7
Network Working Group T. Berners-Lee
8
Request for Comments: 2396 MIT/LCS
9
Updates: 1808, 1738 R. Fielding
10
Category: Standards Track U.C. Irvine
16
Uniform Resource Identifiers (URI): Generic Syntax
20
This document specifies an Internet standards track protocol for the
21
Internet community, and requests discussion and suggestions for
22
improvements. Please refer to the current edition of the "Internet
23
Official Protocol Standards" (STD 1) for the standardization state
24
and status of this protocol. Distribution of this memo is unlimited.
28
Copyright (C) The Internet Society (1998). All Rights Reserved.
32
This paper describes a "superset" of operations that can be applied
33
to URI. It consists of both a grammar and a description of basic
34
functionality for URI. To understand what is a valid URI, both the
35
grammar and the associated description have to be studied. Some of
36
the functionality described is not applicable to all URI schemes, and
37
some operations are only possible when certain media types are
38
retrieved using the URI, regardless of the scheme used.
42
A Uniform Resource Identifier (URI) is a compact string of characters
43
for identifying an abstract or physical resource. This document
44
defines the generic syntax of URI, including both absolute and
45
relative forms, and guidelines for their use; it revises and replaces
46
the generic definitions in RFC 1738 and RFC 1808.
48
This document defines a grammar that is a superset of all valid URI,
49
such that an implementation can parse the common components of a URI
50
reference without knowing the scheme-specific requirements of every
51
possible identifier type. This document does not define a generative
52
grammar for URI; that task will be performed by the individual
53
specifications of each URI scheme.
58
Berners-Lee, et. al. Standards Track [Page 1]
60
RFC 2396 URI Generic Syntax August 1998
65
Uniform Resource Identifiers (URI) provide a simple and extensible
66
means for identifying a resource. This specification of URI syntax
67
and semantics is derived from concepts introduced by the World Wide
68
Web global information initiative, whose use of such objects dates
69
from 1990 and is described in "Universal Resource Identifiers in WWW"
70
[RFC1630]. The specification of URI is designed to meet the
71
recommendations laid out in "Functional Recommendations for Internet
72
Resource Locators" [RFC1736] and "Functional Requirements for Uniform
73
Resource Names" [RFC1737].
75
This document updates and merges "Uniform Resource Locators"
76
[RFC1738] and "Relative Uniform Resource Locators" [RFC1808] in order
77
to define a single, generic syntax for all URI. It excludes those
78
portions of RFC 1738 that defined the specific syntax of individual
79
URL schemes; those portions will be updated as separate documents, as
80
will the process for registration of new URI schemes. This document
81
does not discuss the issues and recommendation for dealing with
82
characters outside of the US-ASCII character set [ASCII]; those
83
recommendations are discussed in a separate document.
85
All significant changes from the prior RFCs are noted in Appendix G.
89
URI are characterized by the following definitions:
92
Uniformity provides several benefits: it allows different types
93
of resource identifiers to be used in the same context, even
94
when the mechanisms used to access those resources may differ;
95
it allows uniform semantic interpretation of common syntactic
96
conventions across different types of resource identifiers; it
97
allows introduction of new types of resource identifiers
98
without interfering with the way that existing identifiers are
99
used; and, it allows the identifiers to be reused in many
100
different contexts, thus permitting new applications or
101
protocols to leverage a pre-existing, large, and widely-used
102
set of resource identifiers.
105
A resource can be anything that has identity. Familiar
106
examples include an electronic document, an image, a service
107
(e.g., "today's weather report for Los Angeles"), and a
108
collection of other resources. Not all resources are network
109
"retrievable"; e.g., human beings, corporations, and bound
110
books in a library can also be considered resources.
114
Berners-Lee, et. al. Standards Track [Page 2]
116
RFC 2396 URI Generic Syntax August 1998
119
The resource is the conceptual mapping to an entity or set of
120
entities, not necessarily the entity which corresponds to that
121
mapping at any particular instance in time. Thus, a resource
122
can remain constant even when its content---the entities to
123
which it currently corresponds---changes over time, provided
124
that the conceptual mapping is not changed in the process.
127
An identifier is an object that can act as a reference to
128
something that has identity. In the case of URI, the object is
129
a sequence of characters with a restricted syntax.
131
Having identified a resource, a system may perform a variety of
132
operations on the resource, as might be characterized by such words
133
as `access', `update', `replace', or `find attributes'.
135
1.2. URI, URL, and URN
137
A URI can be further classified as a locator, a name, or both. The
138
term "Uniform Resource Locator" (URL) refers to the subset of URI
139
that identify resources via a representation of their primary access
140
mechanism (e.g., their network "location"), rather than identifying
141
the resource by name or by some other attribute(s) of that resource.
142
The term "Uniform Resource Name" (URN) refers to the subset of URI
143
that are required to remain globally unique and persistent even when
144
the resource ceases to exist or becomes unavailable.
146
The URI scheme (Section 3.1) defines the namespace of the URI, and
147
thus may further restrict the syntax and semantics of identifiers
148
using that scheme. This specification defines those elements of the
149
URI syntax that are either required of all URI schemes or are common
150
to many URI schemes. It thus defines the syntax and semantics that
151
are needed to implement a scheme-independent parsing mechanism for
152
URI references, such that the scheme-dependent handling of a URI can
153
be postponed until the scheme-dependent semantics are needed. We use
154
the term URL below when describing syntax or semantics that only
157
Although many URL schemes are named after protocols, this does not
158
imply that the only way to access the URL's resource is via the named
159
protocol. Gateways, proxies, caches, and name resolution services
160
might be used to access some resources, independent of the protocol
161
of their origin, and the resolution of some URL may require the use
162
of more than one protocol (e.g., both DNS and HTTP are typically used
163
to access an "http" URL's resource when it can't be found in a local
170
Berners-Lee, et. al. Standards Track [Page 3]
172
RFC 2396 URI Generic Syntax August 1998
175
A URN differs from a URL in that it's primary purpose is persistent
176
labeling of a resource with an identifier. That identifier is drawn
177
from one of a set of defined namespaces, each of which has its own
178
set name structure and assignment procedures. The "urn" scheme has
179
been reserved to establish the requirements for a standardized URN
180
namespace, as defined in "URN Syntax" [RFC2141] and its related
183
Most of the examples in this specification demonstrate URL, since
184
they allow the most varied use of the syntax and often have a
185
hierarchical namespace. A parser of the URI syntax is capable of
186
parsing both URL and URN references as a generic URI; once the scheme
187
is determined, the scheme-specific parsing can be performed on the
188
generic URI components. In other words, the URI syntax is a superset
189
of the syntax of all URI schemes.
193
The following examples illustrate URI that are in common use.
195
ftp://ftp.is.co.za/rfc/rfc1808.txt
196
-- ftp scheme for File Transfer Protocol services
198
gopher://spinaltap.micro.umn.edu/00/Weather/California/Los%20Angeles
199
-- gopher scheme for Gopher and Gopher+ Protocol services
201
http://www.math.uio.no/faq/compression-faq/part1.html
202
-- http scheme for Hypertext Transfer Protocol services
204
mailto:mduerst@ifi.unizh.ch
205
-- mailto scheme for electronic mail addresses
207
news:comp.infosystems.www.servers.unix
208
-- news scheme for USENET news groups and articles
210
telnet://melvyl.ucop.edu/
211
-- telnet scheme for interactive services via the TELNET Protocol
213
1.4. Hierarchical URI and Relative Forms
215
An absolute identifier refers to a resource independent of the
216
context in which the identifier is used. In contrast, a relative
217
identifier refers to a resource by describing the difference within a
218
hierarchical namespace between the current context and an absolute
219
identifier of the resource.
226
Berners-Lee, et. al. Standards Track [Page 4]
228
RFC 2396 URI Generic Syntax August 1998
231
Some URI schemes support a hierarchical naming system, where the
232
hierarchy of the name is denoted by a "/" delimiter separating the
233
components in the scheme. This document defines a scheme-independent
234
`relative' form of URI reference that can be used in conjunction with
235
a `base' URI (of a hierarchical scheme) to produce another URI. The
236
syntax of hierarchical URI is described in Section 3; the relative
237
URI calculation is described in Section 5.
239
1.5. URI Transcribability
241
The URI syntax was designed with global transcribability as one of
242
its main concerns. A URI is a sequence of characters from a very
243
limited set, i.e. the letters of the basic Latin alphabet, digits,
244
and a few special characters. A URI may be represented in a variety
245
of ways: e.g., ink on paper, pixels on a screen, or a sequence of
246
octets in a coded character set. The interpretation of a URI depends
247
only on the characters used and not how those characters are
248
represented in a network protocol.
250
The goal of transcribability can be described by a simple scenario.
251
Imagine two colleagues, Sam and Kim, sitting in a pub at an
252
international conference and exchanging research ideas. Sam asks Kim
253
for a location to get more information, so Kim writes the URI for the
254
research site on a napkin. Upon returning home, Sam takes out the
255
napkin and types the URI into a computer, which then retrieves the
256
information to which Kim referred.
258
There are several design concerns revealed by the scenario:
260
o A URI is a sequence of characters, which is not always
261
represented as a sequence of octets.
263
o A URI may be transcribed from a non-network source, and thus
264
should consist of characters that are most likely to be able to
265
be typed into a computer, within the constraints imposed by
266
keyboards (and related input devices) across languages and
269
o A URI often needs to be remembered by people, and it is easier
270
for people to remember a URI when it consists of meaningful
273
These design concerns are not always in alignment. For example, it
274
is often the case that the most meaningful name for a URI component
275
would require characters that cannot be typed into some systems. The
276
ability to transcribe the resource identifier from one medium to
277
another was considered more important than having its URI consist of
278
the most meaningful of components. In local and regional contexts
282
Berners-Lee, et. al. Standards Track [Page 5]
284
RFC 2396 URI Generic Syntax August 1998
287
and with improving technology, users might benefit from being able to
288
use a wider range of characters; such use is not defined in this
291
1.6. Syntax Notation and Common Elements
293
This document uses two conventions to describe and define the syntax
294
for URI. The first, called the layout form, is a general description
295
of the order of components and component separators, as in
297
<first>/<second>;<third>?<fourth>
299
The component names are enclosed in angle-brackets and any characters
300
outside angle-brackets are literal separators. Whitespace should be
301
ignored. These descriptions are used informally and do not define
302
the syntax requirements.
304
The second convention is a BNF-like grammar, used to define the
305
formal URI syntax. The grammar is that of [RFC822], except that "|"
306
is used to designate alternatives. Briefly, rules are separated from
307
definitions by an equal "=", indentation is used to continue a rule
308
definition over more than one line, literals are quoted with "",
309
parentheses "(" and ")" are used to group elements, optional elements
310
are enclosed in "[" and "]" brackets, and elements may be preceded
311
with <n>* to designate n or more repetitions of the following
312
element; n defaults to 0.
314
Unlike many specifications that use a BNF-like grammar to define the
315
bytes (octets) allowed by a protocol, the URI grammar is defined in
316
terms of characters. Each literal in the grammar corresponds to the
317
character it represents, rather than to the octet encoding of that
318
character in any particular coded character set. How a URI is
319
represented in terms of bits and bytes on the wire is dependent upon
320
the character encoding of the protocol used to transport it, or the
321
charset of the document which contains it.
323
The following definitions are common to many elements:
325
alpha = lowalpha | upalpha
327
lowalpha = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" | "i" |
328
"j" | "k" | "l" | "m" | "n" | "o" | "p" | "q" | "r" |
329
"s" | "t" | "u" | "v" | "w" | "x" | "y" | "z"
331
upalpha = "A" | "B" | "C" | "D" | "E" | "F" | "G" | "H" | "I" |
332
"J" | "K" | "L" | "M" | "N" | "O" | "P" | "Q" | "R" |
333
"S" | "T" | "U" | "V" | "W" | "X" | "Y" | "Z"
338
Berners-Lee, et. al. Standards Track [Page 6]
340
RFC 2396 URI Generic Syntax August 1998
343
digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" |
346
alphanum = alpha | digit
348
The complete URI syntax is collected in Appendix A.
350
2. URI Characters and Escape Sequences
352
URI consist of a restricted set of characters, primarily chosen to
353
aid transcribability and usability both in computer systems and in
354
non-computer communications. Characters used conventionally as
355
delimiters around URI were excluded. The restricted set of
356
characters consists of digits, letters, and a few graphic symbols
357
were chosen from those common to most of the character encodings and
358
input facilities available to Internet users.
360
uric = reserved | unreserved | escaped
362
Within a URI, characters are either used as delimiters, or to
363
represent strings of data (octets) within the delimited portions.
364
Octets are either represented directly by a character (using the US-
365
ASCII character for that octet [ASCII]) or by an escape encoding.
366
This representation is elaborated below.
368
2.1 URI and non-ASCII characters
370
The relationship between URI and characters has been a source of
371
confusion for characters that are not part of US-ASCII. To describe
372
the relationship, it is useful to distinguish between a "character"
373
(as a distinguishable semantic entity) and an "octet" (an 8-bit
374
byte). There are two mappings, one from URI characters to octets, and
375
a second from octets to original characters:
377
URI character sequence->octet sequence->original character sequence
379
A URI is represented as a sequence of characters, not as a sequence
380
of octets. That is because URI might be "transported" by means that
381
are not through a computer network, e.g., printed on paper, read over
384
A URI scheme may define a mapping from URI characters to octets;
385
whether this is done depends on the scheme. Commonly, within a
386
delimited component of a URI, a sequence of characters may be used to
387
represent a sequence of octets. For example, the character "a"
388
represents the octet 97 (decimal), while the character sequence "%",
389
"0", "a" represents the octet 10 (decimal).
394
Berners-Lee, et. al. Standards Track [Page 7]
396
RFC 2396 URI Generic Syntax August 1998
399
There is a second translation for some resources: the sequence of
400
octets defined by a component of the URI is subsequently used to
401
represent a sequence of characters. A 'charset' defines this mapping.
402
There are many charsets in use in Internet protocols. For example,
403
UTF-8 [UTF-8] defines a mapping from sequences of octets to sequences
404
of characters in the repertoire of ISO 10646.
406
In the simplest case, the original character sequence contains only
407
characters that are defined in US-ASCII, and the two levels of
408
mapping are simple and easily invertible: each 'original character'
409
is represented as the octet for the US-ASCII code for it, which is,
410
in turn, represented as either the US-ASCII character, or else the
411
"%" escape sequence for that octet.
413
For original character sequences that contain non-ASCII characters,
414
however, the situation is more difficult. Internet protocols that
415
transmit octet sequences intended to represent character sequences
416
are expected to provide some way of identifying the charset used, if
417
there might be more than one [RFC2277]. However, there is currently
418
no provision within the generic URI syntax to accomplish this
419
identification. An individual URI scheme may require a single
420
charset, define a default charset, or provide a way to indicate the
423
It is expected that a systematic treatment of character encoding
424
within URI will be developed as a future modification of this
427
2.2. Reserved Characters
429
Many URI include components consisting of or delimited by, certain
430
special characters. These characters are called "reserved", since
431
their usage within the URI component is limited to their reserved
432
purpose. If the data for a URI component would conflict with the
433
reserved purpose, then the conflicting data must be escaped before
436
reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |
439
The "reserved" syntax class above refers to those characters that are
440
allowed within a URI, but which may not be allowed within a
441
particular component of the generic URI syntax; they are used as
442
delimiters of the components described in Section 3.
450
Berners-Lee, et. al. Standards Track [Page 8]
452
RFC 2396 URI Generic Syntax August 1998
455
Characters in the "reserved" set are not reserved in all contexts.
456
The set of characters actually reserved within any given URI
457
component is defined by that component. In general, a character is
458
reserved if the semantics of the URI changes if the character is
459
replaced with its escaped US-ASCII encoding.
461
2.3. Unreserved Characters
463
Data characters that are allowed in a URI but do not have a reserved
464
purpose are called unreserved. These include upper and lower case
465
letters, decimal digits, and a limited set of punctuation marks and
468
unreserved = alphanum | mark
470
mark = "-" | "_" | "." | "!" | "~" | "*" | "'" | "(" | ")"
472
Unreserved characters can be escaped without changing the semantics
473
of the URI, but this should not be done unless the URI is being used
474
in a context that does not allow the unescaped character to appear.
476
2.4. Escape Sequences
478
Data must be escaped if it does not have a representation using an
479
unreserved character; this includes data that does not correspond to
480
a printable character of the US-ASCII coded character set, or that
481
corresponds to any US-ASCII character that is disallowed, as
484
2.4.1. Escaped Encoding
486
An escaped octet is encoded as a character triplet, consisting of the
487
percent character "%" followed by the two hexadecimal digits
488
representing the octet code. For example, "%20" is the escaped
489
encoding for the US-ASCII space character.
491
escaped = "%" hex hex
492
hex = digit | "A" | "B" | "C" | "D" | "E" | "F" |
493
"a" | "b" | "c" | "d" | "e" | "f"
495
2.4.2. When to Escape and Unescape
497
A URI is always in an "escaped" form, since escaping or unescaping a
498
completed URI might change its semantics. Normally, the only time
499
escape encodings can safely be made is when the URI is being created
500
from its component parts; each component may have its own set of
501
characters that are reserved, so only the mechanism responsible for
502
generating or interpreting that component can determine whether or
506
Berners-Lee, et. al. Standards Track [Page 9]
508
RFC 2396 URI Generic Syntax August 1998
511
not escaping a character will change its semantics. Likewise, a URI
512
must be separated into its components before the escaped characters
513
within those components can be safely decoded.
515
In some cases, data that could be represented by an unreserved
516
character may appear escaped; for example, some of the unreserved
517
"mark" characters are automatically escaped by some systems. If the
518
given URI scheme defines a canonicalization algorithm, then
519
unreserved characters may be unescaped according to that algorithm.
520
For example, "%7e" is sometimes used instead of "~" in an http URL
521
path, but the two are equivalent for an http URL.
523
Because the percent "%" character always has the reserved purpose of
524
being the escape indicator, it must be escaped as "%25" in order to
525
be used as data within a URI. Implementers should be careful not to
526
escape or unescape the same string more than once, since unescaping
527
an already unescaped string might lead to misinterpreting a percent
528
data character as another escaped character, or vice versa in the
529
case of escaping an already escaped string.
531
2.4.3. Excluded US-ASCII Characters
533
Although they are disallowed within the URI syntax, we include here a
534
description of those US-ASCII characters that have been excluded and
535
the reasons for their exclusion.
537
The control characters in the US-ASCII coded character set are not
538
used within a URI, both because they are non-printable and because
539
they are likely to be misinterpreted by some control mechanisms.
541
control = <US-ASCII coded characters 00-1F and 7F hexadecimal>
543
The space character is excluded because significant spaces may
544
disappear and insignificant spaces may be introduced when URI are
545
transcribed or typeset or subjected to the treatment of word-
546
processing programs. Whitespace is also used to delimit URI in many
549
space = <US-ASCII coded character 20 hexadecimal>
551
The angle-bracket "<" and ">" and double-quote (") characters are
552
excluded because they are often used as the delimiters around URI in
553
text documents and protocol fields. The character "#" is excluded
554
because it is used to delimit a URI from a fragment identifier in URI
555
references (Section 4). The percent character "%" is excluded because
556
it is used for the encoding of escaped characters.
558
delims = "<" | ">" | "#" | "%" | <">
562
Berners-Lee, et. al. Standards Track [Page 10]
564
RFC 2396 URI Generic Syntax August 1998
567
Other characters are excluded because gateways and other transport
568
agents are known to sometimes modify such characters, or they are
571
unwise = "{" | "}" | "|" | "\" | "^" | "[" | "]" | "`"
573
Data corresponding to excluded characters must be escaped in order to
574
be properly represented within a URI.
576
3. URI Syntactic Components
578
The URI syntax is dependent upon the scheme. In general, absolute
579
URI are written as follows:
581
<scheme>:<scheme-specific-part>
583
An absolute URI contains the name of the scheme being used (<scheme>)
584
followed by a colon (":") and then a string (the <scheme-specific-
585
part>) whose interpretation depends on the scheme.
587
The URI syntax does not require that the scheme-specific-part have
588
any general structure or set of semantics which is common among all
589
URI. However, a subset of URI do share a common syntax for
590
representing hierarchical relationships within the namespace. This
591
"generic URI" syntax consists of a sequence of four main components:
593
<scheme>://<authority><path>?<query>
595
each of which, except <scheme>, may be absent from a particular URI.
596
For example, some URI schemes do not allow an <authority> component,
597
and others do not use a <query> component.
599
absoluteURI = scheme ":" ( hier_part | opaque_part )
601
URI that are hierarchical in nature use the slash "/" character for
602
separating hierarchical components. For some file systems, a "/"
603
character (used to denote the hierarchical structure of a URI) is the
604
delimiter used to construct a file name hierarchy, and thus the URI
605
path will look similar to a file pathname. This does NOT imply that
606
the resource is a file or that the URI maps to an actual filesystem
609
hier_part = ( net_path | abs_path ) [ "?" query ]
611
net_path = "//" authority [ abs_path ]
613
abs_path = "/" path_segments
618
Berners-Lee, et. al. Standards Track [Page 11]
620
RFC 2396 URI Generic Syntax August 1998
623
URI that do not make use of the slash "/" character for separating
624
hierarchical components are considered opaque by the generic URI
627
opaque_part = uric_no_slash *uric
629
uric_no_slash = unreserved | escaped | ";" | "?" | ":" | "@" |
630
"&" | "=" | "+" | "$" | ","
632
We use the term <path> to refer to both the <abs_path> and
633
<opaque_part> constructs, since they are mutually exclusive for any
634
given URI and can be parsed as a single component.
636
3.1. Scheme Component
638
Just as there are many different methods of access to resources,
639
there are a variety of schemes for identifying such resources. The
640
URI syntax consists of a sequence of components separated by reserved
641
characters, with the first component defining the semantics for the
642
remainder of the URI string.
644
Scheme names consist of a sequence of characters beginning with a
645
lower case letter and followed by any combination of lower case
646
letters, digits, plus ("+"), period ("."), or hyphen ("-"). For
647
resiliency, programs interpreting URI should treat upper case letters
648
as equivalent to lower case in scheme names (e.g., allow "HTTP" as
651
scheme = alpha *( alpha | digit | "+" | "-" | "." )
653
Relative URI references are distinguished from absolute URI in that
654
they do not begin with a scheme name. Instead, the scheme is
655
inherited from the base URI, as described in Section 5.2.
657
3.2. Authority Component
659
Many URI schemes include a top hierarchical element for a naming
660
authority, such that the namespace defined by the remainder of the
661
URI is governed by that authority. This authority component is
662
typically defined by an Internet-based server or a scheme-specific
663
registry of naming authorities.
665
authority = server | reg_name
667
The authority component is preceded by a double slash "//" and is
668
terminated by the next slash "/", question-mark "?", or by the end of
669
the URI. Within the authority component, the characters ";", ":",
670
"@", "?", and "/" are reserved.
674
Berners-Lee, et. al. Standards Track [Page 12]
676
RFC 2396 URI Generic Syntax August 1998
679
An authority component is not required for a URI scheme to make use
680
of relative references. A base URI without an authority component
681
implies that any relative reference will also be without an authority
684
3.2.1. Registry-based Naming Authority
686
The structure of a registry-based naming authority is specific to the
687
URI scheme, but constrained to the allowed characters for an
690
reg_name = 1*( unreserved | escaped | "$" | "," |
691
";" | ":" | "@" | "&" | "=" | "+" )
693
3.2.2. Server-based Naming Authority
695
URL schemes that involve the direct use of an IP-based protocol to a
696
specified server on the Internet use a common syntax for the server
697
component of the URI's scheme-specific data:
699
<userinfo>@<host>:<port>
701
where <userinfo> may consist of a user name and, optionally, scheme-
702
specific information about how to gain authorization to access the
703
server. The parts "<userinfo>@" and ":<port>" may be omitted.
705
server = [ [ userinfo "@" ] hostport ]
707
The user information, if present, is followed by a commercial at-sign
710
userinfo = *( unreserved | escaped |
711
";" | ":" | "&" | "=" | "+" | "$" | "," )
713
Some URL schemes use the format "user:password" in the userinfo
714
field. This practice is NOT RECOMMENDED, because the passing of
715
authentication information in clear text (such as URI) has proven to
716
be a security risk in almost every case where it has been used.
718
The host is a domain name of a network host, or its IPv4 address as a
719
set of four decimal digit groups separated by ".". Literal IPv6
720
addresses are not supported.
722
hostport = host [ ":" port ]
723
host = hostname | IPv4address
724
hostname = *( domainlabel "." ) toplabel [ "." ]
725
domainlabel = alphanum | alphanum *( alphanum | "-" ) alphanum
726
toplabel = alpha | alpha *( alphanum | "-" ) alphanum
730
Berners-Lee, et. al. Standards Track [Page 13]
732
RFC 2396 URI Generic Syntax August 1998
735
IPv4address = 1*digit "." 1*digit "." 1*digit "." 1*digit
738
Hostnames take the form described in Section 3 of [RFC1034] and
739
Section 2.1 of [RFC1123]: a sequence of domain labels separated by
740
".", each domain label starting and ending with an alphanumeric
741
character and possibly also containing "-" characters. The rightmost
742
domain label of a fully qualified domain name will never start with a
743
digit, thus syntactically distinguishing domain names from IPv4
744
addresses, and may be followed by a single "." if it is necessary to
745
distinguish between the complete domain name and any local domain.
746
To actually be "Uniform" as a resource locator, a URL hostname should
747
be a fully qualified domain name. In practice, however, the host
748
component may be a local domain literal.
750
Note: A suitable representation for including a literal IPv6
751
address as the host part of a URL is desired, but has not yet been
752
determined or implemented in practice.
754
The port is the network port number for the server. Most schemes
755
designate protocols that have a default port number. Another port
756
number may optionally be supplied, in decimal, separated from the
757
host by a colon. If the port is omitted, the default port number is
762
The path component contains data, specific to the authority (or the
763
scheme if there is no authority component), identifying the resource
764
within the scope of that scheme and authority.
766
path = [ abs_path | opaque_part ]
768
path_segments = segment *( "/" segment )
769
segment = *pchar *( ";" param )
772
pchar = unreserved | escaped |
773
":" | "@" | "&" | "=" | "+" | "$" | ","
775
The path may consist of a sequence of path segments separated by a
776
single slash "/" character. Within a path segment, the characters
777
"/", ";", "=", and "?" are reserved. Each path segment may include a
778
sequence of parameters, indicated by the semicolon ";" character.
779
The parameters are not significant to the parsing of relative
786
Berners-Lee, et. al. Standards Track [Page 14]
788
RFC 2396 URI Generic Syntax August 1998
793
The query component is a string of information to be interpreted by
798
Within a query component, the characters ";", "/", "?", ":", "@",
799
"&", "=", "+", ",", and "$" are reserved.
803
The term "URI-reference" is used here to denote the common usage of a
804
resource identifier. A URI reference may be absolute or relative,
805
and may have additional information attached in the form of a
806
fragment identifier. However, "the URI" that results from such a
807
reference includes only the absolute URI after the fragment
808
identifier (if any) is removed and after any relative URI is resolved
809
to its absolute form. Although it is possible to limit the
810
discussion of URI syntax and semantics to that of the absolute
811
result, most usage of URI is within general URI references, and it is
812
impossible to obtain the URI from such a reference without also
813
parsing the fragment and resolving the relative form.
815
URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
817
The syntax for relative URI is a shortened form of that for absolute
818
URI, where some prefix of the URI is missing and certain path
819
components ("." and "..") have a special meaning when, and only when,
820
interpreting a relative path. The relative URI syntax is defined in
823
4.1. Fragment Identifier
825
When a URI reference is used to perform a retrieval action on the
826
identified resource, the optional fragment identifier, separated from
827
the URI by a crosshatch ("#") character, consists of additional
828
reference information to be interpreted by the user agent after the
829
retrieval action has been successfully completed. As such, it is not
830
part of a URI, but is often used in conjunction with a URI.
834
The semantics of a fragment identifier is a property of the data
835
resulting from a retrieval action, regardless of the type of URI used
836
in the reference. Therefore, the format and interpretation of
837
fragment identifiers is dependent on the media type [RFC2046] of the
838
retrieval result. The character restrictions described in Section 2
842
Berners-Lee, et. al. Standards Track [Page 15]
844
RFC 2396 URI Generic Syntax August 1998
847
for URI also apply to the fragment in a URI-reference. Individual
848
media types may define additional restrictions or structure within
849
the fragment for specifying different types of "partial views" that
850
can be identified within that media type.
852
A fragment identifier is only meaningful when a URI reference is
853
intended for retrieval and the result of that retrieval is a document
854
for which the identified fragment is consistently defined.
856
4.2. Same-document References
858
A URI reference that does not contain a URI is a reference to the
859
current document. In other words, an empty URI reference within a
860
document is interpreted as a reference to the start of that document,
861
and a reference containing only a fragment identifier is a reference
862
to the identified fragment of that document. Traversal of such a
863
reference should not result in an additional retrieval action.
864
However, if the URI reference occurs in a context that is always
865
intended to result in a new request, as in the case of HTML's FORM
866
element, then an empty URI reference represents the base URI of the
867
current document and should be replaced by that URI when transformed
870
4.3. Parsing a URI Reference
872
A URI reference is typically parsed according to the four main
873
components and fragment identifier in order to determine what
874
components are present and whether the reference is relative or
875
absolute. The individual components are then parsed for their
876
subparts and, if not opaque, to verify their validity.
878
Although the BNF defines what is allowed in each component, it is
879
ambiguous in terms of differentiating between an authority component
880
and a path component that begins with two slash characters. The
881
greedy algorithm is used for disambiguation: the left-most matching
882
rule soaks up as much of the URI reference string as it is capable of
883
matching. In other words, the authority component wins.
885
Readers familiar with regular expressions should see Appendix B for a
886
concrete parsing example and test oracle.
888
5. Relative URI References
890
It is often the case that a group or "tree" of documents has been
891
constructed to serve a common purpose; the vast majority of URI in
892
these documents point to resources within the tree rather than
898
Berners-Lee, et. al. Standards Track [Page 16]
900
RFC 2396 URI Generic Syntax August 1998
903
outside of it. Similarly, documents located at a particular site are
904
much more likely to refer to other resources at that site than to
905
resources at remote sites.
907
Relative addressing of URI allows document trees to be partially
908
independent of their location and access scheme. For instance, it is
909
possible for a single set of hypertext documents to be simultaneously
910
accessible and traversable via each of the "file", "http", and "ftp"
911
schemes if the documents refer to each other using relative URI.
912
Furthermore, such document trees can be moved, as a whole, without
913
changing any of the relative references. Experience within the WWW
914
has demonstrated that the ability to perform relative referencing is
915
necessary for the long-term usability of embedded URI.
917
The syntax for relative URI takes advantage of the <hier_part> syntax
918
of <absoluteURI> (Section 3) in order to express a reference that is
919
relative to the namespace of another hierarchical URI.
921
relativeURI = ( net_path | abs_path | rel_path ) [ "?" query ]
923
A relative reference beginning with two slash characters is termed a
924
network-path reference, as defined by <net_path> in Section 3. Such
925
references are rarely used.
927
A relative reference beginning with a single slash character is
928
termed an absolute-path reference, as defined by <abs_path> in
931
A relative reference that does not begin with a scheme name or a
932
slash character is termed a relative-path reference.
934
rel_path = rel_segment [ abs_path ]
936
rel_segment = 1*( unreserved | escaped |
937
";" | "@" | "&" | "=" | "+" | "$" | "," )
939
Within a relative-path reference, the complete path segments "." and
940
".." have special meanings: "the current hierarchy level" and "the
941
level above this hierarchy level", respectively. Although this is
942
very similar to their use within Unix-based filesystems to indicate
943
directory levels, these path components are only considered special
944
when resolving a relative-path reference to its absolute form
947
Authors should be aware that a path segment which contains a colon
948
character cannot be used as the first segment of a relative URI path
949
(e.g., "this:that"), because it would be mistaken for a scheme name.
954
Berners-Lee, et. al. Standards Track [Page 17]
956
RFC 2396 URI Generic Syntax August 1998
959
It is therefore necessary to precede such segments with other
960
segments (e.g., "./this:that") in order for them to be referenced as
963
It is not necessary for all URI within a given scheme to be
964
restricted to the <hier_part> syntax, since the hierarchical
965
properties of that syntax are only necessary when relative URI are
966
used within a particular document. Documents can only make use of
967
relative URI when their base URI fits within the <hier_part> syntax.
968
It is assumed that any document which contains a relative reference
969
will also have a base URI that obeys the syntax. In other words,
970
relative URI cannot be used within a document that has an unsuitable
973
Some URI schemes do not allow a hierarchical syntax matching the
974
<hier_part> syntax, and thus cannot use relative references.
976
5.1. Establishing a Base URI
978
The term "relative URI" implies that there exists some absolute "base
979
URI" against which the relative reference is applied. Indeed, the
980
base URI is necessary to define the semantics of any relative URI
981
reference; without it, a relative reference is meaningless. In order
982
for relative URI to be usable within a document, the base URI of that
983
document must be known to the parser.
985
The base URI of a document can be established in one of four ways,
986
listed below in order of precedence. The order of precedence can be
987
thought of in terms of layers, where the innermost defined base URI
988
has the highest precedence. This can be visualized graphically as:
990
.----------------------------------------------------------.
991
| .----------------------------------------------------. |
992
| | .----------------------------------------------. | |
993
| | | .----------------------------------------. | | |
994
| | | | .----------------------------------. | | | |
995
| | | | | <relative_reference> | | | | |
996
| | | | `----------------------------------' | | | |
997
| | | | (5.1.1) Base URI embedded in the | | | |
998
| | | | document's content | | | |
999
| | | `----------------------------------------' | | |
1000
| | | (5.1.2) Base URI of the encapsulating entity | | |
1001
| | | (message, document, or none). | | |
1002
| | `----------------------------------------------' | |
1003
| | (5.1.3) URI used to retrieve the entity | |
1004
| `----------------------------------------------------' |
1005
| (5.1.4) Default Base URI is application-dependent |
1006
`----------------------------------------------------------'
1010
Berners-Lee, et. al. Standards Track [Page 18]
1012
RFC 2396 URI Generic Syntax August 1998
1015
5.1.1. Base URI within Document Content
1017
Within certain document media types, the base URI of the document can
1018
be embedded within the content itself such that it can be readily
1019
obtained by a parser. This can be useful for descriptive documents,
1020
such as tables of content, which may be transmitted to others through
1021
protocols other than their usual retrieval context (e.g., E-Mail or
1024
It is beyond the scope of this document to specify how, for each
1025
media type, the base URI can be embedded. It is assumed that user
1026
agents manipulating such media types will be able to obtain the
1027
appropriate syntax from that media type's specification. An example
1028
of how the base URI can be embedded in the Hypertext Markup Language
1029
(HTML) [RFC1866] is provided in Appendix D.
1031
A mechanism for embedding the base URI within MIME container types
1032
(e.g., the message and multipart types) is defined by MHTML
1033
[RFC2110]. Protocols that do not use the MIME message header syntax,
1034
but which do allow some form of tagged metainformation to be included
1035
within messages, may define their own syntax for defining the base
1036
URI as part of a message.
1038
5.1.2. Base URI from the Encapsulating Entity
1040
If no base URI is embedded, the base URI of a document is defined by
1041
the document's retrieval context. For a document that is enclosed
1042
within another entity (such as a message or another document), the
1043
retrieval context is that entity; thus, the default base URI of the
1044
document is the base URI of the entity in which the document is
1047
5.1.3. Base URI from the Retrieval URI
1049
If no base URI is embedded and the document is not encapsulated
1050
within some other entity (e.g., the top level of a composite entity),
1051
then, if a URI was used to retrieve the base document, that URI shall
1052
be considered the base URI. Note that if the retrieval was the
1053
result of a redirected request, the last URI used (i.e., that which
1054
resulted in the actual retrieval of the document) is the base URI.
1056
5.1.4. Default Base URI
1058
If none of the conditions described in Sections 5.1.1--5.1.3 apply,
1059
then the base URI is defined by the context of the application.
1060
Since this definition is necessarily application-dependent, failing
1066
Berners-Lee, et. al. Standards Track [Page 19]
1068
RFC 2396 URI Generic Syntax August 1998
1071
to define the base URI using one of the other methods may result in
1072
the same content being interpreted differently by different types of
1075
It is the responsibility of the distributor(s) of a document
1076
containing relative URI to ensure that the base URI for that document
1077
can be established. It must be emphasized that relative URI cannot
1078
be used reliably in situations where the document's base URI is not
1081
5.2. Resolving Relative References to Absolute Form
1083
This section describes an example algorithm for resolving URI
1084
references that might be relative to a given base URI.
1086
The base URI is established according to the rules of Section 5.1 and
1087
parsed into the four main components as described in Section 3. Note
1088
that only the scheme component is required to be present in the base
1089
URI; the other components may be empty or undefined. A component is
1090
undefined if its preceding separator does not appear in the URI
1091
reference; the path component is never undefined, though it may be
1092
empty. The base URI's query component is not used by the resolution
1093
algorithm and may be discarded.
1095
For each URI reference, the following steps are performed in order:
1097
1) The URI reference is parsed into the potential four components and
1098
fragment identifier, as described in Section 4.3.
1100
2) If the path component is empty and the scheme, authority, and
1101
query components are undefined, then it is a reference to the
1102
current document and we are done. Otherwise, the reference URI's
1103
query and fragment components are defined as found (or not found)
1104
within the URI reference and not inherited from the base URI.
1106
3) If the scheme component is defined, indicating that the reference
1107
starts with a scheme name, then the reference is interpreted as an
1108
absolute URI and we are done. Otherwise, the reference URI's
1109
scheme is inherited from the base URI's scheme component.
1111
Due to a loophole in prior specifications [RFC1630], some parsers
1112
allow the scheme name to be present in a relative URI if it is the
1113
same as the base URI scheme. Unfortunately, this can conflict
1114
with the correct parsing of non-hierarchical URI. For backwards
1115
compatibility, an implementation may work around such references
1116
by removing the scheme if it matches that of the base URI and the
1117
scheme is known to always use the <hier_part> syntax. The parser
1122
Berners-Lee, et. al. Standards Track [Page 20]
1124
RFC 2396 URI Generic Syntax August 1998
1127
can then continue with the steps below for the remainder of the
1128
reference components. Validating parsers should mark such a
1129
misformed relative reference as an error.
1131
4) If the authority component is defined, then the reference is a
1132
network-path and we skip to step 7. Otherwise, the reference
1133
URI's authority is inherited from the base URI's authority
1134
component, which will also be undefined if the URI scheme does not
1135
use an authority component.
1137
5) If the path component begins with a slash character ("/"), then
1138
the reference is an absolute-path and we skip to step 7.
1140
6) If this step is reached, then we are resolving a relative-path
1141
reference. The relative path needs to be merged with the base
1142
URI's path. Although there are many ways to do this, we will
1143
describe a simple method using a separate string buffer.
1145
a) All but the last segment of the base URI's path component is
1146
copied to the buffer. In other words, any characters after the
1147
last (right-most) slash character, if any, are excluded.
1149
b) The reference's path component is appended to the buffer
1152
c) All occurrences of "./", where "." is a complete path segment,
1153
are removed from the buffer string.
1155
d) If the buffer string ends with "." as a complete path segment,
1156
that "." is removed.
1158
e) All occurrences of "<segment>/../", where <segment> is a
1159
complete path segment not equal to "..", are removed from the
1160
buffer string. Removal of these path segments is performed
1161
iteratively, removing the leftmost matching pattern on each
1162
iteration, until no matching pattern remains.
1164
f) If the buffer string ends with "<segment>/..", where <segment>
1165
is a complete path segment not equal to "..", that
1166
"<segment>/.." is removed.
1168
g) If the resulting buffer string still begins with one or more
1169
complete path segments of "..", then the reference is
1170
considered to be in error. Implementations may handle this
1171
error by retaining these components in the resolved path (i.e.,
1172
treating them as part of the final URI), by removing them from
1173
the resolved path (i.e., discarding relative levels above the
1174
root), or by avoiding traversal of the reference.
1178
Berners-Lee, et. al. Standards Track [Page 21]
1180
RFC 2396 URI Generic Syntax August 1998
1183
h) The remaining buffer string is the reference URI's new path
1186
7) The resulting URI components, including any inherited from the
1187
base URI, are recombined to give the absolute form of the URI
1188
reference. Using pseudocode, this would be
1192
if scheme is defined then
1193
append scheme to result
1194
append ":" to result
1196
if authority is defined then
1197
append "//" to result
1198
append authority to result
1200
append path to result
1202
if query is defined then
1203
append "?" to result
1204
append query to result
1206
if fragment is defined then
1207
append "#" to result
1208
append fragment to result
1212
Note that we must be careful to preserve the distinction between a
1213
component that is undefined, meaning that its separator was not
1214
present in the reference, and a component that is empty, meaning
1215
that the separator was present and was immediately followed by the
1216
next component separator or the end of the reference.
1218
The above algorithm is intended to provide an example by which the
1219
output of implementations can be tested -- implementation of the
1220
algorithm itself is not required. For example, some systems may find
1221
it more efficient to implement step 6 as a pair of segment stacks
1222
being merged, rather than as a series of string pattern replacements.
1224
Note: Some WWW client applications will fail to separate the
1225
reference's query component from its path component before merging
1226
the base and reference paths in step 6 above. This may result in
1227
a loss of information if the query component contains the strings
1230
Resolution examples are provided in Appendix C.
1234
Berners-Lee, et. al. Standards Track [Page 22]
1236
RFC 2396 URI Generic Syntax August 1998
1239
6. URI Normalization and Equivalence
1241
In many cases, different URI strings may actually identify the
1242
identical resource. For example, the host names used in URL are
1243
actually case insensitive, and the URL <http://www.XEROX.com> is
1244
equivalent to <http://www.xerox.com>. In general, the rules for
1245
equivalence and definition of a normal form, if any, are scheme
1246
dependent. When a scheme uses elements of the common syntax, it will
1247
also use the common syntax equivalence rules, namely that the scheme
1248
and hostname are case insensitive and a URL with an explicit ":port",
1249
where the port is the default for the scheme, is equivalent to one
1250
where the port is elided.
1252
7. Security Considerations
1254
A URI does not in itself pose a security threat. Users should beware
1255
that there is no general guarantee that a URL, which at one time
1256
located a given resource, will continue to do so. Nor is there any
1257
guarantee that a URL will not locate a different resource at some
1258
later point in time, due to the lack of any constraint on how a given
1259
authority apportions its namespace. Such a guarantee can only be
1260
obtained from the person(s) controlling that namespace and the
1261
resource in question. A specific URI scheme may include additional
1262
semantics, such as name persistence, if those semantics are required
1263
of all naming authorities for that scheme.
1265
It is sometimes possible to construct a URL such that an attempt to
1266
perform a seemingly harmless, idempotent operation, such as the
1267
retrieval of an entity associated with the resource, will in fact
1268
cause a possibly damaging remote operation to occur. The unsafe URL
1269
is typically constructed by specifying a port number other than that
1270
reserved for the network protocol in question. The client
1271
unwittingly contacts a site that is in fact running a different
1272
protocol. The content of the URL contains instructions that, when
1273
interpreted according to this other protocol, cause an unexpected
1274
operation. An example has been the use of a gopher URL to cause an
1275
unintended or impersonating message to be sent via a SMTP server.
1277
Caution should be used when using any URL that specifies a port
1278
number other than the default for the protocol, especially when it is
1279
a number within the reserved space.
1281
Care should be taken when a URL contains escaped delimiters for a
1282
given protocol (for example, CR and LF characters for telnet
1283
protocols) that these are not unescaped before transmission. This
1284
might violate the protocol, but avoids the potential for such
1290
Berners-Lee, et. al. Standards Track [Page 23]
1292
RFC 2396 URI Generic Syntax August 1998
1295
characters to be used to simulate an extra operation or parameter in
1296
that protocol, which might lead to an unexpected and possibly harmful
1297
remote operation to be performed.
1299
It is clearly unwise to use a URL that contains a password which is
1300
intended to be secret. In particular, the use of a password within
1301
the 'userinfo' component of a URL is strongly disrecommended except
1302
in those rare cases where the 'password' parameter is intended to be
1307
This document was derived from RFC 1738 [RFC1738] and RFC 1808
1308
[RFC1808]; the acknowledgements in those specifications still apply.
1309
In addition, contributions by Gisle Aas, Martin Beet, Martin Duerst,
1310
Jim Gettys, Martijn Koster, Dave Kristol, Daniel LaLiberte, Foteos
1311
Macrides, James Marshall, Ryan Moats, Keith Moore, and Lauren Wood
1312
are gratefully acknowledged.
1316
[RFC2277] Alvestrand, H., "IETF Policy on Character Sets and
1317
Languages", BCP 18, RFC 2277, January 1998.
1319
[RFC1630] Berners-Lee, T., "Universal Resource Identifiers in WWW: A
1320
Unifying Syntax for the Expression of Names and Addresses
1321
of Objects on the Network as used in the World-Wide Web",
1322
RFC 1630, June 1994.
1324
[RFC1738] Berners-Lee, T., Masinter, L., and M. McCahill, Editors,
1325
"Uniform Resource Locators (URL)", RFC 1738, December 1994.
1327
[RFC1866] Berners-Lee T., and D. Connolly, "HyperText Markup Language
1328
Specification -- 2.0", RFC 1866, November 1995.
1330
[RFC1123] Braden, R., Editor, "Requirements for Internet Hosts --
1331
Application and Support", STD 3, RFC 1123, October 1989.
1333
[RFC822] Crocker, D., "Standard for the Format of ARPA Internet Text
1334
Messages", STD 11, RFC 822, August 1982.
1336
[RFC1808] Fielding, R., "Relative Uniform Resource Locators", RFC
1339
[RFC2046] Freed, N., and N. Borenstein, "Multipurpose Internet Mail
1340
Extensions (MIME) Part Two: Media Types", RFC 2046,
1346
Berners-Lee, et. al. Standards Track [Page 24]
1348
RFC 2396 URI Generic Syntax August 1998
1351
[RFC1736] Kunze, J., "Functional Recommendations for Internet
1352
Resource Locators", RFC 1736, February 1995.
1354
[RFC2141] Moats, R., "URN Syntax", RFC 2141, May 1997.
1356
[RFC1034] Mockapetris, P., "Domain Names - Concepts and Facilities",
1357
STD 13, RFC 1034, November 1987.
1359
[RFC2110] Palme, J., and A. Hopmann, "MIME E-mail Encapsulation of
1360
Aggregate Documents, such as HTML (MHTML)", RFC 2110, March
1363
[RFC1737] Sollins, K., and L. Masinter, "Functional Requirements for
1364
Uniform Resource Names", RFC 1737, December 1994.
1366
[ASCII] US-ASCII. "Coded Character Set -- 7-bit American Standard
1367
Code for Information Interchange", ANSI X3.4-1986.
1369
[UTF-8] Yergeau, F., "UTF-8, a transformation format of ISO 10646",
1370
RFC 2279, January 1998.
1402
Berners-Lee, et. al. Standards Track [Page 25]
1404
RFC 2396 URI Generic Syntax August 1998
1407
10. Authors' Addresses
1410
World Wide Web Consortium
1411
MIT Laboratory for Computer Science, NE43-356
1412
545 Technology Square
1415
Fax: +1(617)258-8682
1420
Department of Information and Computer Science
1421
University of California, Irvine
1422
Irvine, CA 92697-3425
1424
Fax: +1(949)824-1715
1425
EMail: fielding@ics.uci.edu
1430
3333 Coyote Hill Road
1433
Fax: +1(415)812-4333
1434
EMail: masinter@parc.xerox.com
1458
Berners-Lee, et. al. Standards Track [Page 26]
1460
RFC 2396 URI Generic Syntax August 1998
1463
A. Collected BNF for URI
1465
URI-reference = [ absoluteURI | relativeURI ] [ "#" fragment ]
1466
absoluteURI = scheme ":" ( hier_part | opaque_part )
1467
relativeURI = ( net_path | abs_path | rel_path ) [ "?" query ]
1469
hier_part = ( net_path | abs_path ) [ "?" query ]
1470
opaque_part = uric_no_slash *uric
1472
uric_no_slash = unreserved | escaped | ";" | "?" | ":" | "@" |
1473
"&" | "=" | "+" | "$" | ","
1475
net_path = "//" authority [ abs_path ]
1476
abs_path = "/" path_segments
1477
rel_path = rel_segment [ abs_path ]
1479
rel_segment = 1*( unreserved | escaped |
1480
";" | "@" | "&" | "=" | "+" | "$" | "," )
1482
scheme = alpha *( alpha | digit | "+" | "-" | "." )
1484
authority = server | reg_name
1486
reg_name = 1*( unreserved | escaped | "$" | "," |
1487
";" | ":" | "@" | "&" | "=" | "+" )
1489
server = [ [ userinfo "@" ] hostport ]
1490
userinfo = *( unreserved | escaped |
1491
";" | ":" | "&" | "=" | "+" | "$" | "," )
1493
hostport = host [ ":" port ]
1494
host = hostname | IPv4address
1495
hostname = *( domainlabel "." ) toplabel [ "." ]
1496
domainlabel = alphanum | alphanum *( alphanum | "-" ) alphanum
1497
toplabel = alpha | alpha *( alphanum | "-" ) alphanum
1498
IPv4address = 1*digit "." 1*digit "." 1*digit "." 1*digit
1501
path = [ abs_path | opaque_part ]
1502
path_segments = segment *( "/" segment )
1503
segment = *pchar *( ";" param )
1505
pchar = unreserved | escaped |
1506
":" | "@" | "&" | "=" | "+" | "$" | ","
1514
Berners-Lee, et. al. Standards Track [Page 27]
1516
RFC 2396 URI Generic Syntax August 1998
1519
uric = reserved | unreserved | escaped
1520
reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |
1522
unreserved = alphanum | mark
1523
mark = "-" | "_" | "." | "!" | "~" | "*" | "'" |
1526
escaped = "%" hex hex
1527
hex = digit | "A" | "B" | "C" | "D" | "E" | "F" |
1528
"a" | "b" | "c" | "d" | "e" | "f"
1530
alphanum = alpha | digit
1531
alpha = lowalpha | upalpha
1533
lowalpha = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" | "i" |
1534
"j" | "k" | "l" | "m" | "n" | "o" | "p" | "q" | "r" |
1535
"s" | "t" | "u" | "v" | "w" | "x" | "y" | "z"
1536
upalpha = "A" | "B" | "C" | "D" | "E" | "F" | "G" | "H" | "I" |
1537
"J" | "K" | "L" | "M" | "N" | "O" | "P" | "Q" | "R" |
1538
"S" | "T" | "U" | "V" | "W" | "X" | "Y" | "Z"
1539
digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" |
1570
Berners-Lee, et. al. Standards Track [Page 28]
1572
RFC 2396 URI Generic Syntax August 1998
1575
B. Parsing a URI Reference with a Regular Expression
1577
As described in Section 4.3, the generic URI syntax is not sufficient
1578
to disambiguate the components of some forms of URI. Since the
1579
"greedy algorithm" described in that section is identical to the
1580
disambiguation method used by POSIX regular expressions, it is
1581
natural and commonplace to use a regular expression for parsing the
1582
potential four components and fragment identifier of a URI reference.
1584
The following line is the regular expression for breaking-down a URI
1585
reference into its components.
1587
^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))?
1590
The numbers in the second line above are only to assist readability;
1591
they indicate the reference points for each subexpression (i.e., each
1592
paired parenthesis). We refer to the value matched for subexpression
1593
<n> as $<n>. For example, matching the above expression to
1595
http://www.ics.uci.edu/pub/ietf/uri/#Related
1597
results in the following subexpression matches:
1601
$3 = //www.ics.uci.edu
1602
$4 = www.ics.uci.edu
1609
where <undefined> indicates that the component is not present, as is
1610
the case for the query component in the above example. Therefore, we
1611
can determine the value of the four components and fragment as
1619
and, going in the opposite direction, we can recreate a URI reference
1620
from its components using the algorithm in step 7 of Section 5.2.
1626
Berners-Lee, et. al. Standards Track [Page 29]
1628
RFC 2396 URI Generic Syntax August 1998
1631
C. Examples of Resolving Relative URI References
1633
Within an object with a well-defined base URI of
1637
the relative URI would be resolved as follows:
1639
C.1. Normal Examples
1643
./g = http://a/b/c/g
1644
g/ = http://a/b/c/g/
1647
?y = http://a/b/c/?y
1648
g?y = http://a/b/c/g?y
1649
#s = (current document)#s
1650
g#s = http://a/b/c/g#s
1651
g?y#s = http://a/b/c/g?y#s
1652
;x = http://a/b/c/;x
1653
g;x = http://a/b/c/g;x
1654
g;x?y#s = http://a/b/c/g;x?y#s
1662
../../g = http://a/g
1664
C.2. Abnormal Examples
1666
Although the following abnormal examples are unlikely to occur in
1667
normal practice, all URI parsers should be capable of resolving them
1668
consistently. Each example uses the same base as above.
1670
An empty reference refers to the start of the current document.
1672
<> = (current document)
1674
Parsers must be careful in handling the case where there are more
1675
relative path ".." segments than there are hierarchical levels in the
1676
base URI's path. Note that the ".." syntax cannot be used to change
1677
the authority component of a URI.
1682
Berners-Lee, et. al. Standards Track [Page 30]
1684
RFC 2396 URI Generic Syntax August 1998
1687
../../../g = http://a/../g
1688
../../../../g = http://a/../../g
1690
In practice, some implementations strip leading relative symbolic
1691
elements (".", "..") after applying a relative URI calculation, based
1692
on the theory that compensating for obvious author errors is better
1693
than allowing the request to fail. Thus, the above two references
1694
will be interpreted as "http://a/g" by some implementations.
1696
Similarly, parsers must avoid treating "." and ".." as special when
1697
they are not complete components of a relative path.
1700
/../g = http://a/../g
1701
g. = http://a/b/c/g.
1702
.g = http://a/b/c/.g
1703
g.. = http://a/b/c/g..
1704
..g = http://a/b/c/..g
1706
Less likely are cases where the relative URI uses unnecessary or
1707
nonsensical forms of the "." and ".." complete path segments.
1709
./../g = http://a/b/g
1710
./g/. = http://a/b/c/g/
1711
g/./h = http://a/b/c/g/h
1712
g/../h = http://a/b/c/h
1713
g;x=1/./y = http://a/b/c/g;x=1/y
1714
g;x=1/../y = http://a/b/c/y
1716
All client applications remove the query component from the base URI
1717
before resolving relative URI. However, some applications fail to
1718
separate the reference's query and/or fragment components from a
1719
relative path before merging it with the base path. This error is
1720
rarely noticed, since typical usage of a fragment never includes the
1721
hierarchy ("/") character, and the query component is not normally
1722
used within relative references.
1724
g?y/./x = http://a/b/c/g?y/./x
1725
g?y/../x = http://a/b/c/g?y/../x
1726
g#s/./x = http://a/b/c/g#s/./x
1727
g#s/../x = http://a/b/c/g#s/../x
1738
Berners-Lee, et. al. Standards Track [Page 31]
1740
RFC 2396 URI Generic Syntax August 1998
1743
Some parsers allow the scheme name to be present in a relative URI if
1744
it is the same as the base URI scheme. This is considered to be a
1745
loophole in prior specifications of partial URI [RFC1630]. Its use
1748
http:g = http:g ; for validating parsers
1749
| http://a/b/c/g ; for backwards compatibility
1794
Berners-Lee, et. al. Standards Track [Page 32]
1796
RFC 2396 URI Generic Syntax August 1998
1799
D. Embedding the Base URI in HTML documents
1801
It is useful to consider an example of how the base URI of a document
1802
can be embedded within the document's content. In this appendix, we
1803
describe how documents written in the Hypertext Markup Language
1804
(HTML) [RFC1866] can include an embedded base URI. This appendix
1805
does not form a part of the URI specification and should not be
1806
considered as anything more than a descriptive example.
1808
HTML defines a special element "BASE" which, when present in the
1809
"HEAD" portion of a document, signals that the parser should use the
1810
BASE element's "HREF" attribute as the base URI for resolving any
1811
relative URI. The "HREF" attribute must be an absolute URI. Note
1812
that, in HTML, element and attribute names are case-insensitive. For
1815
<!doctype html public "-//IETF//DTD HTML//EN">
1817
<TITLE>An example HTML document</TITLE>
1818
<BASE href="http://www.ics.uci.edu/Test/a/b/c">
1820
... <A href="../x">a hypertext anchor</A> ...
1823
A parser reading the example document should interpret the given
1824
relative URI "../x" as representing the absolute URI
1826
<http://www.ics.uci.edu/Test/a/x>
1828
regardless of the context in which the example document was obtained.
1850
Berners-Lee, et. al. Standards Track [Page 33]
1852
RFC 2396 URI Generic Syntax August 1998
1855
E. Recommendations for Delimiting URI in Context
1857
URI are often transmitted through formats that do not provide a clear
1858
context for their interpretation. For example, there are many
1859
occasions when URI are included in plain text; examples include text
1860
sent in electronic mail, USENET news messages, and, most importantly,
1861
printed on paper. In such cases, it is important to be able to
1862
delimit the URI from the rest of the text, and in particular from
1863
punctuation marks that might be mistaken for part of the URI.
1865
In practice, URI are delimited in a variety of ways, but usually
1866
within double-quotes "http://test.com/", angle brackets
1867
<http://test.com/>, or just using whitespace
1871
These wrappers do not form part of the URI.
1873
In the case where a fragment identifier is associated with a URI
1874
reference, the fragment would be placed within the brackets as well
1875
(separated from the URI with a "#" character).
1877
In some cases, extra whitespace (spaces, linebreaks, tabs, etc.) may
1878
need to be added to break long URI across lines. The whitespace
1879
should be ignored when extracting the URI.
1881
No whitespace should be introduced after a hyphen ("-") character.
1882
Because some typesetters and printers may (erroneously) introduce a
1883
hyphen at the end of line when breaking a line, the interpreter of a
1884
URI containing a line break immediately after a hyphen should ignore
1885
all unescaped whitespace around the line break, and should be aware
1886
that the hyphen may or may not actually be part of the URI.
1888
Using <> angle brackets around each URI is especially recommended as
1889
a delimiting style for URI that contain whitespace.
1891
The prefix "URL:" (with or without a trailing space) was recommended
1892
as a way to used to help distinguish a URL from other bracketed
1893
designators, although this is not common in practice.
1895
For robustness, software that accepts user-typed URI should attempt
1896
to recognize and strip both delimiters and embedded whitespace.
1898
For example, the text:
1906
Berners-Lee, et. al. Standards Track [Page 34]
1908
RFC 2396 URI Generic Syntax August 1998
1911
Yes, Jim, I found it under "http://www.w3.org/Addressing/",
1912
but you can probably pick it up from <ftp://ds.internic.
1913
net/rfc/>. Note the warning in <http://www.ics.uci.edu/pub/
1914
ietf/uri/historical.html#WARNING>.
1916
contains the URI references
1918
http://www.w3.org/Addressing/
1919
ftp://ds.internic.net/rfc/
1920
http://www.ics.uci.edu/pub/ietf/uri/historical.html#WARNING
1962
Berners-Lee, et. al. Standards Track [Page 35]
1964
RFC 2396 URI Generic Syntax August 1998
1969
The URL syntax was designed for unambiguous reference to network
1970
resources and extensibility via the URL scheme. However, as URL
1971
identification and usage have become commonplace, traditional media
1972
(television, radio, newspapers, billboards, etc.) have increasingly
1973
used abbreviated URL references. That is, a reference consisting of
1974
only the authority and path portions of the identified resource, such
1977
www.w3.org/Addressing/
1979
or simply the DNS hostname on its own. Such references are primarily
1980
intended for human interpretation rather than machine, with the
1981
assumption that context-based heuristics are sufficient to complete
1982
the URL (e.g., most hostnames beginning with "www" are likely to have
1983
a URL prefix of "http://"). Although there is no standard set of
1984
heuristics for disambiguating abbreviated URL references, many client
1985
implementations allow them to be entered by the user and
1986
heuristically resolved. It should be noted that such heuristics may
1987
change over time, particularly when new URL schemes are introduced.
1989
Since an abbreviated URL has the same syntax as a relative URL path,
1990
abbreviated URL references cannot be used in contexts where relative
1991
URLs are expected. This limits the use of abbreviated URLs to places
1992
where there is no defined base URL, such as dialog boxes and off-line
2018
Berners-Lee, et. al. Standards Track [Page 36]
2020
RFC 2396 URI Generic Syntax August 1998
2023
G. Summary of Non-editorial Changes
2027
Section 4 (URI References) was added to stem the confusion regarding
2028
"what is a URI" and how to describe fragment identifiers given that
2029
they are not part of the URI, but are part of the URI syntax and
2030
parsing concerns. In addition, it provides a reference definition
2031
for use by other IETF specifications (HTML, HTTP, etc.) that have
2032
previously attempted to redefine the URI syntax in order to account
2033
for the presence of fragment identifiers in URI references.
2035
Section 2.4 was rewritten to clarify a number of misinterpretations
2036
and to leave room for fully internationalized URI.
2038
Appendix F on abbreviated URLs was added to describe the shortened
2039
references often seen on television and magazine advertisements and
2040
explain why they are not used in other contexts.
2042
G.2. Modifications from both RFC 1738 and RFC 1808
2044
Changed to URI syntax instead of just URL.
2046
Confusion regarding the terms "character encoding", the URI
2047
"character set", and the escaping of characters with %<hex><hex>
2048
equivalents has (hopefully) been reduced. Many of the BNF rule names
2049
regarding the character sets have been changed to more accurately
2050
describe their purpose and to encompass all "characters" rather than
2051
just US-ASCII octets. Unless otherwise noted here, these
2052
modifications do not affect the URI syntax.
2054
Both RFC 1738 and RFC 1808 refer to the "reserved" set of characters
2055
as if URI-interpreting software were limited to a single set of
2056
characters with a reserved purpose (i.e., as meaning something other
2057
than the data to which the characters correspond), and that this set
2058
was fixed by the URI scheme. However, this has not been true in
2059
practice; any character that is interpreted differently when it is
2060
escaped is, in effect, reserved. Furthermore, the interpreting
2061
engine on a HTTP server is often dependent on the resource, not just
2062
the URI scheme. The description of reserved characters has been
2063
changed accordingly.
2065
The plus "+", dollar "$", and comma "," characters have been added to
2066
those in the "reserved" set, since they are treated as reserved
2067
within the query component.
2074
Berners-Lee, et. al. Standards Track [Page 37]
2076
RFC 2396 URI Generic Syntax August 1998
2079
The tilde "~" character was added to those in the "unreserved" set,
2080
since it is extensively used on the Internet in spite of the
2081
difficulty to transcribe it with some keyboards.
2083
The syntax for URI scheme has been changed to require that all
2084
schemes begin with an alpha character.
2086
The "user:password" form in the previous BNF was changed to a
2087
"userinfo" token, and the possibility that it might be
2088
"user:password" made scheme specific. In particular, the use of
2089
passwords in the clear is not even suggested by the syntax.
2091
The question-mark "?" character was removed from the set of allowed
2092
characters for the userinfo in the authority component, since testing
2093
showed that many applications treat it as reserved for separating the
2094
query component from the rest of the URI.
2096
The semicolon ";" character was added to those stated as being
2097
reserved within the authority component, since several new schemes
2098
are using it as a separator within userinfo to indicate the type of
2099
user authentication.
2101
RFC 1738 specified that the path was separated from the authority
2102
portion of a URI by a slash. RFC 1808 followed suit, but with a
2103
fudge of carrying around the separator as a "prefix" in order to
2104
describe the parsing algorithm. RFC 1630 never had this problem,
2105
since it considered the slash to be part of the path. In writing
2106
this specification, it was found to be impossible to accurately
2107
describe and retain the difference between the two URI
2108
<foo:/bar> and <foo:bar>
2109
without either considering the slash to be part of the path (as
2110
corresponds to actual practice) or creating a separate component just
2111
to hold that slash. We chose the former.
2113
G.3. Modifications from RFC 1738
2115
The definition of specific URL schemes and their scheme-specific
2116
syntax and semantics has been moved to separate documents.
2118
The URL host was defined as a fully-qualified domain name. However,
2119
many URLs are used without fully-qualified domain names (in contexts
2120
for which the full qualification is not necessary), without any host
2121
(as in some file URLs), or with a host of "localhost".
2123
The URL port is now *digit instead of 1*digit, since systems are
2124
expected to handle the case where the ":" separator between host and
2125
port is supplied without a port.
2130
Berners-Lee, et. al. Standards Track [Page 38]
2132
RFC 2396 URI Generic Syntax August 1998
2135
The recommendations for delimiting URI in context (Appendix E) have
2136
been adjusted to reflect current practice.
2138
G.4. Modifications from RFC 1808
2140
RFC 1808 (Section 4) defined an empty URL reference (a reference
2141
containing nothing aside from the fragment identifier) as being a
2142
reference to the base URL. Unfortunately, that definition could be
2143
interpreted, upon selection of such a reference, as a new retrieval
2144
action on that resource. Since the normal intent of such references
2145
is for the user agent to change its view of the current document to
2146
the beginning of the specified fragment within that document, not to
2147
make an additional request of the resource, a description of how to
2148
correctly interpret an empty reference has been added in Section 4.
2150
The description of the mythical Base header field has been replaced
2151
with a reference to the Content-Location header field defined by
2154
RFC 1808 described various schemes as either having or not having the
2155
properties of the generic URI syntax. However, the only requirement
2156
is that the particular document containing the relative references
2157
have a base URI that abides by the generic URI syntax, regardless of
2158
the URI scheme, so the associated description has been updated to
2161
The BNF term <net_loc> has been replaced with <authority>, since the
2162
latter more accurately describes its use and purpose. Likewise, the
2163
authority is no longer restricted to the IP server syntax.
2165
Extensive testing of current client applications demonstrated that
2166
the majority of deployed systems do not use the ";" character to
2167
indicate trailing parameter information, and that the presence of a
2168
semicolon in a path segment does not affect the relative parsing of
2169
that segment. Therefore, parameters have been removed as a separate
2170
component and may now appear in any path segment. Their influence
2171
has been removed from the algorithm for resolving a relative URI
2172
reference. The resolution examples in Appendix C have been modified
2173
to reflect this change.
2175
Implementations are now allowed to work around misformed relative
2176
references that are prefixed by the same scheme as the base URI, but
2177
only for schemes known to use the <hier_part> syntax.
2186
Berners-Lee, et. al. Standards Track [Page 39]
2188
RFC 2396 URI Generic Syntax August 1998
2191
H. Full Copyright Statement
2193
Copyright (C) The Internet Society (1998). All Rights Reserved.
2195
This document and translations of it may be copied and furnished to
2196
others, and derivative works that comment on or otherwise explain it
2197
or assist in its implementation may be prepared, copied, published
2198
and distributed, in whole or in part, without restriction of any
2199
kind, provided that the above copyright notice and this paragraph are
2200
included on all such copies and derivative works. However, this
2201
document itself may not be modified in any way, such as by removing
2202
the copyright notice or references to the Internet Society or other
2203
Internet organizations, except as needed for the purpose of
2204
developing Internet standards in which case the procedures for
2205
copyrights defined in the Internet Standards process must be
2206
followed, or as required to translate it into languages other than
2209
The limited permissions granted above are perpetual and will not be
2210
revoked by the Internet Society or its successors or assigns.
2212
This document and the information contained herein is provided on an
2213
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
2214
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
2215
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
2216
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
2217
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
2242
Berners-Lee, et. al. Standards Track [Page 40]