source: draft-ietf-httpbis/latest/p1-messaging.xml @ 1968

Last change on this file since 1968 was 1968, checked in by julian.reschke@…, 10 years ago

insert paragraph break

  • Property svn:eol-style set to native
  • Property svn:mime-type set to text/xml
File size: 218.7 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 "November">
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' x:rel='#payload' 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 HEAD                   "<xref target='Part2' x:rel='#HEAD' xmlns:x=''/>">
29  <!ENTITY header-allow           "<xref target='Part2' x:rel='#header.allow' xmlns:x=''/>">
30  <!ENTITY header-cache-control   "<xref target='Part6' x:rel='#header.cache-control' xmlns:x=''/>">
31  <!ENTITY header-content-encoding    "<xref target='Part2' x:rel='#header.content-encoding' xmlns:x=''/>">
32  <!ENTITY header-content-location    "<xref target='Part2' x:rel='#header.content-location' xmlns:x=''/>">
33  <!ENTITY header-content-range   "<xref target='Part5' x:rel='#header.content-range' xmlns:x=''/>">
34  <!ENTITY header-content-type    "<xref target='Part2' x:rel='#header.content-type' xmlns:x=''/>">
35  <!ENTITY header-date            "<xref target='Part2' x:rel='' xmlns:x=''/>">
36  <!ENTITY header-etag            "<xref target='Part4' x:rel='#header.etag' 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 qvalue                 "<xref target='Part2' x:rel='#quality.values' xmlns:x=''/>">
49  <!ENTITY resource               "<xref target='Part2' x:rel='#resource' xmlns:x=''/>">
50  <!ENTITY status-codes           "<xref target='Part2' x:rel='' xmlns:x=''/>">
51  <!ENTITY status-1xx             "<xref target='Part2' x:rel='#status.1xx' xmlns:x=''/>">
52  <!ENTITY status-203             "<xref target='Part2' x:rel='#status.203' xmlns:x=''/>">
53  <!ENTITY status-3xx             "<xref target='Part2' x:rel='#status.3xx' xmlns:x=''/>">
54  <!ENTITY status-304             "<xref target='Part4' x:rel='#status.304' xmlns:x=''/>">
55  <!ENTITY status-4xx             "<xref target='Part2' x:rel='#status.4xx' xmlns:x=''/>">
56  <!ENTITY status-414             "<xref target='Part2' x:rel='#status.414' xmlns:x=''/>">
57  <!ENTITY iana-header-registry   "<xref target='Part2' x:rel='#header.field.registry' xmlns:x=''/>">
59<?rfc toc="yes" ?>
60<?rfc symrefs="yes" ?>
61<?rfc sortrefs="yes" ?>
62<?rfc compact="yes"?>
63<?rfc subcompact="no" ?>
64<?rfc linkmailto="no" ?>
65<?rfc editing="no" ?>
66<?rfc comments="yes"?>
67<?rfc inline="yes"?>
68<?rfc rfcedstyle="yes"?>
69<?rfc-ext allow-markup-in-artwork="yes" ?>
70<?rfc-ext include-references-in-index="yes" ?>
71<rfc obsoletes="2145,2616" updates="2817" category="std" x:maturity-level="proposed"
72     ipr="pre5378Trust200902" docName="draft-ietf-httpbis-p1-messaging-&ID-VERSION;"
73     xmlns:x=''>
74<x:link rel="next" basename="p2-semantics"/>
75<x:feedback template="{docname},%20%22{section}%22&amp;body=&lt;{ref}&gt;:"/>
78  <title abbrev="HTTP/1.1 Message Syntax and Routing">Hypertext Transfer Protocol (HTTP/1.1): Message Syntax and Routing</title>
80  <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
81    <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
82    <address>
83      <postal>
84        <street>345 Park Ave</street>
85        <city>San Jose</city>
86        <region>CA</region>
87        <code>95110</code>
88        <country>USA</country>
89      </postal>
90      <email></email>
91      <uri></uri>
92    </address>
93  </author>
95  <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
96    <organization abbrev="greenbytes">greenbytes GmbH</organization>
97    <address>
98      <postal>
99        <street>Hafenweg 16</street>
100        <city>Muenster</city><region>NW</region><code>48155</code>
101        <country>Germany</country>
102      </postal>
103      <email></email>
104      <uri></uri>
105    </address>
106  </author>
108  <date month="&ID-MONTH;" year="&ID-YEAR;"/>
109  <workgroup>HTTPbis Working Group</workgroup>
113   The Hypertext Transfer Protocol (HTTP) is an application-level protocol for
114   distributed, collaborative, hypertext information systems. HTTP has been in
115   use by the World Wide Web global information initiative since 1990.
116   This document provides an overview of HTTP architecture and its associated
117   terminology, defines the "http" and "https" Uniform Resource Identifier
118   (URI) schemes, defines the HTTP/1.1 message syntax and parsing requirements,
119   and describes general security concerns for implementations.
123<note title="Editorial Note (To be removed by RFC Editor)">
124  <t>
125    Discussion of this draft takes place on the HTTPBIS working group
126    mailing list (, which is archived at
127    <eref target=""/>.
128  </t>
129  <t>
130    The current issues list is at
131    <eref target=""/> and related
132    documents (including fancy diffs) can be found at
133    <eref target=""/>.
134  </t>
135  <t>
136    The changes in this draft are summarized in <xref target="changes.since.21"/>.
137  </t>
141<section title="Introduction" anchor="introduction">
143   The Hypertext Transfer Protocol (HTTP) is an application-level
144   request/response protocol that uses extensible semantics and MIME-like
145   message payloads for flexible interaction with network-based hypertext
146   information systems. This document is the first in a series of documents
147   that collectively form the HTTP/1.1 specification:
148   <list style="empty">
149    <t>RFC xxx1: Message Syntax and Routing</t>
150    <t><xref target="Part2" x:fmt="none">RFC xxx2</xref>: Semantics and Content</t>
151    <t><xref target="Part4" x:fmt="none">RFC xxx3</xref>: Conditional Requests</t>
152    <t><xref target="Part5" x:fmt="none">RFC xxx4</xref>: Range Requests</t>
153    <t><xref target="Part6" x:fmt="none">RFC xxx5</xref>: Caching</t>
154    <t><xref target="Part7" x:fmt="none">RFC xxx6</xref>: Authentication</t>
155   </list>
158   This HTTP/1.1 specification obsoletes and moves to historic status
159   <xref target="RFC2616" x:fmt="none">RFC 2616</xref>, its predecessor
160   <xref target="RFC2068" x:fmt="none">RFC 2068</xref>,
161   <xref target="RFC2145" x:fmt="none">RFC 2145</xref> (on HTTP versioning),
162   and <xref target="RFC2817" x:fmt="none">RFC 2817</xref> (on using CONNECT
163   for TLS upgrades).
166   HTTP is a generic interface protocol for information systems. It is
167   designed to hide the details of how a service is implemented by presenting
168   a uniform interface to clients that is independent of the types of
169   resources provided. Likewise, servers do not need to be aware of each
170   client's purpose: an HTTP request can be considered in isolation rather
171   than being associated with a specific type of client or a predetermined
172   sequence of application steps. The result is a protocol that can be used
173   effectively in many different contexts and for which implementations can
174   evolve independently over time.
177   HTTP is also designed for use as an intermediation protocol for translating
178   communication to and from non-HTTP information systems.
179   HTTP proxies and gateways can provide access to alternative information
180   services by translating their diverse protocols into a hypertext
181   format that can be viewed and manipulated by clients in the same way
182   as HTTP services.
185   One consequence of HTTP flexibility is that the protocol cannot be
186   defined in terms of what occurs behind the interface. Instead, we
187   are limited to defining the syntax of communication, the intent
188   of received communication, and the expected behavior of recipients.
189   If the communication is considered in isolation, then successful
190   actions ought to be reflected in corresponding changes to the
191   observable interface provided by servers. However, since multiple
192   clients might act in parallel and perhaps at cross-purposes, we
193   cannot require that such changes be observable beyond the scope
194   of a single response.
197   This document describes the architectural elements that are used or
198   referred to in HTTP, defines the "http" and "https" URI schemes,
199   describes overall network operation and connection management,
200   and defines HTTP message framing and forwarding requirements.
201   Our goal is to define all of the mechanisms necessary for HTTP message
202   handling that are independent of message semantics, thereby defining the
203   complete set of requirements for message parsers and
204   message-forwarding intermediaries.
208<section title="Requirement Notation" anchor="intro.requirements">
210   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
211   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
212   document are to be interpreted as described in <xref target="RFC2119"/>.
215   Conformance criteria and considerations regarding error handling
216   are defined in <xref target="conformance"/>.
220<section title="Syntax Notation" anchor="notation">
221<iref primary="true" item="Grammar" subitem="ALPHA"/>
222<iref primary="true" item="Grammar" subitem="CR"/>
223<iref primary="true" item="Grammar" subitem="CRLF"/>
224<iref primary="true" item="Grammar" subitem="CTL"/>
225<iref primary="true" item="Grammar" subitem="DIGIT"/>
226<iref primary="true" item="Grammar" subitem="DQUOTE"/>
227<iref primary="true" item="Grammar" subitem="HEXDIG"/>
228<iref primary="true" item="Grammar" subitem="HTAB"/>
229<iref primary="true" item="Grammar" subitem="LF"/>
230<iref primary="true" item="Grammar" subitem="OCTET"/>
231<iref primary="true" item="Grammar" subitem="SP"/>
232<iref primary="true" item="Grammar" subitem="VCHAR"/>
234   This specification uses the Augmented Backus-Naur Form (ABNF) notation
235   of <xref target="RFC5234"/> with the list rule extension defined in
236   <xref target="abnf.extension"/>.  <xref target="collected.abnf"/> shows
237   the collected ABNF with the list rule expanded.
239<t anchor="core.rules">
240  <x:anchor-alias value="ALPHA"/>
241  <x:anchor-alias value="CTL"/>
242  <x:anchor-alias value="CR"/>
243  <x:anchor-alias value="CRLF"/>
244  <x:anchor-alias value="DIGIT"/>
245  <x:anchor-alias value="DQUOTE"/>
246  <x:anchor-alias value="HEXDIG"/>
247  <x:anchor-alias value="HTAB"/>
248  <x:anchor-alias value="LF"/>
249  <x:anchor-alias value="OCTET"/>
250  <x:anchor-alias value="SP"/>
251  <x:anchor-alias value="VCHAR"/>
252   The following core rules are included by
253   reference, as defined in <xref target="RFC5234" x:fmt="," x:sec="B.1"/>:
254   ALPHA (letters), CR (carriage return), CRLF (CR LF), CTL (controls),
255   DIGIT (decimal 0-9), DQUOTE (double quote),
256   HEXDIG (hexadecimal 0-9/A-F/a-f), HTAB (horizontal tab), LF (line feed),
257   OCTET (any 8-bit sequence of data), SP (space), and
258   VCHAR (any visible <xref target="USASCII"/> character).
261   As a convention, ABNF rule names prefixed with "obs-" denote
262   "obsolete" grammar rules that appear for historical reasons.
267<section title="Architecture" anchor="architecture">
269   HTTP was created for the World Wide Web architecture
270   and has evolved over time to support the scalability needs of a worldwide
271   hypertext system. Much of that architecture is reflected in the terminology
272   and syntax productions used to define HTTP.
275<section title="Client/Server Messaging" anchor="operation">
276<iref primary="true" item="client"/>
277<iref primary="true" item="server"/>
278<iref primary="true" item="connection"/>
280   HTTP is a stateless request/response protocol that operates by exchanging
281   <x:dfn>messages</x:dfn> (<xref target="http.message"/>) across a reliable
282   transport or session-layer
283   "<x:dfn>connection</x:dfn>" (<xref target=""/>).
284   An HTTP "<x:dfn>client</x:dfn>" is a program that establishes a connection
285   to a server for the purpose of sending one or more HTTP requests.
286   An HTTP "<x:dfn>server</x:dfn>" is a program that accepts connections
287   in order to service HTTP requests by sending HTTP responses.
289<iref primary="true" item="user agent"/>
290<iref primary="true" item="origin server"/>
291<iref primary="true" item="browser"/>
292<iref primary="true" item="spider"/>
293<iref primary="true" item="sender"/>
294<iref primary="true" item="recipient"/>
296   The terms client and server refer only to the roles that
297   these programs perform for a particular connection.  The same program
298   might act as a client on some connections and a server on others.  We use
299   the term "<x:dfn>user agent</x:dfn>" to refer to the program that initiates a request,
300   such as a WWW browser, editor, or spider (web-traversing robot), and
301   the term "<x:dfn>origin server</x:dfn>" to refer to the program that can originate
302   authoritative responses to a request.  For general requirements, we use
303   the term "<x:dfn>sender</x:dfn>" to refer to whichever component sent a given message
304   and the term "<x:dfn>recipient</x:dfn>" to refer to any component that receives the
305   message.
308   HTTP relies upon the Uniform Resource Identifier (URI)
309   standard <xref target="RFC3986"/> to indicate the target resource
310   (<xref target="target-resource"/>) and relationships between resources.
311   Messages are passed in a format similar to that used by Internet mail
312   <xref target="RFC5322"/> and the Multipurpose Internet Mail Extensions
313   (MIME) <xref target="RFC2045"/> (see &diff-mime; for the differences
314   between HTTP and MIME messages).
317   Most HTTP communication consists of a retrieval request (GET) for
318   a representation of some resource identified by a URI.  In the
319   simplest case, this might be accomplished via a single bidirectional
320   connection (===) between the user agent (UA) and the origin server (O).
322<figure><artwork type="drawing">
323         request   &gt;
324    <x:highlight>UA</x:highlight> ======================================= <x:highlight>O</x:highlight>
325                                &lt;   response
327<iref primary="true" item="message"/>
328<iref primary="true" item="request"/>
329<iref primary="true" item="response"/>
331   A client sends an HTTP request to a server in the form of a <x:dfn>request</x:dfn>
332   message, beginning with a request-line that includes a method, URI, and
333   protocol version (<xref target="request.line"/>),
334   followed by header fields containing
335   request modifiers, client information, and representation metadata
336   (<xref target="header.fields"/>),
337   an empty line to indicate the end of the header section, and finally
338   a message body containing the payload body (if any,
339   <xref target="message.body"/>).
342   A server responds to a client's request by sending one or more HTTP
343   <x:dfn>response</x:dfn>
344   messages, each beginning with a status line that
345   includes the protocol version, a success or error code, and textual
346   reason phrase (<xref target="status.line"/>),
347   possibly followed by header fields containing server
348   information, resource metadata, and representation metadata
349   (<xref target="header.fields"/>),
350   an empty line to indicate the end of the header section, and finally
351   a message body containing the payload body (if any,
352   <xref target="message.body"/>).
355   A connection might be used for multiple request/response exchanges,
356   as defined in <xref target="persistent.connections"/>.
359   The following example illustrates a typical message exchange for a
360   GET request on the URI "":
363client request:
364</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
365GET /hello.txt HTTP/1.1
366User-Agent: curl/7.16.3 libcurl/7.16.3 OpenSSL/0.9.7l zlib/1.2.3
368Accept-Language: en, mi
372server response:
373</preamble><artwork type="message/http; msgtype=&#34;response&#34;" x:indent-with="  ">
374HTTP/1.1 200 OK
375Date: Mon, 27 Jul 2009 12:28:53 GMT
376Server: Apache
377Last-Modified: Wed, 22 Jul 2009 19:15:56 GMT
378ETag: "34aa387-d-1568eb00"
379Accept-Ranges: bytes
380Content-Length: <x:length-of target="exbody"/>
381Vary: Accept-Encoding
382Content-Type: text/plain
384<x:span anchor="exbody">Hello World!
388<section title="Implementation Diversity" anchor="implementation-diversity">
390   When considering the design of HTTP, it is easy to fall into a trap of
391   thinking that all user agents are general-purpose browsers and all origin
392   servers are large public websites. That is not the case in practice.
393   Common HTTP user agents include household appliances, stereos, scales,
394   firmware update scripts, command-line programs, mobile apps,
395   and communication devices in a multitude of shapes and sizes.  Likewise,
396   common HTTP origin servers include home automation units, configurable
397   networking components, office machines, autonomous robots, news feeds,
398   traffic cameras, ad selectors, and video delivery platforms.
401   The term "user agent" does not imply that there is a human user directly
402   interacting with the software agent at the time of a request. In many
403   cases, a user agent is installed or configured to run in the background
404   and save its results for later inspection (or save only a subset of those
405   results that might be interesting or erroneous). Spiders, for example, are
406   typically given a start URI and configured to follow certain behavior while
407   crawling the Web as a hypertext graph.
410   The implementation diversity of HTTP means that we cannot assume the
411   user agent can make interactive suggestions to a user or provide adequate
412   warning for security or privacy options.  In the few cases where this
413   specification requires reporting of errors to the user, it is acceptable
414   for such reporting to only be observable in an error console or log file.
415   Likewise, requirements that an automated action be confirmed by the user
416   before proceeding can be met via advance configuration choices,
417   run-time options, or simply not proceeding with the unsafe action.
421<section title="Intermediaries" anchor="intermediaries">
422<iref primary="true" item="intermediary"/>
424   HTTP enables the use of intermediaries to satisfy requests through
425   a chain of connections.  There are three common forms of HTTP
426   <x:dfn>intermediary</x:dfn>: proxy, gateway, and tunnel.  In some cases,
427   a single intermediary might act as an origin server, proxy, gateway,
428   or tunnel, switching behavior based on the nature of each request.
430<figure><artwork type="drawing">
431         &gt;             &gt;             &gt;             &gt;
432    <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>
433               &lt;             &lt;             &lt;             &lt;
436   The figure above shows three intermediaries (A, B, and C) between the
437   user agent and origin server. A request or response message that
438   travels the whole chain will pass through four separate connections.
439   Some HTTP communication options
440   might apply only to the connection with the nearest, non-tunnel
441   neighbor, only to the end-points of the chain, or to all connections
442   along the chain. Although the diagram is linear, each participant might
443   be engaged in multiple, simultaneous communications. For example, B
444   might be receiving requests from many clients other than A, and/or
445   forwarding requests to servers other than C, at the same time that it
446   is handling A's request.
449<iref primary="true" item="upstream"/><iref primary="true" item="downstream"/>
450<iref primary="true" item="inbound"/><iref primary="true" item="outbound"/>
451   We use the terms "<x:dfn>upstream</x:dfn>" and "<x:dfn>downstream</x:dfn>"
452   to describe various requirements in relation to the directional flow of a
453   message: all messages flow from upstream to downstream.
454   Likewise, we use the terms inbound and outbound to refer to
455   directions in relation to the request path:
456   "<x:dfn>inbound</x:dfn>" means toward the origin server and
457   "<x:dfn>outbound</x:dfn>" means toward the user agent.
459<t><iref primary="true" item="proxy"/>
460   A "<x:dfn>proxy</x:dfn>" is a message forwarding agent that is selected by the
461   client, usually via local configuration rules, to receive requests
462   for some type(s) of absolute URI and attempt to satisfy those
463   requests via translation through the HTTP interface.  Some translations
464   are minimal, such as for proxy requests for "http" URIs, whereas
465   other requests might require translation to and from entirely different
466   application-level protocols. Proxies are often used to group an
467   organization's HTTP requests through a common intermediary for the
468   sake of security, annotation services, or shared caching.
471<iref primary="true" item="transforming proxy"/>
472<iref primary="true" item="non-transforming proxy"/>
473   An HTTP-to-HTTP proxy is called a "<x:dfn>transforming proxy</x:dfn>" if it is designed
474   or configured to modify request or response messages in a semantically
475   meaningful way (i.e., modifications, beyond those required by normal
476   HTTP processing, that change the message in a way that would be
477   significant to the original sender or potentially significant to
478   downstream recipients).  For example, a transforming proxy might be
479   acting as a shared annotation server (modifying responses to include
480   references to a local annotation database), a malware filter, a
481   format transcoder, or an intranet-to-Internet privacy filter.  Such
482   transformations are presumed to be desired by the client (or client
483   organization) that selected the proxy and are beyond the scope of
484   this specification.  However, when a proxy is not intended to transform
485   a given message, we use the term "<x:dfn>non-transforming proxy</x:dfn>" to target
486   requirements that preserve HTTP message semantics. See &status-203; and
487   &header-warning; for status and warning codes related to transformations.
489<t><iref primary="true" item="gateway"/><iref primary="true" item="reverse proxy"/>
490<iref primary="true" item="accelerator"/>
491   A "<x:dfn>gateway</x:dfn>" (a.k.a., "<x:dfn>reverse proxy</x:dfn>")
492   is a receiving agent that acts
493   as a layer above some other server(s) and translates the received
494   requests to the underlying server's protocol.  Gateways are often
495   used to encapsulate legacy or untrusted information services, to
496   improve server performance through "<x:dfn>accelerator</x:dfn>" caching, and to
497   enable partitioning or load-balancing of HTTP services across
498   multiple machines.
501   A gateway behaves as an origin server on its outbound connection and
502   as a user agent on its inbound connection.
503   All HTTP requirements applicable to an origin server
504   also apply to the outbound communication of a gateway.
505   A gateway communicates with inbound servers using any protocol that
506   it desires, including private extensions to HTTP that are outside
507   the scope of this specification.  However, an HTTP-to-HTTP gateway
508   that wishes to interoperate with third-party HTTP servers &MUST;
509   conform to HTTP user agent requirements on the gateway's inbound
510   connection and &MUST; implement the <x:ref>Connection</x:ref>
511   (<xref target="header.connection"/>) and <x:ref>Via</x:ref>
512   (<xref target="header.via"/>) header fields for both connections.
514<t><iref primary="true" item="tunnel"/>
515   A "<x:dfn>tunnel</x:dfn>" acts as a blind relay between two connections
516   without changing the messages. Once active, a tunnel is not
517   considered a party to the HTTP communication, though the tunnel might
518   have been initiated by an HTTP request. A tunnel ceases to exist when
519   both ends of the relayed connection are closed. Tunnels are used to
520   extend a virtual connection through an intermediary, such as when
521   Transport Layer Security (TLS, <xref target="RFC5246"/>) is used to
522   establish confidential communication through a shared firewall proxy.
524<t><iref primary="true" item="interception proxy"/>
525<iref primary="true" item="transparent proxy"/>
526<iref primary="true" item="captive portal"/>
527   The above categories for intermediary only consider those acting as
528   participants in the HTTP communication.  There are also intermediaries
529   that can act on lower layers of the network protocol stack, filtering or
530   redirecting HTTP traffic without the knowledge or permission of message
531   senders. Network intermediaries often introduce security flaws or
532   interoperability problems by violating HTTP semantics.  For example, an
533   "<x:dfn>interception proxy</x:dfn>" <xref target="RFC3040"/> (also commonly
534   known as a "<x:dfn>transparent proxy</x:dfn>" <xref target="RFC1919"/> or
535   "<x:dfn>captive portal</x:dfn>")
536   differs from an HTTP proxy because it is not selected by the client.
537   Instead, an interception proxy filters or redirects outgoing TCP port 80
538   packets (and occasionally other common port traffic).
539   Interception proxies are commonly found on public network access points,
540   as a means of enforcing account subscription prior to allowing use of
541   non-local Internet services, and within corporate firewalls to enforce
542   network usage policies.
543   They are indistinguishable from a man-in-the-middle attack.
546   HTTP is defined as a stateless protocol, meaning that each request message
547   can be understood in isolation.  Many implementations depend on HTTP's
548   stateless design in order to reuse proxied connections or dynamically
549   load-balance requests across multiple servers.  Hence, servers &MUST-NOT;
550   assume that two requests on the same connection are from the same user
551   agent unless the connection is secured and specific to that agent.
552   Some non-standard HTTP extensions (e.g., <xref target="RFC4559"/>) have
553   been known to violate this requirement, resulting in security and
554   interoperability problems.
558<section title="Caches" anchor="caches">
559<iref primary="true" item="cache"/>
561   A "<x:dfn>cache</x:dfn>" is a local store of previous response messages and the
562   subsystem that controls its message storage, retrieval, and deletion.
563   A cache stores cacheable responses in order to reduce the response
564   time and network bandwidth consumption on future, equivalent
565   requests. Any client or server &MAY; employ a cache, though a cache
566   cannot be used by a server while it is acting as a tunnel.
569   The effect of a cache is that the request/response chain is shortened
570   if one of the participants along the chain has a cached response
571   applicable to that request. The following illustrates the resulting
572   chain if B has a cached copy of an earlier response from O (via C)
573   for a request which has not been cached by UA or A.
575<figure><artwork type="drawing">
576            &gt;             &gt;
577       <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>
578                  &lt;             &lt;
580<t><iref primary="true" item="cacheable"/>
581   A response is "<x:dfn>cacheable</x:dfn>" if a cache is allowed to store a copy of
582   the response message for use in answering subsequent requests.
583   Even when a response is cacheable, there might be additional
584   constraints placed by the client or by the origin server on when
585   that cached response can be used for a particular request. HTTP
586   requirements for cache behavior and cacheable responses are
587   defined in &caching-overview;. 
590   There are a wide variety of architectures and configurations
591   of caches and proxies deployed across the World Wide Web and
592   inside large organizations. These systems include national hierarchies
593   of proxy caches to save transoceanic bandwidth, systems that
594   broadcast or multicast cache entries, organizations that distribute
595   subsets of cached data via optical media, and so on.
599<section title="Conformance and Error Handling" anchor="conformance">
601   This specification targets conformance criteria according to the role of
602   a participant in HTTP communication.  Hence, HTTP requirements are placed
603   on senders, recipients, clients, servers, user agents, intermediaries,
604   origin servers, proxies, gateways, or caches, depending on what behavior
605   is being constrained by the requirement. Additional (social) requirements
606   are placed on implementations, resource owners, and protocol element
607   registrations when they apply beyond the scope of a single communication.
610   The verb "generate" is used instead of "send" where a requirement
611   differentiates between creating a protocol element and merely forwarding a
612   received element downstream.
615   An implementation is considered conformant if it complies with all of the
616   requirements associated with the roles it partakes in HTTP. Note that
617   SHOULD-level requirements are relevant here, unless one of the documented
618   exceptions is applicable.
621   Conformance applies to both the syntax and semantics of HTTP protocol
622   elements. A sender &MUST-NOT; generate protocol elements that convey a
623   meaning that is known by that sender to be false. A sender &MUST-NOT;
624   generate protocol elements that do not match the grammar defined by the
625   ABNF rules for those protocol elements that are applicable to the sender's
626   role. If a received protocol element is processed, the recipient &MUST; be
627   able to parse any value that would match the ABNF rules for that protocol
628   element, excluding only those rules not applicable to the recipient's role.
631   Unless noted otherwise, a recipient &MAY; attempt to recover a usable
632   protocol element from an invalid construct.  HTTP does not define
633   specific error handling mechanisms except when they have a direct impact
634   on security, since different applications of the protocol require
635   different error handling strategies.  For example, a Web browser might
636   wish to transparently recover from a response where the
637   <x:ref>Location</x:ref> header field doesn't parse according to the ABNF,
638   whereas a systems control client might consider any form of error recovery
639   to be dangerous.
643<section title="Protocol Versioning" anchor="http.version">
644  <x:anchor-alias value="HTTP-version"/>
645  <x:anchor-alias value="HTTP-name"/>
647   HTTP uses a "&lt;major&gt;.&lt;minor&gt;" numbering scheme to indicate
648   versions of the protocol. This specification defines version "1.1".
649   The protocol version as a whole indicates the sender's conformance
650   with the set of requirements laid out in that version's corresponding
651   specification of HTTP.
654   The version of an HTTP message is indicated by an HTTP-version field
655   in the first line of the message. HTTP-version is case-sensitive.
657<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-version"/><iref primary="true" item="Grammar" subitem="HTTP-name"/>
658  <x:ref>HTTP-version</x:ref>  = <x:ref>HTTP-name</x:ref> "/" <x:ref>DIGIT</x:ref> "." <x:ref>DIGIT</x:ref>
659  <x:ref>HTTP-name</x:ref>     = <x:abnf-char-sequence>"HTTP"</x:abnf-char-sequence> ; "HTTP", case-sensitive
662   The HTTP version number consists of two decimal digits separated by a "."
663   (period or decimal point).  The first digit ("major version") indicates the
664   HTTP messaging syntax, whereas the second digit ("minor version") indicates
665   the highest minor version to which the sender is
666   conformant and able to understand for future communication.  The minor
667   version advertises the sender's communication capabilities even when the
668   sender is only using a backwards-compatible subset of the protocol,
669   thereby letting the recipient know that more advanced features can
670   be used in response (by servers) or in future requests (by clients).
673   When an HTTP/1.1 message is sent to an HTTP/1.0 recipient
674   <xref target="RFC1945"/> or a recipient whose version is unknown,
675   the HTTP/1.1 message is constructed such that it can be interpreted
676   as a valid HTTP/1.0 message if all of the newer features are ignored.
677   This specification places recipient-version requirements on some
678   new features so that a conformant sender will only use compatible
679   features until it has determined, through configuration or the
680   receipt of a message, that the recipient supports HTTP/1.1.
683   The interpretation of a header field does not change between minor
684   versions of the same major HTTP version, though the default
685   behavior of a recipient in the absence of such a field can change.
686   Unless specified otherwise, header fields defined in HTTP/1.1 are
687   defined for all versions of HTTP/1.x.  In particular, the <x:ref>Host</x:ref>
688   and <x:ref>Connection</x:ref> header fields ought to be implemented by all
689   HTTP/1.x implementations whether or not they advertise conformance with
690   HTTP/1.1.
693   New header fields can be defined such that, when they are
694   understood by a recipient, they might override or enhance the
695   interpretation of previously defined header fields.  When an
696   implementation receives an unrecognized header field, the recipient
697   &MUST; ignore that header field for local processing regardless of
698   the message's HTTP version.  An unrecognized header field received
699   by a proxy &MUST; be forwarded downstream unless the header field's
700   field-name is listed in the message's <x:ref>Connection</x:ref> header field
701   (see <xref target="header.connection"/>).
702   These requirements allow HTTP's functionality to be enhanced without
703   requiring prior update of deployed intermediaries.
706   Intermediaries that process HTTP messages (i.e., all intermediaries
707   other than those acting as tunnels) &MUST; send their own HTTP-version
708   in forwarded messages.  In other words, they &MUST-NOT; blindly
709   forward the first line of an HTTP message without ensuring that the
710   protocol version in that message matches a version to which that
711   intermediary is conformant for both the receiving and
712   sending of messages.  Forwarding an HTTP message without rewriting
713   the HTTP-version might result in communication errors when downstream
714   recipients use the message sender's version to determine what features
715   are safe to use for later communication with that sender.
718   An HTTP client &SHOULD; send a request version equal to the highest
719   version to which the client is conformant and
720   whose major version is no higher than the highest version supported
721   by the server, if this is known.  An HTTP client &MUST-NOT; send a
722   version to which it is not conformant.
725   An HTTP client &MAY; send a lower request version if it is known that
726   the server incorrectly implements the HTTP specification, but only
727   after the client has attempted at least one normal request and determined
728   from the response status or header fields (e.g., <x:ref>Server</x:ref>) that
729   the server improperly handles higher request versions.
732   An HTTP server &SHOULD; send a response version equal to the highest
733   version to which the server is conformant and
734   whose major version is less than or equal to the one received in the
735   request.  An HTTP server &MUST-NOT; send a version to which it is not
736   conformant.  A server &MAY; send a <x:ref>505 (HTTP Version Not
737   Supported)</x:ref> response if it cannot send a response using the
738   major version used in the client's request.
741   An HTTP server &MAY; send an HTTP/1.0 response to an HTTP/1.0 request
742   if it is known or suspected that the client incorrectly implements the
743   HTTP specification and is incapable of correctly processing later
744   version responses, such as when a client fails to parse the version
745   number correctly or when an intermediary is known to blindly forward
746   the HTTP-version even when it doesn't conform to the given minor
747   version of the protocol. Such protocol downgrades &SHOULD-NOT; be
748   performed unless triggered by specific client attributes, such as when
749   one or more of the request header fields (e.g., <x:ref>User-Agent</x:ref>)
750   uniquely match the values sent by a client known to be in error.
753   The intention of HTTP's versioning design is that the major number
754   will only be incremented if an incompatible message syntax is
755   introduced, and that the minor number will only be incremented when
756   changes made to the protocol have the effect of adding to the message
757   semantics or implying additional capabilities of the sender.  However,
758   the minor version was not incremented for the changes introduced between
759   <xref target="RFC2068"/> and <xref target="RFC2616"/>, and this revision
760   has specifically avoiding any such changes to the protocol.
764<section title="Uniform Resource Identifiers" anchor="uri">
765<iref primary="true" item="resource"/>
767   Uniform Resource Identifiers (URIs) <xref target="RFC3986"/> are used
768   throughout HTTP as the means for identifying resources (&resource;).
769   URI references are used to target requests, indicate redirects, and define
770   relationships.
772  <x:anchor-alias value="URI-reference"/>
773  <x:anchor-alias value="absolute-URI"/>
774  <x:anchor-alias value="relative-part"/>
775  <x:anchor-alias value="authority"/>
776  <x:anchor-alias value="path-abempty"/>
777  <x:anchor-alias value="path-absolute"/>
778  <x:anchor-alias value="port"/>
779  <x:anchor-alias value="query"/>
780  <x:anchor-alias value="uri-host"/>
781  <x:anchor-alias value="partial-URI"/>
783   This specification adopts the definitions of "URI-reference",
784   "absolute-URI", "relative-part", "port", "host",
785   "path-abempty", "path-absolute", "query", and "authority" from the
786   URI generic syntax.
787   In addition, we define a partial-URI rule for protocol elements
788   that allow a relative URI but not a fragment.
790<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="URI-reference"><!--exported production--></iref><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"/><iref primary="true" item="Grammar" subitem="partial-URI"><!--exported production--></iref>
791  <x:ref>URI-reference</x:ref> = &lt;URI-reference, defined in <xref target="RFC3986" x:fmt="," x:sec="4.1"/>&gt;
792  <x:ref>absolute-URI</x:ref>  = &lt;absolute-URI, defined in <xref target="RFC3986" x:fmt="," x:sec="4.3"/>&gt;
793  <x:ref>relative-part</x:ref> = &lt;relative-part, defined in <xref target="RFC3986" x:fmt="," x:sec="4.2"/>&gt;
794  <x:ref>authority</x:ref>     = &lt;authority, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2"/>&gt;
795  <x:ref>path-abempty</x:ref>  = &lt;path-abempty, defined in <xref target="RFC3986" x:fmt="," x:sec="3.3"/>&gt;
796  <x:ref>path-absolute</x:ref> = &lt;path-absolute, defined in <xref target="RFC3986" x:fmt="," x:sec="3.3"/>&gt;
797  <x:ref>port</x:ref>          = &lt;port, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2.3"/>&gt;
798  <x:ref>query</x:ref>         = &lt;query, defined in <xref target="RFC3986" x:fmt="," x:sec="3.4"/>&gt;
799  <x:ref>uri-host</x:ref>      = &lt;host, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2.2"/>&gt;
801  <x:ref>partial-URI</x:ref>   = relative-part [ "?" query ]
804   Each protocol element in HTTP that allows a URI reference will indicate
805   in its ABNF production whether the element allows any form of reference
806   (URI-reference), only a URI in absolute form (absolute-URI), only the
807   path and optional query components, or some combination of the above.
808   Unless otherwise indicated, URI references are parsed
809   relative to the effective request URI
810   (<xref target="effective.request.uri"/>).
813<section title="http URI scheme" anchor="http.uri">
814  <x:anchor-alias value="http-URI"/>
815  <iref item="http URI scheme" primary="true"/>
816  <iref item="URI scheme" subitem="http" primary="true"/>
818   The "http" URI scheme is hereby defined for the purpose of minting
819   identifiers according to their association with the hierarchical
820   namespace governed by a potential HTTP origin server listening for
821   TCP connections on a given port.
823<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="http-URI"><!--terminal production--></iref>
824  <x:ref>http-URI</x:ref> = "http:" "//" <x:ref>authority</x:ref> <x:ref>path-abempty</x:ref> [ "?" <x:ref>query</x:ref> ]
827   The HTTP origin server is identified by the generic syntax's
828   <x:ref>authority</x:ref> component, which includes a host identifier
829   and optional TCP port (<xref target="RFC3986" x:fmt="," x:sec="3.2.2"/>).
830   The remainder of the URI, consisting of both the hierarchical path
831   component and optional query component, serves as an identifier for
832   a potential resource within that origin server's name space.
835   If the host identifier is provided as an IP literal or IPv4 address,
836   then the origin server is any listener on the indicated TCP port at
837   that IP address. If host is a registered name, then that name is
838   considered an indirect identifier and the recipient might use a name
839   resolution service, such as DNS, to find the address of a listener
840   for that host.
841   The host &MUST-NOT; be empty; if an "http" URI is received with an
842   empty host, then it &MUST; be rejected as invalid.
843   If the port subcomponent is empty or not given, then TCP port 80 is
844   assumed (the default reserved port for WWW services).
847   Regardless of the form of host identifier, access to that host is not
848   implied by the mere presence of its name or address. The host might or might
849   not exist and, even when it does exist, might or might not be running an
850   HTTP server or listening to the indicated port. The "http" URI scheme
851   makes use of the delegated nature of Internet names and addresses to
852   establish a naming authority (whatever entity has the ability to place
853   an HTTP server at that Internet name or address) and allows that
854   authority to determine which names are valid and how they might be used.
857   When an "http" URI is used within a context that calls for access to the
858   indicated resource, a client &MAY; attempt access by resolving
859   the host to an IP address, establishing a TCP connection to that address
860   on the indicated port, and sending an HTTP request message
861   (<xref target="http.message"/>) containing the URI's identifying data
862   (<xref target="message.routing"/>) to the server.
863   If the server responds to that request with a non-interim HTTP response
864   message, as described in &status-codes;, then that response
865   is considered an authoritative answer to the client's request.
868   Although HTTP is independent of the transport protocol, the "http"
869   scheme is specific to TCP-based services because the name delegation
870   process depends on TCP for establishing authority.
871   An HTTP service based on some other underlying connection protocol
872   would presumably be identified using a different URI scheme, just as
873   the "https" scheme (below) is used for resources that require an
874   end-to-end secured connection. Other protocols might also be used to
875   provide access to "http" identified resources &mdash; it is only the
876   authoritative interface used for mapping the namespace that is
877   specific to TCP.
880   The URI generic syntax for authority also includes a deprecated
881   userinfo subcomponent (<xref target="RFC3986" x:fmt="," x:sec="3.2.1"/>)
882   for including user authentication information in the URI.  Some
883   implementations make use of the userinfo component for internal
884   configuration of authentication information, such as within command
885   invocation options, configuration files, or bookmark lists, even
886   though such usage might expose a user identifier or password.
887   Senders &MUST-NOT; include a userinfo subcomponent (and its "@"
888   delimiter) when transmitting an "http" URI in a message.  Recipients
889   of HTTP messages that contain a URI reference &SHOULD; parse for the
890   existence of userinfo and treat its presence as an error, likely
891   indicating that the deprecated subcomponent is being used to obscure
892   the authority for the sake of phishing attacks.
896<section title="https URI scheme" anchor="https.uri">
897   <x:anchor-alias value="https-URI"/>
898   <iref item="https URI scheme"/>
899   <iref item="URI scheme" subitem="https"/>
901   The "https" URI scheme is hereby defined for the purpose of minting
902   identifiers according to their association with the hierarchical
903   namespace governed by a potential HTTP origin server listening to a
904   given TCP port for TLS-secured connections <xref target="RFC5246"/>.
907   All of the requirements listed above for the "http" scheme are also
908   requirements for the "https" scheme, except that a default TCP port
909   of 443 is assumed if the port subcomponent is empty or not given,
910   and the TCP connection &MUST; be secured, end-to-end, through the
911   use of strong encryption prior to sending the first HTTP request.
913<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="https-URI"><!--terminal production--></iref>
914  <x:ref>https-URI</x:ref> = "https:" "//" <x:ref>authority</x:ref> <x:ref>path-abempty</x:ref> [ "?" <x:ref>query</x:ref> ]
917   Unlike the "http" scheme, responses to "https" identified requests
918   are never "public" and thus &MUST-NOT; be reused for shared caching.
919   They can, however, be reused in a private cache if the message is
920   cacheable by default in HTTP or specifically indicated as such by
921   the Cache-Control header field (&header-cache-control;).
924   Resources made available via the "https" scheme have no shared
925   identity with the "http" scheme even if their resource identifiers
926   indicate the same authority (the same host listening to the same
927   TCP port).  They are distinct name spaces and are considered to be
928   distinct origin servers.  However, an extension to HTTP that is
929   defined to apply to entire host domains, such as the Cookie protocol
930   <xref target="RFC6265"/>, can allow information
931   set by one service to impact communication with other services
932   within a matching group of host domains.
935   The process for authoritative access to an "https" identified
936   resource is defined in <xref target="RFC2818"/>.
940<section title="http and https URI Normalization and Comparison" anchor="uri.comparison">
942   Since the "http" and "https" schemes conform to the URI generic syntax,
943   such URIs are normalized and compared according to the algorithm defined
944   in <xref target="RFC3986" x:fmt="," x:sec="6"/>, using the defaults
945   described above for each scheme.
948   If the port is equal to the default port for a scheme, the normal
949   form is to elide the port subcomponent. Likewise, an empty path
950   component is equivalent to an absolute path of "/", so the normal
951   form is to provide a path of "/" instead. The scheme and host
952   are case-insensitive and normally provided in lowercase; all
953   other components are compared in a case-sensitive manner.
954   Characters other than those in the "reserved" set are equivalent
955   to their percent-encoded octets (see <xref target="RFC3986"
956   x:fmt="," x:sec="2.1"/>): the normal form is to not encode them.
959   For example, the following three URIs are equivalent:
961<figure><artwork type="example">
970<section title="Message Format" anchor="http.message">
971<x:anchor-alias value="generic-message"/>
972<x:anchor-alias value="message.types"/>
973<x:anchor-alias value="HTTP-message"/>
974<x:anchor-alias value="start-line"/>
975<iref item="header section"/>
976<iref item="headers"/>
977<iref item="header field"/>
979   All HTTP/1.1 messages consist of a start-line followed by a sequence of
980   octets in a format similar to the Internet Message Format
981   <xref target="RFC5322"/>: zero or more header fields (collectively
982   referred to as the "headers" or the "header section"), an empty line
983   indicating the end of the header section, and an optional message body.
985<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-message"><!--terminal production--></iref>
986  <x:ref>HTTP-message</x:ref>   = <x:ref>start-line</x:ref>
987                   *( <x:ref>header-field</x:ref> <x:ref>CRLF</x:ref> )
988                   <x:ref>CRLF</x:ref>
989                   [ <x:ref>message-body</x:ref> ]
992   The normal procedure for parsing an HTTP message is to read the
993   start-line into a structure, read each header field into a hash
994   table by field name until the empty line, and then use the parsed
995   data to determine if a message body is expected.  If a message body
996   has been indicated, then it is read as a stream until an amount
997   of octets equal to the message body length is read or the connection
998   is closed.
1001   Recipients &MUST; parse an HTTP message as a sequence of octets in an
1002   encoding that is a superset of US-ASCII <xref target="USASCII"/>.
1003   Parsing an HTTP message as a stream of Unicode characters, without regard
1004   for the specific encoding, creates security vulnerabilities due to the
1005   varying ways that string processing libraries handle invalid multibyte
1006   character sequences that contain the octet LF (%x0A).  String-based
1007   parsers can only be safely used within protocol elements after the element
1008   has been extracted from the message, such as within a header field-value
1009   after message parsing has delineated the individual fields.
1012   An HTTP message can be parsed as a stream for incremental processing or
1013   forwarding downstream.  However, recipients cannot rely on incremental
1014   delivery of partial messages, since some implementations will buffer or
1015   delay message forwarding for the sake of network efficiency, security
1016   checks, or payload transformations.
1019<section title="Start Line" anchor="start.line">
1020  <x:anchor-alias value="Start-Line"/>
1022   An HTTP message can either be a request from client to server or a
1023   response from server to client.  Syntactically, the two types of message
1024   differ only in the start-line, which is either a request-line (for requests)
1025   or a status-line (for responses), and in the algorithm for determining
1026   the length of the message body (<xref target="message.body"/>).
1029   In theory, a client could receive requests and a server could receive
1030   responses, distinguishing them by their different start-line formats,
1031   but in practice servers are implemented to only expect a request
1032   (a response is interpreted as an unknown or invalid request method)
1033   and clients are implemented to only expect a response.
1035<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="start-line"/>
1036  <x:ref>start-line</x:ref>     = <x:ref>request-line</x:ref> / <x:ref>status-line</x:ref>
1039   A sender &MUST-NOT; send whitespace between the start-line and
1040   the first header field. The presence of such whitespace in a request
1041   might be an attempt to trick a server into ignoring that field or
1042   processing the line after it as a new request, either of which might
1043   result in a security vulnerability if other implementations within
1044   the request chain interpret the same message differently.
1045   Likewise, the presence of such whitespace in a response might be
1046   ignored by some clients or cause others to cease parsing.
1049<section title="Request Line" anchor="request.line">
1050  <x:anchor-alias value="Request"/>
1051  <x:anchor-alias value="request-line"/>
1053   A request-line begins with a method token, followed by a single
1054   space (SP), the request-target, another single space (SP), the
1055   protocol version, and ending with CRLF.
1057<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="request-line"/>
1058  <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>
1061   A server &MUST; be able to parse any received message that begins
1062   with a request-line and matches the ABNF rule for HTTP-message.
1064<iref primary="true" item="method"/>
1065<t anchor="method">
1066   The method token indicates the request method to be performed on the
1067   target resource. The request method is case-sensitive.
1069<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="method"/>
1070  <x:ref>method</x:ref>         = <x:ref>token</x:ref>
1073   The methods defined by this specification can be found in
1074   &methods;, along with information regarding the HTTP method registry
1075   and considerations for defining new methods.
1077<iref item="request-target"/>
1079   The request-target identifies the target resource upon which to apply
1080   the request, as defined in <xref target="request-target"/>.
1083   No whitespace is allowed inside the method, request-target, and
1084   protocol version.  Hence, recipients typically parse the request-line
1085   into its component parts by splitting on the SP characters.
1088   Unfortunately, some user agents fail to properly encode hypertext
1089   references that have embedded whitespace, sending the characters
1090   directly instead of properly percent-encoding the disallowed characters.
1091   Recipients of an invalid request-line &SHOULD; respond with either a
1092   <x:ref>400 (Bad Request)</x:ref> error or a <x:ref>301 (Moved Permanently)</x:ref>
1093   redirect with the request-target properly encoded.  Recipients &SHOULD-NOT;
1094   attempt to autocorrect and then process the request without a redirect,
1095   since the invalid request-line might be deliberately crafted to bypass
1096   security filters along the request chain.
1099   HTTP does not place a pre-defined limit on the length of a request-line.
1100   A server that receives a method longer than any that it implements
1101   &SHOULD; respond with either a <x:ref>405 (Method Not Allowed)</x:ref>, if it is an origin
1102   server, or a <x:ref>501 (Not Implemented)</x:ref> status code.
1103   A server &MUST; be prepared to receive URIs of unbounded length and
1104   respond with the <x:ref>414 (URI Too Long)</x:ref> status code if the received
1105   request-target would be longer than the server wishes to handle
1106   (see &status-414;).
1109   Various ad-hoc limitations on request-line length are found in practice.
1110   It is &RECOMMENDED; that all HTTP senders and recipients support, at a
1111   minimum, request-line lengths of up to 8000 octets.
1115<section title="Status Line" anchor="status.line">
1116  <x:anchor-alias value="response"/>
1117  <x:anchor-alias value="status-line"/>
1118  <x:anchor-alias value="status-code"/>
1119  <x:anchor-alias value="reason-phrase"/>
1121   The first line of a response message is the status-line, consisting
1122   of the protocol version, a space (SP), the status code, another space,
1123   a possibly-empty textual phrase describing the status code, and
1124   ending with CRLF.
1126<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="status-line"/>
1127  <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>
1130   A client &MUST; be able to parse any received message that begins
1131   with a status-line and matches the ABNF rule for HTTP-message.
1134   The status-code element is a 3-digit integer code describing the
1135   result of the server's attempt to understand and satisfy the client's
1136   corresponding request. The rest of the response message is to be
1137   interpreted in light of the semantics defined for that status code.
1138   See &status-codes; for information about the semantics of status codes,
1139   including the classes of status code (indicated by the first digit),
1140   the status codes defined by this specification, considerations for the
1141   definition of new status codes, and the IANA registry.
1143<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="status-code"/>
1144  <x:ref>status-code</x:ref>    = 3<x:ref>DIGIT</x:ref>
1147   The reason-phrase element exists for the sole purpose of providing a
1148   textual description associated with the numeric status code, mostly
1149   out of deference to earlier Internet application protocols that were more
1150   frequently used with interactive text clients. A client &SHOULD; ignore
1151   the reason-phrase content.
1153<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="reason-phrase"/>
1154  <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> )
1159<section title="Header Fields" anchor="header.fields">
1160  <x:anchor-alias value="header-field"/>
1161  <x:anchor-alias value="field-content"/>
1162  <x:anchor-alias value="field-name"/>
1163  <x:anchor-alias value="field-value"/>
1164  <x:anchor-alias value="obs-fold"/>
1166   Each HTTP header field consists of a case-insensitive field name
1167   followed by a colon (":"), optional whitespace, and the field value.
1169<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"/>
1170  <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>
1171  <x:ref>field-name</x:ref>     = <x:ref>token</x:ref>
1172  <x:ref>field-value</x:ref>    = *( <x:ref>field-content</x:ref> / <x:ref>obs-fold</x:ref> )
1173  <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> )
1174  <x:ref>obs-fold</x:ref>       = <x:ref>CRLF</x:ref> ( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1175                 ; obsolete line folding
1176                 ; see <xref target="field.parsing"/>
1179   The field-name token labels the corresponding field-value as having the
1180   semantics defined by that header field.  For example, the <x:ref>Date</x:ref>
1181   header field is defined in &header-date; as containing the origination
1182   timestamp for the message in which it appears.
1185   HTTP header fields are fully extensible: there is no limit on the
1186   introduction of new field names, each presumably defining new semantics,
1187   or on the number of header fields used in a given message.  Existing
1188   fields are defined in each part of this specification and in many other
1189   specifications outside the standards process.
1190   New header fields can be introduced without changing the protocol version
1191   if their defined semantics allow them to be safely ignored by recipients
1192   that do not recognize them.
1195   New HTTP header fields &SHOULD; be registered with IANA in the
1196   Message Header Field Registry, as described in &iana-header-registry;.
1197   Unrecognized header fields &MUST; be forwarded by a proxy unless the
1198   field-name is listed in the <x:ref>Connection</x:ref> header field
1199   (<xref target="header.connection"/>) or the proxy is specifically
1200   configured to block or otherwise transform such fields.
1201   Unrecognized header fields &SHOULD; be ignored by other recipients.
1204   The order in which header fields with differing field names are
1205   received is not significant. However, it is "good practice" to send
1206   header fields that contain control data first, such as <x:ref>Host</x:ref>
1207   on requests and <x:ref>Date</x:ref> on responses, so that implementations
1208   can decide when not to handle a message as early as possible.  A server
1209   &MUST; wait until the entire header section is received before interpreting
1210   a request message, since later header fields might include conditionals,
1211   authentication credentials, or deliberately misleading duplicate
1212   header fields that would impact request processing.
1215   Multiple header fields with the same field name &MUST-NOT; be
1216   sent in a message unless the entire field value for that
1217   header field is defined as a comma-separated list [i.e., #(values)].
1218   Multiple header fields with the same field name can be combined into
1219   one "field-name: field-value" pair, without changing the semantics of the
1220   message, by appending each subsequent field value to the combined
1221   field value in order, separated by a comma. The order in which
1222   header fields with the same field name are received is therefore
1223   significant to the interpretation of the combined field value;
1224   a proxy &MUST-NOT; change the order of these field values when
1225   forwarding a message.
1228  <t>
1229   &Note; The "Set-Cookie" header field as implemented in
1230   practice can occur multiple times, but does not use the list syntax, and
1231   thus cannot be combined into a single line (<xref target="RFC6265"/>). (See Appendix A.2.3 of <xref target="Kri2001"/>
1232   for details.) Also note that the Set-Cookie2 header field specified in
1233   <xref target="RFC2965"/> does not share this problem.
1234  </t>
1237<section title="Whitespace" anchor="whitespace">
1238<t anchor="rule.LWS">
1239   This specification uses three rules to denote the use of linear
1240   whitespace: OWS (optional whitespace), RWS (required whitespace), and
1241   BWS ("bad" whitespace).
1243<t anchor="rule.OWS">
1244   The OWS rule is used where zero or more linear whitespace octets might
1245   appear. OWS &SHOULD; either not be produced or be produced as a single
1246   SP. Multiple OWS octets that occur within field-content &SHOULD; either
1247   be replaced with a single SP or transformed to all SP octets (each
1248   octet other than SP replaced with SP) before interpreting the field value
1249   or forwarding the message downstream.
1251<t anchor="rule.RWS">
1252   RWS is used when at least one linear whitespace octet is required to
1253   separate field tokens. RWS &SHOULD; be produced as a single SP.
1254   Multiple RWS octets that occur within field-content &SHOULD; either
1255   be replaced with a single SP or transformed to all SP octets before
1256   interpreting the field value or forwarding the message downstream.
1258<t anchor="rule.BWS">
1259   BWS is used where the grammar allows optional whitespace, for historical
1260   reasons, but senders &SHOULD-NOT; produce it in messages;
1261   recipients &MUST; accept such bad optional whitespace and remove it before
1262   interpreting the field value or forwarding the message downstream.
1264<t anchor="rule.whitespace">
1265  <x:anchor-alias value="BWS"/>
1266  <x:anchor-alias value="OWS"/>
1267  <x:anchor-alias value="RWS"/>
1269<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"/>
1270  <x:ref>OWS</x:ref>            = *( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1271                 ; optional whitespace
1272  <x:ref>RWS</x:ref>            = 1*( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1273                 ; required whitespace
1274  <x:ref>BWS</x:ref>            = <x:ref>OWS</x:ref>
1275                 ; "bad" whitespace
1279<section title="Field Parsing" anchor="field.parsing">
1281   No whitespace is allowed between the header field-name and colon.
1282   In the past, differences in the handling of such whitespace have led to
1283   security vulnerabilities in request routing and response handling.
1284   Any received request message that contains whitespace between a header
1285   field-name and colon &MUST; be rejected with a response code of 400
1286   (Bad Request).  A proxy &MUST; remove any such whitespace from a response
1287   message before forwarding the message downstream.
1290   A field value is preceded by optional whitespace (OWS); a single SP is
1291   preferred. The field value does not include any leading or trailing white
1292   space: OWS occurring before the first non-whitespace octet of the
1293   field value or after the last non-whitespace octet of the field value
1294   is ignored and &SHOULD; be removed before further processing (as this does
1295   not change the meaning of the header field).
1298   Historically, HTTP header field values could be extended over multiple
1299   lines by preceding each extra line with at least one space or horizontal
1300   tab (obs-fold). This specification deprecates such line
1301   folding except within the message/http media type
1302   (<xref target=""/>).
1303   HTTP senders &MUST-NOT; produce messages that include line folding
1304   (i.e., that contain any field-value that matches the obs-fold rule) unless
1305   the message is intended for packaging within the message/http media type.
1306   HTTP recipients &SHOULD; accept line folding and replace any embedded
1307   obs-fold whitespace with either a single SP or a matching number of SP
1308   octets (to avoid buffer copying) prior to interpreting the field value or
1309   forwarding the message downstream.
1312   Historically, HTTP has allowed field content with text in the ISO-8859-1
1313   <xref target="ISO-8859-1"/> character encoding and supported other
1314   character sets only through use of <xref target="RFC2047"/> encoding.
1315   In practice, most HTTP header field values use only a subset of the
1316   US-ASCII character encoding <xref target="USASCII"/>. Newly defined
1317   header fields &SHOULD; limit their field values to US-ASCII octets.
1318   Recipients &SHOULD; treat other (obs-text) octets in field content as
1319   opaque data.
1323<section title="Field Length" anchor="field.length">
1325   HTTP does not place a pre-defined limit on the length of header fields,
1326   either in isolation or as a set. A server &MUST; be prepared to receive
1327   request header fields of unbounded length and respond with a <x:ref>4xx
1328   (Client Error)</x:ref> status code if the received header field(s) would be
1329   longer than the server wishes to handle.
1332   A client that receives response header fields that are longer than it wishes
1333   to handle can only treat it as a server error.
1336   Various ad-hoc limitations on header field length are found in practice. It
1337   is &RECOMMENDED; that all HTTP senders and recipients support messages whose
1338   combined header fields have 4000 or more octets.
1342<section title="Field value components" anchor="field.components">
1343<t anchor="rule.token.separators">
1344  <x:anchor-alias value="tchar"/>
1345  <x:anchor-alias value="token"/>
1346  <x:anchor-alias value="special"/>
1347  <x:anchor-alias value="word"/>
1348   Many HTTP header field values consist of words (token or quoted-string)
1349   separated by whitespace or special characters. These special characters
1350   &MUST; be in a quoted string to be used within a parameter value (as defined
1351   in <xref target="transfer.codings"/>).
1353<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"><!--unused production--></iref>
1354  <x:ref>word</x:ref>           = <x:ref>token</x:ref> / <x:ref>quoted-string</x:ref>
1356  <x:ref>token</x:ref>          = 1*<x:ref>tchar</x:ref>
1358  IMPORTANT: when editing "tchar" make sure that "special" is updated accordingly!!!
1359 -->
1360  <x:ref>tchar</x:ref>          = "!" / "#" / "$" / "%" / "&amp;" / "'" / "*"
1361                 / "+" / "-" / "." / "^" / "_" / "`" / "|" / "~"
1362                 / <x:ref>DIGIT</x:ref> / <x:ref>ALPHA</x:ref>
1363                 ; any <x:ref>VCHAR</x:ref>, except <x:ref>special</x:ref>
1365  <x:ref>special</x:ref>        = "(" / ")" / "&lt;" / ">" / "@" / ","
1366                 / ";" / ":" / "\" / DQUOTE / "/" / "["
1367                 / "]" / "?" / "=" / "{" / "}"
1369<t anchor="rule.quoted-string">
1370  <x:anchor-alias value="quoted-string"/>
1371  <x:anchor-alias value="qdtext"/>
1372  <x:anchor-alias value="obs-text"/>
1373   A string of text is parsed as a single word if it is quoted using
1374   double-quote marks.
1376<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"/>
1377  <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>
1378  <x:ref>qdtext</x:ref>         = <x:ref>HTAB</x:ref> / <x:ref>SP</x:ref> /%x21 / %x23-5B / %x5D-7E / <x:ref>obs-text</x:ref>
1379  <x:ref>obs-text</x:ref>       = %x80-FF
1381<t anchor="rule.quoted-pair">
1382  <x:anchor-alias value="quoted-pair"/>
1383   The backslash octet ("\") can be used as a single-octet
1384   quoting mechanism within quoted-string constructs:
1386<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-pair"/>
1387  <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> )
1390   Recipients that process the value of the quoted-string &MUST; handle a
1391   quoted-pair as if it were replaced by the octet following the backslash.
1394   Senders &SHOULD-NOT; escape octets in quoted-strings that do not require
1395   escaping (i.e., other than DQUOTE and the backslash octet).
1397<t anchor="rule.comment">
1398  <x:anchor-alias value="comment"/>
1399  <x:anchor-alias value="ctext"/>
1400   Comments can be included in some HTTP header fields by surrounding
1401   the comment text with parentheses. Comments are only allowed in
1402   fields containing "comment" as part of their field value definition.
1404<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="comment"/><iref primary="true" item="Grammar" subitem="ctext"/>
1405  <x:ref>comment</x:ref>        = "(" *( <x:ref>ctext</x:ref> / <x:ref>quoted-cpair</x:ref> / <x:ref>comment</x:ref> ) ")"
1406  <x:ref>ctext</x:ref>          = <x:ref>HTAB</x:ref> / <x:ref>SP</x:ref> / %x21-27 / %x2A-5B / %x5D-7E / <x:ref>obs-text</x:ref>
1408<t anchor="rule.quoted-cpair">
1409  <x:anchor-alias value="quoted-cpair"/>
1410   The backslash octet ("\") can be used as a single-octet
1411   quoting mechanism within comment constructs:
1413<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-cpair"/>
1414  <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> )
1417   Senders &SHOULD-NOT; escape octets in comments that do not require escaping
1418   (i.e., other than the backslash octet "\" and the parentheses "(" and ")").
1424<section title="Message Body" anchor="message.body">
1425  <x:anchor-alias value="message-body"/>
1427   The message body (if any) of an HTTP message is used to carry the
1428   payload body of that request or response.  The message body is
1429   identical to the payload body unless a transfer coding has been
1430   applied, as described in <xref target="header.transfer-encoding"/>.
1432<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="message-body"/>
1433  <x:ref>message-body</x:ref> = *OCTET
1436   The rules for when a message body is allowed in a message differ for
1437   requests and responses.
1440   The presence of a message body in a request is signaled by a
1441   <x:ref>Content-Length</x:ref> or <x:ref>Transfer-Encoding</x:ref> header
1442   field. Request message framing is independent of method semantics,
1443   even if the method does not define any use for a message body.
1446   The presence of a message body in a response depends on both
1447   the request method to which it is responding and the response
1448   status code (<xref target="status.line"/>).
1449   Responses to the HEAD request method never include a message body
1450   because the associated response header fields (e.g.,
1451   <x:ref>Transfer-Encoding</x:ref>, <x:ref>Content-Length</x:ref>, etc.),
1452   if present, indicate only what their values would have been if the request
1453   method had been GET (&HEAD;).
1454   <x:ref>2xx (Successful)</x:ref> responses to CONNECT switch to tunnel
1455   mode instead of having a message body (&CONNECT;).
1456   All <x:ref>1xx (Informational)</x:ref>, <x:ref>204 (No Content)</x:ref>, and
1457   <x:ref>304 (Not Modified)</x:ref> responses &MUST-NOT; include a message body.
1458   All other responses do include a message body, although the body
1459   &MAY; be of zero length.
1462<section title="Transfer-Encoding" anchor="header.transfer-encoding">
1463  <iref primary="true" item="Transfer-Encoding header field" x:for-anchor=""/>
1464  <x:anchor-alias value="Transfer-Encoding"/>
1466   When one or more transfer codings are applied to a payload body in order
1467   to form the message body, a Transfer-Encoding header field &MUST; be sent
1468   in the message and &MUST; contain the list of corresponding
1469   transfer-coding names in the same order that they were applied.
1470   Transfer codings are defined in <xref target="transfer.codings"/>.
1472<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Transfer-Encoding"/>
1473  <x:ref>Transfer-Encoding</x:ref> = 1#<x:ref>transfer-coding</x:ref>
1476   Transfer-Encoding is analogous to the Content-Transfer-Encoding field of
1477   MIME, which was designed to enable safe transport of binary data over a
1478   7-bit transport service (<xref target="RFC2045" x:fmt="," x:sec="6"/>).
1479   However, safe transport has a different focus for an 8bit-clean transfer
1480   protocol. In HTTP's case, Transfer-Encoding is primarily intended to
1481   accurately delimit a dynamically generated payload and to distinguish
1482   payload encodings that are only applied for transport efficiency or
1483   security from those that are characteristics of the target resource.
1486   The "chunked" transfer-coding (<xref target="chunked.encoding"/>)
1487   &MUST; be implemented by all HTTP/1.1 recipients because it plays a
1488   crucial role in delimiting messages when the payload body size is not
1489   known in advance.
1490   When the "chunked" transfer-coding is used, it &MUST; be the last
1491   transfer-coding applied to form the message body and &MUST-NOT;
1492   be applied more than once in a message body.
1493   If any transfer-coding is applied to a request payload body,
1494   the final transfer-coding applied &MUST; be "chunked".
1495   If any transfer-coding is applied to a response payload body, then either
1496   the final transfer-coding applied &MUST; be "chunked" or
1497   the message &MUST; be terminated by closing the connection.
1500   For example,
1501</preamble><artwork type="example">
1502  Transfer-Encoding: gzip, chunked
1504   indicates that the payload body has been compressed using the gzip
1505   coding and then chunked using the chunked coding while forming the
1506   message body.
1509   If more than one Transfer-Encoding header field is present in a message,
1510   the multiple field-values &MUST; be combined into one field-value,
1511   according to the algorithm defined in <xref target="header.fields"/>,
1512   before determining the message body length.
1515   Unlike <x:ref>Content-Encoding</x:ref> (&content-codings;),
1516   Transfer-Encoding is a property of the message, not of the payload, and thus
1517   &MAY; be added or removed by any implementation along the request/response
1518   chain. Additional information about the encoding parameters &MAY; be
1519   provided by other header fields not defined by this specification.
1522   Transfer-Encoding &MAY; be sent in a response to a HEAD request or in a
1523   <x:ref>304 (Not Modified)</x:ref> response (&status-304;) to a GET request,
1524   neither of which includes a message body,
1525   to indicate that the origin server would have applied a transfer coding
1526   to the message body if the request had been an unconditional GET.
1527   This indication is not required, however, because any recipient on
1528   the response chain (including the origin server) can remove transfer
1529   codings when they are not needed.
1532   Transfer-Encoding was added in HTTP/1.1.  It is generally assumed that
1533   implementations advertising only HTTP/1.0 support will not understand
1534   how to process a transfer-encoded payload.
1535   A client &MUST-NOT; send a request containing Transfer-Encoding unless it
1536   knows the server will handle HTTP/1.1 (or later) requests; such knowledge
1537   might be in the form of specific user configuration or by remembering the
1538   version of a prior received response.
1539   A server &MUST-NOT; send a response containing Transfer-Encoding unless
1540   the corresponding request indicates HTTP/1.1 (or later).
1543   A server that receives a request message with a transfer-coding it does
1544   not understand &SHOULD; respond with <x:ref>501 (Not Implemented)</x:ref> and then
1545   close the connection.
1549<section title="Content-Length" anchor="header.content-length">
1550  <iref primary="true" item="Content-Length header field" x:for-anchor=""/>
1551  <x:anchor-alias value="Content-Length"/>
1553   When a message is allowed to contain a message body, does not have a
1554   <x:ref>Transfer-Encoding</x:ref> header field, and has a payload body
1555   length that is known to the sender before the message header section has
1556   been sent, the sender &SHOULD; send a Content-Length header field to
1557   indicate the length of the payload body as a decimal number of octets.
1559<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Content-Length"/>
1560  <x:ref>Content-Length</x:ref> = 1*<x:ref>DIGIT</x:ref>
1563   An example is
1565<figure><artwork type="example">
1566  Content-Length: 3495
1569   A sender &MUST-NOT; send a Content-Length header field in any message that
1570   contains a <x:ref>Transfer-Encoding</x:ref> header field.
1573   A server &MAY; send a Content-Length header field in a response to a HEAD
1574   request (&HEAD;); a server &MUST-NOT; send Content-Length in such a
1575   response unless its field-value equals the decimal number of octets that
1576   would have been sent in the payload body of a response if the same
1577   request had used the GET method.
1580   A server &MAY; send a Content-Length header field in a
1581   <x:ref>304 (Not Modified)</x:ref> response to a conditional GET request
1582   (&status-304;); a server &MUST-NOT; send Content-Length in such a
1583   response unless its field-value equals the decimal number of octets that
1584   would have been sent in the payload body of a <x:ref>200 (OK)</x:ref>
1585   response to the same request.
1588   A server &MUST-NOT; send a Content-Length header field in any response
1589   with a status code of
1590   <x:ref>1xx (Informational)</x:ref> or <x:ref>204 (No Content)</x:ref>.
1591   A server &SHOULD-NOT; send a Content-Length header field in any
1592   <x:ref>2xx (Successful)</x:ref> response to a CONNECT request (&CONNECT;).
1595   Any Content-Length field value greater than or equal to zero is valid.
1596   Since there is no predefined limit to the length of an HTTP payload,
1597   recipients &SHOULD; anticipate potentially large decimal numerals and
1598   prevent parsing errors due to integer conversion overflows
1599   (<xref target="attack.protocol.element.size.overflows"/>).
1602   If a message is received that has multiple Content-Length header fields
1603   with field-values consisting of the same decimal value, or a single
1604   Content-Length header field with a field value containing a list of
1605   identical decimal values (e.g., "Content-Length: 42, 42"), indicating that
1606   duplicate Content-Length header fields have been generated or combined by an
1607   upstream message processor, then the recipient &MUST; either reject the
1608   message as invalid or replace the duplicated field-values with a single
1609   valid Content-Length field containing that decimal value prior to
1610   determining the message body length.
1613  <t>
1614   &Note; HTTP's use of Content-Length for message framing differs
1615   significantly from the same field's use in MIME, where it is an optional
1616   field used only within the "message/external-body" media-type.
1617  </t>
1621<section title="Message Body Length" anchor="message.body.length">
1623   The length of a message body is determined by one of the following
1624   (in order of precedence):
1627  <list style="numbers">
1628    <x:lt><t>
1629     Any response to a HEAD request and any response with a
1630     <x:ref>1xx (Informational)</x:ref>, <x:ref>204 (No Content)</x:ref>, or
1631     <x:ref>304 (Not Modified)</x:ref> status code is always
1632     terminated by the first empty line after the header fields, regardless of
1633     the header fields present in the message, and thus cannot contain a
1634     message body.
1635    </t></x:lt>
1636    <x:lt><t>
1637     Any <x:ref>2xx (Successful)</x:ref> response to a CONNECT request implies that the
1638     connection will become a tunnel immediately after the empty line that
1639     concludes the header fields.  A client &MUST; ignore any
1640     <x:ref>Content-Length</x:ref> or <x:ref>Transfer-Encoding</x:ref> header
1641     fields received in such a message.
1642    </t></x:lt>
1643    <x:lt><t>
1644     If a <x:ref>Transfer-Encoding</x:ref> header field is present
1645     and the "chunked" transfer-coding (<xref target="chunked.encoding"/>)
1646     is the final encoding, the message body length is determined by reading
1647     and decoding the chunked data until the transfer-coding indicates the
1648     data is complete.
1649    </t>
1650    <t>
1651     If a <x:ref>Transfer-Encoding</x:ref> header field is present in a
1652     response and the "chunked" transfer-coding is not the final encoding, the
1653     message body length is determined by reading the connection until it is
1654     closed by the server.
1655     If a Transfer-Encoding header field is present in a request and the
1656     "chunked" transfer-coding is not the final encoding, the message body
1657     length cannot be determined reliably; the server &MUST; respond with
1658     the <x:ref>400 (Bad Request)</x:ref> status code and then close the connection.
1659    </t>
1660    <t>
1661     If a message is received with both a <x:ref>Transfer-Encoding</x:ref>
1662     and a <x:ref>Content-Length</x:ref> header field, the
1663     Transfer-Encoding overrides the Content-Length.
1664     Such a message might indicate an attempt to perform request or response
1665     smuggling (bypass of security-related checks on message routing or content)
1666     and thus ought to be handled as an error.  The provided Content-Length &MUST;
1667     be removed, prior to forwarding the message downstream, or replaced with
1668     the real message body length after the transfer-coding is decoded.
1669    </t></x:lt>
1670    <x:lt><t>
1671     If a message is received without <x:ref>Transfer-Encoding</x:ref> and with
1672     either multiple <x:ref>Content-Length</x:ref> header fields having
1673     differing field-values or a single Content-Length header field having an
1674     invalid value, then the message framing is invalid and &MUST; be treated
1675     as an error to prevent request or response smuggling.
1676     If this is a request message, the server &MUST; respond with
1677     a <x:ref>400 (Bad Request)</x:ref> status code and then close the connection.
1678     If this is a response message received by a proxy, the proxy
1679     &MUST; discard the received response, send a <x:ref>502 (Bad Gateway)</x:ref>
1680     status code as its downstream response, and then close the connection.
1681     If this is a response message received by a user agent, it &MUST; be
1682     treated as an error by discarding the message and closing the connection.
1683    </t></x:lt>
1684    <x:lt><t>
1685     If a valid <x:ref>Content-Length</x:ref> header field is present without
1686     <x:ref>Transfer-Encoding</x:ref>, its decimal value defines the
1687     message body length in octets.  If the actual number of octets sent in
1688     the message is less than the indicated Content-Length, the recipient
1689     &MUST; consider the message to be incomplete and treat the connection
1690     as no longer usable.
1691     If the actual number of octets sent in the message is more than the indicated
1692     Content-Length, the recipient &MUST; only process the message body up to the
1693     field value's number of octets; the remainder of the message &MUST; either
1694     be discarded or treated as the next message in a pipeline.  For the sake of
1695     robustness, a user agent &MAY; attempt to detect and correct such an error
1696     in message framing if it is parsing the response to the last request on
1697     a connection and the connection has been closed by the server.
1698    </t></x:lt>
1699    <x:lt><t>
1700     If this is a request message and none of the above are true, then the
1701     message body length is zero (no message body is present).
1702    </t></x:lt>
1703    <x:lt><t>
1704     Otherwise, this is a response message without a declared message body
1705     length, so the message body length is determined by the number of octets
1706     received prior to the server closing the connection.
1707    </t></x:lt>
1708  </list>
1711   Since there is no way to distinguish a successfully completed,
1712   close-delimited message from a partially-received message interrupted
1713   by network failure, a server &SHOULD; use encoding or
1714   length-delimited messages whenever possible.  The close-delimiting
1715   feature exists primarily for backwards compatibility with HTTP/1.0.
1718   A server &MAY; reject a request that contains a message body but
1719   not a <x:ref>Content-Length</x:ref> by responding with
1720   <x:ref>411 (Length Required)</x:ref>.
1723   Unless a transfer-coding other than "chunked" has been applied,
1724   a client that sends a request containing a message body &SHOULD;
1725   use a valid <x:ref>Content-Length</x:ref> header field if the message body
1726   length is known in advance, rather than the "chunked" encoding, since some
1727   existing services respond to "chunked" with a <x:ref>411 (Length Required)</x:ref>
1728   status code even though they understand the chunked encoding.  This
1729   is typically because such services are implemented via a gateway that
1730   requires a content-length in advance of being called and the server
1731   is unable or unwilling to buffer the entire request before processing.
1734   A client that sends a request containing a message body &MUST; include a
1735   valid <x:ref>Content-Length</x:ref> header field if it does not know the
1736   server will handle HTTP/1.1 (or later) requests; such knowledge can be in
1737   the form of specific user configuration or by remembering the version of a
1738   prior received response.
1743<section anchor="incomplete.messages" title="Handling Incomplete Messages">
1745   Request messages that are prematurely terminated, possibly due to a
1746   canceled connection or a server-imposed time-out exception, &MUST;
1747   result in closure of the connection; sending an error response
1748   prior to closing the connection is &OPTIONAL;.
1751   Response messages that are prematurely terminated, usually by closure
1752   of the connection prior to receiving the expected number of octets or by
1753   failure to decode a transfer-encoded message body, &MUST; be recorded
1754   as incomplete.  A response that terminates in the middle of the header
1755   block (before the empty line is received) cannot be assumed to convey the
1756   full semantics of the response and &MUST; be treated as an error.
1759   A message body that uses the chunked transfer encoding is
1760   incomplete if the zero-sized chunk that terminates the encoding has not
1761   been received.  A message that uses a valid <x:ref>Content-Length</x:ref> is
1762   incomplete if the size of the message body received (in octets) is less than
1763   the value given by Content-Length.  A response that has neither chunked
1764   transfer encoding nor Content-Length is terminated by closure of the
1765   connection, and thus is considered complete regardless of the number of
1766   message body octets received, provided that the header block was received
1767   intact.
1770   A user agent &MUST-NOT; render an incomplete response message body as if
1771   it were complete (i.e., some indication needs to be given to the user that an
1772   error occurred).  Cache requirements for incomplete responses are defined
1773   in &cache-incomplete;.
1776   A server &MUST; read the entire request message body or close
1777   the connection after sending its response, since otherwise the
1778   remaining data on a persistent connection would be misinterpreted
1779   as the next request.  Likewise,
1780   a client &MUST; read the entire response message body if it intends
1781   to reuse the same connection for a subsequent request.  Pipelining
1782   multiple requests on a connection is described in <xref target="pipelining"/>.
1786<section title="Message Parsing Robustness" anchor="message.robustness">
1788   Older HTTP/1.0 client implementations might send an extra CRLF
1789   after a POST request as a lame workaround for some early server
1790   applications that failed to read message body content that was
1791   not terminated by a line-ending. An HTTP/1.1 client &MUST-NOT;
1792   preface or follow a request with an extra CRLF.  If terminating
1793   the request message body with a line-ending is desired, then the
1794   client &MUST; include the terminating CRLF octets as part of the
1795   message body length.
1798   In the interest of robustness, servers &SHOULD; ignore at least one
1799   empty line received where a request-line is expected. In other words, if
1800   the server is reading the protocol stream at the beginning of a
1801   message and receives a CRLF first, it &SHOULD; ignore the CRLF.
1802   Likewise, although the line terminator for the start-line and header
1803   fields is the sequence CRLF, we recommend that recipients recognize a
1804   single LF as a line terminator and ignore any CR.
1807   When a server listening only for HTTP request messages, or processing
1808   what appears from the start-line to be an HTTP request message,
1809   receives a sequence of octets that does not match the HTTP-message
1810   grammar aside from the robustness exceptions listed above, the
1811   server &MUST; respond with an HTTP/1.1 <x:ref>400 (Bad Request)</x:ref> response. 
1816<section title="Transfer Codings" anchor="transfer.codings">
1817  <x:anchor-alias value="transfer-coding"/>
1818  <x:anchor-alias value="transfer-extension"/>
1820   Transfer-coding values are used to indicate an encoding
1821   transformation that has been, can be, or might need to be applied to a
1822   payload body in order to ensure "safe transport" through the network.
1823   This differs from a content coding in that the transfer-coding is a
1824   property of the message rather than a property of the representation
1825   that is being transferred.
1827<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="transfer-coding"/><iref primary="true" item="Grammar" subitem="transfer-extension"/>
1828  <x:ref>transfer-coding</x:ref>    = "chunked" ; <xref target="chunked.encoding"/>
1829                     / "compress" ; <xref target="compress.coding"/>
1830                     / "deflate" ; <xref target="deflate.coding"/>
1831                     / "gzip" ; <xref target="gzip.coding"/>
1832                     / <x:ref>transfer-extension</x:ref>
1833  <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> )
1835<t anchor="rule.parameter">
1836  <x:anchor-alias value="attribute"/>
1837  <x:anchor-alias value="transfer-parameter"/>
1838  <x:anchor-alias value="value"/>
1839   Parameters are in the form of attribute/value pairs.
1841<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"/>
1842  <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>
1843  <x:ref>attribute</x:ref>          = <x:ref>token</x:ref>
1844  <x:ref>value</x:ref>              = <x:ref>word</x:ref>
1847   All transfer-coding values are case-insensitive and &SHOULD; be registered
1848   within the HTTP Transfer Coding registry, as defined in
1849   <xref target="transfer.coding.registry"/>.
1850   They are used in the <x:ref>TE</x:ref> (<xref target="header.te"/>) and
1851   <x:ref>Transfer-Encoding</x:ref> (<xref target="header.transfer-encoding"/>)
1852   header fields.
1855<section title="Chunked Transfer Coding" anchor="chunked.encoding">
1856  <iref item="chunked (Coding Format)"/>
1857  <x:anchor-alias value="chunk"/>
1858  <x:anchor-alias value="chunked-body"/>
1859  <x:anchor-alias value="chunk-data"/>
1860  <x:anchor-alias value="chunk-ext"/>
1861  <x:anchor-alias value="chunk-ext-name"/>
1862  <x:anchor-alias value="chunk-ext-val"/>
1863  <x:anchor-alias value="chunk-size"/>
1864  <x:anchor-alias value="last-chunk"/>
1865  <x:anchor-alias value="trailer-part"/>
1866  <x:anchor-alias value="quoted-str-nf"/>
1867  <x:anchor-alias value="qdtext-nf"/>
1869   The chunked encoding modifies the body of a message in order to
1870   transfer it as a series of chunks, each with its own size indicator,
1871   followed by an &OPTIONAL; trailer containing header fields. This
1872   allows dynamically produced content to be transferred along with the
1873   information necessary for the recipient to verify that it has
1874   received the full message.
1876<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="chunked-body"><!--terminal production--></iref><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"/>
1877  <x:ref>chunked-body</x:ref>   = *<x:ref>chunk</x:ref>
1878                   <x:ref>last-chunk</x:ref>
1879                   <x:ref>trailer-part</x:ref>
1880                   <x:ref>CRLF</x:ref>
1882  <x:ref>chunk</x:ref>          = <x:ref>chunk-size</x:ref> [ <x:ref>chunk-ext</x:ref> ] <x:ref>CRLF</x:ref>
1883                   <x:ref>chunk-data</x:ref> <x:ref>CRLF</x:ref>
1884  <x:ref>chunk-size</x:ref>     = 1*<x:ref>HEXDIG</x:ref>
1885  <x:ref>last-chunk</x:ref>     = 1*("0") [ <x:ref>chunk-ext</x:ref> ] <x:ref>CRLF</x:ref>
1887  <x:ref>chunk-ext</x:ref>      = *( ";" <x:ref>chunk-ext-name</x:ref> [ "=" <x:ref>chunk-ext-val</x:ref> ] )
1888  <x:ref>chunk-ext-name</x:ref> = <x:ref>token</x:ref>
1889  <x:ref>chunk-ext-val</x:ref>  = <x:ref>token</x:ref> / <x:ref>quoted-str-nf</x:ref>
1890  <x:ref>chunk-data</x:ref>     = 1*<x:ref>OCTET</x:ref> ; a sequence of chunk-size octets
1891  <x:ref>trailer-part</x:ref>   = *( <x:ref>header-field</x:ref> <x:ref>CRLF</x:ref> )
1893  <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>
1894                 ; like <x:ref>quoted-string</x:ref>, but disallowing line folding
1895  <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>
1898   Chunk extensions within the chucked encoding are deprecated.
1899   Senders &SHOULD-NOT; send chunk-ext.
1900   Definition of new chunk extensions is discouraged.
1903   The chunk-size field is a string of hex digits indicating the size of
1904   the chunk-data in octets. The chunked encoding is ended by any chunk whose size is
1905   zero, followed by the trailer, which is terminated by an empty line.
1908<section title="Trailer" anchor="header.trailer">
1909  <iref primary="true" item="Trailer header field" x:for-anchor=""/>
1910  <x:anchor-alias value="Trailer"/>
1912   A trailer allows the sender to include additional fields at the end of a
1913   chunked message in order to supply metadata that might be dynamically
1914   generated while the message body is sent, such as a message integrity
1915   check, digital signature, or post-processing status.
1916   The trailer &MUST-NOT; contain fields that need to be known before a
1917   recipient processes the body, such as <x:ref>Transfer-Encoding</x:ref>,
1918   <x:ref>Content-Length</x:ref>, and <x:ref>Trailer</x:ref>.
1921   When a message includes a message body encoded with the chunked
1922   transfer-coding and the sender desires to send metadata in the form of
1923   trailer fields at the end of the message, the sender &SHOULD; send a
1924   <x:ref>Trailer</x:ref> header field before the message body to indicate
1925   which fields will be present in the trailers. This allows the recipient
1926   to prepare for receipt of that metadata before it starts processing the body,
1927   which is useful if the message is being streamed and the recipient wishes
1928   to confirm an integrity check on the fly.
1930<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Trailer"/>
1931  <x:ref>Trailer</x:ref> = 1#<x:ref>field-name</x:ref>
1934   If no <x:ref>Trailer</x:ref> header field is present, the sender of a
1935   chunked message body &SHOULD; send an empty trailer.
1938   A server &MUST; send an empty trailer with the chunked transfer-coding
1939   unless at least one of the following is true:
1940  <list style="numbers">
1941    <t>the request included a <x:ref>TE</x:ref> header field that indicates
1942    "trailers" is acceptable in the transfer-coding of the response, as
1943    described in <xref target="header.te"/>; or,</t>
1945    <t>the trailer fields consist entirely of optional metadata and the
1946    recipient could use the message (in a manner acceptable to the server where
1947    the field originated) without receiving that metadata. In other words,
1948    the server that generated the header field is willing to accept the
1949    possibility that the trailer fields might be silently discarded along
1950    the path to the client.</t>
1951  </list>
1954   The above requirement prevents the need for an infinite buffer when a
1955   message is being received by an HTTP/1.1 (or later) proxy and forwarded to
1956   an HTTP/1.0 recipient.
1960<section title="Decoding chunked" anchor="decoding.chunked">
1962   A process for decoding the "chunked" transfer-coding
1963   can be represented in pseudo-code as:
1965<figure><artwork type="code">
1966  length := 0
1967  read chunk-size, chunk-ext (if any) and CRLF
1968  while (chunk-size &gt; 0) {
1969     read chunk-data and CRLF
1970     append chunk-data to decoded-body
1971     length := length + chunk-size
1972     read chunk-size and CRLF
1973  }
1974  read header-field
1975  while (header-field not empty) {
1976     append header-field to existing header fields
1977     read header-field
1978  }
1979  Content-Length := length
1980  Remove "chunked" from Transfer-Encoding
1981  Remove Trailer from existing header fields
1984   All recipients &MUST; be able to receive and decode the
1985   "chunked" transfer-coding and &MUST; ignore chunk-ext extensions
1986   they do not understand.
1991<section title="Compression Codings" anchor="compression.codings">
1993   The codings defined below can be used to compress the payload of a
1994   message.
1997<section title="Compress Coding" anchor="compress.coding">
1998<iref item="compress (Coding Format)"/>
2000   The "compress" format is produced by the common UNIX file compression
2001   program "compress". This format is an adaptive Lempel-Ziv-Welch
2002   coding (LZW). Recipients &SHOULD; consider "x-compress" to be
2003   equivalent to "compress".
2007<section title="Deflate Coding" anchor="deflate.coding">
2008<iref item="deflate (Coding Format)"/>
2010   The "deflate" format is defined as the "deflate" compression mechanism
2011   (described in <xref target="RFC1951"/>) used inside the "zlib"
2012   data format (<xref target="RFC1950"/>).
2015  <t>
2016    &Note; Some incorrect implementations send the "deflate"
2017    compressed data without the zlib wrapper.
2018   </t>
2022<section title="Gzip Coding" anchor="gzip.coding">
2023<iref item="gzip (Coding Format)"/>
2025   The "gzip" format is produced by the file compression program
2026   "gzip" (GNU zip), as described in <xref target="RFC1952"/>. This format is a
2027   Lempel-Ziv coding (LZ77) with a 32 bit CRC.
2028   Recipients &SHOULD; consider "x-gzip" to be equivalent to "gzip".
2034<section title="TE" anchor="header.te">
2035  <iref primary="true" item="TE header field" x:for-anchor=""/>
2036  <x:anchor-alias value="TE"/>
2037  <x:anchor-alias value="t-codings"/>
2038  <x:anchor-alias value="t-ranking"/>
2039  <x:anchor-alias value="rank"/>
2041   The "TE" header field in a request indicates what transfer-codings,
2042   besides "chunked", the client is willing to accept in response, and
2043   whether or not the client is willing to accept trailer fields in a
2044   chunked transfer-coding.
2047   The TE field-value consists of a comma-separated list of transfer-coding
2048   names, each allowing for optional parameters (as described in
2049   <xref target="transfer.codings"/>), and/or the keyword "trailers".
2050   Clients &MUST-NOT; send the chunked transfer-coding name in TE;
2051   chunked is always acceptable for HTTP/1.1 recipients.
2053<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="t-ranking"/><iref primary="true" item="Grammar" subitem="rank"/>
2054  <x:ref>TE</x:ref>        = #<x:ref>t-codings</x:ref>
2055  <x:ref>t-codings</x:ref> = "trailers" / ( <x:ref>transfer-coding</x:ref> [ <x:ref>t-ranking</x:ref> ] )
2056  <x:ref>t-ranking</x:ref> = <x:ref>OWS</x:ref> ";" <x:ref>OWS</x:ref> "q=" <x:ref>rank</x:ref>
2057  <x:ref>rank</x:ref>      = ( "0" [ "." 0*3<x:ref>DIGIT</x:ref> ] )
2058             / ( "1" [ "." 0*3("0") ] )
2061   Three examples of TE use are below.
2063<figure><artwork type="example">
2064  TE: deflate
2065  TE:
2066  TE: trailers, deflate;q=0.5
2069   The presence of the keyword "trailers" indicates that the client is
2070   willing to accept trailer fields in a chunked transfer-coding,
2071   as defined in <xref target="chunked.encoding"/>, on behalf of itself and
2072   any downstream clients. For chained requests, this implies that either:
2073   (a) all downstream clients are willing to accept trailer fields in the
2074   forwarded response; or,
2075   (b) the client will attempt to buffer the response on behalf of downstream
2076   recipients.
2077   Note that HTTP/1.1 does not define any means to limit the size of a
2078   chunked response such that a client can be assured of buffering the
2079   entire response.
2082   When multiple transfer-codings are acceptable, the client &MAY; rank the
2083   codings by preference using a case-insensitive "q" parameter (similar to
2084   the qvalues used in content negotiation fields, &qvalue;). The rank value
2085   is a real number in the range 0 through 1, where 0.001 is the least
2086   preferred and 1 is the most preferred; a value of 0 means "not acceptable".
2089   If the TE field-value is empty or if no TE field is present, the only
2090   acceptable transfer-coding is "chunked". A message with no transfer-coding
2091   is always acceptable.
2094   Since the TE header field only applies to the immediate connection,
2095   a sender of TE &MUST; also send a "TE" connection option within the
2096   <x:ref>Connection</x:ref> header field (<xref target="header.connection"/>)
2097   in order to prevent the TE field from being forwarded by intermediaries
2098   that do not support its semantics.
2103<section title="Message Routing" anchor="message.routing">
2105   HTTP request message routing is determined by each client based on the
2106   target resource, the client's proxy configuration, and
2107   establishment or reuse of an inbound connection.  The corresponding
2108   response routing follows the same connection chain back to the client.
2111<section title="Identifying a Target Resource" anchor="target-resource">
2112  <iref primary="true" item="target resource"/>
2113  <iref primary="true" item="target URI"/>
2114  <x:anchor-alias value="target resource"/>
2115  <x:anchor-alias value="target URI"/>
2117   HTTP is used in a wide variety of applications, ranging from
2118   general-purpose computers to home appliances.  In some cases,
2119   communication options are hard-coded in a client's configuration.
2120   However, most HTTP clients rely on the same resource identification
2121   mechanism and configuration techniques as general-purpose Web browsers.
2124   HTTP communication is initiated by a user agent for some purpose.
2125   The purpose is a combination of request semantics, which are defined in
2126   <xref target="Part2"/>, and a target resource upon which to apply those
2127   semantics.  A URI reference (<xref target="uri"/>) is typically used as
2128   an identifier for the "<x:dfn>target resource</x:dfn>", which a user agent
2129   would resolve to its absolute form in order to obtain the
2130   "<x:dfn>target URI</x:dfn>".  The target URI
2131   excludes the reference's fragment identifier component, if any,
2132   since fragment identifiers are reserved for client-side processing
2133   (<xref target="RFC3986" x:fmt="," x:sec="3.5"/>).
2137<section title="Connecting Inbound" anchor="connecting.inbound">
2139   Once the target URI is determined, a client needs to decide whether
2140   a network request is necessary to accomplish the desired semantics and,
2141   if so, where that request is to be directed.
2144   If the client has a response cache and the request semantics can be
2145   satisfied by a cache (<xref target="Part6"/>), then the request is
2146   usually directed to the cache first.
2149   If the request is not satisfied by a cache, then a typical client will
2150   check its configuration to determine whether a proxy is to be used to
2151   satisfy the request.  Proxy configuration is implementation-dependent,
2152   but is often based on URI prefix matching, selective authority matching,
2153   or both, and the proxy itself is usually identified by an "http" or
2154   "https" URI.  If a proxy is applicable, the client connects inbound by
2155   establishing (or reusing) a connection to that proxy.
2158   If no proxy is applicable, a typical client will invoke a handler routine,
2159   usually specific to the target URI's scheme, to connect directly
2160   to an authority for the target resource.  How that is accomplished is
2161   dependent on the target URI scheme and defined by its associated
2162   specification, similar to how this specification defines origin server
2163   access for resolution of the "http" (<xref target="http.uri"/>) and
2164   "https" (<xref target="https.uri"/>) schemes.
2167   HTTP requirements regarding connection management are defined in
2168   <xref target=""/>.
2172<section title="Request Target" anchor="request-target">
2174   Once an inbound connection is obtained,
2175   the client sends an HTTP request message (<xref target="http.message"/>)
2176   with a request-target derived from the target URI.
2177   There are four distinct formats for the request-target, depending on both
2178   the method being requested and whether the request is to a proxy.
2180<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"/>
2181  <x:ref>request-target</x:ref> = <x:ref>origin-form</x:ref>
2182                 / <x:ref>absolute-form</x:ref>
2183                 / <x:ref>authority-form</x:ref>
2184                 / <x:ref>asterisk-form</x:ref>
2186  <x:ref>origin-form</x:ref>    = <x:ref>path-absolute</x:ref> [ "?" <x:ref>query</x:ref> ]
2187  <x:ref>absolute-form</x:ref>  = <x:ref>absolute-URI</x:ref>
2188  <x:ref>authority-form</x:ref> = <x:ref>authority</x:ref>
2189  <x:ref>asterisk-form</x:ref>  = "*"
2191<t anchor="origin-form"><iref item="origin-form (of request-target)"/>
2192   The most common form of request-target is the origin-form.
2193   When making a request directly to an origin server, other than a CONNECT
2194   or server-wide OPTIONS request (as detailed below),
2195   a client &MUST; send only the absolute path and query components of
2196   the target URI as the request-target.
2197   If the target URI's path component is empty, then the client &MUST; send
2198   "/" as the path within the origin-form of request-target.
2199   A <x:ref>Host</x:ref> header field is also sent, as defined in
2200   <xref target=""/>, containing the target URI's
2201   authority component (excluding any userinfo).
2204   For example, a client wishing to retrieve a representation of the resource
2205   identified as
2207<figure><artwork x:indent-with="  " type="example">
2211   directly from the origin server would open (or reuse) a TCP connection
2212   to port 80 of the host "" and send the lines:
2214<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2215GET /where?q=now HTTP/1.1
2219   followed by the remainder of the request message.
2221<t anchor="absolute-form"><iref item="absolute-form (of request-target)"/>
2222   When making a request to a proxy, other than a CONNECT or server-wide
2223   OPTIONS request (as detailed below), a client &MUST; send the target URI
2224   in absolute-form as the request-target.
2225   The proxy is requested to either service that request from a valid cache,
2226   if possible, or make the same request on the client's behalf to either
2227   the next inbound proxy server or directly to the origin server indicated
2228   by the request-target.  Requirements on such "forwarding" of messages are
2229   defined in <xref target="message.forwarding"/>.
2232   An example absolute-form of request-line would be:
2234<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2235GET HTTP/1.1
2238   To allow for transition to the absolute-form for all requests in some
2239   future version of HTTP, HTTP/1.1 servers &MUST; accept the absolute-form
2240   in requests, even though HTTP/1.1 clients will only send them in requests
2241   to proxies.
2243<t anchor="authority-form"><iref item="authority-form (of request-target)"/>
2244   The authority-form of request-target is only used for CONNECT requests
2245   (&CONNECT;).  When making a CONNECT request to establish a tunnel through
2246   one or more proxies, a client &MUST; send only the target URI's
2247   authority component (excluding any userinfo) as the request-target.
2248   For example,
2250<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2253<t anchor="asterisk-form"><iref item="asterisk-form (of request-target)"/>
2254   The asterisk-form of request-target is only used for a server-wide
2255   OPTIONS request (&OPTIONS;).  When a client wishes to request OPTIONS
2256   for the server as a whole, as opposed to a specific named resource of
2257   that server, the client &MUST; send only "*" (%x2A) as the request-target.
2258   For example,
2260<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2261OPTIONS * HTTP/1.1
2264   If a proxy receives an OPTIONS request with an absolute-form of
2265   request-target in which the URI has an empty path and no query component,
2266   then the last proxy on the request chain &MUST; send a request-target
2267   of "*" when it forwards the request to the indicated origin server.
2270   For example, the request
2271</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2275  would be forwarded by the final proxy as
2276</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2277OPTIONS * HTTP/1.1
2281   after connecting to port 8001 of host "".
2286<section title="Host" anchor="">
2287  <iref primary="true" item="Host header field" x:for-anchor=""/>
2288  <x:anchor-alias value="Host"/>
2290   The "Host" header field in a request provides the host and port
2291   information from the target URI, enabling the origin
2292   server to distinguish among resources while servicing requests
2293   for multiple host names on a single IP address.  Since the Host
2294   field-value is critical information for handling a request, it
2295   &SHOULD; be sent as the first header field following the request-line.
2297<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Host"/>
2298  <x:ref>Host</x:ref> = <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ; <xref target="http.uri"/>
2301   A client &MUST; send a Host header field in all HTTP/1.1 request
2302   messages.  If the target URI includes an authority component, then
2303   the Host field-value &MUST; be identical to that authority component
2304   after excluding any userinfo (<xref target="http.uri"/>).
2305   If the authority component is missing or undefined for the target URI,
2306   then the Host header field &MUST; be sent with an empty field-value.
2309   For example, a GET request to the origin server for
2310   &lt;; would begin with:
2312<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2313GET /pub/WWW/ HTTP/1.1
2317   The Host header field &MUST; be sent in an HTTP/1.1 request even
2318   if the request-target is in the absolute-form, since this
2319   allows the Host information to be forwarded through ancient HTTP/1.0
2320   proxies that might not have implemented Host.
2323   When a proxy receives a request with an absolute-form of
2324   request-target, the proxy &MUST; ignore the received
2325   Host header field (if any) and instead replace it with the host
2326   information of the request-target.  If the proxy forwards the request,
2327   it &MUST; generate a new Host field-value based on the received
2328   request-target rather than forward the received Host field-value.
2331   Since the Host header field acts as an application-level routing
2332   mechanism, it is a frequent target for malware seeking to poison
2333   a shared cache or redirect a request to an unintended server.
2334   An interception proxy is particularly vulnerable if it relies on
2335   the Host field-value for redirecting requests to internal
2336   servers, or for use as a cache key in a shared cache, without
2337   first verifying that the intercepted connection is targeting a
2338   valid IP address for that host.
2341   A server &MUST; respond with a <x:ref>400 (Bad Request)</x:ref> status code
2342   to any HTTP/1.1 request message that lacks a Host header field and
2343   to any request message that contains more than one Host header field
2344   or a Host header field with an invalid field-value.
2348<section title="Effective Request URI" anchor="effective.request.uri">
2349  <iref primary="true" item="effective request URI"/>
2351   A server that receives an HTTP request message &MUST; reconstruct
2352   the user agent's original target URI, based on the pieces of information
2353   learned from the request-target, <x:ref>Host</x:ref> header field, and
2354   connection context, in order to identify the intended target resource and
2355   properly service the request. The URI derived from this reconstruction
2356   process is referred to as the "<x:dfn>effective request URI</x:dfn>".
2359   For a user agent, the effective request URI is the target URI.
2362   If the request-target is in absolute-form, then the effective request URI
2363   is the same as the request-target.  Otherwise, the effective request URI
2364   is constructed as follows.
2367   If the request is received over a TLS-secured TCP connection,
2368   then the effective request URI's scheme is "https"; otherwise, the
2369   scheme is "http".
2372   If the request-target is in authority-form, then the effective
2373   request URI's authority component is the same as the request-target.
2374   Otherwise, if a <x:ref>Host</x:ref> header field is supplied with a
2375   non-empty field-value, then the authority component is the same as the
2376   Host field-value. Otherwise, the authority component is the concatenation of
2377   the default host name configured for the server, a colon (":"), and the
2378   connection's incoming TCP port number in decimal form.
2381   If the request-target is in authority-form or asterisk-form, then the
2382   effective request URI's combined path and query component is empty.
2383   Otherwise, the combined path and query component is the same as the
2384   request-target.
2387   The components of the effective request URI, once determined as above,
2388   can be combined into absolute-URI form by concatenating the scheme,
2389   "://", authority, and combined path and query component.
2393   Example 1: the following message received over an insecure TCP connection
2395<artwork type="example" x:indent-with="  ">
2396GET /pub/WWW/TheProject.html HTTP/1.1
2402  has an effective request URI of
2404<artwork type="example" x:indent-with="  ">
2410   Example 2: the following message received over a TLS-secured TCP connection
2412<artwork type="example" x:indent-with="  ">
2413OPTIONS * HTTP/1.1
2419  has an effective request URI of
2421<artwork type="example" x:indent-with="  ">
2426   An origin server that does not allow resources to differ by requested
2427   host &MAY; ignore the <x:ref>Host</x:ref> field-value and instead replace it
2428   with a configured server name when constructing the effective request URI.
2431   Recipients of an HTTP/1.0 request that lacks a <x:ref>Host</x:ref> header
2432   field &MAY; attempt to use heuristics (e.g., examination of the URI path for
2433   something unique to a particular host) in order to guess the
2434   effective request URI's authority component.
2438<section title="Message Forwarding" anchor="message.forwarding">
2440   As described in <xref target="intermediaries"/>, intermediaries can serve
2441   a variety of roles in the processing of HTTP requests and responses.
2442   Some intermediaries are used to improve performance or availability.
2443   Others are used for access control or to filter content.
2444   Since an HTTP stream has characteristics similar to a pipe-and-filter
2445   architecture, there are no inherent limits to the extent an intermediary
2446   can enhance (or interfere) with either direction of the stream.
2449   Intermediaries that forward a message &MUST; implement the
2450   <x:ref>Connection</x:ref> header field, as specified in
2451   <xref target="header.connection"/>, to exclude fields that are only
2452   intended for the incoming connection.
2455   In order to avoid request loops, a proxy that forwards requests to other
2456   proxies &MUST; be able to recognize and exclude all of its own server
2457   names, including any aliases, local variations, or literal IP addresses.
2461<section title="Via" anchor="header.via">
2462  <iref primary="true" item="Via header field" x:for-anchor=""/>
2463  <x:anchor-alias value="pseudonym"/>
2464  <x:anchor-alias value="received-by"/>
2465  <x:anchor-alias value="received-protocol"/>
2466  <x:anchor-alias value="Via"/>
2468   The "Via" header field &MUST; be sent by a proxy or gateway
2469   in forwarded messages to
2470   indicate the intermediate protocols and recipients between the user
2471   agent and the server on requests, and between the origin server and
2472   the client on responses. It is analogous to the "Received" field
2473   used by email systems (<xref target="RFC5322" x:fmt="of" x:sec="3.6.7"/>).
2474   Via is used in HTTP for tracking message forwards,
2475   avoiding request loops, and identifying the protocol capabilities of
2476   all senders along the request/response chain.
2478<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"/>
2479  <x:ref>Via</x:ref>               = 1#( <x:ref>received-protocol</x:ref> <x:ref>RWS</x:ref> <x:ref>received-by</x:ref>
2480                          [ <x:ref>RWS</x:ref> <x:ref>comment</x:ref> ] )
2481  <x:ref>received-protocol</x:ref> = [ <x:ref>protocol-name</x:ref> "/" ] <x:ref>protocol-version</x:ref>
2482  <x:ref>received-by</x:ref>       = ( <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ) / <x:ref>pseudonym</x:ref>
2483  <x:ref>pseudonym</x:ref>         = <x:ref>token</x:ref>
2486   The received-protocol indicates the protocol version of the message
2487   received by the server or client along each segment of the
2488   request/response chain. The received-protocol version is appended to
2489   the Via field value when the message is forwarded so that information
2490   about the protocol capabilities of upstream applications remains
2491   visible to all recipients.
2494   The protocol-name is excluded if and only if it would be "HTTP". The
2495   received-by field is normally the host and optional port number of a
2496   recipient server or client that subsequently forwarded the message.
2497   However, if the real host is considered to be sensitive information,
2498   it &MAY; be replaced by a pseudonym. If the port is not given, it &MAY;
2499   be assumed to be the default port of the received-protocol.
2502   Multiple Via field values represent each proxy or gateway that has
2503   forwarded the message. Each recipient &MUST; append its information
2504   such that the end result is ordered according to the sequence of
2505   forwarding applications.
2508   Comments &MAY; be used in the Via header field to identify the software
2509   of each recipient, analogous to the <x:ref>User-Agent</x:ref> and
2510   <x:ref>Server</x:ref> header fields. However, all comments in the Via field
2511   are optional and &MAY; be removed by any recipient prior to forwarding the
2512   message.
2515   For example, a request message could be sent from an HTTP/1.0 user
2516   agent to an internal proxy code-named "fred", which uses HTTP/1.1 to
2517   forward the request to a public proxy at, which completes
2518   the request by forwarding it to the origin server at
2519   The request received by would then have the following
2520   Via header field:
2522<figure><artwork type="example">
2523  Via: 1.0 fred, 1.1 (Apache/1.1)
2526   A proxy or gateway used as a portal through a network firewall
2527   &SHOULD-NOT; forward the names and ports of hosts within the firewall
2528   region unless it is explicitly enabled to do so. If not enabled, the
2529   received-by host of any host behind the firewall &SHOULD; be replaced
2530   by an appropriate pseudonym for that host.
2533   A proxy or gateway &MAY; combine an ordered subsequence of Via header
2534   field entries into a single such entry if the entries have identical
2535   received-protocol values. For example,
2537<figure><artwork type="example">
2538  Via: 1.0 ricky, 1.1 ethel, 1.1 fred, 1.0 lucy
2541  could be collapsed to
2543<figure><artwork type="example">
2544  Via: 1.0 ricky, 1.1 mertz, 1.0 lucy
2547   Senders &SHOULD-NOT; combine multiple entries unless they are all
2548   under the same organizational control and the hosts have already been
2549   replaced by pseudonyms. Senders &MUST-NOT; combine entries which
2550   have different received-protocol values.
2554<section title="Message Transforming" anchor="message.transforming">
2556   If a proxy receives a request-target with a host name that is not a
2557   fully qualified domain name, it &MAY; add its own domain to the host name
2558   it received when forwarding the request.  A proxy &MUST-NOT; change the
2559   host name if it is a fully qualified domain name.
2562   A non-transforming proxy &MUST-NOT; modify the "path-absolute" and "query"
2563   parts of the received request-target when forwarding it to the next inbound
2564   server, except as noted above to replace an empty path with "/" or "*".
2567   A non-transforming proxy &MUST; preserve the message payload (&payload;),
2568   though it &MAY; change the message body through application or removal
2569   of a transfer-coding (<xref target="transfer.codings"/>).
2572   A non-transforming proxy &SHOULD-NOT; modify header fields that provide
2573   information about the end points of the communication chain, the resource
2574   state, or the selected representation.
2577   A non-transforming proxy &MUST-NOT; modify any of the following fields in a
2578   request or response, and it &MUST-NOT; add any of these fields if not
2579   already present:
2580  <list style="symbols">
2581    <t><x:ref>Allow</x:ref> (&header-allow;)</t>
2582    <t><x:ref>Content-Location</x:ref> (&header-content-location;)</t>
2583    <t>Content-MD5 (<xref target="RFC2616" x:fmt="of" x:sec="14.15"/>)</t>
2584    <t><x:ref>ETag</x:ref> (&header-etag;)</t>
2585    <t><x:ref>Last-Modified</x:ref> (&header-last-modified;)</t>
2586    <t><x:ref>Server</x:ref> (&header-server;)</t>
2587  </list>
2590   A non-transforming proxy &MUST-NOT; modify an <x:ref>Expires</x:ref>
2591   header field (&header-expires;) if already present in a response, but
2592   it &MAY; add an <x:ref>Expires</x:ref> header field with a field-value
2593   identical to that of the <x:ref>Date</x:ref> header field.
2596   A proxy &MUST-NOT; modify or add any of the following fields in a
2597   message that contains the no-transform cache-control directive:
2598  <list style="symbols">
2599    <t><x:ref>Content-Encoding</x:ref> (&header-content-encoding;)</t>
2600    <t><x:ref>Content-Range</x:ref> (&header-content-range;)</t>
2601    <t><x:ref>Content-Type</x:ref> (&header-content-type;)</t>
2602  </list>
2605   A transforming proxy &MAY; modify or add these fields to a message
2606   that does not include no-transform, but if it does so, it &MUST; add a
2607   Warning 214 (Transformation applied) if one does not already appear
2608   in the message (see &header-warning;).
2611  <t>
2612    <x:h>Warning:</x:h> Unnecessary modification of header fields might
2613    cause authentication failures if stronger authentication
2614    mechanisms are introduced in later versions of HTTP. Such
2615    authentication mechanisms &MAY; rely on the values of header fields
2616    not listed here.
2617  </t>
2621<section title="Associating a Response to a Request" anchor="">
2623   HTTP does not include a request identifier for associating a given
2624   request message with its corresponding one or more response messages.
2625   Hence, it relies on the order of response arrival to correspond exactly
2626   to the order in which requests are made on the same connection.
2627   More than one response message per request only occurs when one or more
2628   informational responses (<x:ref>1xx</x:ref>, see &status-1xx;) precede a final response
2629   to the same request.
2632   A client that uses persistent connections and sends more than one request
2633   per connection &MUST; maintain a list of outstanding requests in the
2634   order sent on that connection and &MUST; associate each received response
2635   message to the highest ordered request that has not yet received a final
2636   (non-<x:ref>1xx</x:ref>) response.
2641<section title="Connection Management" anchor="">
2643   HTTP messaging is independent of the underlying transport or
2644   session-layer connection protocol(s).  HTTP only presumes a reliable
2645   transport with in-order delivery of requests and the corresponding
2646   in-order delivery of responses.  The mapping of HTTP request and
2647   response structures onto the data units of an underlying transport
2648   protocol is outside the scope of this specification.
2651   As described in <xref target="connecting.inbound"/>, the specific
2652   connection protocols to be used for an HTTP interaction are determined by
2653   client configuration and the <x:ref>target URI</x:ref>.
2654   For example, the "http" URI scheme
2655   (<xref target="http.uri"/>) indicates a default connection of TCP
2656   over IP, with a default TCP port of 80, but the client might be
2657   configured to use a proxy via some other connection, port, or protocol.
2660   HTTP implementations are expected to engage in connection management,
2661   which includes maintaining the state of current connections,
2662   establishing a new connection or reusing an existing connection,
2663   processing messages received on a connection, detecting connection
2664   failures, and closing each connection.
2665   Most clients maintain multiple connections in parallel, including
2666   more than one connection per server endpoint.
2667   Most servers are designed to maintain thousands of concurrent connections,
2668   while controlling request queues to enable fair use and detect
2669   denial of service attacks.
2672<section title="Connection" anchor="header.connection">
2673  <iref primary="true" item="Connection header field" x:for-anchor=""/>
2674  <iref primary="true" item="close" x:for-anchor=""/>
2675  <x:anchor-alias value="Connection"/>
2676  <x:anchor-alias value="connection-option"/>
2677  <x:anchor-alias value="close"/>
2679   The "Connection" header field allows the sender to indicate desired
2680   control options for the current connection.  In order to avoid confusing
2681   downstream recipients, a proxy or gateway &MUST; remove or replace any
2682   received connection options before forwarding the message.
2685   When a header field is used to supply control information for or about
2686   the current connection, the sender &SHOULD; list the corresponding
2687   field-name within the "Connection" header field.
2688   A proxy or gateway &MUST; parse a received Connection
2689   header field before a message is forwarded and, for each
2690   connection-option in this field, remove any header field(s) from
2691   the message with the same name as the connection-option, and then
2692   remove the Connection header field itself (or replace it with the
2693   intermediary's own connection options for the forwarded message).
2696   Hence, the Connection header field provides a declarative way of
2697   distinguishing header fields that are only intended for the
2698   immediate recipient ("hop-by-hop") from those fields that are
2699   intended for all recipients on the chain ("end-to-end"), enabling the
2700   message to be self-descriptive and allowing future connection-specific
2701   extensions to be deployed without fear that they will be blindly
2702   forwarded by older intermediaries.
2705   The Connection header field's value has the following grammar:
2707<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Connection"/><iref primary="true" item="Grammar" subitem="connection-option"/>
2708  <x:ref>Connection</x:ref>        = 1#<x:ref>connection-option</x:ref>
2709  <x:ref>connection-option</x:ref> = <x:ref>token</x:ref>
2712   Connection options are case-insensitive.
2715   A sender &MUST-NOT; include field-names in the Connection header
2716   field-value for fields that are defined as expressing constraints
2717   for all recipients in the request or response chain, such as the
2718   Cache-Control header field (&header-cache-control;).
2721   The connection options do not have to correspond to a header field
2722   present in the message, since a connection-specific header field
2723   might not be needed if there are no parameters associated with that
2724   connection option.  Recipients that trigger certain connection
2725   behavior based on the presence of connection options &MUST; do so
2726   based on the presence of the connection-option rather than only the
2727   presence of the optional header field.  In other words, if the
2728   connection option is received as a header field but not indicated
2729   within the Connection field-value, then the recipient &MUST; ignore
2730   the connection-specific header field because it has likely been
2731   forwarded by an intermediary that is only partially conformant.
2734   When defining new connection options, specifications ought to
2735   carefully consider existing deployed header fields and ensure
2736   that the new connection option does not share the same name as
2737   an unrelated header field that might already be deployed.
2738   Defining a new connection option essentially reserves that potential
2739   field-name for carrying additional information related to the
2740   connection option, since it would be unwise for senders to use
2741   that field-name for anything else.
2744   The "<x:dfn>close</x:dfn>" connection option is defined for a
2745   sender to signal that this connection will be closed after completion of
2746   the response. For example,
2748<figure><artwork type="example">
2749  Connection: close
2752   in either the request or the response header fields indicates that
2753   the connection &SHOULD; be closed after the current request/response
2754   is complete (<xref target="persistent.tear-down"/>).
2757   A client that does not support persistent connections &MUST;
2758   send the "close" connection option in every request message.
2761   A server that does not support persistent connections &MUST;
2762   send the "close" connection option in every response message that
2763   does not have a <x:ref>1xx (Informational)</x:ref> status code.
2767<section title="Persistent Connections" anchor="persistent.connections">
2768  <x:anchor-alias value="persistent connections"/>
2770   HTTP was originally designed to use a separate connection for each
2771   request/response pair. As the Web evolved and embedded requests became
2772   common for inline images, the connection establishment overhead was
2773   a significant drain on performance and a concern for Internet congestion.
2774   Message framing (via <x:ref>Content-Length</x:ref>) and optional
2775   long-lived connections (via Keep-Alive) were added to HTTP/1.0 in order
2776   to improve performance for some requests. However, these extensions were
2777   insufficient for dynamically generated responses and difficult to use
2778   with intermediaries.
2781   HTTP/1.1 defaults to the use of "<x:ref>persistent connections</x:ref>",
2782   which allow multiple requests and responses to be carried over a single
2783   connection. The "<x:ref>close</x:ref>" connection-option is used to
2784   signal that a connection will close after the current request/response.
2785   Persistent connections have a number of advantages:
2786  <list style="symbols">
2787      <t>
2788        By opening and closing fewer connections, CPU time is saved
2789        in routers and hosts (clients, servers, proxies, gateways,
2790        tunnels, or caches), and memory used for protocol control
2791        blocks can be saved in hosts.
2792      </t>
2793      <t>
2794        Most requests and responses can be pipelined on a connection.
2795        Pipelining allows a client to make multiple requests without
2796        waiting for each response, allowing a single connection to
2797        be used much more efficiently and with less overall latency.
2798      </t>
2799      <t>
2800        For TCP connections, network congestion is reduced by eliminating the
2801        packets associated with the three way handshake and graceful close
2802        procedures, and by allowing sufficient time to determine the
2803        congestion state of the network.
2804      </t>
2805      <t>
2806        Latency on subsequent requests is reduced since there is no time
2807        spent in the connection opening handshake.
2808      </t>
2809      <t>
2810        HTTP can evolve more gracefully, since most errors can be reported
2811        without the penalty of closing the connection. Clients using
2812        future versions of HTTP might optimistically try a new feature,
2813        but if communicating with an older server, retry with old
2814        semantics after an error is reported.
2815      </t>
2816    </list>
2819   HTTP implementations &SHOULD; implement persistent connections.
2822<section title="Establishment" anchor="persistent.establishment">
2824   It is beyond the scope of this specification to describe how connections
2825   are established via various transport or session-layer protocols.
2826   Each connection applies to only one transport link.
2829   A recipient determines whether a connection is persistent or not based on
2830   the most recently received message's protocol version and
2831   <x:ref>Connection</x:ref> header field (if any):
2832   <list style="symbols">
2833     <t>If the <x:ref>close</x:ref> connection option is present, the
2834        connection will not persist after the current response; else,</t>
2835     <t>If the received protocol is HTTP/1.1 (or later), the connection will
2836        persist after the current response; else,</t>
2837     <t>If the received protocol is HTTP/1.0, the "keep-alive"
2838        connection option is present, the recipient is not a proxy, and
2839        the recipient wishes to honor the HTTP/1.0 "keep-alive" mechanism,
2840        the connection will persist after the current response; otherwise,</t>
2841     <t>The connection will close after the current response.</t>
2842   </list>
2845   A proxy server &MUST-NOT; maintain a persistent connection with an
2846   HTTP/1.0 client (see <xref x:sec="19.7.1" x:fmt="of" target="RFC2068"/> for
2847   information and discussion of the problems with the Keep-Alive header field
2848   implemented by many HTTP/1.0 clients).
2852<section title="Reuse" anchor="persistent.reuse">
2854   In order to remain persistent, all messages on a connection &MUST;
2855   have a self-defined message length (i.e., one not defined by closure
2856   of the connection), as described in <xref target="message.body"/>.
2859   A server &MAY; assume that an HTTP/1.1 client intends to maintain a
2860   persistent connection until a <x:ref>close</x:ref> connection option
2861   is received in a request.
2864   A client &MAY; reuse a persistent connection until it sends or receives
2865   a <x:ref>close</x:ref> connection option or receives an HTTP/1.0 response
2866   without a "keep-alive" connection option.
2869   Clients and servers &SHOULD-NOT; assume that a persistent connection is
2870   maintained for HTTP versions less than 1.1 unless it is explicitly
2871   signaled.
2872   See <xref target="compatibility.with.http.1.0.persistent.connections"/>
2873   for more information on backward compatibility with HTTP/1.0 clients.
2876<section title="Pipelining" anchor="pipelining">
2878   A client that supports persistent connections &MAY; "pipeline" its
2879   requests (i.e., send multiple requests without waiting for each
2880   response). A server &MUST; send its responses to those requests in the
2881   same order that the requests were received.
2884   Clients which assume persistent connections and pipeline immediately
2885   after connection establishment &SHOULD; be prepared to retry their
2886   connection if the first pipelined attempt fails. If a client does
2887   such a retry, it &MUST-NOT; pipeline before it knows the connection is
2888   persistent. Clients &MUST; also be prepared to resend their requests if
2889   the server closes the connection before sending all of the
2890   corresponding responses.
2893   Clients &SHOULD-NOT; pipeline requests using non-idempotent request methods
2894   or non-idempotent sequences of request methods (see &idempotent-methods;).
2895   Otherwise, a premature termination of the transport connection could lead
2896   to indeterminate results. A client wishing to send a non-idempotent
2897   request &SHOULD; wait to send that request until it has received the
2898   response status line for the previous request.
2902<section title="Retrying Requests" anchor="persistent.retrying.requests">
2904   Senders can close the transport connection at any time. Therefore,
2905   clients, servers, and proxies &MUST; be able to recover
2906   from asynchronous close events. Client software &MAY; reopen the
2907   transport connection and retransmit the aborted sequence of requests
2908   without user interaction so long as the request sequence is
2909   idempotent (see &idempotent-methods;). Non-idempotent request methods or sequences
2910   &MUST-NOT; be automatically retried, although user agents &MAY; offer a
2911   human operator the choice of retrying the request(s). Confirmation by
2912   user agent software with semantic understanding of the application
2913   &MAY; substitute for user confirmation. The automatic retry &SHOULD-NOT;
2914   be repeated if the second sequence of requests fails.
2919<section title="Concurrency" anchor="persistent.concurrency">
2921   Clients &SHOULD; limit the number of simultaneous
2922   connections that they maintain to a given server.
2925   Previous revisions of HTTP gave a specific number of connections as a
2926   ceiling, but this was found to be impractical for many applications. As a
2927   result, this specification does not mandate a particular maximum number of
2928   connections, but instead encourages clients to be conservative when opening
2929   multiple connections.
2932   Multiple connections are typically used to avoid the "head-of-line
2933   blocking" problem, wherein a request that takes significant server-side
2934   processing and/or has a large payload blocks subsequent requests on the
2935   same connection. However, each connection consumes server resources.
2936   Furthermore, using multiple connections can cause undesirable side effects
2937   in congested networks.
2940   Note that servers might reject traffic that they deem abusive, including an
2941   excessive number of connections from a client.
2945<section title="Failures and Time-outs" anchor="persistent.failures">
2947   Servers will usually have some time-out value beyond which they will
2948   no longer maintain an inactive connection. Proxy servers might make
2949   this a higher value since it is likely that the client will be making
2950   more connections through the same server. The use of persistent
2951   connections places no requirements on the length (or existence) of
2952   this time-out for either the client or the server.
2955   When a client or server wishes to time-out it &SHOULD; issue a graceful
2956   close on the transport connection. Clients and servers &SHOULD; both
2957   constantly watch for the other side of the transport close, and
2958   respond to it as appropriate. If a client or server does not detect
2959   the other side's close promptly it could cause unnecessary resource
2960   drain on the network.
2963   A client, server, or proxy &MAY; close the transport connection at any
2964   time. For example, a client might have started to send a new request
2965   at the same time that the server has decided to close the "idle"
2966   connection. From the server's point of view, the connection is being
2967   closed while it was idle, but from the client's point of view, a
2968   request is in progress.
2971   Servers &SHOULD; maintain persistent connections and allow the underlying
2972   transport's flow control mechanisms to resolve temporary overloads, rather
2973   than terminate connections with the expectation that clients will retry.
2974   The latter technique can exacerbate network congestion.
2977   A client sending a message body &SHOULD; monitor
2978   the network connection for an error status code while it is transmitting
2979   the request. If the client sees an error status code, it &SHOULD;
2980   immediately cease transmitting the body and close the connection.
2984<section title="Tear-down" anchor="persistent.tear-down">
2985  <iref primary="false" item="Connection header field" x:for-anchor=""/>
2986  <iref primary="false" item="close" x:for-anchor=""/>
2988   The <x:ref>Connection</x:ref> header field
2989   (<xref target="header.connection"/>) provides a "<x:ref>close</x:ref>"
2990   connection option that a sender &SHOULD; send when it wishes to close
2991   the connection after the current request/response pair.
2994   A client that sends a <x:ref>close</x:ref> connection option &MUST-NOT;
2995   send further requests on that connection (after the one containing
2996   <x:ref>close</x:ref>) and &MUST; close the connection after reading the
2997   final response message corresponding to this request.
3000   A server that receives a <x:ref>close</x:ref> connection option &MUST;
3001   initiate a lingering close (see below) of the connection after it sends the
3002   final response to the request that contained <x:ref>close</x:ref>.
3003   The server &SHOULD; include a <x:ref>close</x:ref> connection option
3004   in its final response on that connection. The server &MUST-NOT; process
3005   any further requests received on that connection.
3008   A server that sends a <x:ref>close</x:ref> connection option &MUST;
3009   initiate a lingering close of the connection after it sends the
3010   response containing <x:ref>close</x:ref>. The server &MUST-NOT; process
3011   any further requests received on that connection.
3014   A client that receives a <x:ref>close</x:ref> connection option &MUST;
3015   cease sending requests on that connection and close the connection
3016   after reading the response message containing the close; if additional
3017   pipelined requests had been sent on the connection, the client &SHOULD;
3018   assume that they will not be processed by the server.
3021   If a server performs an immediate close of a TCP connection, there is a
3022   significant risk that the client will not be able to read the last HTTP
3023   response.  If the server receives additional data from the client on a
3024   fully-closed connection, such as another request that was sent by the
3025   client before receiving the server's response, the server's TCP stack will
3026   send a reset packet to the client; unfortunately, the reset packet might
3027   erase the client's unacknowledged input buffers before they can be read
3028   and interpreted by the client's HTTP parser.
3031   To avoid the TCP reset problem, a server can perform a lingering close on a
3032   connection by closing only the write side of the read/write connection
3033   (a half-close) and continuing to read from the connection until the
3034   connection is closed by the client or the server is reasonably certain
3035   that its own TCP stack has received the client's acknowledgement of the
3036   packet(s) containing the server's last response. It is then safe for the
3037   server to fully close the connection.
3040   It is unknown whether the reset problem is exclusive to TCP or might also
3041   be found in other transport connection protocols.
3046<section title="Upgrade" anchor="header.upgrade">
3047  <iref primary="true" item="Upgrade header field" x:for-anchor=""/>
3048  <x:anchor-alias value="Upgrade"/>
3049  <x:anchor-alias value="protocol"/>
3050  <x:anchor-alias value="protocol-name"/>
3051  <x:anchor-alias value="protocol-version"/>
3053   The "Upgrade" header field is intended to provide a simple mechanism
3054   for transitioning from HTTP/1.1 to some other protocol on the same
3055   connection.  A client &MAY; send a list of protocols in the Upgrade
3056   header field of a request to invite the server to switch to one or
3057   more of those protocols before sending the final response.
3058   A server &MUST; send an Upgrade header field in <x:ref>101 (Switching
3059   Protocols)</x:ref> responses to indicate which protocol(s) are being
3060   switched to, and &MUST; send it in <x:ref>426 (Upgrade Required)</x:ref>
3061   responses to indicate acceptable protocols.
3062   A server &MAY; send an Upgrade header field in any other response to
3063   indicate that they might be willing to upgrade to one of the
3064   specified protocols for a future request.
3066<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Upgrade"/>
3067  <x:ref>Upgrade</x:ref>          = 1#<x:ref>protocol</x:ref>
3069  <x:ref>protocol</x:ref>         = <x:ref>protocol-name</x:ref> ["/" <x:ref>protocol-version</x:ref>]
3070  <x:ref>protocol-name</x:ref>    = <x:ref>token</x:ref>
3071  <x:ref>protocol-version</x:ref> = <x:ref>token</x:ref>
3074   For example,
3076<figure><artwork type="example">
3077  Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11
3080   Upgrade eases the difficult transition between incompatible protocols by
3081   allowing the client to initiate a request in the more commonly
3082   supported protocol while indicating to the server that it would like
3083   to use a "better" protocol if available (where "better" is determined
3084   by the server, possibly according to the nature of the request method
3085   or target resource).
3088   Upgrade cannot be used to insist on a protocol change; its acceptance and
3089   use by the server is optional. The capabilities and nature of the
3090   application-level communication after the protocol change is entirely
3091   dependent upon the new protocol chosen, although the first action
3092   after changing the protocol &MUST; be a response to the initial HTTP
3093   request that contained the Upgrade header field.
3096   For example, if the Upgrade header field is received in a GET request
3097   and the server decides to switch protocols, then it &MUST; first respond
3098   with a <x:ref>101 (Switching Protocols)</x:ref> message in HTTP/1.1 and
3099   then immediately follow that with the new protocol's equivalent of a
3100   response to a GET on the target resource.  This allows a connection to be
3101   upgraded to protocols with the same semantics as HTTP without the
3102   latency cost of an additional round-trip.  A server &MUST-NOT; switch
3103   protocols unless the received message semantics can be honored by the new
3104   protocol; an OPTIONS request can be honored by any protocol.
3107   When Upgrade is sent, a sender &MUST; also send a
3108   <x:ref>Connection</x:ref> header field (<xref target="header.connection"/>)
3109   that contains the "upgrade" connection option, in order to prevent Upgrade
3110   from being accidentally forwarded by intermediaries that might not implement
3111   the listed protocols.  A server &MUST; ignore an Upgrade header field that
3112   is received in an HTTP/1.0 request.
3115   The Upgrade header field only applies to switching application-level
3116   protocols on the existing connection; it cannot be used
3117   to switch to a protocol on a different connection. For that purpose, it is
3118   more appropriate to use a <x:ref>3xx (Redirection)</x:ref> response
3119   (&status-3xx;).
3122   This specification only defines the protocol name "HTTP" for use by
3123   the family of Hypertext Transfer Protocols, as defined by the HTTP
3124   version rules of <xref target="http.version"/> and future updates to this
3125   specification. Additional tokens can be registered with IANA using the
3126   registration procedure defined in <xref target="upgrade.token.registry"/>.
3131<section title="IANA Considerations" anchor="IANA.considerations">
3133<section title="Header Field Registration" anchor="header.field.registration">
3135   HTTP header fields are registered within the Message Header Field Registry
3136   <xref target="RFC3864"/> maintained by IANA at
3137   <eref target=""/>.
3140   This document defines the following HTTP header fields, so their
3141   associated registry entries shall be updated according to the permanent
3142   registrations below:
3144<?BEGININC p1-messaging.iana-headers ?>
3145<!--AUTOGENERATED FROM extract-header-defs.xslt, do not edit manually-->
3146<texttable align="left" suppress-title="true" anchor="iana.header.registration.table">
3147   <ttcol>Header Field Name</ttcol>
3148   <ttcol>Protocol</ttcol>
3149   <ttcol>Status</ttcol>
3150   <ttcol>Reference</ttcol>
3152   <c>Connection</c>
3153   <c>http</c>
3154   <c>standard</c>
3155   <c>
3156      <xref target="header.connection"/>
3157   </c>
3158   <c>Content-Length</c>
3159   <c>http</c>
3160   <c>standard</c>
3161   <c>
3162      <xref target="header.content-length"/>
3163   </c>
3164   <c>Host</c>
3165   <c>http</c>
3166   <c>standard</c>
3167   <c>
3168      <xref target=""/>
3169   </c>
3170   <c>TE</c>
3171   <c>http</c>
3172   <c>standard</c>
3173   <c>
3174      <xref target="header.te"/>
3175   </c>
3176   <c>Trailer</c>
3177   <c>http</c>
3178   <c>standard</c>
3179   <c>
3180      <xref target="header.trailer"/>
3181   </c>
3182   <c>Transfer-Encoding</c>
3183   <c>http</c>
3184   <c>standard</c>
3185   <c>
3186      <xref target="header.transfer-encoding"/>
3187   </c>
3188   <c>Upgrade</c>
3189   <c>http</c>
3190   <c>standard</c>
3191   <c>
3192      <xref target="header.upgrade"/>
3193   </c>
3194   <c>Via</c>
3195   <c>http</c>
3196   <c>standard</c>
3197   <c>
3198      <xref target="header.via"/>
3199   </c>
3202<?ENDINC p1-messaging.iana-headers ?>
3204   Furthermore, the header field-name "Close" shall be registered as
3205   "reserved", since using that name as an HTTP header field might
3206   conflict with the "close" connection option of the "<x:ref>Connection</x:ref>"
3207   header field (<xref target="header.connection"/>).
3209<texttable align="left" suppress-title="true">
3210   <ttcol>Header Field Name</ttcol>
3211   <ttcol>Protocol</ttcol>
3212   <ttcol>Status</ttcol>
3213   <ttcol>Reference</ttcol>
3215   <c>Close</c>
3216   <c>http</c>
3217   <c>reserved</c>
3218   <c>
3219      <xref target="header.field.registration"/>
3220   </c>
3223   The change controller is: "IETF ( - Internet Engineering Task Force".
3227<section title="URI Scheme Registration" anchor="uri.scheme.registration">
3229   IANA maintains the registry of URI Schemes <xref target="RFC4395"/> at
3230   <eref target=""/>.
3233   This document defines the following URI schemes, so their
3234   associated registry entries shall be updated according to the permanent
3235   registrations below:
3237<texttable align="left" suppress-title="true">
3238   <ttcol>URI Scheme</ttcol>
3239   <ttcol>Description</ttcol>
3240   <ttcol>Reference</ttcol>
3242   <c>http</c>
3243   <c>Hypertext Transfer Protocol</c>
3244   <c><xref target="http.uri"/></c>
3246   <c>https</c>
3247   <c>Hypertext Transfer Protocol Secure</c>
3248   <c><xref target="https.uri"/></c>
3252<section title="Internet Media Type Registrations" anchor="">
3254   This document serves as the specification for the Internet media types
3255   "message/http" and "application/http". The following is to be registered with
3256   IANA (see <xref target="RFC4288"/>).
3258<section title="Internet Media Type message/http" anchor="">
3259<iref item="Media Type" subitem="message/http" primary="true"/>
3260<iref item="message/http Media Type" primary="true"/>
3262   The message/http type can be used to enclose a single HTTP request or
3263   response message, provided that it obeys the MIME restrictions for all
3264   "message" types regarding line length and encodings.
3267  <list style="hanging" x:indent="12em">
3268    <t hangText="Type name:">
3269      message
3270    </t>
3271    <t hangText="Subtype name:">
3272      http
3273    </t>
3274    <t hangText="Required parameters:">
3275      none
3276    </t>
3277    <t hangText="Optional parameters:">
3278      version, msgtype
3279      <list style="hanging">
3280        <t hangText="version:">
3281          The HTTP-version number of the enclosed message
3282          (e.g., "1.1"). If not present, the version can be
3283          determined from the first line of the body.
3284        </t>
3285        <t hangText="msgtype:">
3286          The message type &mdash; "request" or "response". If not
3287          present, the type can be determined from the first
3288          line of the body.
3289        </t>
3290      </list>
3291    </t>
3292    <t hangText="Encoding considerations:">
3293      only "7bit", "8bit", or "binary" are permitted
3294    </t>
3295    <t hangText="Security considerations:">
3296      none
3297    </t>
3298    <t hangText="Interoperability considerations:">
3299      none
3300    </t>
3301    <t hangText="Published specification:">
3302      This specification (see <xref target=""/>).
3303    </t>
3304    <t hangText="Applications that use this media type:">
3305    </t>
3306    <t hangText="Additional information:">
3307      <list style="hanging">
3308        <t hangText="Magic number(s):">none</t>
3309        <t hangText="File extension(s):">none</t>
3310        <t hangText="Macintosh file type code(s):">none</t>
3311      </list>
3312    </t>
3313    <t hangText="Person and email address to contact for further information:">
3314      See Authors Section.
3315    </t>
3316    <t hangText="Intended usage:">
3317      COMMON
3318    </t>
3319    <t hangText="Restrictions on usage:">
3320      none
3321    </t>
3322    <t hangText="Author/Change controller:">
3323      IESG
3324    </t>
3325  </list>
3328<section title="Internet Media Type application/http" anchor="">
3329<iref item="Media Type" subitem="application/http" primary="true"/>
3330<iref item="application/http Media Type" primary="true"/>
3332   The application/http type can be used to enclose a pipeline of one or more
3333   HTTP request or response messages (not intermixed).
3336  <list style="hanging" x:indent="12em">
3337    <t hangText="Type name:">
3338      application
3339    </t>
3340    <t hangText="Subtype name:">
3341      http
3342    </t>
3343    <t hangText="Required parameters:">
3344      none
3345    </t>
3346    <t hangText="Optional parameters:">
3347      version, msgtype
3348      <list style="hanging">
3349        <t hangText="version:">
3350          The HTTP-version number of the enclosed messages
3351          (e.g., "1.1"). If not present, the version can be
3352          determined from the first line of the body.
3353        </t>
3354        <t hangText="msgtype:">
3355          The message type &mdash; "request" or "response". If not
3356          present, the type can be determined from the first
3357          line of the body.
3358        </t>
3359      </list>
3360    </t>
3361    <t hangText="Encoding considerations:">
3362      HTTP messages enclosed by this type
3363      are in "binary" format; use of an appropriate
3364      Content-Transfer-Encoding is required when
3365      transmitted via E-mail.
3366    </t>
3367    <t hangText="Security considerations:">
3368      none
3369    </t>
3370    <t hangText="Interoperability considerations:">
3371      none
3372    </t>
3373    <t hangText="Published specification:">
3374      This specification (see <xref target=""/>).
3375    </t>
3376    <t hangText="Applications that use this media type:">
3377    </t>
3378    <t hangText="Additional information:">
3379      <list style="hanging">
3380        <t hangText="Magic number(s):">none</t>
3381        <t hangText="File extension(s):">none</t>
3382        <t hangText="Macintosh file type code(s):">none</t>
3383      </list>
3384    </t>
3385    <t hangText="Person and email address to contact for further information:">
3386      See Authors Section.
3387    </t>
3388    <t hangText="Intended usage:">
3389      COMMON
3390    </t>
3391    <t hangText="Restrictions on usage:">
3392      none
3393    </t>
3394    <t hangText="Author/Change controller:">
3395      IESG
3396    </t>
3397  </list>
3402<section title="Transfer Coding Registry" anchor="transfer.coding.registry">
3404   The HTTP Transfer Coding Registry defines the name space for transfer
3405   coding names.
3408   Registrations &MUST; include the following fields:
3409   <list style="symbols">
3410     <t>Name</t>
3411     <t>Description</t>
3412     <t>Pointer to specification text</t>
3413   </list>
3416   Names of transfer codings &MUST-NOT; overlap with names of content codings
3417   (&content-codings;) unless the encoding transformation is identical, as
3418   is the case for the compression codings defined in
3419   <xref target="compression.codings"/>.
3422   Values to be added to this name space require IETF Review (see
3423   <xref target="RFC5226" x:fmt="of" x:sec="4.1"/>), and &MUST;
3424   conform to the purpose of transfer coding defined in this section.
3425   Use of program names for the identification of encoding formats
3426   is not desirable and is discouraged for future encodings.
3429   The registry itself is maintained at
3430   <eref target=""/>.
3434<section title="Transfer Coding Registrations" anchor="transfer.coding.registration">
3436   The HTTP Transfer Coding Registry shall be updated with the registrations
3437   below:
3439<texttable align="left" suppress-title="true" anchor="iana.transfer.coding.registration.table">
3440   <ttcol>Name</ttcol>
3441   <ttcol>Description</ttcol>
3442   <ttcol>Reference</ttcol>
3443   <c>chunked</c>
3444   <c>Transfer in a series of chunks</c>
3445   <c>
3446      <xref target="chunked.encoding"/>
3447   </c>
3448   <c>compress</c>
3449   <c>UNIX "compress" program method</c>
3450   <c>
3451      <xref target="compress.coding"/>
3452   </c>
3453   <c>deflate</c>
3454   <c>"deflate" compression mechanism (<xref target="RFC1951"/>) used inside
3455   the "zlib" data format (<xref target="RFC1950"/>)
3456   </c>
3457   <c>
3458      <xref target="deflate.coding"/>
3459   </c>
3460   <c>gzip</c>
3461   <c>Same as GNU zip <xref target="RFC1952"/></c>
3462   <c>
3463      <xref target="gzip.coding"/>
3464   </c>
3468<section title="Upgrade Token Registry" anchor="upgrade.token.registry">
3470   The HTTP Upgrade Token Registry defines the name space for protocol-name
3471   tokens used to identify protocols in the <x:ref>Upgrade</x:ref> header
3472   field. Each registered protocol name is associated with contact information
3473   and an optional set of specifications that details how the connection
3474   will be processed after it has been upgraded.
3477   Registrations happen on a "First Come First Served" basis (see
3478   <xref target="RFC5226" x:sec="4.1" x:fmt="of"/>) and are subject to the
3479   following rules:
3480  <list style="numbers">
3481    <t>A protocol-name token, once registered, stays registered forever.</t>
3482    <t>The registration &MUST; name a responsible party for the
3483       registration.</t>
3484    <t>The registration &MUST; name a point of contact.</t>
3485    <t>The registration &MAY; name a set of specifications associated with
3486       that token. Such specifications need not be publicly available.</t>
3487    <t>The registration &SHOULD; name a set of expected "protocol-version"
3488       tokens associated with that token at the time of registration.</t>
3489    <t>The responsible party &MAY; change the registration at any time.
3490       The IANA will keep a record of all such changes, and make them
3491       available upon request.</t>
3492    <t>The IESG &MAY; reassign responsibility for a protocol token.
3493       This will normally only be used in the case when a
3494       responsible party cannot be contacted.</t>
3495  </list>
3498   This registration procedure for HTTP Upgrade Tokens replaces that
3499   previously defined in <xref target="RFC2817" x:fmt="of" x:sec="7.2"/>.
3503<section title="Upgrade Token Registration" anchor="upgrade.token.registration">
3505   The HTTP Upgrade Token Registry shall be updated with the registration
3506   below:
3508<texttable align="left" suppress-title="true">
3509   <ttcol>Value</ttcol>
3510   <ttcol>Description</ttcol>
3511   <ttcol>Expected Version Tokens</ttcol>
3512   <ttcol>Reference</ttcol>
3514   <c>HTTP</c>
3515   <c>Hypertext Transfer Protocol</c>
3516   <c>any DIGIT.DIGIT (e.g, "2.0")</c>
3517   <c><xref target="http.version"/></c>
3520   The responsible party is: "IETF ( - Internet Engineering Task Force".
3526<section title="Security Considerations" anchor="security.considerations">
3528   This section is meant to inform application developers, information
3529   providers, and users of the security limitations in HTTP/1.1 as
3530   described by this document. The discussion does not include
3531   definitive solutions to the problems revealed, though it does make
3532   some suggestions for reducing security risks.
3535<section title="Personal Information" anchor="personal.information">
3537   HTTP clients are often privy to large amounts of personal information,
3538   including both information provided by the user to interact with resources
3539   (e.g., the user's name, location, mail address, passwords, encryption
3540   keys, etc.) and information about the user's browsing activity over
3541   time (e.g., history, bookmarks, etc.). HTTP implementations need to
3542   prevent unintentional leakage of this information.
3546<section title="Abuse of Server Log Information" anchor="abuse.of.server.log.information">
3548   A server is in the position to save personal data about a user's
3549   requests which might identify their reading patterns or subjects of
3550   interest.  In particular, log information gathered at an intermediary
3551   often contains a history of user agent interaction, across a multitude
3552   of sites, that can be traced to individual users.
3555   HTTP log information is confidential in nature; its handling is often
3556   constrained by laws and regulations.  Log information needs to be securely
3557   stored and appropriate guidelines followed for its analysis.
3558   Anonymization of personal information within individual entries helps,
3559   but is generally not sufficient to prevent real log traces from being
3560   re-identified based on correlation with other access characteristics.
3561   As such, access traces that are keyed to a specific client should not
3562   be published even if the key is pseudonymous.
3565   To minimize the risk of theft or accidental publication, log information
3566   should be purged of personally identifiable information, including
3567   user identifiers, IP addresses, and user-provided query parameters,
3568   as soon as that information is no longer necessary to support operational
3569   needs for security, auditing, or fraud control.
3573<section title="Attacks Based On File and Path Names" anchor="attack.pathname">
3575   Origin servers &SHOULD; be careful to restrict
3576   the documents returned by HTTP requests to be only those that were
3577   intended by the server administrators. If an HTTP server translates
3578   HTTP URIs directly into file system calls, the server &MUST; take
3579   special care not to serve files that were not intended to be
3580   delivered to HTTP clients. For example, UNIX, Microsoft Windows, and
3581   other operating systems use ".." as a path component to indicate a
3582   directory level above the current one. On such a system, an HTTP
3583   server &MUST; disallow any such construct in the request-target if it
3584   would otherwise allow access to a resource outside those intended to
3585   be accessible via the HTTP server. Similarly, files intended for
3586   reference only internally to the server (such as access control
3587   files, configuration files, and script code) &MUST; be protected from
3588   inappropriate retrieval, since they might contain sensitive
3589   information.
3593<section title="DNS-related Attacks" anchor="dns.related.attacks">
3595   HTTP clients rely heavily on the Domain Name Service (DNS), and are thus
3596   generally prone to security attacks based on the deliberate misassociation
3597   of IP addresses and DNS names not protected by DNSSEC. Clients need to be
3598   cautious in assuming the validity of an IP number/DNS name association unless
3599   the response is protected by DNSSEC (<xref target="RFC4033"/>).
3603<section title="Intermediaries and Caching" anchor="attack.intermediaries">
3605   By their very nature, HTTP intermediaries are men-in-the-middle, and
3606   represent an opportunity for man-in-the-middle attacks. Compromise of
3607   the systems on which the intermediaries run can result in serious security
3608   and privacy problems. Intermediaries have access to security-related
3609   information, personal information about individual users and
3610   organizations, and proprietary information belonging to users and
3611   content providers. A compromised intermediary, or an intermediary
3612   implemented or configured without regard to security and privacy
3613   considerations, might be used in the commission of a wide range of
3614   potential attacks.
3617   Intermediaries that contain a shared cache are especially vulnerable
3618   to cache poisoning attacks.
3621   Implementers need to consider the privacy and security
3622   implications of their design and coding decisions, and of the
3623   configuration options they provide to operators (especially the
3624   default configuration).
3627   Users need to be aware that intermediaries are no more trustworthy than
3628   the people who run them; HTTP itself cannot solve this problem.
3632<section title="Protocol Element Size Overflows" anchor="attack.protocol.element.size.overflows">
3634   Because HTTP uses mostly textual, character-delimited fields, attackers can
3635   overflow buffers in implementations, and/or perform a Denial of Service
3636   against implementations that accept fields with unlimited lengths.
3639   To promote interoperability, this specification makes specific
3640   recommendations for minimum size limits on request-line
3641   (<xref target="request.line"/>)
3642   and blocks of header fields (<xref target="header.fields"/>). These are
3643   minimum recommendations, chosen to be supportable even by implementations
3644   with limited resources; it is expected that most implementations will
3645   choose substantially higher limits.
3648   This specification also provides a way for servers to reject messages that
3649   have request-targets that are too long (&status-414;) or request entities
3650   that are too large (&status-4xx;).
3653   Recipients &SHOULD; carefully limit the extent to which they read other
3654   fields, including (but not limited to) request methods, response status
3655   phrases, header field-names, and body chunks, so as to avoid denial of
3656   service attacks without impeding interoperability.
3661<section title="Acknowledgments" anchor="acks">
3663   This edition of HTTP/1.1 builds on the many contributions that went into
3664   <xref target="RFC1945" format="none">RFC 1945</xref>,
3665   <xref target="RFC2068" format="none">RFC 2068</xref>,
3666   <xref target="RFC2145" format="none">RFC 2145</xref>, and
3667   <xref target="RFC2616" format="none">RFC 2616</xref>, including
3668   substantial contributions made by the previous authors, editors, and
3669   working group chairs: Tim Berners-Lee, Ari Luotonen, Roy T. Fielding,
3670   Henrik Frystyk Nielsen, Jim Gettys, Jeffrey C. Mogul, Larry Masinter,
3671   and Paul J. Leach. Mark Nottingham oversaw this effort as working group chair.
3674   Since 1999, the following contributors have helped improve the HTTP
3675   specification by reporting bugs, asking smart questions, drafting or
3676   reviewing text, and evaluating open issues:
3678<?BEGININC acks ?>
3679<t>Adam Barth,
3680Adam Roach,
3681Addison Phillips,
3682Adrian Chadd,
3683Adrien W. de Croy,
3684Alan Ford,
3685Alan Ruttenberg,
3686Albert Lunde,
3687Alek Storm,
3688Alex Rousskov,
3689Alexandre Morgaut,
3690Alexey Melnikov,
3691Alisha Smith,
3692Amichai Rothman,
3693Amit Klein,
3694Amos Jeffries,
3695Andreas Maier,
3696Andreas Petersson,
3697Anil Sharma,
3698Anne van Kesteren,
3699Anthony Bryan,
3700Asbjorn Ulsberg,
3701Ashok Kumar,
3702Balachander Krishnamurthy,
3703Barry Leiba,
3704Ben Laurie,
3705Benjamin Niven-Jenkins,
3706Bil Corry,
3707Bill Burke,
3708Bjoern Hoehrmann,
3709Bob Scheifler,
3710Boris Zbarsky,
3711Brett Slatkin,
3712Brian Kell,
3713Brian McBarron,
3714Brian Pane,
3715Brian Smith,
3716Bryce Nesbitt,
3717Cameron Heavon-Jones,
3718Carl Kugler,
3719Carsten Bormann,
3720Charles Fry,
3721Chris Newman,
3722Cyrus Daboo,
3723Dale Robert Anderson,
3724Dan Wing,
3725Dan Winship,
3726Daniel Stenberg,
3727Dave Cridland,
3728Dave Crocker,
3729Dave Kristol,
3730David Booth,
3731David Singer,
3732David W. Morris,
3733Diwakar Shetty,
3734Dmitry Kurochkin,
3735Drummond Reed,
3736Duane Wessels,
3737Edward Lee,
3738Eliot Lear,
3739Eran Hammer-Lahav,
3740Eric D. Williams,
3741Eric J. Bowman,
3742Eric Lawrence,
3743Eric Rescorla,
3744Erik Aronesty,
3745Evan Prodromou,
3746Florian Weimer,
3747Frank Ellermann,
3748Fred Bohle,
3749Gabriel Montenegro,
3750Geoffrey Sneddon,
3751Gervase Markham,
3752Grahame Grieve,
3753Greg Wilkins,
3754Harald Tveit Alvestrand,
3755Harry Halpin,
3756Helge Hess,
3757Henrik Nordstrom,
3758Henry S. Thompson,
3759Henry Story,
3760Herbert van de Sompel,
3761Howard Melman,
3762Hugo Haas,
3763Ian Fette,
3764Ian Hickson,
3765Ido Safruti,
3766Ingo Struck,
3767J. Ross Nicoll,
3768James H. Manger,
3769James Lacey,
3770James M. Snell,
3771Jamie Lokier,
3772Jan Algermissen,
3773Jeff Hodges (who came up with the term 'effective Request-URI'),
3774Jeff Walden,
3775Jim Luther,
3776Joe D. Williams,
3777Joe Gregorio,
3778Joe Orton,
3779John C. Klensin,
3780John C. Mallery,
3781John Cowan,
3782John Kemp,
3783John Panzer,
3784John Schneider,
3785John Stracke,
3786John Sullivan,
3787Jonas Sicking,
3788Jonathan Billington,
3789Jonathan Moore,
3790Jonathan Rees,
3791Jonathan Silvera,
3792Jordi Ros,
3793Joris Dobbelsteen,
3794Josh Cohen,
3795Julien Pierre,
3796Jungshik Shin,
3797Justin Chapweske,
3798Justin Erenkrantz,
3799Justin James,
3800Kalvinder Singh,
3801Karl Dubost,
3802Keith Hoffman,
3803Keith Moore,
3804Ken Murchison,
3805Koen Holtman,
3806Konstantin Voronkov,
3807Kris Zyp,
3808Lisa Dusseault,
3809Maciej Stachowiak,
3810Marc Schneider,
3811Marc Slemko,
3812Mark Baker,
3813Mark Pauley,
3814Mark Watson,
3815Markus Isomaki,
3816Markus Lanthaler,
3817Martin J. Duerst,
3818Martin Musatov,
3819Martin Nilsson,
3820Martin Thomson,
3821Matt Lynch,
3822Matthew Cox,
3823Max Clark,
3824Michael Burrows,
3825Michael Hausenblas,
3826Mike Amundsen,
3827Mike Belshe,
3828Mike Kelly,
3829Mike Schinkel,
3830Miles Sabin,
3831Murray S. Kucherawy,
3832Mykyta Yevstifeyev,
3833Nathan Rixham,
3834Nicholas Shanks,
3835Nico Williams,
3836Nicolas Alvarez,
3837Nicolas Mailhot,
3838Noah Slater,
3839Pablo Castro,
3840Pat Hayes,
3841Patrick R. McManus,
3842Paul E. Jones,
3843Paul Hoffman,
3844Paul Marquess,
3845Peter Lepeska,
3846Peter Saint-Andre,
3847Peter Watkins,
3848Phil Archer,
3849Philippe Mougin,
3850Phillip Hallam-Baker,
3851Poul-Henning Kamp,
3852Preethi Natarajan,
3853Rajeev Bector,
3854Ray Polk,
3855Reto Bachmann-Gmuer,
3856Richard Cyganiak,
3857Robert Brewer,
3858Robert Collins,
3859Robert O'Callahan,
3860Robert Olofsson,
3861Robert Sayre,
3862Robert Siemer,
3863Robert de Wilde,
3864Roberto Javier Godoy,
3865Roberto Peon,
3866Ronny Widjaja,
3867S. Mike Dierken,
3868Salvatore Loreto,
3869Sam Johnston,
3870Sam Ruby,
3871Scott Lawrence (who maintained the original issues list),
3872Sean B. Palmer,
3873Shane McCarron,
3874Stefan Eissing,
3875Stefan Tilkov,
3876Stefanos Harhalakis,
3877Stephane Bortzmeyer,
3878Stephen Farrell,
3879Stephen Ludin,
3880Stuart Williams,
3881Subbu Allamaraju,
3882Sylvain Hellegouarch,
3883Tapan Divekar,
3884Tatsuya Hayashi,
3885Ted Hardie,
3886Thomas Broyer,
3887Thomas Nordin,
3888Thomas Roessler,
3889Tim Bray,
3890Tim Morgan,
3891Tim Olsen,
3892Tom Zhou,
3893Travis Snoozy,
3894Tyler Close,
3895Vincent Murphy,
3896Wenbo Zhu,
3897Werner Baumann,
3898Wilbur Streett,
3899Wilfredo Sanchez Vega,
3900William A. Rowe Jr.,
3901William Chan,
3902Willy Tarreau,
3903Xiaoshu Wang,
3904Yaron Goland,
3905Yngve Nysaeter Pettersen,
3906Yoav Nir,
3907Yogesh Bang,
3908Yutaka Oiwa,
3909Yves Lafon (long-time member of the editor team),
3910Zed A. Shaw, and
3911Zhong Yu.
3913<?ENDINC acks ?>
3915   See <xref target="RFC2616" x:fmt="of" x:sec="16"/> for additional
3916   acknowledgements from prior revisions.
3923<references title="Normative References">
3925<reference anchor="Part2">
3926  <front>
3927    <title>Hypertext Transfer Protocol (HTTP/1.1): Semantics and Content</title>
3928    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3929      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
3930      <address><email></email></address>
3931    </author>
3932    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3933      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3934      <address><email></email></address>
3935    </author>
3936    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3937  </front>
3938  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p2-semantics-&ID-VERSION;"/>
3939  <x:source href="p2-semantics.xml" basename="p2-semantics">
3940    <x:defines>1xx (Informational)</x:defines>
3941    <x:defines>1xx</x:defines>
3942    <x:defines>100 (Continue)</x:defines>
3943    <x:defines>101 (Switching Protocols)</x:defines>
3944    <x:defines>2xx (Successful)</x:defines>
3945    <x:defines>2xx</x:defines>
3946    <x:defines>200 (OK)</x:defines>
3947    <x:defines>204 (No Content)</x:defines>
3948    <x:defines>3xx (Redirection)</x:defines>
3949    <x:defines>3xx</x:defines>
3950    <x:defines>301 (Moved Permanently)</x:defines>
3951    <x:defines>4xx (Client Error)</x:defines>
3952    <x:defines>4xx</x:defines>
3953    <x:defines>400 (Bad Request)</x:defines>
3954    <x:defines>405 (Method Not Allowed)</x:defines>
3955    <x:defines>411 (Length Required)</x:defines>
3956    <x:defines>414 (URI Too Long)</x:defines>
3957    <x:defines>417 (Expectation Failed)</x:defines>
3958    <x:defines>426 (Upgrade Required)</x:defines>
3959    <x:defines>501 (Not Implemented)</x:defines>
3960    <x:defines>502 (Bad Gateway)</x:defines>
3961    <x:defines>505 (HTTP Version Not Supported)</x:defines>
3962    <x:defines>Allow</x:defines>
3963    <x:defines>Content-Encoding</x:defines>
3964    <x:defines>Content-Location</x:defines>
3965    <x:defines>Content-Type</x:defines>
3966    <x:defines>Date</x:defines>
3967    <x:defines>Expect</x:defines>
3968    <x:defines>Location</x:defines>
3969    <x:defines>Server</x:defines>
3970    <x:defines>User-Agent</x:defines>
3971  </x:source>
3974<reference anchor="Part4">
3975  <front>
3976    <title>Hypertext Transfer Protocol (HTTP/1.1): Conditional Requests</title>
3977    <author fullname="Roy T. Fielding" initials="R." role="editor" surname="Fielding">
3978      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
3979      <address><email></email></address>
3980    </author>
3981    <author fullname="Julian F. Reschke" initials="J. F." role="editor" surname="Reschke">
3982      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3983      <address><email></email></address>
3984    </author>
3985    <date month="&ID-MONTH;" year="&ID-YEAR;" />
3986  </front>
3987  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p4-conditional-&ID-VERSION;" />
3988  <x:source basename="p4-conditional" href="p4-conditional.xml">
3989    <x:defines>304 (Not Modified)</x:defines>
3990    <x:defines>ETag</x:defines>
3991    <x:defines>Last-Modified</x:defines>
3992  </x:source>
3995<reference anchor="Part5">
3996  <front>
3997    <title>Hypertext Transfer Protocol (HTTP/1.1): Range Requests</title>
3998    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3999      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4000      <address><email></email></address>
4001    </author>
4002    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
4003      <organization abbrev="W3C">World Wide Web Consortium</organization>
4004      <address><email></email></address>
4005    </author>
4006    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4007      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4008      <address><email></email></address>
4009    </author>
4010    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4011  </front>
4012  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p5-range-&ID-VERSION;"/>
4013  <x:source href="p5-range.xml" basename="p5-range">
4014    <x:defines>Content-Range</x:defines>
4015  </x:source>
4018<reference anchor="Part6">
4019  <front>
4020    <title>Hypertext Transfer Protocol (HTTP/1.1): Caching</title>
4021    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4022      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4023      <address><email></email></address>
4024    </author>
4025    <author initials="M." surname="Nottingham" fullname="Mark Nottingham" role="editor">
4026      <organization>Akamai</organization>
4027      <address><email></email></address>
4028    </author>
4029    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4030      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4031      <address><email></email></address>
4032    </author>
4033    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4034  </front>
4035  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p6-cache-&ID-VERSION;"/>
4036  <x:source href="p6-cache.xml" basename="p6-cache">
4037    <x:defines>Expires</x:defines>
4038  </x:source>
4041<reference anchor="Part7">
4042  <front>
4043    <title>Hypertext Transfer Protocol (HTTP/1.1): Authentication</title>
4044    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4045      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4046      <address><email></email></address>
4047    </author>
4048    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4049      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4050      <address><email></email></address>
4051    </author>
4052    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4053  </front>
4054  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p7-auth-&ID-VERSION;"/>
4055  <x:source href="p7-auth.xml" basename="p7-auth">
4056    <x:defines>Proxy-Authenticate</x:defines>
4057    <x:defines>Proxy-Authorization</x:defines>
4058  </x:source>
4061<reference anchor="RFC5234">
4062  <front>
4063    <title abbrev="ABNF for Syntax Specifications">Augmented BNF for Syntax Specifications: ABNF</title>
4064    <author initials="D." surname="Crocker" fullname="Dave Crocker" role="editor">
4065      <organization>Brandenburg InternetWorking</organization>
4066      <address>
4067        <email></email>
4068      </address> 
4069    </author>
4070    <author initials="P." surname="Overell" fullname="Paul Overell">
4071      <organization>THUS plc.</organization>
4072      <address>
4073        <email></email>
4074      </address>
4075    </author>
4076    <date month="January" year="2008"/>
4077  </front>
4078  <seriesInfo name="STD" value="68"/>
4079  <seriesInfo name="RFC" value="5234"/>
4082<reference anchor="RFC2119">
4083  <front>
4084    <title>Key words for use in RFCs to Indicate Requirement Levels</title>
4085    <author initials="S." surname="Bradner" fullname="Scott Bradner">
4086      <organization>Harvard University</organization>
4087      <address><email></email></address>
4088    </author>
4089    <date month="March" year="1997"/>
4090  </front>
4091  <seriesInfo name="BCP" value="14"/>
4092  <seriesInfo name="RFC" value="2119"/>
4095<reference anchor="RFC3986">
4096 <front>
4097  <title abbrev='URI Generic Syntax'>Uniform Resource Identifier (URI): Generic Syntax</title>
4098  <author initials='T.' surname='Berners-Lee' fullname='Tim Berners-Lee'>
4099    <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
4100    <address>
4101       <email></email>
4102       <uri></uri>
4103    </address>
4104  </author>
4105  <author initials='R.' surname='Fielding' fullname='Roy T. Fielding'>
4106    <organization abbrev="Day Software">Day Software</organization>
4107    <address>
4108      <email></email>
4109      <uri></uri>
4110    </address>
4111  </author>
4112  <author initials='L.' surname='Masinter' fullname='Larry Masinter'>
4113    <organization abbrev="Adobe Systems">Adobe Systems Incorporated</organization>
4114    <address>
4115      <email></email>
4116      <uri></uri>
4117    </address>
4118  </author>
4119  <date month='January' year='2005'></date>
4120 </front>
4121 <seriesInfo name="STD" value="66"/>
4122 <seriesInfo name="RFC" value="3986"/>
4125<reference anchor="USASCII">
4126  <front>
4127    <title>Coded Character Set -- 7-bit American Standard Code for Information Interchange</title>
4128    <author>
4129      <organization>American National Standards Institute</organization>
4130    </author>
4131    <date year="1986"/>
4132  </front>
4133  <seriesInfo name="ANSI" value="X3.4"/>
4136<reference anchor="RFC1950">
4137  <front>
4138    <title>ZLIB Compressed Data Format Specification version 3.3</title>
4139    <author initials="L.P." surname="Deutsch" fullname="L. Peter Deutsch">
4140      <organization>Aladdin Enterprises</organization>
4141      <address><email></email></address>
4142    </author>
4143    <author initials="J-L." surname="Gailly" fullname="Jean-Loup Gailly"/>
4144    <date month="May" year="1996"/>
4145  </front>
4146  <seriesInfo name="RFC" value="1950"/>
4147  <!--<annotation>
4148    RFC 1950 is an Informational RFC, thus it might be less stable than
4149    this specification. On the other hand, this downward reference was
4150    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4151    therefore it is unlikely to cause problems in practice. See also
4152    <xref target="BCP97"/>.
4153  </annotation>-->
4156<reference anchor="RFC1951">
4157  <front>
4158    <title>DEFLATE Compressed Data Format Specification version 1.3</title>
4159    <author initials="P." surname="Deutsch" fullname="L. Peter Deutsch">
4160      <organization>Aladdin Enterprises</organization>
4161      <address><email></email></address>
4162    </author>
4163    <date month="May" year="1996"/>
4164  </front>
4165  <seriesInfo name="RFC" value="1951"/>
4166  <!--<annotation>
4167    RFC 1951 is an Informational RFC, thus it might be less stable than
4168    this specification. On the other hand, this downward reference was
4169    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4170    therefore it is unlikely to cause problems in practice. See also
4171    <xref target="BCP97"/>.
4172  </annotation>-->
4175<reference anchor="RFC1952">
4176  <front>
4177    <title>GZIP file format specification version 4.3</title>
4178    <author initials="P." surname="Deutsch" fullname="L. Peter Deutsch">
4179      <organization>Aladdin Enterprises</organization>
4180      <address><email></email></address>
4181    </author>
4182    <author initials="J-L." surname="Gailly" fullname="Jean-Loup Gailly">
4183      <address><email></email></address>
4184    </author>
4185    <author initials="M." surname="Adler" fullname="Mark Adler">
4186      <address><email></email></address>
4187    </author>
4188    <author initials="L.P." surname="Deutsch" fullname="L. Peter Deutsch">
4189      <address><email></email></address>
4190    </author>
4191    <author initials="G." surname="Randers-Pehrson" fullname="Glenn Randers-Pehrson">
4192      <address><email></email></address>
4193    </author>
4194    <date month="May" year="1996"/>
4195  </front>
4196  <seriesInfo name="RFC" value="1952"/>
4197  <!--<annotation>
4198    RFC 1952 is an Informational RFC, thus it might be less stable than
4199    this specification. On the other hand, this downward reference was
4200    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4201    therefore it is unlikely to cause problems in practice. See also
4202    <xref target="BCP97"/>.
4203  </annotation>-->
4208<references title="Informative References">
4210<reference anchor="ISO-8859-1">
4211  <front>
4212    <title>
4213     Information technology -- 8-bit single-byte coded graphic character sets -- Part 1: Latin alphabet No. 1
4214    </title>
4215    <author>
4216      <organization>International Organization for Standardization</organization>
4217    </author>
4218    <date year="1998"/>
4219  </front>
4220  <seriesInfo name="ISO/IEC" value="8859-1:1998"/>
4223<reference anchor='RFC1919'>
4224  <front>
4225    <title>Classical versus Transparent IP Proxies</title>
4226    <author initials='M.' surname='Chatel' fullname='Marc Chatel'>
4227      <address><email></email></address>
4228    </author>
4229    <date year='1996' month='March' />
4230  </front>
4231  <seriesInfo name='RFC' value='1919' />
4234<reference anchor="RFC1945">
4235  <front>
4236    <title abbrev="HTTP/1.0">Hypertext Transfer Protocol -- HTTP/1.0</title>
4237    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
4238      <organization>MIT, Laboratory for Computer Science</organization>
4239      <address><email></email></address>
4240    </author>
4241    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
4242      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
4243      <address><email></email></address>
4244    </author>
4245    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4246      <organization>W3 Consortium, MIT Laboratory for Computer Science</organization>
4247      <address><email></email></address>
4248    </author>
4249    <date month="May" year="1996"/>
4250  </front>
4251  <seriesInfo name="RFC" value="1945"/>
4254<reference anchor="RFC2045">
4255  <front>
4256    <title abbrev="Internet Message Bodies">Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies</title>
4257    <author initials="N." surname="Freed" fullname="Ned Freed">
4258      <organization>Innosoft International, Inc.</organization>
4259      <address><email></email></address>
4260    </author>
4261    <author initials="N.S." surname="Borenstein" fullname="Nathaniel S. Borenstein">
4262      <organization>First Virtual Holdings</organization>
4263      <address><email></email></address>
4264    </author>
4265    <date month="November" year="1996"/>
4266  </front>
4267  <seriesInfo name="RFC" value="2045"/>
4270<reference anchor="RFC2047">
4271  <front>
4272    <title abbrev="Message Header Extensions">MIME (Multipurpose Internet Mail Extensions) Part Three: Message Header Extensions for Non-ASCII Text</title>
4273    <author initials="K." surname="Moore" fullname="Keith Moore">
4274      <organization>University of Tennessee</organization>
4275      <address><email></email></address>
4276    </author>
4277    <date month="November" year="1996"/>
4278  </front>
4279  <seriesInfo name="RFC" value="2047"/>
4282<reference anchor="RFC2068">
4283  <front>
4284    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
4285    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
4286      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
4287      <address><email></email></address>
4288    </author>
4289    <author initials="J." surname="Gettys" fullname="Jim Gettys">
4290      <organization>MIT Laboratory for Computer Science</organization>
4291      <address><email></email></address>
4292    </author>
4293    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
4294      <organization>Digital Equipment Corporation, Western Research Laboratory</organization>
4295      <address><email></email></address>
4296    </author>
4297    <author initials="H." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4298      <organization>MIT Laboratory for Computer Science</organization>
4299      <address><email></email></address>
4300    </author>
4301    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
4302      <organization>MIT Laboratory for Computer Science</organization>
4303      <address><email></email></address>
4304    </author>
4305    <date month="January" year="1997"/>
4306  </front>
4307  <seriesInfo name="RFC" value="2068"/>
4310<reference anchor="RFC2145">
4311  <front>
4312    <title abbrev="HTTP Version Numbers">Use and Interpretation of HTTP Version Numbers</title>
4313    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
4314      <organization>Western Research Laboratory</organization>
4315      <address><email></email></address>
4316    </author>
4317    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
4318      <organization>Department of Information and Computer Science</organization>
4319      <address><email></email></address>
4320    </author>
4321    <author initials="J." surname="Gettys" fullname="Jim Gettys">
4322      <organization>MIT Laboratory for Computer Science</organization>
4323      <address><email></email></address>
4324    </author>
4325    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4326      <organization>W3 Consortium</organization>
4327      <address><email></email></address>
4328    </author>
4329    <date month="May" year="1997"/>
4330  </front>
4331  <seriesInfo name="RFC" value="2145"/>
4334<reference anchor="RFC2616">
4335  <front>
4336    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
4337    <author initials="R." surname="Fielding" fullname="R. Fielding">
4338      <organization>University of California, Irvine</organization>
4339      <address><email></email></address>
4340    </author>
4341    <author initials="J." surname="Gettys" fullname="J. Gettys">
4342      <organization>W3C</organization>
4343      <address><email></email></address>
4344    </author>
4345    <author initials="J." surname="Mogul" fullname="J. Mogul">
4346      <organization>Compaq Computer Corporation</organization>
4347      <address><email></email></address>
4348    </author>
4349    <author initials="H." surname="Frystyk" fullname="H. Frystyk">
4350      <organization>MIT Laboratory for Computer Science</organization>
4351      <address><email></email></address>
4352    </author>
4353    <author initials="L." surname="Masinter" fullname="L. Masinter">
4354      <organization>Xerox Corporation</organization>
4355      <address><email></email></address>
4356    </author>
4357    <author initials="P." surname="Leach" fullname="P. Leach">
4358      <organization>Microsoft Corporation</organization>
4359      <address><email></email></address>
4360    </author>
4361    <author initials="T." surname="Berners-Lee" fullname="T. Berners-Lee">
4362      <organization>W3C</organization>
4363      <address><email></email></address>
4364    </author>
4365    <date month="June" year="1999"/>
4366  </front>
4367  <seriesInfo name="RFC" value="2616"/>
4370<reference anchor='RFC2817'>
4371  <front>
4372    <title>Upgrading to TLS Within HTTP/1.1</title>
4373    <author initials='R.' surname='Khare' fullname='R. Khare'>
4374      <organization>4K Associates / UC Irvine</organization>
4375      <address><email></email></address>
4376    </author>
4377    <author initials='S.' surname='Lawrence' fullname='S. Lawrence'>
4378      <organization>Agranat Systems, Inc.</organization>
4379      <address><email></email></address>
4380    </author>
4381    <date year='2000' month='May' />
4382  </front>
4383  <seriesInfo name='RFC' value='2817' />
4386<reference anchor='RFC2818'>
4387  <front>
4388    <title>HTTP Over TLS</title>
4389    <author initials='E.' surname='Rescorla' fullname='Eric Rescorla'>
4390      <organization>RTFM, Inc.</organization>
4391      <address><email></email></address>
4392    </author>
4393    <date year='2000' month='May' />
4394  </front>
4395  <seriesInfo name='RFC' value='2818' />
4398<reference anchor='RFC2965'>
4399  <front>
4400    <title>HTTP State Management Mechanism</title>
4401    <author initials='D. M.' surname='Kristol' fullname='David M. Kristol'>
4402      <organization>Bell Laboratories, Lucent Technologies</organization>
4403      <address><email></email></address>
4404    </author>
4405    <author initials='L.' surname='Montulli' fullname='Lou Montulli'>
4406      <organization>, Inc.</organization>
4407      <address><email></email></address>
4408    </author>
4409    <date year='2000' month='October' />
4410  </front>
4411  <seriesInfo name='RFC' value='2965' />
4414<reference anchor='RFC3040'>
4415  <front>
4416    <title>Internet Web Replication and Caching Taxonomy</title>
4417    <author initials='I.' surname='Cooper' fullname='I. Cooper'>
4418      <organization>Equinix, Inc.</organization>
4419    </author>
4420    <author initials='I.' surname='Melve' fullname='I. Melve'>
4421      <organization>UNINETT</organization>
4422    </author>
4423    <author initials='G.' surname='Tomlinson' fullname='G. Tomlinson'>
4424      <organization>CacheFlow Inc.</organization>
4425    </author>
4426    <date year='2001' month='January' />
4427  </front>
4428  <seriesInfo name='RFC' value='3040' />
4431<reference anchor='RFC3864'>
4432  <front>
4433    <title>Registration Procedures for Message Header Fields</title>
4434    <author initials='G.' surname='Klyne' fullname='G. Klyne'>
4435      <organization>Nine by Nine</organization>
4436      <address><email></email></address>
4437    </author>
4438    <author initials='M.' surname='Nottingham' fullname='M. Nottingham'>
4439      <organization>BEA Systems</organization>
4440      <address><email></email></address>
4441    </author>
4442    <author initials='J.' surname='Mogul' fullname='J. Mogul'>
4443      <organization>HP Labs</organization>
4444      <address><email></email></address>
4445    </author>
4446    <date year='2004' month='September' />
4447  </front>
4448  <seriesInfo name='BCP' value='90' />
4449  <seriesInfo name='RFC' value='3864' />
4452<reference anchor='RFC4033'>
4453  <front>
4454    <title>DNS Security Introduction and Requirements</title>
4455    <author initials='R.' surname='Arends' fullname='R. Arends'/>
4456    <author initials='R.' surname='Austein' fullname='R. Austein'/>
4457    <author initials='M.' surname='Larson' fullname='M. Larson'/>
4458    <author initials='D.' surname='Massey' fullname='D. Massey'/>
4459    <author initials='S.' surname='Rose' fullname='S. Rose'/>
4460    <date year='2005' month='March' />
4461  </front>
4462  <seriesInfo name='RFC' value='4033' />
4465<reference anchor="RFC4288">
4466  <front>
4467    <title>Media Type Specifications and Registration Procedures</title>
4468    <author initials="N." surname="Freed" fullname="N. Freed">
4469      <organization>Sun Microsystems</organization>
4470      <address>
4471        <email></email>
4472      </address>
4473    </author>
4474    <author initials="J." surname="Klensin" fullname="J. Klensin">
4475      <address>
4476        <email></email>
4477      </address>
4478    </author>
4479    <date year="2005" month="December"/>
4480  </front>
4481  <seriesInfo name="BCP" value="13"/>
4482  <seriesInfo name="RFC" value="4288"/>
4485<reference anchor='RFC4395'>
4486  <front>
4487    <title>Guidelines and Registration Procedures for New URI Schemes</title>
4488    <author initials='T.' surname='Hansen' fullname='T. Hansen'>
4489      <organization>AT&amp;T Laboratories</organization>
4490      <address>
4491        <email></email>
4492      </address>
4493    </author>
4494    <author initials='T.' surname='Hardie' fullname='T. Hardie'>
4495      <organization>Qualcomm, Inc.</organization>
4496      <address>
4497        <email></email>
4498      </address>
4499    </author>
4500    <author initials='L.' surname='Masinter' fullname='L. Masinter'>
4501      <organization>Adobe Systems</organization>
4502      <address>
4503        <email></email>
4504      </address>
4505    </author>
4506    <date year='2006' month='February' />
4507  </front>
4508  <seriesInfo name='BCP' value='115' />
4509  <seriesInfo name='RFC' value='4395' />
4512<reference anchor='RFC4559'>
4513  <front>
4514    <title>SPNEGO-based Kerberos and NTLM HTTP Authentication in Microsoft Windows</title>
4515    <author initials='K.' surname='Jaganathan' fullname='K. Jaganathan'/>
4516    <author initials='L.' surname='Zhu' fullname='L. Zhu'/>
4517    <author initials='J.' surname='Brezak' fullname='J. Brezak'/>
4518    <date year='2006' month='June' />
4519  </front>
4520  <seriesInfo name='RFC' value='4559' />
4523<reference anchor='RFC5226'>
4524  <front>
4525    <title>Guidelines for Writing an IANA Considerations Section in RFCs</title>
4526    <author initials='T.' surname='Narten' fullname='T. Narten'>
4527      <organization>IBM</organization>
4528      <address><email></email></address>
4529    </author>
4530    <author initials='H.' surname='Alvestrand' fullname='H. Alvestrand'>
4531      <organization>Google</organization>
4532      <address><email></email></address>
4533    </author>
4534    <date year='2008' month='May' />
4535  </front>
4536  <seriesInfo name='BCP' value='26' />
4537  <seriesInfo name='RFC' value='5226' />
4540<reference anchor='RFC5246'>
4541   <front>
4542      <title>The Transport Layer Security (TLS) Protocol Version 1.2</title>
4543      <author initials='T.' surname='Dierks' fullname='T. Dierks'>
4544         <organization />
4545      </author>
4546      <author initials='E.' surname='Rescorla' fullname='E. Rescorla'>
4547         <organization>RTFM, Inc.</organization>
4548      </author>
4549      <date year='2008' month='August' />
4550   </front>
4551   <seriesInfo name='RFC' value='5246' />
4554<reference anchor="RFC5322">
4555  <front>
4556    <title>Internet Message Format</title>
4557    <author initials="P." surname="Resnick" fullname="P. Resnick">
4558      <organization>Qualcomm Incorporated</organization>
4559    </author>
4560    <date year="2008" month="October"/>
4561  </front>
4562  <seriesInfo name="RFC" value="5322"/>
4565<reference anchor="RFC6265">
4566  <front>
4567    <title>HTTP State Management Mechanism</title>
4568    <author initials="A." surname="Barth" fullname="Adam Barth">
4569      <organization abbrev="U.C. Berkeley">
4570        University of California, Berkeley
4571      </organization>
4572      <address><email></email></address>
4573    </author>
4574    <date year="2011" month="April" />
4575  </front>
4576  <seriesInfo name="RFC" value="6265"/>
4579<!--<reference anchor='BCP97'>
4580  <front>
4581    <title>Handling Normative References to Standards-Track Documents</title>
4582    <author initials='J.' surname='Klensin' fullname='J. Klensin'>
4583      <address>
4584        <email></email>
4585      </address>
4586    </author>
4587    <author initials='S.' surname='Hartman' fullname='S. Hartman'>
4588      <organization>MIT</organization>
4589      <address>
4590        <email></email>
4591      </address>
4592    </author>
4593    <date year='2007' month='June' />
4594  </front>
4595  <seriesInfo name='BCP' value='97' />
4596  <seriesInfo name='RFC' value='4897' />
4599<reference anchor="Kri2001" target="">
4600  <front>
4601    <title>HTTP Cookies: Standards, Privacy, and Politics</title>
4602    <author initials="D." surname="Kristol" fullname="David M. Kristol"/>
4603    <date year="2001" month="November"/>
4604  </front>
4605  <seriesInfo name="ACM Transactions on Internet Technology" value="Vol. 1, #2"/>
4611<section title="HTTP Version History" anchor="compatibility">
4613   HTTP has been in use by the World-Wide Web global information initiative
4614   since 1990. The first version of HTTP, later referred to as HTTP/0.9,
4615   was a simple protocol for hypertext data transfer across the Internet
4616   with only a single request method (GET) and no metadata.
4617   HTTP/1.0, as defined by <xref target="RFC1945"/>, added a range of request
4618   methods and MIME-like messaging that could include metadata about the data
4619   transferred and modifiers on the request/response semantics. However,
4620   HTTP/1.0 did not sufficiently take into consideration the effects of
4621   hierarchical proxies, caching, the need for persistent connections, or
4622   name-based virtual hosts. The proliferation of incompletely-implemented
4623   applications calling themselves "HTTP/1.0" further necessitated a
4624   protocol version change in order for two communicating applications
4625   to determine each other's true capabilities.
4628   HTTP/1.1 remains compatible with HTTP/1.0 by including more stringent
4629   requirements that enable reliable implementations, adding only
4630   those new features that will either be safely ignored by an HTTP/1.0
4631   recipient or only sent when communicating with a party advertising
4632   conformance with HTTP/1.1.
4635   It is beyond the scope of a protocol specification to mandate
4636   conformance with previous versions. HTTP/1.1 was deliberately
4637   designed, however, to make supporting previous versions easy.
4638   We would expect a general-purpose HTTP/1.1 server to understand
4639   any valid request in the format of HTTP/1.0 and respond appropriately
4640   with an HTTP/1.1 message that only uses features understood (or
4641   safely ignored) by HTTP/1.0 clients.  Likewise, we would expect
4642   an HTTP/1.1 client to understand any valid HTTP/1.0 response.
4645   Since HTTP/0.9 did not support header fields in a request,
4646   there is no mechanism for it to support name-based virtual
4647   hosts (selection of resource by inspection of the <x:ref>Host</x:ref> header
4648   field).  Any server that implements name-based virtual hosts
4649   ought to disable support for HTTP/0.9.  Most requests that
4650   appear to be HTTP/0.9 are, in fact, badly constructed HTTP/1.x
4651   requests wherein a buggy client failed to properly encode
4652   linear whitespace found in a URI reference and placed in
4653   the request-target.
4656<section title="Changes from HTTP/1.0" anchor="changes.from.1.0">
4658   This section summarizes major differences between versions HTTP/1.0
4659   and HTTP/1.1.
4662<section title="Multi-homed Web Servers" anchor="">
4664   The requirements that clients and servers support the <x:ref>Host</x:ref>
4665   header field (<xref target=""/>), report an error if it is
4666   missing from an HTTP/1.1 request, and accept absolute URIs (<xref target="request-target"/>)
4667   are among the most important changes defined by HTTP/1.1.
4670   Older HTTP/1.0 clients assumed a one-to-one relationship of IP
4671   addresses and servers; there was no other established mechanism for
4672   distinguishing the intended server of a request than the IP address
4673   to which that request was directed. The <x:ref>Host</x:ref> header field was
4674   introduced during the development of HTTP/1.1 and, though it was
4675   quickly implemented by most HTTP/1.0 browsers, additional requirements
4676   were placed on all HTTP/1.1 requests in order to ensure complete
4677   adoption.  At the time of this writing, most HTTP-based services
4678   are dependent upon the Host header field for targeting requests.
4682<section title="Keep-Alive Connections" anchor="compatibility.with.http.1.0.persistent.connections">
4684   In HTTP/1.0, each connection is established by the client prior to the
4685   request and closed by the server after sending the response. However, some
4686   implementations implement the explicitly negotiated ("Keep-Alive") version
4687   of persistent connections described in <xref x:sec="19.7.1" x:fmt="of"
4688   target="RFC2068"/>.
4691   Some clients and servers might wish to be compatible with these previous
4692   approaches to persistent connections, by explicitly negotiating for them
4693   with a "Connection: keep-alive" request header field. However, some
4694   experimental implementations of HTTP/1.0 persistent connections are faulty;
4695   for example, if a HTTP/1.0 proxy server doesn't understand
4696   <x:ref>Connection</x:ref>, it will erroneously forward that header field
4697   to the next inbound server, which would result in a hung connection.
4700   One attempted solution was the introduction of a Proxy-Connection header
4701   field, targeted specifically at proxies. In practice, this was also
4702   unworkable, because proxies are often deployed in multiple layers, bringing
4703   about the same problem discussed above.
4706   As a result, clients are encouraged not to send the Proxy-Connection header
4707   field in any requests.
4710   Clients are also encouraged to consider the use of Connection: keep-alive
4711   in requests carefully; while they can enable persistent connections with
4712   HTTP/1.0 servers, clients using them need will need to monitor the
4713   connection for "hung" requests (which indicate that the client ought stop
4714   sending the header field), and this mechanism ought not be used by clients
4715   at all when a proxy is being used.
4719<section title="Introduction of Transfer-Encoding" anchor="introduction.of.transfer-encoding">
4721   HTTP/1.1 introduces the <x:ref>Transfer-Encoding</x:ref> header field
4722   (<xref target="header.transfer-encoding"/>). Proxies/gateways &MUST; remove
4723   any transfer-coding prior to forwarding a message via a MIME-compliant
4724   protocol.
4730<section title="Changes from RFC 2616" anchor="changes.from.rfc.2616">
4732  Clarify that the string "HTTP" in the HTTP-version ABNF production is case
4733  sensitive. Restrict the version numbers to be single digits due to the fact
4734  that implementations are known to handle multi-digit version numbers
4735  incorrectly.
4736  (<xref target="http.version"/>)
4739  Require that invalid whitespace around field-names be rejected.
4740  Change ABNF productions for header fields to only define the field value.
4741  (<xref target="header.fields"/>)
4744  Rules about implicit linear whitespace between certain grammar productions
4745  have been removed; now whitespace is only allowed where specifically
4746  defined in the ABNF.
4747  (<xref target="whitespace"/>)
4750  The NUL octet is no longer allowed in comment and quoted-string
4751  text. The quoted-pair rule no longer allows escaping control characters other than HTAB.
4752  Non-ASCII content in header fields and reason phrase has been obsoleted and
4753  made opaque (the TEXT rule was removed).
4754  (<xref target="field.components"/>)
4757  Require recipients to handle bogus "<x:ref>Content-Length</x:ref>" header
4758  fields as errors.
4759  (<xref target="message.body"/>)
4762  Remove reference to non-existent identity transfer-coding value tokens.
4763  (Sections <xref format="counter" target="message.body"/> and
4764  <xref format="counter" target="transfer.codings"/>)
4767  Clarification that the chunk length does not include the count of the octets
4768  in the chunk header and trailer. Furthermore disallowed line folding
4769  in chunk extensions, and deprecate their use.
4770  (<xref target="chunked.encoding"/>)
4773  Update use of abs_path production from RFC 1808 to the path-absolute + query
4774  components of RFC 3986. State that the asterisk form is allowed for the OPTIONS
4775  request method only.
4776  (<xref target="request-target"/>)
4779  Clarify exactly when "close" connection options have to be sent; drop
4780  notion of header fields being "hop-by-hop" without being listed in the
4781  Connection header field.
4782  (<xref target="header.connection"/>)
4785  Remove hard limit of two connections per server.
4786  Remove requirement to retry a sequence of requests as long it was idempotent.
4787  Remove requirements about when servers are allowed to close connections
4788  prematurely.
4789  (<xref target="persistent.connections"/>)
4792  Remove requirement to retry requests under certain circumstances when the
4793  server prematurely closes the connection.
4794  (<xref target="persistent.reuse"/>)
4797  Define the semantics of the <x:ref>Upgrade</x:ref> header field in responses
4798  other than 101 (this was incorporated from <xref target="RFC2817"/>).
4799  (<xref target="header.upgrade"/>)
4802  Registration of Transfer Codings now requires IETF Review
4803  (<xref target="transfer.coding.registry"/>)
4806  Take over the Upgrade Token Registry, previously defined in
4807  <xref target="RFC2817" x:fmt="of" x:sec="7.2"/>.
4808  (<xref target="upgrade.token.registry"/>)
4811  Empty list elements in list productions have been deprecated.
4812  (<xref target="abnf.extension"/>)
4817<section title="ABNF list extension: #rule" anchor="abnf.extension">
4819  A #rule extension to the ABNF rules of <xref target="RFC5234"/> is used to
4820  improve readability in the definitions of some header field values.
4823  A construct "#" is defined, similar to "*", for defining comma-delimited
4824  lists of elements. The full form is "&lt;n&gt;#&lt;m&gt;element" indicating
4825  at least &lt;n&gt; and at most &lt;m&gt; elements, each separated by a single
4826  comma (",") and optional whitespace (OWS).   
4829  Thus,
4830</preamble><artwork type="example">
4831  1#element =&gt; element *( OWS "," OWS element )
4834  and:
4835</preamble><artwork type="example">
4836  #element =&gt; [ 1#element ]
4839  and for n &gt;= 1 and m &gt; 1:
4840</preamble><artwork type="example">
4841  &lt;n&gt;#&lt;m&gt;element =&gt; element &lt;n-1&gt;*&lt;m-1&gt;( OWS "," OWS element )
4844  For compatibility with legacy list rules, recipients &SHOULD; accept empty
4845  list elements. In other words, consumers would follow the list productions:
4847<figure><artwork type="example">
4848  #element =&gt; [ ( "," / element ) *( OWS "," [ OWS element ] ) ]
4850  1#element =&gt; *( "," OWS ) element *( OWS "," [ OWS element ] )
4853  Note that empty elements do not contribute to the count of elements present,
4854  though.
4857  For example, given these ABNF productions:
4859<figure><artwork type="example">
4860  example-list      = 1#example-list-elmt
4861  example-list-elmt = token ; see <xref target="field.components"/>
4864  Then these are valid values for example-list (not including the double
4865  quotes, which are present for delimitation only):
4867<figure><artwork type="example">
4868  "foo,bar"
4869  "foo ,bar,"
4870  "foo , ,bar,charlie   "
4873  But these values would be invalid, as at least one non-empty element is
4874  required:
4876<figure><artwork type="example">
4877  ""
4878  ","
4879  ",   ,"
4882  <xref target="collected.abnf"/> shows the collected ABNF, with the list rules
4883  expanded as explained above.
4887<?BEGININC p1-messaging.abnf-appendix ?>
4888<section xmlns:x="" title="Collected ABNF" anchor="collected.abnf">
4890<artwork type="abnf" name="p1-messaging.parsed-abnf">
4891<x:ref>BWS</x:ref> = OWS
4893<x:ref>Connection</x:ref> = *( "," OWS ) connection-option *( OWS "," [ OWS
4894 connection-option ] )
4895<x:ref>Content-Length</x:ref> = 1*DIGIT
4897<x:ref>HTTP-message</x:ref> = start-line *( header-field CRLF ) CRLF [ message-body
4898 ]
4899<x:ref>HTTP-name</x:ref> = %x48.54.54.50 ; HTTP
4900<x:ref>HTTP-version</x:ref> = HTTP-name "/" DIGIT "." DIGIT
4901<x:ref>Host</x:ref> = uri-host [ ":" port ]
4903<x:ref>OWS</x:ref> = *( SP / HTAB )
4905<x:ref>RWS</x:ref> = 1*( SP / HTAB )
4907<x:ref>TE</x:ref> = [ ( "," / t-codings ) *( OWS "," [ OWS t-codings ] ) ]
4908<x:ref>Trailer</x:ref> = *( "," OWS ) field-name *( OWS "," [ OWS field-name ] )
4909<x:ref>Transfer-Encoding</x:ref> = *( "," OWS ) transfer-coding *( OWS "," [ OWS
4910 transfer-coding ] )
4912<x:ref>URI-reference</x:ref> = &lt;URI-reference, defined in [RFC3986], Section 4.1&gt;
4913<x:ref>Upgrade</x:ref> = *( "," OWS ) protocol *( OWS "," [ OWS protocol ] )
4915<x:ref>Via</x:ref> = *( "," OWS ) ( received-protocol RWS received-by [ RWS comment
4916 ] ) *( OWS "," [ OWS ( received-protocol RWS received-by [ RWS
4917 comment ] ) ] )
4919<x:ref>absolute-URI</x:ref> = &lt;absolute-URI, defined in [RFC3986], Section 4.3&gt;
4920<x:ref>absolute-form</x:ref> = absolute-URI
4921<x:ref>asterisk-form</x:ref> = "*"
4922<x:ref>attribute</x:ref> = token
4923<x:ref>authority</x:ref> = &lt;authority, defined in [RFC3986], Section 3.2&gt;
4924<x:ref>authority-form</x:ref> = authority
4926<x:ref>chunk</x:ref> = chunk-size [ chunk-ext ] CRLF chunk-data CRLF
4927<x:ref>chunk-data</x:ref> = 1*OCTET
4928<x:ref>chunk-ext</x:ref> = *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
4929<x:ref>chunk-ext-name</x:ref> = token
4930<x:ref>chunk-ext-val</x:ref> = token / quoted-str-nf
4931<x:ref>chunk-size</x:ref> = 1*HEXDIG
4932<x:ref>chunked-body</x:ref> = *chunk last-chunk trailer-part CRLF
4933<x:ref>comment</x:ref> = "(" *( ctext / quoted-cpair / comment ) ")"
4934<x:ref>connection-option</x:ref> = token
4935<x:ref>ctext</x:ref> = HTAB / SP / %x21-27 ; '!'-'''
4936 / %x2A-5B ; '*'-'['
4937 / %x5D-7E ; ']'-'~'
4938 / obs-text
4940<x:ref>field-content</x:ref> = *( HTAB / SP / VCHAR / obs-text )
4941<x:ref>field-name</x:ref> = token
4942<x:ref>field-value</x:ref> = *( field-content / obs-fold )
4944<x:ref>header-field</x:ref> = field-name ":" OWS field-value BWS
4945<x:ref>http-URI</x:ref> = "http://" authority path-abempty [ "?" query ]
4946<x:ref>https-URI</x:ref> = "https://" authority path-abempty [ "?" query ]
4948<x:ref>last-chunk</x:ref> = 1*"0" [ chunk-ext ] CRLF
4950<x:ref>message-body</x:ref> = *OCTET
4951<x:ref>method</x:ref> = token
4953<x:ref>obs-fold</x:ref> = CRLF ( SP / HTAB )
4954<x:ref>obs-text</x:ref> = %x80-FF
4955<x:ref>origin-form</x:ref> = path-absolute [ "?" query ]
4957<x:ref>partial-URI</x:ref> = relative-part [ "?" query ]
4958<x:ref>path-abempty</x:ref> = &lt;path-abempty, defined in [RFC3986], Section 3.3&gt;
4959<x:ref>path-absolute</x:ref> = &lt;path-absolute, defined in [RFC3986], Section 3.3&gt;
4960<x:ref>port</x:ref> = &lt;port, defined in [RFC3986], Section 3.2.3&gt;
4961<x:ref>protocol</x:ref> = protocol-name [ "/" protocol-version ]
4962<x:ref>protocol-name</x:ref> = token
4963<x:ref>protocol-version</x:ref> = token
4964<x:ref>pseudonym</x:ref> = token
4966<x:ref>qdtext</x:ref> = HTAB / SP / "!" / %x23-5B ; '#'-'['
4967 / %x5D-7E ; ']'-'~'
4968 / obs-text
4969<x:ref>qdtext-nf</x:ref> = HTAB / SP / "!" / %x23-5B ; '#'-'['
4970 / %x5D-7E ; ']'-'~'
4971 / obs-text
4972<x:ref>query</x:ref> = &lt;query, defined in [RFC3986], Section 3.4&gt;
4973<x:ref>quoted-cpair</x:ref> = "\" ( HTAB / SP / VCHAR / obs-text )
4974<x:ref>quoted-pair</x:ref> = "\" ( HTAB / SP / VCHAR / obs-text )
4975<x:ref>quoted-str-nf</x:ref> = DQUOTE *( qdtext-nf / quoted-pair ) DQUOTE
4976<x:ref>quoted-string</x:ref> = DQUOTE *( qdtext / quoted-pair ) DQUOTE
4978<x:ref>rank</x:ref> = ( "0" [ "." *3DIGIT ] ) / ( "1" [ "." *3"0" ] )
4979<x:ref>reason-phrase</x:ref> = *( HTAB / SP / VCHAR / obs-text )
4980<x:ref>received-by</x:ref> = ( uri-host [ ":" port ] ) / pseudonym
4981<x:ref>received-protocol</x:ref> = [ protocol-name "/" ] protocol-version
4982<x:ref>relative-part</x:ref> = &lt;relative-part, defined in [RFC3986], Section 4.2&gt;
4983<x:ref>request-line</x:ref> = method SP request-target SP HTTP-version CRLF
4984<x:ref>request-target</x:ref> = origin-form / absolute-form / authority-form /
4985 asterisk-form
4987<x:ref>special</x:ref> = "(" / ")" / "&lt;" / "&gt;" / "@" / "," / ";" / ":" / "\" /
4988 DQUOTE / "/" / "[" / "]" / "?" / "=" / "{" / "}"
4989<x:ref>start-line</x:ref> = request-line / status-line
4990<x:ref>status-code</x:ref> = 3DIGIT
4991<x:ref>status-line</x:ref> = HTTP-version SP status-code SP reason-phrase CRLF
4993<x:ref>t-codings</x:ref> = "trailers" / ( transfer-coding [ t-ranking ] )
4994<x:ref>t-ranking</x:ref> = OWS ";" OWS "q=" rank
4995<x:ref>tchar</x:ref> = "!" / "#" / "$" / "%" / "&amp;" / "'" / "*" / "+" / "-" / "." /
4996 "^" / "_" / "`" / "|" / "~" / DIGIT / ALPHA
4997<x:ref>token</x:ref> = 1*tchar
4998<x:ref>trailer-part</x:ref> = *( header-field CRLF )
4999<x:ref>transfer-coding</x:ref> = "chunked" / "compress" / "deflate" / "gzip" /
5000 transfer-extension
5001<x:ref>transfer-extension</x:ref> = token *( OWS ";" OWS transfer-parameter )
5002<x:ref>transfer-parameter</x:ref> = attribute BWS "=" BWS value
5004<x:ref>uri-host</x:ref> = &lt;host, defined in [RFC3986], Section 3.2.2&gt;
5006<x:ref>value</x:ref> = word
5008<x:ref>word</x:ref> = token / quoted-string
5012<?ENDINC p1-messaging.abnf-appendix ?>
5014<section title="Change Log (to be removed by RFC Editor before publication)" anchor="change.log">
5016<section title="Since RFC 2616">
5018  Changes up to the first Working Group Last Call draft are summarized
5019  in <eref target=""/>.
5023<section title="Since draft-ietf-httpbis-p1-messaging-21" anchor="changes.since.21">
5025  None yet.
Note: See TracBrowser for help on using the repository browser.