source: draft-ietf-iri-3987bis/draft-ietf-iri-3987bis.xml @ 45

Last change on this file since 45 was 45, checked in by duerst@…, 9 years ago

adapted text from http://lists.w3.org/Archives/Public/public-iri/2010Mar/0000.html, with some tweaks

  • Property svn:executable set to *
File size: 124.6 KB
Line 
1<?xml version="1.0"?>
2<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
3<!ENTITY rfc1738 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.1738.xml">
4<!ENTITY rfc2045 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2045.xml">
5<!ENTITY rfc2119 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml">
6<!ENTITY rfc2130 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2130.xml">
7<!ENTITY rfc2141 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2141.xml">
8<!ENTITY rfc2192 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2192.xml">
9<!ENTITY rfc2277 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2277.xml">
10<!ENTITY rfc2368 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2368.xml">
11<!ENTITY rfc2384 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2384.xml">
12<!ENTITY rfc2396 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2396.xml">
13<!ENTITY rfc2397 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2397.xml">
14<!ENTITY rfc2616 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2616.xml">
15<!ENTITY rfc2640 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2640.xml">
16<!ENTITY rfc3490 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3490.xml">
17<!ENTITY rfc3491 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3491.xml">
18<!ENTITY rfc3629 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3629.xml">
19<!ENTITY rfc3986 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3986.xml">
20<!ENTITY rfc3987 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3987.xml">
21<!ENTITY rfc5890 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5890.xml">
22<!ENTITY rfc5891 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5891.xml">
23]>
24<?rfc strict='yes'?>
25
26<?xml-stylesheet type='text/css' href='rfc2629.css' ?>
27<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
28<?rfc symrefs='yes'?>
29<?rfc sortrefs='yes'?>
30<?rfc iprnotified="no" ?>
31<?rfc toc='yes'?>
32<?rfc compact='yes'?>
33<?rfc subcompact='no'?>
34<rfc ipr="pre5378Trust200902" docName="draft-ietf-iri-3987bis-05" category="std" xml:lang="en" obsoletes="3987">
35<front>
36<title abbrev="IRIs">Internationalized Resource Identifiers (IRIs)</title>
37
38  <author initials="M.J." surname="Duerst" fullname='Martin Duerst'>
39    <!-- (Note: Please write "Duerst" with u-umlaut wherever
40      possible, for example as "D&#252;rst" in XML and HTML) -->
41  <organization abbrev="Aoyama Gakuin University">Aoyama Gakuin University</organization>
42  <address>
43  <postal>
44  <street>5-10-1 Fuchinobe</street>
45  <city>Sagamihara</city>
46  <region>Kanagawa</region>
47  <code>229-8558</code>
48  <country>Japan</country>
49  </postal>
50  <phone>+81 42 759 6329</phone>
51  <facsimile>+81 42 759 6495</facsimile>
52  <email>duerst@it.aoyama.ac.jp</email>
53  <uri>http://www.sw.it.aoyama.ac.jp/D%C3%BCrst/<!-- (Note: This is the percent-encoded form of an IRI)--></uri>
54  </address>
55</author>
56
57<author initials="M.L." surname="Suignard" fullname="Michel Suignard">
58   <organization>Unicode Consortium</organization>
59   <address>
60   <postal>
61   <street></street>
62   <street>P.O. Box 391476</street>
63   <city>Mountain View</city>
64   <region>CA</region>
65   <code>94039-1476</code>
66   <country>U.S.A.</country>
67   </postal>
68   <phone>+1-650-693-3921</phone>
69   <email>michel@unicode.org</email>
70   <uri>http://www.suignard.com</uri>
71   </address>
72</author>
73<author initials="L." surname="Masinter" fullname="Larry Masinter">
74   <organization>Adobe</organization>
75   <address>
76   <postal>
77   <street>345 Park Ave</street>
78   <city>San Jose</city>
79   <region>CA</region>
80   <code>95110</code>
81   <country>U.S.A.</country>
82   </postal>
83   <phone>+1-408-536-3024</phone>
84   <email>masinter@adobe.com</email>
85   <uri>http://larry.masinter.net</uri>
86   </address>
87</author>
88
89<date year="2011" month="March" day="29"/>
90<area>Applications</area>
91<workgroup>Internationalized Resource Identifiers (iri)</workgroup>
92<keyword>IRI</keyword>
93<keyword>Internationalized Resource Identifier</keyword>
94<keyword>UTF-8</keyword>
95<keyword>URI</keyword>
96<keyword>URL</keyword>
97<keyword>IDN</keyword>
98<keyword>LEIRI</keyword>
99
100<abstract>
101<t>This document defines the Internationalized Resource Identifier
102(IRI) protocol element, as an extension of the Uniform Resource
103Identifier (URI).  An IRI is a sequence of characters from the
104Universal Character Set (Unicode/ISO 10646). Grammar and processing
105rules are given for IRIs and related syntactic forms.</t>
106
107<t>In addition, this document provides named additional rule sets
108for processing otherwise invalid IRIs, in a way that supports
109other specifications that wish to mandate common behavior for
110'error' handling. In particular, rules used in some XML languages
111(LEIRI) and web applications are given.</t>
112
113<t>Defining IRI as new protocol element (rather than updating or
114extending the definition of URI) allows independent orderly
115transitions: other protocols and languages that use URIs must
116explicitly choose to allow IRIs.</t>
117
118<t>Guidelines are provided for the use and deployment of IRIs and
119related protocol elements when revising protocols, formats, and
120software components that currently deal only with URIs.</t>
121
122</abstract>
123  <note title='RFC Editor: Please remove the next paragraph before publication.'>
124    <t>This document is intended to update RFC 3987 and move towards IETF
125    Draft Standard.  For discussion and comments on this
126    draft, please join the IETF IRI WG by subscribing to the mailing
127    list public-iri@w3.org. For a list of open issues, please see
128    the issue tracker of the WG at http://trac.tools.ietf.org/wg/iri/trac/report/1.
129    For a list of individual edits, please see the change history at
130    http://trac.tools.ietf.org/wg/iri/trac/log/draft-ietf-iri-3987bis.</t>
131</note>
132</front>
133<middle>
134
135<section title="Introduction">
136
137<section title="Overview and Motivation" anchor="overview">
138
139<t>A Uniform Resource Identifier (URI) is defined in <xref
140target="RFC3986"/> as a sequence of characters chosen from a limited
141subset of the repertoire of US-ASCII <xref target="ASCII"/>
142characters.</t>
143
144<t>The characters in URIs are frequently used for representing words
145of natural languages.  This usage has many advantages: Such URIs are
146easier to memorize, easier to interpret, easier to transcribe, easier
147to create, and easier to guess. For most languages other than English,
148however, the natural script uses characters other than A - Z. For many
149people, handling Latin characters is as difficult as handling the
150characters of other scripts is for those who use only the Latin
151alphabet. Many languages with non-Latin scripts are transcribed with
152Latin letters. These transcriptions are now often used in URIs, but
153they introduce additional difficulties.</t>
154
155<t>The infrastructure for the appropriate handling of characters from
156additional scripts is now widely deployed in operating system and
157application software. Software that can handle a wide variety of
158scripts and languages at the same time is increasingly common. Also,
159an increasing number of protocols and formats can carry a wide range of
160characters.</t>
161
162<t>URIs are used both as a protocol element (for transmission and
163processing by software) and also a presentation element (for display
164and handling by people who read, interpret, coin, or guess them). The
165transition between these roles is more difficult and complex when
166dealing with the larger set of characters than allowed for URIs in
167<xref target="RFC3986"/>. </t>
168
169<t>This document defines the protocol element called Internationalized
170Resource Identifier (IRI), which allow applications of URIs to be
171extended to use resource identifiers that have a much wider repertoire
172of characters. It also provides corresponding "internationalized"
173versions of other constructs from <xref target="RFC3986"/>, such as
174URI references. The syntax of IRIs is defined in <xref
175target="syntax"/>.
176</t>
177
178<t>Using characters outside of A - Z in IRIs adds a number of
179difficulties. <xref target="Bidi"/> discusses the special case of
180bidirectional IRIs using characters from scripts written
181right-to-left.  <xref target="equivalence"/> discusses various forms
182of equivalence between IRIs. <xref target="IRIuse"/> discusses the use
183of IRIs in different situations.  <xref target="guidelines"/> gives
184additional informative guidelines.  <xref target="security"/>
185discusses IRI-specific security considerations.</t>
186
187  <t>When originally defining IRIs, several design alternatives were considered.
188    Historically interested readers can find an overview in Appendix A of <xref target="RFC3987"/>.
189  For some additional background on the design of URIs and IRIs, please also see
190    <xref target="Gettys"/>.</t>
191</section> <!-- overview -->
192
193<section title="Applicability" anchor="Applicability">
194
195<t>IRIs are designed to allow protocols and software that deal with
196URIs to be updated to handle IRIs. A "URI scheme" (as defined by <xref
197target="RFC3986"/> and registered through the IANA process defined in
198<xref target="RFC4395bis"/> also serves as an "IRI scheme". Processing of
199IRIs is accomplished by extending the URI syntax while retaining (and
200not expanding) the set of "reserved" characters, such that the syntax
201for any URI scheme may be uniformly extended to allow non-ASCII
202characters. In addition, following parsing of an IRI, it is possible
203to construct a corresponding URI by first encoding characters outside
204of the allowed URI range and then reassembling the components.
205</t>
206
207<t>Practical use of IRIs forms in place of URIs forms depends on the
208following conditions being met:</t>
209
210<t><list style="hanging">
211   
212<t hangText="a.">A protocol or format element MUST be explicitly designated to be
213  able to carry IRIs. The intent is to avoid introducing IRIs into
214  contexts that are not defined to accept them.  For example, XML
215  schema <xref target="XMLSchema"/> has an explicit type "anyURI" that
216  includes IRIs and IRI references. Therefore, IRIs and IRI references
217  can be in attributes and elements of type "anyURI".  On the other
218  hand, in the <xref target="RFC2616"/> definition of HTTP/1.1, the
219  Request URI is defined as a URI, which means that direct use of IRIs
220  is not allowed in HTTP requests.</t>
221
222<t hangText="b.">The protocol or format carrying the IRIs MUST have a
223  mechanism to represent the wide range of characters used in IRIs,
224  either natively or by some protocol- or format-specific escaping
225  mechanism (for example, numeric character references in <xref
226  target="XML1"/>).</t>
227
228<t hangText="c.">The URI scheme definition, if it explicitly allows a
229  percent sign ("%") in any syntactic component, SHOULD define the
230  interpretation of sequences of percent-encoded octets (using "%XX"
231  hex octets) as octet from sequences of UTF-8 encoded strings; this
232  is recommended in the guidelines for registering new schemes, <xref
233  target="RFC4395bis"/>.  For example, this is the practice for IMAP URLs
234  <xref target="RFC2192"/>, POP URLs <xref target="RFC2384"/> and the
235  URN syntax <xref target="RFC2141"/>). Note that use of
236  percent-encoding may also be restricted in some situations, for
237  example, URI schemes that disallow percent-encoding might still be
238  used with a fragment identifier which is percent-encoded (e.g.,
239  <xref target="XPointer"/>). See <xref target="UTF8use"/> for further
240  discussion.</t>
241</list></t>
242
243</section> <!-- applicability -->
244
245<section title="Definitions" anchor="sec-Definitions">
246 
247<t>The following definitions are used in this document; they follow the
248terms in <xref target="RFC2130"/>, <xref target="RFC2277"/>, and
249<xref target="ISO10646"/>.</t>
250<t><list style="hanging">
251   
252<t hangText="character:">A member of a set of elements used for the
253    organization, control, or representation of data. For example,
254    "LATIN CAPITAL LETTER A" names a character.</t>
255   
256<t hangText="octet:">An ordered sequence of eight bits considered as a
257    unit.</t>
258   
259<t hangText="character repertoire:">A set of characters (set in the
260    mathematical sense).</t>
261   
262<t hangText="sequence of characters:">A sequence of characters (one
263    after another).</t>
264   
265<t hangText="sequence of octets:">A sequence of octets (one after
266    another).</t>
267   
268<t hangText="character encoding:">A method of representing a sequence
269    of characters as a sequence of octets (maybe with variants). Also,
270    a method of (unambiguously) converting a sequence of octets into a
271    sequence of characters.</t>
272   
273<t hangText="charset:">The name of a parameter or attribute used to
274    identify a character encoding.</t>
275   
276<t hangText="UCS:">Universal Character Set. The coded character set
277    defined by ISO/IEC 10646 <xref target="ISO10646"/> and the Unicode
278    Standard <xref target="UNIV6"/>.</t>
279   
280<t hangText="IRI reference:">Denotes the common usage of an
281    Internationalized Resource Identifier. An IRI reference may be
282    absolute or relative.  However, the "IRI" that results from such a
283    reference only includes absolute IRIs; any relative IRI references
284    are resolved to their absolute form.  Note that in <xref
285    target="RFC2396"/> URIs did not include fragment identifiers, but
286    in <xref target="RFC3986"/> fragment identifiers are part of
287    URIs.</t>
288   
289<t hangText="URL:">The term "URL" was originally used <xref
290   target="RFC1738"/> for roughly what is now called a "URI".  Books,
291   software and documentation often refers to URIs and IRIs using the
292   "URL" term. Some usages restrict "URL" to those URIs which are not
293   URNs. Because of the ambiguity of the term using the term "URL" is
294   NOT RECOMMENDED in formal documents.</t>
295
296<t hangText="LEIRI (Legacy Extended IRI) processing:">  This term was used in
297   various XML specifications to refer
298   to strings that, although not valid IRIs, were acceptable input to
299   the processing rules in <xref target="LEIRIspec" />.</t>
300
301<t hangText="(Web Address, Hypertext Reference, HREF):"> These terms have been
302   added in this document for convenience, to allow other
303   specifications to refer to those strings that, although not valid
304   IRIs, are acceptable input to the processing rules in <xref
305   target="webaddress"/>. This usage corresponds to the parsing rules
306   of some popular web browsing applications.
307   ISSUE: Need to find a good name/abbreviation for these.</t>
308   
309<t hangText="running text:">Human text (paragraphs, sentences,
310   phrases) with syntax according to orthographic conventions of a
311   natural language, as opposed to syntax defined for ease of
312   processing by machines (e.g., markup, programming languages).</t>
313   
314<t hangText="protocol element:">Any portion of a message that affects
315    processing of that message by the protocol in question.</t>
316   
317<t hangText="presentation element:">A presentation form corresponding
318    to a protocol element; for example, using a wider range of
319    characters.</t>
320   
321<t hangText="create (a URI or IRI):">With respect to URIs and IRIs,
322     the term is used for the initial creation. This may be the
323     initial creation of a resource with a certain identifier, or the
324     initial exposition of a resource under a particular
325     identifier.</t>
326   
327<t hangText="generate (a URI or IRI):">With respect to URIs and IRIs,
328     the term is used when the identifier is generated by derivation
329     from other information.</t>
330
331<t hangText="parsed URI component:">When a URI processor parses a URI
332   (following the generic syntax or a scheme-specific syntax, the result
333   is a set of parsed URI components, each of which has a type
334   (corresponding to the syntactic definition) and a sequence of URI
335   characters.  </t>
336
337<t hangText="parsed IRI component:">When an IRI processor parses
338   an IRI directly, following the general syntax or a scheme-specific
339   syntax, the result is a set of parsed IRI components, each of
340   which has a type (corresponding to the syntactice definition)
341   and a sequence of IRI characters. (This definition is analogous
342   to "parsed URI component".)</t>
343
344<t hangText="IRI scheme:">A URI scheme may also be known as
345   an "IRI scheme" if the scheme's syntax has been extended to
346   allow non-US-ASCII characters according to the rules in this
347   document.</t>
348
349</list></t>
350</section> <!-- definitions -->
351<section title="Notation" anchor="sec-Notation">
352     
353<t>RFCs and Internet Drafts currently do not allow any characters
354outside the US-ASCII repertoire. Therefore, this document uses various
355special notations to denote such characters in examples.</t>
356     
357<t>In text, characters outside US-ASCII are sometimes referenced by
358using a prefix of 'U+', followed by four to six hexadecimal
359digits.</t>
360
361<t>To represent characters outside US-ASCII in examples, this document
362uses two notations: 'XML Notation' and 'Bidi Notation'.</t>
363
364<t>XML Notation uses a leading '&amp;#x', a trailing ';', and the
365hexadecimal number of the character in the UCS in between. For
366example, &amp;#x44F; stands for CYRILLIC CAPITAL LETTER YA. In this
367notation, an actual '&amp;' is denoted by '&amp;amp;'.</t>
368
369<t>Bidi Notation is used for bidirectional examples: Lower case
370letters stand for Latin letters or other letters that are written left
371to right, whereas upper case letters represent Arabic or Hebrew
372letters that are written right to left.</t>
373
374<t>To denote actual octets in examples (as opposed to percent-encoded
375octets), the two hex digits denoting the octet are enclosed in "&lt;"
376and "&gt;".  For example, the octet often denoted as 0xc9 is denoted
377here as &lt;c9&gt;.</t>
378
379<t> In this document, the key words "MUST", "MUST NOT", "REQUIRED",
380"SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY",
381and "OPTIONAL" are to be interpreted as described in <xref
382target="RFC2119"/>.</t>
383
384</section> <!-- notation -->
385</section> <!-- introduction -->
386
387<section title="IRI Syntax" anchor="syntax">
388<t>This section defines the syntax of Internationalized Resource
389Identifiers (IRIs).</t>
390
391<t>As with URIs, an IRI is defined as a sequence of characters, not as
392a sequence of octets. This definition accommodates the fact that IRIs
393may be written on paper or read over the radio as well as stored or
394transmitted digitally.  The same IRI might be represented as different
395sequences of octets in different protocols or documents if these
396protocols or documents use different character encodings (and/or
397transfer encodings).  Using the same character encoding as the
398containing protocol or document ensures that the characters in the IRI
399can be handled (e.g., searched, converted, displayed) in the same way
400as the rest of the protocol or document.</t>
401
402<section title="Summary of IRI Syntax" anchor="summary">
403
404<t>IRIs are defined by extending the URI syntax in <xref
405target="RFC3986"/>, but extending the class of unreserved characters
406by adding the characters of the UCS (Universal Character Set, <xref
407target="ISO10646"/>) beyond U+007F, subject to the limitations given
408in the syntax rules below and in <xref target="limitations"/>.</t>
409
410<t>The syntax and use of components and reserved characters is the
411same as that in <xref target="RFC3986"/>. Each "URI scheme" thus also
412functions as an "IRI scheme", in that scheme-specific parsing rules
413for URIs of a scheme are be extended to allow parsing of IRIs using
414the same parsing rules.</t>
415
416<t>All the operations defined in <xref target="RFC3986"/>, such as the
417resolution of relative references, can be applied to IRIs by
418IRI-processing software in exactly the same way as they are for URIs
419by URI-processing software.</t>
420
421<t>Characters outside the US-ASCII repertoire MUST NOT be reserved and
422therefore MUST NOT be used for syntactical purposes, such as to
423delimit components in newly defined schemes. For example, U+00A2, CENT
424SIGN, is not allowed as a delimiter in IRIs, because it is in the
425'iunreserved' category. This is similar to the fact that it is not
426possible to use '-' as a delimiter in URIs, because it is in the
427'unreserved' category.</t>
428
429</section> <!-- summary -->
430<section title="ABNF for IRI References and IRIs" anchor="abnf">
431
432<t>An ABNF definition for IRI references (which are the most general
433concept and the start of the grammar) and IRIs is given here. The
434syntax of this ABNF is described in <xref target="STD68"/>. Character
435numbers are taken from the UCS, without implying any actual binary
436encoding. Terminals in the ABNF are characters, not octets.</t>
437
438<t>The following grammar closely follows the URI grammar in <xref
439target="RFC3986"/>, except that the range of unreserved characters is
440expanded to include UCS characters, with the restriction that private
441UCS characters can occur only in query parts. The grammar is split
442into two parts: Rules that differ from <xref target="RFC3986"/>
443because of the above-mentioned expansion, and rules that are the same
444as those in <xref target="RFC3986"/>. For rules that are different
445than those in <xref target="RFC3986"/>, the names of the non-terminals
446have been changed as follows. If the non-terminal contains 'URI', this
447has been changed to 'IRI'. Otherwise, an 'i' has been prefixed.
448The rule &lt;pct-form> has been introduced in order to be able to reference it from other parts of
449the document.</t>
450
451<!--
452for line length measuring in artwork (max 72 chars, three chars at start):
453      1         2         3         4         5         6         7
454456789012345678901234567890123456789012345678901234567890123456789012
455-->
456<figure>
457<preamble>The following rules are different from those in <xref target="RFC3986"/>:</preamble>
458<artwork>
459IRI            = scheme ":" ihier-part [ "?" iquery ]
460                 [ "#" ifragment ]
461
462ihier-part     = "//" iauthority ipath-abempty
463               / ipath-absolute
464               / ipath-rootless
465               / ipath-empty
466
467IRI-reference  = IRI / irelative-ref
468
469absolute-IRI   = scheme ":" ihier-part [ "?" iquery ]
470
471irelative-ref  = irelative-part [ "?" iquery ] [ "#" ifragment ]
472
473irelative-part = "//" iauthority ipath-abempty
474               / ipath-absolute
475               / ipath-noscheme
476               / ipath-empty
477
478iauthority     = [ iuserinfo "@" ] ihost [ ":" port ]
479iuserinfo      = *( iunreserved / pct-form / sub-delims / ":" )
480ihost          = IP-literal / IPv4address / ireg-name
481
482pct-form       = pct-encoded
483
484ireg-name      = *( iunreserved / sub-delims )
485
486ipath          = ipath-abempty   ; begins with "/" or is empty
487               / ipath-absolute  ; begins with "/" but not "//"
488               / ipath-noscheme  ; begins with a non-colon segment
489               / ipath-rootless  ; begins with a segment
490               / ipath-empty     ; zero characters
491
492ipath-abempty  = *( path-sep isegment )
493ipath-absolute = path-sep [ isegment-nz *( path-sep isegment ) ]
494ipath-noscheme = isegment-nz-nc *( path-sep isegment )
495ipath-rootless = isegment-nz *( path-sep isegment )
496ipath-empty    = 0&lt;ipchar&gt;
497path-sep       = "/"
498
499isegment       = *ipchar
500isegment-nz    = 1*ipchar
501isegment-nz-nc = 1*( iunreserved / pct-form / sub-delims
502                     / "@" )
503               ; non-zero-length segment without any colon ":"                     
504
505ipchar         = iunreserved / pct-form / sub-delims / ":"
506               / "@"
507 
508iquery         = *( ipchar / iprivate / "/" / "?" )
509
510ifragment      = *( ipchar / "/" / "?" / "#" )
511
512iunreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~" / ucschar
513
514ucschar        = %xA0-D7FF / %xF900-FDCF / %xFDF0-FFEF
515               / %x10000-1FFFD / %x20000-2FFFD / %x30000-3FFFD
516               / %x40000-4FFFD / %x50000-5FFFD / %x60000-6FFFD
517               / %x70000-7FFFD / %x80000-8FFFD / %x90000-9FFFD
518               / %xA0000-AFFFD / %xB0000-BFFFD / %xC0000-CFFFD
519               / %xD0000-DFFFD / %xE1000-EFFFD
520
521iprivate       = %xE000-F8FF / %xE0000-E0FFF / %xF0000-FFFFD
522               / %x100000-10FFFD
523</artwork>
524</figure>
525
526<t>Some productions are ambiguous. The "first-match-wins" (a.k.a. "greedy")
527algorithm applies. For details, see <xref target="RFC3986"/>.</t>
528
529<figure>
530<preamble>The following rules are the same as those in <xref target="RFC3986"/>:</preamble>
531<artwork>
532scheme         = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
533 
534port           = *DIGIT
535 
536IP-literal     = "[" ( IPv6address / IPvFuture  ) "]"
537 
538IPvFuture      = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" )
539 
540IPv6address    =                            6( h16 ":" ) ls32
541               /                       "::" 5( h16 ":" ) ls32
542               / [               h16 ] "::" 4( h16 ":" ) ls32
543               / [ *1( h16 ":" ) h16 ] "::" 3( h16 ":" ) ls32
544               / [ *2( h16 ":" ) h16 ] "::" 2( h16 ":" ) ls32
545               / [ *3( h16 ":" ) h16 ] "::"    h16 ":"   ls32
546               / [ *4( h16 ":" ) h16 ] "::"              ls32
547               / [ *5( h16 ":" ) h16 ] "::"              h16
548               / [ *6( h16 ":" ) h16 ] "::"
549               
550h16            = 1*4HEXDIG
551ls32           = ( h16 ":" h16 ) / IPv4address
552
553IPv4address    = dec-octet "." dec-octet "." dec-octet "." dec-octet
554
555dec-octet      = DIGIT                 ; 0-9
556               / %x31-39 DIGIT         ; 10-99
557               / "1" 2DIGIT            ; 100-199
558               / "2" %x30-34 DIGIT     ; 200-249
559               / "25" %x30-35          ; 250-255
560           
561pct-encoded    = "%" HEXDIG HEXDIG
562
563unreserved     = ALPHA / DIGIT / "-" / "." / "_" / "~"
564reserved       = gen-delims / sub-delims
565gen-delims     = ":" / "/" / "?" / "#" / "[" / "]" / "@"
566sub-delims     = "!" / "$" / "&amp;" / "'" / "(" / ")"
567               / "*" / "+" / "," / ";" / "="
568</artwork></figure>
569
570<t>This syntax does not support IPv6 scoped addressing zone identifiers.</t>
571
572</section> <!-- abnf -->
573
574</section> <!-- syntax -->
575
576<section title="Processing IRIs and related protocol elements" anchor="processing">
577
578<t>IRIs are meant to replace URIs in identifying resources within new
579versions of protocols, formats, and software components that use a
580UCS-based character repertoire.  Protocols and components may use and
581process IRIs directly. However, there are still numerous systems and
582protocols which only accept URIs or components of parsed URIs; that is,
583they only accept sequences of characters within the subset of US-ASCII
584characters allowed in URIs. </t>
585
586<t>This section defines specific processing steps for IRI consumers
587which establish the relationship between the string given and the
588interpreted derivatives. These
589processing steps apply to both IRIs and IRI references (i.e., absolute
590or relative forms); for IRIs, some steps are scheme specific. </t>
591
592<section title="Converting to UCS" anchor="ucsconv"> 
593 
594<t>Input that is already in a Unicode form (i.e., a sequence of Unicode
595 characters or an octet-stream representing a Unicode-based character
596 encoding such as UTF-8 or UTF-16) should be left as is and not
597 normalized (see (see <xref target="normalization"/>).</t>
598
599  <t>An IRI or IRI reference is a sequence of characters from the UCS.
600    For IRIs that are not already in a Unicode form
601    (as when written on paper, read aloud, or represented in a text stream
602    using a legacy character encoding), convert the IRI to Unicode.
603    Note that some character encodings or transcriptions can be converted
604    to or represented by more than one sequence of Unicode characters.
605    Ideally the resulting IRI would use a normalized form,
606    such as Unicode Normalization Form C <xref target="UTR15"/>
607    (see <xref target='ladder'/> Normalization and Comparison),
608    since that ensures a stable, consistent representation
609    that is most likely to produce the intended results.
610    Implementers and users are cautioned that, while denormalized character sequences are valid,
611    they might be difficult for other users or processes to reproduce
612    and might lead to unexpected results.
613  </t>
614
615<t> In other cases (written on paper, read aloud, or otherwise
616 represented independent of any character encoding) represent the IRI
617 as a sequence of characters from the UCS normalized according to
618 Unicode Normalization Form C (NFC, <xref target="UTR15"/>).</t>
619</section> <!-- ucsconv -->
620
621<section title="Parse the IRI into IRI components">
622
623<t>Parse the IRI, either as a relative reference (no scheme)
624or using scheme specific processing (according to the scheme
625given); the result resulting in a set of parsed IRI components.
626(NOTE: FIX BEFORE RELEASE: INTENT IS THAT ALL IRI SCHEMES
627THAT USE GENERIC SYNTAX AND ALLOW NON-ASCII AUTHORITY CAN
628ONLY USE AUTHORITY FOR NAMES THAT FOLLOW PUNICODE.)
629 </t>
630
631<t>NOTE: The result of parsing into components will correspond result
632in a correspondence of subtrings of the IRI according to the part
633matched.  For example, in <xref target="HTML5"/>, the protocol
634components of interest are SCHEME (scheme), HOST (ireg-name), PORT
635(port), the PATH (ipath after the initial "/"), QUERY (iquery),
636FRAGMENT (ifragment), and AUTHORITY (iauthority).
637</t>
638
639<t>Subsequent processing rules are sometimes used to define other
640syntactic components. For example, <xref target="HTML5"/> defines APIs
641for IRI processing; in these APIs:
642
643<list style="hanging">
644<t hangText="HOSTSPECIFIC"> the substring that follows
645the substring matched by the iauthority production, or the whole
646string if the iauthority production wasn't matched.</t>
647<t hangText="HOSTPORT"> if there is a scheme component and a port
648component and the port given by the port component is different than
649the default port defined for the protocol given by the scheme
650component, then HOSTPORT is the substring that starts with the
651substring matched by the host production and ends with the substring
652matched by the port production, and includes the colon in between the
653two. Otherwise, it is the same as the host component.
654</t>
655</list>
656</t>
657</section> <!-- parse -->
658
659<section title="General percent-encoding of IRI components" anchor="compmapping">
660   
661<t>For most IRI components, it is possible to map the IRI component
662to an equivalent URI component by percent-encoding those characters
663not allowed in URIs. Previous processing steps will have removed
664some characters, and the interpretation of reserved characters will
665have already been done (with the syntactic reserved characters outside
666of the IRI component). This mapping is defined for all sequences
667of Unicode characters, whether or not they are valid for the component
668in question. </t>
669   
670<t>For each character which is not allowed in a valid URI (NOTE: WHAT
671IS THE RIGHT REFERENCE HERE), apply the following steps. </t>
672
673<t><list style="hanging">
674
675<t hangText="Convert to UTF-8">Convert the character to a sequence of
676  one or more octets using UTF-8 <xref target="RFC3629"/>.</t>
677
678<t hangText="Percent encode">Convert each octet of this sequence to %HH,
679   where HH is the hexadecimal notation of the octet value. The
680   hexadecimal notation SHOULD use uppercase letters. (This is the
681   general URI percent-encoding mechanism in Section 2.1 of <xref
682   target="RFC3986"/>.)</t>
683   
684</list></t>
685
686<t>Note that the mapping is an identity transformation for parsed URI
687components of valid URIs, and is idempotent: applying the mapping a
688second time will not change anything.</t>
689</section> <!-- general conversion -->
690
691<section title="Mapping ireg-name" anchor="dnsmapping">
692
693<t>Schemes that allow non-ASCII based characters
694in the reg-name (ireg-name) position MUST convert the ireg-name
695component of an IRI as follows:</t>
696
697<t>Replace the ireg-name part of the IRI by the part converted using
698the ToASCII operation specified in Section 4.1 of <xref
699target="RFC3490"/> on each dot-separated label, and by using U+002E
700(FULL STOP) as a label separator, with the flag UseSTD3ASCIIRules set
701to FALSE, and with the flag AllowUnassigned set to FALSE.
702The ToASCII operation may
703fail, but this would mean that the IRI cannot be resolved.
704In such cases, if the domain name conversion fails, then the
705entire IRI conversion fails. Processors that have no mechanism for
706signalling a failure MAY instead substitute an otherwise
707invalid host name, although such processing SHOULD be avoided.
708 </t>
709
710<t>For example, the IRI
711<vspace/>"http://r&amp;#xE9;sum&amp;#xE9;.example.org"<vspace/> MAY be
712converted to <vspace/>"http://xn--rsum-bad.example.org"<vspace/>;
713conversion to percent-encoded form, e.g.,
714 <vspace/>"http://r%C3%A9sum%C3%A9.example.org", MUST NOT be performed. </t>
715
716<t><list style="hanging"> 
717
718<t hangText="Note:">Domain Names may appear in parts of an IRI other
719than the ireg-name part.  It is the responsibility of scheme-specific
720implementations (if the Internationalized Domain Name is part of the
721scheme syntax) or of server-side implementations (if the
722Internationalized Domain Name is part of 'iquery') to apply the
723necessary conversions at the appropriate point. Example: Trying to
724validate the Web page at<vspace/>
725http://r&amp;#xE9;sum&amp;#xE9;.example.org would lead to an IRI of
726<vspace/>http://validator.w3.org/check?uri=http%3A%2F%2Fr&amp;#xE9;sum&amp;#xE9;.<vspace/>example.org,
727which would convert to a URI
728of<vspace/>http://validator.w3.org/check?uri=http%3A%2F%2Fr%C3%A9sum%C3%A9.<vspace/>example.org.
729The server-side implementation is responsible for making the
730necessary conversions to be able to retrieve the Web page.</t>
731
732<t hangText="Note:">In this process, characters allowed in URI
733references and existing percent-encoded sequences are not encoded further.
734(This mapping is similar to, but different from, the encoding applied
735when arbitrary content is included in some part of a URI.)
736
737For example, an IRI of
738<vspace/>"http://www.example.org/red%09ros&amp;#xE9;#red"
739(in XML notation) is converted to
740<vspace/>"http://www.example.org/red%09ros%C3%A9#red", not to
741something like
742<vspace/>"http%3A%2F%2Fwww.example.org%2Fred%2509ros%C3%A9%23red".
743((DESIGN QUESTION: What about e.g. http://r%C3%A9sum%C3%A9.example.org in an IRI? Will that get converted to punycode, or not?))
744
745</t>
746
747</list></t>
748</section> <!-- dnsmapping -->
749
750<section title="Mapping query components" anchor="querymapping">
751
752<t>((NOTE: SEE ISSUES LIST))
753
754For compatibility with existing deployed HTTP infrastructure,
755the following special case applies for schemes "http" and "https"
756and IRIs whose origin has a document charset other than one which
757is UCS-based (e.g., UTF-8 or UTF-16). In such a case, the "query"
758component of an IRI is mapped into a URI by using the document
759charset rather than UTF-8 as the binary representation before
760pct-encoding. This mapping is not applied for any other scheme
761or component.</t>
762
763</section> <!-- querymapping -->
764
765<section title="Mapping IRIs to URIs" anchor="mapping">
766
767<t>The canonical mapping from a IRI to URI is defined by applying the
768mapping above (from IRI to URI components) and then reassembling a URI
769from the parsed URI components using the original punctuation that
770delimited the IRI components. </t>
771
772</section> <!-- mapping -->
773
774<section title="Converting URIs to IRIs" anchor="URItoIRI">
775
776<t>In some situations, for presentation and further processing,
777it is desirable to convert a URI into an equivalent IRI in which
778natural characters are represented directly rather than
779percent encoded. Of course, every URI is already an IRI in
780its own right without any conversion, and in general there
781This section gives one such procedure for this conversion.
782</t>
783
784<t>
785The conversion described in this section, if given a valid URI, will
786result in an IRI that maps back to the URI used as an input for the
787conversion (except for potential case differences in percent-encoding
788and for potential percent-encoded unreserved characters).
789
790However, the IRI resulting from this conversion may differ
791from the original IRI (if there ever was one).</t> 
792
793<t>URI-to-IRI conversion removes percent-encodings, but not all
794percent-encodings can be eliminated. There are several reasons for
795this:</t>
796
797<t><list style="hanging">
798
799<t hangText="1.">Some percent-encodings are necessary to distinguish
800    percent-encoded and unencoded uses of reserved characters.</t>
801
802<t hangText="2.">Some percent-encodings cannot be interpreted as sequences
803    of UTF-8 octets.<vspace blankLines="1"/>
804    (Note: The octet patterns of UTF-8 are highly regular.
805    Therefore, there is a very high probability, but no guarantee,
806    that percent-encodings that can be interpreted as sequences of UTF-8
807    octets actually originated from UTF-8. For a detailed discussion,
808    see <xref target="Duerst97"/>.)</t>
809
810<t hangText="3.">The conversion may result in a character that is not
811    appropriate in an IRI. See <xref target="abnf"/>, <xref target="visual"/>,
812      and <xref target="limitations"/> for further details.</t>
813
814<t hangText="4.">IRI to URI conversion has different rules for
815    dealing with domain names and query parameters.</t>
816
817</list></t>
818
819<t>Conversion from a URI to an IRI MAY be done by using the following
820steps:
821
822<list style="hanging">
823<t hangText="1.">Represent the URI as a sequence of octets in
824       US-ASCII.</t>
825
826<t hangText="2.">Convert all percent-encodings ("%" followed by two
827      hexadecimal digits) to the corresponding octets, except those
828      corresponding to "%", characters in "reserved", and characters
829      in US-ASCII not allowed in URIs.</t> 
830
831<t hangText="3.">Re-percent-encode any octet produced in step 2 that
832      is not part of a strictly legal UTF-8 octet sequence.</t>
833
834
835<t hangText="4.">Re-percent-encode all octets produced in step 3 that
836      in UTF-8 represent characters that are not appropriate according
837      to <xref target="abnf"/>, <xref target="visual"/>, and <xref
838      target="limitations"/>.</t> 
839
840<t hangText="5.">Interpret the resulting octet sequence as a sequence
841      of characters encoded in UTF-8.</t>
842
843<t hangText="6.">URIs known to contain domain names in the reg-name
844      component SHOULD convert punycode-encoded domain name labels to
845      the corresponding characters using the ToUnicode procedure. </t>
846</list></t>
847
848<t>This procedure will convert as many percent-encoded characters as
849possible to characters in an IRI. Because there are some choices when
850step 4 is applied (see <xref target="limitations"/>), results may
851vary.</t>
852
853<t>Conversions from URIs to IRIs MUST NOT use any character
854encoding other than UTF-8 in steps 3 and 4, even if it might be
855possible to guess from the context that another character encoding
856than UTF-8 was used in the URI.  For example, the URI
857"http://www.example.org/r%E9sum%E9.html" might with some guessing be
858interpreted to contain two e-acute characters encoded as
859iso-8859-1. It must not be converted to an IRI containing these
860e-acute characters. Otherwise, in the future the IRI will be mapped to
861"http://www.example.org/r%C3%A9sum%C3%A9.html", which is a different
862URI from "http://www.example.org/r%E9sum%E9.html".</t>
863
864<section title="Examples">
865
866<t>This section shows various examples of converting URIs to IRIs.
867Each example shows the result after each of the steps 1 through 6 is
868applied. XML Notation is used for the final result.  Octets are
869denoted by "&lt;" followed by two hexadecimal digits followed by
870"&gt;".</t>
871
872<t>The following example contains the sequence "%C3%BC", which is a
873strictly legal UTF-8 sequence, and which is converted into the actual
874character U+00FC, LATIN SMALL LETTER U WITH DIAERESIS (also known as
875u-umlaut).
876
877<list style="hanging">
878<t hangText="1.">http://www.example.org/D%C3%BCrst</t>
879<t hangText="2.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
880<t hangText="3.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
881<t hangText="4.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
882<t hangText="5.">http://www.example.org/D&amp;#xFC;rst</t>
883<t hangText="6.">http://www.example.org/D&amp;#xFC;rst</t>
884</list>
885</t>
886
887<t>The following example contains the sequence "%FC", which might
888represent U+00FC, LATIN SMALL LETTER U WITH DIAERESIS, in
889the<vspace/>iso-8859-1 character encoding.  (It might represent other
890characters in other character encodings. For example, the octet
891&lt;fc&gt; in iso-8859-5 represents U+045C, CYRILLIC SMALL LETTER
892KJE.)  Because &lt;fc&gt; is not part of a strictly legal UTF-8
893sequence, it is re-percent-encoded in step 3.
894
895
896<list style="hanging">
897<t hangText="1.">http://www.example.org/D%FCrst</t>
898<t hangText="2.">http://www.example.org/D&lt;fc&gt;rst</t>
899<t hangText="3.">http://www.example.org/D%FCrst</t>
900<t hangText="4.">http://www.example.org/D%FCrst</t>
901<t hangText="5.">http://www.example.org/D%FCrst</t>
902<t hangText="6.">http://www.example.org/D%FCrst</t>
903</list>
904</t>
905
906<t>The following example contains "%e2%80%ae", which is the percent-encoded<vspace/>UTF-8
907character encoding of U+202E, RIGHT-TO-LEFT OVERRIDE. <xref target="visual"/>
908forbids the direct use of this character in an IRI. Therefore, the
909corresponding octets are re-percent-encoded in step 4. This example shows
910that the case (upper- or lowercase) of letters used in percent-encodings may not be preserved.
911The example also contains a punycode-encoded domain name label (xn--99zt52a),
912which is not converted.
913
914<list style="hanging">
915<t hangText="1.">http://xn--99zt52a.example.org/%e2%80%ae</t>
916<t hangText="2.">http://xn--99zt52a.example.org/&lt;e2&gt;&lt;80&gt;&lt;ae&gt;</t>
917<t hangText="3.">http://xn--99zt52a.example.org/&lt;e2&gt;&lt;80&gt;&lt;ae&gt;</t>
918<t hangText="4.">http://xn--99zt52a.example.org/%E2%80%AE</t>
919<t hangText="5.">http://xn--99zt52a.example.org/%E2%80%AE</t>
920<t hangText="6.">http://&amp;#x7D0D;&amp;#x8C46;.example.org/%E2%80%AE</t>
921</list></t>
922
923<t>Note that the label "xn--99zt52a" is converted to U+7D0D U+8C46
924(Japanese Natto). ((EDITOR NOTE: There is some inconsistency in this note.))</t>
925
926</section> <!-- examples -->
927</section> <!-- URItoIRI -->
928</section> <!-- processing -->
929<section title="Bidirectional IRIs for Right-to-Left Languages" anchor="Bidi">
930
931<t>Some UCS characters, such as those used in the Arabic and Hebrew
932scripts, have an inherent right-to-left (rtl) writing direction. IRIs
933containing these characters (called bidirectional IRIs or Bidi IRIs)
934require additional attention because of the non-trivial relation
935between logical representation (used for digital representation and
936for reading/spelling) and visual representation (used for
937display/printing).</t>
938
939<t>Because of the complex interaction between the logical representation,
940the visual representation, and the syntax of a Bidi IRI, a balance is
941needed between various requirements.
942The main requirements are<list style="hanging">
943<t hangText="1.">user-predictable conversion between visual and
944    logical representation;</t>
945<t hangText="2.">the ability to include a wide range of characters
946    in various parts of the IRI; and</t>
947<t hangText="3.">minor or no changes or restrictions for
948      implementations.</t>
949</list></t>
950
951<section title="Logical Storage and Visual Presentation" anchor="visual">
952
953<t>When stored or transmitted in digital representation, bidirectional
954IRIs MUST be in full logical order and MUST conform to the IRI syntax
955rules (which includes the rules relevant to their scheme). This
956ensures that bidirectional IRIs can be processed in the same way as
957other IRIs.</t> <t>Bidirectional IRIs MUST be rendered by using the
958Unicode Bidirectional Algorithm <xref target="UNIV6"/>, <xref
959target="UNI9"/>.  Bidirectional IRIs MUST be rendered in the same way
960as they would be if they were in a left-to-right embedding; i.e., as
961if they were preceded by U+202A, LEFT-TO-RIGHT EMBEDDING (LRE), and
962followed by U+202C, POP DIRECTIONAL FORMATTING (PDF).  Setting the
963embedding direction can also be done in a higher-level protocol (e.g.,
964the dir='ltr' attribute in HTML).</t> 
965
966<t>There is no requirement to use the above embedding if the display
967is still the same without the embedding. For example, a bidirectional
968IRI in a text with left-to-right base directionality (such as used for
969English or Cyrillic) that is preceded and followed by whitespace and
970strong left-to-right characters does not need an embedding.  Also, a
971bidirectional relative IRI reference that only contains strong
972right-to-left characters and weak characters and that starts and ends
973with a strong right-to-left character and appears in a text with
974right-to-left base directionality (such as used for Arabic or Hebrew)
975and is preceded and followed by whitespace and strong characters does
976not need an embedding.</t>
977
978<t>In some other cases, using U+200E, LEFT-TO-RIGHT MARK (LRM), may be
979sufficient to force the correct display behavior.  However, the
980details of the Unicode Bidirectional algorithm are not always easy to
981understand. Implementers are strongly advised to err on the side of
982caution and to use embedding in all cases where they are not
983completely sure that the display behavior is unaffected without the
984embedding.</t>
985
986<t>The Unicode Bidirectional Algorithm (<xref target="UNI9"/>, section
9874.3) permits higher-level protocols to influence bidirectional
988rendering. Such changes by higher-level protocols MUST NOT be used if
989they change the rendering of IRIs.</t> 
990
991<t>The bidirectional formatting characters that may be used before or
992after the IRI to ensure correct display are not themselves part of the
993IRI.  IRIs MUST NOT contain bidirectional formatting characters (LRM,
994RLM, LRE, RLE, LRO, RLO, and PDF). They affect the visual rendering of
995the IRI but do not appear themselves. It would therefore not be
996possible to input an IRI with such characters correctly.</t>
997
998</section> <!-- visual -->
999<section title="Bidi IRI Structure" anchor="bidi-structure">
1000
1001<t>The Unicode Bidirectional Algorithm is designed mainly for running
1002text.  To make sure that it does not affect the rendering of
1003bidirectional IRIs too much, some restrictions on bidirectional IRIs
1004are necessary. These restrictions are given in terms of delimiters
1005(structural characters, mostly punctuation such as "@", ".", ":",
1006and<vspace/>"/") and components (usually consisting mostly of letters
1007and digits).</t>
1008
1009<t>The following syntax rules from <xref target="abnf"/> correspond to
1010components for the purpose of Bidi behavior: iuserinfo, ireg-name,
1011isegment, isegment-nz, isegment-nz-nc, ireg-name, iquery, and
1012ifragment.</t>
1013
1014<t>Specifications that define the syntax of any of the above
1015components MAY divide them further and define smaller parts to be
1016components according to this document. As an example, the restrictions
1017of <xref target="RFC3490"/> on bidirectional domain names correspond
1018to treating each label of a domain name as a component for schemes
1019with ireg-name as a domain name.  Even where the components are not
1020defined formally, it may be helpful to think about some syntax in
1021terms of components and to apply the relevant restrictions.  For
1022example, for the usual name/value syntax in query parts, it is
1023convenient to treat each name and each value as a component. As
1024another example, the extensions in a resource name can be treated as
1025separate components.</t>
1026
1027<t>For each component, the following restrictions apply:</t>
1028<t>
1029<list style="hanging">
1030
1031<t hangText="1.">A component SHOULD NOT use both right-to-left and
1032  left-to-right characters.</t>
1033
1034<t hangText="2.">A component using right-to-left characters SHOULD
1035  start and end with right-to-left characters.</t>
1036
1037</list></t>
1038
1039<t>The above restrictions are given as "SHOULD"s, rather than as
1040"MUST"s.  For IRIs that are never presented visually, they are not
1041relevant.  However, for IRIs in general, they are very important to
1042ensure consistent conversion between visual presentation and logical
1043representation, in both directions.</t>
1044
1045<t><list style="hanging">
1046
1047<t hangText="Note:">In some components, the above restrictions may
1048  actually be strictly enforced.  For example, <xref
1049  target="RFC3490"></xref> requires that these restrictions apply to
1050  the labels of a host name for those schemes where ireg-name is a
1051  host name.  In some other components (for example, path components)
1052  following these restrictions may not be too difficult.  For other
1053  components, such as parts of the query part, it may be very
1054  difficult to enforce the restrictions because the values of query
1055  parameters may be arbitrary character sequences.</t>
1056
1057</list></t>
1058
1059<t>If the above restrictions cannot be satisfied otherwise, the
1060affected component can always be mapped to URI notation as described
1061in <xref target="compmapping"/>. Please note that the whole component
1062has to be mapped (see also Example 9 below).</t>
1063
1064</section> <!-- bidi-structure -->
1065
1066<section title="Input of Bidi IRIs" anchor="bidiInput">
1067
1068<t>Bidi input methods MUST generate Bidi IRIs in logical order while
1069rendering them according to <xref target="visual"/>.  During input,
1070rendering SHOULD be updated after every new character is input to
1071avoid end-user confusion.</t>
1072
1073</section> <!-- bidiInput -->
1074
1075<section title="Examples">
1076
1077<t>This section gives examples of bidirectional IRIs, in Bidi
1078Notation.  It shows legal IRIs with the relationship between logical
1079and visual representation and explains how certain phenomena in this
1080relationship may look strange to somebody not familiar with
1081bidirectional behavior, but familiar to users of Arabic and Hebrew. It
1082also shows what happens if the restrictions given in <xref
1083target="bidi-structure"/> are not followed. The examples below can be
1084seen at <xref target="BidiEx"/>, in Arabic, Hebrew, and Bidi Notation
1085variants.</t>
1086
1087<t>To read the bidi text in the examples, read the visual
1088representation from left to right until you encounter a block of rtl
1089text. Read the rtl block (including slashes and other special
1090characters) from right to left, then continue at the next unread ltr
1091character.</t>
1092
1093<t>Example 1: A single component with rtl characters is inverted:
1094<vspace/>Logical representation:
1095"http://ab.CDEFGH.ij/kl/mn/op.html"<vspace/>Visual representation:
1096"http://ab.HGFEDC.ij/kl/mn/op.html"<vspace/> Components can be read
1097one by one, and each component can be read in its natural
1098direction.</t>
1099
1100<t>Example 2: More than one consecutive component with rtl characters
1101is inverted as a whole: <vspace/>Logical representation:
1102"http://ab.CDE.FGH/ij/kl/mn/op.html"<vspace/>Visual representation:
1103"http://ab.HGF.EDC/ij/kl/mn/op.html"<vspace/> A sequence of rtl
1104components is read rtl, in the same way as a sequence of rtl words is
1105read rtl in a bidi text.</t>
1106
1107<t>Example 3: All components of an IRI (except for the scheme) are
1108rtl.  All rtl components are inverted overall: <vspace/>Logical
1109representation:
1110"http://AB.CD.EF/GH/IJ/KL?MN=OP;QR=ST#UV"<vspace/>Visual
1111representation: "http://VU#TS=RQ;PO=NM?LK/JI/HG/FE.DC.BA"<vspace/> The
1112whole IRI (except the scheme) is read rtl. Delimiters between rtl
1113components stay between the respective components; delimiters between
1114ltr and rtl components don't move.</t>
1115
1116<t>Example 4: Each of several sequences of rtl components is inverted
1117on its own: <vspace/>Logical representation:
1118"http://AB.CD.ef/gh/IJ/KL.html"<vspace/>Visual representation:
1119"http://DC.BA.ef/gh/LK/JI.html"<vspace/> Each sequence of rtl
1120components is read rtl, in the same way as each sequence of rtl words
1121in an ltr text is read rtl.</t>
1122
1123<t>Example 5: Example 2, applied to components of different kinds:
1124<vspace/>Logical representation: "http://ab.cd.EF/GH/ij/kl.html"
1125<vspace/>Visual representation:
1126"http://ab.cd.HG/FE/ij/kl.html"<vspace/> The inversion of the domain
1127name label and the path component may be unexpected, but it is
1128consistent with other bidi behavior.  For reassurance that the domain
1129component really is "ab.cd.EF", it may be helpful to read aloud the
1130visual representation following the bidi algorithm. After
1131"http://ab.cd." one reads the RTL block "E-F-slash-G-H", which
1132corresponds to the logical representation.
1133</t>
1134
1135<t>Example 6: Same as Example 5, with more rtl components:
1136<vspace/>Logical representation:
1137"http://ab.CD.EF/GH/IJ/kl.html"<vspace/>Visual representation:
1138"http://ab.JI/HG/FE.DC/kl.html"<vspace/> The inversion of the domain
1139name labels and the path components may be easier to identify because
1140the delimiters also move.</t>
1141
1142<t>Example 7: A single rtl component includes digits: <vspace/>Logical
1143representation: "http://ab.CDE123FGH.ij/kl/mn/op.html"<vspace/>Visual
1144representation: "http://ab.HGF123EDC.ij/kl/mn/op.html"<vspace/>
1145Numbers are written ltr in all cases but are treated as an additional
1146embedding inside a run of rtl characters. This is completely
1147consistent with usual bidirectional text.</t>
1148
1149<t>Example 8 (not allowed): Numbers are at the start or end of an rtl
1150component:<vspace/>Logical representation:
1151"http://ab.cd.ef/GH1/2IJ/KL.html"<vspace/>Visual representation:
1152"http://ab.cd.ef/LK/JI1/2HG.html"<vspace/> The sequence "1/2" is
1153interpreted by the bidi algorithm as a fraction, fragmenting the
1154components and leading to confusion. There are other characters that
1155are interpreted in a special way close to numbers; in particular, "+",
1156"-", "#", "$", "%", ",", ".", and ":".</t>
1157
1158<t>Example 9 (not allowed): The numbers in the previous example are
1159percent-encoded: <vspace/>Logical representation:
1160"http://ab.cd.ef/GH%31/%32IJ/KL.html",<vspace/>Visual representation:
1161"http://ab.cd.ef/LK/JI%32/%31HG.html"</t>
1162
1163<t>Example 10 (allowed but not recommended): <vspace/>Logical
1164representation: "http://ab.CDEFGH.123/kl/mn/op.html"<vspace/>Visual
1165representation: "http://ab.123.HGFEDC/kl/mn/op.html"<vspace/>
1166Components consisting of only numbers are allowed (it would be rather
1167difficult to prohibit them), but these may interact with adjacent RTL
1168components in ways that are not easy to predict.</t>
1169
1170<t>Example 11 (allowed but not recommended): <vspace/>Logical
1171representation: "http://ab.CDEFGH.123ij/kl/mn/op.html"<vspace/>Visual
1172representation: "http://ab.123.HGFEDCij/kl/mn/op.html"<vspace/>
1173Components consisting of numbers and left-to-right characters are
1174allowed, but these may interact with adjacent RTL components in ways
1175that are not easy to predict.</t>
1176</section><!-- examples -->
1177</section><!-- bidi -->
1178
1179<section title="Normalization and Comparison" anchor="equivalence">
1180
1181<t><list style="hanging"><t hangText="Note:">The structure and much of
1182  the material for this section is taken from section 6 of <xref
1183  target="RFC3986"></xref>; the differences are due to the specifics
1184  of IRIs.</t></list></t>
1185
1186<t>One of the most common operations on IRIs is simple comparison:
1187Determining whether two IRIs are equivalent, without using the IRIs to
1188access their respective resource(s). A comparison is performed
1189whenever a response cache is accessed, a browser checks its history to
1190color a link, or an XML parser processes tags within a
1191namespace. Extensive normalization prior to comparison of IRIs may be
1192used by spiders and indexing engines to prune a search space or reduce
1193duplication of request actions and response storage.</t>
1194
1195<t>IRI comparison is performed for some particular purpose. Protocols
1196or implementations that compare IRIs for different purposes will often
1197be subject to differing design trade-offs in regards to how much
1198effort should be spent in reducing aliased identifiers. This section
1199describes various methods that may be used to compare IRIs, the
1200trade-offs between them, and the types of applications that might use
1201them.</t>
1202
1203<section title="Equivalence">
1204
1205<t>Because IRIs exist to identify resources, presumably they should be
1206considered equivalent when they identify the same resource. However,
1207this definition of equivalence is not of much practical use, as there
1208is no way for an implementation to compare two resources to determine
1209if they are "the same" unless it has full knowledge or control of
1210them. For this reason, determination of equivalence or difference of
1211IRIs is based on string comparison, perhaps augmented by reference to
1212additional rules provided by URI scheme definitions.  We use the terms
1213"different" and "equivalent" to describe the possible outcomes of such
1214comparisons, but there are many application-dependent versions of
1215equivalence.</t>
1216
1217<t>Even when it is possible to determine that two IRIs are equivalent,
1218IRI comparison is not sufficient to determine whether two IRIs
1219identify different resources. For example, an owner of two different
1220domain names could decide to serve the same resource from both,
1221resulting in two different IRIs. Therefore, comparison methods are
1222designed to minimize false negatives while strictly avoiding false
1223positives.</t>
1224
1225<t>In testing for equivalence, applications should not directly
1226compare relative references; the references should be converted to
1227their respective target IRIs before comparison. When IRIs are compared
1228to select (or avoid) a network action, such as retrieval of a
1229representation, fragment components (if any) should be excluded from
1230the comparison.</t>
1231
1232<t>Applications using IRIs as identity tokens with no relationship to
1233a protocol MUST use the Simple String Comparison (see <xref
1234target="stringcomp"></xref>).  All other applications MUST select one
1235of the comparison practices from the Comparison Ladder (see <xref
1236target="ladder"></xref>.</t>
1237</section> <!-- equivalence -->
1238
1239
1240<section title="Preparation for Comparison">
1241<t>Any kind of IRI comparison REQUIRES that any additional contextual
1242processing is first performed, including undoing higher-level
1243escapings or encodings in the protocol or format that carries an
1244IRI. This preprocessing is usually done when the protocol or format is
1245parsed.</t>
1246
1247<t>Examples of contextual preprocessing steps are described in <xref
1248target="LEIRIHREF"/>. </t>
1249
1250<t>Examples of such escapings or encodings are entities and
1251numeric character references in <xref target="HTML4"></xref> and <xref
1252target="XML1"></xref>. As an example,
1253"http://example.org/ros&amp;eacute;" (in HTML),
1254"http://example.org/ros&amp;#233;" (in HTML or XML), and
1255<vspace/>"http://example.org/ros&amp;#xE9;" (in HTML or XML) are all
1256resolved into what is denoted in this document (see <xref
1257target="sec-Notation"></xref>) as "http://example.org/ros&amp;#xE9;"
1258(the "&amp;#xE9;" here standing for the actual e-acute character, to
1259compensate for the fact that this document cannot contain non-ASCII
1260characters).</t>
1261
1262<t>Similar considerations apply to encodings such as Transfer Codings
1263in HTTP (see <xref target="RFC2616"></xref>) and Content Transfer
1264Encodings in MIME (<xref target="RFC2045"></xref>), although in these
1265cases, the encoding is based not on characters but on octets, and
1266additional care is required to make sure that characters, and not just
1267arbitrary octets, are compared (see <xref
1268target="stringcomp"></xref>).</t>
1269
1270</section> <!-- preparation -->
1271
1272<section title="Comparison Ladder" anchor="ladder">
1273
1274<t>In practice, a variety of methods are used to test IRI
1275equivalence. These methods fall into a range distinguished by the
1276amount of processing required and the degree to which the probability
1277of false negatives is reduced. As noted above, false negatives cannot
1278be eliminated. In practice, their probability can be reduced, but this
1279reduction requires more processing and is not cost-effective for all
1280applications.</t>
1281
1282
1283<t>If this range of comparison practices is considered as a ladder,
1284the following discussion will climb the ladder, starting with
1285practices that are cheap but have a relatively higher chance of
1286producing false negatives, and proceeding to those that have higher
1287computational cost and lower risk of false negatives.</t>
1288
1289<section title="Simple String Comparison" anchor="stringcomp">
1290
1291<t>If two IRIs, when considered as character strings, are identical,
1292then it is safe to conclude that they are equivalent.  This type of
1293equivalence test has very low computational cost and is in wide use in
1294a variety of applications, particularly in the domain of parsing. It
1295is also used when a definitive answer to the question of IRI
1296equivalence is needed that is independent of the scheme used and that
1297can be calculated quickly and without accessing a network. An example
1298of such a case is XML Namespaces (<xref
1299target="XMLNamespace"></xref>).</t>
1300
1301
1302<t>Testing strings for equivalence requires some basic precautions.
1303This procedure is often referred to as "bit-for-bit" or
1304"byte-for-byte" comparison, which is potentially misleading. Testing
1305strings for equality is normally based on pair comparison of the
1306characters that make up the strings, starting from the first and
1307proceeding until both strings are exhausted and all characters are
1308found to be equal, until a pair of characters compares unequal, or
1309until one of the strings is exhausted before the other.</t>
1310
1311<t>This character comparison requires that each pair of characters be
1312put in comparable encoding form. For example, should one IRI be stored
1313in a byte array in UTF-8 encoding form and the second in a UTF-16
1314encoding form, bit-for-bit comparisons applied naively will produce
1315errors. It is better to speak of equality on a character-for-character
1316rather than on a byte-for-byte or bit-for-bit basis.  In practical
1317terms, character-by-character comparisons should be done codepoint by
1318codepoint after conversion to a common character encoding form.
1319
1320When comparing character by character, the comparison function MUST
1321NOT map IRIs to URIs, because such a mapping would create additional
1322spurious equivalences. It follows that an IRI SHOULD NOT be modified
1323when being transported if there is any chance that this IRI might be
1324used in a context that uses Simple String Comparison.</t>
1325
1326
1327<t>False negatives are caused by the production and use of IRI
1328aliases. Unnecessary aliases can be reduced, regardless of the
1329comparison method, by consistently providing IRI references in an
1330already normalized form (i.e., a form identical to what would be
1331produced after normalization is applied, as described below).
1332Protocols and data formats often limit some IRI comparisons to simple
1333string comparison, based on the theory that people and implementations
1334will, in their own best interest, be consistent in providing IRI
1335references, or at least be consistent enough to negate any efficiency
1336that might be obtained from further normalization.</t>
1337</section> <!-- stringcomp -->
1338
1339<section title="Syntax-Based Normalization">
1340
1341<figure><preamble>Implementations may use logic based on the
1342definitions provided by this specification to reduce the probability
1343of false negatives. This processing is moderately higher in cost than
1344character-for-character string comparison. For example, an application
1345using this approach could reasonably consider the following two IRIs
1346equivalent:</preamble>
1347
1348<artwork>
1349   example://a/b/c/%7Bfoo%7D/ros&amp;#xE9;
1350   eXAMPLE://a/./b/../b/%63/%7bfoo%7d/ros%C3%A9
1351</artwork></figure>
1352
1353<t>Web user agents, such as browsers, typically apply this type of IRI
1354normalization when determining whether a cached response is
1355available. Syntax-based normalization includes such techniques as case
1356normalization, character normalization, percent-encoding
1357normalization, and removal of dot-segments.</t>
1358
1359<section title="Case Normalization">
1360
1361<t>For all IRIs, the hexadecimal digits within a percent-encoding
1362triplet (e.g., "%3a" versus "%3A") are case-insensitive and therefore
1363should be normalized to use uppercase letters for the digits A-F.</t>
1364
1365<t>When an IRI uses components of the generic syntax, the component
1366syntax equivalence rules always apply; namely, that the scheme and
1367US-ASCII only host are case insensitive and therefore should be
1368normalized to lowercase. For example, the URI
1369"HTTP://www.EXAMPLE.com/" is equivalent to
1370"http://www.example.com/". Case equivalence for non-ASCII characters
1371in IRI components that are IDNs are discussed in <xref
1372target="schemecomp"></xref>.  The other generic syntax components are
1373assumed to be case sensitive unless specifically defined otherwise by
1374the scheme.</t>
1375
1376<t>Creating schemes that allow case-insensitive syntax components
1377containing non-ASCII characters should be avoided. Case normalization
1378of non-ASCII characters can be culturally dependent and is always a
1379complex operation. The only exception concerns non-ASCII host names
1380for which the character normalization includes a mapping step derived
1381from case folding.</t>
1382
1383</section> <!-- casenorm -->
1384
1385<section title="Character Normalization" anchor="normalization">
1386
1387<t>The Unicode Standard <xref target="UNIV6"></xref> defines various
1388equivalences between sequences of characters for various
1389purposes. Unicode Standard Annex #15 <xref target="UTR15"></xref>
1390defines various Normalization Forms for these equivalences, in
1391particular Normalization Form C (NFC, Canonical Decomposition,
1392followed by Canonical Composition) and Normalization Form KC (NFKC,
1393Compatibility Decomposition, followed by Canonical Composition).</t>
1394
1395<t> IRIs already in Unicode MUST NOT be normalized before parsing or
1396interpreting. In many non-Unicode character encodings, some text
1397cannot be represented directly. For example, the word "Vietnam" is
1398natively written "Vi&amp;#x1EC7;t Nam" (containing a LATIN SMALL
1399LETTER E WITH CIRCUMFLEX AND DOT BELOW) in NFC, but a direct
1400transcoding from the windows-1258 character encoding leads to
1401"Vi&amp;#xEA;&amp;#x323;t Nam" (containing a LATIN SMALL LETTER E WITH
1402CIRCUMFLEX followed by a COMBINING DOT BELOW). Direct transcoding of
1403other 8-bit encodings of Vietnamese may lead to other
1404representations.</t>
1405
1406<t>Equivalence of IRIs MUST rely on the assumption that IRIs are
1407appropriately pre-character-normalized rather than apply character
1408normalization when comparing two IRIs. The exceptions are conversion
1409from a non-digital form, and conversion from a non-UCS-based character
1410encoding to a UCS-based character encoding. In these cases, NFC or a
1411normalizing transcoder using NFC MUST be used for interoperability. To
1412avoid false negatives and problems with transcoding, IRIs SHOULD be
1413created by using NFC. Using NFKC may avoid even more problems; for
1414example, by choosing half-width Latin letters instead of full-width
1415ones, and full-width instead of half-width Katakana.</t>
1416
1417
1418<t>As an example,
1419"http://www.example.org/r&amp;#xE9;sum&amp;#xE9;.html" (in XML
1420Notation) is in NFC. On the other hand,
1421"http://www.example.org/re&amp;#x301;sume&amp;#x301;.html" is not in
1422NFC.</t>
1423
1424<t>The former uses precombined e-acute characters, and the latter uses
1425"e" characters followed by combining acute accents. Both usages are
1426defined as canonically equivalent in <xref target="UNIV6"></xref>.</t>
1427
1428<t><list style="hanging">
1429
1430<t hangText="Note:">
1431Because it is unknown how a particular sequence of characters is being
1432treated with respect to character normalization, it would be
1433inappropriate to allow third parties to normalize an IRI
1434arbitrarily. This does not contradict the recommendation that when a
1435resource is created, its IRI should be as character normalized as
1436possible (i.e., NFC or even NFKC). This is similar to the
1437uppercase/lowercase problems.  Some parts of a URI are case
1438insensitive (for example, the domain name). For others, it is unclear
1439whether they are case sensitive, case insensitive, or something in
1440between (e.g., case sensitive, but with a multiple choice selection if
1441the wrong case is used, instead of a direct negative result).  The
1442best recipe is that the creator use a reasonable capitalization and,
1443when transferring the URI, capitalization never be
1444changed.</t></list></t>
1445
1446<t>Various IRI schemes may allow the usage of Internationalized Domain
1447Names (IDN) <xref target="RFC3490"></xref> either in the ireg-name
1448part or elsewhere. Character Normalization also applies to IDNs, as
1449discussed in <xref target="schemecomp"></xref>.</t>
1450</section> <!-- charnorm -->
1451
1452<section title="Percent-Encoding Normalization">
1453
1454<t>The percent-encoding mechanism (Section 2.1 of <xref
1455target="RFC3986"></xref>) is a frequent source of variance among
1456otherwise identical IRIs. In addition to the case normalization issue
1457noted above, some IRI producers percent-encode octets that do not
1458require percent-encoding, resulting in IRIs that are equivalent to
1459their nonencoded counterparts. These IRIs should be normalized by
1460decoding any percent-encoded octet sequence that corresponds to an
1461unreserved character, as described in section 2.3 of <xref
1462target="RFC3986"></xref>.</t>
1463
1464<t>For actual resolution, differences in percent-encoding (except for
1465the percent-encoding of reserved characters) MUST always result in the
1466same resource.  For example, "http://example.org/~user",
1467"http://example.org/%7euser", and "http://example.org/%7Euser", must
1468resolve to the same resource.</t>
1469
1470<t>If this kind of equivalence is to be tested, the percent-encoding
1471of both IRIs to be compared has to be aligned; for example, by
1472converting both IRIs to URIs (see Section 3.1), eliminating escape
1473differences in the resulting URIs, and making sure that the case of
1474the hexadecimal characters in the percent-encoding is always the same
1475(preferably upper case). If the IRI is to be passed to another
1476application or used further in some other way, its original form MUST
1477be preserved.  The conversion described here should be performed only
1478for local comparison.</t>
1479
1480</section> <!-- pctnorm -->
1481
1482<section title="Path Segment Normalization">
1483
1484<t>The complete path segments "." and ".." are intended only for use
1485within relative references (Section 4.1 of <xref
1486target="RFC3986"></xref>) and are removed as part of the reference
1487resolution process (Section 5.2 of <xref target="RFC3986"></xref>).
1488However, some implementations may incorrectly assume that reference
1489resolution is not necessary when the reference is already an IRI, and
1490thus fail to remove dot-segments when they occur in non-relative
1491paths.  IRI normalizers should remove dot-segments by applying the
1492remove_dot_segments algorithm to the path, as described in Section
14935.2.4 of <xref target="RFC3986"></xref>.</t>
1494
1495</section> <!-- pathnorm -->
1496</section> <!-- ladder -->
1497
1498<section title="Scheme-Based Normalization" anchor="schemecomp">
1499
1500<t>The syntax and semantics of IRIs vary from scheme to scheme, as
1501described by the defining specification for each
1502scheme. Implementations may use scheme-specific rules, at further
1503processing cost, to reduce the probability of false negatives. For
1504example, because the "http" scheme makes use of an authority
1505component, has a default port of "80", and defines an empty path to be
1506equivalent to "/", the following four IRIs are equivalent:</t>
1507
1508<figure><artwork>
1509   http://example.com
1510   http://example.com/
1511   http://example.com:/
1512   http://example.com:80/</artwork></figure>
1513
1514<t>In general, an IRI that uses the generic syntax for authority with
1515an empty path should be normalized to a path of "/". Likewise, an
1516explicit ":port", for which the port is empty or the default for the
1517scheme, is equivalent to one where the port and its ":" delimiter are
1518elided and thus should be removed by scheme-based normalization. For
1519example, the second IRI above is the normal form for the "http"
1520scheme.</t>
1521
1522<t>Another case where normalization varies by scheme is in the
1523handling of an empty authority component or empty host
1524subcomponent. For many scheme specifications, an empty authority or
1525host is considered an error; for others, it is considered equivalent
1526to "localhost" or the end-user's host. When a scheme defines a default
1527for authority and an IRI reference to that default is desired, the
1528reference should be normalized to an empty authority for the sake of
1529uniformity, brevity, and internationalization. If, however, either the
1530userinfo or port subcomponents are non-empty, then the host should be
1531given explicitly even if it matches the default.</t>
1532
1533<t>Normalization should not remove delimiters when their associated
1534component is empty unless it is licensed to do so by the scheme
1535specification. For example, the IRI "http://example.com/?" cannot be
1536assumed to be equivalent to any of the examples above. Likewise, the
1537presence or absence of delimiters within a userinfo subcomponent is
1538usually significant to its interpretation.  The fragment component is
1539not subject to any scheme-based normalization; thus, two IRIs that
1540differ only by the suffix "#" are considered different regardless of
1541the scheme.</t>
1542 
1543<t>Some IRI schemes allow the usage of Internationalized Domain
1544Names (IDN) <xref target='RFC5890'></xref> either in their ireg-name
1545part or elswhere. When in use in IRIs, those names SHOULD
1546conform to the definition of U-Label in <xref
1547target='RFC5890'></xref>. An IRI containing an invalid IDN cannot
1548successfully be resolved. For legibility purposes, they
1549SHOULD NOT be converted into ASCII Compatible Encoding (ACE).</t>
1550
1551<t>Scheme-based normalization may also consider IDN
1552components and their conversions to punycode as equivalent. As an
1553example, "http://r&amp;#xE9;sum&amp;#xE9;.example.org" may be
1554considered equivalent to
1555"http://xn--rsum-bpad.example.org".</t><t>Other scheme-specific
1556normalizations are possible.</t>
1557
1558</section> <!-- schemenorm -->
1559
1560<section title="Protocol-Based Normalization">
1561
1562<t>Substantial effort to reduce the incidence of false negatives is
1563often cost-effective for web spiders. Consequently, they implement
1564even more aggressive techniques in IRI comparison. For example, if
1565they observe that an IRI such as</t>
1566
1567<figure><artwork>
1568   http://example.com/data</artwork></figure>
1569<t>redirects to an IRI differing only in the trailing slash</t>
1570<figure><artwork>
1571   http://example.com/data/</artwork></figure>
1572
1573<t>they will likely regard the two as equivalent in the future.  This
1574kind of technique is only appropriate when equivalence is clearly
1575indicated by both the result of accessing the resources and the common
1576conventions of their scheme's dereference algorithm (in this case, use
1577of redirection by HTTP origin servers to avoid problems with relative
1578references).</t>
1579
1580</section> <!-- protonorm -->
1581</section> <!-- equivalence -->
1582</section> 
1583
1584<section title="Use of IRIs" anchor="IRIuse">
1585
1586<section title="Limitations on UCS Characters Allowed in IRIs" anchor="limitations">
1587
1588<t>This section discusses limitations on characters and character
1589sequences usable for IRIs beyond those given in <xref target="abnf"/>
1590and <xref target="visual"/>. The considerations in this section are
1591relevant when IRIs are created and when URIs are converted to
1592IRIs.</t>
1593
1594<t>
1595
1596<list style="hanging"><t hangText="a.">The repertoire of characters allowed
1597    in each IRI component is limited by the definition of that component.
1598    For example, the definition of the scheme component does not allow
1599    characters beyond US-ASCII.
1600    <vspace blankLines="1"/>
1601    (Note: In accordance with URI practice, generic IRI
1602    software cannot and should not check for such limitations.)</t>
1603
1604<t hangText="b.">The UCS contains many areas of characters for which
1605    there are strong visual look-alikes. Because of the likelihood of
1606    transcription errors, these also should be avoided. This includes
1607    the full-width equivalents of Latin characters, half-width
1608    Katakana characters for Japanese, and many others. It also
1609    includes many look-alikes of "space", "delims", and "unwise",
1610    characters excluded in <xref target="RFC3491"/>.</t>
1611   
1612</list>
1613</t>
1614
1615<t>Additional information is available from <xref target="UNIXML"/>.
1616    <xref target="UNIXML"/> is written in the context of running text
1617    rather than in that of identifiers. Nevertheless, it discusses
1618    many of the categories of characters not appropriate for IRIs.</t>
1619</section> <!-- limitations -->
1620
1621<section title="Software Interfaces and Protocols">
1622
1623<t>Although an IRI is defined as a sequence of characters, software
1624interfaces for URIs typically function on sequences of octets or other
1625kinds of code units. Thus, software interfaces and protocols MUST
1626define which character encoding is used.</t>
1627
1628<t>Intermediate software interfaces between IRI-capable components and
1629URI-only components MUST map the IRIs per <xref target="mapping"/>,
1630when transferring from IRI-capable to URI-only components.
1631
1632This mapping SHOULD be applied as late as possible. It SHOULD NOT be
1633applied between components that are known to be able to handle IRIs.</t>
1634</section> <!-- software -->
1635
1636<section title="Format of URIs and IRIs in Documents and Protocols">
1637
1638<t>Document formats that transport URIs may have to be upgraded to allow
1639the transport of IRIs. In cases where the document as a whole
1640has a native character encoding, IRIs MUST also be encoded in this
1641character encoding and converted accordingly by a parser or interpreter.
1642
1643IRI characters not expressible in the native character encoding SHOULD
1644be escaped by using the escaping conventions of the document format if
1645such conventions are available. Alternatively, they MAY be
1646percent-encoded according to <xref target="mapping"/>. For example, in
1647HTML or XML, numeric character references SHOULD be used. If a
1648document as a whole has a native character encoding and that character
1649encoding is not UTF-8, then IRIs MUST NOT be placed into the document
1650in the UTF-8 character encoding.</t>
1651
1652<t>((UPDATE THIS NOTE)) Note: Some formats already accommodate IRIs,
1653although they use different terminology. HTML 4.0 <xref
1654target="HTML4"/> defines the conversion from IRIs to URIs as
1655error-avoiding behavior. XML 1.0 <xref target="XML1"/>, XLink <xref
1656target="XLink"/>, XML Schema <xref target="XMLSchema"/>, and
1657specifications based upon them allow IRIs. Also, it is expected that
1658all relevant new W3C formats and protocols will be required to handle
1659IRIs <xref target="CharMod"/>.</t>
1660
1661</section> <!-- format -->
1662
1663<section title="Use of UTF-8 for Encoding Original Characters" anchor="UTF8use">
1664
1665<t>This section discusses details and gives examples for point c) in
1666<xref target="Applicability"/>. To be able to use IRIs, the URI
1667corresponding to the IRI in question has to encode original characters
1668into octets by using UTF-8.  This can be specified for all URIs of a
1669URI scheme or can apply to individual URIs for schemes that do not
1670specify how to encode original characters.  It can apply to the whole
1671URI, or only to some part. For background information on encoding
1672characters into URIs, see also Section 2.5 of <xref
1673target="RFC3986"/>.</t>
1674
1675<t>For new URI schemes, using UTF-8 is recommended in <xref
1676target="RFC4395bis"/>.  Examples where UTF-8 is already used are the URN
1677syntax <xref target="RFC2141"/>, IMAP URLs <xref target="RFC2192"/>,
1678and POP URLs <xref target="RFC2384"/>.  On the other hand, because the
1679HTTP URI scheme does not specify how to encode original characters,
1680only some HTTP URLs can have corresponding but different IRIs.</t>
1681
1682<t>For example, for a document with a URI
1683of<vspace/>"http://www.example.org/r%C3%A9sum%C3%A9.html", it is
1684possible to construct a corresponding IRI (in XML notation, see <xref
1685target="sec-Notation"/>):
1686"http://www.example.org/r&amp;#xE9;sum&amp;#xE9;.html" ("&amp;#xE9;"
1687stands for the e-acute character, and "%C3%A9" is the UTF-8 encoded
1688and percent-encoded representation of that character). On the other
1689hand, for a document with a URI of
1690"http://www.example.org/r%E9sum%E9.html", the percent-encoding octets
1691cannot be converted to actual characters in an IRI, as the
1692percent-encoding is not based on UTF-8.</t>
1693
1694<t>For most URI schemes, there is no need to upgrade their scheme
1695definition in order for them to work with IRIs.  The main case where
1696upgrading makes sense is when a scheme definition, or a particular
1697component of a scheme, is strictly limited to the use of US-ASCII
1698characters with no provision to include non-ASCII characters/octets
1699via percent-encoding, or if a scheme definition currently uses highly
1700scheme-specific provisions for the encoding of non-ASCII characters.
1701An example of this is the mailto: scheme <xref target="RFC2368"/>.</t>
1702
1703<t>This specification updates the IANA registry of URI schemes to note
1704their applicability to IRIs, see <xref target="iana"/>.  All IRIs use
1705URI schemes, and all URIs with URI schemes can be used as IRIs, even
1706though in some cases only by using URIs directly as IRIs, without any
1707conversion.</t>
1708
1709<t>Scheme definitions can impose restrictions on the syntax of
1710scheme-specific URIs; i.e., URIs that are admissible under the generic
1711URI syntax <xref target="RFC3986"/> may not be admissible due to
1712narrower syntactic constraints imposed by a URI scheme
1713specification. URI scheme definitions cannot broaden the syntactic
1714restrictions of the generic URI syntax; otherwise, it would be
1715possible to generate URIs that satisfied the scheme-specific syntactic
1716constraints without satisfying the syntactic constraints of the
1717generic URI syntax. However, additional syntactic constraints imposed
1718by URI scheme specifications are applicable to IRI, as the
1719corresponding URI resulting from the mapping defined in <xref
1720target="mapping"/> MUST be a valid URI under the syntactic
1721restrictions of generic URI syntax and any narrower restrictions
1722imposed by the corresponding URI scheme specification.</t>
1723
1724<t>The requirement for the use of UTF-8 generally applies to all parts
1725of a URI.  However, it is possible that the capability of IRIs to
1726represent a wide range of characters directly is used just in some
1727parts of the IRI (or IRI reference). The other parts of the IRI may
1728only contain US-ASCII characters, or they may not be based on
1729UTF-8. They may be based on another character encoding, or they may
1730directly encode raw binary data (see also <xref
1731target="RFC2397"/>). </t>
1732
1733<t>For example, it is possible to have a URI reference
1734of<vspace/>"http://www.example.org/r%E9sum%E9.xml#r%C3%A9sum%C3%A9",
1735where the document name is encoded in iso-8859-1 based on server
1736settings, but where the fragment identifier is encoded in UTF-8 according
1737to <xref target="XPointer"/>. The IRI corresponding to the above
1738URI would be (in XML notation)<vspace/>"http://www.example.org/r%E9sum%E9.xml#r&amp;#xE9;sum&amp;#xE9;".</t>
1739
1740<t>Similar considerations apply to query parts. The functionality
1741of IRIs (namely, to be able to include non-ASCII characters) can
1742only be used if the query part is encoded in UTF-8.</t>
1743
1744</section> <!-- utf8 -->
1745
1746<section title="Relative IRI References">
1747<t>Processing of relative IRI references against a base is handled
1748straightforwardly; the algorithms of <xref target="RFC3986"/> can
1749be applied directly, treating the characters additionally allowed
1750in IRI references in the same way that unreserved characters are in URI
1751references.</t>
1752
1753</section> <!-- relative -->
1754</section> <!-- IRIuse -->
1755
1756<section title="Liberal Handling of Otherwise Invalid IRIs" anchor="LEIRIHREF">
1757
1758<t>(EDITOR NOTE: This Section may move to an appendix.)
1759 
1760Some technical specifications and widely-deployed software have
1761allowed additional variations and extensions of IRIs to be used in
1762syntactic components. This section describes two widely-used
1763preprocessing agreements. Other technical specifications may wish to
1764reference a syntactic component which is "a valid IRI or a string that
1765will map to a valid IRI after this preprocessing algorithm". These two
1766variants are known as <xref target="LEIRI">Legacy Extended IRI or
1767LEIRI</xref>, and <xref target="HTML5">Web Address</xref>).
1768</t>
1769
1770<t>Future technical specifications SHOULD NOT allow conforming
1771producers to produce, or conforming content to contain, such forms,
1772as they are not interoperable with other IRI consuming software.</t>
1773
1774<section title="LEIRI Processing"  anchor="LEIRIspec">
1775  <t>This section defines Legacy Extended IRIs (LEIRIs).
1776    The syntax of Legacy Extended IRIs is the same as that for &lt;IRI-reference>,
1777    except that the ucschar production is replaced by the leiri-ucschar production:</t>
1778<figure>
1779
1780<artwork>
1781  leiri-ucschar  = " " / "&lt;" / "&gt;" / '"' / "{" / "}" / "|"
1782                   / "\" / "^" / "`" / %x0-1F / %x7F-D7FF
1783                   / %xE000-FFFD / %x10000-10FFFF
1784</artwork>
1785
1786<postamble>
1787  Among other extensions, processors based on this specification also
1788  did not enforce the restriction on bidirectional formatting
1789  characters in <xref target="visual"></xref>, and the iprivate
1790  production becomes redundant.</postamble>
1791</figure>
1792
1793<t>To convert a string allowed as a LEIRI to an IRI, each character
1794allowed in leiri-ucschar but not in ucschar must be percent-encoded
1795using <xref target="compmapping"/>.</t>
1796</section> <!-- leiriproc -->
1797
1798<section title="Web Address Processing" anchor="webaddress">
1799
1800<t>Many popular web browsers have taken the approach of being quite
1801liberal in what is accepted as a "URL" or its relative
1802forms. This section describes their behavior in terms of a preprocessor
1803which maps strings into the IRI space for subsequent parsing and
1804interpretation as an IRI.</t>
1805
1806<t>In some situations, it might be appropriate to describe the syntax
1807that a liberal consumer implementation might accept as a "Web
1808Address" or "Hypertext Reference" or "HREF". However,
1809technical specifications SHOULD restrict the syntactic form allowed by compliant producers
1810to the IRI or IRI reference syntax defined in this document
1811even if they want to mandate this processing.</t>
1812
1813<t>
1814Summary:
1815<list style="symbols">
1816   <t>Leading and trailing whitespace is removed.</t>
1817   <t>Some additional characters are removed.</t>
1818   <t>Some additional characters are allowed and escaped (as with LEIRI).</t>
1819   <t>If interpreting an IRI as a URI, the pct-encoding of the query
1820   component of the parsed URI component depends on operational
1821   context.</t>
1822</list>
1823</t>
1824
1825<t>Each string provided may have an associated charset (called
1826the HREF-charset here); this defaults to UTF-8.
1827For web browsers interpreting HTML, the document
1828charset of a string is determined:
1829
1830<list style="hanging">
1831<t hangText="If the string came from a script (e.g. as an argument to
1832 a method)">The HRef-charset is the script's charset.</t>
1833
1834<t hangText="If the string came from a DOM node (e.g. from an
1835  element)">The node has a Document, and the HRef-charset is the
1836  Document's character encoding.</t>
1837
1838<t hangText="If the string had a HRef-charset defined when the string was
1839created or defined">The HRef-charset is as defined.</t>
1840
1841</list></t>
1842
1843<t>If the resulting HRef-charset is a unicode based character encoding
1844(e.g., UTF-16), then use UTF-8 instead.</t>
1845
1846
1847<figure>
1848<preamble>The syntax for Web Addresses is obtained by replacing the 'ucschar',
1849  pct-form, and path-sep rules with the href-ucschar, href-pct-form, and href-path-sep
1850  rules below. In addition, some characters are stripped.</preamble>
1851
1852<artwork type='abnf'>
1853  href-ucschar  = " " / "&lt;" / "&gt;" / DQUOTE / "{" / "}" / "|"
1854                   / "\" / "^" / "`" / %x0-1F / %x7F-D7FF
1855                   / %xE000-FFFD / %x10000-10FFFF
1856  href-pct-form = pct-encoded / "%"
1857  href-path-sep = "/" / "\"
1858  href-strip    = &lt;to be done&gt;
1859</artwork>
1860
1861<postamble>
1862(NOTE: NEED TO FIX THESE SETS TO MATCH HTML5; NOT SURE ABOUT NEXT SENTENCE)
1863browsers did not enforce the restriction on bidirectional formatting
1864  characters in <xref target="visual"></xref>, and the iprivate
1865  production becomes redundant.</postamble>
1866</figure>
1867
1868<t>'Web Address processing' requires the following additional
1869preprocessing steps:
1870
1871<list style="numbers">
1872
1873<t>Leading and trailing instances of space (U+0020),
1874CR (U+000A), LF (U+000D), and TAB (U+0009) characters are removed.</t>
1875
1876<t>strip all characters in href-strip.</t>
1877  <t>Percent-encode all characters in href-ucschar not in ucschar.</t>
1878  <t>Replace occurrences of "%" not followed by two hexadecimal digits by "%25".</t>
1879  <t>Convert backslashes ('\') matching href-path-sep to forward slashes ('/').</t>
1880</list></t>
1881</section> <!-- webaddress -->
1882
1883<section title="Characters Not Allowed in IRIs" anchor="notAllowed">
1884
1885<t>This section provides a list of the groups of characters and code
1886points that are allowed by LEIRI or HREF but are not allowed in IRIs or are
1887allowed in IRIs only in the query part. For each group of characters,
1888advice on the usage of these characters is also given, concentrating
1889on the reasons for why they are excluded from IRI use.</t>
1890
1891<t>
1892
1893<list><t>Space (U+0020): Some formats and applications use space as a
1894delimiter, e.g. for items in a list. Appendix C of <xref
1895target="RFC3986"></xref> also mentions that white space may have to be
1896added when displaying or printing long URIs; the same applies to long
1897IRIs. This means that spaces can disappear, or can make the what is
1898intended as a single IRI or IRI reference to be treated as two or more
1899separate IRIs.</t>
1900
1901<t>Delimiters "&lt;" (U+003C), "&gt;" (U+003E), and '"' (U+0022):
1902Appendix C of <xref target="RFC3986"></xref> suggests the use of
1903double-quotes ("http://example.com/") and angle brackets
1904(&lt;http://example.com/&gt;) as delimiters for URIs in plain
1905text. These conventions are often used, and also apply to IRIs.  Using
1906these characters in strings intended to be IRIs would result in the
1907IRIs being cut off at the wrong place.</t>
1908
1909<t>Unwise characters "\" (U+005C), "^" (U+005E), "`"
1910(U+0060), "{" (U+007B), "|" (U+007C), and "}" (U+007D): These
1911characters originally have been excluded from URIs because the
1912respective codepoints are assigned to different graphic characters in
1913some 7-bit or 8-bit encoding. Despite the move to Unicode, some of
1914these characters are still occasionally displayed differently on some
1915systems, e.g. U+005C may appear as a Japanese Yen symbol on some
1916systems. Also, the fact that these characters are not used in URIs or
1917IRIs has encouraged their use outside URIs or IRIs in contexts that
1918may include URIs or IRIs. If a string with such a character were used
1919as an IRI in such a context, it would likely be interpreted
1920piecemeal.</t>
1921
1922<t>The controls (C0 controls, DEL, and C1 controls, #x0 - #x1F #x7F -
1923#x9F): There is generally no way to transmit these characters reliably
1924as text outside of a charset encoding.  Even when in encoded form,
1925many software components silently filter out some of these characters,
1926or may stop processing alltogether when encountering some of
1927them. These characters may affect text display in subtle, unnoticable
1928ways or in drastic, global, and irreversible ways depending on the
1929hardware and software involved. The use of some of these characters
1930would allow malicious users to manipulate the display of an IRI and
1931its context in many situations.</t>
1932
1933<t>Bidi formatting characters (U+200E, U+200F, U+202A-202E): These
1934characters affect the display ordering of characters. If IRIs were
1935allowed to contain these characters and the resulting visual display
1936transcribed. they could not be converted back to electronic form
1937(logical order) unambiguously. These characters, if allowed in IRIs,
1938might allow malicious users to manipulate the display of IRI and its
1939context.</t>
1940
1941<t>Specials (U+FFF0-FFFD): These code points provide functionality
1942beyond that useful in an IRI, for example byte order identification,
1943annotation, and replacements for unknown characters and objects. Their
1944use and interpretation in an IRI would serve no purpose and might lead
1945to confusing display variations.</t>
1946
1947<t>Private use code points (U+E000-F8FF, U+F0000-FFFFD,
1948U+100000-10FFFD): Display and interpretation of these code points is
1949by definition undefined without private agreement. Therefore, these
1950code points are not suited for use on the Internet. They are not
1951interoperable and may have unpredictable effects.</t>
1952
1953<t>Tags (U+E0000-E0FFF): These characters provide a way to language
1954tag in Unicode plain text. They are not appropriate for IRIs because
1955language information in identifiers cannot reliably be input,
1956transmitted (e.g. on a visual medium such as paper), or
1957recognized.</t>
1958
1959<t>Non-characters (U+FDD0-FDEF, U+1FFFE-1FFFF, U+2FFFE-2FFFF,
1960U+3FFFE-3FFFF, U+4FFFE-4FFFF, U+5FFFE-5FFFF, U+6FFFE-6FFFF,
1961U+7FFFE-7FFFF, U+8FFFE-8FFFF, U+9FFFE-9FFFF, U+AFFFE-AFFFF,
1962U+BFFFE-BFFFF, U+CFFFE-CFFFF, U+DFFFE-DFFFF, U+EFFFE-EFFFF,
1963U+FFFFE-FFFFF, U+10FFFE-10FFFF): These code points are defined as
1964non-characters. Applications may use some of them internally, but are
1965not prepared to interchange them.</t>
1966
1967</list></t>
1968
1969<t>LEIRI preprocessing disallowed some code points and
1970code units:
1971
1972<list><t>Surrogate code units (D800-DFFF): These do not represent
1973Unicode codepoints.</t></list></t>
1974</section> <!-- notallowed -->
1975</section> <!-- lieirihref -->
1976 
1977<section title="URI/IRI Processing Guidelines (Informative)" anchor="guidelines">
1978
1979<t>This informative section provides guidelines for supporting IRIs in
1980the same software components and operations that currently process
1981URIs: Software interfaces that handle URIs, software that allows users
1982to enter URIs, software that creates or generates URIs, software that
1983displays URIs, formats and protocols that transport URIs, and software
1984that interprets URIs. These may all require modification before
1985functioning properly with IRIs. The considerations in this section
1986also apply to URI references and IRI references.</t>
1987
1988<section title="URI/IRI Software Interfaces">
1989<t>Software interfaces that handle URIs, such as URI-handling APIs and
1990protocols transferring URIs, need interfaces and protocol elements
1991that are designed to carry IRIs.</t>
1992
1993<t>In case the current handling in an API or protocol is based on
1994US-ASCII, UTF-8 is recommended as the character encoding for IRIs, as
1995it is compatible with US-ASCII, is in accordance with the
1996recommendations of <xref target="RFC2277"/>, and makes converting to
1997URIs easy. In any case, the API or protocol definition must clearly
1998define the character encoding to be used.</t>
1999
2000<t>The transfer from URI-only to IRI-capable components requires no
2001mapping, although the conversion described in <xref
2002target="URItoIRI"/> above may be performed. It is preferable not to
2003perform this inverse conversion unless it is certain this can be done
2004correctly.</t>
2005</section>
2006
2007<section title="URI/IRI Entry">
2008
2009<t>Some components allow users to enter URIs into the system
2010by typing or dictation, for example. This software must be updated to allow
2011for IRI entry.</t>
2012
2013<t>A person viewing a visual representation of an IRI (as a sequence
2014of glyphs, in some order, in some visual display) or hearing an IRI
2015will use an entry method for characters in the user's language to
2016input the IRI. Depending on the script and the input method used, this
2017may be a more or less complicated process.</t>
2018
2019<t>The process of IRI entry must ensure, as much as possible, that the
2020restrictions defined in <xref target="abnf"/> are met. This may be
2021done by choosing appropriate input methods or variants/settings
2022thereof, by appropriately converting the characters being input, by
2023eliminating characters that cannot be converted, and/or by issuing a
2024warning or error message to the user.</t>
2025
2026<t>As an example of variant settings, input method editors for East
2027Asian Languages usually allow the input of Latin letters and related
2028characters in full-width or half-width versions. For IRI input, the
2029input method editor should be set so that it produces half-width Latin
2030letters and punctuation and full-width Katakana.</t>
2031
2032<t>An input field primarily or solely used for the input of URIs/IRIs
2033might allow the user to view an IRI as it is mapped to a URI.  Places
2034where the input of IRIs is frequent may provide the possibility for
2035viewing an IRI as mapped to a URI. This will help users when some of
2036the software they use does not yet accept IRIs.</t>
2037
2038<t>An IRI input component interfacing to components that handle URIs,
2039but not IRIs, must map the IRI to a URI before passing it to these
2040components.</t>
2041
2042<t>For the input of IRIs with right-to-left characters, please see
2043<xref target="bidiInput"></xref>.</t>
2044</section>
2045
2046<section title="URI/IRI Transfer between Applications">
2047
2048<t>Many applications (for example, mail user agents) try to detect
2049URIs appearing in plain text. For this, they use some heuristics based
2050on URI syntax. They then allow the user to click on such URIs and
2051retrieve the corresponding resource in an appropriate (usually
2052scheme-dependent) application.</t>
2053
2054<t>Such applications would need to be upgraded, in order to use the
2055IRI syntax as a base for heuristics. In particular, a non-ASCII
2056character should not be taken as the indication of the end of an IRI.
2057Such applications also would need to make sure that they correctly
2058convert the detected IRI from the character encoding of the document
2059or application where the IRI appears, to the character encoding used
2060by the system-wide IRI invocation mechanism, or to a URI (according to
2061<xref target="mapping"/>) if the system-wide invocation mechanism only
2062accepts URIs.</t>
2063
2064<t>The clipboard is another frequently used way to transfer URIs and
2065IRIs from one application to another. On most platforms, the clipboard
2066is able to store and transfer text in many languages and scripts.
2067Correctly used, the clipboard transfers characters, not octets, which
2068will do the right thing with IRIs.</t>
2069</section>
2070
2071<section title="URI/IRI Generation">
2072
2073<t>Systems that offer resources through the Internet, where those
2074resources have logical names, sometimes automatically generate URIs
2075for the resources they offer. For example, some HTTP servers can
2076generate a directory listing for a file directory and then respond to
2077the generated URIs with the files.</t>
2078
2079<t>Many legacy character encodings are in use in various file systems.
2080Many currently deployed systems do not transform the local character
2081representation of the underlying system before generating URIs.</t>
2082
2083<t>For maximum interoperability, systems that generate resource
2084identifiers should make the appropriate transformations. For example,
2085if a file system contains a file named
2086"r&amp;#xE9;sum&amp;#xE9;.html", a server should expose this as
2087"r%C3%A9sum%C3%A9.html" in a URI, which allows use of
2088"r&amp;#xE9;sum&amp;#xE9;.html" in an IRI, even if locally the file
2089name is kept in a character encoding other than UTF-8.
2090</t>
2091
2092<t>This recommendation particularly applies to HTTP servers. For FTP
2093servers, similar considerations apply; see <xref target="RFC2640"/>.</t>
2094</section>
2095
2096<section title="URI/IRI Selection" anchor="selection">
2097<t>In some cases, resource owners and publishers have control over the
2098IRIs used to identify their resources. This control is mostly
2099executed by controlling the resource names, such as file names,
2100directly.</t>
2101
2102<t>In these cases, it is recommended to avoid choosing IRIs that are
2103easily confused. For example, for US-ASCII, the lower-case ell ("l") is
2104easily confused with the digit one ("1"), and the upper-case oh ("O") is
2105easily confused with the digit zero ("0"). Publishers should avoid
2106confusing users with "br0ken" or "1ame" identifiers.</t>
2107
2108<t>Outside the US-ASCII repertoire, there are many more opportunities for
2109confusion; a complete set of guidelines is too lengthy to include
2110here. As long as names are limited to characters from a single script,
2111native writers of a given script or language will know best when
2112ambiguities can appear, and how they can be avoided. What may look
2113ambiguous to a stranger may be completely obvious to the average
2114native user. On the other hand, in some cases, the UCS contains
2115variants for compatibility reasons; for example, for typographic purposes.
2116These should be avoided wherever possible. Although there may be exceptions,
2117newly created resource names should generally be in NFKC
2118<xref target="UTR15"></xref> (which means that they are also in NFC).</t>
2119
2120<t>As an example, the UCS contains the "fi" ligature at U+FB01
2121for compatibility reasons.
2122Wherever possible, IRIs should use the two letters "f" and "i" rather
2123than the "fi" ligature. An example where the latter may be used is
2124in the query part of an IRI for an explicit search for a word written
2125containing the "fi" ligature.</t>
2126
2127<t>In certain cases, there is a chance that characters from different
2128scripts look the same. The best known example is the similarity of the
2129Latin "A", the Greek "Alpha", and the Cyrillic "A". To avoid such
2130cases, IRIs should only be created where all the characters in a
2131single component are used together in a given language. This usually
2132means that all of these characters will be from the same script, but
2133there are languages that mix characters from different scripts (such
2134as Japanese).  This is similar to the heuristics used to distinguish
2135between letters and numbers in the examples above. Also, for Latin,
2136Greek, and Cyrillic, using lowercase letters results in fewer
2137ambiguities than using uppercase letters would.</t>
2138</section>
2139
2140<section title="Display of URIs/IRIs" anchor="display">
2141<t>
2142In situations where the rendering software is not expected to display
2143non-ASCII parts of the IRI correctly using the available layout and font
2144resources, these parts should be percent-encoded before being displayed.</t>
2145
2146<t>For display of Bidi IRIs, please see <xref target="visual"/>.</t>
2147</section>
2148
2149<section title="Interpretation of URIs and IRIs">
2150<t>Software that interprets IRIs as the names of local resources should
2151accept IRIs in multiple forms and convert and match them with the
2152appropriate local resource names.</t>
2153
2154<t>First, multiple representations include both IRIs in the native
2155character encoding of the protocol and also their URI counterparts.</t>
2156
2157<t>Second, it may include URIs constructed based on character
2158encodings other than UTF-8. These URIs may be produced by user agents that do
2159not conform to this specification and that use legacy character encodings to
2160convert non-ASCII characters to URIs. Whether this is necessary, and what
2161character encodings to cover, depends on a number of factors, such as
2162the legacy character encodings used locally and the distribution of
2163various versions of user agents. For example, software for Japanese
2164may accept URIs in Shift_JIS and/or EUC-JP in addition to UTF-8.</t>
2165
2166<t>Third, it may include additional mappings to be more user-friendly
2167and robust against transmission errors. These would be similar to how
2168some servers currently treat URIs as case insensitive or perform
2169additional matching to account for spelling errors. For characters
2170beyond the US-ASCII repertoire, this may, for example, include
2171ignoring the accents on received IRIs or resource names. Please note
2172that such mappings, including case mappings, are language
2173dependent.</t>
2174
2175<t>It can be difficult to identify a resource unambiguously if too
2176many mappings are taken into consideration. However, percent-encoded
2177and not percent-encoded parts of IRIs can always be clearly distinguished.
2178Also, the regularity of UTF-8 (see <xref target="Duerst97"/>) makes the
2179potential for collisions lower than it may seem at first.</t>
2180</section>
2181
2182<section title="Upgrading Strategy">
2183<t>Where this recommendation places further constraints on software
2184for which many instances are already deployed, it is important to
2185introduce upgrades carefully and to be aware of the various
2186interdependencies.</t>
2187
2188<t>If IRIs cannot be interpreted correctly, they should not be created,
2189generated, or transported. This suggests that upgrading URI interpreting
2190software to accept IRIs should have highest priority.</t>
2191
2192<t>On the other hand, a single IRI is interpreted only by a single or
2193very few interpreters that are known in advance, although it may be
2194entered and transported very widely.</t>
2195
2196<t>Therefore, IRIs benefit most from a broad upgrade of software to be
2197able to enter and transport IRIs. However, before an
2198individual IRI is published, care should be taken to upgrade the corresponding
2199interpreting software in order to cover the forms expected to be
2200received by various versions of entry and transport software.</t>
2201
2202<t>The upgrade of generating software to generate IRIs instead of using a
2203local character encoding should happen only after the service is upgraded
2204to accept IRIs. Similarly, IRIs should only be generated when the service
2205accepts IRIs and the intervening infrastructure and protocol is known
2206to transport them safely.</t>
2207
2208<t>Software converting from URIs to IRIs for display should be upgraded
2209only after upgraded entry software has been widely deployed to the
2210population that will see the displayed result.</t>
2211
2212
2213<t>Where there is a free choice of character encodings, it is often
2214possible to reduce the effort and dependencies for upgrading to IRIs
2215by using UTF-8 rather than another encoding. For example, when a new
2216file-based Web server is set up, using UTF-8 as the character encoding
2217for file names will make the transition to IRIs easier. Likewise, when
2218a new Web form is set up using UTF-8 as the character encoding of the
2219form page, the returned query URIs will use UTF-8 as the character
2220encoding (unless the user, for whatever reason, changes the character
2221encoding) and will therefore be compatible with IRIs.</t>
2222
2223
2224<t>These recommendations, when taken together, will allow for the
2225extension from URIs to IRIs in order to handle characters other than
2226US-ASCII while minimizing interoperability problems. For
2227considerations regarding the upgrade of URI scheme definitions, see
2228<xref target="UTF8use"/>.</t>
2229
2230</section>
2231</section> <!-- guidelines -->
2232
2233<section title="IANA Considerations" anchor="iana">
2234
2235<t>RFC Editor and IANA note: Please Replace RFC XXXX with the
2236number of this document when it issues as an RFC. </t>
2237
2238<t>IANA maintains a registry of "URI schemes". A "URI scheme" also
2239serves an "IRI scheme". </t>
2240
2241<t>To clarify that the URI scheme registration process also applies to
2242IRIs, change the description of the "URI schemes" registry
2243header to say "[RFC4395] defines an IANA-maintained registry of URI
2244Schemes. These registries include the Permanent and Provisional URI
2245Schemes.  RFC XXXX updates this registry to designate that schemes may
2246also indicate their usability as IRI schemes.</t>
2247
2248<t> Update "per RFC 4395" to "per RFC 4395 and RFC XXXX".
2249</t>
2250
2251</section> <!-- IANA -->
2252   
2253<section title="Security Considerations" anchor="security">
2254<t>The security considerations discussed in <xref target="RFC3986"/>
2255also apply to IRIs. In addition, the following issues require
2256particular care for IRIs.</t>
2257<t>Incorrect encoding or decoding can lead to security problems.
2258For example, some UTF-8 decoders do not check against overlong
2259byte sequences. See <xref target='UTR36'/> Section 3 for details.</t>
2260
2261  <t>There are serious difficulties with relying on a human to verify that a
2262    an IRI (whether presented visually or aurally)
2263    is the same as another IRI or is the one intended.
2264    These problems exist with ASCII-only URIs (bl00mberg.com vs. bloomberg.com)
2265    but are strongly exacerbated when using the much larger character repertoire of Unicode.
2266    For details, see Section 2 of <xref target='UTR36'/>.
2267    There seems to be little hope of relying on either administrative or technical means
2268    to reduce the availability of such exploits, to the extent that user agents SHOULD NOT
2269    relying on visual or perceptual comparison or verification of IRIs
2270    as any means of validating or assuring safety, correctness or appropriateness of an IRI.</t>
2271
2272<t>There are various ways in which "spoofing" can occur with IRIs.
2273"Spoofing" means that somebody may add a resource name that looks the
2274same or similar to the user, but that points to a different resource.
2275The added resource may pretend to be the real resource by looking
2276very similar but may contain all kinds of changes that may be
2277difficult to spot and that can cause all kinds of problems.
2278Most spoofing possibilities for IRIs are extensions of those for URIs.</t>
2279
2280<t>Spoofing can occur for various reasons. First, a user's normalization expectations or actual normalization
2281when entering an IRI or  transcoding an IRI from a legacy character
2282encoding do not match the normalization used on the
2283server side. Conceptually, this is no different from the problems
2284surrounding the use of case-insensitive web servers. For example,
2285a popular web page with a mixed-case name ("http://big.example.com/PopularPage.html")
2286might be "spoofed" by someone who is able to create "http://big.example.com/popularpage.html".
2287However, the use of unnormalized character sequences, and of additional
2288mappings for user convenience, may increase the chance for spoofing.
2289Protocols and servers that allow the creation of resources with
2290names that are not normalized are particularly vulnerable to such
2291attacks. This is an inherent
2292security problem of the relevant protocol, server, or resource
2293and is not specific to IRIs, but it is mentioned here for completeness.</t>
2294
2295<t>Spoofing can occur in various IRI components, such as the
2296domain name part or a path part. For considerations specific
2297to the domain name part, see <xref target="RFC3491"/>.
2298For the path part, administrators of sites that allow independent
2299users to create resources in the same sub area may have to be careful
2300to check for spoofing.</t>
2301
2302<t>Spoofing can occur because in the UCS many characters look very similar. Details are discussed in <xref target="selection"/>.
2303Again, this is very similar to spoofing possibilities on US-ASCII,
2304e.g., using "br0ken" or "1ame" URIs.</t>
2305
2306<t>Spoofing can occur when URIs with percent-encodings based on various
2307character encodings are accepted to deal with older user agents. In some
2308cases, particularly for Latin-based resource names, this is usually easy to
2309detect because UTF-8-encoded names, when interpreted and viewed as
2310legacy character encodings, produce mostly garbage.</t><t>When
2311concurrently used character encodings have a similar structure but there
2312are no characters that have exactly the same encoding, detection is more
2313difficult.</t>
2314
2315<t>Spoofing can occur with bidirectional IRIs, if the restrictions
2316in <xref target="bidi-structure"/> are not followed. The same visual
2317representation may be interpreted as different logical representations,
2318and vice versa. It is also very important that a correct Unicode bidirectional
2319implementation be used.</t>
2320  <t>The characters additionally allowed in Legacy Extended IRIs
2321    introduce additional security issues. For details, see <xref target='#notAllowed'/>.</t>
2322</section><!-- security -->
2323
2324<section title="Acknowledgements">
2325<t>This document was derived from <xref target="RFC3987"/>; the acknowledgments from
2326that specification still apply.</t>
2327<t>We would like to thank Ian Hickson, Michael Sperberg-McQueen,
2328  and Dan Connolly for their work on HyperText References, and Norman Walsh, Richard Tobin,
2329  Henry S. Thomson, John Cowan, Paul Grosso, and the XML Core Working Group of the W3C for their work on LEIRIs.</t>
2330<t>In addition, this document was influenced by contributions from (in no particular order) Chris
2331  Lilley, Bjoern Hoehrmann,
2332Felix Sasaki, Jeremy Carroll, Frank Ellermann, Michael Everson, Cary Karp, Matitiahu Allouche,
2333Richard Ishida, Addison Phillips, Jonathan Rosenne, Najib Tounsi, Debbie Garside, Mark Davis, Sarmad
2334Hussain, Ted Hardie, Konrad Lanz, Thomas Roessler, Lisa Dusseault, Julian Reschke, Giovanni
2335Campagna, Anne van Kesteren, Mark Nottingham, Erik van der Poel, Marcin Hanclik, Marcos Caceres, Roy
2336Fielding, Greg Wilkins, Pieter Hintjens, Daniel R. Tobias, Marko Martin, Maciej Stanchowiak, Wil
2337Tan, Yui Naruse, Michael A. Puls II, Dave Thaler, Tom Perch, John Klensin, Shawn Steele, Peter
2338Saint-Andre, Geoffrey Sneddon, Chris Weber, Alex Melnikov, Slim Amamou, SM, Tim Berners-Lee, Yaron
2339Goland, Sam Ruby, Adam Barth, Abdulrahman I. ALGhadir, Aharon Lanin, Thomas Milo, Murray Sargent,
2340Marc Blanchet, and Mykyta Yevstifeyev.</t>
2341</section>
2342
2343<section title="Main Changes Since RFC 3987">
2344  <t>This section describes the main changes since <xref target="RFC3987"></xref>.</t>
2345  <section title="Major restructuring of IRI processing model" anchor="forkChanges">
2346    <t>Major restructuring of IRI processing model to make scheme-specific translation
2347      necessary to handle IDNA requirements and for consistency with web implementations. </t>
2348    <t>Starting with IRI, you want one of:
2349      <list style="hanging">
2350        <t hangText="a"> IRI components (IRI parsed into UTF8 pieces)</t>
2351        <t hangText="b"> URI components (URI parsed into ASCII pieces, encoded correctly) </t>
2352        <t hangText="c"> whole URI  (for passing on to some other system that wants whole URIs) </t>
2353      </list></t>
2354   
2355    <section title="OLD WAY">
2356      <t><list style="numbers">
2357       
2358        <t>Pct-encoding on the whole thing to a URI.
2359          (c1) If you want a (maybe broken) whole URI, you might
2360          stop here.</t>
2361       
2362        <t>Parsing the URI into URI components.
2363          (b1) If you want (maybe broken) URI components, stop here.</t>
2364       
2365        <t> Decode the components (undoing the pct-encoding).
2366          (a) if you want IRI components, stop here.</t>
2367       
2368        <t> reencode:  Either using a different encoding some components
2369          (for domain names, and query components in web pages, which
2370          depends on the component, scheme and context), and otherwise
2371          using pct-encoding.
2372          (b2) if you want (good) URI components, stop here.</t>
2373       
2374        <t> reassemble the reencoded components.
2375          (c2) if you want a (*good*) whole URI stop here.</t>
2376      </list>
2377       
2378      </t>
2379     
2380    </section>
2381   
2382    <section title="NEW WAY">
2383      <t>
2384        <list style="numbers">
2385         
2386          <t> Parse the IRI into IRI components using the generic syntax.
2387            (a) if you want IRI components, stop here.</t>
2388         
2389          <t> Encode each components, using pct-encoding, IDN encoding, or
2390            special query part encoding depending on the component
2391            scheme or context. (b) If you want URI components, stop here.</t>
2392          <t> reassemble the a whole URI from URI components.
2393            (c) if you want a whole URI stop here.</t>
2394        </list></t>
2395    </section>
2396    <section title="Extension of Syntax">
2397      <t>Added the tag range (U+E0000-E0FFF) to the iprivate production.
2398        Some IRIs generated with the new syntax may fail to pass very strict checks
2399        relying on the old syntax. But characters in this range should be extremely infrequent
2400        anyway.</t>
2401    </section>
2402    <section title="More to be added"><t>TODO: There are more main changes that need to be
2403      documented in this section.</t></section>
2404</section>
2405
2406<section title="Change Log">
2407
2408<t>Note to RFC Editor: Please completely remove this section before publication.</t>
2409
2410<section title='Changes after draft-ietf-iri-3987bis-01'>
2411    <t>Changes from draft-ietf-iri-3987bis-01 onwards are available as changesets
2412      in the IETF tools subversion repository at
2413      http://trac.tools.ietf.org/wg/iri/trac/log/draft-ietf-iri-3987bis/draft-ietf-iri-3987bis.xml.</t>
2414</section>
2415 
2416<section title='Changes from draft-duerst-iri-bis-07 to draft-ietf-iri-3987bis-00'>
2417     <t>Changed draft name, date, last paragraph of abstract, and titles in change log, and added this section
2418     in moving from draft-duerst-iri-bis-07 (personal submission) to draft-ietf-iri-3987bis-00 (WG document).</t>
2419</section>
2420
2421<section title="Changes from -06 to -07 of draft-duerst-iri-bis">
2422  <t>Major restructuring of the processing model, see <xref target="forkChanges"></xref>.</t>
2423</section>
2424</section>
2425
2426<section title='Changes from -00 to -01'><t><list style="symbols">
2427  <t>Removed 'mailto:' before mail addresses of authors.</t>
2428  <t>Added "&lt;to be done&gt;" as right side of 'href-strip' rule. Fixed '|' to '/' for
2429    alternatives.</t>
2430</list></t>
2431</section>
2432
2433<section title="Changes from -05 to -06 of draft-duerst-iri-bis-00"><t><list style="symbols">
2434<t>Add HyperText Reference, change abstract, acks and references for it</t>
2435<t>Add Masinter back as another editor.</t>
2436<t>Masinter integrates HRef material from HTML5 spec.</t>
2437<t>Rewrite introduction sections to modernize.</t>
2438</list></t>
2439</section>
2440
2441<section title="Changes from -04 to -05 of draft-duerst-iri-bis">
2442  <t><list style="symbols">
2443    <t>Updated references.</t>
2444    <t>Changed IPR text to pre5378Trust200902.</t></list></t>
2445</section>
2446
2447<section title="Changes from -03 to -04 of draft-duerst-iri-bis">
2448  <t><list style="symbols">
2449    <t>Added explicit abbreviation for LEIRIs.</t>
2450    <t>Mentioned LEIRI references.</t>
2451    <t>Completed text in LEIRI section about tag characters and about specials.</t></list></t>
2452</section>
2453
2454<section title="Changes from -02 to -03 of draft-duerst-iri-bis">
2455  <t><list style="symbols">
2456    <t>Updated some references.</t>
2457    <t>Updated Michel Suginard's coordinates.</t></list></t>
2458</section>
2459
2460<section title="Changes from -01 to -02 of draft-duerst-iri-bis">
2461  <t><list style="symbols">
2462    <t>Added tag range to iprivate (issue private-include-tags-115).</t>
2463    <t>Added Specials (U+FFF0-FFFD) to Legacy Extended IRIs.</t></list></t>
2464</section>
2465<section title="Changes from -00 to -01 of draft-duerst-iri-bis">
2466  <t><list style="symbols">
2467    <t>Changed from "IRIs with Spaces/Controls" to "Legacy Extended IRI"
2468      based on input from the W3C XML Core WG.
2469      Moved the relevant subsections to the back and promoted them to a section.</t>
2470    <t>Added some text re. Legacy Extended IRIs to the security section.</t>
2471    <t>Added a IANA Consideration Section.</t>
2472    <t>Added this Change Log Section.</t>
2473    <t>Added a section about "IRIs with Spaces/Controls" (converting from a Note in RFC 3987).</t></list></t>
2474</section>
2475<section title="Changes from RFC 3987 to -00 of draft-duerst-iri-bis">
2476  <t><list>
2477    <t>Fixed errata (see http://www.rfc-editor.org/cgi-bin/errataSearch.pl?rfc=3987).</t></list></t>
2478</section>
2479</section>
2480</middle>
2481
2482<back>
2483<references title="Normative References">
2484
2485<reference anchor="ASCII">
2486<front>
2487<title>Coded Character Set -- 7-bit American Standard Code for Information
2488Interchange</title>
2489<author>
2490<organization>American National Standards Institute</organization>
2491</author>
2492<date year="1986"/>
2493</front>
2494<seriesInfo name="ANSI" value="X3.4"/>
2495</reference>
2496
2497<reference anchor="ISO10646">
2498<front>
2499<title>ISO/IEC 10646:2003: Information Technology -
2500Universal Multiple-Octet Coded Character Set (UCS)</title>
2501<author>
2502<organization>International Organization for Standardization</organization>
2503</author>
2504<date month="December" year="2003"/>
2505</front>
2506<seriesInfo name="ISO" value="Standard 10646"/>
2507</reference>
2508
2509&rfc2119;
2510&rfc3490;
2511&rfc3491;
2512&rfc3629;
2513&rfc3986;
2514
2515<reference anchor="STD68">
2516<front>
2517<title abbrev="ABNF">Augmented BNF for Syntax Specifications: ABNF</title>
2518<author initials="D." surname="Crocker" fullname="Dave Crocker"><organization/></author>
2519<author initials="P." surname="Overell" fullname="Paul Overell"><organization/></author>
2520<date month="January" year="2008"/></front>
2521<seriesInfo name="STD" value="68"/><seriesInfo name="RFC" value="5234"/>
2522</reference>
2523 
2524&rfc5890;
2525&rfc5891;
2526
2527<reference anchor="UNIV6">
2528<front>
2529<title>The Unicode Standard, Version 6.0.0 (Mountain View, CA, The Unicode Consortium, 2011, ISBN 978-1-936213-01-6)</title>
2530<author><organization>The Unicode Consortium</organization></author>
2531<date year="2010" month="October"/>
2532</front>
2533</reference>
2534
2535<reference anchor="UNI9" target="http://www.unicode.org/reports/tr9/tr9-13.html">
2536<front>
2537<title>The Bidirectional Algorithm</title>
2538<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2539<date year="2004" month="March"/>
2540</front>
2541<seriesInfo name="Unicode Standard Annex" value="#9"/>
2542</reference>
2543
2544<reference anchor="UTR15" target="http://www.unicode.org/unicode/reports/tr15/tr15-23.html">
2545<front>
2546<title>Unicode Normalization Forms</title>
2547<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2548<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2549<date year="2008" month="March"/>
2550</front>
2551<seriesInfo name="Unicode Standard Annex" value="#15"/>
2552</reference>
2553
2554</references>
2555
2556<references title="Informative References">
2557
2558<reference anchor="BidiEx" target="http://www.w3.org/International/iri-edit/BidiExamples">
2559<front>
2560<title>Examples of bidirectional IRIs</title>
2561<author><organization/></author>
2562<date year="" month=""/>
2563</front>
2564</reference>
2565
2566<reference anchor="CharMod" target="http://www.w3.org/TR/charmod-resid">
2567<front>
2568<title>Character Model for the World Wide Web: Resource Identifiers</title>
2569<author initials="M." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2570<author initials="F." surname="Yergeau" fullname="Francois Yergeau"><organization/></author>
2571<author initials="R." surname="Ishida" fullname="Richard Ishida"><organization/></author>
2572<author initials="M." surname="Wolf" fullname="Misha Wolf"><organization/></author>
2573<author initials="T." surname="Texin" fullname="Tex Texin"><organization/></author>
2574<date year="2004" month="November" day="25"/>
2575</front>
2576<seriesInfo name="World Wide Web Consortium" value="Candidate Recommendation"/>
2577</reference>
2578
2579<reference anchor="Duerst97" target="http://www.ifi.unizh.ch/mml/mduerst/papers/PDF/IUC11-UTF-8.pdf">
2580<front>
2581<title>The Properties and Promises of UTF-8</title>
2582<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2583<date year="1997" month="September"/>
2584</front>
2585<seriesInfo name="Proc. 11th International Unicode Conference, San Jose" value=""/>
2586</reference>
2587
2588<reference anchor="Gettys" target="http://www.w3.org/DesignIssues/ModelConsequences">
2589<front>
2590<title>URI Model Consequences</title>
2591<author initials="J." surname="Gettys" fullname="Jim Gettys"><organization/></author>
2592<date month="" year=""/>
2593</front>
2594</reference>
2595
2596<reference anchor="HTML4" target="http://www.w3.org/TR/html401/appendix/notes.html#h-B.2">
2597<front>
2598<title>HTML 4.01 Specification</title>
2599<author initials="D." surname="Raggett" fullname="Dave Raggett"><organization/></author>
2600<author initials="A." surname="Le Hors" fullname="Arnaud Le Hors"><organization/></author>
2601<author initials="I." surname="Jacobs" fullname="Ian Jacobs"><organization/></author>
2602<date year="1999" month="December" day="24"/>
2603</front>
2604<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2605</reference>
2606
2607<reference anchor="LEIRI" target="http://www.w3.org/TR/leiri/">
2608<front>
2609<title>Legacy extended IRIs for XML resource identification</title>
2610<author initials="H." surname="Thompson" fullname="Henry Thompson"><organization/></author>
2611<author initials="R." surname="Tobin"    fullname="Richard Tobin"><organization/></author>
2612<author initials="N." surname="Walsh" fullname="Norman Walsh"><organization/></author>
2613  <date year="2008" month="November" day="3"/>
2614
2615</front>
2616<seriesInfo name="World Wide Web Consortium" value="Note"/>
2617</reference>
2618
2619
2620&rfc2045;
2621&rfc2130;
2622&rfc2141;
2623&rfc2192;
2624&rfc2277;
2625&rfc2368;
2626&rfc2384;
2627&rfc2396;
2628&rfc2397;
2629&rfc2616;
2630&rfc1738;
2631&rfc2640;
2632&rfc3987;
2633<reference anchor='RFC4395bis'>
2634  <front>
2635    <title>Guidelines and Registration Procedures for New URI/IRI Schemes</title>
2636    <author initials='T.' surname='Hansen' fullname="Tony Hansen"><organization/></author>
2637    <author initials='T.' surname='Hardie' fullname="Ted Hardie"><organization/></author>
2638    <author initials='L.' surname='Masinter' fullname="Larry Masinter"><organization/></author>
2639    <date year="2010" month='September' day="30"/>
2640    <workgroup>IRI</workgroup>
2641  </front>
2642  <seriesInfo name="Internet-Draft" value="draft-hansen-iri-4395bis-irireg-00"/>
2643</reference>
2644 
2645 
2646<reference anchor="UNIXML" target="http://www.w3.org/TR/unicode-xml/">
2647<front>
2648<title>Unicode in XML and other Markup Languages</title>
2649<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2650<author initials="A." surname="Freytag" fullname="Asmus Freytag"><organization/></author>
2651<date year="2003" month="June" day="18"/>
2652</front>
2653<seriesInfo name="Unicode Technical Report" value="#20"/>
2654<seriesInfo name="World Wide Web Consortium" value="Note"/>
2655</reference>
2656 
2657<reference anchor="UTR36" target="http://unicode.org/reports/tr36/">
2658<front>
2659<title>Unicode Security Considerations</title>
2660<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2661<author initials="M." surname="Suignard" fullname="Michel Suignard"><organization/></author>
2662<date year="2010" month="August" day="4"/>
2663</front>
2664<seriesInfo name="Unicode Technical Report" value="#36"/>
2665</reference>
2666
2667<reference anchor="XLink" target="http://www.w3.org/TR/xlink/#link-locators">
2668<front>
2669<title>XML Linking Language (XLink) Version 1.0</title>
2670<author initials="S." surname="DeRose" fullname="Steve DeRose"><organization/></author>
2671<author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2672<author initials="D." surname="Orchard" fullname="David Orchard"><organization/></author>
2673<date year="2001" month="June" day="27"/>
2674</front>
2675<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2676</reference>
2677
2678<reference anchor="XML1" target="http://www.w3.org/TR/REC-xml">
2679  <front>
2680    <title>Extensible Markup Language (XML) 1.0 (Forth Edition)</title>
2681    <author initials="T." surname="Bray" fullname="Tim Bray"><organization/></author>
2682    <author initials="J." surname="Paoli" fullname="Jean Paoli"><organization/></author>
2683    <author initials="C.M." surname="Sperberg-McQueen" fullname="C. M. Sperberg-McQueen">
2684      <organization/></author>
2685    <author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2686    <author initials="F." surname="Yergeau" fullname="Francois Yergeau"><organization/></author>
2687    <date day="16" month="August" year="2006"/>
2688  </front>
2689  <seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2690</reference>
2691
2692<reference anchor="XMLNamespace" target="http://www.w3.org/TR/REC-xml-names">
2693  <front>
2694    <title>Namespaces in XML (Second Edition)</title>
2695    <author initials="T." surname="Bray" fullname="Tim Bray"><organization/></author>
2696    <author initials="D." surname="Hollander" fullname="Dave Hollander"><organization/></author>
2697    <author initials="A." surname="Layman" fullname="Andrew Layman"><organization/></author>
2698    <author initials="R." surname="Tobin" fullname="Richard Tobin"><organization></organization></author>
2699    <date day="16" month="August" year="2006"/>
2700  </front>
2701  <seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2702</reference>
2703
2704<reference anchor="XMLSchema" target="http://www.w3.org/TR/xmlschema-2/#anyURI">
2705<front>
2706<title>XML Schema Part 2: Datatypes</title>
2707<author initials="P." surname="Biron" fullname="Paul Biron"><organization/></author>
2708<author initials="A." surname="Malhotra" fullname="Ashok Malhotra"><organization/></author>
2709<date year="2001" month="May" day="2"/>
2710</front>
2711<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2712</reference>
2713
2714<reference anchor="XPointer" target="http://www.w3.org/TR/xptr-framework/#escaping">
2715<front>
2716<title>XPointer Framework</title>
2717<author initials="P." surname="Grosso" fullname="Paul Grosso"><organization/></author>
2718<author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2719<author initials="J." surname="Marsh" fullname="Jonathan Marsh"><organization/></author>
2720<author initials="N." surname="Walsh" fullname="Norman Walsh"><organization/></author>
2721<date year="2003" month="March" day="25"/>
2722</front>
2723<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2724</reference>
2725
2726<reference anchor="HTML5" target="http://www.w3.org/TR/2009/WD-html5-20090423/">
2727<front>
2728<title>A vocabulary and associated APIs for HTML and XHTML</title>
2729<author initials="I." surname="Hickson" fullname="Ian Hickson"><organization>Google, Inc.</organization></author>
2730<author initials="D." surname="Hyatt" fullname="David Hyatt"><organization>Apple, Inc.</organization></author>
2731<date year="2009"  month="April" day="23"/>
2732</front>
2733<seriesInfo name="World Wide Web Consortium" value="Working Draft"/>
2734</reference>
2735
2736</references>
2737
2738</back>
2739</rfc>
Note: See TracBrowser for help on using the repository browser.