idnits 2.17.1 draft-ietf-httpbis-messaging-10.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: ---------------------------------------------------------------------------- == There is 1 instance of lines with non-ascii characters in the document. 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 (July 12, 2020) is 1382 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 2105, but no explicit reference was found in the text == Unused Reference: 'RFC7231' is defined on line 2194, but no explicit reference was found in the text == Outdated reference: A later version (-19) exists of draft-ietf-httpbis-cache-10 -- Possible downref: Normative reference to a draft: ref.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reference to an Informational RFC: RFC 1951 ** Downref: Normative reference to an Informational RFC: RFC 1952 == Outdated reference: A later version (-19) exists of draft-ietf-httpbis-semantics-10 -- Possible downref: Normative reference to a draft: ref.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 (~~), 7 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: January 13, 2021 J. F. Reschke, Ed. 7 greenbytes 8 July 12, 2020 10 HTTP/1.1 Messaging 11 draft-ietf-httpbis-messaging-10 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.11. 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 January 13, 2021. 54 Copyright Notice 56 Copyright (c) 2020 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 (https://trustee.ietf.org/ 61 license-info) in effect on the date of publication of this document. 62 Please review these documents carefully, as they describe your rights 63 and restrictions with respect to this document. Code Components 64 extracted from this document must include Simplified BSD License text 65 as described in Section 4.e of the Trust Legal Provisions and are 66 provided without warranty as described in the Simplified BSD License. 68 This document may contain material from IETF Documents or IETF 69 Contributions published or made publicly available before November 70 10, 2008. The person(s) controlling the copyright in some of this 71 material may not have granted the IETF Trust the right to allow 72 modifications of such material outside the IETF Standards Process. 73 Without obtaining an adequate license from the person(s) controlling 74 the copyright in such materials, this document may not be modified 75 outside the IETF Standards Process, and derivative works of it may 76 not be created outside the IETF Standards Process, except to format 77 it for publication as an RFC or to translate it into languages other 78 than English. 80 Table of Contents 82 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 83 1.1. Requirements Notation . . . . . . . . . . . . . . . . . . 5 84 1.2. Syntax Notation . . . . . . . . . . . . . . . . . . . . . 5 85 2. Message . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 86 2.1. Message Format . . . . . . . . . . . . . . . . . . . . . 6 87 2.2. Message Parsing . . . . . . . . . . . . . . . . . . . . . 7 88 2.3. HTTP Version . . . . . . . . . . . . . . . . . . . . . . 8 89 3. Request Line . . . . . . . . . . . . . . . . . . . . . . . . 9 90 3.1. Method . . . . . . . . . . . . . . . . . . . . . . . . . 10 91 3.2. Request Target . . . . . . . . . . . . . . . . . . . . . 10 92 3.2.1. origin-form . . . . . . . . . . . . . . . . . . . . . 10 93 3.2.2. absolute-form . . . . . . . . . . . . . . . . . . . . 11 94 3.2.3. authority-form . . . . . . . . . . . . . . . . . . . 12 95 3.2.4. asterisk-form . . . . . . . . . . . . . . . . . . . . 12 96 3.3. Reconstructing the Target URI . . . . . . . . . . . . . . 13 97 4. Status Line . . . . . . . . . . . . . . . . . . . . . . . . . 14 98 5. Field Syntax . . . . . . . . . . . . . . . . . . . . . . . . 15 99 5.1. Field Line Parsing . . . . . . . . . . . . . . . . . . . 16 100 5.2. Obsolete Line Folding . . . . . . . . . . . . . . . . . . 17 101 6. Message Body . . . . . . . . . . . . . . . . . . . . . . . . 17 102 6.1. Transfer-Encoding . . . . . . . . . . . . . . . . . . . . 18 103 6.2. Content-Length . . . . . . . . . . . . . . . . . . . . . 19 104 6.3. Message Body Length . . . . . . . . . . . . . . . . . . . 20 105 7. Transfer Codings . . . . . . . . . . . . . . . . . . . . . . 22 106 7.1. Chunked Transfer Coding . . . . . . . . . . . . . . . . . 23 107 7.1.1. Chunk Extensions . . . . . . . . . . . . . . . . . . 24 108 7.1.2. Chunked Trailer Section . . . . . . . . . . . . . . . 25 109 7.1.3. Decoding Chunked . . . . . . . . . . . . . . . . . . 25 110 7.2. Transfer Codings for Compression . . . . . . . . . . . . 26 111 7.3. Transfer Coding Registry . . . . . . . . . . . . . . . . 26 112 7.4. TE . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 113 8. Handling Incomplete Messages . . . . . . . . . . . . . . . . 28 114 9. Connection Management . . . . . . . . . . . . . . . . . . . . 28 115 9.1. Connection . . . . . . . . . . . . . . . . . . . . . . . 29 116 9.2. Establishment . . . . . . . . . . . . . . . . . . . . . . 30 117 9.3. Associating a Response to a Request . . . . . . . . . . . 31 118 9.4. Persistence . . . . . . . . . . . . . . . . . . . . . . . 31 119 9.4.1. Retrying Requests . . . . . . . . . . . . . . . . . . 32 120 9.4.2. Pipelining . . . . . . . . . . . . . . . . . . . . . 32 121 9.5. Concurrency . . . . . . . . . . . . . . . . . . . . . . . 33 122 9.6. Failures and Timeouts . . . . . . . . . . . . . . . . . . 34 123 9.7. Tear-down . . . . . . . . . . . . . . . . . . . . . . . . 34 124 9.8. TLS Connection Closure . . . . . . . . . . . . . . . . . 35 125 9.9. Upgrade . . . . . . . . . . . . . . . . . . . . . . . . . 36 126 9.9.1. Upgrade Protocol Names . . . . . . . . . . . . . . . 39 127 9.9.2. Upgrade Token Registry . . . . . . . . . . . . . . . 39 128 10. Enclosing Messages as Data . . . . . . . . . . . . . . . . . 40 129 10.1. Media Type message/http . . . . . . . . . . . . . . . . 40 130 10.2. Media Type application/http . . . . . . . . . . . . . . 41 131 11. Security Considerations . . . . . . . . . . . . . . . . . . . 42 132 11.1. Response Splitting . . . . . . . . . . . . . . . . . . . 42 133 11.2. Request Smuggling . . . . . . . . . . . . . . . . . . . 43 134 11.3. Message Integrity . . . . . . . . . . . . . . . . . . . 43 135 11.4. Message Confidentiality . . . . . . . . . . . . . . . . 44 136 12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 44 137 12.1. Field Name Registration . . . . . . . . . . . . . . . . 44 138 12.2. Media Type Registration . . . . . . . . . . . . . . . . 44 139 12.3. Transfer Coding Registration . . . . . . . . . . . . . . 45 140 12.4. Upgrade Token Registration . . . . . . . . . . . . . . . 45 141 12.5. ALPN Protocol ID Registration . . . . . . . . . . . . . 45 142 13. References . . . . . . . . . . . . . . . . . . . . . . . . . 45 143 13.1. Normative References . . . . . . . . . . . . . . . . . . 45 144 13.2. Informative References . . . . . . . . . . . . . . . . . 46 145 Appendix A. Collected ABNF . . . . . . . . . . . . . . . . . . . 48 146 Appendix B. Differences between HTTP and MIME . . . . . . . . . 50 147 B.1. MIME-Version . . . . . . . . . . . . . . . . . . . . . . 50 148 B.2. Conversion to Canonical Form . . . . . . . . . . . . . . 50 149 B.3. Conversion of Date Formats . . . . . . . . . . . . . . . 51 150 B.4. Conversion of Content-Encoding . . . . . . . . . . . . . 51 151 B.5. Conversion of Content-Transfer-Encoding . . . . . . . . . 51 152 B.6. MHTML and Line Length Limitations . . . . . . . . . . . . 51 153 Appendix C. HTTP Version History . . . . . . . . . . . . . . . . 52 154 C.1. Changes from HTTP/1.0 . . . . . . . . . . . . . . . . . . 52 155 C.1.1. Multihomed Web Servers . . . . . . . . . . . . . . . 53 156 C.1.2. Keep-Alive Connections . . . . . . . . . . . . . . . 53 157 C.1.3. Introduction of Transfer-Encoding . . . . . . . . . . 54 158 C.2. Changes from RFC 7230 . . . . . . . . . . . . . . . . . . 54 159 Appendix D. Change Log . . . . . . . . . . . . . . . . . . . . . 54 160 D.1. Between RFC7230 and draft 00 . . . . . . . . . . . . . . 55 161 D.2. Since draft-ietf-httpbis-messaging-00 . . . . . . . . . . 55 162 D.3. Since draft-ietf-httpbis-messaging-01 . . . . . . . . . . 55 163 D.4. Since draft-ietf-httpbis-messaging-02 . . . . . . . . . . 56 164 D.5. Since draft-ietf-httpbis-messaging-03 . . . . . . . . . . 56 165 D.6. Since draft-ietf-httpbis-messaging-04 . . . . . . . . . . 56 166 D.7. Since draft-ietf-httpbis-messaging-05 . . . . . . . . . . 57 167 D.8. Since draft-ietf-httpbis-messaging-06 . . . . . . . . . . 57 168 D.9. Since draft-ietf-httpbis-messaging-07 . . . . . . . . . . 58 169 D.10. Since draft-ietf-httpbis-messaging-08 . . . . . . . . . . 58 170 D.11. Since draft-ietf-httpbis-messaging-09 . . . . . . . . . . 58 171 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 58 172 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 58 174 1. Introduction 176 The Hypertext Transfer Protocol (HTTP) is a stateless application- 177 level request/response protocol that uses extensible semantics and 178 self-descriptive messages for flexible interaction with network-based 179 hypertext information systems. HTTP is defined by a series of 180 documents that collectively form the HTTP/1.1 specification: 182 o "HTTP Semantics" [Semantics] 184 o "HTTP Caching" [Caching] 186 o "HTTP/1.1 Messaging" (this document) 188 This document defines HTTP/1.1 message syntax and framing 189 requirements and their associated connection management. Our goal is 190 to define all of the mechanisms necessary for HTTP/1.1 message 191 handling that are independent of message semantics, thereby defining 192 the complete set of requirements for message parsers and message- 193 forwarding intermediaries. 195 This document obsoletes the portions of RFC 7230 related to HTTP/1.1 196 messaging and connection management, with the changes being 197 summarized in Appendix C.2. The other parts of RFC 7230 are 198 obsoleted by "HTTP Semantics" [Semantics]. 200 1.1. Requirements Notation 202 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 203 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 204 "OPTIONAL" in this document are to be interpreted as described in BCP 205 14 [RFC2119] [RFC8174] when, and only when, they appear in all 206 capitals, as shown here. 208 Conformance criteria and considerations regarding error handling are 209 defined in Section 3 of [Semantics]. 211 1.2. Syntax Notation 213 This specification uses the Augmented Backus-Naur Form (ABNF) 214 notation of [RFC5234], extended with the notation for case- 215 sensitivity in strings defined in [RFC7405]. 217 It also uses a list extension, defined in Section 5.5 of [Semantics], 218 that allows for compact definition of comma-separated lists using a 219 '#' operator (similar to how the '*' operator indicates repetition). 220 Appendix A shows the collected grammar with all list operators 221 expanded to standard ABNF notation. 223 As a convention, ABNF rule names prefixed with "obs-" denote 224 "obsolete" grammar rules that appear for historical reasons. 226 The following core rules are included by reference, as defined in 227 [RFC5234], Appendix B.1: ALPHA (letters), CR (carriage return), CRLF 228 (CR LF), CTL (controls), DIGIT (decimal 0-9), DQUOTE (double quote), 229 HEXDIG (hexadecimal 0-9/A-F/a-f), HTAB (horizontal tab), LF (line 230 feed), OCTET (any 8-bit sequence of data), SP (space), and VCHAR (any 231 visible [USASCII] character). 233 The rules below are defined in [Semantics]: 235 BWS = 236 OWS = 237 RWS = 238 absolute-URI = 239 absolute-path = 240 authority = 241 comment = 242 field-name = 243 field-value = 244 obs-text = 245 port = 246 query = 247 quoted-string = 248 token = 249 uri-host = 251 2. Message 253 2.1. Message Format 255 An HTTP/1.1 message consists of a start-line followed by a CRLF and a 256 sequence of octets in a format similar to the Internet Message Format 257 [RFC5322]: zero or more header field lines (collectively referred to 258 as the "headers" or the "header section"), an empty line indicating 259 the end of the header section, and an optional message body. 261 HTTP-message = start-line CRLF 262 *( field-line CRLF ) 263 CRLF 264 [ message-body ] 266 A message can be either a request from client to server or a response 267 from server to client. Syntactically, the two types of message 268 differ only in the start-line, which is either a request-line (for 269 requests) or a status-line (for responses), and in the algorithm for 270 determining the length of the message body (Section 6). 272 start-line = request-line / status-line 274 In theory, a client could receive requests and a server could receive 275 responses, distinguishing them by their different start-line formats. 276 In practice, servers are implemented to only expect a request (a 277 response is interpreted as an unknown or invalid request method) and 278 clients are implemented to only expect a response. 280 Although HTTP makes use of some protocol elements similar to the 281 Multipurpose Internet Mail Extensions (MIME) [RFC2045], see 282 Appendix B for the differences between HTTP and MIME messages. 284 2.2. Message Parsing 286 The normal procedure for parsing an HTTP message is to read the 287 start-line into a structure, read each header field line into a hash 288 table by field name until the empty line, and then use the parsed 289 data to determine if a message body is expected. If a message body 290 has been indicated, then it is read as a stream until an amount of 291 octets equal to the message body length is read or the connection is 292 closed. 294 A recipient MUST parse an HTTP message as a sequence of octets in an 295 encoding that is a superset of US-ASCII [USASCII]. Parsing an HTTP 296 message as a stream of Unicode characters, without regard for the 297 specific encoding, creates security vulnerabilities due to the 298 varying ways that string processing libraries handle invalid 299 multibyte character sequences that contain the octet LF (%x0A). 300 String-based parsers can only be safely used within protocol elements 301 after the element has been extracted from the message, such as within 302 a header field line value after message parsing has delineated the 303 individual field lines. 305 Although the line terminator for the start-line and header fields is 306 the sequence CRLF, a recipient MAY recognize a single LF as a line 307 terminator and ignore any preceding CR. 309 A sender MUST NOT generate a bare CR (a CR character not immediately 310 followed by LF) within any protocol elements other than the payload 311 body. A recipient of such a bare CR MUST consider that element to be 312 invalid or replace each bare CR with SP before processing the element 313 or forwarding the message. 315 Older HTTP/1.0 user agent implementations might send an extra CRLF 316 after a POST request as a workaround for some early server 317 applications that failed to read message body content that was not 318 terminated by a line-ending. An HTTP/1.1 user agent MUST NOT preface 319 or follow a request with an extra CRLF. If terminating the request 320 message body with a line-ending is desired, then the user agent MUST 321 count the terminating CRLF octets as part of the message body length. 323 In the interest of robustness, a server that is expecting to receive 324 and parse a request-line SHOULD ignore at least one empty line (CRLF) 325 received prior to the request-line. 327 A sender MUST NOT send whitespace between the start-line and the 328 first header field. A recipient that receives whitespace between the 329 start-line and the first header field MUST either reject the message 330 as invalid or consume each whitespace-preceded line without further 331 processing of it (i.e., ignore the entire line, along with any 332 subsequent lines preceded by whitespace, until a properly formed 333 header field is received or the header section is terminated). 335 The presence of such whitespace in a request might be an attempt to 336 trick a server into ignoring that field line or processing the line 337 after it as a new request, either of which might result in a security 338 vulnerability if other implementations within the request chain 339 interpret the same message differently. Likewise, the presence of 340 such whitespace in a response might be ignored by some clients or 341 cause others to cease parsing. 343 When a server listening only for HTTP request messages, or processing 344 what appears from the start-line to be an HTTP request message, 345 receives a sequence of octets that does not match the HTTP-message 346 grammar aside from the robustness exceptions listed above, the server 347 SHOULD respond with a 400 (Bad Request) response. 349 2.3. HTTP Version 351 HTTP uses a "." numbering scheme to indicate versions 352 of the protocol. This specification defines version "1.1". 353 Section 4.2 of [Semantics] specifies the semantics of HTTP version 354 numbers. 356 The version of an HTTP/1.x message is indicated by an HTTP-version 357 field in the start-line. HTTP-version is case-sensitive. 359 HTTP-version = HTTP-name "/" DIGIT "." DIGIT 360 HTTP-name = %s"HTTP" 362 When an HTTP/1.1 message is sent to an HTTP/1.0 recipient [RFC1945] 363 or a recipient whose version is unknown, the HTTP/1.1 message is 364 constructed such that it can be interpreted as a valid HTTP/1.0 365 message if all of the newer features are ignored. This specification 366 places recipient-version requirements on some new features so that a 367 conformant sender will only use compatible features until it has 368 determined, through configuration or the receipt of a message, that 369 the recipient supports HTTP/1.1. 371 Intermediaries that process HTTP messages (i.e., all intermediaries 372 other than those acting as tunnels) MUST send their own HTTP-version 373 in forwarded messages. In other words, they are not allowed to 374 blindly forward the start-line without ensuring that the protocol 375 version in that message matches a version to which that intermediary 376 is conformant for both the receiving and sending of messages. 377 Forwarding an HTTP message without rewriting the HTTP-version might 378 result in communication errors when downstream recipients use the 379 message sender's version to determine what features are safe to use 380 for later communication with that sender. 382 A server MAY send an HTTP/1.0 response to an HTTP/1.1 request if it 383 is known or suspected that the client incorrectly implements the HTTP 384 specification and is incapable of correctly processing later version 385 responses, such as when a client fails to parse the version number 386 correctly or when an intermediary is known to blindly forward the 387 HTTP-version even when it doesn't conform to the given minor version 388 of the protocol. Such protocol downgrades SHOULD NOT be performed 389 unless triggered by specific client attributes, such as when one or 390 more of the request header fields (e.g., User-Agent) uniquely match 391 the values sent by a client known to be in error. 393 3. Request Line 395 A request-line begins with a method token, followed by a single space 396 (SP), the request-target, another single space (SP), and ends with 397 the protocol version. 399 request-line = method SP request-target SP HTTP-version 401 Although the request-line grammar rule requires that each of the 402 component elements be separated by a single SP octet, recipients MAY 403 instead parse on whitespace-delimited word boundaries and, aside from 404 the CRLF terminator, treat any form of whitespace as the SP separator 405 while ignoring preceding or trailing whitespace; such whitespace 406 includes one or more of the following octets: SP, HTAB, VT (%x0B), FF 407 (%x0C), or bare CR. However, lenient parsing can result in request 408 smuggling security vulnerabilities if there are multiple recipients 409 of the message and each has its own unique interpretation of 410 robustness (see Section 11.2). 412 HTTP does not place a predefined limit on the length of a request- 413 line, as described in Section 3 of [Semantics]. A server that 414 receives a method longer than any that it implements SHOULD respond 415 with a 501 (Not Implemented) status code. A server that receives a 416 request-target longer than any URI it wishes to parse MUST respond 417 with a 414 (URI Too Long) status code (see Section 10.5.15 of 418 [Semantics]). 420 Various ad hoc limitations on request-line length are found in 421 practice. It is RECOMMENDED that all HTTP senders and recipients 422 support, at a minimum, request-line lengths of 8000 octets. 424 3.1. Method 426 The method token indicates the request method to be performed on the 427 target resource. The request method is case-sensitive. 429 method = token 431 The request methods defined by this specification can be found in 432 Section 8 of [Semantics], along with information regarding the HTTP 433 method registry and considerations for defining new methods. 435 3.2. Request Target 437 The request-target identifies the target resource upon which to apply 438 the request. The client derives a request-target from its desired 439 target URI. There are four distinct formats for the request-target, 440 depending on both the method being requested and whether the request 441 is to a proxy. 443 request-target = origin-form 444 / absolute-form 445 / authority-form 446 / asterisk-form 448 No whitespace is allowed in the request-target. Unfortunately, some 449 user agents fail to properly encode or exclude whitespace found in 450 hypertext references, resulting in those disallowed characters being 451 sent as the request-target in a malformed request-line. 453 Recipients of an invalid request-line SHOULD respond with either a 454 400 (Bad Request) error or a 301 (Moved Permanently) redirect with 455 the request-target properly encoded. A recipient SHOULD NOT attempt 456 to autocorrect and then process the request without a redirect, since 457 the invalid request-line might be deliberately crafted to bypass 458 security filters along the request chain. 460 3.2.1. origin-form 462 The most common form of request-target is the origin-form. 464 origin-form = absolute-path [ "?" query ] 466 When making a request directly to an origin server, other than a 467 CONNECT or server-wide OPTIONS request (as detailed below), a client 468 MUST send only the absolute path and query components of the target 469 URI as the request-target. If the target URI's path component is 470 empty, the client MUST send "/" as the path within the origin-form of 471 request-target. A Host header field is also sent, as defined in 472 Section 6.6 of [Semantics]. 474 For example, a client wishing to retrieve a representation of the 475 resource identified as 477 http://www.example.org/where?q=now 479 directly from the origin server would open (or reuse) a TCP 480 connection to port 80 of the host "www.example.org" and send the 481 lines: 483 GET /where?q=now HTTP/1.1 484 Host: www.example.org 486 followed by the remainder of the request message. 488 3.2.2. absolute-form 490 When making a request to a proxy, other than a CONNECT or server-wide 491 OPTIONS request (as detailed below), a client MUST send the target 492 URI in absolute-form as the request-target. 494 absolute-form = absolute-URI 496 The proxy is requested to either service that request from a valid 497 cache, if possible, or make the same request on the client's behalf 498 to either the next inbound proxy server or directly to the origin 499 server indicated by the request-target. Requirements on such 500 "forwarding" of messages are defined in Section 6.7 of [Semantics]. 502 An example absolute-form of request-line would be: 504 GET http://www.example.org/pub/WWW/TheProject.html HTTP/1.1 506 A client MUST send a Host header field in an HTTP/1.1 request even if 507 the request-target is in the absolute-form, since this allows the 508 Host information to be forwarded through ancient HTTP/1.0 proxies 509 that might not have implemented Host. 511 When a proxy receives a request with an absolute-form of request- 512 target, the proxy MUST ignore the received Host header field (if any) 513 and instead replace it with the host information of the request- 514 target. A proxy that forwards such a request MUST generate a new 515 Host field value based on the received request-target rather than 516 forward the received Host field value. 518 When an origin server receives a request with an absolute-form of 519 request-target, the origin server MUST ignore the received Host 520 header field (if any) and instead use the host information of the 521 request-target. Note that if the request-target does not have an 522 authority component, an empty Host header field will be sent in this 523 case. 525 To allow for transition to the absolute-form for all requests in some 526 future version of HTTP, a server MUST accept the absolute-form in 527 requests, even though HTTP/1.1 clients will only send them in 528 requests to proxies. 530 3.2.3. authority-form 532 The authority-form of request-target is only used for CONNECT 533 requests (Section 8.3.6 of [Semantics]). 535 authority-form = authority 537 When making a CONNECT request to establish a tunnel through one or 538 more proxies, a client MUST send only the target URI's authority 539 component (excluding any userinfo and its "@" delimiter) as the 540 request-target. For example, 542 CONNECT www.example.com:80 HTTP/1.1 544 3.2.4. asterisk-form 546 The asterisk-form of request-target is only used for a server-wide 547 OPTIONS request (Section 8.3.7 of [Semantics]). 549 asterisk-form = "*" 551 When a client wishes to request OPTIONS for the server as a whole, as 552 opposed to a specific named resource of that server, the client MUST 553 send only "*" (%x2A) as the request-target. For example, 555 OPTIONS * HTTP/1.1 557 If a proxy receives an OPTIONS request with an absolute-form of 558 request-target in which the URI has an empty path and no query 559 component, then the last proxy on the request chain MUST send a 560 request-target of "*" when it forwards the request to the indicated 561 origin server. 563 For example, the request 565 OPTIONS http://www.example.org:8001 HTTP/1.1 567 would be forwarded by the final proxy as 569 OPTIONS * HTTP/1.1 570 Host: www.example.org:8001 572 after connecting to port 8001 of host "www.example.org". 574 3.3. Reconstructing the Target URI 576 Since the request-target often contains only part of the user agent's 577 target URI, a server constructs its value to properly service the 578 request (Section 6.1 of [Semantics]). 580 If the request-target is in absolute-form, the target URI is the same 581 as the request-target. Otherwise, the target URI is constructed as 582 follows: 584 If the server's configuration (or outbound gateway) provides a 585 fixed URI scheme, that scheme is used for the target URI. 586 Otherwise, if the request is received over a TLS-secured TCP 587 connection, the target URI's scheme is "https"; if not, the scheme 588 is "http". 590 If the server's configuration (or outbound gateway) provides a 591 fixed URI authority component, that authority is used for the 592 target URI. If not, then if the request-target is in 593 authority-form, the target URI's authority component is the same 594 as the request-target. If not, then if a Host header field is 595 supplied with a non-empty field-value, the authority component is 596 the same as the Host field-value. Otherwise, the authority 597 component is assigned the default name configured for the server 598 and, if the connection's incoming TCP port number differs from the 599 default port for the target URI's scheme, then a colon (":") and 600 the incoming port number (in decimal form) are appended to the 601 authority component. 603 If the request-target is in authority-form or asterisk-form, the 604 target URI's combined path and query component is empty. 605 Otherwise, the combined path and query component is the same as 606 the request-target. 608 The components of the target URI, once determined as above, can be 609 combined into absolute-URI form by concatenating the scheme, 610 "://", authority, and combined path and query component. 612 Example 1: the following message received over an insecure TCP 613 connection 615 GET /pub/WWW/TheProject.html HTTP/1.1 616 Host: www.example.org:8080 618 has a target URI of 620 http://www.example.org:8080/pub/WWW/TheProject.html 622 Example 2: the following message received over a TLS-secured TCP 623 connection 625 OPTIONS * HTTP/1.1 626 Host: www.example.org 628 has a target URI of 630 https://www.example.org 632 Recipients of an HTTP/1.0 request that lacks a Host header field 633 might need to use heuristics (e.g., examination of the URI path for 634 something unique to a particular host) in order to guess the target 635 URI's authority component. 637 4. Status Line 639 The first line of a response message is the status-line, consisting 640 of the protocol version, a space (SP), the status code, another 641 space, and ending with an OPTIONAL textual phrase describing the 642 status code. 644 status-line = HTTP-version SP status-code SP [reason-phrase] 646 Although the status-line grammar rule requires that each of the 647 component elements be separated by a single SP octet, recipients MAY 648 instead parse on whitespace-delimited word boundaries and, aside from 649 the line terminator, treat any form of whitespace as the SP separator 650 while ignoring preceding or trailing whitespace; such whitespace 651 includes one or more of the following octets: SP, HTAB, VT (%x0B), FF 652 (%x0C), or bare CR. However, lenient parsing can result in response 653 splitting security vulnerabilities if there are multiple recipients 654 of the message and each has its own unique interpretation of 655 robustness (see Section 11.1). 657 The status-code element is a 3-digit integer code describing the 658 result of the server's attempt to understand and satisfy the client's 659 corresponding request. The rest of the response message is to be 660 interpreted in light of the semantics defined for that status code. 661 See Section 10 of [Semantics] for information about the semantics of 662 status codes, including the classes of status code (indicated by the 663 first digit), the status codes defined by this specification, 664 considerations for the definition of new status codes, and the IANA 665 registry. 667 status-code = 3DIGIT 669 The reason-phrase element exists for the sole purpose of providing a 670 textual description associated with the numeric status code, mostly 671 out of deference to earlier Internet application protocols that were 672 more frequently used with interactive text clients. 674 reason-phrase = 1*( HTAB / SP / VCHAR / obs-text ) 676 A client SHOULD ignore the reason-phrase content because it is not a 677 reliable channel for information (it might be translated for a given 678 locale, overwritten by intermediaries, or discarded when the message 679 is forwarded via other versions of HTTP). A server MUST send the 680 space that separates status-code from the reason-phrase even when the 681 reason-phrase is absent (i.e., the status-line would end with the 682 three octets SP CR LF). 684 5. Field Syntax 686 Each field line consists of a case-insensitive field name followed by 687 a colon (":"), optional leading whitespace, the field line value, and 688 optional trailing whitespace. 690 field-line = field-name ":" OWS field-value OWS 692 Most HTTP field names and the rules for parsing within field values 693 are defined in Section 5 of [Semantics]. This section covers the 694 generic syntax for header field inclusion within, and extraction 695 from, HTTP/1.1 messages. In addition, the following header fields 696 are defined by this document because they are specific to HTTP/1.1 697 message processing: 699 +-------------------+----------+--------------+ 700 | Field Name | Status | Reference | 701 | Connection | standard | Section 9.1 | 702 | MIME-Version | standard | Appendix B.1 | 703 | TE | standard | Section 7.4 | 704 | Transfer-Encoding | standard | Section 6.1 | 705 | Upgrade | standard | Section 9.9 | 706 +-------------------+----------+--------------+ 708 Table 1 710 Furthermore, the field name "Close" is reserved, since using that 711 name as an HTTP header field might conflict with the "close" 712 connection option of the Connection header field (Section 9.1). 714 +------------+----------+-----------+------------+ 715 | Field Name | Status | Reference | Comments | 716 | Close | standard | Section 5 | (reserved) | 717 +------------+----------+-----------+------------+ 719 Table 2 721 5.1. Field Line Parsing 723 Messages are parsed using a generic algorithm, independent of the 724 individual field names. The contents within a given field line value 725 are not parsed until a later stage of message interpretation (usually 726 after the message's entire header section has been processed). 728 No whitespace is allowed between the field name and colon. In the 729 past, differences in the handling of such whitespace have led to 730 security vulnerabilities in request routing and response handling. A 731 server MUST reject any received request message that contains 732 whitespace between a header field name and colon with a response 733 status code of 400 (Bad Request). A proxy MUST remove any such 734 whitespace from a response message before forwarding the message 735 downstream. 737 A field line value might be preceded and/or followed by optional 738 whitespace (OWS); a single SP preceding the field line value is 739 preferred for consistent readability by humans. The field line value 740 does not include any leading or trailing whitespace: OWS occurring 741 before the first non-whitespace octet of the field line value or 742 after the last non-whitespace octet of the field line value ought to 743 be excluded by parsers when extracting the field line value from a 744 header field line. 746 5.2. Obsolete Line Folding 748 Historically, HTTP header field line values could be extended over 749 multiple lines by preceding each extra line with at least one space 750 or horizontal tab (obs-fold). This specification deprecates such 751 line folding except within the message/http media type 752 (Section 10.1). 754 obs-fold = OWS CRLF RWS 755 ; obsolete line folding 757 A sender MUST NOT generate a message that includes line folding 758 (i.e., that has any field line value that contains a match to the 759 obs-fold rule) unless the message is intended for packaging within 760 the message/http media type. 762 A server that receives an obs-fold in a request message that is not 763 within a message/http container MUST either reject the message by 764 sending a 400 (Bad Request), preferably with a representation 765 explaining that obsolete line folding is unacceptable, or replace 766 each received obs-fold with one or more SP octets prior to 767 interpreting the field value or forwarding the message downstream. 769 A proxy or gateway that receives an obs-fold in a response message 770 that is not within a message/http container MUST either discard the 771 message and replace it with a 502 (Bad Gateway) response, preferably 772 with a representation explaining that unacceptable line folding was 773 received, or replace each received obs-fold with one or more SP 774 octets prior to interpreting the field value or forwarding the 775 message downstream. 777 A user agent that receives an obs-fold in a response message that is 778 not within a message/http container MUST replace each received 779 obs-fold with one or more SP octets prior to interpreting the field 780 value. 782 6. Message Body 784 The message body (if any) of an HTTP message is used to carry the 785 payload body (Section 7.3.3 of [Semantics]) of that request or 786 response. The message body is identical to the payload body unless a 787 transfer coding has been applied, as described in Section 6.1. 789 message-body = *OCTET 791 The rules for determining when a message body is present in an 792 HTTP/1.1 message differ for requests and responses. 794 The presence of a message body in a request is signaled by a 795 Content-Length or Transfer-Encoding header field. Request message 796 framing is independent of method semantics, even if the method does 797 not define any use for a message body. 799 The presence of a message body in a response depends on both the 800 request method to which it is responding and the response status code 801 (Section 4), and corresponds to when a payload body is allowed; see 802 Section 7.3.3 of [Semantics]. 804 6.1. Transfer-Encoding 806 The Transfer-Encoding header field lists the transfer coding names 807 corresponding to the sequence of transfer codings that have been (or 808 will be) applied to the payload body in order to form the message 809 body. Transfer codings are defined in Section 7. 811 Transfer-Encoding = 1#transfer-coding 813 Transfer-Encoding is analogous to the Content-Transfer-Encoding field 814 of MIME, which was designed to enable safe transport of binary data 815 over a 7-bit transport service ([RFC2045], Section 6). However, safe 816 transport has a different focus for an 8bit-clean transfer protocol. 817 In HTTP's case, Transfer-Encoding is primarily intended to accurately 818 delimit a dynamically generated payload and to distinguish payload 819 encodings that are only applied for transport efficiency or security 820 from those that are characteristics of the selected resource. 822 A recipient MUST be able to parse the chunked transfer coding 823 (Section 7.1) because it plays a crucial role in framing messages 824 when the payload body size is not known in advance. A sender MUST 825 NOT apply chunked more than once to a message body (i.e., chunking an 826 already chunked message is not allowed). If any transfer coding 827 other than chunked is applied to a request payload body, the sender 828 MUST apply chunked as the final transfer coding to ensure that the 829 message is properly framed. If any transfer coding other than 830 chunked is applied to a response payload body, the sender MUST either 831 apply chunked as the final transfer coding or terminate the message 832 by closing the connection. 834 For example, 836 Transfer-Encoding: gzip, chunked 838 indicates that the payload body has been compressed using the gzip 839 coding and then chunked using the chunked coding while forming the 840 message body. 842 Unlike Content-Encoding (Section 7.1.2 of [Semantics]), Transfer- 843 Encoding is a property of the message, not of the representation, and 844 any recipient along the request/response chain MAY decode the 845 received transfer coding(s) or apply additional transfer coding(s) to 846 the message body, assuming that corresponding changes are made to the 847 Transfer-Encoding field value. Additional information about the 848 encoding parameters can be provided by other header fields not 849 defined by this specification. 851 Transfer-Encoding MAY be sent in a response to a HEAD request or in a 852 304 (Not Modified) response (Section 10.4.5 of [Semantics]) to a GET 853 request, neither of which includes a message body, to indicate that 854 the origin server would have applied a transfer coding to the message 855 body if the request had been an unconditional GET. This indication 856 is not required, however, because any recipient on the response chain 857 (including the origin server) can remove transfer codings when they 858 are not needed. 860 A server MUST NOT send a Transfer-Encoding header field in any 861 response with a status code of 1xx (Informational) or 204 (No 862 Content). A server MUST NOT send a Transfer-Encoding header field in 863 any 2xx (Successful) response to a CONNECT request (Section 8.3.6 of 864 [Semantics]). 866 Transfer-Encoding was added in HTTP/1.1. It is generally assumed 867 that implementations advertising only HTTP/1.0 support will not 868 understand how to process a transfer-encoded payload. A client MUST 869 NOT send a request containing Transfer-Encoding unless it knows the 870 server will handle HTTP/1.1 requests (or later minor revisions); such 871 knowledge might be in the form of specific user configuration or by 872 remembering the version of a prior received response. A server MUST 873 NOT send a response containing Transfer-Encoding unless the 874 corresponding request indicates HTTP/1.1 (or later minor revisions). 876 A server that receives a request message with a transfer coding it 877 does not understand SHOULD respond with 501 (Not Implemented). 879 6.2. Content-Length 881 When a message does not have a Transfer-Encoding header field, a 882 Content-Length header field can provide the anticipated size, as a 883 decimal number of octets, for a potential payload body. For messages 884 that do include a payload body, the Content-Length field value 885 provides the framing information necessary for determining where the 886 body (and message) ends. For messages that do not include a payload 887 body, the Content-Length indicates the size of the selected 888 representation (Section 7.2.4 of [Semantics]). 890 | *Note:* HTTP's use of Content-Length for message framing 891 | differs significantly from the same field's use in MIME, where 892 | it is an optional field used only within the "message/external- 893 | body" media-type. 895 6.3. Message Body Length 897 The length of a message body is determined by one of the following 898 (in order of precedence): 900 1. Any response to a HEAD request and any response with a 1xx 901 (Informational), 204 (No Content), or 304 (Not Modified) status 902 code is always terminated by the first empty line after the 903 header fields, regardless of the header fields present in the 904 message, and thus cannot contain a message body. 906 2. Any 2xx (Successful) response to a CONNECT request implies that 907 the connection will become a tunnel immediately after the empty 908 line that concludes the header fields. A client MUST ignore any 909 Content-Length or Transfer-Encoding header fields received in 910 such a message. 912 3. If a Transfer-Encoding header field is present and the chunked 913 transfer coding (Section 7.1) is the final encoding, the message 914 body length is determined by reading and decoding the chunked 915 data until the transfer coding indicates the data is complete. 917 If a Transfer-Encoding header field is present in a response and 918 the chunked transfer coding is not the final encoding, the 919 message body length is determined by reading the connection until 920 it is closed by the server. If a Transfer-Encoding header field 921 is present in a request and the chunked transfer coding is not 922 the final encoding, the message body length cannot be determined 923 reliably; the server MUST respond with the 400 (Bad Request) 924 status code and then close the connection. 926 If a message is received with both a Transfer-Encoding and a 927 Content-Length header field, the Transfer-Encoding overrides the 928 Content-Length. Such a message might indicate an attempt to 929 perform request smuggling (Section 11.2) or response splitting 930 (Section 11.1) and ought to be handled as an error. A sender 931 MUST remove the received Content-Length field prior to forwarding 932 such a message downstream. 934 4. If a message is received without Transfer-Encoding and with an 935 invalid Content-Length header field, then the message framing is 936 invalid and the recipient MUST treat it as an unrecoverable 937 error, unless the field value can be successfully parsed as a 938 comma-separated list (Section 5.5 of [Semantics]), all values in 939 the list are valid, and all values in the list are the same. If 940 this is a request message, the server MUST respond with a 400 941 (Bad Request) status code and then close the connection. If this 942 is a response message received by a proxy, the proxy MUST close 943 the connection to the server, discard the received response, and 944 send a 502 (Bad Gateway) response to the client. If this is a 945 response message received by a user agent, the user agent MUST 946 close the connection to the server and discard the received 947 response. 949 5. If a valid Content-Length header field is present without 950 Transfer-Encoding, its decimal value defines the expected message 951 body length in octets. If the sender closes the connection or 952 the recipient times out before the indicated number of octets are 953 received, the recipient MUST consider the message to be 954 incomplete and close the connection. 956 6. If this is a request message and none of the above are true, then 957 the message body length is zero (no message body is present). 959 7. Otherwise, this is a response message without a declared message 960 body length, so the message body length is determined by the 961 number of octets received prior to the server closing the 962 connection. 964 Since there is no way to distinguish a successfully completed, close- 965 delimited message from a partially received message interrupted by 966 network failure, a server SHOULD generate encoding or length- 967 delimited messages whenever possible. The close-delimiting feature 968 exists primarily for backwards compatibility with HTTP/1.0. 970 A server MAY reject a request that contains a message body but not a 971 Content-Length by responding with 411 (Length Required). 973 Unless a transfer coding other than chunked has been applied, a 974 client that sends a request containing a message body SHOULD use a 975 valid Content-Length header field if the message body length is known 976 in advance, rather than the chunked transfer coding, since some 977 existing services respond to chunked with a 411 (Length Required) 978 status code even though they understand the chunked transfer coding. 979 This is typically because such services are implemented via a gateway 980 that requires a content-length in advance of being called and the 981 server is unable or unwilling to buffer the entire request before 982 processing. 984 A user agent that sends a request containing a message body MUST send 985 a valid Content-Length header field if it does not know the server 986 will handle HTTP/1.1 (or later) requests; such knowledge can be in 987 the form of specific user configuration or by remembering the version 988 of a prior received response. 990 If the final response to the last request on a connection has been 991 completely received and there remains additional data to read, a user 992 agent MAY discard the remaining data or attempt to determine if that 993 data belongs as part of the prior response body, which might be the 994 case if the prior message's Content-Length value is incorrect. A 995 client MUST NOT process, cache, or forward such extra data as a 996 separate response, since such behavior would be vulnerable to cache 997 poisoning. 999 7. Transfer Codings 1001 Transfer coding names are used to indicate an encoding transformation 1002 that has been, can be, or might need to be applied to a payload body 1003 in order to ensure "safe transport" through the network. This 1004 differs from a content coding in that the transfer coding is a 1005 property of the message rather than a property of the representation 1006 that is being transferred. 1008 transfer-coding = token *( OWS ";" OWS transfer-parameter ) 1010 Parameters are in the form of a name=value pair. 1012 transfer-parameter = token BWS "=" BWS ( token / quoted-string ) 1014 All transfer-coding names are case-insensitive and ought to be 1015 registered within the HTTP Transfer Coding registry, as defined in 1016 Section 7.3. They are used in the TE (Section 7.4) and 1017 Transfer-Encoding (Section 6.1) header fields. 1019 +------------+-------------------------------+-----------+ 1020 | Name | Description | Reference | 1021 | chunked | Transfer in a series of | Section | 1022 | | chunks | 7.1 | 1023 | compress | UNIX "compress" data format | Section | 1024 | | [Welch] | 7.2 | 1025 | deflate | "deflate" compressed data | Section | 1026 | | ([RFC1951]) inside the "zlib" | 7.2 | 1027 | | data format ([RFC1950]) | | 1028 | gzip | GZIP file format [RFC1952] | Section | 1029 | | | 7.2 | 1030 | trailers | (reserved) | Section 7 | 1031 | x-compress | Deprecated (alias for | Section | 1032 | | compress) | 7.2 | 1033 | x-gzip | Deprecated (alias for gzip) | Section | 1034 | | | 7.2 | 1035 +------------+-------------------------------+-----------+ 1037 Table 3 1039 | *Note:* the coding name "trailers" is reserved because its use 1040 | would conflict with the keyword "trailers" in the TE header 1041 | field (Section 7.4). 1043 7.1. Chunked Transfer Coding 1045 The chunked transfer coding wraps the payload body in order to 1046 transfer it as a series of chunks, each with its own size indicator, 1047 followed by an OPTIONAL trailer section containing trailer fields. 1048 Chunked enables content streams of unknown size to be transferred as 1049 a sequence of length-delimited buffers, which enables the sender to 1050 retain connection persistence and the recipient to know when it has 1051 received the entire message. 1053 chunked-body = *chunk 1054 last-chunk 1055 trailer-section 1056 CRLF 1058 chunk = chunk-size [ chunk-ext ] CRLF 1059 chunk-data CRLF 1060 chunk-size = 1*HEXDIG 1061 last-chunk = 1*("0") [ chunk-ext ] CRLF 1063 chunk-data = 1*OCTET ; a sequence of chunk-size octets 1065 The chunk-size field is a string of hex digits indicating the size of 1066 the chunk-data in octets. The chunked transfer coding is complete 1067 when a chunk with a chunk-size of zero is received, possibly followed 1068 by a trailer section, and finally terminated by an empty line. 1070 A recipient MUST be able to parse and decode the chunked transfer 1071 coding. 1073 Note that HTTP/1.1 does not define any means to limit the size of a 1074 chunked response such that an intermediary can be assured of 1075 buffering the entire response. 1077 The chunked encoding does not define any parameters. Their presence 1078 SHOULD be treated as an error. 1080 7.1.1. Chunk Extensions 1082 The chunked encoding allows each chunk to include zero or more chunk 1083 extensions, immediately following the chunk-size, for the sake of 1084 supplying per-chunk metadata (such as a signature or hash), mid- 1085 message control information, or randomization of message body size. 1087 chunk-ext = *( BWS ";" BWS chunk-ext-name 1088 [ BWS "=" BWS chunk-ext-val ] ) 1090 chunk-ext-name = token 1091 chunk-ext-val = token / quoted-string 1093 The chunked encoding is specific to each connection and is likely to 1094 be removed or recoded by each recipient (including intermediaries) 1095 before any higher-level application would have a chance to inspect 1096 the extensions. Hence, use of chunk extensions is generally limited 1097 to specialized HTTP services such as "long polling" (where client and 1098 server can have shared expectations regarding the use of chunk 1099 extensions) or for padding within an end-to-end secured connection. 1101 A recipient MUST ignore unrecognized chunk extensions. A server 1102 ought to limit the total length of chunk extensions received in a 1103 request to an amount reasonable for the services provided, in the 1104 same way that it applies length limitations and timeouts for other 1105 parts of a message, and generate an appropriate 4xx (Client Error) 1106 response if that amount is exceeded. 1108 7.1.2. Chunked Trailer Section 1110 A trailer section allows the sender to include additional fields at 1111 the end of a chunked message in order to supply metadata that might 1112 be dynamically generated while the message body is sent, such as a 1113 message integrity check, digital signature, or post-processing 1114 status. The proper use and limitations of trailer fields are defined 1115 in Section 5.6 of [Semantics]. 1117 trailer-section = *( field-line CRLF ) 1119 A recipient that decodes and removes the chunked encoding from a 1120 message (e.g., for storage or forwarding to a non-HTTP/1.1 peer) MUST 1121 discard any received trailer fields, store/forward them separately 1122 from the header fields, or selectively merge into the header section 1123 only those trailer fields corresponding to header field definitions 1124 that are understood by the recipient to explicitly permit and define 1125 how their corresponding trailer field value can be safely merged. 1127 7.1.3. Decoding Chunked 1129 A process for decoding the chunked transfer coding can be represented 1130 in pseudo-code as: 1132 length := 0 1133 read chunk-size, chunk-ext (if any), and CRLF 1134 while (chunk-size > 0) { 1135 read chunk-data and CRLF 1136 append chunk-data to decoded-body 1137 length := length + chunk-size 1138 read chunk-size, chunk-ext (if any), and CRLF 1139 } 1140 read trailer field 1141 while (trailer field is not empty) { 1142 if (trailer fields are stored/forwarded separately) { 1143 append trailer field to existing trailer fields 1144 } 1145 else if (trailer field is understood and defined as mergeable) { 1146 merge trailer field with existing header fields 1147 } 1148 else { 1149 discard trailer field 1150 } 1151 read trailer field 1152 } 1153 Content-Length := length 1154 Remove "chunked" from Transfer-Encoding 1155 Remove Trailer from existing header fields 1157 7.2. Transfer Codings for Compression 1159 The following transfer coding names for compression are defined by 1160 the same algorithm as their corresponding content coding: 1162 compress (and x-compress) 1163 See Section 7.1.2.1 of [Semantics]. 1165 deflate 1166 See Section 7.1.2.2 of [Semantics]. 1168 gzip (and x-gzip) 1169 See Section 7.1.2.3 of [Semantics]. 1171 The compression codings do not define any parameters. Their presence 1172 SHOULD be treated as an error. 1174 7.3. Transfer Coding Registry 1176 The "HTTP Transfer Coding Registry" defines the namespace for 1177 transfer coding names. It is maintained at 1178 . 1180 Registrations MUST include the following fields: 1182 o Name 1184 o Description 1186 o Pointer to specification text 1188 Names of transfer codings MUST NOT overlap with names of content 1189 codings (Section 7.1.2 of [Semantics]) unless the encoding 1190 transformation is identical, as is the case for the compression 1191 codings defined in Section 7.2. 1193 The TE header field (Section 7.4) uses a pseudo parameter named "q" 1194 as rank value when multiple transfer codings are acceptable. Future 1195 registrations of transfer codings SHOULD NOT define parameters called 1196 "q" (case-insensitively) in order to avoid ambiguities. 1198 Values to be added to this namespace require IETF Review (see 1199 Section 4.8 of [RFC8126]), and MUST conform to the purpose of 1200 transfer coding defined in this specification. 1202 Use of program names for the identification of encoding formats is 1203 not desirable and is discouraged for future encodings. 1205 7.4. TE 1207 The "TE" header field in a request indicates what transfer codings, 1208 besides chunked, the client is willing to accept in response, and 1209 whether or not the client is willing to accept trailer fields in a 1210 chunked transfer coding. 1212 The TE field-value consists of a list of transfer coding names, each 1213 allowing for optional parameters (as described in Section 7), and/or 1214 the keyword "trailers". A client MUST NOT send the chunked transfer 1215 coding name in TE; chunked is always acceptable for HTTP/1.1 1216 recipients. 1218 TE = #t-codings 1219 t-codings = "trailers" / ( transfer-coding [ t-ranking ] ) 1220 t-ranking = OWS ";" OWS "q=" rank 1221 rank = ( "0" [ "." 0*3DIGIT ] ) 1222 / ( "1" [ "." 0*3("0") ] ) 1224 Three examples of TE use are below. 1226 TE: deflate 1227 TE: 1228 TE: trailers, deflate;q=0.5 1230 When multiple transfer codings are acceptable, the client MAY rank 1231 the codings by preference using a case-insensitive "q" parameter 1232 (similar to the qvalues used in content negotiation fields, 1233 Section 7.4.4 of [Semantics]). The rank value is a real number in 1234 the range 0 through 1, where 0.001 is the least preferred and 1 is 1235 the most preferred; a value of 0 means "not acceptable". 1237 If the TE field value is empty or if no TE field is present, the only 1238 acceptable transfer coding is chunked. A message with no transfer 1239 coding is always acceptable. 1241 The keyword "trailers" indicates that the sender will not discard 1242 trailer fields, as described in Section 5.6 of [Semantics]. 1244 Since the TE header field only applies to the immediate connection, a 1245 sender of TE MUST also send a "TE" connection option within the 1246 Connection header field (Section 9.1) in order to prevent the TE 1247 field from being forwarded by intermediaries that do not support its 1248 semantics. 1250 8. Handling Incomplete Messages 1252 A server that receives an incomplete request message, usually due to 1253 a canceled request or a triggered timeout exception, MAY send an 1254 error response prior to closing the connection. 1256 A client that receives an incomplete response message, which can 1257 occur when a connection is closed prematurely or when decoding a 1258 supposedly chunked transfer coding fails, MUST record the message as 1259 incomplete. Cache requirements for incomplete responses are defined 1260 in Section 3 of [Caching]. 1262 If a response terminates in the middle of the header section (before 1263 the empty line is received) and the status code might rely on header 1264 fields to convey the full meaning of the response, then the client 1265 cannot assume that meaning has been conveyed; the client might need 1266 to repeat the request in order to determine what action to take next. 1268 A message body that uses the chunked transfer coding is incomplete if 1269 the zero-sized chunk that terminates the encoding has not been 1270 received. A message that uses a valid Content-Length is incomplete 1271 if the size of the message body received (in octets) is less than the 1272 value given by Content-Length. A response that has neither chunked 1273 transfer coding nor Content-Length is terminated by closure of the 1274 connection and, thus, is considered complete regardless of the number 1275 of message body octets received, provided that the header section was 1276 received intact. 1278 9. Connection Management 1280 HTTP messaging is independent of the underlying transport- or 1281 session-layer connection protocol(s). HTTP only presumes a reliable 1282 transport with in-order delivery of requests and the corresponding 1283 in-order delivery of responses. The mapping of HTTP request and 1284 response structures onto the data units of an underlying transport 1285 protocol is outside the scope of this specification. 1287 As described in Section 6.3 of [Semantics], the specific connection 1288 protocols to be used for an HTTP interaction are determined by client 1289 configuration and the target URI. For example, the "http" URI scheme 1290 (Section 2.5.1 of [Semantics]) indicates a default connection of TCP 1291 over IP, with a default TCP port of 80, but the client might be 1292 configured to use a proxy via some other connection, port, or 1293 protocol. 1295 HTTP implementations are expected to engage in connection management, 1296 which includes maintaining the state of current connections, 1297 establishing a new connection or reusing an existing connection, 1298 processing messages received on a connection, detecting connection 1299 failures, and closing each connection. Most clients maintain 1300 multiple connections in parallel, including more than one connection 1301 per server endpoint. Most servers are designed to maintain thousands 1302 of concurrent connections, while controlling request queues to enable 1303 fair use and detect denial-of-service attacks. 1305 9.1. Connection 1307 The "Connection" header field allows the sender to list desired 1308 control options for the current connection. 1310 When a field aside from Connection is used to supply control 1311 information for or about the current connection, the sender MUST list 1312 the corresponding field name within the Connection header field. 1314 Intermediaries MUST parse a received Connection header field before a 1315 message is forwarded and, for each connection-option in this field, 1316 remove any header or trailer field(s) from the message with the same 1317 name as the connection-option, and then remove the Connection header 1318 field itself (or replace it with the intermediary's own connection 1319 options for the forwarded message). 1321 Hence, the Connection header field provides a declarative way of 1322 distinguishing fields that are only intended for the immediate 1323 recipient ("hop-by-hop") from those fields that are intended for all 1324 recipients on the chain ("end-to-end"), enabling the message to be 1325 self-descriptive and allowing future connection-specific extensions 1326 to be deployed without fear that they will be blindly forwarded by 1327 older intermediaries. 1329 Furthermore, intermediaries SHOULD remove or replace field(s) whose 1330 semantics are known to require removal before forwarding, whether or 1331 not they appear as a Connection option, after applying those fields' 1332 semantics. This includes but is not limited to: 1334 o Proxy-Connection (Appendix C.1.2) 1336 o Keep-Alive (Section 19.7.1 of [RFC2068]) 1338 o TE (Section 7.4) 1340 o Trailer (Section 5.6.3 of [Semantics]) 1342 o Transfer-Encoding (Section 6.1) 1344 o Upgrade (Section 9.9) 1345 The Connection header field's value has the following grammar: 1347 Connection = 1#connection-option 1348 connection-option = token 1350 Connection options are case-insensitive. 1352 A sender MUST NOT send a connection option corresponding to a field 1353 that is intended for all recipients of the payload. For example, 1354 Cache-Control is never appropriate as a connection option 1355 (Section 5.2 of [Caching]). 1357 The connection options do not always correspond to a field present in 1358 the message, since a connection-specific field might not be needed if 1359 there are no parameters associated with a connection option. In 1360 contrast, a connection-specific field that is received without a 1361 corresponding connection option usually indicates that the field has 1362 been improperly forwarded by an intermediary and ought to be ignored 1363 by the recipient. 1365 When defining new connection options, specification authors ought to 1366 document it as reserved field name and register that definition in 1367 the Hypertext Transfer Protocol (HTTP) Field Name Registry 1368 (Section 5.3.2 of [Semantics]), to avoid collisions. 1370 The "close" connection option is defined for a sender to signal that 1371 this connection will be closed after completion of the response. For 1372 example, 1374 Connection: close 1376 in either the request or the response header fields indicates that 1377 the sender is going to close the connection after the current 1378 request/response is complete (Section 9.7). 1380 A client that does not support persistent connections MUST send the 1381 "close" connection option in every request message. 1383 A server that does not support persistent connections MUST send the 1384 "close" connection option in every response message that does not 1385 have a 1xx (Informational) status code. 1387 9.2. Establishment 1389 It is beyond the scope of this specification to describe how 1390 connections are established via various transport- or session-layer 1391 protocols. Each connection applies to only one transport link. 1393 9.3. Associating a Response to a Request 1395 HTTP/1.1 does not include a request identifier for associating a 1396 given request message with its corresponding one or more response 1397 messages. Hence, it relies on the order of response arrival to 1398 correspond exactly to the order in which requests are made on the 1399 same connection. More than one response message per request only 1400 occurs when one or more informational responses (1xx, see 1401 Section 10.2 of [Semantics]) precede a final response to the same 1402 request. 1404 A client that has more than one outstanding request on a connection 1405 MUST maintain a list of outstanding requests in the order sent and 1406 MUST associate each received response message on that connection to 1407 the highest ordered request that has not yet received a final (non- 1408 1xx) response. 1410 If an HTTP/1.1 client receives data on a connection that doesn't have 1411 any outstanding requests, it MUST NOT consider them to be a response 1412 to a not-yet-issued request; it SHOULD close the connection, since 1413 message delimitation is now ambiguous, unless the data consists only 1414 of one or more CRLF (which can be discarded, as per Section 2.2). 1416 9.4. Persistence 1418 HTTP/1.1 defaults to the use of "persistent connections", allowing 1419 multiple requests and responses to be carried over a single 1420 connection. The "close" connection option is used to signal that a 1421 connection will not persist after the current request/response. HTTP 1422 implementations SHOULD support persistent connections. 1424 A recipient determines whether a connection is persistent or not 1425 based on the most recently received message's protocol version and 1426 Connection header field (if any): 1428 o If the "close" connection option is present, the connection will 1429 not persist after the current response; else, 1431 o If the received protocol is HTTP/1.1 (or later), the connection 1432 will persist after the current response; else, 1434 o If the received protocol is HTTP/1.0, the "keep-alive" connection 1435 option is present, either the recipient is not a proxy or the 1436 message is a response, and the recipient wishes to honor the 1437 HTTP/1.0 "keep-alive" mechanism, the connection will persist after 1438 the current response; otherwise, 1440 o The connection will close after the current response. 1442 A client MAY send additional requests on a persistent connection 1443 until it sends or receives a "close" connection option or receives an 1444 HTTP/1.0 response without a "keep-alive" connection option. 1446 In order to remain persistent, all messages on a connection need to 1447 have a self-defined message length (i.e., one not defined by closure 1448 of the connection), as described in Section 6. A server MUST read 1449 the entire request message body or close the connection after sending 1450 its response, since otherwise the remaining data on a persistent 1451 connection would be misinterpreted as the next request. Likewise, a 1452 client MUST read the entire response message body if it intends to 1453 reuse the same connection for a subsequent request. 1455 A proxy server MUST NOT maintain a persistent connection with an 1456 HTTP/1.0 client (see Section 19.7.1 of [RFC2068] for information and 1457 discussion of the problems with the Keep-Alive header field 1458 implemented by many HTTP/1.0 clients). 1460 See Appendix C.1.2 for more information on backwards compatibility 1461 with HTTP/1.0 clients. 1463 9.4.1. Retrying Requests 1465 Connections can be closed at any time, with or without intention. 1466 Implementations ought to anticipate the need to recover from 1467 asynchronous close events. The conditions under which a client can 1468 automatically retry a sequence of outstanding requests are defined in 1469 Section 8.2.2 of [Semantics]. 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 8.2.1 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 8.2.2 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. TLS Connection Closure 1619 TLS provides a facility for secure connection closure. When a valid 1620 closure alert is received, an implementation can be assured that no 1621 further data will be received on that connection. TLS 1622 implementations MUST initiate an exchange of closure alerts before 1623 closing a connection. A TLS implementation MAY, after sending a 1624 closure alert, close the connection without waiting for the peer to 1625 send its closure alert, generating an "incomplete close". Note that 1626 an implementation which does this MAY choose to reuse the session. 1627 This SHOULD only be done when the application knows (typically 1628 through detecting HTTP message boundaries) that it has received all 1629 the message data that it cares about. 1631 As specified in [RFC8446], any implementation which receives a 1632 connection close without first receiving a valid closure alert (a 1633 "premature close") MUST NOT reuse that session. Note that a 1634 premature close does not call into question the security of the data 1635 already received, but simply indicates that subsequent data might 1636 have been truncated. Because TLS is oblivious to HTTP request/ 1637 response boundaries, it is necessary to examine the HTTP data itself 1638 (specifically the Content-Length header) to determine whether the 1639 truncation occurred inside a message or between messages. 1641 When encountering a premature close, a client SHOULD treat as 1642 completed all requests for which it has received as much data as 1643 specified in the Content-Length header. 1645 A client detecting an incomplete close SHOULD recover gracefully. It 1646 MAY resume a TLS session closed in this fashion. 1648 Clients MUST send a closure alert before closing the connection. 1649 Clients which are unprepared to receive any more data MAY choose not 1650 to wait for the server's closure alert and simply close the 1651 connection, thus generating an incomplete close on the server side. 1653 Servers SHOULD be prepared to receive an incomplete close from the 1654 client, since the client can often determine when the end of server 1655 data is. Servers SHOULD be willing to resume TLS sessions closed in 1656 this fashion. 1658 Servers MUST attempt to initiate an exchange of closure alerts with 1659 the client before closing the connection. Servers MAY close the 1660 connection after sending the closure alert, thus generating an 1661 incomplete close on the client side. 1663 9.9. Upgrade 1665 The "Upgrade" header field is intended to provide a simple mechanism 1666 for transitioning from HTTP/1.1 to some other protocol on the same 1667 connection. 1669 A client MAY send a list of protocol names in the Upgrade header 1670 field of a request to invite the server to switch to one or more of 1671 the named protocols, in order of descending preference, before 1672 sending the final response. A server MAY ignore a received Upgrade 1673 header field if it wishes to continue using the current protocol on 1674 that connection. Upgrade cannot be used to insist on a protocol 1675 change. 1677 Upgrade = 1#protocol 1679 protocol = protocol-name ["/" protocol-version] 1680 protocol-name = token 1681 protocol-version = token 1683 Although protocol names are registered with a preferred case, 1684 recipients SHOULD use case-insensitive comparison when matching each 1685 protocol-name to supported protocols. 1687 A server that sends a 101 (Switching Protocols) response MUST send an 1688 Upgrade header field to indicate the new protocol(s) to which the 1689 connection is being switched; if multiple protocol layers are being 1690 switched, the sender MUST list the protocols in layer-ascending 1691 order. A server MUST NOT switch to a protocol that was not indicated 1692 by the client in the corresponding request's Upgrade header field. A 1693 server MAY choose to ignore the order of preference indicated by the 1694 client and select the new protocol(s) based on other factors, such as 1695 the nature of the request or the current load on the server. 1697 A server that sends a 426 (Upgrade Required) response MUST send an 1698 Upgrade header field to indicate the acceptable protocols, in order 1699 of descending preference. 1701 A server MAY send an Upgrade header field in any other response to 1702 advertise that it implements support for upgrading to the listed 1703 protocols, in order of descending preference, when appropriate for a 1704 future request. 1706 The following is a hypothetical example sent by a client: 1708 GET /hello HTTP/1.1 1709 Host: www.example.com 1710 Connection: upgrade 1711 Upgrade: websocket, IRC/6.9, RTA/x11 1713 The capabilities and nature of the application-level communication 1714 after the protocol change is entirely dependent upon the new 1715 protocol(s) chosen. However, immediately after sending the 101 1716 (Switching Protocols) response, the server is expected to continue 1717 responding to the original request as if it had received its 1718 equivalent within the new protocol (i.e., the server still has an 1719 outstanding request to satisfy after the protocol has been changed, 1720 and is expected to do so without requiring the request to be 1721 repeated). 1723 For example, if the Upgrade header field is received in a GET request 1724 and the server decides to switch protocols, it first responds with a 1725 101 (Switching Protocols) message in HTTP/1.1 and then immediately 1726 follows that with the new protocol's equivalent of a response to a 1727 GET on the target resource. This allows a connection to be upgraded 1728 to protocols with the same semantics as HTTP without the latency cost 1729 of an additional round trip. A server MUST NOT switch protocols 1730 unless the received message semantics can be honored by the new 1731 protocol; an OPTIONS request can be honored by any protocol. 1733 The following is an example response to the above hypothetical 1734 request: 1736 HTTP/1.1 101 Switching Protocols 1737 Connection: upgrade 1738 Upgrade: websocket 1740 [... data stream switches to websocket with an appropriate response 1741 (as defined by new protocol) to the "GET /hello" request ...] 1743 When Upgrade is sent, the sender MUST also send a Connection header 1744 field (Section 9.1) that contains an "upgrade" connection option, in 1745 order to prevent Upgrade from being accidentally forwarded by 1746 intermediaries that might not implement the listed protocols. A 1747 server MUST ignore an Upgrade header field that is received in an 1748 HTTP/1.0 request. 1750 A client cannot begin using an upgraded protocol on the connection 1751 until it has completely sent the request message (i.e., the client 1752 can't change the protocol it is sending in the middle of a message). 1753 If a server receives both an Upgrade and an Expect header field with 1754 the "100-continue" expectation (Section 9.1.1 of [Semantics]), the 1755 server MUST send a 100 (Continue) response before sending a 101 1756 (Switching Protocols) response. 1758 The Upgrade header field only applies to switching protocols on top 1759 of the existing connection; it cannot be used to switch the 1760 underlying connection (transport) protocol, nor to switch the 1761 existing communication to a different connection. For those 1762 purposes, it is more appropriate to use a 3xx (Redirection) response 1763 (Section 10.4 of [Semantics]). 1765 9.9.1. Upgrade Protocol Names 1767 This specification only defines the protocol name "HTTP" for use by 1768 the family of Hypertext Transfer Protocols, as defined by the HTTP 1769 version rules of Section 4.2 of [Semantics] and future updates to 1770 this specification. Additional protocol names ought to be registered 1771 using the registration procedure defined in Section 9.9.2. 1773 +------+-------------------+-----------------+----------------+ 1774 | Name | Description | Expected | Reference | 1775 | | | Version Tokens | | 1776 | HTTP | Hypertext | any DIGIT.DIGIT | Section 4.2 of | 1777 | | Transfer Protocol | (e.g, "2.0") | [Semantics] | 1778 +------+-------------------+-----------------+----------------+ 1780 Table 4 1782 9.9.2. Upgrade Token Registry 1784 The "Hypertext Transfer Protocol (HTTP) Upgrade Token Registry" 1785 defines the namespace for protocol-name tokens used to identify 1786 protocols in the Upgrade header field. The registry is maintained at 1787 . 1789 Each registered protocol name is associated with contact information 1790 and an optional set of specifications that details how the connection 1791 will be processed after it has been upgraded. 1793 Registrations happen on a "First Come First Served" basis (see 1794 Section 4.4 of [RFC8126]) and are subject to the following rules: 1796 1. A protocol-name token, once registered, stays registered forever. 1798 2. A protocol-name token is case-insensitive and registered with the 1799 preferred case to be generated by senders. 1801 3. The registration MUST name a responsible party for the 1802 registration. 1804 4. The registration MUST name a point of contact. 1806 5. The registration MAY name a set of specifications associated with 1807 that token. Such specifications need not be publicly available. 1809 6. The registration SHOULD name a set of expected "protocol-version" 1810 tokens associated with that token at the time of registration. 1812 7. The responsible party MAY change the registration at any time. 1813 The IANA will keep a record of all such changes, and make them 1814 available upon request. 1816 8. The IESG MAY reassign responsibility for a protocol token. This 1817 will normally only be used in the case when a responsible party 1818 cannot be contacted. 1820 10. Enclosing Messages as Data 1822 10.1. Media Type message/http 1824 The message/http media type can be used to enclose a single HTTP 1825 request or response message, provided that it obeys the MIME 1826 restrictions for all "message" types regarding line length and 1827 encodings. 1829 Type name: message 1831 Subtype name: http 1833 Required parameters: N/A 1835 Optional parameters: version, msgtype 1837 version: The HTTP-version number of the enclosed message (e.g., 1838 "1.1"). If not present, the version can be determined from the 1839 first line of the body. 1841 msgtype: The message type - "request" or "response". If not 1842 present, the type can be determined from the first line of the 1843 body. 1845 Encoding considerations: only "7bit", "8bit", or "binary" are 1846 permitted 1848 Security considerations: see Section 11 1850 Interoperability considerations: N/A 1852 Published specification: This specification (see Section 10.1). 1854 Applications that use this media type: N/A 1856 Fragment identifier considerations: N/A 1857 Additional information: Magic number(s): N/A 1859 Deprecated alias names for this type: N/A 1861 File extension(s): N/A 1863 Macintosh file type code(s): N/A 1865 Person and email address to contact for further information: See Aut 1866 hors' Addresses section. 1868 Intended usage: COMMON 1870 Restrictions on usage: N/A 1872 Author: See Authors' Addresses section. 1874 Change controller: IESG 1876 10.2. Media Type application/http 1878 The application/http media type can be used to enclose a pipeline of 1879 one or more HTTP request or response messages (not intermixed). 1881 Type name: application 1883 Subtype name: http 1885 Required parameters: N/A 1887 Optional parameters: version, msgtype 1889 version: The HTTP-version number of the enclosed messages (e.g., 1890 "1.1"). If not present, the version can be determined from the 1891 first line of the body. 1893 msgtype: The message type - "request" or "response". If not 1894 present, the type can be determined from the first line of the 1895 body. 1897 Encoding considerations: HTTP messages enclosed by this type are in 1898 "binary" format; use of an appropriate Content-Transfer-Encoding 1899 is required when transmitted via email. 1901 Security considerations: see Section 11 1903 Interoperability considerations: N/A 1904 Published specification: This specification (see Section 10.2). 1906 Applications that use this media type: N/A 1908 Fragment identifier considerations: N/A 1910 Additional information: Deprecated alias names for this type: N/A 1912 Magic number(s): N/A 1914 File extension(s): N/A 1916 Macintosh file type code(s): N/A 1918 Person and email address to contact for further information: See Aut 1919 hors' Addresses section. 1921 Intended usage: COMMON 1923 Restrictions on usage: N/A 1925 Author: See Authors' Addresses section. 1927 Change controller: IESG 1929 11. Security Considerations 1931 This section is meant to inform developers, information providers, 1932 and users of known security considerations relevant to HTTP message 1933 syntax, parsing, and routing. Security considerations about HTTP 1934 semantics and payloads are addressed in [Semantics]. 1936 11.1. Response Splitting 1938 Response splitting (a.k.a, CRLF injection) is a common technique, 1939 used in various attacks on Web usage, that exploits the line-based 1940 nature of HTTP message framing and the ordered association of 1941 requests to responses on persistent connections [Klein]. This 1942 technique can be particularly damaging when the requests pass through 1943 a shared cache. 1945 Response splitting exploits a vulnerability in servers (usually 1946 within an application server) where an attacker can send encoded data 1947 within some parameter of the request that is later decoded and echoed 1948 within any of the response header fields of the response. If the 1949 decoded data is crafted to look like the response has ended and a 1950 subsequent response has begun, the response has been split and the 1951 content within the apparent second response is controlled by the 1952 attacker. The attacker can then make any other request on the same 1953 persistent connection and trick the recipients (including 1954 intermediaries) into believing that the second half of the split is 1955 an authoritative answer to the second request. 1957 For example, a parameter within the request-target might be read by 1958 an application server and reused within a redirect, resulting in the 1959 same parameter being echoed in the Location header field of the 1960 response. If the parameter is decoded by the application and not 1961 properly encoded when placed in the response field, the attacker can 1962 send encoded CRLF octets and other content that will make the 1963 application's single response look like two or more responses. 1965 A common defense against response splitting is to filter requests for 1966 data that looks like encoded CR and LF (e.g., "%0D" and "%0A"). 1967 However, that assumes the application server is only performing URI 1968 decoding, rather than more obscure data transformations like charset 1969 transcoding, XML entity translation, base64 decoding, sprintf 1970 reformatting, etc. A more effective mitigation is to prevent 1971 anything other than the server's core protocol libraries from sending 1972 a CR or LF within the header section, which means restricting the 1973 output of header fields to APIs that filter for bad octets and not 1974 allowing application servers to write directly to the protocol 1975 stream. 1977 11.2. Request Smuggling 1979 Request smuggling ([Linhart]) is a technique that exploits 1980 differences in protocol parsing among various recipients to hide 1981 additional requests (which might otherwise be blocked or disabled by 1982 policy) within an apparently harmless request. Like response 1983 splitting, request smuggling can lead to a variety of attacks on HTTP 1984 usage. 1986 This specification has introduced new requirements on request 1987 parsing, particularly with regard to message framing in Section 6.3, 1988 to reduce the effectiveness of request smuggling. 1990 11.3. Message Integrity 1992 HTTP does not define a specific mechanism for ensuring message 1993 integrity, instead relying on the error-detection ability of 1994 underlying transport protocols and the use of length or chunk- 1995 delimited framing to detect completeness. Additional integrity 1996 mechanisms, such as hash functions or digital signatures applied to 1997 the content, can be selectively added to messages via extensible 1998 metadata fields. Historically, the lack of a single integrity 1999 mechanism has been justified by the informal nature of most HTTP 2000 communication. However, the prevalence of HTTP as an information 2001 access mechanism has resulted in its increasing use within 2002 environments where verification of message integrity is crucial. 2004 User agents are encouraged to implement configurable means for 2005 detecting and reporting failures of message integrity such that those 2006 means can be enabled within environments for which integrity is 2007 necessary. For example, a browser being used to view medical history 2008 or drug interaction information needs to indicate to the user when 2009 such information is detected by the protocol to be incomplete, 2010 expired, or corrupted during transfer. Such mechanisms might be 2011 selectively enabled via user agent extensions or the presence of 2012 message integrity metadata in a response. At a minimum, user agents 2013 ought to provide some indication that allows a user to distinguish 2014 between a complete and incomplete response message (Section 8) when 2015 such verification is desired. 2017 11.4. Message Confidentiality 2019 HTTP relies on underlying transport protocols to provide message 2020 confidentiality when that is desired. HTTP has been specifically 2021 designed to be independent of the transport protocol, such that it 2022 can be used over many different forms of encrypted connection, with 2023 the selection of such transports being identified by the choice of 2024 URI scheme or within user agent configuration. 2026 The "https" scheme can be used to identify resources that require a 2027 confidential connection, as described in Section 2.5.2 of 2028 [Semantics]. 2030 12. IANA Considerations 2032 The change controller for the following registrations is: "IETF 2033 (iesg@ietf.org) - Internet Engineering Task Force". 2035 12.1. Field Name Registration 2037 Please update the "Hypertext Transfer Protocol (HTTP) Field Name 2038 Registry" at with the 2039 field names listed in the two tables of Section 5. 2041 12.2. Media Type Registration 2043 Please update the "Media Types" registry at 2044 with the registration 2045 information in Section 10.1 and Section 10.2 for the media types 2046 "message/http" and "application/http", respectively. 2048 12.3. Transfer Coding Registration 2050 Please update the "HTTP Transfer Coding Registry" at 2051 with the 2052 registration procedure of Section 7.3 and the content coding names 2053 summarized in the table of Section 7. 2055 12.4. Upgrade Token Registration 2057 Please update the "Hypertext Transfer Protocol (HTTP) Upgrade Token 2058 Registry" at 2059 with the registration procedure of Section 9.9.2 and the upgrade 2060 token names summarized in the table of Section 9.9.1. 2062 12.5. ALPN Protocol ID Registration 2064 Please update the "TLS Application-Layer Protocol Negotiation (ALPN) 2065 Protocol IDs" registry at with the 2067 registration below: 2069 +----------+-----------------------------+----------------+ 2070 | Protocol | Identification Sequence | Reference | 2071 | HTTP/1.1 | 0x68 0x74 0x74 0x70 0x2f | (this | 2072 | | 0x31 0x2e 0x31 ("http/1.1") | specification) | 2073 +----------+-----------------------------+----------------+ 2075 Table 5 2077 13. References 2079 13.1. Normative References 2081 [Caching] Fielding, R., Ed., Nottingham, M., Ed., and J. F. Reschke, 2082 Ed., "HTTP Caching", Work in Progress, Internet-Draft, 2083 draft-ietf-httpbis-cache-10, July 12, 2020, 2084 . 2086 [RFC1950] Deutsch, L.P. and J-L. Gailly, "ZLIB Compressed Data 2087 Format Specification version 3.3", RFC 1950, 2088 DOI 10.17487/RFC1950, May 1996, 2089 . 2091 [RFC1951] Deutsch, P., "DEFLATE Compressed Data Format Specification 2092 version 1.3", RFC 1951, DOI 10.17487/RFC1951, May 1996, 2093 . 2095 [RFC1952] Deutsch, P., Gailly, J-L., Adler, M., Deutsch, L.P., and 2096 G. Randers-Pehrson, "GZIP file format specification 2097 version 4.3", RFC 1952, DOI 10.17487/RFC1952, May 1996, 2098 . 2100 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2101 Requirement Levels", BCP 14, RFC 2119, 2102 DOI 10.17487/RFC2119, March 1997, 2103 . 2105 [RFC3986] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform 2106 Resource Identifier (URI): Generic Syntax", STD 66, 2107 RFC 3986, DOI 10.17487/RFC3986, January 2005, 2108 . 2110 [RFC5234] Crocker, D., Ed. and P. Overell, "Augmented BNF for Syntax 2111 Specifications: ABNF", STD 68, RFC 5234, 2112 DOI 10.17487/RFC5234, January 2008, 2113 . 2115 [RFC7405] Kyzivat, P., "Case-Sensitive String Support in ABNF", 2116 RFC 7405, DOI 10.17487/RFC7405, December 2014, 2117 . 2119 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2120 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 2121 May 2017, . 2123 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 2124 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 2125 . 2127 [Semantics] 2128 Fielding, R., Ed., Nottingham, M., Ed., and J. F. Reschke, 2129 Ed., "HTTP Semantics", Work in Progress, Internet-Draft, 2130 draft-ietf-httpbis-semantics-10, July 12, 2020, 2131 . 2134 [USASCII] American National Standards Institute, "Coded Character 2135 Set -- 7-bit American Standard Code for Information 2136 Interchange", ANSI X3.4, 1986. 2138 [Welch] Welch, T. A., "A Technique for High-Performance Data 2139 Compression", IEEE Computer 17(6), June 1984. 2141 13.2. Informative References 2143 [Err4667] RFC Errata, Erratum ID 4667, RFC 7230, 2144 . 2146 [Klein] Klein, A., "Divide and Conquer - HTTP Response Splitting, 2147 Web Cache Poisoning Attacks, and Related Topics", March 2148 2004, . 2151 [Linhart] Linhart, C., Klein, A., Heled, R., and S. Orrin, "HTTP 2152 Request Smuggling", June 2005, 2153 . 2155 [RFC1945] Berners-Lee, T., Fielding, R.T., and H.F. Nielsen, 2156 "Hypertext Transfer Protocol -- HTTP/1.0", RFC 1945, 2157 DOI 10.17487/RFC1945, May 1996, 2158 . 2160 [RFC2045] Freed, N. and N.S. Borenstein, "Multipurpose Internet Mail 2161 Extensions (MIME) Part One: Format of Internet Message 2162 Bodies", RFC 2045, DOI 10.17487/RFC2045, November 1996, 2163 . 2165 [RFC2046] Freed, N. and N. Borenstein, "Multipurpose Internet Mail 2166 Extensions (MIME) Part Two: Media Types", RFC 2046, 2167 DOI 10.17487/RFC2046, November 1996, 2168 . 2170 [RFC2049] Freed, N. and N.S. Borenstein, "Multipurpose Internet Mail 2171 Extensions (MIME) Part Five: Conformance Criteria and 2172 Examples", RFC 2049, DOI 10.17487/RFC2049, November 1996, 2173 . 2175 [RFC2068] Fielding, R., Gettys, J., Mogul, J., Nielsen, H., and T. 2176 Berners-Lee, "Hypertext Transfer Protocol -- HTTP/1.1", 2177 RFC 2068, DOI 10.17487/RFC2068, January 1997, 2178 . 2180 [RFC2557] Palme, F., Hopmann, A., Shelness, N., and E. Stefferud, 2181 "MIME Encapsulation of Aggregate Documents, such as HTML 2182 (MHTML)", RFC 2557, DOI 10.17487/RFC2557, March 1999, 2183 . 2185 [RFC5322] Resnick, P., "Internet Message Format", RFC 5322, 2186 DOI 10.17487/RFC5322, October 2008, 2187 . 2189 [RFC7230] Fielding, R., Ed. and J. F. Reschke, Ed., "Hypertext 2190 Transfer Protocol (HTTP/1.1): Message Syntax and Routing", 2191 RFC 7230, DOI 10.17487/RFC7230, June 2014, 2192 . 2194 [RFC7231] Fielding, R., Ed. and J. F. Reschke, Ed., "Hypertext 2195 Transfer Protocol (HTTP/1.1): Semantics and Content", 2196 RFC 7231, DOI 10.17487/RFC7231, June 2014, 2197 . 2199 [RFC8126] Cotton, M., Leiba, B., and T. Narten, "Guidelines for 2200 Writing an IANA Considerations Section in RFCs", BCP 26, 2201 RFC 8126, DOI 10.17487/RFC8126, June 2017, 2202 . 2204 Appendix A. Collected ABNF 2206 In the collected ABNF below, list rules are expanded as per 2207 Section 5.5.1 of [Semantics]. 2209 BWS = 2211 Connection = connection-option *( OWS "," OWS connection-option ) 2213 HTTP-message = start-line CRLF *( field-line CRLF ) CRLF [ 2214 message-body ] 2215 HTTP-name = %x48.54.54.50 ; HTTP 2216 HTTP-version = HTTP-name "/" DIGIT "." DIGIT 2218 OWS = 2220 RWS = 2222 TE = [ t-codings *( OWS "," OWS t-codings ) ] 2223 Transfer-Encoding = transfer-coding *( OWS "," OWS transfer-coding ) 2225 Upgrade = protocol *( OWS "," OWS protocol ) 2227 absolute-URI = 2228 absolute-form = absolute-URI 2229 absolute-path = 2230 asterisk-form = "*" 2231 authority = 2232 authority-form = authority 2234 chunk = chunk-size [ chunk-ext ] CRLF chunk-data CRLF 2235 chunk-data = 1*OCTET 2236 chunk-ext = *( BWS ";" BWS chunk-ext-name [ BWS "=" BWS chunk-ext-val 2237 ] ) 2238 chunk-ext-name = token 2239 chunk-ext-val = token / quoted-string 2240 chunk-size = 1*HEXDIG 2241 chunked-body = *chunk last-chunk trailer-section CRLF 2242 comment = 2243 connection-option = token 2245 field-line = field-name ":" OWS field-value OWS 2246 field-name = 2247 field-value = 2249 last-chunk = 1*"0" [ chunk-ext ] CRLF 2251 message-body = *OCTET 2252 method = token 2254 obs-fold = OWS CRLF RWS 2255 obs-text = 2256 origin-form = absolute-path [ "?" query ] 2258 port = 2259 protocol = protocol-name [ "/" protocol-version ] 2260 protocol-name = token 2261 protocol-version = token 2263 query = 2264 quoted-string = 2266 rank = ( "0" [ "." *3DIGIT ] ) / ( "1" [ "." *3"0" ] ) 2267 reason-phrase = 1*( HTAB / SP / VCHAR / obs-text ) 2268 request-line = method SP request-target SP HTTP-version 2269 request-target = origin-form / absolute-form / authority-form / 2270 asterisk-form 2272 start-line = request-line / status-line 2273 status-code = 3DIGIT 2274 status-line = HTTP-version SP status-code SP [ reason-phrase ] 2276 t-codings = "trailers" / ( transfer-coding [ t-ranking ] ) 2277 t-ranking = OWS ";" OWS "q=" rank 2278 token = 2279 trailer-section = *( field-line CRLF ) 2280 transfer-coding = token *( OWS ";" OWS transfer-parameter ) 2281 transfer-parameter = token BWS "=" BWS ( token / quoted-string ) 2283 uri-host = 2285 Appendix B. Differences between HTTP and MIME 2287 HTTP/1.1 uses many of the constructs defined for the Internet Message 2288 Format [RFC5322] and the Multipurpose Internet Mail Extensions (MIME) 2289 [RFC2045] to allow a message body to be transmitted in an open 2290 variety of representations and with extensible fields. However, RFC 2291 2045 is focused only on email; applications of HTTP have many 2292 characteristics that differ from email; hence, HTTP has features that 2293 differ from MIME. These differences were carefully chosen to 2294 optimize performance over binary connections, to allow greater 2295 freedom in the use of new media types, to make date comparisons 2296 easier, and to acknowledge the practice of some early HTTP servers 2297 and clients. 2299 This appendix describes specific areas where HTTP differs from MIME. 2300 Proxies and gateways to and from strict MIME environments need to be 2301 aware of these differences and provide the appropriate conversions 2302 where necessary. 2304 B.1. MIME-Version 2306 HTTP is not a MIME-compliant protocol. However, messages can include 2307 a single MIME-Version header field to indicate what version of the 2308 MIME protocol was used to construct the message. Use of the MIME- 2309 Version header field indicates that the message is in full 2310 conformance with the MIME protocol (as defined in [RFC2045]). 2311 Senders are responsible for ensuring full conformance (where 2312 possible) when exporting HTTP messages to strict MIME environments. 2314 B.2. Conversion to Canonical Form 2316 MIME requires that an Internet mail body part be converted to 2317 canonical form prior to being transferred, as described in Section 4 2318 of [RFC2049]. Section 7.1.1.2 of [Semantics] describes the forms 2319 allowed for subtypes of the "text" media type when transmitted over 2320 HTTP. [RFC2046] requires that content with a type of "text" 2321 represent line breaks as CRLF and forbids the use of CR or LF outside 2322 of line break sequences. HTTP allows CRLF, bare CR, and bare LF to 2323 indicate a line break within text content. 2325 A proxy or gateway from HTTP to a strict MIME environment ought to 2326 translate all line breaks within text media types to the RFC 2049 2327 canonical form of CRLF. Note, however, this might be complicated by 2328 the presence of a Content-Encoding and by the fact that HTTP allows 2329 the use of some charsets that do not use octets 13 and 10 to 2330 represent CR and LF, respectively. 2332 Conversion will break any cryptographic checksums applied to the 2333 original content unless the original content is already in canonical 2334 form. Therefore, the canonical form is recommended for any content 2335 that uses such checksums in HTTP. 2337 B.3. Conversion of Date Formats 2339 HTTP/1.1 uses a restricted set of date formats (Section 5.4.1.5 of 2340 [Semantics]) to simplify the process of date comparison. Proxies and 2341 gateways from other protocols ought to ensure that any Date header 2342 field present in a message conforms to one of the HTTP/1.1 formats 2343 and rewrite the date if necessary. 2345 B.4. Conversion of Content-Encoding 2347 MIME does not include any concept equivalent to HTTP/1.1's Content- 2348 Encoding header field. Since this acts as a modifier on the media 2349 type, proxies and gateways from HTTP to MIME-compliant protocols 2350 ought to either change the value of the Content-Type header field or 2351 decode the representation before forwarding the message. (Some 2352 experimental applications of Content-Type for Internet mail have used 2353 a media-type parameter of ";conversions=" to perform 2354 a function equivalent to Content-Encoding. However, this parameter 2355 is not part of the MIME standards). 2357 B.5. Conversion of Content-Transfer-Encoding 2359 HTTP does not use the Content-Transfer-Encoding field of MIME. 2360 Proxies and gateways from MIME-compliant protocols to HTTP need to 2361 remove any Content-Transfer-Encoding prior to delivering the response 2362 message to an HTTP client. 2364 Proxies and gateways from HTTP to MIME-compliant protocols are 2365 responsible for ensuring that the message is in the correct format 2366 and encoding for safe transport on that protocol, where "safe 2367 transport" is defined by the limitations of the protocol being used. 2368 Such a proxy or gateway ought to transform and label the data with an 2369 appropriate Content-Transfer-Encoding if doing so will improve the 2370 likelihood of safe transport over the destination protocol. 2372 B.6. MHTML and Line Length Limitations 2374 HTTP implementations that share code with MHTML [RFC2557] 2375 implementations need to be aware of MIME line length limitations. 2376 Since HTTP does not have this limitation, HTTP does not fold long 2377 lines. MHTML messages being transported by HTTP follow all 2378 conventions of MHTML, including line length limitations and folding, 2379 canonicalization, etc., since HTTP transfers message-bodies as 2380 payload and, aside from the "multipart/byteranges" type 2381 (Section 7.3.5 of [Semantics]), does not interpret the content or any 2382 MIME header lines that might be contained therein. 2384 Appendix C. HTTP Version History 2386 HTTP has been in use since 1990. The first version, later referred 2387 to as HTTP/0.9, was a simple protocol for hypertext data transfer 2388 across the Internet, using only a single request method (GET) and no 2389 metadata. HTTP/1.0, as defined by [RFC1945], added a range of 2390 request methods and MIME-like messaging, allowing for metadata to be 2391 transferred and modifiers placed on the request/response semantics. 2392 However, HTTP/1.0 did not sufficiently take into consideration the 2393 effects of hierarchical proxies, caching, the need for persistent 2394 connections, or name-based virtual hosts. The proliferation of 2395 incompletely implemented applications calling themselves "HTTP/1.0" 2396 further necessitated a protocol version change in order for two 2397 communicating applications to determine each other's true 2398 capabilities. 2400 HTTP/1.1 remains compatible with HTTP/1.0 by including more stringent 2401 requirements that enable reliable implementations, adding only those 2402 features that can either be safely ignored by an HTTP/1.0 recipient 2403 or only be sent when communicating with a party advertising 2404 conformance with HTTP/1.1. 2406 HTTP/1.1 has been designed to make supporting previous versions easy. 2407 A general-purpose HTTP/1.1 server ought to be able to understand any 2408 valid request in the format of HTTP/1.0, responding appropriately 2409 with an HTTP/1.1 message that only uses features understood (or 2410 safely ignored) by HTTP/1.0 clients. Likewise, an HTTP/1.1 client 2411 can be expected to understand any valid HTTP/1.0 response. 2413 Since HTTP/0.9 did not support header fields in a request, there is 2414 no mechanism for it to support name-based virtual hosts (selection of 2415 resource by inspection of the Host header field). Any server that 2416 implements name-based virtual hosts ought to disable support for 2417 HTTP/0.9. Most requests that appear to be HTTP/0.9 are, in fact, 2418 badly constructed HTTP/1.x requests caused by a client failing to 2419 properly encode the request-target. 2421 C.1. Changes from HTTP/1.0 2423 This section summarizes major differences between versions HTTP/1.0 2424 and HTTP/1.1. 2426 C.1.1. Multihomed Web Servers 2428 The requirements that clients and servers support the Host header 2429 field (Section 6.6 of [Semantics]), report an error if it is missing 2430 from an HTTP/1.1 request, and accept absolute URIs (Section 3.2) are 2431 among the most important changes defined by HTTP/1.1. 2433 Older HTTP/1.0 clients assumed a one-to-one relationship of IP 2434 addresses and servers; there was no other established mechanism for 2435 distinguishing the intended server of a request than the IP address 2436 to which that request was directed. The Host header field was 2437 introduced during the development of HTTP/1.1 and, though it was 2438 quickly implemented by most HTTP/1.0 browsers, additional 2439 requirements were placed on all HTTP/1.1 requests in order to ensure 2440 complete adoption. At the time of this writing, most HTTP-based 2441 services are dependent upon the Host header field for targeting 2442 requests. 2444 C.1.2. Keep-Alive Connections 2446 In HTTP/1.0, each connection is established by the client prior to 2447 the request and closed by the server after sending the response. 2448 However, some implementations implement the explicitly negotiated 2449 ("Keep-Alive") version of persistent connections described in 2450 Section 19.7.1 of [RFC2068]. 2452 Some clients and servers might wish to be compatible with these 2453 previous approaches to persistent connections, by explicitly 2454 negotiating for them with a "Connection: keep-alive" request header 2455 field. However, some experimental implementations of HTTP/1.0 2456 persistent connections are faulty; for example, if an HTTP/1.0 proxy 2457 server doesn't understand Connection, it will erroneously forward 2458 that header field to the next inbound server, which would result in a 2459 hung connection. 2461 One attempted solution was the introduction of a Proxy-Connection 2462 header field, targeted specifically at proxies. In practice, this 2463 was also unworkable, because proxies are often deployed in multiple 2464 layers, bringing about the same problem discussed above. 2466 As a result, clients are encouraged not to send the Proxy-Connection 2467 header field in any requests. 2469 Clients are also encouraged to consider the use of Connection: keep- 2470 alive in requests carefully; while they can enable persistent 2471 connections with HTTP/1.0 servers, clients using them will need to 2472 monitor the connection for "hung" requests (which indicate that the 2473 client ought stop sending the header field), and this mechanism ought 2474 not be used by clients at all when a proxy is being used. 2476 C.1.3. Introduction of Transfer-Encoding 2478 HTTP/1.1 introduces the Transfer-Encoding header field (Section 6.1). 2479 Transfer codings need to be decoded prior to forwarding an HTTP 2480 message over a MIME-compliant protocol. 2482 C.2. Changes from RFC 7230 2484 Most of the sections introducing HTTP's design goals, history, 2485 architecture, conformance criteria, protocol versioning, URIs, 2486 message routing, and header fields have been moved to [Semantics]. 2487 This document has been reduced to just the messaging syntax and 2488 connection management requirements specific to HTTP/1.1. 2490 Prohibited generation of bare CRs outside of payload body. 2491 (Section 2.2) 2493 In the ABNF for chunked extensions, re-introduced (bad) whitespace 2494 around ";" and "=". Whitespace was removed in [RFC7230], but that 2495 change was found to break existing implementations (see [Err4667]). 2496 (Section 7.1.1) 2498 Trailer field semantics now transcend the specifics of chunked 2499 encoding. The decoding algorithm for chunked (Section 7.1.3) has 2500 been updated to encourage storage/forwarding of trailer fields 2501 separately from the header section, to only allow merging into the 2502 header section if the recipient knows the corresponding field 2503 definition permits and defines how to merge, and otherwise to discard 2504 the trailer fields instead of merging. The trailer part is now 2505 called the trailer section to be more consistent with the header 2506 section and more distinct from a body part. (Section 7.1.2) 2508 Disallowed transfer coding parameters called "q" in order to avoid 2509 conflicts with the use of ranks in the TE header field. 2510 (Section 7.3) 2512 Appendix D. Change Log 2514 This section is to be removed before publishing as an RFC. 2516 D.1. Between RFC7230 and draft 00 2518 The changes were purely editorial: 2520 o Change boilerplate and abstract to indicate the "draft" status, 2521 and update references to ancestor specifications. 2523 o Adjust historical notes. 2525 o Update links to sibling specifications. 2527 o Replace sections listing changes from RFC 2616 by new empty 2528 sections referring to RFC 723x. 2530 o Remove acknowledgements specific to RFC 723x. 2532 o Move "Acknowledgements" to the very end and make them unnumbered. 2534 D.2. Since draft-ietf-httpbis-messaging-00 2536 The changes in this draft are editorial, with respect to HTTP as a 2537 whole, to move all core HTTP semantics into [Semantics]: 2539 o Moved introduction, architecture, conformance, and ABNF extensions 2540 from RFC 7230 (Messaging) to semantics [Semantics]. 2542 o Moved discussion of MIME differences from RFC 7231 (Semantics) to 2543 Appendix B since they mostly cover transforming 1.1 messages. 2545 o Moved all extensibility tips, registration procedures, and 2546 registry tables from the IANA considerations to normative 2547 sections, reducing the IANA considerations to just instructions 2548 that will be removed prior to publication as an RFC. 2550 D.3. Since draft-ietf-httpbis-messaging-01 2552 o Cite RFC 8126 instead of RFC 5226 () 2555 o Resolved erratum 4779, no change needed here 2556 (, 2557 ) 2559 o In Section 7, fixed prose claiming transfer parameters allow bare 2560 names (, 2561 ) 2563 o Resolved erratum 4225, no change needed here 2564 (, 2565 ) 2567 o Replace "response code" with "response status code" 2568 (, 2569 ) 2571 o In Section 9.4, clarify statement about HTTP/1.0 keep-alive 2572 (, 2573 ) 2575 o In Section 7.1.1, re-introduce (bad) whitespace around ";" and "=" 2576 (, 2577 , ) 2580 o In Section 7.3, state that transfer codings should not use 2581 parameters named "q" (, ) 2584 o In Section 7, mark coding name "trailers" as reserved in the IANA 2585 registry () 2587 D.4. Since draft-ietf-httpbis-messaging-02 2589 o In Section 4, explain why the reason phrase should be ignored by 2590 clients (). 2592 o Add Section 9.3 to explain how request/response correlation is 2593 performed () 2595 D.5. Since draft-ietf-httpbis-messaging-03 2597 o In Section 9.3, caution against treating data on a connection as 2598 part of a not-yet-issued request () 2601 o In Section 7, remove the predefined codings from the ABNF and make 2602 it generic instead () 2605 o Use RFC 7405 ABNF notation for case-sensitive string constants 2606 () 2608 D.6. Since draft-ietf-httpbis-messaging-04 2609 o In Section 9.9, clarify that protocol-name is to be matched case- 2610 insensitively () 2612 o In Section 5.2, add leading optional whitespace to obs-fold ABNF 2613 (, 2614 ) 2616 o In Section 4, add clarifications about empty reason phrases 2617 () 2619 o Move discussion of retries from Section 9.4.1 into [Semantics] 2620 () 2622 D.7. Since draft-ietf-httpbis-messaging-05 2624 o In Section 7.1.2, the trailer part has been renamed the trailer 2625 section (for consistency with the header section) and trailers are 2626 no longer merged as header fields by default, but rather can be 2627 discarded, kept separate from header fields, or merged with header 2628 fields only if understood and defined as being mergeable 2629 () 2631 o In Section 2.1 and related Sections, move the trailing CRLF from 2632 the line grammars into the message format 2633 () 2635 o Moved Section 2.3 down () 2638 o In Section 9.9, use 'websocket' instead of 'HTTP/2.0' in examples 2639 () 2641 o Move version non-specific text from Section 6 into semantics as 2642 "payload body" () 2644 o In Section 9.8, add text from RFC 2818 2645 () 2647 D.8. Since draft-ietf-httpbis-messaging-06 2649 o In Section 12.5, update the APLN protocol id for HTTP/1.1 2650 () 2652 o In Section 5, align with updates to field terminology in semantics 2653 () 2655 o In Section 9.1, clarify that new connection options indeed need to 2656 be registered () 2658 o In Section 1.1, reference RFC 8174 as well 2659 () 2661 D.9. Since draft-ietf-httpbis-messaging-07 2663 o Move TE: trailers into [Semantics] () 2666 o In Section 6.3, adjust requirements for handling multiple content- 2667 length values () 2669 o Throughout, replace "effective request URI" with "target URI" 2670 () 2672 o In Section 6.1, don't claim Transfer-Encoding is supported by 2673 HTTP/2 or later () 2675 D.10. Since draft-ietf-httpbis-messaging-08 2677 o In Section 2.2, disallow bare CRs () 2680 o Appendix A now uses the sender variant of the "#" list expansion 2681 () 2683 o In Section 5, adjust IANA "Close" entry for new registry format 2684 () 2686 D.11. Since draft-ietf-httpbis-messaging-09 2688 o Switch to xml2rfc v3 mode for draft generation 2689 () 2691 Acknowledgments 2693 See Appendix "Acknowledgments" of [Semantics]. 2695 Authors' Addresses 2697 Roy T. Fielding (editor) 2698 Adobe 2699 345 Park Ave 2700 San Jose, CA 95110 2701 United States of America 2703 Email: fielding@gbiv.com 2704 URI: https://roy.gbiv.com/ 2705 Mark Nottingham (editor) 2706 Fastly 2708 Email: mnot@mnot.net 2709 URI: https://www.mnot.net/ 2711 Julian F. Reschke (editor) 2712 greenbytes GmbH 2713 Hafenweg 16 2714 48155 Münster 2715 Germany 2717 Email: julian.reschke@greenbytes.de 2718 URI: https://greenbytes.de/tech/webdav/