>
4782    <author initials='J.' surname='Brezak' fullname='J. Brezak'/>
4783    <date year='2006' month='June' />
4784  </front>
4785  <seriesInfo name='RFC' value='4559' />
4788<reference anchor='RFC5226'>
4789  <front>
4790    <title>Guidelines for Writing an IANA Considerations Section in RFCs</title>
4791    <author initials='T.' surname='Narten' fullname='T. Narten'>
4792      <organization>IBM</organization>
4793      <address><email></email></address>
4794    </author>
4795    <author initials='H.' surname='Alvestrand' fullname='H. Alvestrand'>
4796      <organization>Google</organization>
4797      <address><email></email></address>
4798    </author>
4799    <date year='2008' month='May' />
4800  </front>
4801  <seriesInfo name='BCP' value='26' />
4802  <seriesInfo name='RFC' value='5226' />
4805<reference anchor="RFC5322">
4806  <front>
4807    <title>Internet Message Format</title>
4808    <author initials="P." surname="Resnick" fullname="P. Resnick">
4809      <organization>Qualcomm Incorporated</organization>
4810    </author>
4811    <date year="2008" month="October"/>
4812  </front>
4813  <seriesInfo name="RFC" value="5322"/>
4816<reference anchor="RFC6265">
4817  <front>
4818    <title>HTTP State Management Mechanism</title>
4819    <author initials="A." surname="Barth" fullname="Adam Barth">
4820      <organization abbrev="U.C. Berkeley">
4821        University of California, Berkeley
4822      </organization>
4823      <address><email></email></address>
4824    </author>
4825    <date year="2011" month="April" />
4826  </front>
4827  <seriesInfo name="RFC" value="6265"/>
4830<!--<reference anchor='BCP97'>
4831  <front>
4832    <title>Handling Normative References to Standards-Track Documents</title>
4833    <author initials='J.' surname='Klensin' fullname='J. Klensin'>
4834      <address>
4835        <email></email>
4836      </address>
4837    </author>
4838    <author initials='S.' surname='Hartman' fullname='S. Hartman'>
4839      <organization>MIT</organization>
4840      <address>
4841        <email></email>
4842      </address>
4843    </author>
4844    <date year='2007' month='June' />
4845  </front>
4846  <seriesInfo name='BCP' value='97' />
4847  <seriesInfo name='RFC' value='4897' />
4850<reference anchor="Kri2001" target="">
4851  <front>
4852    <title>HTTP Cookies: Standards, Privacy, and Politics</title>
4853    <author initials="D." surname="Kristol" fullname="David M. Kristol"/>
4854    <date year="2001" month="November"/>
4855  </front>
4856  <seriesInfo name="ACM Transactions on Internet Technology" value="Vol. 1, #2"/>
4859<reference anchor="Spe" target="">
4860  <front>
4861    <title>Analysis of HTTP Performance Problems</title>
4862    <author initials="S." surname="Spero" fullname="Simon E. Spero"/>
4863    <date/>
4864  </front>
4867<reference anchor="Tou1998" target="">
4868  <front>
4869  <title>Analysis of HTTP Performance</title>
4870  <author initials="J." surname="Touch" fullname="Joe Touch">
4871    <organization>USC/Information Sciences Institute</organization>
4872    <address><email></email></address>
4873  </author>
4874  <author initials="J." surname="Heidemann" fullname="John Heidemann">
4875    <organization>USC/Information Sciences Institute</organization>
4876    <address><email></email></address>
4877  </author>
4878  <author initials="K." surname="Obraczka" fullname="Katia Obraczka">
4879    <organization>USC/Information Sciences Institute</organization>
4880    <address><email></email></address>
4881  </author>
4882  <date year="1998" month="Aug"/>
4883  </front>
4884  <seriesInfo name="ISI Research Report" value="ISI/RR-98-463"/>
4885  <annotation>(original report dated Aug. 1996)</annotation>
4891<section title="HTTP Version History" anchor="compatibility">
4893   HTTP has been in use by the World-Wide Web global information initiative
4894   since 1990. The first version of HTTP, later referred to as HTTP/0.9,
4895   was a simple protocol for hypertext data transfer across the Internet
4896   with only a single request method (GET) and no metadata.
4897   HTTP/1.0, as defined by <xref target="RFC1945"/>, added a range of request
4898   methods and MIME-like messaging that could include metadata about the data
4899   transferred and modifiers on the request/response semantics. However,
4900   HTTP/1.0 did not sufficiently take into consideration the effects of
4901   hierarchical proxies, caching, the need for persistent connections, or
4902   name-based virtual hosts. The proliferation of incompletely-implemented
4903   applications calling themselves "HTTP/1.0" further necessitated a
4904   protocol version change in order for two communicating applications
4905   to determine each other's true capabilities.
4908   HTTP/1.1 remains compatible with HTTP/1.0 by including more stringent
4909   requirements that enable reliable implementations, adding only
4910   those new features that will either be safely ignored by an HTTP/1.0
4911   recipient or only sent when communicating with a party advertising
4912   conformance with HTTP/1.1.
4915   It is beyond the scope of a protocol specification to mandate
4916   conformance with previous versions. HTTP/1.1 was deliberately
4917   designed, however, to make supporting previous versions easy.
4918   We would expect a general-purpose HTTP/1.1 server to understand
4919   any valid request in the format of HTTP/1.0 and respond appropriately
4920   with an HTTP/1.1 message that only uses features understood (or
4921   safely ignored) by HTTP/1.0 clients.  Likewise, we would expect
4922   an HTTP/1.1 client to understand any valid HTTP/1.0 response.
4925   Since HTTP/0.9 did not support header fields in a request,
4926   there is no mechanism for it to support name-based virtual
4927   hosts (selection of resource by inspection of the <x:ref>Host</x:ref> header
4928   field).  Any server that implements name-based virtual hosts
4929   ought to disable support for HTTP/0.9.  Most requests that
4930   appear to be HTTP/0.9 are, in fact, badly constructed HTTP/1.x
4931   requests wherein a buggy client failed to properly encode
4932   linear whitespace found in a URI reference and placed in
4933   the request-target.
4936<section title="Changes from HTTP/1.0" anchor="changes.from.1.0">
4938   This section summarizes major differences between versions HTTP/1.0
4939   and HTTP/1.1.
4942<section title="Multi-homed Web Servers" anchor="">
4944   The requirements that clients and servers support the <x:ref>Host</x:ref>
4945   header field (<xref target=""/>), report an error if it is
4946   missing from an HTTP/1.1 request, and accept absolute URIs (<xref target="request-target"/>)
4947   are among the most important changes defined by HTTP/1.1.
4950   Older HTTP/1.0 clients assumed a one-to-one relationship of IP
4951   addresses and servers; there was no other established mechanism for
4952   distinguishing the intended server of a request than the IP address
4953   to which that request was directed. The <x:ref>Host</x:ref> header field was
4954   introduced during the development of HTTP/1.1 and, though it was
4955   quickly implemented by most HTTP/1.0 browsers, additional requirements
4956   were placed on all HTTP/1.1 requests in order to ensure complete
4957   adoption.  At the time of this writing, most HTTP-based services
4958   are dependent upon the Host header field for targeting requests.
4962<section title="Keep-Alive Connections" anchor="compatibility.with.http.1.0.persistent.connections">
4964   In HTTP/1.0, each connection is established by the client prior to the
4965   request and closed by the server after sending the response. However, some
4966   implementations implement the explicitly negotiated ("Keep-Alive") version
4967   of persistent connections described in <xref x:sec="19.7.1" x:fmt="of"
4968   target="RFC2068"/>.
4971   Some clients and servers might wish to be compatible with these previous
4972   approaches to persistent connections, by explicitly negotiating for them
4973   with a "Connection: keep-alive" request header field. However, some
4974   experimental implementations of HTTP/1.0 persistent connections are faulty;
4975   for example, if a HTTP/1.0 proxy server doesn't understand
4976   <x:ref>Connection</x:ref>, it will erroneously forward that header field
4977   to the next inbound server, which would result in a hung connection.
4980   One attempted solution was the introduction of a Proxy-Connection header
4981   field, targeted specifically at proxies. In practice, this was also
4982   unworkable, because proxies are often deployed in multiple layers, bringing
4983   about the same problem discussed above.
4986   As a result, clients are encouraged not to send the Proxy-Connection header
4987   field in any requests.
4990   Clients are also encouraged to consider the use of Connection: keep-alive
4991   in requests carefully; while they can enable persistent connections with
4992   HTTP/1.0 servers, clients using them need will need to monitor the
4993   connection for "hung" requests (which indicate that the client ought stop
4994   sending the header field), and this mechanism ought not be used by clients
4995   at all when a proxy is being used.
4999<section title="Introduction of Transfer-Encoding" anchor="introduction.of.transfer-encoding">
5001   HTTP/1.1 introduces the <x:ref>Transfer-Encoding</x:ref> header field
5002   (<xref target="header.transfer-encoding"/>). Proxies/gateways &MUST; remove
5003   any transfer-coding prior to forwarding a message via a MIME-compliant
5004   protocol.
5010<section title="Changes from RFC 2616" anchor="changes.from.rfc.2616">
5012  Clarify that the string "HTTP" in the HTTP-version ABFN production is case
5013  sensitive. Restrict the version numbers to be single digits due to the fact
5014  that implementations are known to handle multi-digit version numbers
5015  incorrectly.
5016  (<xref target="http.version"/>)
5019  Update use of abs_path production from RFC 1808 to the path-absolute + query
5020  components of RFC 3986. State that the asterisk form is allowed for the OPTIONS
5021  request method only.
5022  (<xref target="request-target"/>)
5025  Require that invalid whitespace around field-names be rejected.
5026  (<xref target="header.fields"/>)
5029  Rules about implicit linear whitespace between certain grammar productions
5030  have been removed; now whitespace is only allowed where specifically
5031  defined in the ABNF.
5032  (<xref target="whitespace"/>)
5035  The NUL octet is no longer allowed in comment and quoted-string
5036  text. The quoted-pair rule no longer allows escaping control characters other than HTAB.
5037  Non-ASCII content in header fields and reason phrase has been obsoleted and
5038  made opaque (the TEXT rule was removed).
5039  (<xref target="field.components"/>)
5042  Empty list elements in list productions have been deprecated.
5043  (<xref target="abnf.extension"/>)
5046  Require recipients to handle bogus <x:ref>Content-Length</x:ref> header
5047  fields as errors.
5048  (<xref target="message.body"/>)
5051  Remove reference to non-existent identity transfer-coding value tokens.
5052  (Sections <xref format="counter" target="message.body"/> and
5053  <xref format="counter" target="transfer.codings"/>)
5056  Clarification that the chunk length does not include the count of the octets
5057  in the chunk header and trailer. Furthermore disallowed line folding
5058  in chunk extensions, and deprecate their use.
5059  (<xref target="chunked.encoding"/>)
5062  Registration of Transfer Codings now requires IETF Review
5063  (<xref target="transfer.coding.registry"/>)
5066  Remove hard limit of two connections per server.
5067  Remove requirement to retry a sequence of requests as long it was idempotent.
5068  Remove requirements about when servers are allowed to close connections
5069  prematurely.
5070  (<xref target="persistent.practical"/>)
5073  Remove requirement to retry requests under certain cirumstances when the
5074  server prematurely closes the connection.
5075  (<xref target="message.transmission.requirements"/>)
5078  Change ABNF productions for header fields to only define the field value.
5081  Clarify exactly when "close" connection options have to be sent.
5082  (<xref target="header.connection"/>)
5085  Define the semantics of the <x:ref>Upgrade</x:ref> header field in responses
5086  other than 101 (this was incorporated from <xref target="RFC2817"/>).
5087  (<xref target="header.upgrade"/>)
5090  Take over the Upgrade Token Registry, previously defined in
5091  <xref target="RFC2817" x:fmt="of" x:sec="7.2"/>.
5092  (<xref target="upgrade.token.registry"/>)
5097<?BEGININC p1-messaging.abnf-appendix ?>
5098<section xmlns:x="" title="Collected ABNF" anchor="collected.abnf">
5100<artwork type="abnf" name="p1-messaging.parsed-abnf">
5101<x:ref>BWS</x:ref> = OWS
5103<x:ref>Connection</x:ref> = *( "," OWS ) connection-option *( OWS "," [ OWS
5104 connection-option ] )
5105<x:ref>Content-Length</x:ref> = 1*DIGIT
5107<x:ref>HTTP-message</x:ref> = start-line *( header-field CRLF ) CRLF [ message-body
5108 ]
5109<x:ref>HTTP-name</x:ref> = %x48.54.54.50 ; HTTP
5110<x:ref>HTTP-version</x:ref> = HTTP-name "/" DIGIT "." DIGIT
5111<x:ref>Host</x:ref> = uri-host [ ":" port ]
5113<x:ref>OWS</x:ref> = *( SP / HTAB )
5115<x:ref>RWS</x:ref> = 1*( SP / HTAB )
5117<x:ref>TE</x:ref> = [ ( "," / t-codings ) *( OWS "," [ OWS t-codings ] ) ]
5118<x:ref>Trailer</x:ref> = *( "," OWS ) field-name *( OWS "," [ OWS field-name ] )
5119<x:ref>Transfer-Encoding</x:ref> = *( "," OWS ) transfer-coding *( OWS "," [ OWS
5120 transfer-coding ] )
5122<x:ref>URI-reference</x:ref> = &lt;URI-reference, defined in [RFC3986], Section 4.1&gt;
5123<x:ref>Upgrade</x:ref> = *( "," OWS ) protocol *( OWS "," [ OWS protocol ] )
5125<x:ref>Via</x:ref> = *( "," OWS ) ( received-protocol RWS received-by [ RWS comment
5126 ] ) *( OWS "," [ OWS ( received-protocol RWS received-by [ RWS
5127 comment ] ) ] )
5129<x:ref>absolute-URI</x:ref> = &lt;absolute-URI, defined in [RFC3986], Section 4.3&gt;
5130<x:ref>absolute-form</x:ref> = absolute-URI
5131<x:ref>asterisk-form</x:ref> = "*"
5132<x:ref>attribute</x:ref> = token
5133<x:ref>authority</x:ref> = &lt;authority, defined in [RFC3986], Section 3.2&gt;
5134<x:ref>authority-form</x:ref> = authority
5136<x:ref>chunk</x:ref> = chunk-size [ chunk-ext ] CRLF chunk-data CRLF
5137<x:ref>chunk-data</x:ref> = 1*OCTET
5138<x:ref>chunk-ext</x:ref> = *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
5139<x:ref>chunk-ext-name</x:ref> = token
5140<x:ref>chunk-ext-val</x:ref> = token / quoted-str-nf
5141<x:ref>chunk-size</x:ref> = 1*HEXDIG
5142<x:ref>chunked-body</x:ref> = *chunk last-chunk trailer-part CRLF
5143<x:ref>comment</x:ref> = "(" *( ctext / quoted-cpair / comment ) ")"
5144<x:ref>connection-option</x:ref> = token
5145<x:ref>ctext</x:ref> = OWS / %x21-27 ; '!'-'''
5146 / %x2A-5B ; '*'-'['
5147 / %x5D-7E ; ']'-'~'
5148 / obs-text
5150<x:ref>field-content</x:ref> = *( HTAB / SP / VCHAR / obs-text )
5151<x:ref>field-name</x:ref> = token
5152<x:ref>field-value</x:ref> = *( field-content / obs-fold )
5154<x:ref>header-field</x:ref> = field-name ":" OWS field-value BWS
5155<x:ref>http-URI</x:ref> = "http://" authority path-abempty [ "?" query ]
5156<x:ref>https-URI</x:ref> = "https://" authority path-abempty [ "?" query ]
5158<x:ref>last-chunk</x:ref> = 1*"0" [ chunk-ext ] CRLF
5160<x:ref>message-body</x:ref> = *OCTET
5161<x:ref>method</x:ref> = token
5163<x:ref>obs-fold</x:ref> = CRLF ( SP / HTAB )
5164<x:ref>obs-text</x:ref> = %x80-FF
5165<x:ref>origin-form</x:ref> = path-absolute [ "?" query ]
5167<x:ref>partial-URI</x:ref> = relative-part [ "?" query ]
5168<x:ref>path-abempty</x:ref> = &lt;path-abempty, defined in [RFC3986], Section 3.3&gt;
5169<x:ref>path-absolute</x:ref> = &lt;path-absolute, defined in [RFC3986], Section 3.3&gt;
5170<x:ref>port</x:ref> = &lt;port, defined in [RFC3986], Section 3.2.3&gt;
5171<x:ref>protocol</x:ref> = protocol-name [ "/" protocol-version ]
5172<x:ref>protocol-name</x:ref> = token
5173<x:ref>protocol-version</x:ref> = token
5174<x:ref>pseudonym</x:ref> = token
5176<x:ref>qdtext</x:ref> = OWS / "!" / %x23-5B ; '#'-'['
5177 / %x5D-7E ; ']'-'~'
5178 / obs-text
5179<x:ref>qdtext-nf</x:ref> = HTAB / SP / "!" / %x23-5B ; '#'-'['
5180 / %x5D-7E ; ']'-'~'
5181 / obs-text
5182<x:ref>query</x:ref> = &lt;query, defined in [RFC3986], Section 3.4&gt;
5183<x:ref>quoted-cpair</x:ref> = "\" ( HTAB / SP / VCHAR / obs-text )
5184<x:ref>quoted-pair</x:ref> = "\" ( HTAB / SP / VCHAR / obs-text )
5185<x:ref>quoted-str-nf</x:ref> = DQUOTE *( qdtext-nf / quoted-pair ) DQUOTE
5186<x:ref>quoted-string</x:ref> = DQUOTE *( qdtext / quoted-pair ) DQUOTE
5187<x:ref>qvalue</x:ref> = ( "0" [ "." *3DIGIT ] ) / ( "1" [ "." *3"0" ] )
5189<x:ref>reason-phrase</x:ref> = *( HTAB / SP / VCHAR / obs-text )
5190<x:ref>received-by</x:ref> = ( uri-host [ ":" port ] ) / pseudonym
5191<x:ref>received-protocol</x:ref> = [ protocol-name "/" ] protocol-version
5192<x:ref>relative-part</x:ref> = &lt;relative-part, defined in [RFC3986], Section 4.2&gt;
5193<x:ref>request-line</x:ref> = method SP request-target SP HTTP-version CRLF
5194<x:ref>request-target</x:ref> = origin-form / absolute-form / authority-form /
5195 asterisk-form
5197<x:ref>special</x:ref> = "(" / ")" / "&lt;" / "&gt;" / "@" / "," / ";" / ":" / "\" /
5198 DQUOTE / "/" / "[" / "]" / "?" / "=" / "{" / "}"
5199<x:ref>start-line</x:ref> = request-line / status-line
5200<x:ref>status-code</x:ref> = 3DIGIT
5201<x:ref>status-line</x:ref> = HTTP-version SP status-code SP reason-phrase CRLF
5203<x:ref>t-codings</x:ref> = "trailers" / ( transfer-extension [ te-params ] )
5204<x:ref>tchar</x:ref> = "!" / "#" / "$" / "%" / "&amp;" / "'" / "*" / "+" / "-" / "." /
5205 "^" / "_" / "`" / "|" / "~" / DIGIT / ALPHA
5206<x:ref>te-ext</x:ref> = OWS ";" OWS token [ "=" word ]
5207<x:ref>te-params</x:ref> = OWS ";" OWS "q=" qvalue *te-ext
5208<x:ref>token</x:ref> = 1*tchar
5209<x:ref>trailer-part</x:ref> = *( header-field CRLF )
5210<x:ref>transfer-coding</x:ref> = "chunked" / "compress" / "deflate" / "gzip" /
5211 transfer-extension
5212<x:ref>transfer-extension</x:ref> = token *( OWS ";" OWS transfer-parameter )
5213<x:ref>transfer-parameter</x:ref> = attribute BWS "=" BWS value
5215<x:ref>uri-host</x:ref> = &lt;host, defined in [RFC3986], Section 3.2.2&gt;
5217<x:ref>value</x:ref> = word
5219<x:ref>word</x:ref> = token / quoted-string
5222<figure><preamble>ABNF diagnostics:</preamble><artwork type="inline">
5223; HTTP-message defined but not used
5224; URI-reference defined but not used
5225; chunked-body defined but not used
5226; http-URI defined but not used
5227; https-URI defined but not used
5228; partial-URI defined but not used
5229; special defined but not used
5231<?ENDINC p1-messaging.abnf-appendix ?>
5233<section title="Change Log (to be removed by RFC Editor before publication)" anchor="change.log">
5235<section title="Since RFC 2616">
5237  Extracted relevant partitions from <xref target="RFC2616"/>.
5241<section title="Since draft-ietf-httpbis-p1-messaging-00">
5243  Closed issues:
5244  <list style="symbols">
5245    <t>
5246      <eref target=""/>:
5247      "HTTP Version should be case sensitive"
5248      (<eref target=""/>)
5249    </t>
5250    <t>
5251      <eref target=""/>:
5252      "'unsafe' characters"
5253      (<eref target=""/>)
5254    </t>
5255    <t>
5256      <eref target=""/>:
5257      "Chunk Size Definition"
5258      (<eref target=""/>)
5259    </t>
5260    <t>
5261      <eref target=""/>:
5262      "Message Length"
5263      (<eref target=""/>)
5264    </t>
5265    <t>
5266      <eref target=""/>:
5267      "Media Type Registrations"
5268      (<eref target=""/>)
5269    </t>
5270    <t>
5271      <eref target=""/>:
5272      "URI includes query"
5273      (<eref target=""/>)
5274    </t>
5275    <t>
5276      <eref target=""/>:
5277      "No close on 1xx responses"
5278      (<eref target=""/>)
5279    </t>
5280    <t>
5281      <eref target=""/>:
5282      "Remove 'identity' token references"
5283      (<eref target=""/>)
5284    </t>
5285    <t>
5286      <eref target=""/>:
5287      "Import query BNF"
5288    </t>
5289    <t>
5290      <eref target=""/>:
5291      "qdtext BNF"
5292    </t>
5293    <t>
5294      <eref target=""/>:
5295      "Normative and Informative references"
5296    </t>
5297    <t>
5298      <eref target=""/>:
5299      "RFC2606 Compliance"
5300    </t>
5301    <t>
5302      <eref target=""/>:
5303      "RFC977 reference"
5304    </t>
5305    <t>
5306      <eref target=""/>:
5307      "RFC1700 references"
5308    </t>
5309    <t>
5310      <eref target=""/>:
5311      "inconsistency in date format explanation"
5312    </t>
5313    <t>
5314      <eref target=""/>:
5315      "Date reference typo"
5316    </t>
5317    <t>
5318      <eref target=""/>:
5319      "Informative references"
5320    </t>
5321    <t>
5322      <eref target=""/>:
5323      "ISO-8859-1 Reference"
5324    </t>
5325    <t>
5326      <eref target=""/>:
5327      "Normative up-to-date references"
5328    </t>
5329  </list>
5332  Other changes:
5333  <list style="symbols">
5334    <t>
5335      Update media type registrations to use RFC4288 template.
5336    </t>
5337    <t>
5338      Use names of RFC4234 core rules DQUOTE and HTAB,
5339      fix broken ABNF for chunk-data
5340      (work in progress on <eref target=""/>)
5341    </t>
5342  </list>
5346<section title="Since draft-ietf-httpbis-p1-messaging-01">
5348  Closed issues:
5349  <list style="symbols">
5350    <t>
5351      <eref target=""/>:
5352      "Bodies on GET (and other) requests"
5353    </t>
5354    <t>
5355      <eref target=""/>:
5356      "Updating to RFC4288"
5357    </t>
5358    <t>
5359      <eref target=""/>:
5360      "Status Code and Reason Phrase"
5361    </t>
5362    <t>
5363      <eref target=""/>:
5364      "rel_path not used"
5365    </t>
5366  </list>
5369  Ongoing work on ABNF conversion (<eref target=""/>):
5370  <list style="symbols">
5371    <t>
5372      Get rid of duplicate BNF rule names ("host" -> "uri-host", "trailer" ->
5373      "trailer-part").
5374    </t>
5375    <t>
5376      Avoid underscore character in rule names ("http_URL" ->
5377      "http-URL", "abs_path" -> "path-absolute").
5378    </t>
5379    <t>
5380      Add rules for terms imported from URI spec ("absoluteURI", "authority",
5381      "path-absolute", "port", "query", "relativeURI", "host) &mdash; these will
5382      have to be updated when switching over to RFC3986.
5383    </t>
5384    <t>
5385      Synchronize core rules with RFC5234.
5386    </t>
5387    <t>
5388      Get rid of prose rules that span multiple lines.
5389    </t>
5390    <t>
5391      Get rid of unused rules LOALPHA and UPALPHA.
5392    </t>
5393    <t>
5394      Move "Product Tokens" section (back) into Part 1, as "token" is used
5395      in the definition of the Upgrade header field.
5396    </t>
5397    <t>
5398      Add explicit references to BNF syntax and rules imported from other parts of the specification.
5399    </t>
5400    <t>
5401      Rewrite prose rule "token" in terms of "tchar", rewrite prose rule "TEXT".
5402    </t>
5403  </list>
5407<section title="Since draft-ietf-httpbis-p1-messaging-02" anchor="changes.since.02">
5409  Closed issues:
5410  <list style="symbols">
5411    <t>
5412      <eref target=""/>:
5413      "HTTP-date vs. rfc1123-date"
5414    </t>
5415    <t>
5416      <eref target=""/>:
5417      "WS in quoted-pair"
5418    </t>
5419  </list>
5422  Ongoing work on IANA Message Header Field Registration (<eref target=""/>):
5423  <list style="symbols">
5424    <t>
5425      Reference RFC 3984, and update header field registrations for header
5426      fields defined in this document.
5427    </t>
5428  </list>
5431  Ongoing work on ABNF conversion (<eref target=""/>):
5432  <list style="symbols">
5433    <t>
5434      Replace string literals when the string really is case-sensitive (HTTP-version).
5435    </t>
5436  </list>
5440<section title="Since draft-ietf-httpbis-p1-messaging-03" anchor="changes.since.03">
5442  Closed issues:
5443  <list style="symbols">
5444    <t>
5445      <eref target=""/>:
5446      "Connection closing"
5447    </t>
5448    <t>
5449      <eref target=""/>:
5450      "Move registrations and registry information to IANA Considerations"
5451    </t>
5452    <t>
5453      <eref target=""/>:
5454      "need new URL for PAD1995 reference"
5455    </t>
5456    <t>
5457      <eref target=""/>:
5458      "IANA Considerations: update HTTP URI scheme registration"
5459    </t>
5460    <t>
5461      <eref target=""/>:
5462      "Cite HTTPS URI scheme definition"
5463    </t>
5464    <t>
5465      <eref target=""/>:
5466      "List-type header fields vs Set-Cookie"
5467    </t>
5468  </list>
5471  Ongoing work on ABNF conversion (<eref target=""/>):
5472  <list style="symbols">
5473    <t>
5474      Replace string literals when the string really is case-sensitive (HTTP-Date).
5475    </t>
5476    <t>
5477      Replace HEX by HEXDIG for future consistence with RFC 5234's core rules.
5478    </t>
5479  </list>
5483<section title="Since draft-ietf-httpbis-p1-messaging-04" anchor="changes.since.04">
5485  Closed issues:
5486  <list style="symbols">
5487    <t>
5488      <eref target=""/>:
5489      "Out-of-date reference for URIs"
5490    </t>
5491    <t>
5492      <eref target=""/>:
5493      "RFC 2822 is updated by RFC 5322"
5494    </t>
5495  </list>
5498  Ongoing work on ABNF conversion (<eref target=""/>):
5499  <list style="symbols">
5500    <t>
5501      Use "/" instead of "|" for alternatives.
5502    </t>
5503    <t>
5504      Get rid of RFC822 dependency; use RFC5234 plus extensions instead.
5505    </t>
5506    <t>
5507      Only reference RFC 5234's core rules.
5508    </t>
5509    <t>
5510      Introduce new ABNF rules for "bad" whitespace ("BWS"), optional
5511      whitespace ("OWS") and required whitespace ("RWS").
5512    </t>
5513    <t>
5514      Rewrite ABNFs to spell out whitespace rules, factor out
5515      header field value format definitions.
5516    </t>
5517  </list>
5521<section title="Since draft-ietf-httpbis-p1-messaging-05" anchor="changes.since.05">
5523  Closed issues:
5524  <list style="symbols">
5525    <t>
5526      <eref target=""/>:
5527      "Header LWS"
5528    </t>
5529    <t>
5530      <eref target=""/>:
5531      "Sort 1.3 Terminology"
5532    </t>
5533    <t>
5534      <eref target=""/>:
5535      "RFC2047 encoded words"
5536    </t>
5537    <t>
5538      <eref target=""/>:
5539      "Character Encodings in TEXT"
5540    </t>
5541    <t>
5542      <eref target=""/>:
5543      "Line Folding"
5544    </t>
5545    <t>
5546      <eref target=""/>:
5547      "OPTIONS * and proxies"
5548    </t>
5549    <t>
5550      <eref target=""/>:
5551      "reason-phrase BNF"
5552    </t>
5553    <t>
5554      <eref target=""/>:
5555      "Use of TEXT"
5556    </t>
5557    <t>
5558      <eref target=""/>:
5559      "Join "Differences Between HTTP Entities and RFC 2045 Entities"?"
5560    </t>
5561    <t>
5562      <eref target=""/>:
5563      "RFC822 reference left in discussion of date formats"
5564    </t>
5565  </list>
5568  Final work on ABNF conversion (<eref target=""/>):
5569  <list style="symbols">
5570    <t>
5571      Rewrite definition of list rules, deprecate empty list elements.
5572    </t>
5573    <t>
5574      Add appendix containing collected and expanded ABNF.
5575    </t>
5576  </list>
5579  Other changes:
5580  <list style="symbols">
5581    <t>
5582      Rewrite introduction; add mostly new Architecture Section.
5583    </t>
5584    <t>
5585      Move definition of quality values from Part 3 into Part 1;
5586      make TE request header field grammar independent of accept-params (defined in Part 3).
5587    </t>
5588  </list>
5592<section title="Since draft-ietf-httpbis-p1-messaging-06" anchor="changes.since.06">
5594  Closed issues:
5595  <list style="symbols">
5596    <t>
5597      <eref target=""/>:
5598      "base for numeric protocol elements"
5599    </t>
5600    <t>
5601      <eref target=""/>:
5602      "comment ABNF"
5603    </t>
5604  </list>
5607  Partly resolved issues:
5608  <list style="symbols">
5609    <t>
5610      <eref target=""/>:
5611      "205 Bodies" (took out language that implied that there might be
5612      methods for which a request body MUST NOT be included)
5613    </t>
5614    <t>
5615      <eref target=""/>:
5616      "editorial improvements around HTTP-date"
5617    </t>
5618  </list>
5622<section title="Since draft-ietf-httpbis-p1-messaging-07" anchor="changes.since.07">
5624  Closed issues:
5625  <list style="symbols">
5626    <t>
5627      <eref target=""/>:
5628      "Repeating single-value header fields"
5629    </t>
5630    <t>
5631      <eref target=""/>:
5632      "increase connection limit"
5633    </t>
5634    <t>
5635      <eref target=""/>:
5636      "IP addresses in URLs"
5637    </t>
5638    <t>
5639      <eref target=""/>:
5640      "take over HTTP Upgrade Token Registry"
5641    </t>
5642    <t>
5643      <eref target=""/>:
5644      "CR and LF in chunk extension values"
5645    </t>
5646    <t>
5647      <eref target=""/>:
5648      "HTTP/0.9 support"
5649    </t>
5650    <t>
5651      <eref target=""/>:
5652      "pick IANA policy (RFC5226) for Transfer Coding / Content Coding"
5653    </t>
5654    <t>
5655      <eref target=""/>:
5656      "move definitions of gzip/deflate/compress to part 1"
5657    </t>
5658    <t>
5659      <eref target=""/>:
5660      "disallow control characters in quoted-pair"
5661    </t>
5662  </list>
5665  Partly resolved issues:
5666  <list style="symbols">
5667    <t>
5668      <eref target=""/>:
5669      "update IANA requirements wrt Transfer-Coding values" (add the
5670      IANA Considerations subsection)
5671    </t>
5672  </list>
5676<section title="Since draft-ietf-httpbis-p1-messaging-08" anchor="changes.since.08">
5678  Closed issues:
5679  <list style="symbols">
5680    <t>
5681      <eref target=""/>:
5682      "header parsing, treatment of leading and trailing OWS"
5683    </t>
5684  </list>
5687  Partly resolved issues:
5688  <list style="symbols">
5689    <t>
5690      <eref target=""/>:
5691      "Placement of 13.5.1 and 13.5.2"
5692    </t>
5693    <t>
5694      <eref target=""/>:
5695      "use of term "word" when talking about header field structure"
5696    </t>
5697  </list>
5701<section title="Since draft-ietf-httpbis-p1-messaging-09" anchor="changes.since.09">
5703  Closed issues:
5704  <list style="symbols">
5705    <t>
5706      <eref target=""/>:
5707      "Clarification of the term 'deflate'"
5708    </t>
5709    <t>
5710      <eref target=""/>:
5711      "OPTIONS * and proxies"
5712    </t>
5713    <t>
5714      <eref target=""/>:
5715      "MIME-Version not listed in P1, general header fields"
5716    </t>
5717    <t>
5718      <eref target=""/>:
5719      "IANA registry for content/transfer encodings"
5720    </t>
5721    <t>
5722      <eref target=""/>:
5723      "Case-sensitivity of HTTP-date"
5724    </t>
5725    <t>
5726      <eref target=""/>:
5727      "use of term "word" when talking about header field structure"
5728    </t>
5729  </list>
5732  Partly resolved issues:
5733  <list style="symbols">
5734    <t>
5735      <eref target=""/>:
5736      "Term for the requested resource's URI"
5737    </t>
5738  </list>
5742<section title="Since draft-ietf-httpbis-p1-messaging-10" anchor="changes.since.10">
5744  Closed issues:
5745  <list style="symbols">
5746    <t>
5747      <eref target=""/>:
5748      "Connection Closing"
5749    </t>
5750    <t>
5751      <eref target=""/>:
5752      "Delimiting messages with multipart/byteranges"
5753    </t>
5754    <t>
5755      <eref target=""/>:
5756      "Handling multiple Content-Length header fields"
5757    </t>
5758    <t>
5759      <eref target=""/>:
5760      "Clarify entity / representation / variant terminology"
5761    </t>
5762    <t>
5763      <eref target=""/>:
5764      "consider removing the 'changes from 2068' sections"
5765    </t>
5766  </list>
5769  Partly resolved issues:
5770  <list style="symbols">
5771    <t>
5772      <eref target=""/>:
5773      "HTTP(s) URI scheme definitions"
5774    </t>
5775  </list>
5779<section title="Since draft-ietf-httpbis-p1-messaging-11" anchor="changes.since.11">
5781  Closed issues:
5782  <list style="symbols">
5783    <t>
5784      <eref target=""/>:
5785      "Trailer requirements"
5786    </t>
5787    <t>
5788      <eref target=""/>:
5789      "Text about clock requirement for caches belongs in p6"
5790    </t>
5791    <t>
5792      <eref target=""/>:
5793      "effective request URI: handling of missing host in HTTP/1.0"
5794    </t>
5795    <t>
5796      <eref target=""/>:
5797      "confusing Date requirements for clients"
5798    </t>
5799  </list>
5802  Partly resolved issues:
5803  <list style="symbols">
5804    <t>
5805      <eref target=""/>:
5806      "Handling multiple Content-Length header fields"
5807    </t>
5808  </list>
5812<section title="Since draft-ietf-httpbis-p1-messaging-12" anchor="changes.since.12">
5814  Closed issues:
5815  <list style="symbols">
5816    <t>
5817      <eref target=""/>:
5818      "RFC2145 Normative"
5819    </t>
5820    <t>
5821      <eref target=""/>:
5822      "HTTP(s) URI scheme definitions" (tune the requirements on userinfo)
5823    </t>
5824    <t>
5825      <eref target=""/>:
5826      "define 'transparent' proxy"
5827    </t>
5828    <t>
5829      <eref target=""/>:
5830      "Header Field Classification"
5831    </t>
5832    <t>
5833      <eref target=""/>:
5834      "Is * usable as a request-uri for new methods?"
5835    </t>
5836    <t>
5837      <eref target=""/>:
5838      "Migrate Upgrade details from RFC2817"
5839    </t>
5840    <t>
5841      <eref target=""/>:
5842      "untangle ABNFs for header fields"
5843    </t>
5844    <t>
5845      <eref target=""/>:
5846      "update RFC 2109 reference"
5847    </t>
5848  </list>
5852<section title="Since draft-ietf-httpbis-p1-messaging-13" anchor="changes.since.13">
5854  Closed issues:
5855  <list style="symbols">
5856    <t>
5857      <eref target=""/>:
5858      "Allow is not in 13.5.2"
5859    </t>
5860    <t>
5861      <eref target=""/>:
5862      "Handling multiple Content-Length header fields"
5863    </t>
5864    <t>
5865      <eref target=""/>:
5866      "untangle ABNFs for header fields"
5867    </t>
5868    <t>
5869      <eref target=""/>:
5870      "Content-Length ABNF broken"
5871    </t>
5872  </list>
5876<section title="Since draft-ietf-httpbis-p1-messaging-14" anchor="changes.since.14">
5878  Closed issues:
5879  <list style="symbols">
5880    <t>
5881      <eref target=""/>:
5882      "HTTP-version should be redefined as fixed length pair of DIGIT . DIGIT"
5883    </t>
5884    <t>
5885      <eref target=""/>:
5886      "Recommend minimum sizes for protocol elements"
5887    </t>
5888    <t>
5889      <eref target=""/>:
5890      "Set expectations around buffering"
5891    </t>
5892    <t>
5893      <eref target=""/>:
5894      "Considering messages in isolation"
5895    </t>
5896  </list>
5900<section title="Since draft-ietf-httpbis-p1-messaging-15" anchor="changes.since.15">
5902  Closed issues:
5903  <list style="symbols">
5904    <t>
5905      <eref target=""/>:
5906      "DNS Spoofing / DNS Binding advice"
5907    </t>
5908    <t>
5909      <eref target=""/>:
5910      "move RFCs 2145, 2616, 2817 to Historic status"
5911    </t>
5912    <t>
5913      <eref target=""/>:
5914      "\-escaping in quoted strings"
5915    </t>
5916    <t>
5917      <eref target=""/>:
5918      "'Close' should be reserved in the HTTP header field registry"
5919    </t>
5920  </list>
5924<section title="Since draft-ietf-httpbis-p1-messaging-16" anchor="changes.since.16">
5926  Closed issues:
5927  <list style="symbols">
5928    <t>
5929      <eref target=""/>:
5930      "Document HTTP's error-handling philosophy"
5931    </t>
5932    <t>
5933      <eref target=""/>:
5934      "Explain header field registration"
5935    </t>
5936    <t>
5937      <eref target=""/>:
5938      "Revise Acknowledgements Sections"
5939    </t>
5940    <t>
5941      <eref target=""/>:
5942      "Retrying Requests"
5943    </t>
5944    <t>
5945      <eref target=""/>:
5946      "Closing the connection on server error"
5947    </t>
5948  </list>
5952<section title="Since draft-ietf-httpbis-p1-messaging-17" anchor="changes.since.17">
5954  Closed issues:
5955  <list style="symbols">
5956    <t>
5957      <eref target=""/>:
5958      "Proxy-Connection and Keep-Alive"
5959    </t>
5960    <t>
5961      <eref target=""/>:
5962      "Clarify 'User Agent'"
5963    </t>
5964    <t>
5965      <eref target=""/>:
5966      "Define non-final responses"
5967    </t>
5968    <t>
5969      <eref target=""/>:
5970      "intended maturity level vs normative references"
5971    </t>
5972    <t>
5973      <eref target=""/>:
5974      "Intermediary rewriting of queries"
5975    </t>
5976  </list>
5980<section title="Since draft-ietf-httpbis-p1-messaging-18" anchor="changes.since.18">
5982  Closed issues:
5983  <list style="symbols">
5984    <t>
5985      <eref target=""/>:
5986      "message-body in CONNECT response"
5987    </t>
5988    <t>
5989      <eref target=""/>:
5990      "Misplaced text on connection handling in p2"
5991    </t>
5992    <t>
5993      <eref target=""/>:
5994      "wording of line folding rule"
5995    </t>
5996    <t>
5997      <eref target=""/>:
5998      "chunk-extensions"
5999    </t>
6000    <t>
6001      <eref target=""/>:
6002      "make IANA policy definitions consistent"
6003    </t>
6004  </list>
6008<section title="Since draft-ietf-httpbis-p1-messaging-19" anchor="changes.since.19">
6010  Closed issues:
6011  <list style="symbols">
6012    <t>
6013      <eref target=""/>:
6014      "make IANA policy definitions consistent"
6015    </t>
6016    <t>
6017      <eref target=""/>:
6018      "clarify connection header field values are case-insensitive"
6019    </t>
6020    <t>
6021      <eref target=""/>:
6022      "ABNF requirements for recipients"
6023    </t>
6024    <t>
6025      <eref target=""/>:
6026      "note introduction of new IANA registries as normative changes"
6027    </t>
6028  </list>
Note: See TracBrowser for help on using the repository browser.