source: draft-ietf-httpbis/latest/p1-messaging.xml @ 2252

Last change on this file since 2252 was 2252, checked in by fielding@…, 10 years ago

(editorial) #433 make example a more self-descriptive message; replaces [2185]

  • Property svn:eol-style set to native
  • Property svn:mime-type set to text/xml
File size: 227.0 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 "May">
16  <!ENTITY ID-YEAR "2013">
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-type            "<xref target='Part2' x:rel='#media.type' 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='#representations' 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 safe-methods           "<xref target='Part2' x:rel='#safe.methods' xmlns:x=''/>">
47  <!ENTITY methods                "<xref target='Part2' x:rel='#methods' xmlns:x=''/>">
48  <!ENTITY OPTIONS                "<xref target='Part2' x:rel='#OPTIONS' xmlns:x=''/>">
49  <!ENTITY qvalue                 "<xref target='Part2' x:rel='#quality.values' xmlns:x=''/>">
50  <!ENTITY resource               "<xref target='Part2' x:rel='#resources' xmlns:x=''/>">
51  <!ENTITY status-codes           "<xref target='Part2' x:rel='' xmlns:x=''/>">
52  <!ENTITY status-1xx             "<xref target='Part2' x:rel='#status.1xx' xmlns:x=''/>">
53  <!ENTITY status-203             "<xref target='Part2' x:rel='#status.203' xmlns:x=''/>">
54  <!ENTITY status-3xx             "<xref target='Part2' x:rel='#status.3xx' xmlns:x=''/>">
55  <!ENTITY status-304             "<xref target='Part4' x:rel='#status.304' xmlns:x=''/>">
56  <!ENTITY status-4xx             "<xref target='Part2' x:rel='#status.4xx' xmlns:x=''/>">
57  <!ENTITY status-414             "<xref target='Part2' x:rel='#status.414' xmlns:x=''/>">
58  <!ENTITY iana-header-registry   "<xref target='Part2' x:rel='#header.field.registry' xmlns:x=''/>">
60<?rfc toc="yes" ?>
61<?rfc symrefs="yes" ?>
62<?rfc sortrefs="yes" ?>
63<?rfc compact="yes"?>
64<?rfc subcompact="no" ?>
65<?rfc linkmailto="no" ?>
66<?rfc editing="no" ?>
67<?rfc comments="yes"?>
68<?rfc inline="yes"?>
69<?rfc rfcedstyle="yes"?>
70<?rfc-ext allow-markup-in-artwork="yes" ?>
71<?rfc-ext include-references-in-index="yes" ?>
72<rfc obsoletes="2145,2616" updates="2817,2818" category="std" x:maturity-level="proposed"
73     ipr="pre5378Trust200902" docName="draft-ietf-httpbis-p1-messaging-&ID-VERSION;"
74     xmlns:x=''>
75<x:link rel="next" basename="p2-semantics"/>
76<x:feedback template="{docname},%20%22{section}%22&amp;body=&lt;{ref}&gt;:"/>
79  <title abbrev="HTTP/1.1 Message Syntax and Routing">Hypertext Transfer Protocol (HTTP/1.1): Message Syntax and Routing</title>
81  <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
82    <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
83    <address>
84      <postal>
85        <street>345 Park Ave</street>
86        <city>San Jose</city>
87        <region>CA</region>
88        <code>95110</code>
89        <country>USA</country>
90      </postal>
91      <email></email>
92      <uri></uri>
93    </address>
94  </author>
96  <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
97    <organization abbrev="greenbytes">greenbytes GmbH</organization>
98    <address>
99      <postal>
100        <street>Hafenweg 16</street>
101        <city>Muenster</city><region>NW</region><code>48155</code>
102        <country>Germany</country>
103      </postal>
104      <email></email>
105      <uri></uri>
106    </address>
107  </author>
109  <date month="&ID-MONTH;" year="&ID-YEAR;"/>
110  <workgroup>HTTPbis Working Group</workgroup>
114   The Hypertext Transfer Protocol (HTTP) is an application-level protocol for
115   distributed, collaborative, hypertext information systems. HTTP has been in
116   use by the World Wide Web global information initiative since 1990.
117   This document provides an overview of HTTP architecture and its associated
118   terminology, defines the "http" and "https" Uniform Resource Identifier
119   (URI) schemes, defines the HTTP/1.1 message syntax and parsing requirements,
120   and describes general security concerns for implementations.
124<note title="Editorial Note (To be removed by RFC Editor)">
125  <t>
126    Discussion of this draft takes place on the HTTPBIS working group
127    mailing list (, which is archived at
128    <eref target=""/>.
129  </t>
130  <t>
131    The current issues list is at
132    <eref target=""/> and related
133    documents (including fancy diffs) can be found at
134    <eref target=""/>.
135  </t>
136  <t>
137    The changes in this draft are summarized in <xref target="changes.since.22"/>.
138  </t>
142<section title="Introduction" anchor="introduction">
144   The Hypertext Transfer Protocol (HTTP) is an application-level
145   request/response protocol that uses extensible semantics and self-descriptive
146   message payloads for flexible interaction with network-based hypertext
147   information systems. This document is the first in a series of documents
148   that collectively form the HTTP/1.1 specification:
149   <list style="empty">
150    <t>RFC xxx1: Message Syntax and Routing</t>
151    <t><xref target="Part2" x:fmt="none">RFC xxx2</xref>: Semantics and Content</t>
152    <t><xref target="Part4" x:fmt="none">RFC xxx3</xref>: Conditional Requests</t>
153    <t><xref target="Part5" x:fmt="none">RFC xxx4</xref>: Range Requests</t>
154    <t><xref target="Part6" x:fmt="none">RFC xxx5</xref>: Caching</t>
155    <t><xref target="Part7" x:fmt="none">RFC xxx6</xref>: Authentication</t>
156   </list>
159   This HTTP/1.1 specification obsoletes and moves to historic status
160   <xref target="RFC2616" x:fmt="none">RFC 2616</xref>, its predecessor
161   <xref target="RFC2068" x:fmt="none">RFC 2068</xref>, and
162   <xref target="RFC2145" x:fmt="none">RFC 2145</xref> (on HTTP versioning).
163   This specification also updates the use of CONNECT to establish a tunnel,
164   previously defined in <xref target="RFC2817" x:fmt="none">RFC 2817</xref>,
165   and defines the "https" URI scheme that was described informally in
166   <xref target="RFC2818" x:fmt="none">RFC 2818</xref>.
169   HTTP is a generic interface protocol for information systems. It is
170   designed to hide the details of how a service is implemented by presenting
171   a uniform interface to clients that is independent of the types of
172   resources provided. Likewise, servers do not need to be aware of each
173   client's purpose: an HTTP request can be considered in isolation rather
174   than being associated with a specific type of client or a predetermined
175   sequence of application steps. The result is a protocol that can be used
176   effectively in many different contexts and for which implementations can
177   evolve independently over time.
180   HTTP is also designed for use as an intermediation protocol for translating
181   communication to and from non-HTTP information systems.
182   HTTP proxies and gateways can provide access to alternative information
183   services by translating their diverse protocols into a hypertext
184   format that can be viewed and manipulated by clients in the same way
185   as HTTP services.
188   One consequence of this flexibility is that the protocol cannot be
189   defined in terms of what occurs behind the interface. Instead, we
190   are limited to defining the syntax of communication, the intent
191   of received communication, and the expected behavior of recipients.
192   If the communication is considered in isolation, then successful
193   actions ought to be reflected in corresponding changes to the
194   observable interface provided by servers. However, since multiple
195   clients might act in parallel and perhaps at cross-purposes, we
196   cannot require that such changes be observable beyond the scope
197   of a single response.
200   This document describes the architectural elements that are used or
201   referred to in HTTP, defines the "http" and "https" URI schemes,
202   describes overall network operation and connection management,
203   and defines HTTP message framing and forwarding requirements.
204   Our goal is to define all of the mechanisms necessary for HTTP message
205   handling that are independent of message semantics, thereby defining the
206   complete set of requirements for message parsers and
207   message-forwarding intermediaries.
211<section title="Requirement Notation" anchor="intro.requirements">
213   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
214   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
215   document are to be interpreted as described in <xref target="RFC2119"/>.
218   Conformance criteria and considerations regarding error handling
219   are defined in <xref target="conformance"/>.
223<section title="Syntax Notation" anchor="notation">
224<iref primary="true" item="Grammar" subitem="ALPHA"/>
225<iref primary="true" item="Grammar" subitem="CR"/>
226<iref primary="true" item="Grammar" subitem="CRLF"/>
227<iref primary="true" item="Grammar" subitem="CTL"/>
228<iref primary="true" item="Grammar" subitem="DIGIT"/>
229<iref primary="true" item="Grammar" subitem="DQUOTE"/>
230<iref primary="true" item="Grammar" subitem="HEXDIG"/>
231<iref primary="true" item="Grammar" subitem="HTAB"/>
232<iref primary="true" item="Grammar" subitem="LF"/>
233<iref primary="true" item="Grammar" subitem="OCTET"/>
234<iref primary="true" item="Grammar" subitem="SP"/>
235<iref primary="true" item="Grammar" subitem="VCHAR"/>
237   This specification uses the Augmented Backus-Naur Form (ABNF) notation
238   of <xref target="RFC5234"/> with the list rule extension defined in
239   <xref target="abnf.extension"/>.  <xref target="collected.abnf"/> shows
240   the collected ABNF with the list rule expanded.
242<t anchor="core.rules">
243  <x:anchor-alias value="ALPHA"/>
244  <x:anchor-alias value="CTL"/>
245  <x:anchor-alias value="CR"/>
246  <x:anchor-alias value="CRLF"/>
247  <x:anchor-alias value="DIGIT"/>
248  <x:anchor-alias value="DQUOTE"/>
249  <x:anchor-alias value="HEXDIG"/>
250  <x:anchor-alias value="HTAB"/>
251  <x:anchor-alias value="LF"/>
252  <x:anchor-alias value="OCTET"/>
253  <x:anchor-alias value="SP"/>
254  <x:anchor-alias value="VCHAR"/>
255   The following core rules are included by
256   reference, as defined in <xref target="RFC5234" x:fmt="," x:sec="B.1"/>:
257   ALPHA (letters), CR (carriage return), CRLF (CR LF), CTL (controls),
258   DIGIT (decimal 0-9), DQUOTE (double quote),
259   HEXDIG (hexadecimal 0-9/A-F/a-f), HTAB (horizontal tab), LF (line feed),
260   OCTET (any 8-bit sequence of data), SP (space), and
261   VCHAR (any visible <xref target="USASCII"/> character).
264   As a convention, ABNF rule names prefixed with "obs-" denote
265   "obsolete" grammar rules that appear for historical reasons.
270<section title="Architecture" anchor="architecture">
272   HTTP was created for the World Wide Web architecture
273   and has evolved over time to support the scalability needs of a worldwide
274   hypertext system. Much of that architecture is reflected in the terminology
275   and syntax productions used to define HTTP.
278<section title="Client/Server Messaging" anchor="operation">
279<iref primary="true" item="client"/>
280<iref primary="true" item="server"/>
281<iref primary="true" item="connection"/>
283   HTTP is a stateless request/response protocol that operates by exchanging
284   <x:dfn>messages</x:dfn> (<xref target="http.message"/>) across a reliable
285   transport or session-layer
286   "<x:dfn>connection</x:dfn>" (<xref target=""/>).
287   An HTTP "<x:dfn>client</x:dfn>" is a program that establishes a connection
288   to a server for the purpose of sending one or more HTTP requests.
289   An HTTP "<x:dfn>server</x:dfn>" is a program that accepts connections
290   in order to service HTTP requests by sending HTTP responses.
292<iref primary="true" item="user agent"/>
293<iref primary="true" item="origin server"/>
294<iref primary="true" item="browser"/>
295<iref primary="true" item="spider"/>
296<iref primary="true" item="sender"/>
297<iref primary="true" item="recipient"/>
299   The terms client and server refer only to the roles that
300   these programs perform for a particular connection.  The same program
301   might act as a client on some connections and a server on others.
302   We use the term "<x:dfn>user agent</x:dfn>" to refer to any of the various
303   client programs that initiate a request, including (but not limited to)
304   browsers, spiders (web-based robots), command-line tools, native
305   applications, and mobile apps.  The term "<x:dfn>origin server</x:dfn>" is
306   used to refer to the program that can originate authoritative responses to
307   a request. For general requirements, we use the terms
308   "<x:dfn>sender</x:dfn>" and "<x:dfn>recipient</x:dfn>" to refer to any
309   component that sends or receives, respectively, a given message.
312   HTTP relies upon the Uniform Resource Identifier (URI)
313   standard <xref target="RFC3986"/> to indicate the target resource
314   (<xref target="target-resource"/>) and relationships between resources.
315   Messages are passed in a format similar to that used by Internet mail
316   <xref target="RFC5322"/> and the Multipurpose Internet Mail Extensions
317   (MIME) <xref target="RFC2045"/> (see &diff-mime; for the differences
318   between HTTP and MIME messages).
321   Most HTTP communication consists of a retrieval request (GET) for
322   a representation of some resource identified by a URI.  In the
323   simplest case, this might be accomplished via a single bidirectional
324   connection (===) between the user agent (UA) and the origin server (O).
326<figure><artwork type="drawing">
327         request   &gt;
328    <x:highlight>UA</x:highlight> ======================================= <x:highlight>O</x:highlight>
329                                &lt;   response
331<iref primary="true" item="message"/>
332<iref primary="true" item="request"/>
333<iref primary="true" item="response"/>
335   A client sends an HTTP request to a server in the form of a <x:dfn>request</x:dfn>
336   message, beginning with a request-line that includes a method, URI, and
337   protocol version (<xref target="request.line"/>),
338   followed by header fields containing
339   request modifiers, client information, and representation metadata
340   (<xref target="header.fields"/>),
341   an empty line to indicate the end of the header section, and finally
342   a message body containing the payload body (if any,
343   <xref target="message.body"/>).
346   A server responds to a client's request by sending one or more HTTP
347   <x:dfn>response</x:dfn>
348   messages, each beginning with a status line that
349   includes the protocol version, a success or error code, and textual
350   reason phrase (<xref target="status.line"/>),
351   possibly followed by header fields containing server
352   information, resource metadata, and representation metadata
353   (<xref target="header.fields"/>),
354   an empty line to indicate the end of the header section, and finally
355   a message body containing the payload body (if any,
356   <xref target="message.body"/>).
359   A connection might be used for multiple request/response exchanges,
360   as defined in <xref target="persistent.connections"/>.
363   The following example illustrates a typical message exchange for a
364   GET request on the URI "":
367client request:
368</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
369GET /hello.txt HTTP/1.1
370User-Agent: curl/7.16.3 libcurl/7.16.3 OpenSSL/0.9.7l zlib/1.2.3
372Accept-Language: en, mi
376server response:
377</preamble><artwork type="message/http; msgtype=&#34;response&#34;" x:indent-with="  ">
378HTTP/1.1 200 OK
379Date: Mon, 27 Jul 2009 12:28:53 GMT
380Server: Apache
381Last-Modified: Wed, 22 Jul 2009 19:15:56 GMT
382ETag: "34aa387-d-1568eb00"
383Accept-Ranges: bytes
384Content-Length: <x:length-of target="exbody"/>
385Vary: Accept-Encoding
386Content-Type: text/plain
388<x:span anchor="exbody">Hello World! My payload includes a trailing CRLF.
393<section title="Implementation Diversity" anchor="implementation-diversity">
395   When considering the design of HTTP, it is easy to fall into a trap of
396   thinking that all user agents are general-purpose browsers and all origin
397   servers are large public websites. That is not the case in practice.
398   Common HTTP user agents include household appliances, stereos, scales,
399   firmware update scripts, command-line programs, mobile apps,
400   and communication devices in a multitude of shapes and sizes.  Likewise,
401   common HTTP origin servers include home automation units, configurable
402   networking components, office machines, autonomous robots, news feeds,
403   traffic cameras, ad selectors, and video delivery platforms.
406   The term "user agent" does not imply that there is a human user directly
407   interacting with the software agent at the time of a request. In many
408   cases, a user agent is installed or configured to run in the background
409   and save its results for later inspection (or save only a subset of those
410   results that might be interesting or erroneous). Spiders, for example, are
411   typically given a start URI and configured to follow certain behavior while
412   crawling the Web as a hypertext graph.
415   The implementation diversity of HTTP means that we cannot assume the
416   user agent can make interactive suggestions to a user or provide adequate
417   warning for security or privacy options.  In the few cases where this
418   specification requires reporting of errors to the user, it is acceptable
419   for such reporting to only be observable in an error console or log file.
420   Likewise, requirements that an automated action be confirmed by the user
421   before proceeding can be met via advance configuration choices,
422   run-time options, or simply not proceeding with the unsafe action.
426<section title="Intermediaries" anchor="intermediaries">
427<iref primary="true" item="intermediary"/>
429   HTTP enables the use of intermediaries to satisfy requests through
430   a chain of connections.  There are three common forms of HTTP
431   <x:dfn>intermediary</x:dfn>: proxy, gateway, and tunnel.  In some cases,
432   a single intermediary might act as an origin server, proxy, gateway,
433   or tunnel, switching behavior based on the nature of each request.
435<figure><artwork type="drawing">
436         &gt;             &gt;             &gt;             &gt;
437    <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>
438               &lt;             &lt;             &lt;             &lt;
441   The figure above shows three intermediaries (A, B, and C) between the
442   user agent and origin server. A request or response message that
443   travels the whole chain will pass through four separate connections.
444   Some HTTP communication options
445   might apply only to the connection with the nearest, non-tunnel
446   neighbor, only to the end-points of the chain, or to all connections
447   along the chain. Although the diagram is linear, each participant might
448   be engaged in multiple, simultaneous communications. For example, B
449   might be receiving requests from many clients other than A, and/or
450   forwarding requests to servers other than C, at the same time that it
451   is handling A's request.
454<iref primary="true" item="upstream"/><iref primary="true" item="downstream"/>
455<iref primary="true" item="inbound"/><iref primary="true" item="outbound"/>
456   We use the terms "<x:dfn>upstream</x:dfn>" and "<x:dfn>downstream</x:dfn>"
457   to describe various requirements in relation to the directional flow of a
458   message: all messages flow from upstream to downstream.
459   Likewise, we use the terms inbound and outbound to refer to
460   directions in relation to the request path:
461   "<x:dfn>inbound</x:dfn>" means toward the origin server and
462   "<x:dfn>outbound</x:dfn>" means toward the user agent.
464<t><iref primary="true" item="proxy"/>
465   A "<x:dfn>proxy</x:dfn>" is a message forwarding agent that is selected by the
466   client, usually via local configuration rules, to receive requests
467   for some type(s) of absolute URI and attempt to satisfy those
468   requests via translation through the HTTP interface.  Some translations
469   are minimal, such as for proxy requests for "http" URIs, whereas
470   other requests might require translation to and from entirely different
471   application-level protocols. Proxies are often used to group an
472   organization's HTTP requests through a common intermediary for the
473   sake of security, annotation services, or shared caching.
476<iref primary="true" item="transforming proxy"/>
477<iref primary="true" item="non-transforming proxy"/>
478   An HTTP-to-HTTP proxy is called a "<x:dfn>transforming proxy</x:dfn>" if it is designed
479   or configured to modify request or response messages in a semantically
480   meaningful way (i.e., modifications, beyond those required by normal
481   HTTP processing, that change the message in a way that would be
482   significant to the original sender or potentially significant to
483   downstream recipients).  For example, a transforming proxy might be
484   acting as a shared annotation server (modifying responses to include
485   references to a local annotation database), a malware filter, a
486   format transcoder, or an intranet-to-Internet privacy filter.  Such
487   transformations are presumed to be desired by the client (or client
488   organization) that selected the proxy and are beyond the scope of
489   this specification.  However, when a proxy is not intended to transform
490   a given message, we use the term "<x:dfn>non-transforming proxy</x:dfn>" to target
491   requirements that preserve HTTP message semantics. See &status-203; and
492   &header-warning; for status and warning codes related to transformations.
494<t><iref primary="true" item="gateway"/><iref primary="true" item="reverse proxy"/>
495<iref primary="true" item="accelerator"/>
496   A "<x:dfn>gateway</x:dfn>" (a.k.a., "<x:dfn>reverse proxy</x:dfn>")
497   is a receiving agent that acts
498   as a layer above some other server(s) and translates the received
499   requests to the underlying server's protocol.  Gateways are often
500   used to encapsulate legacy or untrusted information services, to
501   improve server performance through "<x:dfn>accelerator</x:dfn>" caching, and to
502   enable partitioning or load-balancing of HTTP services across
503   multiple machines.
506   A gateway behaves as an origin server on its outbound connection and
507   as a user agent on its inbound connection.
508   All HTTP requirements applicable to an origin server
509   also apply to the outbound communication of a gateway.
510   A gateway communicates with inbound servers using any protocol that
511   it desires, including private extensions to HTTP that are outside
512   the scope of this specification.  However, an HTTP-to-HTTP gateway
513   that wishes to interoperate with third-party HTTP servers &MUST;
514   conform to HTTP user agent requirements on the gateway's inbound
515   connection and &MUST; implement the <x:ref>Connection</x:ref>
516   (<xref target="header.connection"/>) and <x:ref>Via</x:ref>
517   (<xref target="header.via"/>) header fields for both connections.
519<t><iref primary="true" item="tunnel"/>
520   A "<x:dfn>tunnel</x:dfn>" acts as a blind relay between two connections
521   without changing the messages. Once active, a tunnel is not
522   considered a party to the HTTP communication, though the tunnel might
523   have been initiated by an HTTP request. A tunnel ceases to exist when
524   both ends of the relayed connection are closed. Tunnels are used to
525   extend a virtual connection through an intermediary, such as when
526   Transport Layer Security (TLS, <xref target="RFC5246"/>) is used to
527   establish confidential communication through a shared firewall proxy.
529<t><iref primary="true" item="interception proxy"/>
530<iref primary="true" item="transparent proxy"/>
531<iref primary="true" item="captive portal"/>
532   The above categories for intermediary only consider those acting as
533   participants in the HTTP communication.  There are also intermediaries
534   that can act on lower layers of the network protocol stack, filtering or
535   redirecting HTTP traffic without the knowledge or permission of message
536   senders. Network intermediaries often introduce security flaws or
537   interoperability problems by violating HTTP semantics.  For example, an
538   "<x:dfn>interception proxy</x:dfn>" <xref target="RFC3040"/> (also commonly
539   known as a "<x:dfn>transparent proxy</x:dfn>" <xref target="RFC1919"/> or
540   "<x:dfn>captive portal</x:dfn>")
541   differs from an HTTP proxy because it is not selected by the client.
542   Instead, an interception proxy filters or redirects outgoing TCP port 80
543   packets (and occasionally other common port traffic).
544   Interception proxies are commonly found on public network access points,
545   as a means of enforcing account subscription prior to allowing use of
546   non-local Internet services, and within corporate firewalls to enforce
547   network usage policies.
548   They are indistinguishable from a man-in-the-middle attack.
551   HTTP is defined as a stateless protocol, meaning that each request message
552   can be understood in isolation.  Many implementations depend on HTTP's
553   stateless design in order to reuse proxied connections or dynamically
554   load-balance requests across multiple servers.  Hence, servers &MUST-NOT;
555   assume that two requests on the same connection are from the same user
556   agent unless the connection is secured and specific to that agent.
557   Some non-standard HTTP extensions (e.g., <xref target="RFC4559"/>) have
558   been known to violate this requirement, resulting in security and
559   interoperability problems.
563<section title="Caches" anchor="caches">
564<iref primary="true" item="cache"/>
566   A "<x:dfn>cache</x:dfn>" is a local store of previous response messages and the
567   subsystem that controls its message storage, retrieval, and deletion.
568   A cache stores cacheable responses in order to reduce the response
569   time and network bandwidth consumption on future, equivalent
570   requests. Any client or server &MAY; employ a cache, though a cache
571   cannot be used by a server while it is acting as a tunnel.
574   The effect of a cache is that the request/response chain is shortened
575   if one of the participants along the chain has a cached response
576   applicable to that request. The following illustrates the resulting
577   chain if B has a cached copy of an earlier response from O (via C)
578   for a request that has not been cached by UA or A.
580<figure><artwork type="drawing">
581            &gt;             &gt;
582       <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>
583                  &lt;             &lt;
585<t><iref primary="true" item="cacheable"/>
586   A response is "<x:dfn>cacheable</x:dfn>" if a cache is allowed to store a copy of
587   the response message for use in answering subsequent requests.
588   Even when a response is cacheable, there might be additional
589   constraints placed by the client or by the origin server on when
590   that cached response can be used for a particular request. HTTP
591   requirements for cache behavior and cacheable responses are
592   defined in &caching-overview;. 
595   There are a wide variety of architectures and configurations
596   of caches deployed across the World Wide Web and
597   inside large organizations. These include national hierarchies
598   of proxy caches to save transoceanic bandwidth, collaborative systems that
599   broadcast or multicast cache entries, archives of pre-fetched cache
600   entries for use in off-line or high-latency environments, and so on.
604<section title="Conformance and Error Handling" anchor="conformance">
606   This specification targets conformance criteria according to the role of
607   a participant in HTTP communication.  Hence, HTTP requirements are placed
608   on senders, recipients, clients, servers, user agents, intermediaries,
609   origin servers, proxies, gateways, or caches, depending on what behavior
610   is being constrained by the requirement. Additional (social) requirements
611   are placed on implementations, resource owners, and protocol element
612   registrations when they apply beyond the scope of a single communication.
615   The verb "generate" is used instead of "send" where a requirement
616   differentiates between creating a protocol element and merely forwarding a
617   received element downstream.
620   An implementation is considered conformant if it complies with all of the
621   requirements associated with the roles it partakes in HTTP. Note that
622   SHOULD-level requirements are relevant here, unless one of the documented
623   exceptions is applicable.
626   Conformance applies to both the syntax and semantics of HTTP protocol
627   elements. A sender &MUST-NOT; generate protocol elements that convey a
628   meaning that is known by that sender to be false. A sender &MUST-NOT;
629   generate protocol elements that do not match the grammar defined by the
630   ABNF rules for those protocol elements that are applicable to the sender's
631   role. If a received protocol element is processed, the recipient &MUST; be
632   able to parse any value that would match the ABNF rules for that protocol
633   element, excluding only those rules not applicable to the recipient's role.
636   Unless noted otherwise, a recipient &MAY; attempt to recover a usable
637   protocol element from an invalid construct.  HTTP does not define
638   specific error handling mechanisms except when they have a direct impact
639   on security, since different applications of the protocol require
640   different error handling strategies.  For example, a Web browser might
641   wish to transparently recover from a response where the
642   <x:ref>Location</x:ref> header field doesn't parse according to the ABNF,
643   whereas a systems control client might consider any form of error recovery
644   to be dangerous.
648<section title="Protocol Versioning" anchor="http.version">
649  <x:anchor-alias value="HTTP-version"/>
650  <x:anchor-alias value="HTTP-name"/>
652   HTTP uses a "&lt;major&gt;.&lt;minor&gt;" numbering scheme to indicate
653   versions of the protocol. This specification defines version "1.1".
654   The protocol version as a whole indicates the sender's conformance
655   with the set of requirements laid out in that version's corresponding
656   specification of HTTP.
659   The version of an HTTP message is indicated by an HTTP-version field
660   in the first line of the message. HTTP-version is case-sensitive.
662<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-version"/><iref primary="true" item="Grammar" subitem="HTTP-name"/>
663  <x:ref>HTTP-version</x:ref>  = <x:ref>HTTP-name</x:ref> "/" <x:ref>DIGIT</x:ref> "." <x:ref>DIGIT</x:ref>
664  <x:ref>HTTP-name</x:ref>     = <x:abnf-char-sequence>"HTTP"</x:abnf-char-sequence> ; "HTTP", case-sensitive
667   The HTTP version number consists of two decimal digits separated by a "."
668   (period or decimal point).  The first digit ("major version") indicates the
669   HTTP messaging syntax, whereas the second digit ("minor version") indicates
670   the highest minor version to which the sender is
671   conformant and able to understand for future communication.  The minor
672   version advertises the sender's communication capabilities even when the
673   sender is only using a backwards-compatible subset of the protocol,
674   thereby letting the recipient know that more advanced features can
675   be used in response (by servers) or in future requests (by clients).
678   When an HTTP/1.1 message is sent to an HTTP/1.0 recipient
679   <xref target="RFC1945"/> or a recipient whose version is unknown,
680   the HTTP/1.1 message is constructed such that it can be interpreted
681   as a valid HTTP/1.0 message if all of the newer features are ignored.
682   This specification places recipient-version requirements on some
683   new features so that a conformant sender will only use compatible
684   features until it has determined, through configuration or the
685   receipt of a message, that the recipient supports HTTP/1.1.
688   The interpretation of a header field does not change between minor
689   versions of the same major HTTP version, though the default
690   behavior of a recipient in the absence of such a field can change.
691   Unless specified otherwise, header fields defined in HTTP/1.1 are
692   defined for all versions of HTTP/1.x.  In particular, the <x:ref>Host</x:ref>
693   and <x:ref>Connection</x:ref> header fields ought to be implemented by all
694   HTTP/1.x implementations whether or not they advertise conformance with
695   HTTP/1.1.
698   New header fields can be defined such that, when they are
699   understood by a recipient, they might override or enhance the
700   interpretation of previously defined header fields.  When an
701   implementation receives an unrecognized header field, the recipient
702   &MUST; ignore that header field for local processing regardless of
703   the message's HTTP version.  An unrecognized header field received
704   by a proxy &MUST; be forwarded downstream unless the header field's
705   field-name is listed in the message's <x:ref>Connection</x:ref> header field
706   (see <xref target="header.connection"/>).
707   These requirements allow HTTP's functionality to be enhanced without
708   requiring prior update of deployed intermediaries.
711   Intermediaries that process HTTP messages (i.e., all intermediaries
712   other than those acting as tunnels) &MUST; send their own HTTP-version
713   in forwarded messages.  In other words, they &MUST-NOT; blindly
714   forward the first line of an HTTP message without ensuring that the
715   protocol version in that message matches a version to which that
716   intermediary is conformant for both the receiving and
717   sending of messages.  Forwarding an HTTP message without rewriting
718   the HTTP-version might result in communication errors when downstream
719   recipients use the message sender's version to determine what features
720   are safe to use for later communication with that sender.
723   An HTTP client &SHOULD; send a request version equal to the highest
724   version to which the client is conformant and
725   whose major version is no higher than the highest version supported
726   by the server, if this is known.  An HTTP client &MUST-NOT; send a
727   version to which it is not conformant.
730   An HTTP client &MAY; send a lower request version if it is known that
731   the server incorrectly implements the HTTP specification, but only
732   after the client has attempted at least one normal request and determined
733   from the response status or header fields (e.g., <x:ref>Server</x:ref>) that
734   the server improperly handles higher request versions.
737   An HTTP server &SHOULD; send a response version equal to the highest
738   version to which the server is conformant and
739   whose major version is less than or equal to the one received in the
740   request.  An HTTP server &MUST-NOT; send a version to which it is not
741   conformant.  A server &MAY; send a <x:ref>505 (HTTP Version Not
742   Supported)</x:ref> response if it cannot send a response using the
743   major version used in the client's request.
746   An HTTP server &MAY; send an HTTP/1.0 response to an HTTP/1.0 request
747   if it is known or suspected that the client incorrectly implements the
748   HTTP specification and is incapable of correctly processing later
749   version responses, such as when a client fails to parse the version
750   number correctly or when an intermediary is known to blindly forward
751   the HTTP-version even when it doesn't conform to the given minor
752   version of the protocol. Such protocol downgrades &SHOULD-NOT; be
753   performed unless triggered by specific client attributes, such as when
754   one or more of the request header fields (e.g., <x:ref>User-Agent</x:ref>)
755   uniquely match the values sent by a client known to be in error.
758   The intention of HTTP's versioning design is that the major number
759   will only be incremented if an incompatible message syntax is
760   introduced, and that the minor number will only be incremented when
761   changes made to the protocol have the effect of adding to the message
762   semantics or implying additional capabilities of the sender.  However,
763   the minor version was not incremented for the changes introduced between
764   <xref target="RFC2068"/> and <xref target="RFC2616"/>, and this revision
765   has specifically avoided any such changes to the protocol.
769<section title="Uniform Resource Identifiers" anchor="uri">
770<iref primary="true" item="resource"/>
772   Uniform Resource Identifiers (URIs) <xref target="RFC3986"/> are used
773   throughout HTTP as the means for identifying resources (&resource;).
774   URI references are used to target requests, indicate redirects, and define
775   relationships.
777  <x:anchor-alias value="URI-reference"/>
778  <x:anchor-alias value="absolute-URI"/>
779  <x:anchor-alias value="relative-part"/>
780  <x:anchor-alias value="authority"/>
781  <x:anchor-alias value="path-abempty"/>
782  <x:anchor-alias value="port"/>
783  <x:anchor-alias value="query"/>
784  <x:anchor-alias value="segment"/>
785  <x:anchor-alias value="uri-host"/>
786  <x:anchor-alias value="absolute-path"/>
787  <x:anchor-alias value="partial-URI"/>
789   This specification adopts the definitions of "URI-reference",
790   "absolute-URI", "relative-part", "port", "host",
791   "path-abempty", "query", "segment", and "authority" from the
792   URI generic syntax.
793   In addition, we define an "absolute-path" rule (that differs from
794   RFC 3986's "path-absolute" in that it allows a leading "//")
795   and a "partial-URI" rule for protocol elements
796   that allow a relative URI but not a fragment.
798<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="absolute-path"/><iref primary="true" item="Grammar" subitem="port"/><iref primary="true" item="Grammar" subitem="query"/><iref primary="true" item="Grammar" subitem="segment"/><iref primary="true" item="Grammar" subitem="uri-host"/><iref primary="true" item="Grammar" subitem="partial-URI"><!--exported production--></iref>
799  <x:ref>URI-reference</x:ref> = &lt;URI-reference, defined in <xref target="RFC3986" x:fmt="," x:sec="4.1"/>&gt;
800  <x:ref>absolute-URI</x:ref>  = &lt;absolute-URI, defined in <xref target="RFC3986" x:fmt="," x:sec="4.3"/>&gt;
801  <x:ref>relative-part</x:ref> = &lt;relative-part, defined in <xref target="RFC3986" x:fmt="," x:sec="4.2"/>&gt;
802  <x:ref>authority</x:ref>     = &lt;authority, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2"/>&gt;
803  <x:ref>path-abempty</x:ref>  = &lt;path-abempty, defined in <xref target="RFC3986" x:fmt="," x:sec="3.3"/>&gt;
804  <x:ref>port</x:ref>          = &lt;port, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2.3"/>&gt;
805  <x:ref>query</x:ref>         = &lt;query, defined in <xref target="RFC3986" x:fmt="," x:sec="3.4"/>&gt;
806  <x:ref>segment</x:ref>       = &lt;segment, defined in <xref target="RFC3986" x:fmt="," x:sec="3.3"/>&gt;
807  <x:ref>uri-host</x:ref>      = &lt;host, defined in <xref target="RFC3986" x:fmt="," x:sec="3.2.2"/>&gt;
809  <x:ref>absolute-path</x:ref> = 1*( "/" segment )
810  <x:ref>partial-URI</x:ref>   = relative-part [ "?" query ]
813   Each protocol element in HTTP that allows a URI reference will indicate
814   in its ABNF production whether the element allows any form of reference
815   (URI-reference), only a URI in absolute form (absolute-URI), only the
816   path and optional query components, or some combination of the above.
817   Unless otherwise indicated, URI references are parsed
818   relative to the effective request URI
819   (<xref target="effective.request.uri"/>).
822<section title="http URI scheme" anchor="http.uri">
823  <x:anchor-alias value="http-URI"/>
824  <iref item="http URI scheme" primary="true"/>
825  <iref item="URI scheme" subitem="http" primary="true"/>
827   The "http" URI scheme is hereby defined for the purpose of minting
828   identifiers according to their association with the hierarchical
829   namespace governed by a potential HTTP origin server listening for
830   TCP (<xref target="RFC0793"/>) connections on a given port.
832<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="http-URI"><!--terminal production--></iref>
833  <x:ref>http-URI</x:ref> = "http:" "//" <x:ref>authority</x:ref> <x:ref>path-abempty</x:ref> [ "?" <x:ref>query</x:ref> ]
836   The HTTP origin server is identified by the generic syntax's
837   <x:ref>authority</x:ref> component, which includes a host identifier
838   and optional TCP port (<xref target="RFC3986" x:fmt="," x:sec="3.2.2"/>).
839   The remainder of the URI, consisting of both the hierarchical path
840   component and optional query component, serves as an identifier for
841   a potential resource within that origin server's name space.
844   If the host identifier is provided as an IP address,
845   then the origin server is any listener on the indicated TCP port at
846   that IP address. If host is a registered name, then that name is
847   considered an indirect identifier and the recipient might use a name
848   resolution service, such as DNS, to find the address of a listener
849   for that host.
850   The host &MUST-NOT; be empty; if an "http" URI is received with an
851   empty host, then it &MUST; be rejected as invalid.
852   If the port subcomponent is empty or not given, then TCP port 80 is
853   assumed (the default reserved port for WWW services).
856   Regardless of the form of host identifier, access to that host is not
857   implied by the mere presence of its name or address. The host might or might
858   not exist and, even when it does exist, might or might not be running an
859   HTTP server or listening to the indicated port. The "http" URI scheme
860   makes use of the delegated nature of Internet names and addresses to
861   establish a naming authority (whatever entity has the ability to place
862   an HTTP server at that Internet name or address) and allows that
863   authority to determine which names are valid and how they might be used.
866   When an "http" URI is used within a context that calls for access to the
867   indicated resource, a client &MAY; attempt access by resolving
868   the host to an IP address, establishing a TCP connection to that address
869   on the indicated port, and sending an HTTP request message
870   (<xref target="http.message"/>) containing the URI's identifying data
871   (<xref target="message.routing"/>) to the server.
872   If the server responds to that request with a non-interim HTTP response
873   message, as described in &status-codes;, then that response
874   is considered an authoritative answer to the client's request.
877   Although HTTP is independent of the transport protocol, the "http"
878   scheme is specific to TCP-based services because the name delegation
879   process depends on TCP for establishing authority.
880   An HTTP service based on some other underlying connection protocol
881   would presumably be identified using a different URI scheme, just as
882   the "https" scheme (below) is used for resources that require an
883   end-to-end secured connection. Other protocols might also be used to
884   provide access to "http" identified resources &mdash; it is only the
885   authoritative interface used for mapping the namespace that is
886   specific to TCP.
889   The URI generic syntax for authority also includes a deprecated
890   userinfo subcomponent (<xref target="RFC3986" x:fmt="," x:sec="3.2.1"/>)
891   for including user authentication information in the URI.  Some
892   implementations make use of the userinfo component for internal
893   configuration of authentication information, such as within command
894   invocation options, configuration files, or bookmark lists, even
895   though such usage might expose a user identifier or password.
896   Senders &MUST; exclude the userinfo subcomponent (and its "@"
897   delimiter) when an "http" URI is transmitted within a message as a
898   request target or header field value.
899   Recipients of an "http" URI reference &SHOULD; parse for userinfo and
900   treat its presence as an error, since it is likely being used to obscure
901   the authority for the sake of phishing attacks.
905<section title="https URI scheme" anchor="https.uri">
906   <x:anchor-alias value="https-URI"/>
907   <iref item="https URI scheme"/>
908   <iref item="URI scheme" subitem="https"/>
910   The "https" URI scheme is hereby defined for the purpose of minting
911   identifiers according to their association with the hierarchical
912   namespace governed by a potential HTTP origin server listening to a
913   given TCP port for TLS-secured connections
914   (<xref target="RFC0793"/>, <xref target="RFC5246"/>).
917   All of the requirements listed above for the "http" scheme are also
918   requirements for the "https" scheme, except that a default TCP port
919   of 443 is assumed if the port subcomponent is empty or not given,
920   and the TCP connection &MUST; be secured, end-to-end, through the
921   use of strong encryption prior to sending the first HTTP request.
923<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="https-URI"><!--terminal production--></iref>
924  <x:ref>https-URI</x:ref> = "https:" "//" <x:ref>authority</x:ref> <x:ref>path-abempty</x:ref> [ "?" <x:ref>query</x:ref> ]
927   Resources made available via the "https" scheme have no shared
928   identity with the "http" scheme even if their resource identifiers
929   indicate the same authority (the same host listening to the same
930   TCP port).  They are distinct name spaces and are considered to be
931   distinct origin servers.  However, an extension to HTTP that is
932   defined to apply to entire host domains, such as the Cookie protocol
933   <xref target="RFC6265"/>, can allow information
934   set by one service to impact communication with other services
935   within a matching group of host domains.
938   The process for authoritative access to an "https" identified
939   resource is defined in <xref target="RFC2818"/>.
943<section title="http and https URI Normalization and Comparison" anchor="uri.comparison">
945   Since the "http" and "https" schemes conform to the URI generic syntax,
946   such URIs are normalized and compared according to the algorithm defined
947   in <xref target="RFC3986" x:fmt="," x:sec="6"/>, using the defaults
948   described above for each scheme.
951   If the port is equal to the default port for a scheme, the normal form is
952   to elide the port subcomponent. When not being used in absolute form as the
953   request target of an OPTIONS request, an empty path component is equivalent
954   to an absolute path of "/", so the normal form is to provide a path of "/"
955   instead. The scheme and host are case-insensitive and normally provided in
956   lowercase; all other components are compared in a case-sensitive manner.
957   Characters other than those in the "reserved" set are equivalent to their
958   percent-encoded octets (see <xref target="RFC3986" x:fmt=","
959   x:sec="2.1"/>): the normal form is to not encode them.
962   For example, the following three URIs are equivalent:
964<figure><artwork type="example">
973<section title="Message Format" anchor="http.message">
974<x:anchor-alias value="generic-message"/>
975<x:anchor-alias value="message.types"/>
976<x:anchor-alias value="HTTP-message"/>
977<x:anchor-alias value="start-line"/>
978<iref item="header section"/>
979<iref item="headers"/>
980<iref item="header field"/>
982   All HTTP/1.1 messages consist of a start-line followed by a sequence of
983   octets in a format similar to the Internet Message Format
984   <xref target="RFC5322"/>: zero or more header fields (collectively
985   referred to as the "headers" or the "header section"), an empty line
986   indicating the end of the header section, and an optional message body.
988<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="HTTP-message"><!--terminal production--></iref>
989  <x:ref>HTTP-message</x:ref>   = <x:ref>start-line</x:ref>
990                   *( <x:ref>header-field</x:ref> <x:ref>CRLF</x:ref> )
991                   <x:ref>CRLF</x:ref>
992                   [ <x:ref>message-body</x:ref> ]
995   The normal procedure for parsing an HTTP message is to read the
996   start-line into a structure, read each header field into a hash
997   table by field name until the empty line, and then use the parsed
998   data to determine if a message body is expected.  If a message body
999   has been indicated, then it is read as a stream until an amount
1000   of octets equal to the message body length is read or the connection
1001   is closed.
1004   Recipients &MUST; parse an HTTP message as a sequence of octets in an
1005   encoding that is a superset of US-ASCII <xref target="USASCII"/>.
1006   Parsing an HTTP message as a stream of Unicode characters, without regard
1007   for the specific encoding, creates security vulnerabilities due to the
1008   varying ways that string processing libraries handle invalid multibyte
1009   character sequences that contain the octet LF (%x0A).  String-based
1010   parsers can only be safely used within protocol elements after the element
1011   has been extracted from the message, such as within a header field-value
1012   after message parsing has delineated the individual fields.
1015   An HTTP message can be parsed as a stream for incremental processing or
1016   forwarding downstream.  However, recipients cannot rely on incremental
1017   delivery of partial messages, since some implementations will buffer or
1018   delay message forwarding for the sake of network efficiency, security
1019   checks, or payload transformations.
1022<section title="Start Line" anchor="start.line">
1023  <x:anchor-alias value="Start-Line"/>
1025   An HTTP message can either be a request from client to server or a
1026   response from server to client.  Syntactically, the two types of message
1027   differ only in the start-line, which is either a request-line (for requests)
1028   or a status-line (for responses), and in the algorithm for determining
1029   the length of the message body (<xref target="message.body"/>).
1032   In theory, a client could receive requests and a server could receive
1033   responses, distinguishing them by their different start-line formats,
1034   but in practice servers are implemented to only expect a request
1035   (a response is interpreted as an unknown or invalid request method)
1036   and clients are implemented to only expect a response.
1038<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="start-line"/>
1039  <x:ref>start-line</x:ref>     = <x:ref>request-line</x:ref> / <x:ref>status-line</x:ref>
1042   A sender &MUST-NOT; send whitespace between the start-line and
1043   the first header field. The presence of such whitespace in a request
1044   might be an attempt to trick a server into ignoring that field or
1045   processing the line after it as a new request, either of which might
1046   result in a security vulnerability if other implementations within
1047   the request chain interpret the same message differently.
1048   Likewise, the presence of such whitespace in a response might be
1049   ignored by some clients or cause others to cease parsing.
1052   A recipient that receives whitespace between the start-line and
1053   the first header field &MUST; either reject the message as invalid or
1054   consume each whitespace-preceded line without further processing of it
1055   (i.e., ignore the entire line, along with any subsequent lines preceded
1056   by whitespace, until a properly formed header field is received or the
1057   header block is terminated).
1060<section title="Request Line" anchor="request.line">
1061  <x:anchor-alias value="Request"/>
1062  <x:anchor-alias value="request-line"/>
1064   A request-line begins with a method token, followed by a single
1065   space (SP), the request-target, another single space (SP), the
1066   protocol version, and ending with CRLF.
1068<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="request-line"/>
1069  <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>
1071<iref primary="true" item="method"/>
1072<t anchor="method">
1073   The method token indicates the request method to be performed on the
1074   target resource. The request method is case-sensitive.
1076<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="method"/>
1077  <x:ref>method</x:ref>         = <x:ref>token</x:ref>
1080   The methods defined by this specification can be found in
1081   &methods;, along with information regarding the HTTP method registry
1082   and considerations for defining new methods.
1084<iref item="request-target"/>
1086   The request-target identifies the target resource upon which to apply
1087   the request, as defined in <xref target="request-target"/>.
1090   No whitespace is allowed inside the method, request-target, and
1091   protocol version.  Hence, recipients typically parse the request-line
1092   into its component parts by splitting on whitespace
1093   (see <xref target="message.robustness"/>).
1096   Unfortunately, some user agents fail to properly encode hypertext
1097   references that have embedded whitespace, sending the characters directly
1098   instead of properly encoding or excluding the disallowed characters.
1099   Recipients of an invalid request-line &SHOULD; respond with either a
1100   <x:ref>400 (Bad Request)</x:ref> error or a <x:ref>301 (Moved Permanently)</x:ref>
1101   redirect with the request-target properly encoded.  Recipients &SHOULD-NOT;
1102   attempt to autocorrect and then process the request without a redirect,
1103   since the invalid request-line might be deliberately crafted to bypass
1104   security filters along the request chain.
1107   HTTP does not place a pre-defined limit on the length of a request-line.
1108   A server that receives a method longer than any that it implements
1109   &SHOULD; respond with a <x:ref>501 (Not Implemented)</x:ref> status code.
1110   A server &MUST; be prepared to receive URIs of unbounded length and
1111   respond with the <x:ref>414 (URI Too Long)</x:ref> status code if the received
1112   request-target would be longer than the server wishes to handle
1113   (see &status-414;).
1116   Various ad-hoc limitations on request-line length are found in practice.
1117   It is &RECOMMENDED; that all HTTP senders and recipients support, at a
1118   minimum, request-line lengths of 8000 octets.
1122<section title="Status Line" anchor="status.line">
1123  <x:anchor-alias value="response"/>
1124  <x:anchor-alias value="status-line"/>
1125  <x:anchor-alias value="status-code"/>
1126  <x:anchor-alias value="reason-phrase"/>
1128   The first line of a response message is the status-line, consisting
1129   of the protocol version, a space (SP), the status code, another space,
1130   a possibly-empty textual phrase describing the status code, and
1131   ending with CRLF.
1133<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="status-line"/>
1134  <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>
1137   The status-code element is a 3-digit integer code describing the
1138   result of the server's attempt to understand and satisfy the client's
1139   corresponding request. The rest of the response message is to be
1140   interpreted in light of the semantics defined for that status code.
1141   See &status-codes; for information about the semantics of status codes,
1142   including the classes of status code (indicated by the first digit),
1143   the status codes defined by this specification, considerations for the
1144   definition of new status codes, and the IANA registry.
1146<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="status-code"/>
1147  <x:ref>status-code</x:ref>    = 3<x:ref>DIGIT</x:ref>
1150   The reason-phrase element exists for the sole purpose of providing a
1151   textual description associated with the numeric status code, mostly
1152   out of deference to earlier Internet application protocols that were more
1153   frequently used with interactive text clients. A client &SHOULD; ignore
1154   the reason-phrase content.
1156<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="reason-phrase"/>
1157  <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> )
1162<section title="Header Fields" anchor="header.fields">
1163  <x:anchor-alias value="header-field"/>
1164  <x:anchor-alias value="field-content"/>
1165  <x:anchor-alias value="field-name"/>
1166  <x:anchor-alias value="field-value"/>
1167  <x:anchor-alias value="obs-fold"/>
1169   Each HTTP header field consists of a case-insensitive field name
1170   followed by a colon (":"), optional whitespace, and the field value.
1172<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"/>
1173  <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>
1174  <x:ref>field-name</x:ref>     = <x:ref>token</x:ref>
1175  <x:ref>field-value</x:ref>    = *( <x:ref>field-content</x:ref> / <x:ref>obs-fold</x:ref> )
1176  <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> )
1177  <x:ref>obs-fold</x:ref>       = <x:ref>CRLF</x:ref> ( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1178                 ; obsolete line folding
1179                 ; see <xref target="field.parsing"/>
1182   The field-name token labels the corresponding field-value as having the
1183   semantics defined by that header field.  For example, the <x:ref>Date</x:ref>
1184   header field is defined in &header-date; as containing the origination
1185   timestamp for the message in which it appears.
1188<section title="Field Extensibility" anchor="field.extensibility">
1190   HTTP header fields are fully extensible: there is no limit on the
1191   introduction of new field names, each presumably defining new semantics,
1192   nor on the number of header fields used in a given message.  Existing
1193   fields are defined in each part of this specification and in many other
1194   specifications outside the core standard.
1195   New header fields can be introduced without changing the protocol version
1196   if their defined semantics allow them to be safely ignored by recipients
1197   that do not recognize them.
1200   New HTTP header fields ought to be registered with IANA in the
1201   Message Header Field Registry, as described in &iana-header-registry;.
1202   A proxy &MUST; forward unrecognized header fields unless the
1203   field-name is listed in the <x:ref>Connection</x:ref> header field
1204   (<xref target="header.connection"/>) or the proxy is specifically
1205   configured to block, or otherwise transform, such fields.
1206   Other recipients &SHOULD; ignore unrecognized header fields.
1210<section title="Field Order" anchor="field.order">
1212   The order in which header fields with differing field names are
1213   received is not significant. However, it is "good practice" to send
1214   header fields that contain control data first, such as <x:ref>Host</x:ref>
1215   on requests and <x:ref>Date</x:ref> on responses, so that implementations
1216   can decide when not to handle a message as early as possible.  A server
1217   &MUST; wait until the entire header section is received before interpreting
1218   a request message, since later header fields might include conditionals,
1219   authentication credentials, or deliberately misleading duplicate
1220   header fields that would impact request processing.
1223   A sender &MUST-NOT; generate multiple header fields with the same field
1224   name in a message unless either the entire field value for that
1225   header field is defined as a comma-separated list [i.e., #(values)]
1226   or the header field is a well-known exception (as noted below).
1229   Multiple header fields with the same field name can be combined into
1230   one "field-name: field-value" pair, without changing the semantics of the
1231   message, by appending each subsequent field value to the combined
1232   field value in order, separated by a comma. The order in which
1233   header fields with the same field name are received is therefore
1234   significant to the interpretation of the combined field value;
1235   a proxy &MUST-NOT; change the order of these field values when
1236   forwarding a message.
1239  <t>
1240   &Note; In practice, the "Set-Cookie" header field (<xref target="RFC6265"/>)
1241   often appears multiple times in a response message and does not use the
1242   list syntax, violating the above requirements on multiple header fields
1243   with the same name. Since it cannot be combined into a single field-value,
1244   recipients ought to handle "Set-Cookie" as a special case while processing
1245   header fields. (See Appendix A.2.3 of <xref target="Kri2001"/> for details.)
1246  </t>
1250<section title="Whitespace" anchor="whitespace">
1251<t anchor="rule.LWS">
1252   This specification uses three rules to denote the use of linear
1253   whitespace: OWS (optional whitespace), RWS (required whitespace), and
1254   BWS ("bad" whitespace).
1256<t anchor="rule.OWS">
1257   The OWS rule is used where zero or more linear whitespace octets might
1258   appear. OWS &SHOULD; either not be generated or be generated as a single
1259   SP. Multiple OWS octets that occur within field-content &SHOULD; either
1260   be replaced with a single SP or transformed to all SP octets (each
1261   octet other than SP replaced with SP) before interpreting the field value
1262   or forwarding the message downstream.
1264<t anchor="rule.RWS">
1265   RWS is used when at least one linear whitespace octet is required to
1266   separate field tokens. RWS &SHOULD; be generated as a single SP.
1267   Multiple RWS octets that occur within field-content &SHOULD; either
1268   be replaced with a single SP or transformed to all SP octets before
1269   interpreting the field value or forwarding the message downstream.
1271<t anchor="rule.BWS">
1272   BWS is used where the grammar allows optional whitespace, for historical
1273   reasons, but it &MUST-NOT; be generated in messages; recipients &MUST;
1274   accept such bad optional whitespace and remove it before interpreting the
1275   field value.
1277<t anchor="rule.whitespace">
1278  <x:anchor-alias value="BWS"/>
1279  <x:anchor-alias value="OWS"/>
1280  <x:anchor-alias value="RWS"/>
1282<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"/>
1283  <x:ref>OWS</x:ref>            = *( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1284                 ; optional whitespace
1285  <x:ref>RWS</x:ref>            = 1*( <x:ref>SP</x:ref> / <x:ref>HTAB</x:ref> )
1286                 ; required whitespace
1287  <x:ref>BWS</x:ref>            = <x:ref>OWS</x:ref>
1288                 ; "bad" whitespace
1292<section title="Field Parsing" anchor="field.parsing">
1294   No whitespace is allowed between the header field-name and colon.
1295   In the past, differences in the handling of such whitespace have led to
1296   security vulnerabilities in request routing and response handling.
1297   A server &MUST; reject any received request message that contains
1298   whitespace between a header field-name and colon with a response code of
1299   <x:ref>400 (Bad Request)</x:ref>. A proxy &MUST; remove any such whitespace
1300   from a response message before forwarding the message downstream.
1303   A field value is preceded by optional whitespace (OWS); a single SP is
1304   preferred. The field value does not include any leading or trailing white
1305   space: OWS occurring before the first non-whitespace octet of the
1306   field value or after the last non-whitespace octet of the field value
1307   is ignored and &SHOULD; be removed before further processing (as this does
1308   not change the meaning of the header field).
1311   Historically, HTTP header field values could be extended over multiple
1312   lines by preceding each extra line with at least one space or horizontal
1313   tab (obs-fold). This specification deprecates such line folding except
1314   within the message/http media type
1315   (<xref target=""/>).
1316   Senders &MUST-NOT; generate messages that include line folding
1317   (i.e., that contain any field-value that contains a match to the
1318   <x:ref>obs-fold</x:ref> rule) unless the message is intended for packaging
1319   within the message/http media type. When an <x:ref>obs-fold</x:ref> is
1320   received in a message, recipients &MUST; do one of:
1321   <list style="symbols">
1322      <t>accept the message and replace any embedded <x:ref>obs-fold</x:ref>
1323         whitespace with either a single <x:ref>SP</x:ref> or a matching
1324         number of <x:ref>SP</x:ref> octets (to avoid buffer copying) prior to
1325         interpreting the field value or forwarding the message
1326         downstream;</t>
1328      <t>if it is a request, reject the message by sending a
1329         <x:ref>400 (Bad Request)</x:ref> response with a representation
1330         explaining that obsolete line folding is unacceptable; or,</t>
1332      <t>if it is a response, discard the message and generate a
1333         <x:ref>502 (Bad Gateway)</x:ref> response with a representation
1334         explaining that unacceptable line folding was received.</t>
1335   </list>
1336   Recipients that choose not to implement <x:ref>obs-fold</x:ref> processing
1337   (as described above) &MUST-NOT; accept messages containing header fields
1338   with leading whitespace, as this can expose them to attacks that exploit
1339   this difference in processing.
1342   Historically, HTTP has allowed field content with text in the ISO-8859-1
1343   <xref target="ISO-8859-1"/> charset, supporting other charsets only
1344   through use of <xref target="RFC2047"/> encoding.
1345   In practice, most HTTP header field values use only a subset of the
1346   US-ASCII charset <xref target="USASCII"/>. Newly defined
1347   header fields &SHOULD; limit their field values to US-ASCII octets.
1348   Recipients &SHOULD; treat other octets in field content (obs-text) as
1349   opaque data.
1353<section title="Field Limits" anchor="field.limits">
1355   HTTP does not place a pre-defined limit on the length of each header field
1356   or on the length of the header block as a whole.  Various ad-hoc
1357   limitations on individual header field length are found in practice,
1358   often depending on the specific field semantics.
1361   A server &MUST; be prepared to receive request header fields of unbounded
1362   length and respond with an appropriate <x:ref>4xx (Client Error)</x:ref>
1363   status code if the received header field(s) are larger than the server
1364   wishes to process.
1367   A client &MUST; be prepared to receive response header fields of unbounded
1368   length. A client &MAY; discard or truncate received header fields that are
1369   larger than the client wishes to process if the field semantics are such
1370   that the dropped value(s) can be safely ignored without changing the
1371   response semantics.
1375<section title="Field value components" anchor="field.components">
1376<t anchor="rule.token.separators">
1377  <x:anchor-alias value="tchar"/>
1378  <x:anchor-alias value="token"/>
1379  <x:anchor-alias value="special"/>
1380  <x:anchor-alias value="word"/>
1381   Many HTTP header field values consist of words (token or quoted-string)
1382   separated by whitespace or special characters. These special characters
1383   &MUST; be in a quoted string to be used within a parameter value (as defined
1384   in <xref target="transfer.codings"/>).
1386<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>
1387  <x:ref>word</x:ref>           = <x:ref>token</x:ref> / <x:ref>quoted-string</x:ref>
1389  <x:ref>token</x:ref>          = 1*<x:ref>tchar</x:ref>
1391  IMPORTANT: when editing "tchar" make sure that "special" is updated accordingly!!!
1392 -->
1393  <x:ref>tchar</x:ref>          = "!" / "#" / "$" / "%" / "&amp;" / "'" / "*"
1394                 / "+" / "-" / "." / "^" / "_" / "`" / "|" / "~"
1395                 / <x:ref>DIGIT</x:ref> / <x:ref>ALPHA</x:ref>
1396                 ; any <x:ref>VCHAR</x:ref>, except <x:ref>special</x:ref>
1398  <x:ref>special</x:ref>        = "(" / ")" / "&lt;" / ">" / "@" / ","
1399                 / ";" / ":" / "\" / DQUOTE / "/" / "["
1400                 / "]" / "?" / "=" / "{" / "}"
1402<t anchor="rule.quoted-string">
1403  <x:anchor-alias value="quoted-string"/>
1404  <x:anchor-alias value="qdtext"/>
1405  <x:anchor-alias value="obs-text"/>
1406   A string of text is parsed as a single word if it is quoted using
1407   double-quote marks.
1409<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"/>
1410  <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>
1411  <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>
1412  <x:ref>obs-text</x:ref>       = %x80-FF
1414<t anchor="rule.quoted-pair">
1415  <x:anchor-alias value="quoted-pair"/>
1416   The backslash octet ("\") can be used as a single-octet
1417   quoting mechanism within quoted-string constructs:
1419<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-pair"/>
1420  <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> )
1423   Recipients that process the value of a quoted-string &MUST; handle a
1424   quoted-pair as if it were replaced by the octet following the backslash.
1427   Senders &SHOULD-NOT; generate a quoted-pair in a quoted-string except where
1428   necessary to quote DQUOTE and backslash octets occurring within that string.
1430<t anchor="rule.comment">
1431  <x:anchor-alias value="comment"/>
1432  <x:anchor-alias value="ctext"/>
1433   Comments can be included in some HTTP header fields by surrounding
1434   the comment text with parentheses. Comments are only allowed in
1435   fields containing "comment" as part of their field value definition.
1437<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="comment"/><iref primary="true" item="Grammar" subitem="ctext"/>
1438  <x:ref>comment</x:ref>        = "(" *( <x:ref>ctext</x:ref> / <x:ref>quoted-cpair</x:ref> / <x:ref>comment</x:ref> ) ")"
1439  <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>
1441<t anchor="rule.quoted-cpair">
1442  <x:anchor-alias value="quoted-cpair"/>
1443   The backslash octet ("\") can be used as a single-octet
1444   quoting mechanism within comment constructs:
1446<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="quoted-cpair"/>
1447  <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> )
1450   Senders &SHOULD-NOT; escape octets in comments that do not require escaping
1451   (i.e., other than the backslash octet "\" and the parentheses "(" and ")").
1457<section title="Message Body" anchor="message.body">
1458  <x:anchor-alias value="message-body"/>
1460   The message body (if any) of an HTTP message is used to carry the
1461   payload body of that request or response.  The message body is
1462   identical to the payload body unless a transfer coding has been
1463   applied, as described in <xref target="header.transfer-encoding"/>.
1465<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="message-body"/>
1466  <x:ref>message-body</x:ref> = *OCTET
1469   The rules for when a message body is allowed in a message differ for
1470   requests and responses.
1473   The presence of a message body in a request is signaled by a
1474   <x:ref>Content-Length</x:ref> or <x:ref>Transfer-Encoding</x:ref> header
1475   field. Request message framing is independent of method semantics,
1476   even if the method does not define any use for a message body.
1479   The presence of a message body in a response depends on both
1480   the request method to which it is responding and the response
1481   status code (<xref target="status.line"/>).
1482   Responses to the HEAD request method never include a message body
1483   because the associated response header fields (e.g.,
1484   <x:ref>Transfer-Encoding</x:ref>, <x:ref>Content-Length</x:ref>, etc.),
1485   if present, indicate only what their values would have been if the request
1486   method had been GET (&HEAD;).
1487   <x:ref>2xx (Successful)</x:ref> responses to CONNECT switch to tunnel
1488   mode instead of having a message body (&CONNECT;).
1489   All <x:ref>1xx (Informational)</x:ref>, <x:ref>204 (No Content)</x:ref>, and
1490   <x:ref>304 (Not Modified)</x:ref> responses do not include a message body.
1491   All other responses do include a message body, although the body
1492   might be of zero length.
1495<section title="Transfer-Encoding" anchor="header.transfer-encoding">
1496  <iref primary="true" item="Transfer-Encoding header field" x:for-anchor=""/>
1497  <iref item="chunked (Coding Format)"/>
1498  <x:anchor-alias value="Transfer-Encoding"/>
1500   The Transfer-Encoding header field lists the transfer coding names
1501   corresponding to the sequence of transfer codings that have been
1502   (or will be) applied to the payload body in order to form the message body.
1503   Transfer codings are defined in <xref target="transfer.codings"/>.
1505<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Transfer-Encoding"/>
1506  <x:ref>Transfer-Encoding</x:ref> = 1#<x:ref>transfer-coding</x:ref>
1509   Transfer-Encoding is analogous to the Content-Transfer-Encoding field of
1510   MIME, which was designed to enable safe transport of binary data over a
1511   7-bit transport service (<xref target="RFC2045" x:fmt="," x:sec="6"/>).
1512   However, safe transport has a different focus for an 8bit-clean transfer
1513   protocol. In HTTP's case, Transfer-Encoding is primarily intended to
1514   accurately delimit a dynamically generated payload and to distinguish
1515   payload encodings that are only applied for transport efficiency or
1516   security from those that are characteristics of the selected resource.
1519   All HTTP/1.1 recipients &MUST; implement the chunked transfer coding
1520   (<xref target="chunked.encoding"/>) because it plays a crucial role in
1521   framing messages when the payload body size is not known in advance.
1522   If chunked is applied to a payload body, the sender &MUST-NOT; apply
1523   chunked more than once (i.e., chunking an already chunked message is not
1524   allowed).
1525   If any transfer coding is applied to a request payload body, the
1526   sender &MUST; apply chunked as the final transfer coding to ensure that
1527   the message is properly framed.
1528   If any transfer coding is applied to a response payload body, the
1529   sender &MUST; either apply chunked as the final transfer coding or
1530   terminate the message by closing the connection.
1533   For example,
1534</preamble><artwork type="example">
1535  Transfer-Encoding: gzip, chunked
1537   indicates that the payload body has been compressed using the gzip
1538   coding and then chunked using the chunked coding while forming the
1539   message body.
1542   Unlike <x:ref>Content-Encoding</x:ref> (&content-codings;),
1543   Transfer-Encoding is a property of the message, not of the representation, and
1544   any recipient along the request/response chain &MAY; decode the received
1545   transfer coding(s) or apply additional transfer coding(s) to the message
1546   body, assuming that corresponding changes are made to the Transfer-Encoding
1547   field-value. Additional information about the encoding parameters &MAY; be
1548   provided by other header fields not defined by this specification.
1551   Transfer-Encoding &MAY; be sent in a response to a HEAD request or in a
1552   <x:ref>304 (Not Modified)</x:ref> response (&status-304;) to a GET request,
1553   neither of which includes a message body,
1554   to indicate that the origin server would have applied a transfer coding
1555   to the message body if the request had been an unconditional GET.
1556   This indication is not required, however, because any recipient on
1557   the response chain (including the origin server) can remove transfer
1558   codings when they are not needed.
1561   Transfer-Encoding was added in HTTP/1.1.  It is generally assumed that
1562   implementations advertising only HTTP/1.0 support will not understand
1563   how to process a transfer-encoded payload.
1564   A client &MUST-NOT; send a request containing Transfer-Encoding unless it
1565   knows the server will handle HTTP/1.1 (or later) requests; such knowledge
1566   might be in the form of specific user configuration or by remembering the
1567   version of a prior received response.
1568   A server &MUST-NOT; send a response containing Transfer-Encoding unless
1569   the corresponding request indicates HTTP/1.1 (or later).
1572   A server that receives a request message with a transfer coding it does
1573   not understand &SHOULD; respond with <x:ref>501 (Not Implemented)</x:ref>.
1577<section title="Content-Length" anchor="header.content-length">
1578  <iref primary="true" item="Content-Length header field" x:for-anchor=""/>
1579  <x:anchor-alias value="Content-Length"/>
1581   When a message does not have a <x:ref>Transfer-Encoding</x:ref> header
1582   field, a Content-Length header field can provide the anticipated size,
1583   as a decimal number of octets, for a potential payload body.
1584   For messages that do include a payload body, the Content-Length field-value
1585   provides the framing information necessary for determining where the body
1586   (and message) ends.  For messages that do not include a payload body, the
1587   Content-Length indicates the size of the selected representation
1588   (&representation;).
1590<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Content-Length"/>
1591  <x:ref>Content-Length</x:ref> = 1*<x:ref>DIGIT</x:ref>
1594   An example is
1596<figure><artwork type="example">
1597  Content-Length: 3495
1600   A sender &MUST-NOT; send a Content-Length header field in any message that
1601   contains a <x:ref>Transfer-Encoding</x:ref> header field.
1604   A user agent &SHOULD; send a Content-Length in a request message when no
1605   <x:ref>Transfer-Encoding</x:ref> is sent and the request method defines
1606   a meaning for an enclosed payload body. For example, a Content-Length
1607   header field is normally sent in a POST request even when the value is
1608   0 (indicating an empty payload body).  A user agent &SHOULD-NOT; send a
1609   Content-Length header field when the request message does not contain a
1610   payload body and the method semantics do not anticipate such a body.
1613   A server &MAY; send a Content-Length header field in a response to a HEAD
1614   request (&HEAD;); a server &MUST-NOT; send Content-Length in such a
1615   response unless its field-value equals the decimal number of octets that
1616   would have been sent in the payload body of a response if the same
1617   request had used the GET method.
1620   A server &MAY; send a Content-Length header field in a
1621   <x:ref>304 (Not Modified)</x:ref> response to a conditional GET request
1622   (&status-304;); a server &MUST-NOT; send Content-Length in such a
1623   response unless its field-value equals the decimal number of octets that
1624   would have been sent in the payload body of a <x:ref>200 (OK)</x:ref>
1625   response to the same request.
1628   A server &MUST-NOT; send a Content-Length header field in any response
1629   with a status code of
1630   <x:ref>1xx (Informational)</x:ref> or <x:ref>204 (No Content)</x:ref>.
1631   A server &SHOULD-NOT; send a Content-Length header field in any
1632   <x:ref>2xx (Successful)</x:ref> response to a CONNECT request (&CONNECT;).
1635   Aside from the cases defined above, in the absence of Transfer-Encoding,
1636   an origin server &SHOULD; send a Content-Length header field when the
1637   payload body size is known prior to sending the complete header block.
1638   This will allow downstream recipients to measure transfer progress,
1639   know when a received message is complete, and potentially reuse the
1640   connection for additional requests.
1643   Any Content-Length field value greater than or equal to zero is valid.
1644   Since there is no predefined limit to the length of a payload,
1645   recipients &SHOULD; anticipate potentially large decimal numerals and
1646   prevent parsing errors due to integer conversion overflows
1647   (<xref target="attack.protocol.element.size.overflows"/>).
1650   If a message is received that has multiple Content-Length header fields
1651   with field-values consisting of the same decimal value, or a single
1652   Content-Length header field with a field value containing a list of
1653   identical decimal values (e.g., "Content-Length: 42, 42"), indicating that
1654   duplicate Content-Length header fields have been generated or combined by an
1655   upstream message processor, then the recipient &MUST; either reject the
1656   message as invalid or replace the duplicated field-values with a single
1657   valid Content-Length field containing that decimal value prior to
1658   determining the message body length.
1661  <t>
1662   &Note; HTTP's use of Content-Length for message framing differs
1663   significantly from the same field's use in MIME, where it is an optional
1664   field used only within the "message/external-body" media-type.
1665  </t>
1669<section title="Message Body Length" anchor="message.body.length">
1670  <iref item="chunked (Coding Format)"/>
1672   The length of a message body is determined by one of the following
1673   (in order of precedence):
1676  <list style="numbers">
1677    <x:lt><t>
1678     Any response to a HEAD request and any response with a
1679     <x:ref>1xx (Informational)</x:ref>, <x:ref>204 (No Content)</x:ref>, or
1680     <x:ref>304 (Not Modified)</x:ref> status code is always
1681     terminated by the first empty line after the header fields, regardless of
1682     the header fields present in the message, and thus cannot contain a
1683     message body.
1684    </t></x:lt>
1685    <x:lt><t>
1686     Any <x:ref>2xx (Successful)</x:ref> response to a CONNECT request implies that the
1687     connection will become a tunnel immediately after the empty line that
1688     concludes the header fields.  A client &MUST; ignore any
1689     <x:ref>Content-Length</x:ref> or <x:ref>Transfer-Encoding</x:ref> header
1690     fields received in such a message.
1691    </t></x:lt>
1692    <x:lt><t>
1693     If a <x:ref>Transfer-Encoding</x:ref> header field is present
1694     and the chunked transfer coding (<xref target="chunked.encoding"/>)
1695     is the final encoding, the message body length is determined by reading
1696     and decoding the chunked data until the transfer coding indicates the
1697     data is complete.
1698    </t>
1699    <t>
1700     If a <x:ref>Transfer-Encoding</x:ref> header field is present in a
1701     response and the chunked transfer coding is not the final encoding, the
1702     message body length is determined by reading the connection until it is
1703     closed by the server.
1704     If a <x:ref>Transfer-Encoding</x:ref> header field is present in a request and the
1705     chunked transfer coding is not the final encoding, the message body
1706     length cannot be determined reliably; the server &MUST; respond with
1707     the <x:ref>400 (Bad Request)</x:ref> status code and then close the connection.
1708    </t>
1709    <t>
1710     If a message is received with both a <x:ref>Transfer-Encoding</x:ref>
1711     and a <x:ref>Content-Length</x:ref> header field, the Transfer-Encoding
1712     overrides the Content-Length. Such a message might indicate an attempt
1713     to perform request or response smuggling (bypass of security-related
1714     checks on message routing or content) and thus ought to be handled as
1715     an error.  A sender &MUST; remove the received Content-Length field
1716     prior to forwarding such a message downstream.
1717    </t></x:lt>
1718    <x:lt><t>
1719     If a message is received without <x:ref>Transfer-Encoding</x:ref> and with
1720     either multiple <x:ref>Content-Length</x:ref> header fields having
1721     differing field-values or a single Content-Length header field having an
1722     invalid value, then the message framing is invalid and &MUST; be treated
1723     as an error to prevent request or response smuggling.
1724     If this is a request message, the server &MUST; respond with
1725     a <x:ref>400 (Bad Request)</x:ref> status code and then close the connection.
1726     If this is a response message received by a proxy, the proxy
1727     &MUST; discard the received response, send a <x:ref>502 (Bad Gateway)</x:ref>
1728     status code as its downstream response, and then close the connection.
1729     If this is a response message received by a user agent, it &MUST; be
1730     treated as an error by discarding the message and closing the connection.
1731    </t></x:lt>
1732    <x:lt><t>
1733     If a valid <x:ref>Content-Length</x:ref> header field is present without
1734     <x:ref>Transfer-Encoding</x:ref>, its decimal value defines the
1735     expected message body length in octets.
1736     If the sender closes the connection or the recipient times out before the
1737     indicated number of octets are received, the recipient &MUST; consider
1738     the message to be incomplete and close the connection.
1739    </t></x:lt>
1740    <x:lt><t>
1741     If this is a request message and none of the above are true, then the
1742     message body length is zero (no message body is present).
1743    </t></x:lt>
1744    <x:lt><t>
1745     Otherwise, this is a response message without a declared message body
1746     length, so the message body length is determined by the number of octets
1747     received prior to the server closing the connection.
1748    </t></x:lt>
1749  </list>
1752   Since there is no way to distinguish a successfully completed,
1753   close-delimited message from a partially-received message interrupted
1754   by network failure, a server &SHOULD; use encoding or
1755   length-delimited messages whenever possible.  The close-delimiting
1756   feature exists primarily for backwards compatibility with HTTP/1.0.
1759   A server &MAY; reject a request that contains a message body but
1760   not a <x:ref>Content-Length</x:ref> by responding with
1761   <x:ref>411 (Length Required)</x:ref>.
1764   Unless a transfer coding other than chunked has been applied,
1765   a client that sends a request containing a message body &SHOULD;
1766   use a valid <x:ref>Content-Length</x:ref> header field if the message body
1767   length is known in advance, rather than the chunked transfer coding, since some
1768   existing services respond to chunked with a <x:ref>411 (Length Required)</x:ref>
1769   status code even though they understand the chunked transfer coding.  This
1770   is typically because such services are implemented via a gateway that
1771   requires a content-length in advance of being called and the server
1772   is unable or unwilling to buffer the entire request before processing.
1775   A user agent that sends a request containing a message body &MUST; send a
1776   valid <x:ref>Content-Length</x:ref> header field if it does not know the
1777   server will handle HTTP/1.1 (or later) requests; such knowledge can be in
1778   the form of specific user configuration or by remembering the version of a
1779   prior received response.
1782   If the final response to the last request on a connection has been
1783   completely received and there remains additional data to read, a user agent
1784   &MAY; discard the remaining data or attempt to determine if that data
1785   belongs as part of the prior response body, which might be the case if the
1786   prior message's Content-Length value is incorrect. A client &MUST-NOT;
1787   process, cache, or forward such extra data as a separate response, since
1788   such behavior would be vulnerable to cache poisoning.
1793<section anchor="incomplete.messages" title="Handling Incomplete Messages">
1795   A server that receives an incomplete request message, usually due to a
1796   canceled request or a triggered time-out exception, &MAY; send an error
1797   response prior to closing the connection.
1800   A client that receives an incomplete response message, which can occur
1801   when a connection is closed prematurely or when decoding a supposedly
1802   chunked transfer coding fails, &MUST; record the message as incomplete.
1803   Cache requirements for incomplete responses are defined in
1804   &cache-incomplete;.
1807   If a response terminates in the middle of the header block (before the
1808   empty line is received) and the status code might rely on header fields to
1809   convey the full meaning of the response, then the client cannot assume
1810   that meaning has been conveyed; the client might need to repeat the
1811   request in order to determine what action to take next.
1814   A message body that uses the chunked transfer coding is
1815   incomplete if the zero-sized chunk that terminates the encoding has not
1816   been received.  A message that uses a valid <x:ref>Content-Length</x:ref> is
1817   incomplete if the size of the message body received (in octets) is less than
1818   the value given by Content-Length.  A response that has neither chunked
1819   transfer coding nor Content-Length is terminated by closure of the
1820   connection, and thus is considered complete regardless of the number of
1821   message body octets received, provided that the header block was received
1822   intact.
1826<section title="Message Parsing Robustness" anchor="message.robustness">
1828   Older HTTP/1.0 user agent implementations might send an extra CRLF
1829   after a POST request as a lame workaround for some early server
1830   applications that failed to read message body content that was
1831   not terminated by a line-ending. An HTTP/1.1 user agent &MUST-NOT;
1832   preface or follow a request with an extra CRLF.  If terminating
1833   the request message body with a line-ending is desired, then the
1834   user agent &MUST; count the terminating CRLF octets as part of the
1835   message body length.
1838   In the interest of robustness, servers &SHOULD; ignore at least one
1839   empty line received where a request-line is expected. In other words, if
1840   a server is reading the protocol stream at the beginning of a
1841   message and receives a CRLF first, the server &SHOULD; ignore the CRLF.
1844   Although the line terminator for the start-line and header
1845   fields is the sequence CRLF, recipients &MAY; recognize a
1846   single LF as a line terminator and ignore any preceding CR.
1849   Although the request-line and status-line grammar rules require that each
1850   of the component elements be separated by a single SP octet, recipients
1851   &MAY; instead parse on whitespace-delimited word boundaries and, aside
1852   from the CRLF terminator, treat any form of whitespace as the SP separator
1853   while ignoring preceding or trailing whitespace;
1854   such whitespace includes one or more of the following octets:
1855   SP, HTAB, VT (%x0B), FF (%x0C), or bare CR.
1858   When a server listening only for HTTP request messages, or processing
1859   what appears from the start-line to be an HTTP request message,
1860   receives a sequence of octets that does not match the HTTP-message
1861   grammar aside from the robustness exceptions listed above, the
1862   server &SHOULD; respond with a <x:ref>400 (Bad Request)</x:ref> response. 
1867<section title="Transfer Codings" anchor="transfer.codings">
1868  <x:anchor-alias value="transfer-coding"/>
1869  <x:anchor-alias value="transfer-extension"/>
1871   Transfer coding names are used to indicate an encoding
1872   transformation that has been, can be, or might need to be applied to a
1873   payload body in order to ensure "safe transport" through the network.
1874   This differs from a content coding in that the transfer coding is a
1875   property of the message rather than a property of the representation
1876   that is being transferred.
1878<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="transfer-coding"/><iref primary="true" item="Grammar" subitem="transfer-extension"/>
1879  <x:ref>transfer-coding</x:ref>    = "chunked" ; <xref target="chunked.encoding"/>
1880                     / "compress" ; <xref target="compress.coding"/>
1881                     / "deflate" ; <xref target="deflate.coding"/>
1882                     / "gzip" ; <xref target="gzip.coding"/>
1883                     / <x:ref>transfer-extension</x:ref>
1884  <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> )
1886<t anchor="rule.parameter">
1887  <x:anchor-alias value="attribute"/>
1888  <x:anchor-alias value="transfer-parameter"/>
1889  <x:anchor-alias value="value"/>
1890   Parameters are in the form of attribute/value pairs.
1892<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"/>
1893  <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>
1894  <x:ref>attribute</x:ref>          = <x:ref>token</x:ref>
1895  <x:ref>value</x:ref>              = <x:ref>word</x:ref>
1898   All transfer-coding names are case-insensitive and ought to be registered
1899   within the HTTP Transfer Coding registry, as defined in
1900   <xref target="transfer.coding.registry"/>.
1901   They are used in the <x:ref>TE</x:ref> (<xref target="header.te"/>) and
1902   <x:ref>Transfer-Encoding</x:ref> (<xref target="header.transfer-encoding"/>)
1903   header fields.
1906<section title="Chunked Transfer Coding" anchor="chunked.encoding">
1907  <iref primary="true" item="chunked (Coding Format)"/>
1908  <x:anchor-alias value="chunk"/>
1909  <x:anchor-alias value="chunked-body"/>
1910  <x:anchor-alias value="chunk-data"/>
1911  <x:anchor-alias value="chunk-ext"/>
1912  <x:anchor-alias value="chunk-ext-name"/>
1913  <x:anchor-alias value="chunk-ext-val"/>
1914  <x:anchor-alias value="chunk-size"/>
1915  <x:anchor-alias value="last-chunk"/>
1916  <x:anchor-alias value="trailer-part"/>
1917  <x:anchor-alias value="quoted-str-nf"/>
1918  <x:anchor-alias value="qdtext-nf"/>
1920   The chunked transfer coding modifies the body of a message in order to
1921   transfer it as a series of chunks, each with its own size indicator,
1922   followed by an &OPTIONAL; trailer containing header fields. This
1923   allows dynamically generated content to be transferred along with the
1924   information necessary for the recipient to verify that it has
1925   received the full message.
1927<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"/>
1928  <x:ref>chunked-body</x:ref>   = *<x:ref>chunk</x:ref>
1929                   <x:ref>last-chunk</x:ref>
1930                   <x:ref>trailer-part</x:ref>
1931                   <x:ref>CRLF</x:ref>
1933  <x:ref>chunk</x:ref>          = <x:ref>chunk-size</x:ref> [ <x:ref>chunk-ext</x:ref> ] <x:ref>CRLF</x:ref>
1934                   <x:ref>chunk-data</x:ref> <x:ref>CRLF</x:ref>
1935  <x:ref>chunk-size</x:ref>     = 1*<x:ref>HEXDIG</x:ref>
1936  <x:ref>last-chunk</x:ref>     = 1*("0") [ <x:ref>chunk-ext</x:ref> ] <x:ref>CRLF</x:ref>
1938  <x:ref>chunk-ext</x:ref>      = *( ";" <x:ref>chunk-ext-name</x:ref> [ "=" <x:ref>chunk-ext-val</x:ref> ] )
1939  <x:ref>chunk-ext-name</x:ref> = <x:ref>token</x:ref>
1940  <x:ref>chunk-ext-val</x:ref>  = <x:ref>token</x:ref> / <x:ref>quoted-str-nf</x:ref>
1941  <x:ref>chunk-data</x:ref>     = 1*<x:ref>OCTET</x:ref> ; a sequence of chunk-size octets
1942  <x:ref>trailer-part</x:ref>   = *( <x:ref>header-field</x:ref> <x:ref>CRLF</x:ref> )
1944  <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>
1945                 ; like <x:ref>quoted-string</x:ref>, but disallowing line folding
1946  <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>
1949   Chunk extensions within the chunked transfer coding are deprecated.
1950   Senders &SHOULD-NOT; send chunk-ext.
1951   Definition of new chunk extensions is discouraged.
1954   The chunk-size field is a string of hex digits indicating the size of
1955   the chunk-data in octets. The chunked transfer coding is complete when a
1956   chunk with a chunk-size of zero is received, possibly followed by a
1957   trailer, and finally terminated by an empty line.
1960<section title="Trailer" anchor="header.trailer">
1961  <iref primary="true" item="Trailer header field" x:for-anchor=""/>
1962  <x:anchor-alias value="Trailer"/>
1964   A trailer allows the sender to include additional fields at the end of a
1965   chunked message in order to supply metadata that might be dynamically
1966   generated while the message body is sent, such as a message integrity
1967   check, digital signature, or post-processing status.
1968   The trailer &MUST-NOT; contain fields that need to be known before a
1969   recipient processes the body, such as <x:ref>Transfer-Encoding</x:ref>,
1970   <x:ref>Content-Length</x:ref>, and <x:ref>Trailer</x:ref>.
1973   When a message includes a message body encoded with the chunked
1974   transfer coding and the sender desires to send metadata in the form of
1975   trailer fields at the end of the message, the sender &SHOULD; send a
1976   <x:ref>Trailer</x:ref> header field before the message body to indicate
1977   which fields will be present in the trailers. This allows the recipient
1978   to prepare for receipt of that metadata before it starts processing the body,
1979   which is useful if the message is being streamed and the recipient wishes
1980   to confirm an integrity check on the fly.
1982<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Trailer"/>
1983  <x:ref>Trailer</x:ref> = 1#<x:ref>field-name</x:ref>
1986   If no <x:ref>Trailer</x:ref> header field is present, the sender of a
1987   chunked message body &SHOULD; send an empty trailer.
1990   A server &MUST; send an empty trailer with the chunked transfer coding
1991   unless at least one of the following is true:
1992  <list style="numbers">
1993    <t>the request included a <x:ref>TE</x:ref> header field that indicates
1994    "trailers" is acceptable in the transfer coding of the response, as
1995    described in <xref target="header.te"/>; or,</t>
1997    <t>the trailer fields consist entirely of optional metadata and the
1998    recipient could use the message (in a manner acceptable to the server where
1999    the field originated) without receiving that metadata. In other words,
2000    the server that generated the header field is willing to accept the
2001    possibility that the trailer fields might be silently discarded along
2002    the path to the client.</t>
2003  </list>
2006   The above requirement prevents the need for an infinite buffer when a
2007   message is being received by an HTTP/1.1 (or later) proxy and forwarded to
2008   an HTTP/1.0 recipient.
2012<section title="Decoding chunked" anchor="decoding.chunked">
2014   A process for decoding the chunked transfer coding
2015   can be represented in pseudo-code as:
2017<figure><artwork type="code">
2018  length := 0
2019  read chunk-size, chunk-ext (if any) and CRLF
2020  while (chunk-size &gt; 0) {
2021     read chunk-data and CRLF
2022     append chunk-data to decoded-body
2023     length := length + chunk-size
2024     read chunk-size and CRLF
2025  }
2026  read header-field
2027  while (header-field not empty) {
2028     append header-field to existing header fields
2029     read header-field
2030  }
2031  Content-Length := length
2032  Remove "chunked" from Transfer-Encoding
2033  Remove Trailer from existing header fields
2036   All recipients &MUST; be able to receive and decode the
2037   chunked transfer coding and &MUST; ignore chunk-ext extensions
2038   they do not understand.
2043<section title="Compression Codings" anchor="compression.codings">
2045   The codings defined below can be used to compress the payload of a
2046   message.
2049<section title="Compress Coding" anchor="compress.coding">
2050<iref item="compress (Coding Format)"/>
2052   The "compress" format is produced by the common UNIX file compression
2053   program "compress". This format is an adaptive Lempel-Ziv-Welch
2054   coding (LZW). Recipients &SHOULD; consider "x-compress" to be
2055   equivalent to "compress".
2059<section title="Deflate Coding" anchor="deflate.coding">
2060<iref item="deflate (Coding Format)"/>
2062   The "deflate" format is defined as the "deflate" compression mechanism
2063   (described in <xref target="RFC1951"/>) used inside the "zlib"
2064   data format (<xref target="RFC1950"/>).
2067  <t>
2068    &Note; Some incorrect implementations send the "deflate"
2069    compressed data without the zlib wrapper.
2070   </t>
2074<section title="Gzip Coding" anchor="gzip.coding">
2075<iref item="gzip (Coding Format)"/>
2077   The "gzip" format is produced by the file compression program
2078   "gzip" (GNU zip), as described in <xref target="RFC1952"/>. This format is a
2079   Lempel-Ziv coding (LZ77) with a 32 bit CRC.
2080   Recipients &SHOULD; consider "x-gzip" to be equivalent to "gzip".
2086<section title="TE" anchor="header.te">
2087  <iref primary="true" item="TE header field" x:for-anchor=""/>
2088  <x:anchor-alias value="TE"/>
2089  <x:anchor-alias value="t-codings"/>
2090  <x:anchor-alias value="t-ranking"/>
2091  <x:anchor-alias value="rank"/>
2093   The "TE" header field in a request indicates what transfer codings,
2094   besides chunked, the client is willing to accept in response, and
2095   whether or not the client is willing to accept trailer fields in a
2096   chunked transfer coding.
2099   The TE field-value consists of a comma-separated list of transfer coding
2100   names, each allowing for optional parameters (as described in
2101   <xref target="transfer.codings"/>), and/or the keyword "trailers".
2102   Clients &MUST-NOT; send the chunked transfer coding name in TE;
2103   chunked is always acceptable for HTTP/1.1 recipients.
2105<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"/>
2106  <x:ref>TE</x:ref>        = #<x:ref>t-codings</x:ref>
2107  <x:ref>t-codings</x:ref> = "trailers" / ( <x:ref>transfer-coding</x:ref> [ <x:ref>t-ranking</x:ref> ] )
2108  <x:ref>t-ranking</x:ref> = <x:ref>OWS</x:ref> ";" <x:ref>OWS</x:ref> "q=" <x:ref>rank</x:ref>
2109  <x:ref>rank</x:ref>      = ( "0" [ "." 0*3<x:ref>DIGIT</x:ref> ] )
2110             / ( "1" [ "." 0*3("0") ] )
2113   Three examples of TE use are below.
2115<figure><artwork type="example">
2116  TE: deflate
2117  TE:
2118  TE: trailers, deflate;q=0.5
2121   The presence of the keyword "trailers" indicates that the client is
2122   willing to accept trailer fields in a chunked transfer coding,
2123   as defined in <xref target="chunked.encoding"/>, on behalf of itself and
2124   any downstream clients. For chained requests, this implies that either:
2125   (a) all downstream clients are willing to accept trailer fields in the
2126   forwarded response; or,
2127   (b) the client will attempt to buffer the response on behalf of downstream
2128   recipients.
2129   Note that HTTP/1.1 does not define any means to limit the size of a
2130   chunked response such that a client can be assured of buffering the
2131   entire response.
2134   When multiple transfer codings are acceptable, the client &MAY; rank the
2135   codings by preference using a case-insensitive "q" parameter (similar to
2136   the qvalues used in content negotiation fields, &qvalue;). The rank value
2137   is a real number in the range 0 through 1, where 0.001 is the least
2138   preferred and 1 is the most preferred; a value of 0 means "not acceptable".
2141   If the TE field-value is empty or if no TE field is present, the only
2142   acceptable transfer coding is chunked. A message with no transfer coding
2143   is always acceptable.
2146   Since the TE header field only applies to the immediate connection,
2147   a sender of TE &MUST; also send a "TE" connection option within the
2148   <x:ref>Connection</x:ref> header field (<xref target="header.connection"/>)
2149   in order to prevent the TE field from being forwarded by intermediaries
2150   that do not support its semantics.
2155<section title="Message Routing" anchor="message.routing">
2157   HTTP request message routing is determined by each client based on the
2158   target resource, the client's proxy configuration, and
2159   establishment or reuse of an inbound connection.  The corresponding
2160   response routing follows the same connection chain back to the client.
2163<section title="Identifying a Target Resource" anchor="target-resource">
2164  <iref primary="true" item="target resource"/>
2165  <iref primary="true" item="target URI"/>
2166  <x:anchor-alias value="target resource"/>
2167  <x:anchor-alias value="target URI"/>
2169   HTTP is used in a wide variety of applications, ranging from
2170   general-purpose computers to home appliances.  In some cases,
2171   communication options are hard-coded in a client's configuration.
2172   However, most HTTP clients rely on the same resource identification
2173   mechanism and configuration techniques as general-purpose Web browsers.
2176   HTTP communication is initiated by a user agent for some purpose.
2177   The purpose is a combination of request semantics, which are defined in
2178   <xref target="Part2"/>, and a target resource upon which to apply those
2179   semantics.  A URI reference (<xref target="uri"/>) is typically used as
2180   an identifier for the "<x:dfn>target resource</x:dfn>", which a user agent
2181   would resolve to its absolute form in order to obtain the
2182   "<x:dfn>target URI</x:dfn>".  The target URI
2183   excludes the reference's fragment identifier component, if any,
2184   since fragment identifiers are reserved for client-side processing
2185   (<xref target="RFC3986" x:fmt="," x:sec="3.5"/>).
2189<section title="Connecting Inbound" anchor="connecting.inbound">
2191   Once the target URI is determined, a client needs to decide whether
2192   a network request is necessary to accomplish the desired semantics and,
2193   if so, where that request is to be directed.
2196   If the client has a response cache and the request semantics can be
2197   satisfied by a cache (<xref target="Part6"/>), then the request is
2198   usually directed to the cache first.
2201   If the request is not satisfied by a cache, then a typical client will
2202   check its configuration to determine whether a proxy is to be used to
2203   satisfy the request.  Proxy configuration is implementation-dependent,
2204   but is often based on URI prefix matching, selective authority matching,
2205   or both, and the proxy itself is usually identified by an "http" or
2206   "https" URI.  If a proxy is applicable, the client connects inbound by
2207   establishing (or reusing) a connection to that proxy.
2210   If no proxy is applicable, a typical client will invoke a handler routine,
2211   usually specific to the target URI's scheme, to connect directly
2212   to an authority for the target resource.  How that is accomplished is
2213   dependent on the target URI scheme and defined by its associated
2214   specification, similar to how this specification defines origin server
2215   access for resolution of the "http" (<xref target="http.uri"/>) and
2216   "https" (<xref target="https.uri"/>) schemes.
2219   HTTP requirements regarding connection management are defined in
2220   <xref target=""/>.
2224<section title="Request Target" anchor="request-target">
2226   Once an inbound connection is obtained,
2227   the client sends an HTTP request message (<xref target="http.message"/>)
2228   with a request-target derived from the target URI.
2229   There are four distinct formats for the request-target, depending on both
2230   the method being requested and whether the request is to a proxy.
2232<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"/>
2233  <x:ref>request-target</x:ref> = <x:ref>origin-form</x:ref>
2234                 / <x:ref>absolute-form</x:ref>
2235                 / <x:ref>authority-form</x:ref>
2236                 / <x:ref>asterisk-form</x:ref>
2238  <x:ref>origin-form</x:ref>    = <x:ref>absolute-path</x:ref> [ "?" <x:ref>query</x:ref> ]
2239  <x:ref>absolute-form</x:ref>  = <x:ref>absolute-URI</x:ref>
2240  <x:ref>authority-form</x:ref> = <x:ref>authority</x:ref>
2241  <x:ref>asterisk-form</x:ref>  = "*"
2243<t anchor="origin-form"><iref item="origin-form (of request-target)"/>
2244  <x:h>origin-form</x:h>
2247   The most common form of request-target is the <x:dfn>origin-form</x:dfn>.
2248   When making a request directly to an origin server, other than a CONNECT
2249   or server-wide OPTIONS request (as detailed below),
2250   a client &MUST; send only the absolute path and query components of
2251   the target URI as the request-target.
2252   If the target URI's path component is empty, then the client &MUST; send
2253   "/" as the path within the origin-form of request-target.
2254   A <x:ref>Host</x:ref> header field is also sent, as defined in
2255   <xref target=""/>, containing the target URI's
2256   authority component (excluding any userinfo).
2259   For example, a client wishing to retrieve a representation of the resource
2260   identified as
2262<figure><artwork x:indent-with="  " type="example">
2266   directly from the origin server would open (or reuse) a TCP connection
2267   to port 80 of the host "" and send the lines:
2269<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2270GET /where?q=now HTTP/1.1
2274   followed by the remainder of the request message.
2276<t anchor="absolute-form"><iref item="absolute-form (of request-target)"/>
2277  <x:h>absolute-form</x:h>
2280   When making a request to a proxy, other than a CONNECT or server-wide
2281   OPTIONS request (as detailed below), a client &MUST; send the target URI
2282   in <x:dfn>absolute-form</x:dfn> as the request-target.
2283   The proxy is requested to either service that request from a valid cache,
2284   if possible, or make the same request on the client's behalf to either
2285   the next inbound proxy server or directly to the origin server indicated
2286   by the request-target.  Requirements on such "forwarding" of messages are
2287   defined in <xref target="message.forwarding"/>.
2290   An example absolute-form of request-line would be:
2292<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2293GET HTTP/1.1
2296   To allow for transition to the absolute-form for all requests in some
2297   future version of HTTP, HTTP/1.1 servers &MUST; accept the absolute-form
2298   in requests, even though HTTP/1.1 clients will only send them in requests
2299   to proxies.
2301<t anchor="authority-form"><iref item="authority-form (of request-target)"/>
2302  <x:h>authority-form</x:h>
2305   The <x:dfn>authority-form</x:dfn> of request-target is only used for CONNECT requests
2306   (&CONNECT;).  When making a CONNECT request to establish a tunnel through
2307   one or more proxies, a client &MUST; send only the target URI's
2308   authority component (excluding any userinfo) as the request-target.
2309   For example,
2311<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2314<t anchor="asterisk-form"><iref item="asterisk-form (of request-target)"/>
2315  <x:h>asterisk-form</x:h>
2318   The <x:dfn>asterisk-form</x:dfn> of request-target is only used for a server-wide
2319   OPTIONS request (&OPTIONS;).  When a client wishes to request OPTIONS
2320   for the server as a whole, as opposed to a specific named resource of
2321   that server, the client &MUST; send only "*" (%x2A) as the request-target.
2322   For example,
2324<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2325OPTIONS * HTTP/1.1
2328   If a proxy receives an OPTIONS request with an absolute-form of
2329   request-target in which the URI has an empty path and no query component,
2330   then the last proxy on the request chain &MUST; send a request-target
2331   of "*" when it forwards the request to the indicated origin server.
2334   For example, the request
2335</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2339  would be forwarded by the final proxy as
2340</preamble><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2341OPTIONS * HTTP/1.1
2345   after connecting to port 8001 of host "".
2350<section title="Host" anchor="">
2351  <iref primary="true" item="Host header field" x:for-anchor=""/>
2352  <x:anchor-alias value="Host"/>
2354   The "Host" header field in a request provides the host and port
2355   information from the target URI, enabling the origin
2356   server to distinguish among resources while servicing requests
2357   for multiple host names on a single IP address.  Since the Host
2358   field-value is critical information for handling a request, it
2359   &SHOULD; be sent as the first header field following the request-line.
2361<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Host"/>
2362  <x:ref>Host</x:ref> = <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ; <xref target="http.uri"/>
2365   A client &MUST; send a Host header field in all HTTP/1.1 request
2366   messages.  If the target URI includes an authority component, then
2367   the Host field-value &MUST; be identical to that authority component
2368   after excluding any userinfo (<xref target="http.uri"/>).
2369   If the authority component is missing or undefined for the target URI,
2370   then the Host header field &MUST; be sent with an empty field-value.
2373   For example, a GET request to the origin server for
2374   &lt;; would begin with:
2376<figure><artwork type="message/http; msgtype=&#34;request&#34;" x:indent-with="  ">
2377GET /pub/WWW/ HTTP/1.1
2381   The Host header field &MUST; be sent in an HTTP/1.1 request even
2382   if the request-target is in the absolute-form, since this
2383   allows the Host information to be forwarded through ancient HTTP/1.0
2384   proxies that might not have implemented Host.
2387   When a proxy receives a request with an absolute-form of
2388   request-target, the proxy &MUST; ignore the received
2389   Host header field (if any) and instead replace it with the host
2390   information of the request-target.  If the proxy forwards the request,
2391   it &MUST; generate a new Host field-value based on the received
2392   request-target rather than forward the received Host field-value.
2395   Since the Host header field acts as an application-level routing
2396   mechanism, it is a frequent target for malware seeking to poison
2397   a shared cache or redirect a request to an unintended server.
2398   An interception proxy is particularly vulnerable if it relies on
2399   the Host field-value for redirecting requests to internal
2400   servers, or for use as a cache key in a shared cache, without
2401   first verifying that the intercepted connection is targeting a
2402   valid IP address for that host.
2405   A server &MUST; respond with a <x:ref>400 (Bad Request)</x:ref> status code
2406   to any HTTP/1.1 request message that lacks a Host header field and
2407   to any request message that contains more than one Host header field
2408   or a Host header field with an invalid field-value.
2412<section title="Effective Request URI" anchor="effective.request.uri">
2413  <iref primary="true" item="effective request URI"/>
2414  <x:anchor-alias value="effective request URI"/>
2416   A server that receives an HTTP request message &MUST; reconstruct
2417   the user agent's original target URI, based on the pieces of information
2418   learned from the request-target, <x:ref>Host</x:ref> header field, and
2419   connection context, in order to identify the intended target resource and
2420   properly service the request. The URI derived from this reconstruction
2421   process is referred to as the "<x:dfn>effective request URI</x:dfn>".
2424   For a user agent, the effective request URI is the target URI.
2427   If the request-target is in absolute-form, then the effective request URI
2428   is the same as the request-target.  Otherwise, the effective request URI
2429   is constructed as follows.
2432   If the request is received over a TLS-secured TCP connection,
2433   then the effective request URI's scheme is "https"; otherwise, the
2434   scheme is "http".
2437   If the request-target is in authority-form, then the effective
2438   request URI's authority component is the same as the request-target.
2439   Otherwise, if a <x:ref>Host</x:ref> header field is supplied with a
2440   non-empty field-value, then the authority component is the same as the
2441   Host field-value. Otherwise, the authority component is the concatenation of
2442   the default host name configured for the server, a colon (":"), and the
2443   connection's incoming TCP port number in decimal form.
2446   If the request-target is in authority-form or asterisk-form, then the
2447   effective request URI's combined path and query component is empty.
2448   Otherwise, the combined path and query component is the same as the
2449   request-target.
2452   The components of the effective request URI, once determined as above,
2453   can be combined into absolute-URI form by concatenating the scheme,
2454   "://", authority, and combined path and query component.
2458   Example 1: the following message received over an insecure TCP connection
2460<artwork type="example" x:indent-with="  ">
2461GET /pub/WWW/TheProject.html HTTP/1.1
2467  has an effective request URI of
2469<artwork type="example" x:indent-with="  ">
2475   Example 2: the following message received over a TLS-secured TCP connection
2477<artwork type="example" x:indent-with="  ">
2478OPTIONS * HTTP/1.1
2484  has an effective request URI of
2486<artwork type="example" x:indent-with="  ">
2491   An origin server that does not allow resources to differ by requested
2492   host &MAY; ignore the <x:ref>Host</x:ref> field-value and instead replace it
2493   with a configured server name when constructing the effective request URI.
2496   Recipients of an HTTP/1.0 request that lacks a <x:ref>Host</x:ref> header
2497   field &MAY; attempt to use heuristics (e.g., examination of the URI path for
2498   something unique to a particular host) in order to guess the
2499   effective request URI's authority component.
2503<section title="Associating a Response to a Request" anchor="">
2505   HTTP does not include a request identifier for associating a given
2506   request message with its corresponding one or more response messages.
2507   Hence, it relies on the order of response arrival to correspond exactly
2508   to the order in which requests are made on the same connection.
2509   More than one response message per request only occurs when one or more
2510   informational responses (<x:ref>1xx</x:ref>, see &status-1xx;) precede a
2511   final response to the same request.
2514   A client that has more than one outstanding request on a connection &MUST;
2515   maintain a list of outstanding requests in the order sent and &MUST;
2516   associate each received response message on that connection to the highest
2517   ordered request that has not yet received a final (non-<x:ref>1xx</x:ref>)
2518   response.
2522<section title="Message Forwarding" anchor="message.forwarding">
2524   As described in <xref target="intermediaries"/>, intermediaries can serve
2525   a variety of roles in the processing of HTTP requests and responses.
2526   Some intermediaries are used to improve performance or availability.
2527   Others are used for access control or to filter content.
2528   Since an HTTP stream has characteristics similar to a pipe-and-filter
2529   architecture, there are no inherent limits to the extent an intermediary
2530   can enhance (or interfere) with either direction of the stream.
2533   Intermediaries that forward a message &MUST; implement the
2534   <x:ref>Connection</x:ref> header field, as specified in
2535   <xref target="header.connection"/>, to exclude fields that are only
2536   intended for the incoming connection.
2539   In order to avoid request loops, a proxy that forwards requests to other
2540   proxies &MUST; be able to recognize and exclude all of its own server
2541   names, including any aliases, local variations, or literal IP addresses.
2544<section title="Via" anchor="header.via">
2545  <iref primary="true" item="Via header field" x:for-anchor=""/>
2546  <x:anchor-alias value="pseudonym"/>
2547  <x:anchor-alias value="received-by"/>
2548  <x:anchor-alias value="received-protocol"/>
2549  <x:anchor-alias value="Via"/>
2551   The "Via" header field &MUST; be sent by a proxy or gateway in forwarded
2552   messages to indicate the intermediate protocols and recipients between the
2553   user agent and the server on requests, and between the origin server and
2554   the client on responses. It is analogous to the "Received" field
2555   used by email systems (<xref target="RFC5322" x:fmt="of" x:sec="3.6.7"/>).
2556   Via is used in HTTP for tracking message forwards,
2557   avoiding request loops, and identifying the protocol capabilities of
2558   all senders along the request/response chain.
2560<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"/>
2561  <x:ref>Via</x:ref>               = 1#( <x:ref>received-protocol</x:ref> <x:ref>RWS</x:ref> <x:ref>received-by</x:ref>
2562                          [ <x:ref>RWS</x:ref> <x:ref>comment</x:ref> ] )
2563  <x:ref>received-protocol</x:ref> = [ <x:ref>protocol-name</x:ref> "/" ] <x:ref>protocol-version</x:ref>
2564                      ; see <xref target="header.upgrade"/>
2565  <x:ref>received-by</x:ref>       = ( <x:ref>uri-host</x:ref> [ ":" <x:ref>port</x:ref> ] ) / <x:ref>pseudonym</x:ref>
2566  <x:ref>pseudonym</x:ref>         = <x:ref>token</x:ref>
2569   The received-protocol indicates the protocol version of the message
2570   received by the server or client along each segment of the
2571   request/response chain. The received-protocol version is appended to
2572   the Via field value when the message is forwarded so that information
2573   about the protocol capabilities of upstream applications remains
2574   visible to all recipients.
2577   The protocol-name is excluded if and only if it would be "HTTP". The
2578   received-by field is normally the host and optional port number of a
2579   recipient server or client that subsequently forwarded the message.
2580   However, if the real host is considered to be sensitive information,
2581   it &MAY; be replaced by a pseudonym. If the port is not given, it &MAY;
2582   be assumed to be the default port of the received-protocol.
2585   Multiple Via field values represent each proxy or gateway that has
2586   forwarded the message. Each recipient &MUST; append its information
2587   such that the end result is ordered according to the sequence of
2588   forwarding applications.
2591   Comments &MAY; be used in the Via header field to identify the software
2592   of each recipient, analogous to the <x:ref>User-Agent</x:ref> and
2593   <x:ref>Server</x:ref> header fields. However, all comments in the Via field
2594   are optional and &MAY; be removed by any recipient prior to forwarding the
2595   message.
2598   For example, a request message could be sent from an HTTP/1.0 user
2599   agent to an internal proxy code-named "fred", which uses HTTP/1.1 to
2600   forward the request to a public proxy at, which completes
2601   the request by forwarding it to the origin server at
2602   The request received by would then have the following
2603   Via header field:
2605<figure><artwork type="example">
2606  Via: 1.0 fred, 1.1 (Apache/1.1)
2609   A proxy or gateway used as a portal through a network firewall
2610   &SHOULD-NOT; forward the names and ports of hosts within the firewall
2611   region unless it is explicitly enabled to do so. If not enabled, the
2612   received-by host of any host behind the firewall &SHOULD; be replaced
2613   by an appropriate pseudonym for that host.
2616   A proxy or gateway &MAY; combine an ordered subsequence of Via header
2617   field entries into a single such entry if the entries have identical
2618   received-protocol values. For example,
2620<figure><artwork type="example">
2621  Via: 1.0 ricky, 1.1 ethel, 1.1 fred, 1.0 lucy
2624  could be collapsed to
2626<figure><artwork type="example">
2627  Via: 1.0 ricky, 1.1 mertz, 1.0 lucy
2630   Senders &SHOULD-NOT; combine multiple entries unless they are all
2631   under the same organizational control and the hosts have already been
2632   replaced by pseudonyms. Senders &MUST-NOT; combine entries that
2633   have different received-protocol values.
2637<section title="Transformations" anchor="message.transformations">
2639   Some intermediaries include features for transforming messages and their
2640   payloads.  A transforming proxy might, for example, convert between image
2641   formats in order to save cache space or to reduce the amount of traffic on
2642   a slow link. However, operational problems might occur when these
2643   transformations are applied to payloads intended for critical applications,
2644   such as medical imaging or scientific data analysis, particularly when
2645   integrity checks or digital signatures are used to ensure that the payload
2646   received is identical to the original.
2649   If a proxy receives a request-target with a host name that is not a
2650   fully qualified domain name, it &MAY; add its own domain to the host name
2651   it received when forwarding the request.  A proxy &MUST-NOT; change the
2652   host name if it is a fully qualified domain name.
2655   A proxy &MUST-NOT; modify the "absolute-path" and "query" parts of the
2656   received request-target when forwarding it to the next inbound server,
2657   except as noted above to replace an empty path with "/" or "*".
2660   A proxy &MUST-NOT; modify header fields that provide information about the
2661   end points of the communication chain, the resource state, or the selected
2662   representation. A proxy &MAY; change the message body through application
2663   or removal of a transfer coding (<xref target="transfer.codings"/>).
2666   A non-transforming proxy &MUST; preserve the message payload (&payload;).
2667   A transforming proxy &MUST; preserve the payload of a message that
2668   contains the no-transform cache-control directive.
2671   A transforming proxy &MAY; transform the payload of a message
2672   that does not contain the no-transform cache-control directive;
2673   if the payload is transformed, the transforming proxy &MUST; add a
2674   Warning 214 (Transformation applied) header field if one does not
2675   already appear in the message (see &header-warning;).
2681<section title="Connection Management" anchor="">
2683   HTTP messaging is independent of the underlying transport or
2684   session-layer connection protocol(s).  HTTP only presumes a reliable
2685   transport with in-order delivery of requests and the corresponding
2686   in-order delivery of responses.  The mapping of HTTP request and
2687   response structures onto the data units of an underlying transport
2688   protocol is outside the scope of this specification.
2691   As described in <xref target="connecting.inbound"/>, the specific
2692   connection protocols to be used for an HTTP interaction are determined by
2693   client configuration and the <x:ref>target URI</x:ref>.
2694   For example, the "http" URI scheme
2695   (<xref target="http.uri"/>) indicates a default connection of TCP
2696   over IP, with a default TCP port of 80, but the client might be
2697   configured to use a proxy via some other connection, port, or protocol.
2700   HTTP implementations are expected to engage in connection management,
2701   which includes maintaining the state of current connections,
2702   establishing a new connection or reusing an existing connection,
2703   processing messages received on a connection, detecting connection
2704   failures, and closing each connection.
2705   Most clients maintain multiple connections in parallel, including
2706   more than one connection per server endpoint.
2707   Most servers are designed to maintain thousands of concurrent connections,
2708   while controlling request queues to enable fair use and detect
2709   denial of service attacks.
2712<section title="Connection" anchor="header.connection">
2713  <iref primary="true" item="Connection header field" x:for-anchor=""/>
2714  <iref primary="true" item="close" x:for-anchor=""/>
2715  <x:anchor-alias value="Connection"/>
2716  <x:anchor-alias value="connection-option"/>
2717  <x:anchor-alias value="close"/>
2719   The "Connection" header field allows the sender to indicate desired
2720   control options for the current connection.  In order to avoid confusing
2721   downstream recipients, a proxy or gateway &MUST; remove or replace any
2722   received connection options before forwarding the message.
2725   When a header field aside from Connection is used to supply control
2726   information for or about the current connection, the sender &MUST; list
2727   the corresponding field-name within the "Connection" header field.
2728   A proxy or gateway &MUST; parse a received Connection
2729   header field before a message is forwarded and, for each
2730   connection-option in this field, remove any header field(s) from
2731   the message with the same name as the connection-option, and then
2732   remove the Connection header field itself (or replace it with the
2733   intermediary's own connection options for the forwarded message).
2736   Hence, the Connection header field provides a declarative way of
2737   distinguishing header fields that are only intended for the
2738   immediate recipient ("hop-by-hop") from those fields that are
2739   intended for all recipients on the chain ("end-to-end"), enabling the
2740   message to be self-descriptive and allowing future connection-specific
2741   extensions to be deployed without fear that they will be blindly
2742   forwarded by older intermediaries.
2745   The Connection header field's value has the following grammar:
2747<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Connection"/><iref primary="true" item="Grammar" subitem="connection-option"/>
2748  <x:ref>Connection</x:ref>        = 1#<x:ref>connection-option</x:ref>
2749  <x:ref>connection-option</x:ref> = <x:ref>token</x:ref>
2752   Connection options are case-insensitive.
2755   A sender &MUST-NOT; send a connection option corresponding to a header
2756   field that is intended for all recipients of the payload.
2757   For example, <x:ref>Cache-Control</x:ref> is never appropriate as a
2758   connection option (&header-cache-control;).
2761   The connection options do not have to correspond to a header field
2762   present in the message, since a connection-specific header field
2763   might not be needed if there are no parameters associated with that
2764   connection option.  Recipients that trigger certain connection
2765   behavior based on the presence of connection options &MUST; do so
2766   based on the presence of the connection-option rather than only the
2767   presence of the optional header field.  In other words, if the
2768   connection option is received as a header field but not indicated
2769   within the Connection field-value, then the recipient &MUST; ignore
2770   the connection-specific header field because it has likely been
2771   forwarded by an intermediary that is only partially conformant.
2774   When defining new connection options, specifications ought to
2775   carefully consider existing deployed header fields and ensure
2776   that the new connection option does not share the same name as
2777   an unrelated header field that might already be deployed.
2778   Defining a new connection option essentially reserves that potential
2779   field-name for carrying additional information related to the
2780   connection option, since it would be unwise for senders to use
2781   that field-name for anything else.
2784   The "<x:dfn>close</x:dfn>" connection option is defined for a
2785   sender to signal that this connection will be closed after completion of
2786   the response. For example,
2788<figure><artwork type="example">
2789  Connection: close
2792   in either the request or the response header fields indicates that
2793   the connection &MUST; be closed after the current request/response
2794   is complete (<xref target="persistent.tear-down"/>).
2797   A client that does not support <x:ref>persistent connections</x:ref> &MUST;
2798   send the "close" connection option in every request message.
2801   A server that does not support <x:ref>persistent connections</x:ref> &MUST;
2802   send the "close" connection option in every response message that
2803   does not have a <x:ref>1xx (Informational)</x:ref> status code.
2807<section title="Establishment" anchor="persistent.establishment">
2809   It is beyond the scope of this specification to describe how connections
2810   are established via various transport or session-layer protocols.
2811   Each connection applies to only one transport link.
2815<section title="Persistence" anchor="persistent.connections">
2816   <x:anchor-alias value="persistent connections"/>
2818   HTTP/1.1 defaults to the use of "<x:dfn>persistent connections</x:dfn>",
2819   allowing multiple requests and responses to be carried over a single
2820   connection. The "<x:ref>close</x:ref>" connection-option is used to signal
2821   that a connection will not persist after the current request/response.
2822   HTTP implementations &SHOULD; support persistent connections.
2825   A recipient determines whether a connection is persistent or not based on
2826   the most recently received message's protocol version and
2827   <x:ref>Connection</x:ref> header field (if any):
2828   <list style="symbols">
2829     <t>If the <x:ref>close</x:ref> connection option is present, the
2830        connection will not persist after the current response; else,</t>
2831     <t>If the received protocol is HTTP/1.1 (or later), the connection will
2832        persist after the current response; else,</t>
2833     <t>If the received protocol is HTTP/1.0, the "keep-alive"
2834        connection option is present, the recipient is not a proxy, and
2835        the recipient wishes to honor the HTTP/1.0 "keep-alive" mechanism,
2836        the connection will persist after the current response; otherwise,</t>
2837     <t>The connection will close after the current response.</t>
2838   </list>
2841   A server &MAY; assume that an HTTP/1.1 client intends to maintain a
2842   persistent connection until a <x:ref>close</x:ref> connection option
2843   is received in a request.
2846   A client &MAY; reuse a persistent connection until it sends or receives
2847   a <x:ref>close</x:ref> connection option or receives an HTTP/1.0 response
2848   without a "keep-alive" connection option.
2851   In order to remain persistent, all messages on a connection &MUST;
2852   have a self-defined message length (i.e., one not defined by closure
2853   of the connection), as described in <xref target="message.body"/>.
2854   A server &MUST; read the entire request message body or close
2855   the connection after sending its response, since otherwise the
2856   remaining data on a persistent connection would be misinterpreted
2857   as the next request.  Likewise,
2858   a client &MUST; read the entire response message body if it intends
2859   to reuse the same connection for a subsequent request.
2862   A proxy server &MUST-NOT; maintain a persistent connection with an
2863   HTTP/1.0 client (see <xref x:sec="19.7.1" x:fmt="of" target="RFC2068"/> for
2864   information and discussion of the problems with the Keep-Alive header field
2865   implemented by many HTTP/1.0 clients).
2868   Clients and servers &SHOULD-NOT; assume that a persistent connection is
2869   maintained for HTTP versions less than 1.1 unless it is explicitly
2870   signaled.
2871   See <xref target="compatibility.with.http.1.0.persistent.connections"/>
2872   for more information on backward compatibility with HTTP/1.0 clients.
2875<section title="Retrying Requests" anchor="persistent.retrying.requests">
2877   Connections can be closed at any time, with or without intention.
2878   Implementations ought to anticipate the need to recover
2879   from asynchronous close events.
2882   When an inbound connection is closed prematurely, a client &MAY; open a new
2883   connection and automatically retransmit an aborted sequence of requests if
2884   all of those requests have idempotent methods (&idempotent-methods;).
2885   A proxy &MUST-NOT; automatically retry non-idempotent requests.
2888   A user agent &MUST-NOT; automatically retry a request with a non-idempotent
2889   method unless it has some means to know that the request semantics are
2890   actually idempotent, regardless of the method, or some means to detect that
2891   the original request was never applied. For example, a user agent that
2892   knows (through design or configuration) that a POST request to a given
2893   resource is safe can repeat that request automatically.
2894   Likewise, a user agent designed specifically to operate on a version
2895   control repository might be able to recover from partial failure conditions
2896   by checking the target resource revision(s) after a failed connection,
2897   reverting or fixing any changes that were partially applied, and then
2898   automatically retrying the requests that failed.
2901   An automatic retry &SHOULD-NOT; be repeated if it fails.
2905<section title="Pipelining" anchor="pipelining">
2906   <x:anchor-alias value="pipeline"/>
2908   A client that supports persistent connections &MAY; "<x:dfn>pipeline</x:dfn>"
2909   its requests (i.e., send multiple requests without waiting for each
2910   response). A server &MAY; process a sequence of pipelined requests in
2911   parallel if they all have safe methods (&safe-methods;), but &MUST; send
2912   the corresponding responses in the same order that the requests were
2913   received.
2916   A client that pipelines requests &MUST; be prepared to retry those
2917   requests if the connection closes before it receives all of the
2918   corresponding responses. A client that assumes a persistent connection and
2919   pipelines immediately after connection establishment &MUST-NOT; pipeline
2920   on a retry connection until it knows the connection is persistent.
2923   Idempotent methods (&idempotent-methods;) are significant to pipelining
2924   because they can be automatically retried after a connection failure.
2925   A user agent &SHOULD-NOT; pipeline requests after a non-idempotent method
2926   until the final response status code for that method has been received,
2927   unless the user agent has a means to detect and recover from partial
2928   failure conditions involving the pipelined sequence.
2931   An intermediary that receives pipelined requests &MAY; pipeline those
2932   requests when forwarding them inbound, since it can rely on the outbound
2933   user agent(s) to determine what requests can be safely pipelined. If the
2934   inbound connection fails before receiving a response, the pipelining
2935   intermediary &MAY; attempt to retry a sequence of requests that have yet
2936   to receive a response if the requests all have idempotent methods;
2937   otherwise, the pipelining intermediary &SHOULD; forward any received
2938   responses and then close the corresponding outbound connection(s) so that
2939   the outbound user agent(s) can recover accordingly.
2944<section title="Concurrency" anchor="persistent.concurrency">
2946   Clients &SHOULD; limit the number of simultaneous
2947   connections that they maintain to a given server.
2950   Previous revisions of HTTP gave a specific number of connections as a
2951   ceiling, but this was found to be impractical for many applications. As a
2952   result, this specification does not mandate a particular maximum number of
2953   connections, but instead encourages clients to be conservative when opening
2954   multiple connections.
2957   Multiple connections are typically used to avoid the "head-of-line
2958   blocking" problem, wherein a request that takes significant server-side
2959   processing and/or has a large payload blocks subsequent requests on the
2960   same connection. However, each connection consumes server resources.
2961   Furthermore, using multiple connections can cause undesirable side effects
2962   in congested networks.
2965   Note that servers might reject traffic that they deem abusive, including an
2966   excessive number of connections from a client.
2970<section title="Failures and Time-outs" anchor="persistent.failures">
2972   Servers will usually have some time-out value beyond which they will
2973   no longer maintain an inactive connection. Proxy servers might make
2974   this a higher value since it is likely that the client will be making
2975   more connections through the same server. The use of persistent
2976   connections places no requirements on the length (or existence) of
2977   this time-out for either the client or the server.
2980   When a client or server wishes to time-out it &SHOULD; issue a graceful
2981   close on the transport connection. Clients and servers &SHOULD; both
2982   constantly watch for the other side of the transport close, and
2983   respond to it as appropriate. If a client or server does not detect
2984   the other side's close promptly it could cause unnecessary resource
2985   drain on the network.
2988   A client, server, or proxy &MAY; close the transport connection at any
2989   time. For example, a client might have started to send a new request
2990   at the same time that the server has decided to close the "idle"
2991   connection. From the server's point of view, the connection is being
2992   closed while it was idle, but from the client's point of view, a
2993   request is in progress.
2996   Servers &SHOULD; maintain persistent connections and allow the underlying
2997   transport's flow control mechanisms to resolve temporary overloads, rather
2998   than terminate connections with the expectation that clients will retry.
2999   The latter technique can exacerbate network congestion.
3002   A client sending a message body &SHOULD; monitor
3003   the network connection for an error status code while it is transmitting
3004   the request. If the client sees an error status code, it &SHOULD;
3005   immediately cease transmitting the body and close the connection.
3009<section title="Tear-down" anchor="persistent.tear-down">
3010  <iref primary="false" item="Connection header field" x:for-anchor=""/>
3011  <iref primary="false" item="close" x:for-anchor=""/>
3013   The <x:ref>Connection</x:ref> header field
3014   (<xref target="header.connection"/>) provides a "<x:ref>close</x:ref>"
3015   connection option that a sender &SHOULD; send when it wishes to close
3016   the connection after the current request/response pair.
3019   A client that sends a <x:ref>close</x:ref> connection option &MUST-NOT;
3020   send further requests on that connection (after the one containing
3021   <x:ref>close</x:ref>) and &MUST; close the connection after reading the
3022   final response message corresponding to this request.
3025   A server that receives a <x:ref>close</x:ref> connection option &MUST;
3026   initiate a lingering close (see below) of the connection after it sends the
3027   final response to the request that contained <x:ref>close</x:ref>.
3028   The server &SHOULD; send a <x:ref>close</x:ref> connection option
3029   in its final response on that connection. The server &MUST-NOT; process
3030   any further requests received on that connection.
3033   A server that sends a <x:ref>close</x:ref> connection option &MUST;
3034   initiate a lingering close of the connection after it sends the
3035   response containing <x:ref>close</x:ref>. The server &MUST-NOT; process
3036   any further requests received on that connection.
3039   A client that receives a <x:ref>close</x:ref> connection option &MUST;
3040   cease sending requests on that connection and close the connection
3041   after reading the response message containing the close; if additional
3042   pipelined requests had been sent on the connection, the client &SHOULD;
3043   assume that they will not be processed by the server.
3046   If a server performs an immediate close of a TCP connection, there is a
3047   significant risk that the client will not be able to read the last HTTP
3048   response.  If the server receives additional data from the client on a
3049   fully-closed connection, such as another request that was sent by the
3050   client before receiving the server's response, the server's TCP stack will
3051   send a reset packet to the client; unfortunately, the reset packet might
3052   erase the client's unacknowledged input buffers before they can be read
3053   and interpreted by the client's HTTP parser.
3056   To avoid the TCP reset problem, a server can perform a lingering close on a
3057   connection by closing only the write side of the read/write connection
3058   (a half-close) and continuing to read from the connection until the
3059   connection is closed by the client or the server is reasonably certain
3060   that its own TCP stack has received the client's acknowledgement of the
3061   packet(s) containing the server's last response. It is then safe for the
3062   server to fully close the connection.
3065   It is unknown whether the reset problem is exclusive to TCP or might also
3066   be found in other transport connection protocols.
3070<section title="Upgrade" anchor="header.upgrade">
3071  <iref primary="true" item="Upgrade header field" x:for-anchor=""/>
3072  <x:anchor-alias value="Upgrade"/>
3073  <x:anchor-alias value="protocol"/>
3074  <x:anchor-alias value="protocol-name"/>
3075  <x:anchor-alias value="protocol-version"/>
3077   The "Upgrade" header field is intended to provide a simple mechanism
3078   for transitioning from HTTP/1.1 to some other protocol on the same
3079   connection.  A client &MAY; send a list of protocols in the Upgrade
3080   header field of a request to invite the server to switch to one or
3081   more of those protocols before sending the final response.
3082   A server &MUST; send an Upgrade header field in <x:ref>101 (Switching
3083   Protocols)</x:ref> responses to indicate which protocol(s) are being
3084   switched to, and &MUST; send it in <x:ref>426 (Upgrade Required)</x:ref>
3085   responses to indicate acceptable protocols.
3086   A server &MAY; send an Upgrade header field in any other response to
3087   indicate that they might be willing to upgrade to one of the
3088   specified protocols for a future request.
3090<figure><artwork type="abnf2616"><iref primary="true" item="Grammar" subitem="Upgrade"/>
3091  <x:ref>Upgrade</x:ref>          = 1#<x:ref>protocol</x:ref>
3093  <x:ref>protocol</x:ref>         = <x:ref>protocol-name</x:ref> ["/" <x:ref>protocol-version</x:ref>]
3094  <x:ref>protocol-name</x:ref>    = <x:ref>token</x:ref>
3095  <x:ref>protocol-version</x:ref> = <x:ref>token</x:ref>
3098   For example,
3100<figure><artwork type="example">
3101  Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11
3104   Upgrade eases the difficult transition between incompatible protocols by
3105   allowing the client to initiate a request in the more commonly
3106   supported protocol while indicating to the server that it would like
3107   to use a "better" protocol if available (where "better" is determined
3108   by the server, possibly according to the nature of the request method
3109   or target resource).
3112   Upgrade cannot be used to insist on a protocol change; its acceptance and
3113   use by the server is optional. The capabilities and nature of the
3114   application-level communication after the protocol change is entirely
3115   dependent upon the new protocol chosen, although the first action
3116   after changing the protocol &MUST; be a response to the initial HTTP
3117   request that contained the Upgrade header field.
3120   For example, if the Upgrade header field is received in a GET request
3121   and the server decides to switch protocols, then it first responds
3122   with a <x:ref>101 (Switching Protocols)</x:ref> message in HTTP/1.1 and
3123   then immediately follows that with the new protocol's equivalent of a
3124   response to a GET on the target resource.  This allows a connection to be
3125   upgraded to protocols with the same semantics as HTTP without the
3126   latency cost of an additional round-trip.  A server &MUST-NOT; switch
3127   protocols unless the received message semantics can be honored by the new
3128   protocol; an OPTIONS request can be honored by any protocol.
3131   When Upgrade is sent, a sender &MUST; also send a
3132   <x:ref>Connection</x:ref> header field (<xref target="header.connection"/>)
3133   that contains the "upgrade" connection option, in order to prevent Upgrade
3134   from being accidentally forwarded by intermediaries that might not implement
3135   the listed protocols.  A server &MUST; ignore an Upgrade header field that
3136   is received in an HTTP/1.0 request.
3139   The Upgrade header field only applies to switching application-level
3140   protocols on the existing connection; it cannot be used
3141   to switch to a protocol on a different connection. For that purpose, it is
3142   more appropriate to use a <x:ref>3xx (Redirection)</x:ref> response
3143   (&status-3xx;).
3146   This specification only defines the protocol name "HTTP" for use by
3147   the family of Hypertext Transfer Protocols, as defined by the HTTP
3148   version rules of <xref target="http.version"/> and future updates to this
3149   specification. Additional tokens ought to be registered with IANA using the
3150   registration procedure defined in <xref target="upgrade.token.registry"/>.
3155<section title="IANA Considerations" anchor="IANA.considerations">
3157<section title="Header Field Registration" anchor="header.field.registration">
3159   HTTP header fields are registered within the Message Header Field Registry
3160   maintained at
3161   <eref target=""/>.
3164   This document defines the following HTTP header fields, so their
3165   associated registry entries shall be updated according to the permanent
3166   registrations below (see <xref target="BCP90"/>):
3168<?BEGININC p1-messaging.iana-headers ?>
3169<!--AUTOGENERATED FROM extract-header-defs.xslt, do not edit manually-->
3170<texttable align="left" suppress-title="true" anchor="iana.header.registration.table">
3171   <ttcol>Header Field Name</ttcol>
3172   <ttcol>Protocol</ttcol>
3173   <ttcol>Status</ttcol>
3174   <ttcol>Reference</ttcol>
3176   <c>Connection</c>
3177   <c>http</c>
3178   <c>standard</c>
3179   <c>
3180      <xref target="header.connection"/>
3181   </c>
3182   <c>Content-Length</c>
3183   <c>http</c>
3184   <c>standard</c>
3185   <c>
3186      <xref target="header.content-length"/>
3187   </c>
3188   <c>Host</c>
3189   <c>http</c>
3190   <c>standard</c>
3191   <c>
3192      <xref target=""/>
3193   </c>
3194   <c>TE</c>
3195   <c>http</c>
3196   <c>standard</c>
3197   <c>
3198      <xref target="header.te"/>
3199   </c>
3200   <c>Trailer</c>
3201   <c>http</c>
3202   <c>standard</c>
3203   <c>
3204      <xref target="header.trailer"/>
3205   </c>
3206   <c>Transfer-Encoding</c>
3207   <c>http</c>
3208   <c>standard</c>
3209   <c>
3210      <xref target="header.transfer-encoding"/>
3211   </c>
3212   <c>Upgrade</c>
3213   <c>http</c>
3214   <c>standard</c>
3215   <c>
3216      <xref target="header.upgrade"/>
3217   </c>
3218   <c>Via</c>
3219   <c>http</c>
3220   <c>standard</c>
3221   <c>
3222      <xref target="header.via"/>
3223   </c>
3226<?ENDINC p1-messaging.iana-headers ?>
3228   Furthermore, the header field-name "Close" shall be registered as
3229   "reserved", since using that name as an HTTP header field might
3230   conflict with the "close" connection option of the "<x:ref>Connection</x:ref>"
3231   header field (<xref target="header.connection"/>).
3233<texttable align="left" suppress-title="true">
3234   <ttcol>Header Field Name</ttcol>
3235   <ttcol>Protocol</ttcol>
3236   <ttcol>Status</ttcol>
3237   <ttcol>Reference</ttcol>
3239   <c>Close</c>
3240   <c>http</c>
3241   <c>reserved</c>
3242   <c>
3243      <xref target="header.field.registration"/>
3244   </c>
3247   The change controller is: "IETF ( - Internet Engineering Task Force".
3251<section title="URI Scheme Registration" anchor="uri.scheme.registration">
3253   IANA maintains the registry of URI Schemes <xref target="BCP115"/> at
3254   <eref target=""/>.
3257   This document defines the following URI schemes, so their
3258   associated registry entries shall be updated according to the permanent
3259   registrations below:
3261<texttable align="left" suppress-title="true">
3262   <ttcol>URI Scheme</ttcol>
3263   <ttcol>Description</ttcol>
3264   <ttcol>Reference</ttcol>
3266   <c>http</c>
3267   <c>Hypertext Transfer Protocol</c>
3268   <c><xref target="http.uri"/></c>
3270   <c>https</c>
3271   <c>Hypertext Transfer Protocol Secure</c>
3272   <c><xref target="https.uri"/></c>
3276<section title="Internet Media Type Registration" anchor="">
3278   This document serves as the specification for the Internet media types
3279   "message/http" and "application/http". The following is to be registered with
3280   IANA (see <xref target="BCP13"/>).
3282<section title="Internet Media Type message/http" anchor="">
3283<iref item="Media Type" subitem="message/http" primary="true"/>
3284<iref item="message/http Media Type" primary="true"/>
3286   The message/http type can be used to enclose a single HTTP request or
3287   response message, provided that it obeys the MIME restrictions for all
3288   "message" types regarding line length and encodings.
3291  <list style="hanging" x:indent="12em">
3292    <t hangText="Type name:">
3293      message
3294    </t>
3295    <t hangText="Subtype name:">
3296      http
3297    </t>
3298    <t hangText="Required parameters:">
3299      none
3300    </t>
3301    <t hangText="Optional parameters:">
3302      version, msgtype
3303      <list style="hanging">
3304        <t hangText="version:">
3305          The HTTP-version number of the enclosed message
3306          (e.g., "1.1"). If not present, the version can be
3307          determined from the first line of the body.
3308        </t>
3309        <t hangText="msgtype:">
3310          The message type &mdash; "request" or "response". If not
3311          present, the type can be determined from the first
3312          line of the body.
3313        </t>
3314      </list>
3315    </t>
3316    <t hangText="Encoding considerations:">
3317      only "7bit", "8bit", or "binary" are permitted
3318    </t>
3319    <t hangText="Security considerations:">
3320      none
3321    </t>
3322    <t hangText="Interoperability considerations:">
3323      none
3324    </t>
3325    <t hangText="Published specification:">
3326      This specification (see <xref target=""/>).
3327    </t>
3328    <t hangText="Applications that use this media type:">
3329    </t>
3330    <t hangText="Additional information:">
3331      <list style="hanging">
3332        <t hangText="Magic number(s):">none</t>
3333        <t hangText="File extension(s):">none</t>
3334        <t hangText="Macintosh file type code(s):">none</t>
3335      </list>
3336    </t>
3337    <t hangText="Person and email address to contact for further information:">
3338      See Authors Section.
3339    </t>
3340    <t hangText="Intended usage:">
3341      COMMON
3342    </t>
3343    <t hangText="Restrictions on usage:">
3344      none
3345    </t>
3346    <t hangText="Author:">
3347      See Authors Section.
3348    </t>
3349    <t hangText="Change controller:">
3350      IESG
3351    </t>
3352  </list>
3355<section title="Internet Media Type application/http" anchor="">
3356<iref item="Media Type" subitem="application/http" primary="true"/>
3357<iref item="application/http Media Type" primary="true"/>
3359   The application/http type can be used to enclose a pipeline of one or more
3360   HTTP request or response messages (not intermixed).
3363  <list style="hanging" x:indent="12em">
3364    <t hangText="Type name:">
3365      application
3366    </t>
3367    <t hangText="Subtype name:">
3368      http
3369    </t>
3370    <t hangText="Required parameters:">
3371      none
3372    </t>
3373    <t hangText="Optional parameters:">
3374      version, msgtype
3375      <list style="hanging">
3376        <t hangText="version:">
3377          The HTTP-version number of the enclosed messages
3378          (e.g., "1.1"). If not present, the version can be
3379          determined from the first line of the body.
3380        </t>
3381        <t hangText="msgtype:">
3382          The message type &mdash; "request" or "response". If not
3383          present, the type can be determined from the first
3384          line of the body.
3385        </t>
3386      </list>
3387    </t>
3388    <t hangText="Encoding considerations:">
3389      HTTP messages enclosed by this type
3390      are in "binary" format; use of an appropriate
3391      Content-Transfer-Encoding is required when
3392      transmitted via E-mail.
3393    </t>
3394    <t hangText="Security considerations:">
3395      none
3396    </t>
3397    <t hangText="Interoperability considerations:">
3398      none
3399    </t>
3400    <t hangText="Published specification:">
3401      This specification (see <xref target=""/>).
3402    </t>
3403    <t hangText="Applications that use this media type:">
3404    </t>
3405    <t hangText="Additional information:">
3406      <list style="hanging">
3407        <t hangText="Magic number(s):">none</t>
3408        <t hangText="File extension(s):">none</t>
3409        <t hangText="Macintosh file type code(s):">none</t>
3410      </list>
3411    </t>
3412    <t hangText="Person and email address to contact for further information:">
3413      See Authors Section.
3414    </t>
3415    <t hangText="Intended usage:">
3416      COMMON
3417    </t>
3418    <t hangText="Restrictions on usage:">
3419      none
3420    </t>
3421    <t hangText="Author:">
3422      See Authors Section.
3423    </t>
3424    <t hangText="Change controller:">
3425      IESG
3426    </t>
3427  </list>
3432<section title="Transfer Coding Registry" anchor="transfer.coding.registry">
3434   The HTTP Transfer Coding Registry defines the name space for transfer
3435   coding names. It is maintained at <eref target=""/>.
3438<section title="Procedure" anchor="transfer.coding.registry.procedure">
3440   Registrations &MUST; include the following fields:
3441   <list style="symbols">
3442     <t>Name</t>
3443     <t>Description</t>
3444     <t>Pointer to specification text</t>
3445   </list>
3448   Names of transfer codings &MUST-NOT; overlap with names of content codings
3449   (&content-codings;) unless the encoding transformation is identical, as
3450   is the case for the compression codings defined in
3451   <xref target="compression.codings"/>.
3454   Values to be added to this name space require IETF Review (see
3455   <xref target="RFC5226" x:fmt="of" x:sec="4.1"/>), and &MUST;
3456   conform to the purpose of transfer coding defined in this specification.
3459   Use of program names for the identification of encoding formats
3460   is not desirable and is discouraged for future encodings.
3464<section title="Registration" anchor="transfer.coding.registration">
3466   The HTTP Transfer Coding Registry shall be updated with the registrations
3467   below:
3469<texttable align="left" suppress-title="true" anchor="iana.transfer.coding.registration.table">
3470   <ttcol>Name</ttcol>
3471   <ttcol>Description</ttcol>
3472   <ttcol>Reference</ttcol>
3473   <c>chunked</c>
3474   <c>Transfer in a series of chunks</c>
3475   <c>
3476      <xref target="chunked.encoding"/>
3477   </c>
3478   <c>compress</c>
3479   <c>UNIX "compress" program method</c>
3480   <c>
3481      <xref target="compress.coding"/>
3482   </c>
3483   <c>deflate</c>
3484   <c>"deflate" compression mechanism (<xref target="RFC1951"/>) used inside
3485   the "zlib" data format (<xref target="RFC1950"/>)
3486   </c>
3487   <c>
3488      <xref target="deflate.coding"/>
3489   </c>
3490   <c>gzip</c>
3491   <c>Same as GNU zip <xref target="RFC1952"/></c>
3492   <c>
3493      <xref target="gzip.coding"/>
3494   </c>
3499<section title="Upgrade Token Registry" anchor="upgrade.token.registry">
3501   The HTTP Upgrade Token Registry defines the name space for protocol-name
3502   tokens used to identify protocols in the <x:ref>Upgrade</x:ref> header
3503   field. The registry is maintained at <eref target=""/>.
3506<section title="Procedure" anchor="upgrade.token.registry.procedure">  
3508   Each registered protocol name is associated with contact information
3509   and an optional set of specifications that details how the connection
3510   will be processed after it has been upgraded.
3513   Registrations happen on a "First Come First Served" basis (see
3514   <xref target="RFC5226" x:sec="4.1" x:fmt="of"/>) and are subject to the
3515   following rules:
3516  <list style="numbers">
3517    <t>A protocol-name token, once registered, stays registered forever.</t>
3518    <t>The registration &MUST; name a responsible party for the
3519       registration.</t>
3520    <t>The registration &MUST; name a point of contact.</t>
3521    <t>The registration &MAY; name a set of specifications associated with
3522       that token. Such specifications need not be publicly available.</t>
3523    <t>The registration &SHOULD; name a set of expected "protocol-version"
3524       tokens associated with that token at the time of registration.</t>
3525    <t>The responsible party &MAY; change the registration at any time.
3526       The IANA will keep a record of all such changes, and make them
3527       available upon request.</t>
3528    <t>The IESG &MAY; reassign responsibility for a protocol token.
3529       This will normally only be used in the case when a
3530       responsible party cannot be contacted.</t>
3531  </list>
3534   This registration procedure for HTTP Upgrade Tokens replaces that
3535   previously defined in <xref target="RFC2817" x:fmt="of" x:sec="7.2"/>.
3539<section title="Upgrade Token Registration" anchor="upgrade.token.registration">
3541   The HTTP Upgrade Token Registry shall be updated with the registration
3542   below:
3544<texttable align="left" suppress-title="true">
3545   <ttcol>Value</ttcol>
3546   <ttcol>Description</ttcol>
3547   <ttcol>Expected Version Tokens</ttcol>
3548   <ttcol>Reference</ttcol>
3550   <c>HTTP</c>
3551   <c>Hypertext Transfer Protocol</c>
3552   <c>any DIGIT.DIGIT (e.g, "2.0")</c>
3553   <c><xref target="http.version"/></c>
3556   The responsible party is: "IETF ( - Internet Engineering Task Force".
3563<section title="Security Considerations" anchor="security.considerations">
3565   This section is meant to inform developers, information providers, and
3566   users of known security concerns relevant to HTTP/1.1 message syntax,
3567   parsing, and routing.
3570<section title="DNS-related Attacks" anchor="dns.related.attacks">
3572   HTTP clients rely heavily on the Domain Name Service (DNS), and are thus
3573   generally prone to security attacks based on the deliberate misassociation
3574   of IP addresses and DNS names not protected by DNSSEC. Clients need to be
3575   cautious in assuming the validity of an IP number/DNS name association unless
3576   the response is protected by DNSSEC (<xref target="RFC4033"/>).
3580<section title="Intermediaries and Caching" anchor="attack.intermediaries">
3582   By their very nature, HTTP intermediaries are men-in-the-middle, and
3583   represent an opportunity for man-in-the-middle attacks. Compromise of
3584   the systems on which the intermediaries run can result in serious security
3585   and privacy problems. Intermediaries have access to security-related
3586   information, personal information about individual users and
3587   organizations, and proprietary information belonging to users and
3588   content providers. A compromised intermediary, or an intermediary
3589   implemented or configured without regard to security and privacy
3590   considerations, might be used in the commission of a wide range of
3591   potential attacks.
3594   Intermediaries that contain a shared cache are especially vulnerable
3595   to cache poisoning attacks.
3598   Implementers need to consider the privacy and security
3599   implications of their design and coding decisions, and of the
3600   configuration options they provide to operators (especially the
3601   default configuration).
3604   Users need to be aware that intermediaries are no more trustworthy than
3605   the people who run them; HTTP itself cannot solve this problem.
3609<section title="Buffer Overflows" anchor="attack.protocol.element.size.overflows">
3611   Because HTTP uses mostly textual, character-delimited fields, attackers can
3612   overflow buffers in implementations, and/or perform a Denial of Service
3613   against implementations that accept fields with unlimited lengths.
3616   To promote interoperability, this specification makes specific
3617   recommendations for minimum size limits on request-line
3618   (<xref target="request.line"/>)
3619   and blocks of header fields (<xref target="header.fields"/>). These are
3620   minimum recommendations, chosen to be supportable even by implementations
3621   with limited resources; it is expected that most implementations will
3622   choose substantially higher limits.
3625   This specification also provides a way for servers to reject messages that
3626   have request-targets that are too long (&status-414;) or request entities
3627   that are too large (&status-4xx;).
3630   Recipients &SHOULD; carefully limit the extent to which they read other
3631   fields, including (but not limited to) request methods, response status
3632   phrases, header field-names, and body chunks, so as to avoid denial of
3633   service attacks without impeding interoperability.
3637<section title="Message Integrity" anchor="message.integrity">
3639   HTTP does not define a specific mechanism for ensuring message integrity,
3640   instead relying on the error-detection ability of underlying transport
3641   protocols and the use of length or chunk-delimited framing to detect
3642   completeness. Additional integrity mechanisms, such as hash functions or
3643   digital signatures applied to the content, can be selectively added to
3644   messages via extensible metadata header fields. Historically, the lack of
3645   a single integrity mechanism has been justified by the informal nature of
3646   most HTTP communication.  However, the prevalence of HTTP as an information
3647   access mechanism has resulted in its increasing use within environments
3648   where verification of message integrity is crucial.
3651   User agents are encouraged to implement configurable means for detecting
3652   and reporting failures of message integrity such that those means can be
3653   enabled within environments for which integrity is necessary. For example,
3654   a browser being used to view medical history or drug interaction
3655   information needs to indicate to the user when such information is detected
3656   by the protocol to be incomplete, expired, or corrupted during transfer.
3657   Such mechanisms might be selectively enabled via user agent extensions or
3658   the presence of message integrity metadata in a response.
3659   At a minimum, user agents ought to provide some indication that allows a
3660   user to distinguish between a complete and incomplete response message
3661   (<xref target="incomplete.messages"/>) when such verification is desired.
3665<section title="Server Log Information" anchor="abuse.of.server.log.information">
3667   A server is in the position to save personal data about a user's requests
3668   over time, which might identify their reading patterns or subjects of
3669   interest.  In particular, log information gathered at an intermediary
3670   often contains a history of user agent interaction, across a multitude
3671   of sites, that can be traced to individual users.
3674   HTTP log information is confidential in nature; its handling is often
3675   constrained by laws and regulations.  Log information needs to be securely
3676   stored and appropriate guidelines followed for its analysis.
3677   Anonymization of personal information within individual entries helps,
3678   but is generally not sufficient to prevent real log traces from being
3679   re-identified based on correlation with other access characteristics.
3680   As such, access traces that are keyed to a specific client should not
3681   be published even if the key is pseudonymous.
3684   To minimize the risk of theft or accidental publication, log information
3685   should be purged of personally identifiable information, including
3686   user identifiers, IP addresses, and user-provided query parameters,
3687   as soon as that information is no longer necessary to support operational
3688   needs for security, auditing, or fraud control.
3693<section title="Acknowledgments" anchor="acks">
3695   This edition of HTTP/1.1 builds on the many contributions that went into
3696   <xref target="RFC1945" format="none">RFC 1945</xref>,
3697   <xref target="RFC2068" format="none">RFC 2068</xref>,
3698   <xref target="RFC2145" format="none">RFC 2145</xref>, and
3699   <xref target="RFC2616" format="none">RFC 2616</xref>, including
3700   substantial contributions made by the previous authors, editors, and
3701   working group chairs: Tim Berners-Lee, Ari Luotonen, Roy T. Fielding,
3702   Henrik Frystyk Nielsen, Jim Gettys, Jeffrey C. Mogul, Larry Masinter,
3703   and Paul J. Leach. Mark Nottingham oversaw this effort as working group chair.
3706   Since 1999, the following contributors have helped improve the HTTP
3707   specification by reporting bugs, asking smart questions, drafting or
3708   reviewing text, and evaluating open issues:
3710<?BEGININC acks ?>
3711<t>Adam Barth,
3712Adam Roach,
3713Addison Phillips,
3714Adrian Chadd,
3715Adrien W. de Croy,
3716Alan Ford,
3717Alan Ruttenberg,
3718Albert Lunde,
3719Alek Storm,
3720Alex Rousskov,
3721Alexandre Morgaut,
3722Alexey Melnikov,
3723Alisha Smith,
3724Amichai Rothman,
3725Amit Klein,
3726Amos Jeffries,
3727Andreas Maier,
3728Andreas Petersson,
3729Anil Sharma,
3730Anne van Kesteren,
3731Anthony Bryan,
3732Asbjorn Ulsberg,
3733Ashok Kumar,
3734Balachander Krishnamurthy,
3735Barry Leiba,
3736Ben Laurie,
3737Benjamin Carlyle,
3738Benjamin Niven-Jenkins,
3739Bil Corry,
3740Bill Burke,
3741Bjoern Hoehrmann,
3742Bob Scheifler,
3743Boris Zbarsky,
3744Brett Slatkin,
3745Brian Kell,
3746Brian McBarron,
3747Brian Pane,
3748Brian Raymor,
3749Brian Smith,
3750Bryce Nesbitt,
3751Cameron Heavon-Jones,
3752Carl Kugler,
3753Carsten Bormann,
3754Charles Fry,
3755Chris Newman,
3756Cyrus Daboo,
3757Dale Robert Anderson,
3758Dan Wing,
3759Dan Winship,
3760Daniel Stenberg,
3761Darrel Miller,
3762Dave Cridland,
3763Dave Crocker,
3764Dave Kristol,
3765Dave Thaler,
3766David Booth,
3767David Singer,
3768David W. Morris,
3769Diwakar Shetty,
3770Dmitry Kurochkin,
3771Drummond Reed,
3772Duane Wessels,
3773Edward Lee,
3774Eitan Adler,
3775Eliot Lear,
3776Eran Hammer-Lahav,
3777Eric D. Williams,
3778Eric J. Bowman,
3779Eric Lawrence,
3780Eric Rescorla,
3781Erik Aronesty,
3782Evan Prodromou,
3783Felix Geisendoerfer,
3784Florian Weimer,
3785Frank Ellermann,
3786Fred Bohle,
3787Frederic Kayser,
3788Gabriel Montenegro,
3789Geoffrey Sneddon,
3790Gervase Markham,
3791Grahame Grieve,
3792Greg Wilkins,
3793Grzegorz Calkowski,
3794Harald Tveit Alvestrand,
3795Harry Halpin,
3796Helge Hess,
3797Henrik Nordstrom,
3798Henry S. Thompson,
3799Henry Story,
3800Herbert van de Sompel,
3801Herve Ruellan,
3802Howard Melman,
3803Hugo Haas,
3804Ian Fette,
3805Ian Hickson,
3806Ido Safruti,
3807Ilari Liusvaara,
3808Ilya Grigorik,
3809Ingo Struck,
3810J. Ross Nicoll,
3811James Cloos,
3812James H. Manger,
3813James Lacey,
3814James M. Snell,
3815Jamie Lokier,
3816Jan Algermissen,
3817Jeff Hodges (who came up with the term 'effective Request-URI'),
3818Jeff Pinner,
3819Jeff Walden,
3820Jim Luther,
3821Jitu Padhye,
3822Joe D. Williams,
3823Joe Gregorio,
3824Joe Orton,
3825John C. Klensin,
3826John C. Mallery,
3827John Cowan,
3828John Kemp,
3829John Panzer,
3830John Schneider,
3831John Stracke,
3832John Sullivan,
3833Jonas Sicking,
3834Jonathan A. Rees,
3835Jonathan Billington,
3836Jonathan Moore,
3837Jonathan Silvera,
3838Jordi Ros,
3839Joris Dobbelsteen,
3840Josh Cohen,
3841Julien Pierre,
3842Jungshik Shin,
3843Justin Chapweske,
3844Justin Erenkrantz,
3845Justin James,
3846Kalvinder Singh,
3847Karl Dubost,
3848Keith Hoffman,
3849Keith Moore,
3850Ken Murchison,
3851Koen Holtman,
3852Konstantin Voronkov,
3853Kris Zyp,
3854Lisa Dusseault,
3855Maciej Stachowiak,
3856Manu Sporny,
3857Marc Schneider,
3858Marc Slemko,
3859Mark Baker,
3860Mark Pauley,
3861Mark Watson,
3862Markus Isomaki,
3863Markus Lanthaler,
3864Martin J. Duerst,
3865Martin Musatov,
3866Martin Nilsson,
3867Martin Thomson,
3868Matt Lynch,
3869Matthew Cox,
3870Max Clark,
3871Michael Burrows,
3872Michael Hausenblas,
3873Mike Amundsen,
3874Mike Belshe,
3875Mike Kelly,
3876Mike Schinkel,
3877Miles Sabin,
3878Murray S. Kucherawy,
3879Mykyta Yevstifeyev,
3880Nathan Rixham,
3881Nicholas Shanks,
3882Nico Williams,
3883Nicolas Alvarez,
3884Nicolas Mailhot,
3885Noah Slater,
3886Osama Mazahir,
3887Pablo Castro,
3888Pat Hayes,
3889Patrick R. McManus,
3890Paul E. Jones,
3891Paul Hoffman,
3892Paul Marquess,
3893Peter Lepeska,
3894Peter Saint-Andre,
3895Peter Watkins,
3896Phil Archer,
3897Philippe Mougin,
3898Phillip Hallam-Baker,
3899Piotr Dobrogost,
3900Poul-Henning Kamp,
3901Preethi Natarajan,
3902Rajeev Bector,
3903Ray Polk,
3904Reto Bachmann-Gmuer,
3905Richard Cyganiak,
3906Robby Simpson,
3907Robert Brewer,
3908Robert Collins,
3909Robert Mattson,
3910Robert O'Callahan,
3911Robert Olofsson,
3912Robert Sayre,
3913Robert Siemer,
3914Robert de Wilde,
3915Roberto Javier Godoy,
3916Roberto Peon,
3917Roland Zink,
3918Ronny Widjaja,
3919S. Mike Dierken,
3920Salvatore Loreto,
3921Sam Johnston,
3922Sam Ruby,
3923Scott Lawrence (who maintained the original issues list),
3924Sean B. Palmer,
3925Shane McCarron,
3926Stefan Eissing,
3927Stefan Tilkov,
3928Stefanos Harhalakis,
3929Stephane Bortzmeyer,
3930Stephen Farrell,
3931Stephen Ludin,
3932Stuart Williams,
3933Subbu Allamaraju,
3934Sylvain Hellegouarch,
3935Tapan Divekar,
3936Tatsuya Hayashi,
3937Ted Hardie,
3938Thomas Broyer,
3939Thomas Fossati,
3940Thomas Maslen,
3941Thomas Nordin,
3942Thomas Roessler,
3943Tim Bray,
3944Tim Morgan,
3945Tim Olsen,
3946Tom Zhou,
3947Travis Snoozy,
3948Tyler Close,
3949Vincent Murphy,
3950Wenbo Zhu,
3951Werner Baumann,
3952Wilbur Streett,
3953Wilfredo Sanchez Vega,
3954William A. Rowe Jr.,
3955William Chan,
3956Willy Tarreau,
3957Xiaoshu Wang,
3958Yaron Goland,
3959Yngve Nysaeter Pettersen,
3960Yoav Nir,
3961Yogesh Bang,
3962Yutaka Oiwa,
3963Yves Lafon (long-time member of the editor team),
3964Zed A. Shaw, and
3965Zhong Yu.
3967<?ENDINC acks ?>
3969   See <xref target="RFC2616" x:fmt="of" x:sec="16"/> for additional
3970   acknowledgements from prior revisions.
3977<references title="Normative References">
3979<reference anchor="Part2">
3980  <front>
3981    <title>Hypertext Transfer Protocol (HTTP/1.1): Semantics and Content</title>
3982    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
3983      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
3984      <address><email></email></address>
3985    </author>
3986    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
3987      <organization abbrev="greenbytes">greenbytes GmbH</organization>
3988      <address><email></email></address>
3989    </author>
3990    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
3991  </front>
3992  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p2-semantics-&ID-VERSION;"/>
3993  <x:source href="p2-semantics.xml" basename="p2-semantics">
3994    <x:defines>1xx (Informational)</x:defines>
3995    <x:defines>1xx</x:defines>
3996    <x:defines>100 (Continue)</x:defines>
3997    <x:defines>101 (Switching Protocols)</x:defines>
3998    <x:defines>2xx (Successful)</x:defines>
3999    <x:defines>2xx</x:defines>
4000    <x:defines>200 (OK)</x:defines>
4001    <x:defines>204 (No Content)</x:defines>
4002    <x:defines>3xx (Redirection)</x:defines>
4003    <x:defines>3xx</x:defines>
4004    <x:defines>301 (Moved Permanently)</x:defines>
4005    <x:defines>4xx (Client Error)</x:defines>
4006    <x:defines>4xx</x:defines>
4007    <x:defines>400 (Bad Request)</x:defines>
4008    <x:defines>411 (Length Required)</x:defines>
4009    <x:defines>414 (URI Too Long)</x:defines>
4010    <x:defines>417 (Expectation Failed)</x:defines>
4011    <x:defines>426 (Upgrade Required)</x:defines>
4012    <x:defines>501 (Not Implemented)</x:defines>
4013    <x:defines>502 (Bad Gateway)</x:defines>
4014    <x:defines>505 (HTTP Version Not Supported)</x:defines>
4015    <x:defines>Allow</x:defines>
4016    <x:defines>Content-Encoding</x:defines>
4017    <x:defines>Content-Location</x:defines>
4018    <x:defines>Content-Type</x:defines>
4019    <x:defines>Date</x:defines>
4020    <x:defines>Expect</x:defines>
4021    <x:defines>Location</x:defines>
4022    <x:defines>Server</x:defines>
4023    <x:defines>User-Agent</x:defines>
4024  </x:source>
4027<reference anchor="Part4">
4028  <front>
4029    <title>Hypertext Transfer Protocol (HTTP/1.1): Conditional Requests</title>
4030    <author fullname="Roy T. Fielding" initials="R." role="editor" surname="Fielding">
4031      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4032      <address><email></email></address>
4033    </author>
4034    <author fullname="Julian F. Reschke" initials="J. F." role="editor" surname="Reschke">
4035      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4036      <address><email></email></address>
4037    </author>
4038    <date month="&ID-MONTH;" year="&ID-YEAR;" />
4039  </front>
4040  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p4-conditional-&ID-VERSION;" />
4041  <x:source basename="p4-conditional" href="p4-conditional.xml">
4042    <x:defines>304 (Not Modified)</x:defines>
4043    <x:defines>ETag</x:defines>
4044    <x:defines>Last-Modified</x:defines>
4045  </x:source>
4048<reference anchor="Part5">
4049  <front>
4050    <title>Hypertext Transfer Protocol (HTTP/1.1): Range Requests</title>
4051    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4052      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4053      <address><email></email></address>
4054    </author>
4055    <author initials="Y." surname="Lafon" fullname="Yves Lafon" role="editor">
4056      <organization abbrev="W3C">World Wide Web Consortium</organization>
4057      <address><email></email></address>
4058    </author>
4059    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4060      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4061      <address><email></email></address>
4062    </author>
4063    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4064  </front>
4065  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p5-range-&ID-VERSION;"/>
4066  <x:source href="p5-range.xml" basename="p5-range">
4067    <x:defines>Content-Range</x:defines>
4068  </x:source>
4071<reference anchor="Part6">
4072  <front>
4073    <title>Hypertext Transfer Protocol (HTTP/1.1): Caching</title>
4074    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4075      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4076      <address><email></email></address>
4077    </author>
4078    <author initials="M." surname="Nottingham" fullname="Mark Nottingham" role="editor">
4079      <organization>Akamai</organization>
4080      <address><email></email></address>
4081    </author>
4082    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4083      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4084      <address><email></email></address>
4085    </author>
4086    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4087  </front>
4088  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p6-cache-&ID-VERSION;"/>
4089  <x:source href="p6-cache.xml" basename="p6-cache">
4090    <x:defines>Cache-Control</x:defines>
4091    <x:defines>Expires</x:defines>
4092  </x:source>
4095<reference anchor="Part7">
4096  <front>
4097    <title>Hypertext Transfer Protocol (HTTP/1.1): Authentication</title>
4098    <author initials="R." surname="Fielding" fullname="Roy T. Fielding" role="editor">
4099      <organization abbrev="Adobe">Adobe Systems Incorporated</organization>
4100      <address><email></email></address>
4101    </author>
4102    <author initials="J. F." surname="Reschke" fullname="Julian F. Reschke" role="editor">
4103      <organization abbrev="greenbytes">greenbytes GmbH</organization>
4104      <address><email></email></address>
4105    </author>
4106    <date month="&ID-MONTH;" year="&ID-YEAR;"/>
4107  </front>
4108  <seriesInfo name="Internet-Draft" value="draft-ietf-httpbis-p7-auth-&ID-VERSION;"/>
4109  <x:source href="p7-auth.xml" basename="p7-auth">
4110    <x:defines>Proxy-Authenticate</x:defines>
4111    <x:defines>Proxy-Authorization</x:defines>
4112  </x:source>
4115<reference anchor="RFC5234">
4116  <front>
4117    <title abbrev="ABNF for Syntax Specifications">Augmented BNF for Syntax Specifications: ABNF</title>
4118    <author initials="D." surname="Crocker" fullname="Dave Crocker" role="editor">
4119      <organization>Brandenburg InternetWorking</organization>
4120      <address>
4121        <email></email>
4122      </address> 
4123    </author>
4124    <author initials="P." surname="Overell" fullname="Paul Overell">
4125      <organization>THUS plc.</organization>
4126      <address>
4127        <email></email>
4128      </address>
4129    </author>
4130    <date month="January" year="2008"/>
4131  </front>
4132  <seriesInfo name="STD" value="68"/>
4133  <seriesInfo name="RFC" value="5234"/>
4136<reference anchor="RFC2119">
4137  <front>
4138    <title>Key words for use in RFCs to Indicate Requirement Levels</title>
4139    <author initials="S." surname="Bradner" fullname="Scott Bradner">
4140      <organization>Harvard University</organization>
4141      <address><email></email></address>
4142    </author>
4143    <date month="March" year="1997"/>
4144  </front>
4145  <seriesInfo name="BCP" value="14"/>
4146  <seriesInfo name="RFC" value="2119"/>
4149<reference anchor="RFC3986">
4150 <front>
4151  <title abbrev='URI Generic Syntax'>Uniform Resource Identifier (URI): Generic Syntax</title>
4152  <author initials='T.' surname='Berners-Lee' fullname='Tim Berners-Lee'>
4153    <organization abbrev="W3C/MIT">World Wide Web Consortium</organization>
4154    <address>
4155       <email></email>
4156       <uri></uri>
4157    </address>
4158  </author>
4159  <author initials='R.' surname='Fielding' fullname='Roy T. Fielding'>
4160    <organization abbrev="Day Software">Day Software</organization>
4161    <address>
4162      <email></email>
4163      <uri></uri>
4164    </address>
4165  </author>
4166  <author initials='L.' surname='Masinter' fullname='Larry Masinter'>
4167    <organization abbrev="Adobe Systems">Adobe Systems Incorporated</organization>
4168    <address>
4169      <email></email>
4170      <uri></uri>
4171    </address>
4172  </author>
4173  <date month='January' year='2005'></date>
4174 </front>
4175 <seriesInfo name="STD" value="66"/>
4176 <seriesInfo name="RFC" value="3986"/>
4179<reference anchor="RFC0793">
4180  <front>
4181    <title>Transmission Control Protocol</title>
4182    <author initials='J.' surname='Postel' fullname='Jon Postel'>
4183      <organization>University of Southern California (USC)/Information Sciences Institute</organization>
4184    </author>
4185    <date year='1981' month='September' />
4186  </front>
4187  <seriesInfo name='STD' value='7' />
4188  <seriesInfo name='RFC' value='793' />
4191<reference anchor="USASCII">
4192  <front>
4193    <title>Coded Character Set -- 7-bit American Standard Code for Information Interchange</title>
4194    <author>
4195      <organization>American National Standards Institute</organization>
4196    </author>
4197    <date year="1986"/>
4198  </front>
4199  <seriesInfo name="ANSI" value="X3.4"/>
4202<reference anchor="RFC1950">
4203  <front>
4204    <title>ZLIB Compressed Data Format Specification version 3.3</title>
4205    <author initials="L.P." surname="Deutsch" fullname="L. Peter Deutsch">
4206      <organization>Aladdin Enterprises</organization>
4207      <address><email></email></address>
4208    </author>
4209    <author initials="J-L." surname="Gailly" fullname="Jean-Loup Gailly"/>
4210    <date month="May" year="1996"/>
4211  </front>
4212  <seriesInfo name="RFC" value="1950"/>
4213  <!--<annotation>
4214    RFC 1950 is an Informational RFC, thus it might be less stable than
4215    this specification. On the other hand, this downward reference was
4216    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4217    therefore it is unlikely to cause problems in practice. See also
4218    <xref target="BCP97"/>.
4219  </annotation>-->
4222<reference anchor="RFC1951">
4223  <front>
4224    <title>DEFLATE Compressed Data Format Specification version 1.3</title>
4225    <author initials="P." surname="Deutsch" fullname="L. Peter Deutsch">
4226      <organization>Aladdin Enterprises</organization>
4227      <address><email></email></address>
4228    </author>
4229    <date month="May" year="1996"/>
4230  </front>
4231  <seriesInfo name="RFC" value="1951"/>
4232  <!--<annotation>
4233    RFC 1951 is an Informational RFC, thus it might be less stable than
4234    this specification. On the other hand, this downward reference was
4235    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4236    therefore it is unlikely to cause problems in practice. See also
4237    <xref target="BCP97"/>.
4238  </annotation>-->
4241<reference anchor="RFC1952">
4242  <front>
4243    <title>GZIP file format specification version 4.3</title>
4244    <author initials="P." surname="Deutsch" fullname="L. Peter Deutsch">
4245      <organization>Aladdin Enterprises</organization>
4246      <address><email></email></address>
4247    </author>
4248    <author initials="J-L." surname="Gailly" fullname="Jean-Loup Gailly">
4249      <address><email></email></address>
4250    </author>
4251    <author initials="M." surname="Adler" fullname="Mark Adler">
4252      <address><email></email></address>
4253    </author>
4254    <author initials="L.P." surname="Deutsch" fullname="L. Peter Deutsch">
4255      <address><email></email></address>
4256    </author>
4257    <author initials="G." surname="Randers-Pehrson" fullname="Glenn Randers-Pehrson">
4258      <address><email></email></address>
4259    </author>
4260    <date month="May" year="1996"/>
4261  </front>
4262  <seriesInfo name="RFC" value="1952"/>
4263  <!--<annotation>
4264    RFC 1952 is an Informational RFC, thus it might be less stable than
4265    this specification. On the other hand, this downward reference was
4266    present since the publication of <xref target="RFC2068" x:fmt="none">RFC 2068</xref> in 1997,
4267    therefore it is unlikely to cause problems in practice. See also
4268    <xref target="BCP97"/>.
4269  </annotation>-->
4274<references title="Informative References">
4276<reference anchor="ISO-8859-1">
4277  <front>
4278    <title>
4279     Information technology -- 8-bit single-byte coded graphic character sets -- Part 1: Latin alphabet No. 1
4280    </title>
4281    <author>
4282      <organization>International Organization for Standardization</organization>
4283    </author>
4284    <date year="1998"/>
4285  </front>
4286  <seriesInfo name="ISO/IEC" value="8859-1:1998"/>
4289<reference anchor='RFC1919'>
4290  <front>
4291    <title>Classical versus Transparent IP Proxies</title>
4292    <author initials='M.' surname='Chatel' fullname='Marc Chatel'>
4293      <address><email></email></address>
4294    </author>
4295    <date year='1996' month='March' />
4296  </front>
4297  <seriesInfo name='RFC' value='1919' />
4300<reference anchor="RFC1945">
4301  <front>
4302    <title abbrev="HTTP/1.0">Hypertext Transfer Protocol -- HTTP/1.0</title>
4303    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
4304      <organization>MIT, Laboratory for Computer Science</organization>
4305      <address><email></email></address>
4306    </author>
4307    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
4308      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
4309      <address><email></email></address>
4310    </author>
4311    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4312      <organization>W3 Consortium, MIT Laboratory for Computer Science</organization>
4313      <address><email></email></address>
4314    </author>
4315    <date month="May" year="1996"/>
4316  </front>
4317  <seriesInfo name="RFC" value="1945"/>
4320<reference anchor="RFC2045">
4321  <front>
4322    <title abbrev="Internet Message Bodies">Multipurpose Internet Mail Extensions (MIME) Part One: Format of Internet Message Bodies</title>
4323    <author initials="N." surname="Freed" fullname="Ned Freed">
4324      <organization>Innosoft International, Inc.</organization>
4325      <address><email></email></address>
4326    </author>
4327    <author initials="N.S." surname="Borenstein" fullname="Nathaniel S. Borenstein">
4328      <organization>First Virtual Holdings</organization>
4329      <address><email></email></address>
4330    </author>
4331    <date month="November" year="1996"/>
4332  </front>
4333  <seriesInfo name="RFC" value="2045"/>
4336<reference anchor="RFC2047">
4337  <front>
4338    <title abbrev="Message Header Extensions">MIME (Multipurpose Internet Mail Extensions) Part Three: Message Header Extensions for Non-ASCII Text</title>
4339    <author initials="K." surname="Moore" fullname="Keith Moore">
4340      <organization>University of Tennessee</organization>
4341      <address><email></email></address>
4342    </author>
4343    <date month="November" year="1996"/>
4344  </front>
4345  <seriesInfo name="RFC" value="2047"/>
4348<reference anchor="RFC2068">
4349  <front>
4350    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
4351    <author initials="R." surname="Fielding" fullname="Roy T. Fielding">
4352      <organization>University of California, Irvine, Department of Information and Computer Science</organization>
4353      <address><email></email></address>
4354    </author>
4355    <author initials="J." surname="Gettys" fullname="Jim Gettys">
4356      <organization>MIT Laboratory for Computer Science</organization>
4357      <address><email></email></address>
4358    </author>
4359    <author initials="J." surname="Mogul" fullname="Jeffrey C. Mogul">
4360      <organization>Digital Equipment Corporation, Western Research Laboratory</organization>
4361      <address><email></email></address>
4362    </author>
4363    <author initials="H." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4364      <organization>MIT Laboratory for Computer Science</organization>
4365      <address><email></email></address>
4366    </author>
4367    <author initials="T." surname="Berners-Lee" fullname="Tim Berners-Lee">
4368      <organization>MIT Laboratory for Computer Science</organization>
4369      <address><email></email></address>
4370    </author>
4371    <date month="January" year="1997"/>
4372  </front>
4373  <seriesInfo name="RFC" value="2068"/>
4376<reference anchor="RFC2145">
4377  <front>
4378    <title abbrev="HTTP Version Numbers">Use and Interpretation of HTTP Version Numbers</title>
4379    <author initials="J.C." surname="Mogul" fullname="Jeffrey C. Mogul">
4380      <organization>Western Research Laboratory</organization>
4381      <address><email></email></address>
4382    </author>
4383    <author initials="R.T." surname="Fielding" fullname="Roy T. Fielding">
4384      <organization>Department of Information and Computer Science</organization>
4385      <address><email></email></address>
4386    </author>
4387    <author initials="J." surname="Gettys" fullname="Jim Gettys">
4388      <organization>MIT Laboratory for Computer Science</organization>
4389      <address><email></email></address>
4390    </author>
4391    <author initials="H.F." surname="Nielsen" fullname="Henrik Frystyk Nielsen">
4392      <organization>W3 Consortium</organization>
4393      <address><email></email></address>
4394    </author>
4395    <date month="May" year="1997"/>
4396  </front>
4397  <seriesInfo name="RFC" value="2145"/>
4400<reference anchor="RFC2616">
4401  <front>
4402    <title>Hypertext Transfer Protocol -- HTTP/1.1</title>
4403    <author initials="R." surname="Fielding" fullname="R. Fielding">
4404      <organization>University of California, Irvine</organization>
4405      <address><email></email></address>
4406    </author>
4407    <author initials="J." surname="Gettys" fullname="J. Gettys">
4408      <organization>W3C</organization>
4409      <address><email></email></address>
4410    </author>
4411    <author initials="J." surname="Mogul" fullname="J. Mogul">
4412      <organization>Compaq Computer Corporation</organization>
4413      <address><email></email></address>
4414    </author>
4415    <author initials="H." surname="Frystyk" fullname="H. Frystyk">
4416      <organization>MIT Laboratory for Computer Science</organization>
4417      <address><email></email></address>
4418    </author>
4419    <author initials="L." surname="Masinter" fullname="L. Masinter">
4420      <organization>Xerox Corporation</organization>
4421      <address><email></email></address>
4422    </author>
4423    <author initials="P." surname="Leach" fullname="P. Leach">
4424      <organization>Microsoft Corporation</organization>
4425      <address><email></email></address>
4426    </author>
4427    <author initials="T." surname="Berners-Lee" fullname="T. Berners-Lee">
4428      <organization>W3C</organization>
4429      <address><email></email></address>
4430    </author>
4431    <date month="June" year="1999"/>
4432  </front>
4433  <seriesInfo name="RFC" value="2616"/>
4436<reference anchor='RFC2817'>
4437  <front>
4438    <title>Upgrading to TLS Within HTTP/1.1</title>
4439    <author initials='R.' surname='Khare' fullname='R. Khare'>
4440      <organization>4K Associates / UC Irvine</organization>
4441      <address><email></email></address>
4442    </author>
4443    <author initials='S.' surname='Lawrence' fullname='S. Lawrence'>
4444      <organization>Agranat Systems, Inc.</organization>
4445      <address><email></email></address>
4446    </author>
4447    <date year='2000' month='May' />
4448  </front>
4449  <seriesInfo name='RFC' value='2817' />
4452<reference anchor='RFC2818'>
4453  <front>
4454    <title>HTTP Over TLS</title>
4455    <author initials='E.' surname='Rescorla' fullname='Eric Rescorla'>
4456      <organization>RTFM, Inc.</organization>
4457      <address><email></email></address>
4458    </author>
4459    <date year='2000' month='May' />
4460  </front>
4461  <seriesInfo name='RFC' value='2818' />
4464<reference anchor='RFC3040'>
4465  <front>
4466    <title>Internet Web Replication and Caching Taxonomy</title>
4467    <author initials='I.' surname='Cooper' fullname='I. Cooper'>
4468      <organization>Equinix, Inc.</organization>
4469    </author>
4470    <author initials='I.' surname='Melve' fullname='I. Melve'>
4471      <organization>UNINETT</organization>
4472    </author>
4473    <author initials='G.' surname='Tomlinson' fullname='G. Tomlinson'>
4474      <organization>CacheFlow Inc.</organization>
4475    </author>
4476    <date year='2001' month='January' />
4477  </front>
4478  <seriesInfo name='RFC' value='3040' />
4481<reference anchor='BCP90'>
4482  <front>
4483    <title>Registration Procedures for Message Header Fields</title>
4484    <author initials='G.' surname='Klyne' fullname='G. Klyne'>
4485      <organization>Nine by Nine</organization>
4486      <address><email></email></address>
4487    </author>
4488    <author initials='M.' surname='Nottingham' fullname='M. Nottingham'>
4489      <organization>BEA Systems</organization>
4490      <address><email></email></address>
4491    </author>
4492    <author initials='J.' surname='Mogul' fullname='J. Mogul'>
4493      <organization>HP Labs</organization>
4494      <address><email></email></address>
4495    </author>
4496    <date year='2004' month='September' />
4497  </front>
4498  <seriesInfo name='BCP' value='90' />
4499  <seriesInfo name='RFC' value='3864' />
4502<reference anchor='RFC4033'>
4503  <front>
4504    <title>DNS Security Introduction and Requirements</title>
4505    <author initials='R.' surname='Arends' fullname='R. Arends'/>
4506    <author initials='R.' surname='Austein' fullname='R. Austein'/>
4507    <author initials='M.' surname='Larson' fullname='M. Larson'/>
4508    <author initials='D.' surname='Massey' fullname='D. Massey'/>
4509    <author initials='S.' surname='Rose' fullname='S. Rose'/>
4510    <date year='2005' month='March' />
4511  </front>
4512  <seriesInfo name='RFC' value='4033' />
4515<reference anchor="BCP13">
4516  <front>
4517    <title>Media Type Specifications and Registration Procedures</title>
4518    <author initials="N." surname="Freed" fullname="Ned Freed">
4519      <organization>Oracle</organization>
4520      <address>
4521        <email></email>
4522      </address>
4523    </author>
4524    <author initials="J." surname="Klensin" fullname="John C. Klensin">
4525      <address>
4526        <email></email>
4527      </address>
4528    </author>
4529    <author initials="T." surname="Hansen" fullname="Tony Hansen">
4530      <organization>AT&amp;T Laboratories</organization>
4531      <address>
4532        <email></email>
4533      </address>
4534    </author>
4535    <date year="2013" month="January"/>
4536  </front>
4537  <seriesInfo name="BCP" value="13"/>
4538  <seriesInfo name="RFC" value="6838"/>
4541<reference anchor='BCP115'>
4542  <front>
4543    <title>Guidelines and Registration Procedures for New URI Schemes</title>
4544    <author initials='T.' surname='Hansen' fullname='T. Hansen'>
4545      <organization>AT&amp;T Laboratories</organization>
4546      <address>
4547        <email></email>
4548      </address>
4549    </author>
4550    <author initials='T.' surname='Hardie' fullname='T. Hardie'>
4551      <organization>Qualcomm, Inc.</organization>
4552      <address>
4553        <email></email>
4554      </address>
4555    </author>
4556    <author initials='L.' surname='Masinter' fullname='L. Masinter'>
4557      <organization>Adobe Systems</organization>
4558      <address>
4559        <email></email>
4560      </address>
4561    </author>
4562    <date year='2006' month='February' />
4563  </front>
4564  <seriesInfo name='BCP' value='115' />
4565  <seriesInfo name='RFC' value='4395' />
4568<reference anchor='RFC4559'>
4569  <front>
4570    <title>SPNEGO-based Kerberos and NTLM HTTP Authentication in Microsoft Windows</title>
4571    <author initials='K.' surname='Jaganathan' fullname='K. Jaganathan'/>
4572    <author initials='L.' surname='Zhu' fullname='L. Zhu'/>
4573    <author initials='J.' surname='Brezak' fullname='J. Brezak'/>
4574    <date year='2006' month='June' />
4575  </front>
4576  <seriesInfo name='RFC' value='4559' />
4579<reference anchor='RFC5226'>
4580  <front>
4581    <title>Guidelines for Writing an IANA Considerations Section in RFCs</title>
4582    <author initials='T.' surname='Narten' fullname='T. Narten'>
4583      <organization>IBM</organization>
4584      <address><email></email></address>
4585    </author>
4586    <author initials='H.' surname='Alvestrand' fullname='H. Alvestrand'>
4587      <organization>Google</organization>
4588      <address><email></email></address>
4589    </author>
4590    <date year='2008' month='May' />
4591  </front>
4592  <seriesInfo name='BCP' value='26' />
4593  <seriesInfo name='RFC' value='5226' />
4596<reference anchor='RFC5246'>
4597   <front>
4598      <title>The Transport Layer Security (TLS) Protocol Version 1.2</title>
4599      <author initials='T.' surname='Dierks' fullname='T. Dierks'>
4600         <organization />
4601      </author>
4602      <author initials='E.' surname='Rescorla' fullname='E. Rescorla'>
4603         <organization>RTFM, Inc.</organization>
4604      </author>
4605      <date year='2008' month='August' />
4606   </front>
4607   <seriesInfo name='RFC' value='5246' />
4610<reference anchor="RFC5322">
4611  <front>
4612    <title>Internet Message Format</title>
4613    <author initials="P." surname="Resnick" fullname="P. Resnick">
4614      <organization>Qualcomm Incorporated</organization>
4615    </author>
4616    <date year="2008" month="October"/>
4617  </front>
4618  <seriesInfo name="RFC" value="5322"/>
4621<reference anchor="RFC6265">
4622  <front>
4623    <title>HTTP State Management Mechanism</title>
4624    <author initials="A." surname="Barth" fullname="Adam Barth">
4625      <organization abbrev="U.C. Berkeley">
4626        University of California, Berkeley
4627      </organization>
4628      <address><email></email></address>
4629    </author>
4630    <date year="2011" month="April" />
4631  </front>
4632  <seriesInfo name="RFC" value="6265"/>
4635<!--<reference anchor='BCP97'>
4636  <front>
4637    <title>Handling Normative References to Standards-Track Documents</title>
4638    <author initials='J.' surname='Klensin' fullname='J. Klensin'>
4639      <address>
4640        <email></email>
4641      </address>
4642    </author>
4643    <author initials='S.' surname='Hartman' fullname='S. Hartman'>
4644      <organization>MIT</organization>
4645      <address>
4646        <email></email>
4647      </address>
4648    </author>
4649    <date year='2007' month='June' />
4650  </front>
4651  <seriesInfo name='BCP' value='97' />
4652  <seriesInfo name='RFC' value='4897' />
4655<reference anchor="Kri2001" target="">
4656  <front>
4657    <title>HTTP Cookies: Standards, Privacy, and Politics</title>
4658    <author initials="D." surname="Kristol" fullname="David M. Kristol"/>
4659    <date year="2001" month="November"/>
4660  </front>
4661  <seriesInfo name="ACM Transactions on Internet Technology" value="Vol. 1, #2"/>
4667<section title="HTTP Version History" anchor="compatibility">
4669   HTTP has been in use by the World-Wide Web global information initiative
4670   since 1990. The first version of HTTP, later referred to as HTTP/0.9,
4671   was a simple protocol for hypertext data transfer across the Internet
4672   with only a single request method (GET) and no metadata.
4673   HTTP/1.0, as defined by <xref target="RFC1945"/>, added a range of request
4674   methods and MIME-like messaging that could include metadata about the data
4675   transferred and modifiers on the request/response semantics. However,
4676   HTTP/1.0 did not sufficiently take into consideration the effects of
4677   hierarchical proxies, caching, the need for persistent connections, or
4678   name-based virtual hosts. The proliferation of incompletely-implemented
4679   applications calling themselves "HTTP/1.0" further necessitated a
4680   protocol version change in order for two communicating applications
4681   to determine each other's true capabilities.
4684   HTTP/1.1 remains compatible with HTTP/1.0 by including more stringent
4685   requirements that enable reliable implementations, adding only
4686   those new features that will either be safely ignored by an HTTP/1.0
4687   recipient or only sent when communicating with a party advertising
4688   conformance with HTTP/1.1.
4691   It is beyond the scope of a protocol specification to mandate
4692   conformance with previous versions. HTTP/1.1 was deliberately
4693   designed, however, to make supporting previous versions easy.
4694   We would expect a general-purpose HTTP/1.1 server to understand
4695   any valid request in the format of HTTP/1.0 and respond appropriately
4696   with an HTTP/1.1 message that only uses features understood (or
4697   safely ignored) by HTTP/1.0 clients.  Likewise, we would expect
4698   an HTTP/1.1 client to understand any valid HTTP/1.0 response.
4701   Since HTTP/0.9 did not support header fields in a request,
4702   there is no mechanism for it to support name-based virtual
4703   hosts (selection of resource by inspection of the <x:ref>Host</x:ref> header
4704   field).  Any server that implements name-based virtual hosts
4705   ought to disable support for HTTP/0.9.  Most requests that
4706   appear to be HTTP/0.9 are, in fact, badly constructed HTTP/1.x
4707   requests wherein a buggy client failed to properly encode
4708   linear whitespace found in a URI reference and placed in
4709   the request-target.
4712<section title="Changes from HTTP/1.0" anchor="changes.from.1.0">
4714   This section summarizes major differences between versions HTTP/1.0
4715   and HTTP/1.1.
4718<section title="Multi-homed Web Servers" anchor="">
4720   The requirements that clients and servers support the <x:ref>Host</x:ref>
4721   header field (<xref target=""/>), report an error if it is
4722   missing from an HTTP/1.1 request, and accept absolute URIs (<xref target="request-target"/>)
4723   are among the most important changes defined by HTTP/1.1.
4726   Older HTTP/1.0 clients assumed a one-to-one relationship of IP
4727   addresses and servers; there was no other established mechanism for
4728   distinguishing the intended server of a request than the IP address
4729   to which that request was directed. The <x:ref>Host</x:ref> header field was
4730   introduced during the development of HTTP/1.1 and, though it was
4731   quickly implemented by most HTTP/1.0 browsers, additional requirements
4732   were placed on all HTTP/1.1 requests in order to ensure complete
4733   adoption.  At the time of this writing, most HTTP-based services
4734   are dependent upon the Host header field for targeting requests.
4738<section title="Keep-Alive Connections" anchor="compatibility.with.http.1.0.persistent.connections">
4740   In HTTP/1.0, each connection is established by the client prior to the
4741   request and closed by the server after sending the response. However, some
4742   implementations implement the explicitly negotiated ("Keep-Alive") version
4743   of persistent connections described in <xref x:sec="19.7.1" x:fmt="of"
4744   target="RFC2068"/>.
4747   Some clients and servers might wish to be compatible with these previous
4748   approaches to persistent connections, by explicitly negotiating for them
4749   with a "Connection: keep-alive" request header field. However, some
4750   experimental implementations of HTTP/1.0 persistent connections are faulty;
4751   for example, if an HTTP/1.0 proxy server doesn't understand
4752   <x:ref>Connection</x:ref>, it will erroneously forward that header field
4753   to the next inbound server, which would result in a hung connection.
4756   One attempted solution was the introduction of a Proxy-Connection header
4757   field, targeted specifically at proxies. In practice, this was also
4758   unworkable, because proxies are often deployed in multiple layers, bringing
4759   about the same problem discussed above.
4762   As a result, clients are encouraged not to send the Proxy-Connection header
4763   field in any requests.
4766   Clients are also encouraged to consider the use of Connection: keep-alive
4767   in requests carefully; while they can enable persistent connections with
4768   HTTP/1.0 servers, clients using them will need to monitor the
4769   connection for "hung" requests (which indicate that the client ought stop
4770   sending the header field), and this mechanism ought not be used by clients
4771   at all when a proxy is being used.
4775<section title="Introduction of Transfer-Encoding" anchor="introduction.of.transfer-encoding">
4777   HTTP/1.1 introduces the <x:ref>Transfer-Encoding</x:ref> header field
4778   (<xref target="header.transfer-encoding"/>).
4779   Transfer codings need to be decoded prior to forwarding an HTTP message
4780   over a MIME-compliant protocol.
4786<section title="Changes from RFC 2616" anchor="changes.from.rfc.2616">
4788  HTTP's approach to error handling has been explained.
4789  (<xref target="conformance"/>)
4792  The expectation to support HTTP/0.9 requests has been removed.
4795  The term "Effective Request URI" has been introduced.
4796  (<xref target="effective.request.uri" />)
4799  HTTP messages can be (and often are) buffered by implementations; despite
4800  it sometimes being available as a stream, HTTP is fundamentally a
4801  message-oriented protocol.
4802  (<xref target="http.message" />)
4805  Minimum supported sizes for various protocol elements have been
4806  suggested, to improve interoperability.
4809  Header fields that span multiple lines ("line folding") are deprecated.
4810  (<xref target="field.parsing" />)
4813  The HTTP-version ABNF production has been clarified to be case-sensitive.
4814  Additionally, version numbers has been restricted to single digits, due
4815  to the fact that implementations are known to handle multi-digit version
4816  numbers incorrectly.
4817  (<xref target="http.version"/>)
4820  The HTTPS URI scheme is now defined by this specification; previously,
4821  it was done in  <xref target="RFC2818" x:fmt="of" x:sec="2.4"/>.
4822  (<xref target="https.uri"/>)
4825  The HTTPS URI scheme implies end-to-end security.
4826  (<xref target="https.uri"/>)
4829  Userinfo (i.e., username and password) are now disallowed in HTTP and
4830  HTTPS URIs, because of security issues related to their transmission on the
4831  wire.
4832  (<xref target="http.uri" />)
4835  Invalid whitespace around field-names is now required to be rejected,
4836  because accepting it represents a security vulnerability.
4837  (<xref target="header.fields"/>)
4840  The ABNF productions defining header fields now only list the field value.
4841  (<xref target="header.fields"/>)
4844  Rules about implicit linear whitespace between certain grammar productions
4845  have been removed; now whitespace is only allowed where specifically
4846  defined in the ABNF.
4847  (<xref target="whitespace"/>)
4850  The NUL octet is no longer allowed in comment and quoted-string text, and
4851  handling of backslash-escaping in them has been clarified.
4852  (<xref target="field.components"/>)
4855  The quoted-pair rule no longer allows escaping control characters other than
4856  HTAB.
4857  (<xref target="field.components"/>)
4860  Non-ASCII content in header fields and the reason phrase has been obsoleted
4861  and made opaque (the TEXT rule was removed).
4862  (<xref target="field.components"/>)
4865  Bogus "<x:ref>Content-Length</x:ref>" header fields are now required to be
4866  handled as errors by recipients.
4867  (<xref target="header.content-length"/>)
4870  The "identity" transfer coding token has been removed.
4871  (Sections <xref format="counter" target="message.body"/> and
4872  <xref format="counter" target="transfer.codings"/>)
4875  The algorithm for determining the message body length has been clarified
4876  to indicate all of the special cases (e.g., driven by methods or status
4877  codes) that affect it, and that new protocol elements cannot define such
4878  special cases.
4879  (<xref target="message.body.length"/>)
4882  "multipart/byteranges" is no longer a way of determining message body length
4883  detection.
4884  (<xref target="message.body.length"/>)
4887  CONNECT is a new, special case in determining message body length.
4888  (<xref target="message.body.length"/>)
4891  Chunk length does not include the count of the octets in the
4892  chunk header and trailer.
4893  (<xref target="chunked.encoding"/>)
4896  Use of chunk extensions is deprecated, and line folding in them is
4897  disallowed.
4898  (<xref target="chunked.encoding"/>)
4901  The segment + query components of RFC3986 have been used to define the
4902  request-target, instead of abs_path from RFC 1808.
4903  (<xref target="request-target"/>)
4906  The asterisk form of the request-target is only allowed in the OPTIONS
4907  method.
4908  (<xref target="request-target"/>)
4911  Exactly when "close" connection options have to be sent has been clarified.
4912  (<xref target="header.connection"/>)
4915  "hop-by-hop" header fields are required to appear in the Connection header
4916  field; just because they're defined as hop-by-hop in this specification
4917  doesn't exempt them.
4918  (<xref target="header.connection"/>)
4921  The limit of two connections per server has been removed.
4922  (<xref target="persistent.connections"/>)
4925  An idempotent sequence of requests is no longer required to be retried.
4926  (<xref target="persistent.connections"/>)
4929  The requirement to retry requests under certain circumstances when the
4930  server prematurely closes the connection has been removed.
4931  (<xref target="persistent.connections"/>)
4934  Some extraneous requirements about when servers are allowed to close
4935  connections prematurely have been removed.
4936  (<xref target="persistent.connections"/>)
4939  The semantics of the <x:ref>Upgrade</x:ref> header field is now defined in
4940  responses other than 101 (this was incorporated from <xref
4941  target="RFC2817"/>).
4942  (<xref target="header.upgrade"/>)
4945  Registration of Transfer Codings now requires IETF Review
4946  (<xref target="transfer.coding.registry"/>)
4949  The meaning of the "deflate" content coding has been clarified.
4950  (<xref target="deflate.coding" />)
4953  This specification now defines the Upgrade Token Registry, previously
4954  defined in <xref target="RFC2817" x:fmt="of" x:sec="7.2"/>.
4955  (<xref target="upgrade.token.registry"/>)
4958  Empty list elements in list productions (e.g., a list header containing
4959  ", ,") have been deprecated.
4960  (<xref target="abnf.extension"/>)
4963  Issues with the Keep-Alive and Proxy-Connection headers in requests
4964  are pointed out, with use of the latter being discouraged altogether.
4965  (<xref target="compatibility.with.http.1.0.persistent.connections" />)
4970<section title="ABNF list extension: #rule" anchor="abnf.extension">
4972  A #rule extension to the ABNF rules of <xref target="RFC5234"/> is used to
4973  improve readability in the definitions of some header field values.
4976  A construct "#" is defined, similar to "*", for defining comma-delimited
4977  lists of elements. The full form is "&lt;n&gt;#&lt;m&gt;element" indicating
4978  at least &lt;n&gt; and at most &lt;m&gt; elements, each separated by a single
4979  comma (",") and optional whitespace (OWS).   
4982  Thus,
4983</preamble><artwork type="example">
4984  1#element =&gt; element *( OWS "," OWS element )
4987  and:
4988</preamble><artwork type="example">
4989  #element =&gt; [ 1#element ]
4992  and for n &gt;= 1 and m &gt; 1:
4993</preamble><artwork type="example">
4994  &lt;n&gt;#&lt;m&gt;element =&gt; element &lt;n-1&gt;*&lt;m-1&gt;( OWS "," OWS element )
4997  For compatibility with legacy list rules, recipients &SHOULD; accept empty
4998  list elements. In other words, consumers would follow the list productions:
5000<figure><artwork type="example">
5001  #element =&gt; [ ( "," / element ) *( OWS "," [ OWS element ] ) ]
5003  1#element =&gt; *( "," OWS ) element *( OWS "," [ OWS element ] )
5006  Note that empty elements do not contribute to the count of elements present,
5007  though.
5010  For example, given these ABNF productions:
5012<figure><artwork type="example">
5013  example-list      = 1#example-list-elmt
5014  example-list-elmt = token ; see <xref target="field.components"/>
5017  Then these are valid values for example-list (not including the double
5018  quotes, which are present for delimitation only):
5020<figure><artwork type="example">
5021  "foo,bar"
5022  "foo ,bar,"
5023  "foo , ,bar,charlie   "
5026  But these values would be invalid, as at least one non-empty element is
5027  required:
5029<figure><artwork type="example">
5030  ""
5031  ","
5032  ",   ,"
5035  <xref target="collected.abnf"/> shows the collected ABNF, with the list rules
5036  expanded as explained above.
5040<?BEGININC p1-messaging.abnf-appendix ?>
5041<section xmlns:x="" title="Collected ABNF" anchor="collected.abnf">
5043<artwork type="abnf" name="p1-messaging.parsed-abnf">
5044<x:ref>BWS</x:ref> = OWS
5046<x:ref>Connection</x:ref> = *( "," OWS ) connection-option *( OWS "," [ OWS
5047 connection-option ] )
5048<x:ref>Content-Length</x:ref> = 1*DIGIT
5050<x:ref>HTTP-message</x:ref> = start-line *( header-field CRLF ) CRLF [ message-body
5051 ]
5052<x:ref>HTTP-name</x:ref> = %x48.54.54.50 ; HTTP
5053<x:ref>HTTP-version</x:ref> = HTTP-name "/" DIGIT "." DIGIT
5054<x:ref>Host</x:ref> = uri-host [ ":" port ]
5056<x:ref>OWS</x:ref> = *( SP / HTAB )
5058<x:ref>RWS</x:ref> = 1*( SP / HTAB )
5060<x:ref>TE</x:ref> = [ ( "," / t-codings ) *( OWS "," [ OWS t-codings ] ) ]
5061<x:ref>Trailer</x:ref> = *( "," OWS ) field-name *( OWS "," [ OWS field-name ] )
5062<x:ref>Transfer-Encoding</x:ref> = *( "," OWS ) transfer-coding *( OWS "," [ OWS
5063 transfer-coding ] )
5065<x:ref>URI-reference</x:ref> = &lt;URI-reference, defined in [RFC3986], Section 4.1&gt;
5066<x:ref>Upgrade</x:ref> = *( "," OWS ) protocol *( OWS "," [ OWS protocol ] )
5068<x:ref>Via</x:ref> = *( "," OWS ) ( received-protocol RWS received-by [ RWS comment
5069 ] ) *( OWS "," [ OWS ( received-protocol RWS received-by [ RWS
5070 comment ] ) ] )
5072<x:ref>absolute-URI</x:ref> = &lt;absolute-URI, defined in [RFC3986], Section 4.3&gt;
5073<x:ref>absolute-form</x:ref> = absolute-URI
5074<x:ref>absolute-path</x:ref> = 1*( "/" segment )
5075<x:ref>asterisk-form</x:ref> = "*"
5076<x:ref>attribute</x:ref> = token
5077<x:ref>authority</x:ref> = &lt;authority, defined in [RFC3986], Section 3.2&gt;
5078<x:ref>authority-form</x:ref> = authority
5080<x:ref>chunk</x:ref> = chunk-size [ chunk-ext ] CRLF chunk-data CRLF
5081<x:ref>chunk-data</x:ref> = 1*OCTET
5082<x:ref>chunk-ext</x:ref> = *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
5083<x:ref>chunk-ext-name</x:ref> = token
5084<x:ref>chunk-ext-val</x:ref> = token / quoted-str-nf
5085<x:ref>chunk-size</x:ref> = 1*HEXDIG
5086<x:ref>chunked-body</x:ref> = *chunk last-chunk trailer-part CRLF
5087<x:ref>comment</x:ref> = "(" *( ctext / quoted-cpair / comment ) ")"
5088<x:ref>connection-option</x:ref> = token
5089<x:ref>ctext</x:ref> = HTAB / SP / %x21-27 ; '!'-'''
5090 / %x2A-5B ; '*'-'['
5091 / %x5D-7E ; ']'-'~'
5092 / obs-text
5094<x:ref>field-content</x:ref> = *( HTAB / SP / VCHAR / obs-text )
5095<x:ref>field-name</x:ref> = token
5096<x:ref>field-value</x:ref> = *( field-content / obs-fold )
5098<x:ref>header-field</x:ref> = field-name ":" OWS field-value BWS
5099<x:ref>http-URI</x:ref> = "http://" authority path-abempty [ "?" query ]
5100<x:ref>https-URI</x:ref> = "https://" authority path-abempty [ "?" query ]
5102<x:ref>last-chunk</x:ref> = 1*"0" [ chunk-ext ] CRLF
5104<x:ref>message-body</x:ref> = *OCTET
5105<x:ref>method</x:ref> = token
5107<x:ref>obs-fold</x:ref> = CRLF ( SP / HTAB )
5108<x:ref>obs-text</x:ref> = %x80-FF
5109<x:ref>origin-form</x:ref> = absolute-path [ "?" query ]
5111<x:ref>partial-URI</x:ref> = relative-part [ "?" query ]
5112<x:ref>path-abempty</x:ref> = &lt;path-abempty, defined in [RFC3986], Section 3.3&gt;
5113<x:ref>port</x:ref> = &lt;port, defined in [RFC3986], Section 3.2.3&gt;
5114<x:ref>protocol</x:ref> = protocol-name [ "/" protocol-version ]
5115<x:ref>protocol-name</x:ref> = token
5116<x:ref>protocol-version</x:ref> = token
5117<x:ref>pseudonym</x:ref> = token
5119<x:ref>qdtext</x:ref> = HTAB / SP / "!" / %x23-5B ; '#'-'['
5120 / %x5D-7E ; ']'-'~'
5121 / obs-text
5122<x:ref>qdtext-nf</x:ref> = HTAB / SP / "!" / %x23-5B ; '#'-'['
5123 / %x5D-7E ; ']'-'~'
5124 / obs-text
5125<x:ref>query</x:ref> = &lt;query, defined in [RFC3986], Section 3.4&gt;
5126<x:ref>quoted-cpair</x:ref> = "\" ( HTAB / SP / VCHAR / obs-text )
5127<x:ref>quoted-pair</x:ref> = "\" ( HTAB / SP / VCHAR / obs-text )
5128<x:ref>quoted-str-nf</x:ref> = DQUOTE *( qdtext-nf / quoted-pair ) DQUOTE
5129<x:ref>quoted-string</x:ref> = DQUOTE *( qdtext / quoted-pair ) DQUOTE
5131<x:ref>rank</x:ref> = ( "0" [ "." *3DIGIT ] ) / ( "1" [ "." *3"0" ] )
5132<x:ref>reason-phrase</x:ref> = *( HTAB / SP / VCHAR / obs-text )
5133<x:ref>received-by</x:ref> = ( uri-host [ ":" port ] ) / pseudonym
5134<x:ref>received-protocol</x:ref> = [ protocol-name "/" ] protocol-version
5135<x:ref>relative-part</x:ref> = &lt;relative-part, defined in [RFC3986], Section 4.2&gt;
5136<x:ref>request-line</x:ref> = method SP request-target SP HTTP-version CRLF
5137<x:ref>request-target</x:ref> = origin-form / absolute-form / authority-form /
5138 asterisk-form
5140<x:ref>segment</x:ref> = &lt;segment, defined in [RFC3986], Section 3.3&gt;
5141<x:ref>special</x:ref> = "(" / ")" / "&lt;" / "&gt;" / "@" / "," / ";" / ":" / "\" /
5142 DQUOTE / "/" / "[" / "]" / "?" / "=" / "{" / "}"
5143<x:ref>start-line</x:ref> = request-line / status-line
5144<x:ref>status-code</x:ref> = 3DIGIT
5145<x:ref>status-line</x:ref> = HTTP-version SP status-code SP reason-phrase CRLF
5147<x:ref>t-codings</x:ref> = "trailers" / ( transfer-coding [ t-ranking ] )
5148<x:ref>t-ranking</x:ref> = OWS ";" OWS "q=" rank
5149<x:ref>tchar</x:ref> = "!" / "#" / "$" / "%" / "&amp;" / "'" / "*" / "+" / "-" / "." /
5150 "^" / "_" / "`" / "|" / "~" / DIGIT / ALPHA
5151<x:ref>token</x:ref> = 1*tchar
5152<x:ref>trailer-part</x:ref> = *( header-field CRLF )
5153<x:ref>transfer-coding</x:ref> = "chunked" / "compress" / "deflate" / "gzip" /
5154 transfer-extension
5155<x:ref>transfer-extension</x:ref> = token *( OWS ";" OWS transfer-parameter )
5156<x:ref>transfer-parameter</x:ref> = attribute BWS "=" BWS value
5158<x:ref>uri-host</x:ref> = &lt;host, defined in [RFC3986], Section 3.2.2&gt;
5160<x:ref>value</x:ref> = word
5162<x:ref>word</x:ref> = token / quoted-string
5166<?ENDINC p1-messaging.abnf-appendix ?>
5168<section title="Change Log (to be removed by RFC Editor before publication)" anchor="change.log">
5170<section title="Since RFC 2616">
5172  Changes up to the first Working Group Last Call draft are summarized
5173  in <eref target=""/>.
5177<section title="Since draft-ietf-httpbis-p1-messaging-21" anchor="changes.since.21">
5179  Closed issues:
5180  <list style="symbols">
5181    <t>
5182      <eref target=""/>:
5183      "Cite HTTPS URI scheme definition" (the spec now includes the HTTPs
5184      scheme definition and thus updates RFC 2818)
5185    </t>
5186    <t>
5187      <eref target=""/>:
5188      "mention of 'proxies' in section about caches"
5189    </t>
5190    <t>
5191      <eref target=""/>:
5192      "use of ABNF terms from RFC 3986"
5193    </t>
5194    <t>
5195      <eref target=""/>:
5196      "transferring URIs with userinfo in payload"
5197    </t>
5198    <t>
5199      <eref target=""/>:
5200      "editorial improvements to message length definition"
5201    </t>
5202    <t>
5203      <eref target=""/>:
5204      "Connection header field MUST vs SHOULD"
5205    </t>
5206    <t>
5207      <eref target=""/>:
5208      "editorial improvements to persistent connections section"
5209    </t>
5210    <t>
5211      <eref target=""/>:
5212      "URI normalization vs empty path"
5213    </t>
5214    <t>
5215      <eref target=""/>:
5216      "p1 feedback"
5217    </t>
5218    <t>
5219      <eref target=""/>:
5220      "is parsing OBS-FOLD mandatory?"
5221    </t>
5222    <t>
5223      <eref target=""/>:
5224      "HTTPS and Shared Caching"
5225    </t>
5226    <t>
5227      <eref target=""/>:
5228      "Requirements for recipients of ws between start-line and first header field"
5229    </t>
5230    <t>
5231      <eref target=""/>:
5232      "SP and HT when being tolerant"
5233    </t>
5234    <t>
5235      <eref target=""/>:
5236      "Message Parsing Strictness"
5237    </t>
5238    <t>
5239      <eref target=""/>:
5240      "'Render'"
5241    </t>
5242    <t>
5243      <eref target=""/>:
5244      "No-Transform"
5245    </t>
5246    <t>
5247      <eref target=""/>:
5248      "p2 editorial feedback"
5249    </t>
5250    <t>
5251      <eref target=""/>:
5252      "Content-Length SHOULD be sent"
5253    </t>
5254    <t>
5255      <eref target=""/>:
5256      "origin-form does not allow path starting with "//""
5257    </t>
5258    <t>
5259      <eref target=""/>:
5260      "ambiguity in part 1 example"
5261    </t>
5262  </list>
5266<section title="Since draft-ietf-httpbis-p1-messaging-22" anchor="changes.since.22">
5268  Closed issues:
5269  <list style="symbols">
5270    <t>
5271      <eref target=""/>:
5272      "Part1 should have a reference to TCP (RFC 793)"
5273    </t>
5274    <t>
5275      <eref target=""/>:
5276      "media type registration template issues"
5277    </t>
5278    <t>
5279      <eref target=""/>:
5280      "BWS" (vs conformance)
5281    </t>
5282  </list>
Note: See TracBrowser for help on using the repository browser.