idnits 2.17.1 draft-ietf-httpbis-messaging-04.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- -- The draft header indicates that this document obsoletes RFC7230, but the abstract doesn't seem to directly say this. It does mention RFC7230 though, so this could be OK. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document seems to contain a disclaimer for pre-RFC5378 work, but was first submitted on or after 10 November 2008. The disclaimer is usually necessary only for documents that revise or obsolete older RFCs, and that take significant amounts of text from those RFCs. If you can contact all authors of the source material and they are willing to grant the BCP78 rights to the IETF Trust, you can and should remove the disclaimer. Otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (March 9, 2019) is 1868 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Unused Reference: 'RFC3986' is defined on line 2037, but no explicit reference was found in the text == Unused Reference: 'RFC7231' is defined on line 2120, but no explicit reference was found in the text == Outdated reference: A later version (-19) exists of draft-ietf-httpbis-cache-04 -- Possible downref: Normative reference to a draft: ref. 'Caching' ** Downref: Normative reference to an Informational RFC: RFC 1950 ** Downref: Normative reference to an Informational RFC: RFC 1951 ** Downref: Normative reference to an Informational RFC: RFC 1952 == Outdated reference: A later version (-19) exists of draft-ietf-httpbis-semantics-04 -- Possible downref: Normative reference to a draft: ref. 'Semantics' -- Possible downref: Non-RFC (?) normative reference: ref. 'USASCII' -- Possible downref: Non-RFC (?) normative reference: ref. 'Welch' -- Obsolete informational reference (is this intentional?): RFC 7230 (ref. 'Err4667') (Obsoleted by RFC 9110, RFC 9112) -- Obsolete informational reference (is this intentional?): RFC 2068 (Obsoleted by RFC 2616) -- Duplicate reference: RFC7230, mentioned in 'RFC7230', was also mentioned in 'Err4667'. -- Obsolete informational reference (is this intentional?): RFC 7230 (Obsoleted by RFC 9110, RFC 9112) -- Obsolete informational reference (is this intentional?): RFC 7231 (Obsoleted by RFC 9110) Summary: 3 errors (**), 0 flaws (~~), 6 warnings (==), 11 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 HTTP Working Group R. Fielding, Ed. 3 Internet-Draft Adobe 4 Obsoletes: 7230 (if approved) M. Nottingham, Ed. 5 Intended status: Standards Track Fastly 6 Expires: September 10, 2019 J. Reschke, Ed. 7 greenbytes 8 March 9, 2019 10 HTTP/1.1 Messaging 11 draft-ietf-httpbis-messaging-04 13 Abstract 15 The Hypertext Transfer Protocol (HTTP) is a stateless application- 16 level protocol for distributed, collaborative, hypertext information 17 systems. This document specifies the HTTP/1.1 message syntax, 18 message parsing, connection management, and related security 19 concerns. 21 This document obsoletes portions of RFC 7230. 23 Editorial Note 25 This note is to be removed before publishing as an RFC. 27 Discussion of this draft takes place on the HTTP working group 28 mailing list (ietf-http-wg@w3.org), which is archived at 29 . 31 Working Group information can be found at ; 32 source code and issues list for this draft can be found at 33 . 35 The changes in this draft are summarized in Appendix D.5. 37 Status of This Memo 39 This Internet-Draft is submitted in full conformance with the 40 provisions of BCP 78 and BCP 79. 42 Internet-Drafts are working documents of the Internet Engineering 43 Task Force (IETF). Note that other groups may also distribute 44 working documents as Internet-Drafts. The list of current Internet- 45 Drafts is at https://datatracker.ietf.org/drafts/current/. 47 Internet-Drafts are draft documents valid for a maximum of six months 48 and may be updated, replaced, or obsoleted by other documents at any 49 time. It is inappropriate to use Internet-Drafts as reference 50 material or to cite them other than as "work in progress." 52 This Internet-Draft will expire on September 10, 2019. 54 Copyright Notice 56 Copyright (c) 2019 IETF Trust and the persons identified as the 57 document authors. All rights reserved. 59 This document is subject to BCP 78 and the IETF Trust's Legal 60 Provisions Relating to IETF Documents 61 (https://trustee.ietf.org/license-info) in effect on the date of 62 publication of this document. Please review these documents 63 carefully, as they describe your rights and restrictions with respect 64 to this document. Code Components extracted from this document must 65 include Simplified BSD License text as described in Section 4.e of 66 the Trust Legal Provisions and are provided without warranty as 67 described in the Simplified BSD License. 69 This document may contain material from IETF Documents or IETF 70 Contributions published or made publicly available before November 71 10, 2008. The person(s) controlling the copyright in some of this 72 material may not have granted the IETF Trust the right to allow 73 modifications of such material outside the IETF Standards Process. 74 Without obtaining an adequate license from the person(s) controlling 75 the copyright in such materials, this document may not be modified 76 outside the IETF Standards Process, and derivative works of it may 77 not be created outside the IETF Standards Process, except to format 78 it for publication as an RFC or to translate it into languages other 79 than English. 81 Table of Contents 83 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 84 1.1. Requirements Notation . . . . . . . . . . . . . . . . . . 5 85 1.2. Syntax Notation . . . . . . . . . . . . . . . . . . . . . 5 86 2. Message . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 87 2.1. Message Format . . . . . . . . . . . . . . . . . . . . . 6 88 2.2. HTTP Version . . . . . . . . . . . . . . . . . . . . . . 6 89 2.3. Message Parsing . . . . . . . . . . . . . . . . . . . . . 7 90 3. Request Line . . . . . . . . . . . . . . . . . . . . . . . . 8 91 3.1. Method . . . . . . . . . . . . . . . . . . . . . . . . . 9 92 3.2. Request Target . . . . . . . . . . . . . . . . . . . . . 9 93 3.2.1. origin-form . . . . . . . . . . . . . . . . . . . . . 10 94 3.2.2. absolute-form . . . . . . . . . . . . . . . . . . . . 10 95 3.2.3. authority-form . . . . . . . . . . . . . . . . . . . 11 96 3.2.4. asterisk-form . . . . . . . . . . . . . . . . . . . . 11 98 3.3. Effective Request URI . . . . . . . . . . . . . . . . . . 12 99 4. Status Line . . . . . . . . . . . . . . . . . . . . . . . . . 13 100 5. Header Fields . . . . . . . . . . . . . . . . . . . . . . . . 14 101 5.1. Header Field Parsing . . . . . . . . . . . . . . . . . . 15 102 5.2. Obsolete Line Folding . . . . . . . . . . . . . . . . . . 15 103 6. Message Body . . . . . . . . . . . . . . . . . . . . . . . . 16 104 6.1. Transfer-Encoding . . . . . . . . . . . . . . . . . . . . 17 105 6.2. Content-Length . . . . . . . . . . . . . . . . . . . . . 18 106 6.3. Message Body Length . . . . . . . . . . . . . . . . . . . 19 107 7. Transfer Codings . . . . . . . . . . . . . . . . . . . . . . 21 108 7.1. Chunked Transfer Coding . . . . . . . . . . . . . . . . . 22 109 7.1.1. Chunk Extensions . . . . . . . . . . . . . . . . . . 22 110 7.1.2. Chunked Trailer Part . . . . . . . . . . . . . . . . 23 111 7.1.3. Decoding Chunked . . . . . . . . . . . . . . . . . . 24 112 7.2. Transfer Codings for Compression . . . . . . . . . . . . 24 113 7.3. Transfer Coding Registry . . . . . . . . . . . . . . . . 25 114 7.4. TE . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 115 8. Handling Incomplete Messages . . . . . . . . . . . . . . . . 26 116 9. Connection Management . . . . . . . . . . . . . . . . . . . . 27 117 9.1. Connection . . . . . . . . . . . . . . . . . . . . . . . 28 118 9.2. Establishment . . . . . . . . . . . . . . . . . . . . . . 29 119 9.3. Associating a Response to a Request . . . . . . . . . . . 29 120 9.4. Persistence . . . . . . . . . . . . . . . . . . . . . . . 30 121 9.4.1. Retrying Requests . . . . . . . . . . . . . . . . . . 31 122 9.4.2. Pipelining . . . . . . . . . . . . . . . . . . . . . 31 123 9.5. Concurrency . . . . . . . . . . . . . . . . . . . . . . . 32 124 9.6. Failures and Timeouts . . . . . . . . . . . . . . . . . . 32 125 9.7. Tear-down . . . . . . . . . . . . . . . . . . . . . . . . 33 126 9.8. Upgrade . . . . . . . . . . . . . . . . . . . . . . . . . 34 127 9.8.1. Upgrade Protocol Names . . . . . . . . . . . . . . . 36 128 9.8.2. Upgrade Token Registry . . . . . . . . . . . . . . . 37 129 10. Enclosing Messages as Data . . . . . . . . . . . . . . . . . 37 130 10.1. Media Type message/http . . . . . . . . . . . . . . . . 38 131 10.2. Media Type application/http . . . . . . . . . . . . . . 39 132 11. Security Considerations . . . . . . . . . . . . . . . . . . . 40 133 11.1. Response Splitting . . . . . . . . . . . . . . . . . . . 40 134 11.2. Request Smuggling . . . . . . . . . . . . . . . . . . . 41 135 11.3. Message Integrity . . . . . . . . . . . . . . . . . . . 41 136 11.4. Message Confidentiality . . . . . . . . . . . . . . . . 42 137 12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 42 138 12.1. Header Field Registration . . . . . . . . . . . . . . . 42 139 12.2. Media Type Registration . . . . . . . . . . . . . . . . 42 140 12.3. Transfer Coding Registration . . . . . . . . . . . . . . 43 141 12.4. Upgrade Token Registration . . . . . . . . . . . . . . . 43 142 13. References . . . . . . . . . . . . . . . . . . . . . . . . . 43 143 13.1. Normative References . . . . . . . . . . . . . . . . . . 43 144 13.2. Informative References . . . . . . . . . . . . . . . . . 44 145 Appendix A. Collected ABNF . . . . . . . . . . . . . . . . . . . 46 146 Appendix B. Differences between HTTP and MIME . . . . . . . . . 47 147 B.1. MIME-Version . . . . . . . . . . . . . . . . . . . . . . 48 148 B.2. Conversion to Canonical Form . . . . . . . . . . . . . . 48 149 B.3. Conversion of Date Formats . . . . . . . . . . . . . . . 48 150 B.4. Conversion of Content-Encoding . . . . . . . . . . . . . 49 151 B.5. Conversion of Content-Transfer-Encoding . . . . . . . . . 49 152 B.6. MHTML and Line Length Limitations . . . . . . . . . . . . 49 153 Appendix C. HTTP Version History . . . . . . . . . . . . . . . . 49 154 C.1. Changes from HTTP/1.0 . . . . . . . . . . . . . . . . . . 50 155 C.1.1. Multihomed Web Servers . . . . . . . . . . . . . . . 50 156 C.1.2. Keep-Alive Connections . . . . . . . . . . . . . . . 51 157 C.1.3. Introduction of Transfer-Encoding . . . . . . . . . . 51 158 C.2. Changes from RFC 7230 . . . . . . . . . . . . . . . . . . 51 159 Appendix D. Change Log . . . . . . . . . . . . . . . . . . . . . 52 160 D.1. Between RFC7230 and draft 00 . . . . . . . . . . . . . . 52 161 D.2. Since draft-ietf-httpbis-messaging-00 . . . . . . . . . . 52 162 D.3. Since draft-ietf-httpbis-messaging-01 . . . . . . . . . . 53 163 D.4. Since draft-ietf-httpbis-messaging-02 . . . . . . . . . . 53 164 D.5. Since draft-ietf-httpbis-messaging-03 . . . . . . . . . . 54 165 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 166 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 56 167 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 56 169 1. Introduction 171 The Hypertext Transfer Protocol (HTTP) is a stateless application- 172 level request/response protocol that uses extensible semantics and 173 self-descriptive messages for flexible interaction with network-based 174 hypertext information systems. HTTP is defined by a series of 175 documents that collectively form the HTTP/1.1 specification: 177 o "HTTP Semantics" [Semantics] 179 o "HTTP Caching" [Caching] 181 o "HTTP/1.1 Messaging" (this document) 183 This document defines HTTP/1.1 message syntax and framing 184 requirements and their associated connection management. Our goal is 185 to define all of the mechanisms necessary for HTTP/1.1 message 186 handling that are independent of message semantics, thereby defining 187 the complete set of requirements for message parsers and message- 188 forwarding intermediaries. 190 This document obsoletes the portions of RFC 7230 related to HTTP/1.1 191 messaging and connection management, with the changes being 192 summarized in Appendix C.2. The other parts of RFC 7230 are 193 obsoleted by "HTTP Semantics" [Semantics]. 195 1.1. Requirements Notation 197 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 198 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 199 document are to be interpreted as described in [RFC2119]. 201 Conformance criteria and considerations regarding error handling are 202 defined in Section 3 of [Semantics]. 204 1.2. Syntax Notation 206 This specification uses the Augmented Backus-Naur Form (ABNF) 207 notation of [RFC5234], extended with the notation for case- 208 sensitivity in strings defined in [RFC7405]. 210 It also uses a list extension, defined in Section 11 of [Semantics], 211 that allows for compact definition of comma-separated lists using a 212 '#' operator (similar to how the '*' operator indicates repetition). 213 Appendix A shows the collected grammar with all list operators 214 expanded to standard ABNF notation. 216 As a convention, ABNF rule names prefixed with "obs-" denote 217 "obsolete" grammar rules that appear for historical reasons. 219 The following core rules are included by reference, as defined in 220 [RFC5234], Appendix B.1: ALPHA (letters), CR (carriage return), CRLF 221 (CR LF), CTL (controls), DIGIT (decimal 0-9), DQUOTE (double quote), 222 HEXDIG (hexadecimal 0-9/A-F/a-f), HTAB (horizontal tab), LF (line 223 feed), OCTET (any 8-bit sequence of data), SP (space), and VCHAR (any 224 visible [USASCII] character). 226 The rules below are defined in [Semantics]: 228 BWS = 229 OWS = 230 RWS = 231 absolute-URI = 232 absolute-path = 233 authority = 234 comment = 235 field-name = 236 field-value = 237 obs-text = 238 port = 239 query = 240 quoted-string = 241 token = 242 uri-host = 244 2. Message 246 2.1. Message Format 248 All HTTP/1.1 messages consist of a start-line followed by a sequence 249 of octets in a format similar to the Internet Message Format 250 [RFC5322]: zero or more header fields (collectively referred to as 251 the "headers" or the "header section"), an empty line indicating the 252 end of the header section, and an optional message body. 254 HTTP-message = start-line 255 *( header-field CRLF ) 256 CRLF 257 [ message-body ] 259 An HTTP message can be either a request from client to server or a 260 response from server to client. Syntactically, the two types of 261 message differ only in the start-line, which is either a request-line 262 (for requests) or a status-line (for responses), and in the algorithm 263 for determining the length of the message body (Section 6). 265 start-line = request-line / status-line 267 In theory, a client could receive requests and a server could receive 268 responses, distinguishing them by their different start-line formats. 269 In practice, servers are implemented to only expect a request (a 270 response is interpreted as an unknown or invalid request method) and 271 clients are implemented to only expect a response. 273 Although HTTP makes use of some protocol elements similar to the 274 Multipurpose Internet Mail Extensions (MIME) [RFC2045], see 275 Appendix B for the differences between HTTP and MIME messages. 277 2.2. HTTP Version 279 HTTP uses a "." numbering scheme to indicate versions 280 of the protocol. This specification defines version "1.1". 281 Section 3.5 of [Semantics] specifies the semantics of HTTP version 282 numbers. 284 The version of an HTTP/1.x message is indicated by an HTTP-version 285 field in the start-line. HTTP-version is case-sensitive. 287 HTTP-version = HTTP-name "/" DIGIT "." DIGIT 288 HTTP-name = %s"HTTP" 290 When an HTTP/1.1 message is sent to an HTTP/1.0 recipient [RFC1945] 291 or a recipient whose version is unknown, the HTTP/1.1 message is 292 constructed such that it can be interpreted as a valid HTTP/1.0 293 message if all of the newer features are ignored. This specification 294 places recipient-version requirements on some new features so that a 295 conformant sender will only use compatible features until it has 296 determined, through configuration or the receipt of a message, that 297 the recipient supports HTTP/1.1. 299 Intermediaries that process HTTP messages (i.e., all intermediaries 300 other than those acting as tunnels) MUST send their own HTTP-version 301 in forwarded messages. In other words, they are not allowed to 302 blindly forward the start-line without ensuring that the protocol 303 version in that message matches a version to which that intermediary 304 is conformant for both the receiving and sending of messages. 305 Forwarding an HTTP message without rewriting the HTTP-version might 306 result in communication errors when downstream recipients use the 307 message sender's version to determine what features are safe to use 308 for later communication with that sender. 310 A server MAY send an HTTP/1.0 response to an HTTP/1.1 request if it 311 is known or suspected that the client incorrectly implements the HTTP 312 specification and is incapable of correctly processing later version 313 responses, such as when a client fails to parse the version number 314 correctly or when an intermediary is known to blindly forward the 315 HTTP-version even when it doesn't conform to the given minor version 316 of the protocol. Such protocol downgrades SHOULD NOT be performed 317 unless triggered by specific client attributes, such as when one or 318 more of the request header fields (e.g., User-Agent) uniquely match 319 the values sent by a client known to be in error. 321 2.3. Message Parsing 323 The normal procedure for parsing an HTTP message is to read the 324 start-line into a structure, read each header field into a hash table 325 by field name until the empty line, and then use the parsed data to 326 determine if a message body is expected. If a message body has been 327 indicated, then it is read as a stream until an amount of octets 328 equal to the message body length is read or the connection is closed. 330 A recipient MUST parse an HTTP message as a sequence of octets in an 331 encoding that is a superset of US-ASCII [USASCII]. Parsing an HTTP 332 message as a stream of Unicode characters, without regard for the 333 specific encoding, creates security vulnerabilities due to the 334 varying ways that string processing libraries handle invalid 335 multibyte character sequences that contain the octet LF (%x0A). 336 String-based parsers can only be safely used within protocol elements 337 after the element has been extracted from the message, such as within 338 a header field-value after message parsing has delineated the 339 individual fields. 341 Although the line terminator for the start-line and header fields is 342 the sequence CRLF, a recipient MAY recognize a single LF as a line 343 terminator and ignore any preceding CR. 345 Older HTTP/1.0 user agent implementations might send an extra CRLF 346 after a POST request as a workaround for some early server 347 applications that failed to read message body content that was not 348 terminated by a line-ending. An HTTP/1.1 user agent MUST NOT preface 349 or follow a request with an extra CRLF. If terminating the request 350 message body with a line-ending is desired, then the user agent MUST 351 count the terminating CRLF octets as part of the message body length. 353 In the interest of robustness, a server that is expecting to receive 354 and parse a request-line SHOULD ignore at least one empty line (CRLF) 355 received prior to the request-line. 357 A sender MUST NOT send whitespace between the start-line and the 358 first header field. A recipient that receives whitespace between the 359 start-line and the first header field MUST either reject the message 360 as invalid or consume each whitespace-preceded line without further 361 processing of it (i.e., ignore the entire line, along with any 362 subsequent lines preceded by whitespace, until a properly formed 363 header field is received or the header section is terminated). 365 The presence of such whitespace in a request might be an attempt to 366 trick a server into ignoring that field or processing the line after 367 it as a new request, either of which might result in a security 368 vulnerability if other implementations within the request chain 369 interpret the same message differently. Likewise, the presence of 370 such whitespace in a response might be ignored by some clients or 371 cause others to cease parsing. 373 When a server listening only for HTTP request messages, or processing 374 what appears from the start-line to be an HTTP request message, 375 receives a sequence of octets that does not match the HTTP-message 376 grammar aside from the robustness exceptions listed above, the server 377 SHOULD respond with a 400 (Bad Request) response. 379 3. Request Line 381 A request-line begins with a method token, followed by a single space 382 (SP), the request-target, another single space (SP), the protocol 383 version, and ends with CRLF. 385 request-line = method SP request-target SP HTTP-version CRLF 387 Although the request-line grammar rule requires that each of the 388 component elements be separated by a single SP octet, recipients MAY 389 instead parse on whitespace-delimited word boundaries and, aside from 390 the CRLF terminator, treat any form of whitespace as the SP separator 391 while ignoring preceding or trailing whitespace; such whitespace 392 includes one or more of the following octets: SP, HTAB, VT (%x0B), FF 393 (%x0C), or bare CR. However, lenient parsing can result in request 394 smuggling security vulnerabilities if there are multiple recipients 395 of the message and each has its own unique interpretation of 396 robustness (see Section 11.2). 398 HTTP does not place a predefined limit on the length of a request- 399 line, as described in Section 3 of [Semantics]. A server that 400 receives a method longer than any that it implements SHOULD respond 401 with a 501 (Not Implemented) status code. A server that receives a 402 request-target longer than any URI it wishes to parse MUST respond 403 with a 414 (URI Too Long) status code (see Section 9.5.15 of 404 [Semantics]). 406 Various ad hoc limitations on request-line length are found in 407 practice. It is RECOMMENDED that all HTTP senders and recipients 408 support, at a minimum, request-line lengths of 8000 octets. 410 3.1. Method 412 The method token indicates the request method to be performed on the 413 target resource. The request method is case-sensitive. 415 method = token 417 The request methods defined by this specification can be found in 418 Section 7 of [Semantics], along with information regarding the HTTP 419 method registry and considerations for defining new methods. 421 3.2. Request Target 423 The request-target identifies the target resource upon which to apply 424 the request. The client derives a request-target from its desired 425 target URI. There are four distinct formats for the request-target, 426 depending on both the method being requested and whether the request 427 is to a proxy. 429 request-target = origin-form 430 / absolute-form 431 / authority-form 432 / asterisk-form 434 No whitespace is allowed in the request-target. Unfortunately, some 435 user agents fail to properly encode or exclude whitespace found in 436 hypertext references, resulting in those disallowed characters being 437 sent as the request-target in a malformed request-line. 439 Recipients of an invalid request-line SHOULD respond with either a 440 400 (Bad Request) error or a 301 (Moved Permanently) redirect with 441 the request-target properly encoded. A recipient SHOULD NOT attempt 442 to autocorrect and then process the request without a redirect, since 443 the invalid request-line might be deliberately crafted to bypass 444 security filters along the request chain. 446 3.2.1. origin-form 448 The most common form of request-target is the origin-form. 450 origin-form = absolute-path [ "?" query ] 452 When making a request directly to an origin server, other than a 453 CONNECT or server-wide OPTIONS request (as detailed below), a client 454 MUST send only the absolute path and query components of the target 455 URI as the request-target. If the target URI's path component is 456 empty, the client MUST send "/" as the path within the origin-form of 457 request-target. A Host header field is also sent, as defined in 458 Section 5.4 of [Semantics]. 460 For example, a client wishing to retrieve a representation of the 461 resource identified as 463 http://www.example.org/where?q=now 465 directly from the origin server would open (or reuse) a TCP 466 connection to port 80 of the host "www.example.org" and send the 467 lines: 469 GET /where?q=now HTTP/1.1 470 Host: www.example.org 472 followed by the remainder of the request message. 474 3.2.2. absolute-form 476 When making a request to a proxy, other than a CONNECT or server-wide 477 OPTIONS request (as detailed below), a client MUST send the target 478 URI in absolute-form as the request-target. 480 absolute-form = absolute-URI 482 The proxy is requested to either service that request from a valid 483 cache, if possible, or make the same request on the client's behalf 484 to either the next inbound proxy server or directly to the origin 485 server indicated by the request-target. Requirements on such 486 "forwarding" of messages are defined in Section 5.5 of [Semantics]. 488 An example absolute-form of request-line would be: 490 GET http://www.example.org/pub/WWW/TheProject.html HTTP/1.1 492 To allow for transition to the absolute-form for all requests in some 493 future version of HTTP, a server MUST accept the absolute-form in 494 requests, even though HTTP/1.1 clients will only send them in 495 requests to proxies. 497 3.2.3. authority-form 499 The authority-form of request-target is only used for CONNECT 500 requests (Section 7.3.6 of [Semantics]). 502 authority-form = authority 504 When making a CONNECT request to establish a tunnel through one or 505 more proxies, a client MUST send only the target URI's authority 506 component (excluding any userinfo and its "@" delimiter) as the 507 request-target. For example, 509 CONNECT www.example.com:80 HTTP/1.1 511 3.2.4. asterisk-form 513 The asterisk-form of request-target is only used for a server-wide 514 OPTIONS request (Section 7.3.7 of [Semantics]). 516 asterisk-form = "*" 518 When a client wishes to request OPTIONS for the server as a whole, as 519 opposed to a specific named resource of that server, the client MUST 520 send only "*" (%x2A) as the request-target. For example, 522 OPTIONS * HTTP/1.1 524 If a proxy receives an OPTIONS request with an absolute-form of 525 request-target in which the URI has an empty path and no query 526 component, then the last proxy on the request chain MUST send a 527 request-target of "*" when it forwards the request to the indicated 528 origin server. 530 For example, the request 531 OPTIONS http://www.example.org:8001 HTTP/1.1 533 would be forwarded by the final proxy as 535 OPTIONS * HTTP/1.1 536 Host: www.example.org:8001 538 after connecting to port 8001 of host "www.example.org". 540 3.3. Effective Request URI 542 Since the request-target often contains only part of the user agent's 543 target URI, a server reconstructs the intended target as an effective 544 request URI to properly service the request (Section 5.3 of 545 [Semantics]). 547 If the request-target is in absolute-form, the effective request URI 548 is the same as the request-target. Otherwise, the effective request 549 URI is constructed as follows: 551 If the server's configuration (or outbound gateway) provides a 552 fixed URI scheme, that scheme is used for the effective request 553 URI. Otherwise, if the request is received over a TLS-secured TCP 554 connection, the effective request URI's scheme is "https"; if not, 555 the scheme is "http". 557 If the server's configuration (or outbound gateway) provides a 558 fixed URI authority component, that authority is used for the 559 effective request URI. If not, then if the request-target is in 560 authority-form, the effective request URI's authority component is 561 the same as the request-target. If not, then if a Host header 562 field is supplied with a non-empty field-value, the authority 563 component is the same as the Host field-value. Otherwise, the 564 authority component is assigned the default name configured for 565 the server and, if the connection's incoming TCP port number 566 differs from the default port for the effective request URI's 567 scheme, then a colon (":") and the incoming port number (in 568 decimal form) are appended to the authority component. 570 If the request-target is in authority-form or asterisk-form, the 571 effective request URI's combined path and query component is 572 empty. Otherwise, the combined path and query component is the 573 same as the request-target. 575 The components of the effective request URI, once determined as 576 above, can be combined into absolute-URI form by concatenating the 577 scheme, "://", authority, and combined path and query component. 579 Example 1: the following message received over an insecure TCP 580 connection 582 GET /pub/WWW/TheProject.html HTTP/1.1 583 Host: www.example.org:8080 585 has an effective request URI of 587 http://www.example.org:8080/pub/WWW/TheProject.html 589 Example 2: the following message received over a TLS-secured TCP 590 connection 592 OPTIONS * HTTP/1.1 593 Host: www.example.org 595 has an effective request URI of 597 https://www.example.org 599 Recipients of an HTTP/1.0 request that lacks a Host header field 600 might need to use heuristics (e.g., examination of the URI path for 601 something unique to a particular host) in order to guess the 602 effective request URI's authority component. 604 4. Status Line 606 The first line of a response message is the status-line, consisting 607 of the protocol version, a space (SP), the status code, another 608 space, a possibly empty textual phrase describing the status code, 609 and ending with CRLF. 611 status-line = HTTP-version SP status-code SP reason-phrase CRLF 613 Although the status-line grammar rule requires that each of the 614 component elements be separated by a single SP octet, recipients MAY 615 instead parse on whitespace-delimited word boundaries and, aside from 616 the line terminator, treat any form of whitespace as the SP separator 617 while ignoring preceding or trailing whitespace; such whitespace 618 includes one or more of the following octets: SP, HTAB, VT (%x0B), FF 619 (%x0C), or bare CR. However, lenient parsing can result in response 620 splitting security vulnerabilities if there are multiple recipients 621 of the message and each has its own unique interpretation of 622 robustness (see Section 11.1). 624 The status-code element is a 3-digit integer code describing the 625 result of the server's attempt to understand and satisfy the client's 626 corresponding request. The rest of the response message is to be 627 interpreted in light of the semantics defined for that status code. 628 See Section 9 of [Semantics] for information about the semantics of 629 status codes, including the classes of status code (indicated by the 630 first digit), the status codes defined by this specification, 631 considerations for the definition of new status codes, and the IANA 632 registry. 634 status-code = 3DIGIT 636 The reason-phrase element exists for the sole purpose of providing a 637 textual description associated with the numeric status code, mostly 638 out of deference to earlier Internet application protocols that were 639 more frequently used with interactive text clients. 641 A client SHOULD ignore the reason-phrase content because it is not a 642 reliable channel for information (it might be discarded or 643 overwritten by intermediaries, and it is not transmitted in other 644 versions of HTTP). 646 reason-phrase = *( HTAB / SP / VCHAR / obs-text ) 648 5. Header Fields 650 Each header field consists of a case-insensitive field name followed 651 by a colon (":"), optional leading whitespace, the field value, and 652 optional trailing whitespace. 654 header-field = field-name ":" OWS field-value OWS 656 Most HTTP field names and the rules for parsing within field values 657 are defined in Section 4 of [Semantics]. This section covers the 658 generic syntax for header field inclusion within, and extraction 659 from, HTTP/1.1 messages. In addition, the following header fields 660 are defined by this document because they are specific to HTTP/1.1 661 message processing: 663 +-------------------+----------+---------------+ 664 | Header Field Name | Status | Reference | 665 +-------------------+----------+---------------+ 666 | Connection | standard | Section 9.1 | 667 | MIME-Version | standard | Appendix B.1 | 668 | TE | standard | Section 7.4 | 669 | Transfer-Encoding | standard | Section 6.1 | 670 | Upgrade | standard | Section 9.8 | 671 +-------------------+----------+---------------+ 672 Furthermore, the field name "Close" is reserved, since using that 673 name as an HTTP header field might conflict with the "close" 674 connection option of the Connection header field (Section 9.1). 676 +-------------------+----------+----------+------------+ 677 | Header Field Name | Protocol | Status | Reference | 678 +-------------------+----------+----------+------------+ 679 | Close | http | reserved | Section 5 | 680 +-------------------+----------+----------+------------+ 682 5.1. Header Field Parsing 684 Messages are parsed using a generic algorithm, independent of the 685 individual header field names. The contents within a given field 686 value are not parsed until a later stage of message interpretation 687 (usually after the message's entire header section has been 688 processed). 690 No whitespace is allowed between the header field-name and colon. In 691 the past, differences in the handling of such whitespace have led to 692 security vulnerabilities in request routing and response handling. A 693 server MUST reject any received request message that contains 694 whitespace between a header field-name and colon with a response 695 status code of 400 (Bad Request). A proxy MUST remove any such 696 whitespace from a response message before forwarding the message 697 downstream. 699 A field value might be preceded and/or followed by optional 700 whitespace (OWS); a single SP preceding the field-value is preferred 701 for consistent readability by humans. The field value does not 702 include any leading or trailing whitespace: OWS occurring before the 703 first non-whitespace octet of the field value or after the last non- 704 whitespace octet of the field value ought to be excluded by parsers 705 when extracting the field value from a header field. 707 5.2. Obsolete Line Folding 709 Historically, HTTP header field values could be extended over 710 multiple lines by preceding each extra line with at least one space 711 or horizontal tab (obs-fold). This specification deprecates such 712 line folding except within the message/http media type 713 (Section 10.1). 715 obs-fold = CRLF 1*( SP / HTAB ) 716 ; obsolete line folding 718 A sender MUST NOT generate a message that includes line folding 719 (i.e., that has any field-value that contains a match to the obs-fold 720 rule) unless the message is intended for packaging within the 721 message/http media type. 723 A server that receives an obs-fold in a request message that is not 724 within a message/http container MUST either reject the message by 725 sending a 400 (Bad Request), preferably with a representation 726 explaining that obsolete line folding is unacceptable, or replace 727 each received obs-fold with one or more SP octets prior to 728 interpreting the field value or forwarding the message downstream. 730 A proxy or gateway that receives an obs-fold in a response message 731 that is not within a message/http container MUST either discard the 732 message and replace it with a 502 (Bad Gateway) response, preferably 733 with a representation explaining that unacceptable line folding was 734 received, or replace each received obs-fold with one or more SP 735 octets prior to interpreting the field value or forwarding the 736 message downstream. 738 A user agent that receives an obs-fold in a response message that is 739 not within a message/http container MUST replace each received obs- 740 fold with one or more SP octets prior to interpreting the field 741 value. 743 6. Message Body 745 The message body (if any) of an HTTP message is used to carry the 746 payload body of that request or response. The message body is 747 identical to the payload body unless a transfer coding has been 748 applied, as described in Section 6.1. 750 message-body = *OCTET 752 The rules for when a message body is allowed in a message differ for 753 requests and responses. 755 The presence of a message body in a request is signaled by a Content- 756 Length or Transfer-Encoding header field. Request message framing is 757 independent of method semantics, even if the method does not define 758 any use for a message body. 760 The presence of a message body in a response depends on both the 761 request method to which it is responding and the response status code 762 (Section 4). Responses to the HEAD request method (Section 7.3.2 of 763 [Semantics]) never include a message body because the associated 764 response header fields (e.g., Transfer-Encoding, Content-Length, 765 etc.), if present, indicate only what their values would have been if 766 the request method had been GET (Section 7.3.1 of [Semantics]). 2xx 767 (Successful) responses to a CONNECT request method (Section 7.3.6 of 769 [Semantics]) switch to tunnel mode instead of having a message body. 770 All 1xx (Informational), 204 (No Content), and 304 (Not Modified) 771 responses do not include a message body. All other responses do 772 include a message body, although the body might be of zero length. 774 6.1. Transfer-Encoding 776 The Transfer-Encoding header field lists the transfer coding names 777 corresponding to the sequence of transfer codings that have been (or 778 will be) applied to the payload body in order to form the message 779 body. Transfer codings are defined in Section 7. 781 Transfer-Encoding = 1#transfer-coding 783 Transfer-Encoding is analogous to the Content-Transfer-Encoding field 784 of MIME, which was designed to enable safe transport of binary data 785 over a 7-bit transport service ([RFC2045], Section 6). However, safe 786 transport has a different focus for an 8bit-clean transfer protocol. 787 In HTTP's case, Transfer-Encoding is primarily intended to accurately 788 delimit a dynamically generated payload and to distinguish payload 789 encodings that are only applied for transport efficiency or security 790 from those that are characteristics of the selected resource. 792 A recipient MUST be able to parse the chunked transfer coding 793 (Section 7.1) because it plays a crucial role in framing messages 794 when the payload body size is not known in advance. A sender MUST 795 NOT apply chunked more than once to a message body (i.e., chunking an 796 already chunked message is not allowed). If any transfer coding 797 other than chunked is applied to a request payload body, the sender 798 MUST apply chunked as the final transfer coding to ensure that the 799 message is properly framed. If any transfer coding other than 800 chunked is applied to a response payload body, the sender MUST either 801 apply chunked as the final transfer coding or terminate the message 802 by closing the connection. 804 For example, 806 Transfer-Encoding: gzip, chunked 808 indicates that the payload body has been compressed using the gzip 809 coding and then chunked using the chunked coding while forming the 810 message body. 812 Unlike Content-Encoding (Section 6.1.2 of [Semantics]), Transfer- 813 Encoding is a property of the message, not of the representation, and 814 any recipient along the request/response chain MAY decode the 815 received transfer coding(s) or apply additional transfer coding(s) to 816 the message body, assuming that corresponding changes are made to the 817 Transfer-Encoding field-value. Additional information about the 818 encoding parameters can be provided by other header fields not 819 defined by this specification. 821 Transfer-Encoding MAY be sent in a response to a HEAD request or in a 822 304 (Not Modified) response (Section 9.4.5 of [Semantics]) to a GET 823 request, neither of which includes a message body, to indicate that 824 the origin server would have applied a transfer coding to the message 825 body if the request had been an unconditional GET. This indication 826 is not required, however, because any recipient on the response chain 827 (including the origin server) can remove transfer codings when they 828 are not needed. 830 A server MUST NOT send a Transfer-Encoding header field in any 831 response with a status code of 1xx (Informational) or 204 (No 832 Content). A server MUST NOT send a Transfer-Encoding header field in 833 any 2xx (Successful) response to a CONNECT request (Section 7.3.6 of 834 [Semantics]). 836 Transfer-Encoding was added in HTTP/1.1. It is generally assumed 837 that implementations advertising only HTTP/1.0 support will not 838 understand how to process a transfer-encoded payload. A client MUST 839 NOT send a request containing Transfer-Encoding unless it knows the 840 server will handle HTTP/1.1 (or later) requests; such knowledge might 841 be in the form of specific user configuration or by remembering the 842 version of a prior received response. A server MUST NOT send a 843 response containing Transfer-Encoding unless the corresponding 844 request indicates HTTP/1.1 (or later). 846 A server that receives a request message with a transfer coding it 847 does not understand SHOULD respond with 501 (Not Implemented). 849 6.2. Content-Length 851 When a message does not have a Transfer-Encoding header field, a 852 Content-Length header field can provide the anticipated size, as a 853 decimal number of octets, for a potential payload body. For messages 854 that do include a payload body, the Content-Length field-value 855 provides the framing information necessary for determining where the 856 body (and message) ends. For messages that do not include a payload 857 body, the Content-Length indicates the size of the selected 858 representation (Section 6.2.4 of [Semantics]). 860 Note: HTTP's use of Content-Length for message framing differs 861 significantly from the same field's use in MIME, where it is an 862 optional field used only within the "message/external-body" media- 863 type. 865 6.3. Message Body Length 867 The length of a message body is determined by one of the following 868 (in order of precedence): 870 1. Any response to a HEAD request and any response with a 1xx 871 (Informational), 204 (No Content), or 304 (Not Modified) status 872 code is always terminated by the first empty line after the 873 header fields, regardless of the header fields present in the 874 message, and thus cannot contain a message body. 876 2. Any 2xx (Successful) response to a CONNECT request implies that 877 the connection will become a tunnel immediately after the empty 878 line that concludes the header fields. A client MUST ignore any 879 Content-Length or Transfer-Encoding header fields received in 880 such a message. 882 3. If a Transfer-Encoding header field is present and the chunked 883 transfer coding (Section 7.1) is the final encoding, the message 884 body length is determined by reading and decoding the chunked 885 data until the transfer coding indicates the data is complete. 887 If a Transfer-Encoding header field is present in a response and 888 the chunked transfer coding is not the final encoding, the 889 message body length is determined by reading the connection until 890 it is closed by the server. If a Transfer-Encoding header field 891 is present in a request and the chunked transfer coding is not 892 the final encoding, the message body length cannot be determined 893 reliably; the server MUST respond with the 400 (Bad Request) 894 status code and then close the connection. 896 If a message is received with both a Transfer-Encoding and a 897 Content-Length header field, the Transfer-Encoding overrides the 898 Content-Length. Such a message might indicate an attempt to 899 perform request smuggling (Section 11.2) or response splitting 900 (Section 11.1) and ought to be handled as an error. A sender 901 MUST remove the received Content-Length field prior to forwarding 902 such a message downstream. 904 4. If a message is received without Transfer-Encoding and with 905 either multiple Content-Length header fields having differing 906 field-values or a single Content-Length header field having an 907 invalid value, then the message framing is invalid and the 908 recipient MUST treat it as an unrecoverable error. If this is a 909 request message, the server MUST respond with a 400 (Bad Request) 910 status code and then close the connection. If this is a response 911 message received by a proxy, the proxy MUST close the connection 912 to the server, discard the received response, and send a 502 (Bad 913 Gateway) response to the client. If this is a response message 914 received by a user agent, the user agent MUST close the 915 connection to the server and discard the received response. 917 5. If a valid Content-Length header field is present without 918 Transfer-Encoding, its decimal value defines the expected message 919 body length in octets. If the sender closes the connection or 920 the recipient times out before the indicated number of octets are 921 received, the recipient MUST consider the message to be 922 incomplete and close the connection. 924 6. If this is a request message and none of the above are true, then 925 the message body length is zero (no message body is present). 927 7. Otherwise, this is a response message without a declared message 928 body length, so the message body length is determined by the 929 number of octets received prior to the server closing the 930 connection. 932 Since there is no way to distinguish a successfully completed, close- 933 delimited message from a partially received message interrupted by 934 network failure, a server SHOULD generate encoding or length- 935 delimited messages whenever possible. The close-delimiting feature 936 exists primarily for backwards compatibility with HTTP/1.0. 938 A server MAY reject a request that contains a message body but not a 939 Content-Length by responding with 411 (Length Required). 941 Unless a transfer coding other than chunked has been applied, a 942 client that sends a request containing a message body SHOULD use a 943 valid Content-Length header field if the message body length is known 944 in advance, rather than the chunked transfer coding, since some 945 existing services respond to chunked with a 411 (Length Required) 946 status code even though they understand the chunked transfer coding. 947 This is typically because such services are implemented via a gateway 948 that requires a content-length in advance of being called and the 949 server is unable or unwilling to buffer the entire request before 950 processing. 952 A user agent that sends a request containing a message body MUST send 953 a valid Content-Length header field if it does not know the server 954 will handle HTTP/1.1 (or later) requests; such knowledge can be in 955 the form of specific user configuration or by remembering the version 956 of a prior received response. 958 If the final response to the last request on a connection has been 959 completely received and there remains additional data to read, a user 960 agent MAY discard the remaining data or attempt to determine if that 961 data belongs as part of the prior response body, which might be the 962 case if the prior message's Content-Length value is incorrect. A 963 client MUST NOT process, cache, or forward such extra data as a 964 separate response, since such behavior would be vulnerable to cache 965 poisoning. 967 7. Transfer Codings 969 Transfer coding names are used to indicate an encoding transformation 970 that has been, can be, or might need to be applied to a payload body 971 in order to ensure "safe transport" through the network. This 972 differs from a content coding in that the transfer coding is a 973 property of the message rather than a property of the representation 974 that is being transferred. 976 transfer-coding = token *( OWS ";" OWS transfer-parameter ) 978 Parameters are in the form of a name=value pair. 980 transfer-parameter = token BWS "=" BWS ( token / quoted-string ) 982 All transfer-coding names are case-insensitive and ought to be 983 registered within the HTTP Transfer Coding registry, as defined in 984 Section 7.3. They are used in the TE (Section 7.4) and Transfer- 985 Encoding (Section 6.1) header fields. 987 +------------+------------------------------------------+-----------+ 988 | Name | Description | Reference | 989 +------------+------------------------------------------+-----------+ 990 | chunked | Transfer in a series of chunks | Section | 991 | | | 7.1 | 992 | compress | UNIX "compress" data format [Welch] | Section | 993 | | | 7.2 | 994 | deflate | "deflate" compressed data ([RFC1951]) | Section | 995 | | inside the "zlib" data format | 7.2 | 996 | | ([RFC1950]) | | 997 | gzip | GZIP file format [RFC1952] | Section | 998 | | | 7.2 | 999 | trailers | (reserved) | Section 7 | 1000 | x-compress | Deprecated (alias for compress) | Section | 1001 | | | 7.2 | 1002 | x-gzip | Deprecated (alias for gzip) | Section | 1003 | | | 7.2 | 1004 +------------+------------------------------------------+-----------+ 1006 Note: the coding name "trailers" is reserved because it would 1007 clash with the use of the keyword "trailers" in the TE header 1008 field (Section 7.4). 1010 7.1. Chunked Transfer Coding 1012 The chunked transfer coding wraps the payload body in order to 1013 transfer it as a series of chunks, each with its own size indicator, 1014 followed by an OPTIONAL trailer containing header fields. Chunked 1015 enables content streams of unknown size to be transferred as a 1016 sequence of length-delimited buffers, which enables the sender to 1017 retain connection persistence and the recipient to know when it has 1018 received the entire message. 1020 chunked-body = *chunk 1021 last-chunk 1022 trailer-part 1023 CRLF 1025 chunk = chunk-size [ chunk-ext ] CRLF 1026 chunk-data CRLF 1027 chunk-size = 1*HEXDIG 1028 last-chunk = 1*("0") [ chunk-ext ] CRLF 1030 chunk-data = 1*OCTET ; a sequence of chunk-size octets 1032 The chunk-size field is a string of hex digits indicating the size of 1033 the chunk-data in octets. The chunked transfer coding is complete 1034 when a chunk with a chunk-size of zero is received, possibly followed 1035 by a trailer, and finally terminated by an empty line. 1037 A recipient MUST be able to parse and decode the chunked transfer 1038 coding. 1040 The chunked encoding does not define any parameters. Their presence 1041 SHOULD be treated as an error. 1043 7.1.1. Chunk Extensions 1045 The chunked encoding allows each chunk to include zero or more chunk 1046 extensions, immediately following the chunk-size, for the sake of 1047 supplying per-chunk metadata (such as a signature or hash), mid- 1048 message control information, or randomization of message body size. 1050 chunk-ext = *( BWS ";" BWS chunk-ext-name 1051 [ BWS "=" BWS chunk-ext-val ] ) 1053 chunk-ext-name = token 1054 chunk-ext-val = token / quoted-string 1056 The chunked encoding is specific to each connection and is likely to 1057 be removed or recoded by each recipient (including intermediaries) 1058 before any higher-level application would have a chance to inspect 1059 the extensions. Hence, use of chunk extensions is generally limited 1060 to specialized HTTP services such as "long polling" (where client and 1061 server can have shared expectations regarding the use of chunk 1062 extensions) or for padding within an end-to-end secured connection. 1064 A recipient MUST ignore unrecognized chunk extensions. A server 1065 ought to limit the total length of chunk extensions received in a 1066 request to an amount reasonable for the services provided, in the 1067 same way that it applies length limitations and timeouts for other 1068 parts of a message, and generate an appropriate 4xx (Client Error) 1069 response if that amount is exceeded. 1071 7.1.2. Chunked Trailer Part 1073 A trailer allows the sender to include additional fields at the end 1074 of a chunked message in order to supply metadata that might be 1075 dynamically generated while the message body is sent, such as a 1076 message integrity check, digital signature, or post-processing 1077 status. The trailer fields are identical to header fields, except 1078 they are sent in a chunked trailer instead of the message's header 1079 section. 1081 trailer-part = *( header-field CRLF ) 1083 A sender MUST NOT generate a trailer that contains a field necessary 1084 for message framing (e.g., Transfer-Encoding and Content-Length), 1085 routing (e.g., Host), request modifiers (e.g., controls and 1086 conditionals in Section 8 of [Semantics]), authentication (e.g., see 1087 Section 8.5 of [Semantics] and [RFC6265]), response control data 1088 (e.g., see Section 10.1 of [Semantics]), or determining how to 1089 process the payload (e.g., Content-Encoding, Content-Type, Content- 1090 Range, and Trailer). 1092 When a chunked message containing a non-empty trailer is received, 1093 the recipient MAY process the fields (aside from those forbidden 1094 above) as if they were appended to the message's header section. A 1095 recipient MUST ignore (or consider as an error) any fields that are 1096 forbidden to be sent in a trailer, since processing them as if they 1097 were present in the header section might bypass external security 1098 filters. 1100 Unless the request includes a TE header field indicating "trailers" 1101 is acceptable, as described in Section 7.4, a server SHOULD NOT 1102 generate trailer fields that it believes are necessary for the user 1103 agent to receive. Without a TE containing "trailers", the server 1104 ought to assume that the trailer fields might be silently discarded 1105 along the path to the user agent. This requirement allows 1106 intermediaries to forward a de-chunked message to an HTTP/1.0 1107 recipient without buffering the entire response. 1109 When a message includes a message body encoded with the chunked 1110 transfer coding and the sender desires to send metadata in the form 1111 of trailer fields at the end of the message, the sender SHOULD 1112 generate a Trailer header field before the message body to indicate 1113 which fields will be present in the trailers. This allows the 1114 recipient to prepare for receipt of that metadata before it starts 1115 processing the body, which is useful if the message is being streamed 1116 and the recipient wishes to confirm an integrity check on the fly. 1118 7.1.3. Decoding Chunked 1120 A process for decoding the chunked transfer coding can be represented 1121 in pseudo-code as: 1123 length := 0 1124 read chunk-size, chunk-ext (if any), and CRLF 1125 while (chunk-size > 0) { 1126 read chunk-data and CRLF 1127 append chunk-data to decoded-body 1128 length := length + chunk-size 1129 read chunk-size, chunk-ext (if any), and CRLF 1130 } 1131 read trailer field 1132 while (trailer field is not empty) { 1133 if (trailer field is allowed to be sent in a trailer) { 1134 append trailer field to existing header fields 1135 } 1136 read trailer-field 1137 } 1138 Content-Length := length 1139 Remove "chunked" from Transfer-Encoding 1140 Remove Trailer from existing header fields 1142 7.2. Transfer Codings for Compression 1144 The following transfer coding names for compression are defined by 1145 the same algorithm as their corresponding content coding: 1147 compress (and x-compress) 1148 See Section 6.1.2.1 of [Semantics]. 1150 deflate 1151 See Section 6.1.2.2 of [Semantics]. 1153 gzip (and x-gzip) 1154 See Section 6.1.2.3 of [Semantics]. 1156 The compression codings do not define any parameters. Their presence 1157 SHOULD be treated as an error. 1159 7.3. Transfer Coding Registry 1161 The "HTTP Transfer Coding Registry" defines the namespace for 1162 transfer coding names. It is maintained at 1163 . 1165 Registrations MUST include the following fields: 1167 o Name 1169 o Description 1171 o Pointer to specification text 1173 Names of transfer codings MUST NOT overlap with names of content 1174 codings (Section 6.1.2 of [Semantics]) unless the encoding 1175 transformation is identical, as is the case for the compression 1176 codings defined in Section 7.2. 1178 The TE header field (Section 7.4) uses a pseudo parameter named "q" 1179 as rank value when multiple transfer codings are acceptable. Future 1180 registrations of transfer codings SHOULD NOT define parameters called 1181 "q" (case-insensitively) in order to avoid ambiguities. 1183 Values to be added to this namespace require IETF Review (see 1184 Section 4.8 of [RFC8126]), and MUST conform to the purpose of 1185 transfer coding defined in this specification. 1187 Use of program names for the identification of encoding formats is 1188 not desirable and is discouraged for future encodings. 1190 7.4. TE 1192 The "TE" header field in a request indicates what transfer codings, 1193 besides chunked, the client is willing to accept in response, and 1194 whether or not the client is willing to accept trailer fields in a 1195 chunked transfer coding. 1197 The TE field-value consists of a comma-separated list of transfer 1198 coding names, each allowing for optional parameters (as described in 1199 Section 7), and/or the keyword "trailers". A client MUST NOT send 1200 the chunked transfer coding name in TE; chunked is always acceptable 1201 for HTTP/1.1 recipients. 1203 TE = #t-codings 1204 t-codings = "trailers" / ( transfer-coding [ t-ranking ] ) 1205 t-ranking = OWS ";" OWS "q=" rank 1206 rank = ( "0" [ "." 0*3DIGIT ] ) 1207 / ( "1" [ "." 0*3("0") ] ) 1209 Three examples of TE use are below. 1211 TE: deflate 1212 TE: 1213 TE: trailers, deflate;q=0.5 1215 The presence of the keyword "trailers" indicates that the client is 1216 willing to accept trailer fields in a chunked transfer coding, as 1217 defined in Section 7.1.2, on behalf of itself and any downstream 1218 clients. For requests from an intermediary, this implies that 1219 either: (a) all downstream clients are willing to accept trailer 1220 fields in the forwarded response; or, (b) the intermediary will 1221 attempt to buffer the response on behalf of downstream recipients. 1222 Note that HTTP/1.1 does not define any means to limit the size of a 1223 chunked response such that an intermediary can be assured of 1224 buffering the entire response. 1226 When multiple transfer codings are acceptable, the client MAY rank 1227 the codings by preference using a case-insensitive "q" parameter 1228 (similar to the qvalues used in content negotiation fields, 1229 Section 8.4.1 of [Semantics]). The rank value is a real number in 1230 the range 0 through 1, where 0.001 is the least preferred and 1 is 1231 the most preferred; a value of 0 means "not acceptable". 1233 If the TE field-value is empty or if no TE field is present, the only 1234 acceptable transfer coding is chunked. A message with no transfer 1235 coding is always acceptable. 1237 Since the TE header field only applies to the immediate connection, a 1238 sender of TE MUST also send a "TE" connection option within the 1239 Connection header field (Section 9.1) in order to prevent the TE 1240 field from being forwarded by intermediaries that do not support its 1241 semantics. 1243 8. Handling Incomplete Messages 1245 A server that receives an incomplete request message, usually due to 1246 a canceled request or a triggered timeout exception, MAY send an 1247 error response prior to closing the connection. 1249 A client that receives an incomplete response message, which can 1250 occur when a connection is closed prematurely or when decoding a 1251 supposedly chunked transfer coding fails, MUST record the message as 1252 incomplete. Cache requirements for incomplete responses are defined 1253 in Section 3 of [Caching]. 1255 If a response terminates in the middle of the header section (before 1256 the empty line is received) and the status code might rely on header 1257 fields to convey the full meaning of the response, then the client 1258 cannot assume that meaning has been conveyed; the client might need 1259 to repeat the request in order to determine what action to take next. 1261 A message body that uses the chunked transfer coding is incomplete if 1262 the zero-sized chunk that terminates the encoding has not been 1263 received. A message that uses a valid Content-Length is incomplete 1264 if the size of the message body received (in octets) is less than the 1265 value given by Content-Length. A response that has neither chunked 1266 transfer coding nor Content-Length is terminated by closure of the 1267 connection and, thus, is considered complete regardless of the number 1268 of message body octets received, provided that the header section was 1269 received intact. 1271 9. Connection Management 1273 HTTP messaging is independent of the underlying transport- or 1274 session-layer connection protocol(s). HTTP only presumes a reliable 1275 transport with in-order delivery of requests and the corresponding 1276 in-order delivery of responses. The mapping of HTTP request and 1277 response structures onto the data units of an underlying transport 1278 protocol is outside the scope of this specification. 1280 As described in Section 5.2 of [Semantics], the specific connection 1281 protocols to be used for an HTTP interaction are determined by client 1282 configuration and the target URI. For example, the "http" URI scheme 1283 (Section 2.5.1 of [Semantics]) indicates a default connection of TCP 1284 over IP, with a default TCP port of 80, but the client might be 1285 configured to use a proxy via some other connection, port, or 1286 protocol. 1288 HTTP implementations are expected to engage in connection management, 1289 which includes maintaining the state of current connections, 1290 establishing a new connection or reusing an existing connection, 1291 processing messages received on a connection, detecting connection 1292 failures, and closing each connection. Most clients maintain 1293 multiple connections in parallel, including more than one connection 1294 per server endpoint. Most servers are designed to maintain thousands 1295 of concurrent connections, while controlling request queues to enable 1296 fair use and detect denial-of-service attacks. 1298 9.1. Connection 1300 The "Connection" header field allows the sender to indicate desired 1301 control options for the current connection. In order to avoid 1302 confusing downstream recipients, a proxy or gateway MUST remove or 1303 replace any received connection options before forwarding the 1304 message. 1306 When a header field aside from Connection is used to supply control 1307 information for or about the current connection, the sender MUST list 1308 the corresponding field-name within the Connection header field. A 1309 proxy or gateway MUST parse a received Connection header field before 1310 a message is forwarded and, for each connection-option in this field, 1311 remove any header field(s) from the message with the same name as the 1312 connection-option, and then remove the Connection header field itself 1313 (or replace it with the intermediary's own connection options for the 1314 forwarded message). 1316 Hence, the Connection header field provides a declarative way of 1317 distinguishing header fields that are only intended for the immediate 1318 recipient ("hop-by-hop") from those fields that are intended for all 1319 recipients on the chain ("end-to-end"), enabling the message to be 1320 self-descriptive and allowing future connection-specific extensions 1321 to be deployed without fear that they will be blindly forwarded by 1322 older intermediaries. 1324 The Connection header field's value has the following grammar: 1326 Connection = 1#connection-option 1327 connection-option = token 1329 Connection options are case-insensitive. 1331 A sender MUST NOT send a connection option corresponding to a header 1332 field that is intended for all recipients of the payload. For 1333 example, Cache-Control is never appropriate as a connection option 1334 (Section 5.2 of [Caching]). 1336 The connection options do not always correspond to a header field 1337 present in the message, since a connection-specific header field 1338 might not be needed if there are no parameters associated with a 1339 connection option. In contrast, a connection-specific header field 1340 that is received without a corresponding connection option usually 1341 indicates that the field has been improperly forwarded by an 1342 intermediary and ought to be ignored by the recipient. 1344 When defining new connection options, specification authors ought to 1345 survey existing header field names and ensure that the new connection 1346 option does not share the same name as an already deployed header 1347 field. Defining a new connection option essentially reserves that 1348 potential field-name for carrying additional information related to 1349 the connection option, since it would be unwise for senders to use 1350 that field-name for anything else. 1352 The "close" connection option is defined for a sender to signal that 1353 this connection will be closed after completion of the response. For 1354 example, 1356 Connection: close 1358 in either the request or the response header fields indicates that 1359 the sender is going to close the connection after the current 1360 request/response is complete (Section 9.7). 1362 A client that does not support persistent connections MUST send the 1363 "close" connection option in every request message. 1365 A server that does not support persistent connections MUST send the 1366 "close" connection option in every response message that does not 1367 have a 1xx (Informational) status code. 1369 9.2. Establishment 1371 It is beyond the scope of this specification to describe how 1372 connections are established via various transport- or session-layer 1373 protocols. Each connection applies to only one transport link. 1375 9.3. Associating a Response to a Request 1377 HTTP/1.1 does not include a request identifier for associating a 1378 given request message with its corresponding one or more response 1379 messages. Hence, it relies on the order of response arrival to 1380 correspond exactly to the order in which requests are made on the 1381 same connection. More than one response message per request only 1382 occurs when one or more informational responses (1xx, see Section 9.2 1383 of [Semantics]) precede a final response to the same request. 1385 A client that has more than one outstanding request on a connection 1386 MUST maintain a list of outstanding requests in the order sent and 1387 MUST associate each received response message on that connection to 1388 the highest ordered request that has not yet received a final (non- 1389 1xx) response. 1391 If an HTTP/1.1 client receives data on a connection that doesn't have 1392 any outstanding requests, it MUST NOT consider them to be a response 1393 to a not-yet-issued request; it SHOULD close the connection, since 1394 message delimitation is now ambiguous, unless the data consists only 1395 of one or more CRLF (which can be discarded, as per Section 2.3). 1397 9.4. Persistence 1399 HTTP/1.1 defaults to the use of "persistent connections", allowing 1400 multiple requests and responses to be carried over a single 1401 connection. The "close" connection option is used to signal that a 1402 connection will not persist after the current request/response. HTTP 1403 implementations SHOULD support persistent connections. 1405 A recipient determines whether a connection is persistent or not 1406 based on the most recently received message's protocol version and 1407 Connection header field (if any): 1409 o If the "close" connection option is present, the connection will 1410 not persist after the current response; else, 1412 o If the received protocol is HTTP/1.1 (or later), the connection 1413 will persist after the current response; else, 1415 o If the received protocol is HTTP/1.0, the "keep-alive" connection 1416 option is present, either the recipient is not a proxy or the 1417 message is a response, and the recipient wishes to honor the 1418 HTTP/1.0 "keep-alive" mechanism, the connection will persist after 1419 the current response; otherwise, 1421 o The connection will close after the current response. 1423 A client MAY send additional requests on a persistent connection 1424 until it sends or receives a "close" connection option or receives an 1425 HTTP/1.0 response without a "keep-alive" connection option. 1427 In order to remain persistent, all messages on a connection need to 1428 have a self-defined message length (i.e., one not defined by closure 1429 of the connection), as described in Section 6. A server MUST read 1430 the entire request message body or close the connection after sending 1431 its response, since otherwise the remaining data on a persistent 1432 connection would be misinterpreted as the next request. Likewise, a 1433 client MUST read the entire response message body if it intends to 1434 reuse the same connection for a subsequent request. 1436 A proxy server MUST NOT maintain a persistent connection with an 1437 HTTP/1.0 client (see Section 19.7.1 of [RFC2068] for information and 1438 discussion of the problems with the Keep-Alive header field 1439 implemented by many HTTP/1.0 clients). 1441 See Appendix C.1.2 for more information on backwards compatibility 1442 with HTTP/1.0 clients. 1444 9.4.1. Retrying Requests 1446 Connections can be closed at any time, with or without intention. 1447 Implementations ought to anticipate the need to recover from 1448 asynchronous close events. 1450 When an inbound connection is closed prematurely, a client MAY open a 1451 new connection and automatically retransmit an aborted sequence of 1452 requests if all of those requests have idempotent methods 1453 (Section 7.2.2 of [Semantics]). A proxy MUST NOT automatically retry 1454 non-idempotent requests. 1456 A user agent MUST NOT automatically retry a request with a non- 1457 idempotent method unless it has some means to know that the request 1458 semantics are actually idempotent, regardless of the method, or some 1459 means to detect that the original request was never applied. For 1460 example, a user agent that knows (through design or configuration) 1461 that a POST request to a given resource is safe can repeat that 1462 request automatically. Likewise, a user agent designed specifically 1463 to operate on a version control repository might be able to recover 1464 from partial failure conditions by checking the target resource 1465 revision(s) after a failed connection, reverting or fixing any 1466 changes that were partially applied, and then automatically retrying 1467 the requests that failed. 1469 A client SHOULD NOT automatically retry a failed automatic retry. 1471 9.4.2. Pipelining 1473 A client that supports persistent connections MAY "pipeline" its 1474 requests (i.e., send multiple requests without waiting for each 1475 response). A server MAY process a sequence of pipelined requests in 1476 parallel if they all have safe methods (Section 7.2.1 of 1477 [Semantics]), but it MUST send the corresponding responses in the 1478 same order that the requests were received. 1480 A client that pipelines requests SHOULD retry unanswered requests if 1481 the connection closes before it receives all of the corresponding 1482 responses. When retrying pipelined requests after a failed 1483 connection (a connection not explicitly closed by the server in its 1484 last complete response), a client MUST NOT pipeline immediately after 1485 connection establishment, since the first remaining request in the 1486 prior pipeline might have caused an error response that can be lost 1487 again if multiple requests are sent on a prematurely closed 1488 connection (see the TCP reset problem described in Section 9.7). 1490 Idempotent methods (Section 7.2.2 of [Semantics]) are significant to 1491 pipelining because they can be automatically retried after a 1492 connection failure. A user agent SHOULD NOT pipeline requests after 1493 a non-idempotent method, until the final response status code for 1494 that method has been received, unless the user agent has a means to 1495 detect and recover from partial failure conditions involving the 1496 pipelined sequence. 1498 An intermediary that receives pipelined requests MAY pipeline those 1499 requests when forwarding them inbound, since it can rely on the 1500 outbound user agent(s) to determine what requests can be safely 1501 pipelined. If the inbound connection fails before receiving a 1502 response, the pipelining intermediary MAY attempt to retry a sequence 1503 of requests that have yet to receive a response if the requests all 1504 have idempotent methods; otherwise, the pipelining intermediary 1505 SHOULD forward any received responses and then close the 1506 corresponding outbound connection(s) so that the outbound user 1507 agent(s) can recover accordingly. 1509 9.5. Concurrency 1511 A client ought to limit the number of simultaneous open connections 1512 that it maintains to a given server. 1514 Previous revisions of HTTP gave a specific number of connections as a 1515 ceiling, but this was found to be impractical for many applications. 1516 As a result, this specification does not mandate a particular maximum 1517 number of connections but, instead, encourages clients to be 1518 conservative when opening multiple connections. 1520 Multiple connections are typically used to avoid the "head-of-line 1521 blocking" problem, wherein a request that takes significant server- 1522 side processing and/or has a large payload blocks subsequent requests 1523 on the same connection. However, each connection consumes server 1524 resources. Furthermore, using multiple connections can cause 1525 undesirable side effects in congested networks. 1527 Note that a server might reject traffic that it deems abusive or 1528 characteristic of a denial-of-service attack, such as an excessive 1529 number of open connections from a single client. 1531 9.6. Failures and Timeouts 1533 Servers will usually have some timeout value beyond which they will 1534 no longer maintain an inactive connection. Proxy servers might make 1535 this a higher value since it is likely that the client will be making 1536 more connections through the same proxy server. The use of 1537 persistent connections places no requirements on the length (or 1538 existence) of this timeout for either the client or the server. 1540 A client or server that wishes to time out SHOULD issue a graceful 1541 close on the connection. Implementations SHOULD constantly monitor 1542 open connections for a received closure signal and respond to it as 1543 appropriate, since prompt closure of both sides of a connection 1544 enables allocated system resources to be reclaimed. 1546 A client, server, or proxy MAY close the transport connection at any 1547 time. For example, a client might have started to send a new request 1548 at the same time that the server has decided to close the "idle" 1549 connection. From the server's point of view, the connection is being 1550 closed while it was idle, but from the client's point of view, a 1551 request is in progress. 1553 A server SHOULD sustain persistent connections, when possible, and 1554 allow the underlying transport's flow-control mechanisms to resolve 1555 temporary overloads, rather than terminate connections with the 1556 expectation that clients will retry. The latter technique can 1557 exacerbate network congestion. 1559 A client sending a message body SHOULD monitor the network connection 1560 for an error response while it is transmitting the request. If the 1561 client sees a response that indicates the server does not wish to 1562 receive the message body and is closing the connection, the client 1563 SHOULD immediately cease transmitting the body and close its side of 1564 the connection. 1566 9.7. Tear-down 1568 The Connection header field (Section 9.1) provides a "close" 1569 connection option that a sender SHOULD send when it wishes to close 1570 the connection after the current request/response pair. 1572 A client that sends a "close" connection option MUST NOT send further 1573 requests on that connection (after the one containing "close") and 1574 MUST close the connection after reading the final response message 1575 corresponding to this request. 1577 A server that receives a "close" connection option MUST initiate a 1578 close of the connection (see below) after it sends the final response 1579 to the request that contained "close". The server SHOULD send a 1580 "close" connection option in its final response on that connection. 1581 The server MUST NOT process any further requests received on that 1582 connection. 1584 A server that sends a "close" connection option MUST initiate a close 1585 of the connection (see below) after it sends the response containing 1586 "close". The server MUST NOT process any further requests received 1587 on that connection. 1589 A client that receives a "close" connection option MUST cease sending 1590 requests on that connection and close the connection after reading 1591 the response message containing the "close"; if additional pipelined 1592 requests had been sent on the connection, the client SHOULD NOT 1593 assume that they will be processed by the server. 1595 If a server performs an immediate close of a TCP connection, there is 1596 a significant risk that the client will not be able to read the last 1597 HTTP response. If the server receives additional data from the 1598 client on a fully closed connection, such as another request that was 1599 sent by the client before receiving the server's response, the 1600 server's TCP stack will send a reset packet to the client; 1601 unfortunately, the reset packet might erase the client's 1602 unacknowledged input buffers before they can be read and interpreted 1603 by the client's HTTP parser. 1605 To avoid the TCP reset problem, servers typically close a connection 1606 in stages. First, the server performs a half-close by closing only 1607 the write side of the read/write connection. The server then 1608 continues to read from the connection until it receives a 1609 corresponding close by the client, or until the server is reasonably 1610 certain that its own TCP stack has received the client's 1611 acknowledgement of the packet(s) containing the server's last 1612 response. Finally, the server fully closes the connection. 1614 It is unknown whether the reset problem is exclusive to TCP or might 1615 also be found in other transport connection protocols. 1617 9.8. Upgrade 1619 The "Upgrade" header field is intended to provide a simple mechanism 1620 for transitioning from HTTP/1.1 to some other protocol on the same 1621 connection. A client MAY send a list of protocols in the Upgrade 1622 header field of a request to invite the server to switch to one or 1623 more of those protocols, in order of descending preference, before 1624 sending the final response. A server MAY ignore a received Upgrade 1625 header field if it wishes to continue using the current protocol on 1626 that connection. Upgrade cannot be used to insist on a protocol 1627 change. 1629 Upgrade = 1#protocol 1631 protocol = protocol-name ["/" protocol-version] 1632 protocol-name = token 1633 protocol-version = token 1635 A server that sends a 101 (Switching Protocols) response MUST send an 1636 Upgrade header field to indicate the new protocol(s) to which the 1637 connection is being switched; if multiple protocol layers are being 1638 switched, the sender MUST list the protocols in layer-ascending 1639 order. A server MUST NOT switch to a protocol that was not indicated 1640 by the client in the corresponding request's Upgrade header field. A 1641 server MAY choose to ignore the order of preference indicated by the 1642 client and select the new protocol(s) based on other factors, such as 1643 the nature of the request or the current load on the server. 1645 A server that sends a 426 (Upgrade Required) response MUST send an 1646 Upgrade header field to indicate the acceptable protocols, in order 1647 of descending preference. 1649 A server MAY send an Upgrade header field in any other response to 1650 advertise that it implements support for upgrading to the listed 1651 protocols, in order of descending preference, when appropriate for a 1652 future request. 1654 The following is a hypothetical example sent by a client: 1656 GET /hello.txt HTTP/1.1 1657 Host: www.example.com 1658 Connection: upgrade 1659 Upgrade: HTTP/2.0, SHTTP/1.3, IRC/6.9, RTA/x11 1661 The capabilities and nature of the application-level communication 1662 after the protocol change is entirely dependent upon the new 1663 protocol(s) chosen. However, immediately after sending the 101 1664 (Switching Protocols) response, the server is expected to continue 1665 responding to the original request as if it had received its 1666 equivalent within the new protocol (i.e., the server still has an 1667 outstanding request to satisfy after the protocol has been changed, 1668 and is expected to do so without requiring the request to be 1669 repeated). 1671 For example, if the Upgrade header field is received in a GET request 1672 and the server decides to switch protocols, it first responds with a 1673 101 (Switching Protocols) message in HTTP/1.1 and then immediately 1674 follows that with the new protocol's equivalent of a response to a 1675 GET on the target resource. This allows a connection to be upgraded 1676 to protocols with the same semantics as HTTP without the latency cost 1677 of an additional round trip. A server MUST NOT switch protocols 1678 unless the received message semantics can be honored by the new 1679 protocol; an OPTIONS request can be honored by any protocol. 1681 The following is an example response to the above hypothetical 1682 request: 1684 HTTP/1.1 101 Switching Protocols 1685 Connection: upgrade 1686 Upgrade: HTTP/2.0 1688 [... data stream switches to HTTP/2.0 with an appropriate response 1689 (as defined by new protocol) to the "GET /hello.txt" request ...] 1691 When Upgrade is sent, the sender MUST also send a Connection header 1692 field (Section 9.1) that contains an "upgrade" connection option, in 1693 order to prevent Upgrade from being accidentally forwarded by 1694 intermediaries that might not implement the listed protocols. A 1695 server MUST ignore an Upgrade header field that is received in an 1696 HTTP/1.0 request. 1698 A client cannot begin using an upgraded protocol on the connection 1699 until it has completely sent the request message (i.e., the client 1700 can't change the protocol it is sending in the middle of a message). 1701 If a server receives both an Upgrade and an Expect header field with 1702 the "100-continue" expectation (Section 8.1.1 of [Semantics]), the 1703 server MUST send a 100 (Continue) response before sending a 101 1704 (Switching Protocols) response. 1706 The Upgrade header field only applies to switching protocols on top 1707 of the existing connection; it cannot be used to switch the 1708 underlying connection (transport) protocol, nor to switch the 1709 existing communication to a different connection. For those 1710 purposes, it is more appropriate to use a 3xx (Redirection) response 1711 (Section 9.4 of [Semantics]). 1713 9.8.1. Upgrade Protocol Names 1715 This specification only defines the protocol name "HTTP" for use by 1716 the family of Hypertext Transfer Protocols, as defined by the HTTP 1717 version rules of Section 3.5 of [Semantics] and future updates to 1718 this specification. Additional protocol names ought to be registered 1719 using the registration procedure defined in Section 9.8.2. 1721 +------+-------------------+--------------------+-------------------+ 1722 | Name | Description | Expected Version | Reference | 1723 | | | Tokens | | 1724 +------+-------------------+--------------------+-------------------+ 1725 | HTTP | Hypertext | any DIGIT.DIGIT | Section 3.5 of | 1726 | | Transfer Protocol | (e.g, "2.0") | [Semantics] | 1727 +------+-------------------+--------------------+-------------------+ 1729 9.8.2. Upgrade Token Registry 1731 The "Hypertext Transfer Protocol (HTTP) Upgrade Token Registry" 1732 defines the namespace for protocol-name tokens used to identify 1733 protocols in the Upgrade header field. The registry is maintained at 1734 . 1736 Each registered protocol name is associated with contact information 1737 and an optional set of specifications that details how the connection 1738 will be processed after it has been upgraded. 1740 Registrations happen on a "First Come First Served" basis (see 1741 Section 4.4 of [RFC8126]) and are subject to the following rules: 1743 1. A protocol-name token, once registered, stays registered forever. 1745 2. The registration MUST name a responsible party for the 1746 registration. 1748 3. The registration MUST name a point of contact. 1750 4. The registration MAY name a set of specifications associated with 1751 that token. Such specifications need not be publicly available. 1753 5. The registration SHOULD name a set of expected "protocol-version" 1754 tokens associated with that token at the time of registration. 1756 6. The responsible party MAY change the registration at any time. 1757 The IANA will keep a record of all such changes, and make them 1758 available upon request. 1760 7. The IESG MAY reassign responsibility for a protocol token. This 1761 will normally only be used in the case when a responsible party 1762 cannot be contacted. 1764 10. Enclosing Messages as Data 1765 10.1. Media Type message/http 1767 The message/http media type can be used to enclose a single HTTP 1768 request or response message, provided that it obeys the MIME 1769 restrictions for all "message" types regarding line length and 1770 encodings. 1772 Type name: message 1774 Subtype name: http 1776 Required parameters: N/A 1778 Optional parameters: version, msgtype 1780 version: The HTTP-version number of the enclosed message (e.g., 1781 "1.1"). If not present, the version can be determined from the 1782 first line of the body. 1784 msgtype: The message type -- "request" or "response". If not 1785 present, the type can be determined from the first line of the 1786 body. 1788 Encoding considerations: only "7bit", "8bit", or "binary" are 1789 permitted 1791 Security considerations: see Section 11 1793 Interoperability considerations: N/A 1795 Published specification: This specification (see Section 10.1). 1797 Applications that use this media type: N/A 1799 Fragment identifier considerations: N/A 1801 Additional information: 1803 Magic number(s): N/A 1805 Deprecated alias names for this type: N/A 1807 File extension(s): N/A 1809 Macintosh file type code(s): N/A 1811 Person and email address to contact for further information: 1812 See Authors' Addresses section. 1814 Intended usage: COMMON 1816 Restrictions on usage: N/A 1818 Author: See Authors' Addresses section. 1820 Change controller: IESG 1822 10.2. Media Type application/http 1824 The application/http media type can be used to enclose a pipeline of 1825 one or more HTTP request or response messages (not intermixed). 1827 Type name: application 1829 Subtype name: http 1831 Required parameters: N/A 1833 Optional parameters: version, msgtype 1835 version: The HTTP-version number of the enclosed messages (e.g., 1836 "1.1"). If not present, the version can be determined from the 1837 first line of the body. 1839 msgtype: The message type -- "request" or "response". If not 1840 present, the type can be determined from the first line of the 1841 body. 1843 Encoding considerations: HTTP messages enclosed by this type are in 1844 "binary" format; use of an appropriate Content-Transfer-Encoding 1845 is required when transmitted via email. 1847 Security considerations: see Section 11 1849 Interoperability considerations: N/A 1851 Published specification: This specification (see Section 10.2). 1853 Applications that use this media type: N/A 1855 Fragment identifier considerations: N/A 1856 Additional information: 1858 Deprecated alias names for this type: N/A 1860 Magic number(s): N/A 1862 File extension(s): N/A 1864 Macintosh file type code(s): N/A 1866 Person and email address to contact for further information: 1867 See Authors' Addresses section. 1869 Intended usage: COMMON 1871 Restrictions on usage: N/A 1873 Author: See Authors' Addresses section. 1875 Change controller: IESG 1877 11. Security Considerations 1879 This section is meant to inform developers, information providers, 1880 and users of known security considerations relevant to HTTP message 1881 syntax, parsing, and routing. Security considerations about HTTP 1882 semantics and payloads are addressed in [Semantics]. 1884 11.1. Response Splitting 1886 Response splitting (a.k.a, CRLF injection) is a common technique, 1887 used in various attacks on Web usage, that exploits the line-based 1888 nature of HTTP message framing and the ordered association of 1889 requests to responses on persistent connections [Klein]. This 1890 technique can be particularly damaging when the requests pass through 1891 a shared cache. 1893 Response splitting exploits a vulnerability in servers (usually 1894 within an application server) where an attacker can send encoded data 1895 within some parameter of the request that is later decoded and echoed 1896 within any of the response header fields of the response. If the 1897 decoded data is crafted to look like the response has ended and a 1898 subsequent response has begun, the response has been split and the 1899 content within the apparent second response is controlled by the 1900 attacker. The attacker can then make any other request on the same 1901 persistent connection and trick the recipients (including 1902 intermediaries) into believing that the second half of the split is 1903 an authoritative answer to the second request. 1905 For example, a parameter within the request-target might be read by 1906 an application server and reused within a redirect, resulting in the 1907 same parameter being echoed in the Location header field of the 1908 response. If the parameter is decoded by the application and not 1909 properly encoded when placed in the response field, the attacker can 1910 send encoded CRLF octets and other content that will make the 1911 application's single response look like two or more responses. 1913 A common defense against response splitting is to filter requests for 1914 data that looks like encoded CR and LF (e.g., "%0D" and "%0A"). 1915 However, that assumes the application server is only performing URI 1916 decoding, rather than more obscure data transformations like charset 1917 transcoding, XML entity translation, base64 decoding, sprintf 1918 reformatting, etc. A more effective mitigation is to prevent 1919 anything other than the server's core protocol libraries from sending 1920 a CR or LF within the header section, which means restricting the 1921 output of header fields to APIs that filter for bad octets and not 1922 allowing application servers to write directly to the protocol 1923 stream. 1925 11.2. Request Smuggling 1927 Request smuggling ([Linhart]) is a technique that exploits 1928 differences in protocol parsing among various recipients to hide 1929 additional requests (which might otherwise be blocked or disabled by 1930 policy) within an apparently harmless request. Like response 1931 splitting, request smuggling can lead to a variety of attacks on HTTP 1932 usage. 1934 This specification has introduced new requirements on request 1935 parsing, particularly with regard to message framing in Section 6.3, 1936 to reduce the effectiveness of request smuggling. 1938 11.3. Message Integrity 1940 HTTP does not define a specific mechanism for ensuring message 1941 integrity, instead relying on the error-detection ability of 1942 underlying transport protocols and the use of length or chunk- 1943 delimited framing to detect completeness. Additional integrity 1944 mechanisms, such as hash functions or digital signatures applied to 1945 the content, can be selectively added to messages via extensible 1946 metadata header fields. Historically, the lack of a single integrity 1947 mechanism has been justified by the informal nature of most HTTP 1948 communication. However, the prevalence of HTTP as an information 1949 access mechanism has resulted in its increasing use within 1950 environments where verification of message integrity is crucial. 1952 User agents are encouraged to implement configurable means for 1953 detecting and reporting failures of message integrity such that those 1954 means can be enabled within environments for which integrity is 1955 necessary. For example, a browser being used to view medical history 1956 or drug interaction information needs to indicate to the user when 1957 such information is detected by the protocol to be incomplete, 1958 expired, or corrupted during transfer. Such mechanisms might be 1959 selectively enabled via user agent extensions or the presence of 1960 message integrity metadata in a response. At a minimum, user agents 1961 ought to provide some indication that allows a user to distinguish 1962 between a complete and incomplete response message (Section 8) when 1963 such verification is desired. 1965 11.4. Message Confidentiality 1967 HTTP relies on underlying transport protocols to provide message 1968 confidentiality when that is desired. HTTP has been specifically 1969 designed to be independent of the transport protocol, such that it 1970 can be used over many different forms of encrypted connection, with 1971 the selection of such transports being identified by the choice of 1972 URI scheme or within user agent configuration. 1974 The "https" scheme can be used to identify resources that require a 1975 confidential connection, as described in Section 2.5.2 of 1976 [Semantics]. 1978 12. IANA Considerations 1980 The change controller for the following registrations is: "IETF 1981 (iesg@ietf.org) - Internet Engineering Task Force". 1983 12.1. Header Field Registration 1985 Please update the "Hypertext Transfer Protocol (HTTP) Header Field 1986 Registry" registry at 1987 with the header field names listed in the two tables of Section 5. 1989 12.2. Media Type Registration 1991 Please update the "Media Types" registry at 1992 with the registration 1993 information in Section 10.1 and Section 10.2 for the media types 1994 "message/http" and "application/http", respectively. 1996 12.3. Transfer Coding Registration 1998 Please update the "HTTP Transfer Coding Registry" at 1999 with the 2000 registration procedure of Section 7.3 and the content coding names 2001 summarized in the table of Section 7. 2003 12.4. Upgrade Token Registration 2005 Please update the "Hypertext Transfer Protocol (HTTP) Upgrade Token 2006 Registry" at 2007 with the registration procedure of Section 9.8.2 and the upgrade 2008 token names summarized in the table of Section 9.8.1. 2010 13. References 2012 13.1. Normative References 2014 [Caching] Fielding, R., Ed., Nottingham, M., Ed., and J. Reschke, 2015 Ed., "HTTP Caching", draft-ietf-httpbis-cache-04 (work in 2016 progress), March 2019. 2018 [RFC1950] Deutsch, L. and J-L. Gailly, "ZLIB Compressed Data Format 2019 Specification version 3.3", RFC 1950, 2020 DOI 10.17487/RFC1950, May 1996, 2021 . 2023 [RFC1951] Deutsch, P., "DEFLATE Compressed Data Format Specification 2024 version 1.3", RFC 1951, DOI 10.17487/RFC1951, May 1996, 2025 . 2027 [RFC1952] Deutsch, P., Gailly, J-L., Adler, M., Deutsch, L., and G. 2028 Randers-Pehrson, "GZIP file format specification version 2029 4.3", RFC 1952, DOI 10.17487/RFC1952, May 1996, 2030 . 2032 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2033 Requirement Levels", BCP 14, RFC 2119, 2034 DOI 10.17487/RFC2119, March 1997, 2035 . 2037 [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform 2038 Resource Identifier (URI): Generic Syntax", STD 66, 2039 RFC 3986, DOI 10.17487/RFC3986, January 2005, 2040 . 2042 [RFC5234] Crocker, D., Ed. and P. Overell, "Augmented BNF for Syntax 2043 Specifications: ABNF", STD 68, RFC 5234, 2044 DOI 10.17487/RFC5234, January 2008, 2045 . 2047 [RFC7405] Kyzivat, P., "Case-Sensitive String Support in ABNF", 2048 RFC 7405, DOI 10.17487/RFC7405, December 2014, 2049 . 2051 [Semantics] 2052 Fielding, R., Ed., Nottingham, M., Ed., and J. Reschke, 2053 Ed., "HTTP Semantics", draft-ietf-httpbis-semantics-04 2054 (work in progress), March 2019. 2056 [USASCII] American National Standards Institute, "Coded Character 2057 Set -- 7-bit American Standard Code for Information 2058 Interchange", ANSI X3.4, 1986. 2060 [Welch] Welch, T., "A Technique for High-Performance Data 2061 Compression", IEEE Computer 17(6), June 1984. 2063 13.2. Informative References 2065 [Err4667] RFC Errata, Erratum ID 4667, RFC 7230, 2066 . 2068 [Klein] Klein, A., "Divide and Conquer - HTTP Response Splitting, 2069 Web Cache Poisoning Attacks, and Related Topics", March 2070 2004, . 2073 [Linhart] Linhart, C., Klein, A., Heled, R., and S. Orrin, "HTTP 2074 Request Smuggling", June 2005, 2075 . 2077 [RFC1945] Berners-Lee, T., Fielding, R., and H. Nielsen, "Hypertext 2078 Transfer Protocol -- HTTP/1.0", RFC 1945, 2079 DOI 10.17487/RFC1945, May 1996, 2080 . 2082 [RFC2045] Freed, N. and N. Borenstein, "Multipurpose Internet Mail 2083 Extensions (MIME) Part One: Format of Internet Message 2084 Bodies", RFC 2045, DOI 10.17487/RFC2045, November 1996, 2085 . 2087 [RFC2046] Freed, N. and N. Borenstein, "Multipurpose Internet Mail 2088 Extensions (MIME) Part Two: Media Types", RFC 2046, 2089 DOI 10.17487/RFC2046, November 1996, 2090 . 2092 [RFC2049] Freed, N. and N. Borenstein, "Multipurpose Internet Mail 2093 Extensions (MIME) Part Five: Conformance Criteria and 2094 Examples", RFC 2049, DOI 10.17487/RFC2049, November 1996, 2095 . 2097 [RFC2068] Fielding, R., Gettys, J., Mogul, J., Nielsen, H., and T. 2098 Berners-Lee, "Hypertext Transfer Protocol -- HTTP/1.1", 2099 RFC 2068, DOI 10.17487/RFC2068, January 1997, 2100 . 2102 [RFC2557] Palme, F., Hopmann, A., Shelness, N., and E. Stefferud, 2103 "MIME Encapsulation of Aggregate Documents, such as HTML 2104 (MHTML)", RFC 2557, DOI 10.17487/RFC2557, March 1999, 2105 . 2107 [RFC5322] Resnick, P., "Internet Message Format", RFC 5322, 2108 DOI 10.17487/RFC5322, October 2008, 2109 . 2111 [RFC6265] Barth, A., "HTTP State Management Mechanism", RFC 6265, 2112 DOI 10.17487/RFC6265, April 2011, 2113 . 2115 [RFC7230] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 2116 Protocol (HTTP/1.1): Message Syntax and Routing", 2117 RFC 7230, DOI 10.17487/RFC7230, June 2014, 2118 . 2120 [RFC7231] Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer 2121 Protocol (HTTP/1.1): Semantics and Content", RFC 7231, 2122 DOI 10.17487/RFC7231, June 2014, 2123 . 2125 [RFC8126] Cotton, M., Leiba, B., and T. Narten, "Guidelines for 2126 Writing an IANA Considerations Section in RFCs", BCP 26, 2127 RFC 8126, DOI 10.17487/RFC8126, June 2017, 2128 . 2130 Appendix A. Collected ABNF 2132 In the collected ABNF below, list rules are expanded as per 2133 Section 11 of [Semantics]. 2135 BWS = 2137 Connection = *( "," OWS ) connection-option *( OWS "," [ OWS 2138 connection-option ] ) 2140 HTTP-message = start-line *( header-field CRLF ) CRLF [ message-body 2141 ] 2142 HTTP-name = %x48.54.54.50 ; HTTP 2143 HTTP-version = HTTP-name "/" DIGIT "." DIGIT 2145 OWS = 2147 RWS = 2149 TE = [ ( "," / t-codings ) *( OWS "," [ OWS t-codings ] ) ] 2150 Transfer-Encoding = *( "," OWS ) transfer-coding *( OWS "," [ OWS 2151 transfer-coding ] ) 2153 Upgrade = *( "," OWS ) protocol *( OWS "," [ OWS protocol ] ) 2155 absolute-URI = 2156 absolute-form = absolute-URI 2157 absolute-path = 2158 asterisk-form = "*" 2159 authority = 2160 authority-form = authority 2162 chunk = chunk-size [ chunk-ext ] CRLF chunk-data CRLF 2163 chunk-data = 1*OCTET 2164 chunk-ext = *( BWS ";" BWS chunk-ext-name [ BWS "=" BWS chunk-ext-val 2165 ] ) 2166 chunk-ext-name = token 2167 chunk-ext-val = token / quoted-string 2168 chunk-size = 1*HEXDIG 2169 chunked-body = *chunk last-chunk trailer-part CRLF 2170 comment = 2171 connection-option = token 2173 field-name = 2174 field-value = 2176 header-field = field-name ":" OWS field-value OWS 2177 last-chunk = 1*"0" [ chunk-ext ] CRLF 2179 message-body = *OCTET 2180 method = token 2182 obs-fold = CRLF 1*( SP / HTAB ) 2183 obs-text = 2184 origin-form = absolute-path [ "?" query ] 2186 port = 2187 protocol = protocol-name [ "/" protocol-version ] 2188 protocol-name = token 2189 protocol-version = token 2191 query = 2192 quoted-string = 2194 rank = ( "0" [ "." *3DIGIT ] ) / ( "1" [ "." *3"0" ] ) 2195 reason-phrase = *( HTAB / SP / VCHAR / obs-text ) 2196 request-line = method SP request-target SP HTTP-version CRLF 2197 request-target = origin-form / absolute-form / authority-form / 2198 asterisk-form 2200 start-line = request-line / status-line 2201 status-code = 3DIGIT 2202 status-line = HTTP-version SP status-code SP reason-phrase CRLF 2204 t-codings = "trailers" / ( transfer-coding [ t-ranking ] ) 2205 t-ranking = OWS ";" OWS "q=" rank 2206 token = 2207 trailer-part = *( header-field CRLF ) 2208 transfer-coding = token *( OWS ";" OWS transfer-parameter ) 2209 transfer-parameter = token BWS "=" BWS ( token / quoted-string ) 2211 uri-host = 2213 Appendix B. Differences between HTTP and MIME 2215 HTTP/1.1 uses many of the constructs defined for the Internet Message 2216 Format [RFC5322] and the Multipurpose Internet Mail Extensions (MIME) 2217 [RFC2045] to allow a message body to be transmitted in an open 2218 variety of representations and with extensible header fields. 2219 However, RFC 2045 is focused only on email; applications of HTTP have 2220 many characteristics that differ from email; hence, HTTP has features 2221 that differ from MIME. These differences were carefully chosen to 2222 optimize performance over binary connections, to allow greater 2223 freedom in the use of new media types, to make date comparisons 2224 easier, and to acknowledge the practice of some early HTTP servers 2225 and clients. 2227 This appendix describes specific areas where HTTP differs from MIME. 2228 Proxies and gateways to and from strict MIME environments need to be 2229 aware of these differences and provide the appropriate conversions 2230 where necessary. 2232 B.1. MIME-Version 2234 HTTP is not a MIME-compliant protocol. However, messages can include 2235 a single MIME-Version header field to indicate what version of the 2236 MIME protocol was used to construct the message. Use of the MIME- 2237 Version header field indicates that the message is in full 2238 conformance with the MIME protocol (as defined in [RFC2045]). 2239 Senders are responsible for ensuring full conformance (where 2240 possible) when exporting HTTP messages to strict MIME environments. 2242 B.2. Conversion to Canonical Form 2244 MIME requires that an Internet mail body part be converted to 2245 canonical form prior to being transferred, as described in Section 4 2246 of [RFC2049]. Section 6.1.1.2 of [Semantics] describes the forms 2247 allowed for subtypes of the "text" media type when transmitted over 2248 HTTP. [RFC2046] requires that content with a type of "text" 2249 represent line breaks as CRLF and forbids the use of CR or LF outside 2250 of line break sequences. HTTP allows CRLF, bare CR, and bare LF to 2251 indicate a line break within text content. 2253 A proxy or gateway from HTTP to a strict MIME environment ought to 2254 translate all line breaks within text media types to the RFC 2049 2255 canonical form of CRLF. Note, however, this might be complicated by 2256 the presence of a Content-Encoding and by the fact that HTTP allows 2257 the use of some charsets that do not use octets 13 and 10 to 2258 represent CR and LF, respectively. 2260 Conversion will break any cryptographic checksums applied to the 2261 original content unless the original content is already in canonical 2262 form. Therefore, the canonical form is recommended for any content 2263 that uses such checksums in HTTP. 2265 B.3. Conversion of Date Formats 2267 HTTP/1.1 uses a restricted set of date formats (Section 10.1.1.1 of 2268 [Semantics]) to simplify the process of date comparison. Proxies and 2269 gateways from other protocols ought to ensure that any Date header 2270 field present in a message conforms to one of the HTTP/1.1 formats 2271 and rewrite the date if necessary. 2273 B.4. Conversion of Content-Encoding 2275 MIME does not include any concept equivalent to HTTP/1.1's Content- 2276 Encoding header field. Since this acts as a modifier on the media 2277 type, proxies and gateways from HTTP to MIME-compliant protocols 2278 ought to either change the value of the Content-Type header field or 2279 decode the representation before forwarding the message. (Some 2280 experimental applications of Content-Type for Internet mail have used 2281 a media-type parameter of ";conversions=" to perform 2282 a function equivalent to Content-Encoding. However, this parameter 2283 is not part of the MIME standards). 2285 B.5. Conversion of Content-Transfer-Encoding 2287 HTTP does not use the Content-Transfer-Encoding field of MIME. 2288 Proxies and gateways from MIME-compliant protocols to HTTP need to 2289 remove any Content-Transfer-Encoding prior to delivering the response 2290 message to an HTTP client. 2292 Proxies and gateways from HTTP to MIME-compliant protocols are 2293 responsible for ensuring that the message is in the correct format 2294 and encoding for safe transport on that protocol, where "safe 2295 transport" is defined by the limitations of the protocol being used. 2296 Such a proxy or gateway ought to transform and label the data with an 2297 appropriate Content-Transfer-Encoding if doing so will improve the 2298 likelihood of safe transport over the destination protocol. 2300 B.6. MHTML and Line Length Limitations 2302 HTTP implementations that share code with MHTML [RFC2557] 2303 implementations need to be aware of MIME line length limitations. 2304 Since HTTP does not have this limitation, HTTP does not fold long 2305 lines. MHTML messages being transported by HTTP follow all 2306 conventions of MHTML, including line length limitations and folding, 2307 canonicalization, etc., since HTTP transfers message-bodies as 2308 payload and, aside from the "multipart/byteranges" type 2309 (Section 6.3.4 of [Semantics]), does not interpret the content or any 2310 MIME header lines that might be contained therein. 2312 Appendix C. HTTP Version History 2314 HTTP has been in use since 1990. The first version, later referred 2315 to as HTTP/0.9, was a simple protocol for hypertext data transfer 2316 across the Internet, using only a single request method (GET) and no 2317 metadata. HTTP/1.0, as defined by [RFC1945], added a range of 2318 request methods and MIME-like messaging, allowing for metadata to be 2319 transferred and modifiers placed on the request/response semantics. 2320 However, HTTP/1.0 did not sufficiently take into consideration the 2321 effects of hierarchical proxies, caching, the need for persistent 2322 connections, or name-based virtual hosts. The proliferation of 2323 incompletely implemented applications calling themselves "HTTP/1.0" 2324 further necessitated a protocol version change in order for two 2325 communicating applications to determine each other's true 2326 capabilities. 2328 HTTP/1.1 remains compatible with HTTP/1.0 by including more stringent 2329 requirements that enable reliable implementations, adding only those 2330 features that can either be safely ignored by an HTTP/1.0 recipient 2331 or only be sent when communicating with a party advertising 2332 conformance with HTTP/1.1. 2334 HTTP/1.1 has been designed to make supporting previous versions easy. 2335 A general-purpose HTTP/1.1 server ought to be able to understand any 2336 valid request in the format of HTTP/1.0, responding appropriately 2337 with an HTTP/1.1 message that only uses features understood (or 2338 safely ignored) by HTTP/1.0 clients. Likewise, an HTTP/1.1 client 2339 can be expected to understand any valid HTTP/1.0 response. 2341 Since HTTP/0.9 did not support header fields in a request, there is 2342 no mechanism for it to support name-based virtual hosts (selection of 2343 resource by inspection of the Host header field). Any server that 2344 implements name-based virtual hosts ought to disable support for 2345 HTTP/0.9. Most requests that appear to be HTTP/0.9 are, in fact, 2346 badly constructed HTTP/1.x requests caused by a client failing to 2347 properly encode the request-target. 2349 C.1. Changes from HTTP/1.0 2351 This section summarizes major differences between versions HTTP/1.0 2352 and HTTP/1.1. 2354 C.1.1. Multihomed Web Servers 2356 The requirements that clients and servers support the Host header 2357 field (Section 5.4 of [Semantics]), report an error if it is missing 2358 from an HTTP/1.1 request, and accept absolute URIs (Section 3.2) are 2359 among the most important changes defined by HTTP/1.1. 2361 Older HTTP/1.0 clients assumed a one-to-one relationship of IP 2362 addresses and servers; there was no other established mechanism for 2363 distinguishing the intended server of a request than the IP address 2364 to which that request was directed. The Host header field was 2365 introduced during the development of HTTP/1.1 and, though it was 2366 quickly implemented by most HTTP/1.0 browsers, additional 2367 requirements were placed on all HTTP/1.1 requests in order to ensure 2368 complete adoption. At the time of this writing, most HTTP-based 2369 services are dependent upon the Host header field for targeting 2370 requests. 2372 C.1.2. Keep-Alive Connections 2374 In HTTP/1.0, each connection is established by the client prior to 2375 the request and closed by the server after sending the response. 2376 However, some implementations implement the explicitly negotiated 2377 ("Keep-Alive") version of persistent connections described in 2378 Section 19.7.1 of [RFC2068]. 2380 Some clients and servers might wish to be compatible with these 2381 previous approaches to persistent connections, by explicitly 2382 negotiating for them with a "Connection: keep-alive" request header 2383 field. However, some experimental implementations of HTTP/1.0 2384 persistent connections are faulty; for example, if an HTTP/1.0 proxy 2385 server doesn't understand Connection, it will erroneously forward 2386 that header field to the next inbound server, which would result in a 2387 hung connection. 2389 One attempted solution was the introduction of a Proxy-Connection 2390 header field, targeted specifically at proxies. In practice, this 2391 was also unworkable, because proxies are often deployed in multiple 2392 layers, bringing about the same problem discussed above. 2394 As a result, clients are encouraged not to send the Proxy-Connection 2395 header field in any requests. 2397 Clients are also encouraged to consider the use of Connection: keep- 2398 alive in requests carefully; while they can enable persistent 2399 connections with HTTP/1.0 servers, clients using them will need to 2400 monitor the connection for "hung" requests (which indicate that the 2401 client ought stop sending the header field), and this mechanism ought 2402 not be used by clients at all when a proxy is being used. 2404 C.1.3. Introduction of Transfer-Encoding 2406 HTTP/1.1 introduces the Transfer-Encoding header field (Section 6.1). 2407 Transfer codings need to be decoded prior to forwarding an HTTP 2408 message over a MIME-compliant protocol. 2410 C.2. Changes from RFC 7230 2412 Most of the sections introducing HTTP's design goals, history, 2413 architecture, conformance criteria, protocol versioning, URIs, 2414 message routing, and header field values have been moved to 2415 [Semantics]. This document has been reduced to just the messaging 2416 syntax and connection management requirements specific to HTTP/1.1. 2418 Furthermore: 2420 In the ABNF for chunked extensions, re-introduce (bad) whitespace 2421 around ";" and "=". Whitespace was removed in [RFC7230], but later 2422 this change was found to break existing implementations (see 2423 [Err4667]). (Section 7.1.1) 2425 Disallow transfer coding parameters called "q" in order to avoid 2426 conflicts with the use of ranks in the TE header field. 2427 (Section 7.3) 2429 Appendix D. Change Log 2431 This section is to be removed before publishing as an RFC. 2433 D.1. Between RFC7230 and draft 00 2435 The changes were purely editorial: 2437 o Change boilerplate and abstract to indicate the "draft" status, 2438 and update references to ancestor specifications. 2440 o Adjust historical notes. 2442 o Update links to sibling specifications. 2444 o Replace sections listing changes from RFC 2616 by new empty 2445 sections referring to RFC 723x. 2447 o Remove acknowledgements specific to RFC 723x. 2449 o Move "Acknowledgements" to the very end and make them unnumbered. 2451 D.2. Since draft-ietf-httpbis-messaging-00 2453 The changes in this draft are editorial, with respect to HTTP as a 2454 whole, to move all core HTTP semantics into [Semantics]: 2456 o Moved introduction, architecture, conformance, and ABNF extensions 2457 from RFC 7230 (Messaging) to semantics [Semantics]. 2459 o Moved discussion of MIME differences from RFC 7231 (Semantics) to 2460 Appendix B since they mostly cover transforming 1.1 messages. 2462 o Moved all extensibility tips, registration procedures, and 2463 registry tables from the IANA considerations to normative 2464 sections, reducing the IANA considerations to just instructions 2465 that will be removed prior to publication as an RFC. 2467 D.3. Since draft-ietf-httpbis-messaging-01 2469 o Cite RFC 8126 instead of RFC 5226 () 2472 o Resolved erratum 4779, no change needed here 2473 (, 2474 ) 2476 o In Section 7, fixed prose claiming transfer parameters allow bare 2477 names (, 2478 ) 2480 o Resolved erratum 4225, no change needed here 2481 (, 2482 ) 2484 o Replace "response code" with "response status code" 2485 (, 2486 ) 2488 o In Section 9.4, clarify statement about HTTP/1.0 keep-alive 2489 (, 2490 ) 2492 o In Section 7.1.1, re-introduce (bad) whitespace around ";" and "=" 2493 (, 2494 , ) 2497 o In Section 7.3, state that transfer codings should not use 2498 parameters named "q" (, ) 2501 o In Section 7, mark coding name "trailers" as reserved in the IANA 2502 registry () 2504 D.4. Since draft-ietf-httpbis-messaging-02 2506 o In Section 4, explain why the reason phrase should be ignored by 2507 clients (). 2509 o Add Section 9.3 to explain how request/response correlation is 2510 performed () 2512 D.5. Since draft-ietf-httpbis-messaging-03 2514 o In Section 9.3, caution against treating data on a connection as 2515 part of a not-yet-issued request () 2518 o In Section 7, remove the predefined codings from the ABNF and make 2519 it generic instead () 2522 o Use RFC 7405 ABNF notation for case-sensitive string constants 2523 () 2525 Index 2527 A 2528 absolute-form (of request-target) 10 2529 application/http Media Type 39 2530 asterisk-form (of request-target) 11 2531 authority-form (of request-target) 11 2533 C 2534 Connection header field 28, 33 2535 Content-Length header field 18 2536 Content-Transfer-Encoding header field 49 2537 chunked (Coding Format) 17, 19 2538 chunked (transfer coding) 22 2539 close 28, 33 2540 compress (transfer coding) 24 2542 D 2543 deflate (transfer coding) 24 2545 E 2546 effective request URI 12 2548 G 2549 Grammar 2550 absolute-form 9-10 2551 ALPHA 5 2552 asterisk-form 9, 11 2553 authority-form 9, 11 2554 chunk 22 2555 chunk-data 22 2556 chunk-ext 22 2557 chunk-ext-name 22 2558 chunk-ext-val 22 2559 chunk-size 22 2560 chunked-body 22 2561 Connection 28 2562 connection-option 28 2563 CR 5 2564 CRLF 5 2565 CTL 5 2566 DIGIT 5 2567 DQUOTE 5 2568 field-name 14 2569 field-value 14 2570 header-field 14, 23 2571 HEXDIG 5 2572 HTAB 5 2573 HTTP-message 6 2574 HTTP-name 6 2575 HTTP-version 6 2576 last-chunk 22 2577 LF 5 2578 message-body 16 2579 method 9 2580 obs-fold 15 2581 OCTET 5 2582 origin-form 9-10 2583 rank 26 2584 reason-phrase 14 2585 request-line 8 2586 request-target 9 2587 SP 5 2588 start-line 6 2589 status-code 14 2590 status-line 13 2591 t-codings 26 2592 t-ranking 26 2593 TE 26 2594 trailer-part 22-23 2595 transfer-coding 21 2596 Transfer-Encoding 17 2597 transfer-parameter 21 2598 Upgrade 35 2599 VCHAR 5 2600 gzip (transfer coding) 24 2602 H 2603 header field 6 2604 header section 6 2605 headers 6 2607 M 2608 MIME-Version header field 48 2609 Media Type 2610 application/http 39 2611 message/http 38 2612 message/http Media Type 38 2613 method 9 2615 O 2616 origin-form (of request-target) 10 2618 R 2619 request-target 9 2621 T 2622 TE header field 25 2623 Transfer-Encoding header field 17 2625 U 2626 Upgrade header field 34 2628 X 2629 x-compress (transfer coding) 24 2630 x-gzip (transfer coding) 24 2632 Acknowledgments 2634 See Appendix "Acknowledgments" of [Semantics]. 2636 Authors' Addresses 2638 Roy T. Fielding (editor) 2639 Adobe 2640 345 Park Ave 2641 San Jose, CA 95110 2642 USA 2644 EMail: fielding@gbiv.com 2645 URI: https://roy.gbiv.com/ 2647 Mark Nottingham (editor) 2648 Fastly 2650 EMail: mnot@mnot.net 2651 URI: https://www.mnot.net/ 2652 Julian F. Reschke (editor) 2653 greenbytes GmbH 2654 Hafenweg 16 2655 Muenster, NW 48155 2656 Germany 2658 EMail: julian.reschke@greenbytes.de 2659 URI: https://greenbytes.de/tech/webdav/