source: draft-ietf-iri-3987bis/draft-ietf-iri-3987bis.xml @ 40

Last change on this file since 40 was 40, checked in by duerst@…, 9 years ago

Added a pointer to change history to the note at the end of the Abstract.

  • Property svn:executable set to *
File size: 124.0 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-04" 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="14"/>
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.</t>
448
449<!--
450for line length measuring in artwork (max 72 chars, three chars at start):
451      1         2         3         4         5         6         7
452456789012345678901234567890123456789012345678901234567890123456789012
453-->
454<figure>
455<preamble>The following rules are different from those in <xref target="RFC3986"/>:</preamble>
456<artwork>
457IRI            = scheme ":" ihier-part [ "?" iquery ]
458                 [ "#" ifragment ]
459
460ihier-part     = "//" iauthority ipath-abempty
461               / ipath-absolute
462               / ipath-rootless
463               / ipath-empty
464
465IRI-reference  = IRI / irelative-ref
466
467absolute-IRI   = scheme ":" ihier-part [ "?" iquery ]
468
469irelative-ref  = irelative-part [ "?" iquery ] [ "#" ifragment ]
470
471irelative-part = "//" iauthority ipath-abempty
472               / ipath-absolute
473               / ipath-noscheme
474               / ipath-empty
475
476iauthority     = [ iuserinfo "@" ] ihost [ ":" port ]
477iuserinfo      = *( iunreserved / pct-form / sub-delims / ":" )
478ihost          = IP-literal / IPv4address / ireg-name
479
480pct-form       = pct-encoded
481
482ireg-name      = *( iunreserved / sub-delims )
483
484ipath          = ipath-abempty   ; begins with "/" or is empty
485               / ipath-absolute  ; begins with "/" but not "//"
486               / ipath-noscheme  ; begins with a non-colon segment
487               / ipath-rootless  ; begins with a segment
488               / ipath-empty     ; zero characters
489
490ipath-abempty  = *( path-sep isegment )
491ipath-absolute = path-sep [ isegment-nz *( path-sep isegment ) ]
492ipath-noscheme = isegment-nz-nc *( path-sep isegment )
493ipath-rootless = isegment-nz *( path-sep isegment )
494ipath-empty    = 0&lt;ipchar&gt;
495path-sep       = "/"
496
497isegment       = *ipchar
498isegment-nz    = 1*ipchar
499isegment-nz-nc = 1*( iunreserved / pct-form / sub-delims
500                     / "@" )
501               ; non-zero-length segment without any colon ":"                     
502
503ipchar         = iunreserved / pct-form / sub-delims / ":"
504               / "@"
505 
506iquery         = *( ipchar / iprivate / "/" / "?" )
507
508ifragment      = *( ipchar / "/" / "?" / "#" )
509
510iunreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~" / ucschar
511
512ucschar        = %xA0-D7FF / %xF900-FDCF / %xFDF0-FFEF
513               / %x10000-1FFFD / %x20000-2FFFD / %x30000-3FFFD
514               / %x40000-4FFFD / %x50000-5FFFD / %x60000-6FFFD
515               / %x70000-7FFFD / %x80000-8FFFD / %x90000-9FFFD
516               / %xA0000-AFFFD / %xB0000-BFFFD / %xC0000-CFFFD
517               / %xD0000-DFFFD / %xE1000-EFFFD
518
519iprivate       = %xE000-F8FF / %xE0000-E0FFF / %xF0000-FFFFD
520               / %x100000-10FFFD
521</artwork>
522</figure>
523
524<t>Some productions are ambiguous. The "first-match-wins" (a.k.a. "greedy")
525algorithm applies. For details, see <xref target="RFC3986"/>.</t>
526
527<figure>
528<preamble>The following rules are the same as those in <xref target="RFC3986"/>:</preamble>
529<artwork>
530scheme         = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
531 
532port           = *DIGIT
533 
534IP-literal     = "[" ( IPv6address / IPvFuture  ) "]"
535 
536IPvFuture      = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" )
537 
538IPv6address    =                            6( h16 ":" ) ls32
539               /                       "::" 5( h16 ":" ) ls32
540               / [               h16 ] "::" 4( h16 ":" ) ls32
541               / [ *1( h16 ":" ) h16 ] "::" 3( h16 ":" ) ls32
542               / [ *2( h16 ":" ) h16 ] "::" 2( h16 ":" ) ls32
543               / [ *3( h16 ":" ) h16 ] "::"    h16 ":"   ls32
544               / [ *4( h16 ":" ) h16 ] "::"              ls32
545               / [ *5( h16 ":" ) h16 ] "::"              h16
546               / [ *6( h16 ":" ) h16 ] "::"
547               
548h16            = 1*4HEXDIG
549ls32           = ( h16 ":" h16 ) / IPv4address
550
551IPv4address    = dec-octet "." dec-octet "." dec-octet "." dec-octet
552
553dec-octet      = DIGIT                 ; 0-9
554               / %x31-39 DIGIT         ; 10-99
555               / "1" 2DIGIT            ; 100-199
556               / "2" %x30-34 DIGIT     ; 200-249
557               / "25" %x30-35          ; 250-255
558           
559pct-encoded    = "%" HEXDIG HEXDIG
560
561unreserved     = ALPHA / DIGIT / "-" / "." / "_" / "~"
562reserved       = gen-delims / sub-delims
563gen-delims     = ":" / "/" / "?" / "#" / "[" / "]" / "@"
564sub-delims     = "!" / "$" / "&amp;" / "'" / "(" / ")"
565               / "*" / "+" / "," / ";" / "="
566</artwork></figure>
567
568<t>This syntax does not support IPv6 scoped addressing zone identifiers.</t>
569
570</section> <!-- abnf -->
571
572</section> <!-- syntax -->
573
574<section title="Processing IRIs and related protocol elements" anchor="processing">
575
576<t>IRIs are meant to replace URIs in identifying resources within new
577versions of protocols, formats, and software components that use a
578UCS-based character repertoire.  Protocols and components may use and
579process IRIs directly. However, there are still numerous systems and
580protocols which only accept URIs or components of parsed URIs; that is,
581they only accept sequences of characters within the subset of US-ASCII
582characters allowed in URIs. </t>
583
584<t>This section defines specific processing steps for IRI consumers
585which establish the relationship between the string given and the
586interpreted derivatives. These
587processing steps apply to both IRIs and IRI references (i.e., absolute
588or relative forms); for IRIs, some steps are scheme specific. </t>
589
590<section title="Converting to UCS" anchor="ucsconv"> 
591 
592<t>Input that is already in a Unicode form (i.e., a sequence of Unicode
593 characters or an octet-stream representing a Unicode-based character
594 encoding such as UTF-8 or UTF-16) should be left as is and not
595 normalized (see (see <xref target="normalization"/>).</t>
596
597  <t>An IRI or IRI reference is a sequence of characters from the UCS.
598    For IRIs that are not already in a Unicode form
599    (as when written on paper, read aloud, or represented in a text stream
600    using a legacy character encoding), convert the IRI to Unicode.
601    Note that some character encodings or transcriptions can be converted
602    to or represented by more than one sequence of Unicode characters.
603    Ideally the resulting IRI would use a normalized form,
604    such as Unicode Normalization Form C <xref target="UTR15"/>
605    (see <xref target='ladder'/> Normalization and Comparison),
606    since that ensures a stable, consistent representation
607    that is most likely to produce the intended results.
608    Implementers and users are cautioned that, while denormalized character sequences are valid,
609    they might be difficult for other users or processes to reproduce
610    and might lead to unexpected results.
611  </t>
612
613<t> In other cases (written on paper, read aloud, or otherwise
614 represented independent of any character encoding) represent the IRI
615 as a sequence of characters from the UCS normalized according to
616 Unicode Normalization Form C (NFC, <xref target="UTR15"/>).</t>
617</section> <!-- ucsconv -->
618
619<section title="Parse the IRI into IRI components">
620
621<t>Parse the IRI, either as a relative reference (no scheme)
622or using scheme specific processing (according to the scheme
623given); the result resulting in a set of parsed IRI components.
624(NOTE: FIX BEFORE RELEASE: INTENT IS THAT ALL IRI SCHEMES
625THAT USE GENERIC SYNTAX AND ALLOW NON-ASCII AUTHORITY CAN
626ONLY USE AUTHORITY FOR NAMES THAT FOLLOW PUNICODE.)
627 </t>
628
629<t>NOTE: The result of parsing into components will correspond result
630in a correspondence of subtrings of the IRI according to the part
631matched.  For example, in <xref target="HTML5"/>, the protocol
632components of interest are SCHEME (scheme), HOST (ireg-name), PORT
633(port), the PATH (ipath after the initial "/"), QUERY (iquery),
634FRAGMENT (ifragment), and AUTHORITY (iauthority).
635</t>
636
637<t>Subsequent processing rules are sometimes used to define other
638syntactic components. For example, <xref target="HTML5"/> defines APIs
639for IRI processing; in these APIs:
640
641<list style="hanging">
642<t hangText="HOSTSPECIFIC"> the substring that follows
643the substring matched by the iauthority production, or the whole
644string if the iauthority production wasn't matched.</t>
645<t hangText="HOSTPORT"> if there is a scheme component and a port
646component and the port given by the port component is different than
647the default port defined for the protocol given by the scheme
648component, then HOSTPORT is the substring that starts with the
649substring matched by the host production and ends with the substring
650matched by the port production, and includes the colon in between the
651two. Otherwise, it is the same as the host component.
652</t>
653</list>
654</t>
655</section> <!-- parse -->
656
657<section title="General percent-encoding of IRI components" anchor="compmapping">
658   
659<t>For most IRI components, it is possible to map the IRI component
660to an equivalent URI component by percent-encoding those characters
661not allowed in URIs. Previous processing steps will have removed
662some characters, and the interpretation of reserved characters will
663have already been done (with the syntactic reserved characters outside
664of the IRI component). This mapping is defined for all sequences
665of Unicode characters, whether or not they are valid for the component
666in question. </t>
667   
668<t>For each character which is not allowed in a valid URI (NOTE: WHAT
669IS THE RIGHT REFERENCE HERE), apply the following steps. </t>
670
671<t><list style="hanging">
672
673<t hangText="Convert to UTF-8">Convert the character to a sequence of
674  one or more octets using UTF-8 <xref target="RFC3629"/>.</t>
675
676<t hangText="Percent encode">Convert each octet of this sequence to %HH,
677   where HH is the hexadecimal notation of the octet value. The
678   hexadecimal notation SHOULD use uppercase letters. (This is the
679   general URI percent-encoding mechanism in Section 2.1 of <xref
680   target="RFC3986"/>.)</t>
681   
682</list></t>
683
684<t>Note that the mapping is an identity transformation for parsed URI
685components of valid URIs, and is idempotent: applying the mapping a
686second time will not change anything.</t>
687</section> <!-- general conversion -->
688
689<section title="Mapping ireg-name" anchor="dnsmapping">
690
691<t>Schemes that allow non-ASCII based characters
692in the reg-name (ireg-name) position MUST convert the ireg-name
693component of an IRI as follows:</t>
694
695<t>Replace the ireg-name part of the IRI by the part converted using
696the ToASCII operation specified in Section 4.1 of <xref
697target="RFC3490"/> on each dot-separated label, and by using U+002E
698(FULL STOP) as a label separator, with the flag UseSTD3ASCIIRules set
699to FALSE, and with the flag AllowUnassigned set to FALSE.
700The ToASCII operation may
701fail, but this would mean that the IRI cannot be resolved.
702In such cases, if the domain name conversion fails, then the
703entire IRI conversion fails. Processors that have no mechanism for
704signalling a failure MAY instead substitute an otherwise
705invalid host name, although such processing SHOULD be avoided.
706 </t>
707
708<t>For example, the IRI
709<vspace/>"http://r&amp;#xE9;sum&amp;#xE9;.example.org"<vspace/> MAY be
710converted to <vspace/>"http://xn--rsum-bad.example.org"<vspace/>;
711conversion to percent-encoded form, e.g.,
712 <vspace/>"http://r%C3%A9sum%C3%A9.example.org", MUST NOT be performed. </t>
713
714<t><list style="hanging"> 
715
716<t hangText="Note:">Domain Names may appear in parts of an IRI other
717than the ireg-name part.  It is the responsibility of scheme-specific
718implementations (if the Internationalized Domain Name is part of the
719scheme syntax) or of server-side implementations (if the
720Internationalized Domain Name is part of 'iquery') to apply the
721necessary conversions at the appropriate point. Example: Trying to
722validate the Web page at<vspace/>
723http://r&amp;#xE9;sum&amp;#xE9;.example.org would lead to an IRI of
724<vspace/>http://validator.w3.org/check?uri=http%3A%2F%2Fr&amp;#xE9;sum&amp;#xE9;.<vspace/>example.org,
725which would convert to a URI
726of<vspace/>http://validator.w3.org/check?uri=http%3A%2F%2Fr%C3%A9sum%C3%A9.<vspace/>example.org.
727The server-side implementation is responsible for making the
728necessary conversions to be able to retrieve the Web page.</t>
729
730<t hangText="Note:">In this process, characters allowed in URI
731references and existing percent-encoded sequences are not encoded further.
732(This mapping is similar to, but different from, the encoding applied
733when arbitrary content is included in some part of a URI.)
734
735For example, an IRI of
736<vspace/>"http://www.example.org/red%09ros&amp;#xE9;#red"
737(in XML notation) is converted to
738<vspace/>"http://www.example.org/red%09ros%C3%A9#red", not to
739something like
740<vspace/>"http%3A%2F%2Fwww.example.org%2Fred%2509ros%C3%A9%23red".
741((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?))
742
743</t>
744
745</list></t>
746</section> <!-- dnsmapping -->
747
748<section title="Mapping query components" anchor="querymapping">
749
750<t>((NOTE: SEE ISSUES LIST))
751
752For compatibility with existing deployed HTTP infrastructure,
753the following special case applies for schemes "http" and "https"
754and IRIs whose origin has a document charset other than one which
755is UCS-based (e.g., UTF-8 or UTF-16). In such a case, the "query"
756component of an IRI is mapped into a URI by using the document
757charset rather than UTF-8 as the binary representation before
758pct-encoding. This mapping is not applied for any other scheme
759or component.</t>
760
761</section> <!-- querymapping -->
762
763<section title="Mapping IRIs to URIs" anchor="mapping">
764
765<t>The canonical mapping from a IRI to URI is defined by applying the
766mapping above (from IRI to URI components) and then reassembling a URI
767from the parsed URI components using the original punctuation that
768delimited the IRI components. </t>
769
770</section> <!-- mapping -->
771
772<section title="Converting URIs to IRIs" anchor="URItoIRI">
773
774<t>In some situations, for presentation and further processing,
775it is desirable to convert a URI into an equivalent IRI in which
776natural characters are represented directly rather than
777percent encoded. Of course, every URI is already an IRI in
778its own right without any conversion, and in general there
779This section gives one such procedure for this conversion.
780</t>
781
782<t>
783The conversion described in this section, if given a valid URI, will
784result in an IRI that maps back to the URI used as an input for the
785conversion (except for potential case differences in percent-encoding
786and for potential percent-encoded unreserved characters).
787
788However, the IRI resulting from this conversion may differ
789from the original IRI (if there ever was one).</t> 
790
791<t>URI-to-IRI conversion removes percent-encodings, but not all
792percent-encodings can be eliminated. There are several reasons for
793this:</t>
794
795<t><list style="hanging">
796
797<t hangText="1.">Some percent-encodings are necessary to distinguish
798    percent-encoded and unencoded uses of reserved characters.</t>
799
800<t hangText="2.">Some percent-encodings cannot be interpreted as sequences
801    of UTF-8 octets.<vspace blankLines="1"/>
802    (Note: The octet patterns of UTF-8 are highly regular.
803    Therefore, there is a very high probability, but no guarantee,
804    that percent-encodings that can be interpreted as sequences of UTF-8
805    octets actually originated from UTF-8. For a detailed discussion,
806    see <xref target="Duerst97"/>.)</t>
807
808<t hangText="3.">The conversion may result in a character that is not
809    appropriate in an IRI. See <xref target="abnf"/>, <xref target="visual"/>,
810      and <xref target="limitations"/> for further details.</t>
811
812<t hangText="4.">IRI to URI conversion has different rules for
813    dealing with domain names and query parameters.</t>
814
815</list></t>
816
817<t>Conversion from a URI to an IRI MAY be done by using the following
818steps:
819
820<list style="hanging">
821<t hangText="1.">Represent the URI as a sequence of octets in
822       US-ASCII.</t>
823
824<t hangText="2.">Convert all percent-encodings ("%" followed by two
825      hexadecimal digits) to the corresponding octets, except those
826      corresponding to "%", characters in "reserved", and characters
827      in US-ASCII not allowed in URIs.</t> 
828
829<t hangText="3.">Re-percent-encode any octet produced in step 2 that
830      is not part of a strictly legal UTF-8 octet sequence.</t>
831
832
833<t hangText="4.">Re-percent-encode all octets produced in step 3 that
834      in UTF-8 represent characters that are not appropriate according
835      to <xref target="abnf"/>, <xref target="visual"/>, and <xref
836      target="limitations"/>.</t> 
837
838<t hangText="5.">Interpret the resulting octet sequence as a sequence
839      of characters encoded in UTF-8.</t>
840
841<t hangText="6.">URIs known to contain domain names in the reg-name
842      component SHOULD convert punycode-encoded domain name labels to
843      the corresponding characters using the ToUnicode procedure. </t>
844</list></t>
845
846<t>This procedure will convert as many percent-encoded characters as
847possible to characters in an IRI. Because there are some choices when
848step 4 is applied (see <xref target="limitations"/>), results may
849vary.</t>
850
851<t>Conversions from URIs to IRIs MUST NOT use any character
852encoding other than UTF-8 in steps 3 and 4, even if it might be
853possible to guess from the context that another character encoding
854than UTF-8 was used in the URI.  For example, the URI
855"http://www.example.org/r%E9sum%E9.html" might with some guessing be
856interpreted to contain two e-acute characters encoded as
857iso-8859-1. It must not be converted to an IRI containing these
858e-acute characters. Otherwise, in the future the IRI will be mapped to
859"http://www.example.org/r%C3%A9sum%C3%A9.html", which is a different
860URI from "http://www.example.org/r%E9sum%E9.html".</t>
861
862<section title="Examples">
863
864<t>This section shows various examples of converting URIs to IRIs.
865Each example shows the result after each of the steps 1 through 6 is
866applied. XML Notation is used for the final result.  Octets are
867denoted by "&lt;" followed by two hexadecimal digits followed by
868"&gt;".</t>
869
870<t>The following example contains the sequence "%C3%BC", which is a
871strictly legal UTF-8 sequence, and which is converted into the actual
872character U+00FC, LATIN SMALL LETTER U WITH DIAERESIS (also known as
873u-umlaut).
874
875<list style="hanging">
876<t hangText="1.">http://www.example.org/D%C3%BCrst</t>
877<t hangText="2.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
878<t hangText="3.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
879<t hangText="4.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
880<t hangText="5.">http://www.example.org/D&amp;#xFC;rst</t>
881<t hangText="6.">http://www.example.org/D&amp;#xFC;rst</t>
882</list>
883</t>
884
885<t>The following example contains the sequence "%FC", which might
886represent U+00FC, LATIN SMALL LETTER U WITH DIAERESIS, in
887the<vspace/>iso-8859-1 character encoding.  (It might represent other
888characters in other character encodings. For example, the octet
889&lt;fc&gt; in iso-8859-5 represents U+045C, CYRILLIC SMALL LETTER
890KJE.)  Because &lt;fc&gt; is not part of a strictly legal UTF-8
891sequence, it is re-percent-encoded in step 3.
892
893
894<list style="hanging">
895<t hangText="1.">http://www.example.org/D%FCrst</t>
896<t hangText="2.">http://www.example.org/D&lt;fc&gt;rst</t>
897<t hangText="3.">http://www.example.org/D%FCrst</t>
898<t hangText="4.">http://www.example.org/D%FCrst</t>
899<t hangText="5.">http://www.example.org/D%FCrst</t>
900<t hangText="6.">http://www.example.org/D%FCrst</t>
901</list>
902</t>
903
904<t>The following example contains "%e2%80%ae", which is the percent-encoded<vspace/>UTF-8
905character encoding of U+202E, RIGHT-TO-LEFT OVERRIDE. <xref target="visual"/>
906forbids the direct use of this character in an IRI. Therefore, the
907corresponding octets are re-percent-encoded in step 4. This example shows
908that the case (upper- or lowercase) of letters used in percent-encodings may not be preserved.
909The example also contains a punycode-encoded domain name label (xn--99zt52a),
910which is not converted.
911
912<list style="hanging">
913<t hangText="1.">http://xn--99zt52a.example.org/%e2%80%ae</t>
914<t hangText="2.">http://xn--99zt52a.example.org/&lt;e2&gt;&lt;80&gt;&lt;ae&gt;</t>
915<t hangText="3.">http://xn--99zt52a.example.org/&lt;e2&gt;&lt;80&gt;&lt;ae&gt;</t>
916<t hangText="4.">http://xn--99zt52a.example.org/%E2%80%AE</t>
917<t hangText="5.">http://xn--99zt52a.example.org/%E2%80%AE</t>
918<t hangText="6.">http://&amp;#x7D0D;&amp;#x8C46;.example.org/%E2%80%AE</t>
919</list></t>
920
921<t>Note that the label "xn--99zt52a" is converted to U+7D0D U+8C46
922(Japanese Natto). ((EDITOR NOTE: There is some inconsistency in this note.))</t>
923
924</section> <!-- examples -->
925</section> <!-- URItoIRI -->
926</section> <!-- processing -->
927<section title="Bidirectional IRIs for Right-to-Left Languages" anchor="Bidi">
928
929<t>Some UCS characters, such as those used in the Arabic and Hebrew
930scripts, have an inherent right-to-left (rtl) writing direction. IRIs
931containing these characters (called bidirectional IRIs or Bidi IRIs)
932require additional attention because of the non-trivial relation
933between logical representation (used for digital representation and
934for reading/spelling) and visual representation (used for
935display/printing).</t>
936
937<t>Because of the complex interaction between the logical representation,
938the visual representation, and the syntax of a Bidi IRI, a balance is
939needed between various requirements.
940The main requirements are<list style="hanging">
941<t hangText="1.">user-predictable conversion between visual and
942    logical representation;</t>
943<t hangText="2.">the ability to include a wide range of characters
944    in various parts of the IRI; and</t>
945<t hangText="3.">minor or no changes or restrictions for
946      implementations.</t>
947</list></t>
948
949<section title="Logical Storage and Visual Presentation" anchor="visual">
950
951<t>When stored or transmitted in digital representation, bidirectional
952IRIs MUST be in full logical order and MUST conform to the IRI syntax
953rules (which includes the rules relevant to their scheme). This
954ensures that bidirectional IRIs can be processed in the same way as
955other IRIs.</t> <t>Bidirectional IRIs MUST be rendered by using the
956Unicode Bidirectional Algorithm <xref target="UNIV6"/>, <xref
957target="UNI9"/>.  Bidirectional IRIs MUST be rendered in the same way
958as they would be if they were in a left-to-right embedding; i.e., as
959if they were preceded by U+202A, LEFT-TO-RIGHT EMBEDDING (LRE), and
960followed by U+202C, POP DIRECTIONAL FORMATTING (PDF).  Setting the
961embedding direction can also be done in a higher-level protocol (e.g.,
962the dir='ltr' attribute in HTML).</t> 
963
964<t>There is no requirement to use the above embedding if the display
965is still the same without the embedding. For example, a bidirectional
966IRI in a text with left-to-right base directionality (such as used for
967English or Cyrillic) that is preceded and followed by whitespace and
968strong left-to-right characters does not need an embedding.  Also, a
969bidirectional relative IRI reference that only contains strong
970right-to-left characters and weak characters and that starts and ends
971with a strong right-to-left character and appears in a text with
972right-to-left base directionality (such as used for Arabic or Hebrew)
973and is preceded and followed by whitespace and strong characters does
974not need an embedding.</t>
975
976<t>In some other cases, using U+200E, LEFT-TO-RIGHT MARK (LRM), may be
977sufficient to force the correct display behavior.  However, the
978details of the Unicode Bidirectional algorithm are not always easy to
979understand. Implementers are strongly advised to err on the side of
980caution and to use embedding in all cases where they are not
981completely sure that the display behavior is unaffected without the
982embedding.</t>
983
984<t>The Unicode Bidirectional Algorithm (<xref target="UNI9"/>, section
9854.3) permits higher-level protocols to influence bidirectional
986rendering. Such changes by higher-level protocols MUST NOT be used if
987they change the rendering of IRIs.</t> 
988
989<t>The bidirectional formatting characters that may be used before or
990after the IRI to ensure correct display are not themselves part of the
991IRI.  IRIs MUST NOT contain bidirectional formatting characters (LRM,
992RLM, LRE, RLE, LRO, RLO, and PDF). They affect the visual rendering of
993the IRI but do not appear themselves. It would therefore not be
994possible to input an IRI with such characters correctly.</t>
995
996</section> <!-- visual -->
997<section title="Bidi IRI Structure" anchor="bidi-structure">
998
999<t>The Unicode Bidirectional Algorithm is designed mainly for running
1000text.  To make sure that it does not affect the rendering of
1001bidirectional IRIs too much, some restrictions on bidirectional IRIs
1002are necessary. These restrictions are given in terms of delimiters
1003(structural characters, mostly punctuation such as "@", ".", ":",
1004and<vspace/>"/") and components (usually consisting mostly of letters
1005and digits).</t>
1006
1007<t>The following syntax rules from <xref target="abnf"/> correspond to
1008components for the purpose of Bidi behavior: iuserinfo, ireg-name,
1009isegment, isegment-nz, isegment-nz-nc, ireg-name, iquery, and
1010ifragment.</t>
1011
1012<t>Specifications that define the syntax of any of the above
1013components MAY divide them further and define smaller parts to be
1014components according to this document. As an example, the restrictions
1015of <xref target="RFC3490"/> on bidirectional domain names correspond
1016to treating each label of a domain name as a component for schemes
1017with ireg-name as a domain name.  Even where the components are not
1018defined formally, it may be helpful to think about some syntax in
1019terms of components and to apply the relevant restrictions.  For
1020example, for the usual name/value syntax in query parts, it is
1021convenient to treat each name and each value as a component. As
1022another example, the extensions in a resource name can be treated as
1023separate components.</t>
1024
1025<t>For each component, the following restrictions apply:</t>
1026<t>
1027<list style="hanging">
1028
1029<t hangText="1.">A component SHOULD NOT use both right-to-left and
1030  left-to-right characters.</t>
1031
1032<t hangText="2.">A component using right-to-left characters SHOULD
1033  start and end with right-to-left characters.</t>
1034
1035</list></t>
1036
1037<t>The above restrictions are given as "SHOULD"s, rather than as
1038"MUST"s.  For IRIs that are never presented visually, they are not
1039relevant.  However, for IRIs in general, they are very important to
1040ensure consistent conversion between visual presentation and logical
1041representation, in both directions.</t>
1042
1043<t><list style="hanging">
1044
1045<t hangText="Note:">In some components, the above restrictions may
1046  actually be strictly enforced.  For example, <xref
1047  target="RFC3490"></xref> requires that these restrictions apply to
1048  the labels of a host name for those schemes where ireg-name is a
1049  host name.  In some other components (for example, path components)
1050  following these restrictions may not be too difficult.  For other
1051  components, such as parts of the query part, it may be very
1052  difficult to enforce the restrictions because the values of query
1053  parameters may be arbitrary character sequences.</t>
1054
1055</list></t>
1056
1057<t>If the above restrictions cannot be satisfied otherwise, the
1058affected component can always be mapped to URI notation as described
1059in <xref target="compmapping"/>. Please note that the whole component
1060has to be mapped (see also Example 9 below).</t>
1061
1062</section> <!-- bidi-structure -->
1063
1064<section title="Input of Bidi IRIs" anchor="bidiInput">
1065
1066<t>Bidi input methods MUST generate Bidi IRIs in logical order while
1067rendering them according to <xref target="visual"/>.  During input,
1068rendering SHOULD be updated after every new character is input to
1069avoid end-user confusion.</t>
1070
1071</section> <!-- bidiInput -->
1072
1073<section title="Examples">
1074
1075<t>This section gives examples of bidirectional IRIs, in Bidi
1076Notation.  It shows legal IRIs with the relationship between logical
1077and visual representation and explains how certain phenomena in this
1078relationship may look strange to somebody not familiar with
1079bidirectional behavior, but familiar to users of Arabic and Hebrew. It
1080also shows what happens if the restrictions given in <xref
1081target="bidi-structure"/> are not followed. The examples below can be
1082seen at <xref target="BidiEx"/>, in Arabic, Hebrew, and Bidi Notation
1083variants.</t>
1084
1085<t>To read the bidi text in the examples, read the visual
1086representation from left to right until you encounter a block of rtl
1087text. Read the rtl block (including slashes and other special
1088characters) from right to left, then continue at the next unread ltr
1089character.</t>
1090
1091<t>Example 1: A single component with rtl characters is inverted:
1092<vspace/>Logical representation:
1093"http://ab.CDEFGH.ij/kl/mn/op.html"<vspace/>Visual representation:
1094"http://ab.HGFEDC.ij/kl/mn/op.html"<vspace/> Components can be read
1095one by one, and each component can be read in its natural
1096direction.</t>
1097
1098<t>Example 2: More than one consecutive component with rtl characters
1099is inverted as a whole: <vspace/>Logical representation:
1100"http://ab.CDE.FGH/ij/kl/mn/op.html"<vspace/>Visual representation:
1101"http://ab.HGF.EDC/ij/kl/mn/op.html"<vspace/> A sequence of rtl
1102components is read rtl, in the same way as a sequence of rtl words is
1103read rtl in a bidi text.</t>
1104
1105<t>Example 3: All components of an IRI (except for the scheme) are
1106rtl.  All rtl components are inverted overall: <vspace/>Logical
1107representation:
1108"http://AB.CD.EF/GH/IJ/KL?MN=OP;QR=ST#UV"<vspace/>Visual
1109representation: "http://VU#TS=RQ;PO=NM?LK/JI/HG/FE.DC.BA"<vspace/> The
1110whole IRI (except the scheme) is read rtl. Delimiters between rtl
1111components stay between the respective components; delimiters between
1112ltr and rtl components don't move.</t>
1113
1114<t>Example 4: Each of several sequences of rtl components is inverted
1115on its own: <vspace/>Logical representation:
1116"http://AB.CD.ef/gh/IJ/KL.html"<vspace/>Visual representation:
1117"http://DC.BA.ef/gh/LK/JI.html"<vspace/> Each sequence of rtl
1118components is read rtl, in the same way as each sequence of rtl words
1119in an ltr text is read rtl.</t>
1120
1121<t>Example 5: Example 2, applied to components of different kinds:
1122<vspace/>Logical representation: "http://ab.cd.EF/GH/ij/kl.html"
1123<vspace/>Visual representation:
1124"http://ab.cd.HG/FE/ij/kl.html"<vspace/> The inversion of the domain
1125name label and the path component may be unexpected, but it is
1126consistent with other bidi behavior.  For reassurance that the domain
1127component really is "ab.cd.EF", it may be helpful to read aloud the
1128visual representation following the bidi algorithm. After
1129"http://ab.cd." one reads the RTL block "E-F-slash-G-H", which
1130corresponds to the logical representation.
1131</t>
1132
1133<t>Example 6: Same as Example 5, with more rtl components:
1134<vspace/>Logical representation:
1135"http://ab.CD.EF/GH/IJ/kl.html"<vspace/>Visual representation:
1136"http://ab.JI/HG/FE.DC/kl.html"<vspace/> The inversion of the domain
1137name labels and the path components may be easier to identify because
1138the delimiters also move.</t>
1139
1140<t>Example 7: A single rtl component includes digits: <vspace/>Logical
1141representation: "http://ab.CDE123FGH.ij/kl/mn/op.html"<vspace/>Visual
1142representation: "http://ab.HGF123EDC.ij/kl/mn/op.html"<vspace/>
1143Numbers are written ltr in all cases but are treated as an additional
1144embedding inside a run of rtl characters. This is completely
1145consistent with usual bidirectional text.</t>
1146
1147<t>Example 8 (not allowed): Numbers are at the start or end of an rtl
1148component:<vspace/>Logical representation:
1149"http://ab.cd.ef/GH1/2IJ/KL.html"<vspace/>Visual representation:
1150"http://ab.cd.ef/LK/JI1/2HG.html"<vspace/> The sequence "1/2" is
1151interpreted by the bidi algorithm as a fraction, fragmenting the
1152components and leading to confusion. There are other characters that
1153are interpreted in a special way close to numbers; in particular, "+",
1154"-", "#", "$", "%", ",", ".", and ":".</t>
1155
1156<t>Example 9 (not allowed): The numbers in the previous example are
1157percent-encoded: <vspace/>Logical representation:
1158"http://ab.cd.ef/GH%31/%32IJ/KL.html",<vspace/>Visual representation:
1159"http://ab.cd.ef/LK/JI%32/%31HG.html"</t>
1160
1161<t>Example 10 (allowed but not recommended): <vspace/>Logical
1162representation: "http://ab.CDEFGH.123/kl/mn/op.html"<vspace/>Visual
1163representation: "http://ab.123.HGFEDC/kl/mn/op.html"<vspace/>
1164Components consisting of only numbers are allowed (it would be rather
1165difficult to prohibit them), but these may interact with adjacent RTL
1166components in ways that are not easy to predict.</t>
1167
1168<t>Example 11 (allowed but not recommended): <vspace/>Logical
1169representation: "http://ab.CDEFGH.123ij/kl/mn/op.html"<vspace/>Visual
1170representation: "http://ab.123.HGFEDCij/kl/mn/op.html"<vspace/>
1171Components consisting of numbers and left-to-right characters are
1172allowed, but these may interact with adjacent RTL components in ways
1173that are not easy to predict.</t>
1174</section><!-- examples -->
1175</section><!-- bidi -->
1176
1177<section title="Normalization and Comparison" anchor="equivalence">
1178
1179<t><list style="hanging"><t hangText="Note:">The structure and much of
1180  the material for this section is taken from section 6 of <xref
1181  target="RFC3986"></xref>; the differences are due to the specifics
1182  of IRIs.</t></list></t>
1183
1184<t>One of the most common operations on IRIs is simple comparison:
1185Determining whether two IRIs are equivalent, without using the IRIs to
1186access their respective resource(s). A comparison is performed
1187whenever a response cache is accessed, a browser checks its history to
1188color a link, or an XML parser processes tags within a
1189namespace. Extensive normalization prior to comparison of IRIs may be
1190used by spiders and indexing engines to prune a search space or reduce
1191duplication of request actions and response storage.</t>
1192
1193<t>IRI comparison is performed for some particular purpose. Protocols
1194or implementations that compare IRIs for different purposes will often
1195be subject to differing design trade-offs in regards to how much
1196effort should be spent in reducing aliased identifiers. This section
1197describes various methods that may be used to compare IRIs, the
1198trade-offs between them, and the types of applications that might use
1199them.</t>
1200
1201<section title="Equivalence">
1202
1203<t>Because IRIs exist to identify resources, presumably they should be
1204considered equivalent when they identify the same resource. However,
1205this definition of equivalence is not of much practical use, as there
1206is no way for an implementation to compare two resources to determine
1207if they are "the same" unless it has full knowledge or control of
1208them. For this reason, determination of equivalence or difference of
1209IRIs is based on string comparison, perhaps augmented by reference to
1210additional rules provided by URI scheme definitions.  We use the terms
1211"different" and "equivalent" to describe the possible outcomes of such
1212comparisons, but there are many application-dependent versions of
1213equivalence.</t>
1214
1215<t>Even when it is possible to determine that two IRIs are equivalent,
1216IRI comparison is not sufficient to determine whether two IRIs
1217identify different resources. For example, an owner of two different
1218domain names could decide to serve the same resource from both,
1219resulting in two different IRIs. Therefore, comparison methods are
1220designed to minimize false negatives while strictly avoiding false
1221positives.</t>
1222
1223<t>In testing for equivalence, applications should not directly
1224compare relative references; the references should be converted to
1225their respective target IRIs before comparison. When IRIs are compared
1226to select (or avoid) a network action, such as retrieval of a
1227representation, fragment components (if any) should be excluded from
1228the comparison.</t>
1229
1230<t>Applications using IRIs as identity tokens with no relationship to
1231a protocol MUST use the Simple String Comparison (see <xref
1232target="stringcomp"></xref>).  All other applications MUST select one
1233of the comparison practices from the Comparison Ladder (see <xref
1234target="ladder"></xref>.</t>
1235</section> <!-- equivalence -->
1236
1237
1238<section title="Preparation for Comparison">
1239<t>Any kind of IRI comparison REQUIRES that any additional contextual
1240processing is first performed, including undoing higher-level
1241escapings or encodings in the protocol or format that carries an
1242IRI. This preprocessing is usually done when the protocol or format is
1243parsed.</t>
1244
1245<t>Examples of contextual preprocessing steps are described in <xref
1246target="LEIRIHREF"/>. </t>
1247
1248<t>Examples of such escapings or encodings are entities and
1249numeric character references in <xref target="HTML4"></xref> and <xref
1250target="XML1"></xref>. As an example,
1251"http://example.org/ros&amp;eacute;" (in HTML),
1252"http://example.org/ros&amp;#233;" (in HTML or XML), and
1253<vspace/>"http://example.org/ros&amp;#xE9;" (in HTML or XML) are all
1254resolved into what is denoted in this document (see <xref
1255target="sec-Notation"></xref>) as "http://example.org/ros&amp;#xE9;"
1256(the "&amp;#xE9;" here standing for the actual e-acute character, to
1257compensate for the fact that this document cannot contain non-ASCII
1258characters).</t>
1259
1260<t>Similar considerations apply to encodings such as Transfer Codings
1261in HTTP (see <xref target="RFC2616"></xref>) and Content Transfer
1262Encodings in MIME (<xref target="RFC2045"></xref>), although in these
1263cases, the encoding is based not on characters but on octets, and
1264additional care is required to make sure that characters, and not just
1265arbitrary octets, are compared (see <xref
1266target="stringcomp"></xref>).</t>
1267
1268</section> <!-- preparation -->
1269
1270<section title="Comparison Ladder" anchor="ladder">
1271
1272<t>In practice, a variety of methods are used to test IRI
1273equivalence. These methods fall into a range distinguished by the
1274amount of processing required and the degree to which the probability
1275of false negatives is reduced. As noted above, false negatives cannot
1276be eliminated. In practice, their probability can be reduced, but this
1277reduction requires more processing and is not cost-effective for all
1278applications.</t>
1279
1280
1281<t>If this range of comparison practices is considered as a ladder,
1282the following discussion will climb the ladder, starting with
1283practices that are cheap but have a relatively higher chance of
1284producing false negatives, and proceeding to those that have higher
1285computational cost and lower risk of false negatives.</t>
1286
1287<section title="Simple String Comparison" anchor="stringcomp">
1288
1289<t>If two IRIs, when considered as character strings, are identical,
1290then it is safe to conclude that they are equivalent.  This type of
1291equivalence test has very low computational cost and is in wide use in
1292a variety of applications, particularly in the domain of parsing. It
1293is also used when a definitive answer to the question of IRI
1294equivalence is needed that is independent of the scheme used and that
1295can be calculated quickly and without accessing a network. An example
1296of such a case is XML Namespaces (<xref
1297target="XMLNamespace"></xref>).</t>
1298
1299
1300<t>Testing strings for equivalence requires some basic precautions.
1301This procedure is often referred to as "bit-for-bit" or
1302"byte-for-byte" comparison, which is potentially misleading. Testing
1303strings for equality is normally based on pair comparison of the
1304characters that make up the strings, starting from the first and
1305proceeding until both strings are exhausted and all characters are
1306found to be equal, until a pair of characters compares unequal, or
1307until one of the strings is exhausted before the other.</t>
1308
1309<t>This character comparison requires that each pair of characters be
1310put in comparable encoding form. For example, should one IRI be stored
1311in a byte array in UTF-8 encoding form and the second in a UTF-16
1312encoding form, bit-for-bit comparisons applied naively will produce
1313errors. It is better to speak of equality on a character-for-character
1314rather than on a byte-for-byte or bit-for-bit basis.  In practical
1315terms, character-by-character comparisons should be done codepoint by
1316codepoint after conversion to a common character encoding form.
1317
1318When comparing character by character, the comparison function MUST
1319NOT map IRIs to URIs, because such a mapping would create additional
1320spurious equivalences. It follows that an IRI SHOULD NOT be modified
1321when being transported if there is any chance that this IRI might be
1322used in a context that uses Simple String Comparison.</t>
1323
1324
1325<t>False negatives are caused by the production and use of IRI
1326aliases. Unnecessary aliases can be reduced, regardless of the
1327comparison method, by consistently providing IRI references in an
1328already normalized form (i.e., a form identical to what would be
1329produced after normalization is applied, as described below).
1330Protocols and data formats often limit some IRI comparisons to simple
1331string comparison, based on the theory that people and implementations
1332will, in their own best interest, be consistent in providing IRI
1333references, or at least be consistent enough to negate any efficiency
1334that might be obtained from further normalization.</t>
1335</section> <!-- stringcomp -->
1336
1337<section title="Syntax-Based Normalization">
1338
1339<figure><preamble>Implementations may use logic based on the
1340definitions provided by this specification to reduce the probability
1341of false negatives. This processing is moderately higher in cost than
1342character-for-character string comparison. For example, an application
1343using this approach could reasonably consider the following two IRIs
1344equivalent:</preamble>
1345
1346<artwork>
1347   example://a/b/c/%7Bfoo%7D/ros&amp;#xE9;
1348   eXAMPLE://a/./b/../b/%63/%7bfoo%7d/ros%C3%A9
1349</artwork></figure>
1350
1351<t>Web user agents, such as browsers, typically apply this type of IRI
1352normalization when determining whether a cached response is
1353available. Syntax-based normalization includes such techniques as case
1354normalization, character normalization, percent-encoding
1355normalization, and removal of dot-segments.</t>
1356
1357<section title="Case Normalization">
1358
1359<t>For all IRIs, the hexadecimal digits within a percent-encoding
1360triplet (e.g., "%3a" versus "%3A") are case-insensitive and therefore
1361should be normalized to use uppercase letters for the digits A-F.</t>
1362
1363<t>When an IRI uses components of the generic syntax, the component
1364syntax equivalence rules always apply; namely, that the scheme and
1365US-ASCII only host are case insensitive and therefore should be
1366normalized to lowercase. For example, the URI
1367"HTTP://www.EXAMPLE.com/" is equivalent to
1368"http://www.example.com/". Case equivalence for non-ASCII characters
1369in IRI components that are IDNs are discussed in <xref
1370target="schemecomp"></xref>.  The other generic syntax components are
1371assumed to be case sensitive unless specifically defined otherwise by
1372the scheme.</t>
1373
1374<t>Creating schemes that allow case-insensitive syntax components
1375containing non-ASCII characters should be avoided. Case normalization
1376of non-ASCII characters can be culturally dependent and is always a
1377complex operation. The only exception concerns non-ASCII host names
1378for which the character normalization includes a mapping step derived
1379from case folding.</t>
1380
1381</section> <!-- casenorm -->
1382
1383<section title="Character Normalization" anchor="normalization">
1384
1385<t>The Unicode Standard <xref target="UNIV6"></xref> defines various
1386equivalences between sequences of characters for various
1387purposes. Unicode Standard Annex #15 <xref target="UTR15"></xref>
1388defines various Normalization Forms for these equivalences, in
1389particular Normalization Form C (NFC, Canonical Decomposition,
1390followed by Canonical Composition) and Normalization Form KC (NFKC,
1391Compatibility Decomposition, followed by Canonical Composition).</t>
1392
1393<t> IRIs already in Unicode MUST NOT be normalized before parsing or
1394interpreting. In many non-Unicode character encodings, some text
1395cannot be represented directly. For example, the word "Vietnam" is
1396natively written "Vi&amp;#x1EC7;t Nam" (containing a LATIN SMALL
1397LETTER E WITH CIRCUMFLEX AND DOT BELOW) in NFC, but a direct
1398transcoding from the windows-1258 character encoding leads to
1399"Vi&amp;#xEA;&amp;#x323;t Nam" (containing a LATIN SMALL LETTER E WITH
1400CIRCUMFLEX followed by a COMBINING DOT BELOW). Direct transcoding of
1401other 8-bit encodings of Vietnamese may lead to other
1402representations.</t>
1403
1404<t>Equivalence of IRIs MUST rely on the assumption that IRIs are
1405appropriately pre-character-normalized rather than apply character
1406normalization when comparing two IRIs. The exceptions are conversion
1407from a non-digital form, and conversion from a non-UCS-based character
1408encoding to a UCS-based character encoding. In these cases, NFC or a
1409normalizing transcoder using NFC MUST be used for interoperability. To
1410avoid false negatives and problems with transcoding, IRIs SHOULD be
1411created by using NFC. Using NFKC may avoid even more problems; for
1412example, by choosing half-width Latin letters instead of full-width
1413ones, and full-width instead of half-width Katakana.</t>
1414
1415
1416<t>As an example,
1417"http://www.example.org/r&amp;#xE9;sum&amp;#xE9;.html" (in XML
1418Notation) is in NFC. On the other hand,
1419"http://www.example.org/re&amp;#x301;sume&amp;#x301;.html" is not in
1420NFC.</t>
1421
1422<t>The former uses precombined e-acute characters, and the latter uses
1423"e" characters followed by combining acute accents. Both usages are
1424defined as canonically equivalent in <xref target="UNIV6"></xref>.</t>
1425
1426<t><list style="hanging">
1427
1428<t hangText="Note:">
1429Because it is unknown how a particular sequence of characters is being
1430treated with respect to character normalization, it would be
1431inappropriate to allow third parties to normalize an IRI
1432arbitrarily. This does not contradict the recommendation that when a
1433resource is created, its IRI should be as character normalized as
1434possible (i.e., NFC or even NFKC). This is similar to the
1435uppercase/lowercase problems.  Some parts of a URI are case
1436insensitive (for example, the domain name). For others, it is unclear
1437whether they are case sensitive, case insensitive, or something in
1438between (e.g., case sensitive, but with a multiple choice selection if
1439the wrong case is used, instead of a direct negative result).  The
1440best recipe is that the creator use a reasonable capitalization and,
1441when transferring the URI, capitalization never be
1442changed.</t></list></t>
1443
1444<t>Various IRI schemes may allow the usage of Internationalized Domain
1445Names (IDN) <xref target="RFC3490"></xref> either in the ireg-name
1446part or elsewhere. Character Normalization also applies to IDNs, as
1447discussed in <xref target="schemecomp"></xref>.</t>
1448</section> <!-- charnorm -->
1449
1450<section title="Percent-Encoding Normalization">
1451
1452<t>The percent-encoding mechanism (Section 2.1 of <xref
1453target="RFC3986"></xref>) is a frequent source of variance among
1454otherwise identical IRIs. In addition to the case normalization issue
1455noted above, some IRI producers percent-encode octets that do not
1456require percent-encoding, resulting in IRIs that are equivalent to
1457their nonencoded counterparts. These IRIs should be normalized by
1458decoding any percent-encoded octet sequence that corresponds to an
1459unreserved character, as described in section 2.3 of <xref
1460target="RFC3986"></xref>.</t>
1461
1462<t>For actual resolution, differences in percent-encoding (except for
1463the percent-encoding of reserved characters) MUST always result in the
1464same resource.  For example, "http://example.org/~user",
1465"http://example.org/%7euser", and "http://example.org/%7Euser", must
1466resolve to the same resource.</t>
1467
1468<t>If this kind of equivalence is to be tested, the percent-encoding
1469of both IRIs to be compared has to be aligned; for example, by
1470converting both IRIs to URIs (see Section 3.1), eliminating escape
1471differences in the resulting URIs, and making sure that the case of
1472the hexadecimal characters in the percent-encoding is always the same
1473(preferably upper case). If the IRI is to be passed to another
1474application or used further in some other way, its original form MUST
1475be preserved.  The conversion described here should be performed only
1476for local comparison.</t>
1477
1478</section> <!-- pctnorm -->
1479
1480<section title="Path Segment Normalization">
1481
1482<t>The complete path segments "." and ".." are intended only for use
1483within relative references (Section 4.1 of <xref
1484target="RFC3986"></xref>) and are removed as part of the reference
1485resolution process (Section 5.2 of <xref target="RFC3986"></xref>).
1486However, some implementations may incorrectly assume that reference
1487resolution is not necessary when the reference is already an IRI, and
1488thus fail to remove dot-segments when they occur in non-relative
1489paths.  IRI normalizers should remove dot-segments by applying the
1490remove_dot_segments algorithm to the path, as described in Section
14915.2.4 of <xref target="RFC3986"></xref>.</t>
1492
1493</section> <!-- pathnorm -->
1494</section> <!-- ladder -->
1495
1496<section title="Scheme-Based Normalization" anchor="schemecomp">
1497
1498<t>The syntax and semantics of IRIs vary from scheme to scheme, as
1499described by the defining specification for each
1500scheme. Implementations may use scheme-specific rules, at further
1501processing cost, to reduce the probability of false negatives. For
1502example, because the "http" scheme makes use of an authority
1503component, has a default port of "80", and defines an empty path to be
1504equivalent to "/", the following four IRIs are equivalent:</t>
1505
1506<figure><artwork>
1507   http://example.com
1508   http://example.com/
1509   http://example.com:/
1510   http://example.com:80/</artwork></figure>
1511
1512<t>In general, an IRI that uses the generic syntax for authority with
1513an empty path should be normalized to a path of "/". Likewise, an
1514explicit ":port", for which the port is empty or the default for the
1515scheme, is equivalent to one where the port and its ":" delimiter are
1516elided and thus should be removed by scheme-based normalization. For
1517example, the second IRI above is the normal form for the "http"
1518scheme.</t>
1519
1520<t>Another case where normalization varies by scheme is in the
1521handling of an empty authority component or empty host
1522subcomponent. For many scheme specifications, an empty authority or
1523host is considered an error; for others, it is considered equivalent
1524to "localhost" or the end-user's host. When a scheme defines a default
1525for authority and an IRI reference to that default is desired, the
1526reference should be normalized to an empty authority for the sake of
1527uniformity, brevity, and internationalization. If, however, either the
1528userinfo or port subcomponents are non-empty, then the host should be
1529given explicitly even if it matches the default.</t>
1530
1531<t>Normalization should not remove delimiters when their associated
1532component is empty unless it is licensed to do so by the scheme
1533specification. For example, the IRI "http://example.com/?" cannot be
1534assumed to be equivalent to any of the examples above. Likewise, the
1535presence or absence of delimiters within a userinfo subcomponent is
1536usually significant to its interpretation.  The fragment component is
1537not subject to any scheme-based normalization; thus, two IRIs that
1538differ only by the suffix "#" are considered different regardless of
1539the scheme.</t>
1540 
1541<t>Some IRI schemes allow the usage of Internationalized Domain
1542Names (IDN) <xref target='RFC5890'></xref> either in their ireg-name
1543part or elswhere. When in use in IRIs, those names SHOULD
1544conform to the definition of U-Label in <xref
1545target='RFC5890'></xref>. An IRI containing an invalid IDN cannot
1546successfully be resolved. For legibility purposes, they
1547SHOULD NOT be converted into ASCII Compatible Encoding (ACE).</t>
1548
1549<t>Scheme-based normalization may also consider IDN
1550components and their conversions to punycode as equivalent. As an
1551example, "http://r&amp;#xE9;sum&amp;#xE9;.example.org" may be
1552considered equivalent to
1553"http://xn--rsum-bpad.example.org".</t><t>Other scheme-specific
1554normalizations are possible.</t>
1555
1556</section> <!-- schemenorm -->
1557
1558<section title="Protocol-Based Normalization">
1559
1560<t>Substantial effort to reduce the incidence of false negatives is
1561often cost-effective for web spiders. Consequently, they implement
1562even more aggressive techniques in IRI comparison. For example, if
1563they observe that an IRI such as</t>
1564
1565<figure><artwork>
1566   http://example.com/data</artwork></figure>
1567<t>redirects to an IRI differing only in the trailing slash</t>
1568<figure><artwork>
1569   http://example.com/data/</artwork></figure>
1570
1571<t>they will likely regard the two as equivalent in the future.  This
1572kind of technique is only appropriate when equivalence is clearly
1573indicated by both the result of accessing the resources and the common
1574conventions of their scheme's dereference algorithm (in this case, use
1575of redirection by HTTP origin servers to avoid problems with relative
1576references).</t>
1577
1578</section> <!-- protonorm -->
1579</section> <!-- equivalence -->
1580</section> 
1581
1582<section title="Use of IRIs" anchor="IRIuse">
1583
1584<section title="Limitations on UCS Characters Allowed in IRIs" anchor="limitations">
1585
1586<t>This section discusses limitations on characters and character
1587sequences usable for IRIs beyond those given in <xref target="abnf"/>
1588and <xref target="visual"/>. The considerations in this section are
1589relevant when IRIs are created and when URIs are converted to
1590IRIs.</t>
1591
1592<t>
1593
1594<list style="hanging"><t hangText="a.">The repertoire of characters allowed
1595    in each IRI component is limited by the definition of that component.
1596    For example, the definition of the scheme component does not allow
1597    characters beyond US-ASCII.
1598    <vspace blankLines="1"/>
1599    (Note: In accordance with URI practice, generic IRI
1600    software cannot and should not check for such limitations.)</t>
1601
1602<t hangText="b.">The UCS contains many areas of characters for which
1603    there are strong visual look-alikes. Because of the likelihood of
1604    transcription errors, these also should be avoided. This includes
1605    the full-width equivalents of Latin characters, half-width
1606    Katakana characters for Japanese, and many others. It also
1607    includes many look-alikes of "space", "delims", and "unwise",
1608    characters excluded in <xref target="RFC3491"/>.</t>
1609   
1610</list>
1611</t>
1612
1613<t>Additional information is available from <xref target="UNIXML"/>.
1614    <xref target="UNIXML"/> is written in the context of running text
1615    rather than in that of identifiers. Nevertheless, it discusses
1616    many of the categories of characters not appropriate for IRIs.</t>
1617</section> <!-- limitations -->
1618
1619<section title="Software Interfaces and Protocols">
1620
1621<t>Although an IRI is defined as a sequence of characters, software
1622interfaces for URIs typically function on sequences of octets or other
1623kinds of code units. Thus, software interfaces and protocols MUST
1624define which character encoding is used.</t>
1625
1626<t>Intermediate software interfaces between IRI-capable components and
1627URI-only components MUST map the IRIs per <xref target="mapping"/>,
1628when transferring from IRI-capable to URI-only components.
1629
1630This mapping SHOULD be applied as late as possible. It SHOULD NOT be
1631applied between components that are known to be able to handle IRIs.</t>
1632</section> <!-- software -->
1633
1634<section title="Format of URIs and IRIs in Documents and Protocols">
1635
1636<t>Document formats that transport URIs may have to be upgraded to allow
1637the transport of IRIs. In cases where the document as a whole
1638has a native character encoding, IRIs MUST also be encoded in this
1639character encoding and converted accordingly by a parser or interpreter.
1640
1641IRI characters not expressible in the native character encoding SHOULD
1642be escaped by using the escaping conventions of the document format if
1643such conventions are available. Alternatively, they MAY be
1644percent-encoded according to <xref target="mapping"/>. For example, in
1645HTML or XML, numeric character references SHOULD be used. If a
1646document as a whole has a native character encoding and that character
1647encoding is not UTF-8, then IRIs MUST NOT be placed into the document
1648in the UTF-8 character encoding.</t>
1649
1650<t>((UPDATE THIS NOTE)) Note: Some formats already accommodate IRIs,
1651although they use different terminology. HTML 4.0 <xref
1652target="HTML4"/> defines the conversion from IRIs to URIs as
1653error-avoiding behavior. XML 1.0 <xref target="XML1"/>, XLink <xref
1654target="XLink"/>, XML Schema <xref target="XMLSchema"/>, and
1655specifications based upon them allow IRIs. Also, it is expected that
1656all relevant new W3C formats and protocols will be required to handle
1657IRIs <xref target="CharMod"/>.</t>
1658
1659</section> <!-- format -->
1660
1661<section title="Use of UTF-8 for Encoding Original Characters" anchor="UTF8use">
1662
1663<t>This section discusses details and gives examples for point c) in
1664<xref target="Applicability"/>. To be able to use IRIs, the URI
1665corresponding to the IRI in question has to encode original characters
1666into octets by using UTF-8.  This can be specified for all URIs of a
1667URI scheme or can apply to individual URIs for schemes that do not
1668specify how to encode original characters.  It can apply to the whole
1669URI, or only to some part. For background information on encoding
1670characters into URIs, see also Section 2.5 of <xref
1671target="RFC3986"/>.</t>
1672
1673<t>For new URI schemes, using UTF-8 is recommended in <xref
1674target="RFC4395bis"/>.  Examples where UTF-8 is already used are the URN
1675syntax <xref target="RFC2141"/>, IMAP URLs <xref target="RFC2192"/>,
1676and POP URLs <xref target="RFC2384"/>.  On the other hand, because the
1677HTTP URI scheme does not specify how to encode original characters,
1678only some HTTP URLs can have corresponding but different IRIs.</t>
1679
1680<t>For example, for a document with a URI
1681of<vspace/>"http://www.example.org/r%C3%A9sum%C3%A9.html", it is
1682possible to construct a corresponding IRI (in XML notation, see <xref
1683target="sec-Notation"/>):
1684"http://www.example.org/r&amp;#xE9;sum&amp;#xE9;.html" ("&amp;#xE9;"
1685stands for the e-acute character, and "%C3%A9" is the UTF-8 encoded
1686and percent-encoded representation of that character). On the other
1687hand, for a document with a URI of
1688"http://www.example.org/r%E9sum%E9.html", the percent-encoding octets
1689cannot be converted to actual characters in an IRI, as the
1690percent-encoding is not based on UTF-8.</t>
1691
1692<t>For most URI schemes, there is no need to upgrade their scheme
1693definition in order for them to work with IRIs.  The main case where
1694upgrading makes sense is when a scheme definition, or a particular
1695component of a scheme, is strictly limited to the use of US-ASCII
1696characters with no provision to include non-ASCII characters/octets
1697via percent-encoding, or if a scheme definition currently uses highly
1698scheme-specific provisions for the encoding of non-ASCII characters.
1699An example of this is the mailto: scheme <xref target="RFC2368"/>.</t>
1700
1701<t>This specification updates the IANA registry of URI schemes to note
1702their applicability to IRIs, see <xref target="iana"/>.  All IRIs use
1703URI schemes, and all URIs with URI schemes can be used as IRIs, even
1704though in some cases only by using URIs directly as IRIs, without any
1705conversion.</t>
1706
1707<t>Scheme definitions can impose restrictions on the syntax of
1708scheme-specific URIs; i.e., URIs that are admissible under the generic
1709URI syntax <xref target="RFC3986"/> may not be admissible due to
1710narrower syntactic constraints imposed by a URI scheme
1711specification. URI scheme definitions cannot broaden the syntactic
1712restrictions of the generic URI syntax; otherwise, it would be
1713possible to generate URIs that satisfied the scheme-specific syntactic
1714constraints without satisfying the syntactic constraints of the
1715generic URI syntax. However, additional syntactic constraints imposed
1716by URI scheme specifications are applicable to IRI, as the
1717corresponding URI resulting from the mapping defined in <xref
1718target="mapping"/> MUST be a valid URI under the syntactic
1719restrictions of generic URI syntax and any narrower restrictions
1720imposed by the corresponding URI scheme specification.</t>
1721
1722<t>The requirement for the use of UTF-8 generally applies to all parts
1723of a URI.  However, it is possible that the capability of IRIs to
1724represent a wide range of characters directly is used just in some
1725parts of the IRI (or IRI reference). The other parts of the IRI may
1726only contain US-ASCII characters, or they may not be based on
1727UTF-8. They may be based on another character encoding, or they may
1728directly encode raw binary data (see also <xref
1729target="RFC2397"/>). </t>
1730
1731<t>For example, it is possible to have a URI reference
1732of<vspace/>"http://www.example.org/r%E9sum%E9.xml#r%C3%A9sum%C3%A9",
1733where the document name is encoded in iso-8859-1 based on server
1734settings, but where the fragment identifier is encoded in UTF-8 according
1735to <xref target="XPointer"/>. The IRI corresponding to the above
1736URI would be (in XML notation)<vspace/>"http://www.example.org/r%E9sum%E9.xml#r&amp;#xE9;sum&amp;#xE9;".</t>
1737
1738<t>Similar considerations apply to query parts. The functionality
1739of IRIs (namely, to be able to include non-ASCII characters) can
1740only be used if the query part is encoded in UTF-8.</t>
1741
1742</section> <!-- utf8 -->
1743
1744<section title="Relative IRI References">
1745<t>Processing of relative IRI references against a base is handled
1746straightforwardly; the algorithms of <xref target="RFC3986"/> can
1747be applied directly, treating the characters additionally allowed
1748in IRI references in the same way that unreserved characters are in URI
1749references.</t>
1750
1751</section> <!-- relative -->
1752</section> <!-- IRIuse -->
1753
1754<section title="Liberal Handling of Otherwise Invalid IRIs" anchor="LEIRIHREF">
1755
1756<t>(EDITOR NOTE: This Section may move to an appendix.)
1757 
1758Some technical specifications and widely-deployed software have
1759allowed additional variations and extensions of IRIs to be used in
1760syntactic components. This section describes two widely-used
1761preprocessing agreements. Other technical specifications may wish to
1762reference a syntactic component which is "a valid IRI or a string that
1763will map to a valid IRI after this preprocessing algorithm". These two
1764variants are known as <xref target="LEIRI">Legacy Extended IRI or
1765LEIRI</xref>, and <xref target="HTML5">Web Address</xref>).
1766</t>
1767
1768<t>Future technical specifications SHOULD NOT allow conforming
1769producers to produce, or conforming content to contain, such forms,
1770as they are not interoperable with other IRI consuming software.</t>
1771
1772<section title="LEIRI Processing"  anchor="LEIRIspec">
1773  <t>This section defines Legacy Extended IRIs (LEIRIs).
1774    The syntax of Legacy Extended IRIs is the same as that for &lt;IRI-reference>,
1775    except that the ucschar production is replaced by the leiri-ucschar production:</t>
1776<figure>
1777
1778<artwork>
1779  leiri-ucschar  = " " / "&lt;" / "&gt;" / '"' / "{" / "}" / "|"
1780                   / "\" / "^" / "`" / %x0-1F / %x7F-D7FF
1781                   / %xE000-FFFD / %x10000-10FFFF
1782</artwork>
1783
1784<postamble>
1785  Among other extensions, processors based on this specification also
1786  did not enforce the restriction on bidirectional formatting
1787  characters in <xref target="visual"></xref>, and the iprivate
1788  production becomes redundant.</postamble>
1789</figure>
1790
1791<t>To convert a string allowed as a LEIRI to an IRI, each character
1792allowed in leiri-ucschar but not in ucschar must be percent-encoded
1793using <xref target="compmapping"/>.</t>
1794</section> <!-- leiriproc -->
1795
1796<section title="Web Address Processing" anchor="webaddress">
1797
1798<t>Many popular web browsers have taken the approach of being quite
1799liberal in what is accepted as a "URL" or its relative
1800forms. This section describes their behavior in terms of a preprocessor
1801which maps strings into the IRI space for subsequent parsing and
1802interpretation as an IRI.</t>
1803
1804<t>In some situations, it might be appropriate to describe the syntax
1805that a liberal consumer implementation might accept as a "Web
1806Address" or "Hypertext Reference" or "HREF". However,
1807technical specifications SHOULD restrict the syntactic form allowed by compliant producers
1808to the IRI or IRI reference syntax defined in this document
1809even if they want to mandate this processing.</t>
1810
1811<t>
1812Summary:
1813<list style="symbols">
1814   <t>Leading and trailing whitespace is removed.</t>
1815   <t>Some additional characters are removed.</t>
1816   <t>Some additional characters are allowed and escaped (as with LEIRI).</t>
1817   <t>If interpreting an IRI as a URI, the pct-encoding of the query
1818   component of the parsed URI component depends on operational
1819   context.</t>
1820</list>
1821</t>
1822
1823<t>Each string provided may have an associated charset (called
1824the HREF-charset here); this defaults to UTF-8.
1825For web browsers interpreting HTML, the document
1826charset of a string is determined:
1827
1828<list style="hanging">
1829<t hangText="If the string came from a script (e.g. as an argument to
1830 a method)">The HRef-charset is the script's charset.</t>
1831
1832<t hangText="If the string came from a DOM node (e.g. from an
1833  element)">The node has a Document, and the HRef-charset is the
1834  Document's character encoding.</t>
1835
1836<t hangText="If the string had a HRef-charset defined when the string was
1837created or defined">The HRef-charset is as defined.</t>
1838
1839</list></t>
1840
1841<t>If the resulting HRef-charset is a unicode based character encoding
1842(e.g., UTF-16), then use UTF-8 instead.</t>
1843
1844
1845<figure>
1846<preamble>The syntax for Web Addresses is obtained by replacing the 'ucschar',
1847  pct-form, and path-sep rules with the href-ucschar, href-pct-form, and href-path-sep
1848  rules below. In addition, some characters are stripped.</preamble>
1849
1850<artwork type='abnf'>
1851  href-ucschar  = " " / "&lt;" / "&gt;" / DQUOTE / "{" / "}" / "|"
1852                   / "\" / "^" / "`" / %x0-1F / %x7F-D7FF
1853                   / %xE000-FFFD / %x10000-10FFFF
1854  href-pct-form = pct-encoded / "%"
1855  href-path-sep = "/" / "\"
1856  href-strip    = &lt;to be done&gt;
1857</artwork>
1858
1859<postamble>
1860(NOTE: NEED TO FIX THESE SETS TO MATCH HTML5; NOT SURE ABOUT NEXT SENTENCE)
1861browsers did not enforce the restriction on bidirectional formatting
1862  characters in <xref target="visual"></xref>, and the iprivate
1863  production becomes redundant.</postamble>
1864</figure>
1865
1866<t>'Web Address processing' requires the following additional
1867preprocessing steps:
1868
1869<list style="numbers">
1870
1871<t>Leading and trailing instances of space (U+0020),
1872CR (U+000A), LF (U+000D), and TAB (U+0009) characters are removed.</t>
1873
1874<t>strip all characters in href-strip.</t>
1875  <t>Percent-encode all characters in href-ucschar not in ucschar.</t>
1876  <t>Replace occurrences of "%" not followed by two hexadecimal digits by "%25".</t>
1877  <t>Convert backslashes ('\') matching href-path-sep to forward slashes ('/').</t>
1878</list></t>
1879</section> <!-- webaddress -->
1880
1881<section title="Characters Not Allowed in IRIs" anchor="notAllowed">
1882
1883<t>This section provides a list of the groups of characters and code
1884points that are allowed by LEIRI or HREF but are not allowed in IRIs or are
1885allowed in IRIs only in the query part. For each group of characters,
1886advice on the usage of these characters is also given, concentrating
1887on the reasons for why they are excluded from IRI use.</t>
1888
1889<t>
1890
1891<list><t>Space (U+0020): Some formats and applications use space as a
1892delimiter, e.g. for items in a list. Appendix C of <xref
1893target="RFC3986"></xref> also mentions that white space may have to be
1894added when displaying or printing long URIs; the same applies to long
1895IRIs. This means that spaces can disappear, or can make the what is
1896intended as a single IRI or IRI reference to be treated as two or more
1897separate IRIs.</t>
1898
1899<t>Delimiters "&lt;" (U+003C), "&gt;" (U+003E), and '"' (U+0022):
1900Appendix C of <xref target="RFC3986"></xref> suggests the use of
1901double-quotes ("http://example.com/") and angle brackets
1902(&lt;http://example.com/&gt;) as delimiters for URIs in plain
1903text. These conventions are often used, and also apply to IRIs.  Using
1904these characters in strings intended to be IRIs would result in the
1905IRIs being cut off at the wrong place.</t>
1906
1907<t>Unwise characters "\" (U+005C), "^" (U+005E), "`"
1908(U+0060), "{" (U+007B), "|" (U+007C), and "}" (U+007D): These
1909characters originally have been excluded from URIs because the
1910respective codepoints are assigned to different graphic characters in
1911some 7-bit or 8-bit encoding. Despite the move to Unicode, some of
1912these characters are still occasionally displayed differently on some
1913systems, e.g. U+005C may appear as a Japanese Yen symbol on some
1914systems. Also, the fact that these characters are not used in URIs or
1915IRIs has encouraged their use outside URIs or IRIs in contexts that
1916may include URIs or IRIs. If a string with such a character were used
1917as an IRI in such a context, it would likely be interpreted
1918piecemeal.</t>
1919
1920<t>The controls (C0 controls, DEL, and C1 controls, #x0 - #x1F #x7F -
1921#x9F): There is generally no way to transmit these characters reliably
1922as text outside of a charset encoding.  Even when in encoded form,
1923many software components silently filter out some of these characters,
1924or may stop processing alltogether when encountering some of
1925them. These characters may affect text display in subtle, unnoticable
1926ways or in drastic, global, and irreversible ways depending on the
1927hardware and software involved. The use of some of these characters
1928would allow malicious users to manipulate the display of an IRI and
1929its context in many situations.</t>
1930
1931<t>Bidi formatting characters (U+200E, U+200F, U+202A-202E): These
1932characters affect the display ordering of characters. If IRIs were
1933allowed to contain these characters and the resulting visual display
1934transcribed. they could not be converted back to electronic form
1935(logical order) unambiguously. These characters, if allowed in IRIs,
1936might allow malicious users to manipulate the display of IRI and its
1937context.</t>
1938
1939<t>Specials (U+FFF0-FFFD): These code points provide functionality
1940beyond that useful in an IRI, for example byte order identification,
1941annotation, and replacements for unknown characters and objects. Their
1942use and interpretation in an IRI would serve no purpose and might lead
1943to confusing display variations.</t>
1944
1945<t>Private use code points (U+E000-F8FF, U+F0000-FFFFD,
1946U+100000-10FFFD): Display and interpretation of these code points is
1947by definition undefined without private agreement. Therefore, these
1948code points are not suited for use on the Internet. They are not
1949interoperable and may have unpredictable effects.</t>
1950
1951<t>Tags (U+E0000-E0FFF): These characters provide a way to language
1952tag in Unicode plain text. They are not appropriate for IRIs because
1953language information in identifiers cannot reliably be input,
1954transmitted (e.g. on a visual medium such as paper), or
1955recognized.</t>
1956
1957<t>Non-characters (U+FDD0-FDEF, U+1FFFE-1FFFF, U+2FFFE-2FFFF,
1958U+3FFFE-3FFFF, U+4FFFE-4FFFF, U+5FFFE-5FFFF, U+6FFFE-6FFFF,
1959U+7FFFE-7FFFF, U+8FFFE-8FFFF, U+9FFFE-9FFFF, U+AFFFE-AFFFF,
1960U+BFFFE-BFFFF, U+CFFFE-CFFFF, U+DFFFE-DFFFF, U+EFFFE-EFFFF,
1961U+FFFFE-FFFFF, U+10FFFE-10FFFF): These code points are defined as
1962non-characters. Applications may use some of them internally, but are
1963not prepared to interchange them.</t>
1964
1965</list></t>
1966
1967<t>LEIRI preprocessing disallowed some code points and
1968code units:
1969
1970<list><t>Surrogate code units (D800-DFFF): These do not represent
1971Unicode codepoints.</t></list></t>
1972</section> <!-- notallowed -->
1973</section> <!-- lieirihref -->
1974 
1975<section title="URI/IRI Processing Guidelines (Informative)" anchor="guidelines">
1976
1977<t>This informative section provides guidelines for supporting IRIs in
1978the same software components and operations that currently process
1979URIs: Software interfaces that handle URIs, software that allows users
1980to enter URIs, software that creates or generates URIs, software that
1981displays URIs, formats and protocols that transport URIs, and software
1982that interprets URIs. These may all require modification before
1983functioning properly with IRIs. The considerations in this section
1984also apply to URI references and IRI references.</t>
1985
1986<section title="URI/IRI Software Interfaces">
1987<t>Software interfaces that handle URIs, such as URI-handling APIs and
1988protocols transferring URIs, need interfaces and protocol elements
1989that are designed to carry IRIs.</t>
1990
1991<t>In case the current handling in an API or protocol is based on
1992US-ASCII, UTF-8 is recommended as the character encoding for IRIs, as
1993it is compatible with US-ASCII, is in accordance with the
1994recommendations of <xref target="RFC2277"/>, and makes converting to
1995URIs easy. In any case, the API or protocol definition must clearly
1996define the character encoding to be used.</t>
1997
1998<t>The transfer from URI-only to IRI-capable components requires no
1999mapping, although the conversion described in <xref
2000target="URItoIRI"/> above may be performed. It is preferable not to
2001perform this inverse conversion unless it is certain this can be done
2002correctly.</t>
2003</section>
2004
2005<section title="URI/IRI Entry">
2006
2007<t>Some components allow users to enter URIs into the system
2008by typing or dictation, for example. This software must be updated to allow
2009for IRI entry.</t>
2010
2011<t>A person viewing a visual representation of an IRI (as a sequence
2012of glyphs, in some order, in some visual display) or hearing an IRI
2013will use an entry method for characters in the user's language to
2014input the IRI. Depending on the script and the input method used, this
2015may be a more or less complicated process.</t>
2016
2017<t>The process of IRI entry must ensure, as much as possible, that the
2018restrictions defined in <xref target="abnf"/> are met. This may be
2019done by choosing appropriate input methods or variants/settings
2020thereof, by appropriately converting the characters being input, by
2021eliminating characters that cannot be converted, and/or by issuing a
2022warning or error message to the user.</t>
2023
2024<t>As an example of variant settings, input method editors for East
2025Asian Languages usually allow the input of Latin letters and related
2026characters in full-width or half-width versions. For IRI input, the
2027input method editor should be set so that it produces half-width Latin
2028letters and punctuation and full-width Katakana.</t>
2029
2030<t>An input field primarily or solely used for the input of URIs/IRIs
2031might allow the user to view an IRI as it is mapped to a URI.  Places
2032where the input of IRIs is frequent may provide the possibility for
2033viewing an IRI as mapped to a URI. This will help users when some of
2034the software they use does not yet accept IRIs.</t>
2035
2036<t>An IRI input component interfacing to components that handle URIs,
2037but not IRIs, must map the IRI to a URI before passing it to these
2038components.</t>
2039
2040<t>For the input of IRIs with right-to-left characters, please see
2041<xref target="bidiInput"></xref>.</t>
2042</section>
2043
2044<section title="URI/IRI Transfer between Applications">
2045
2046<t>Many applications (for example, mail user agents) try to detect
2047URIs appearing in plain text. For this, they use some heuristics based
2048on URI syntax. They then allow the user to click on such URIs and
2049retrieve the corresponding resource in an appropriate (usually
2050scheme-dependent) application.</t>
2051
2052<t>Such applications would need to be upgraded, in order to use the
2053IRI syntax as a base for heuristics. In particular, a non-ASCII
2054character should not be taken as the indication of the end of an IRI.
2055Such applications also would need to make sure that they correctly
2056convert the detected IRI from the character encoding of the document
2057or application where the IRI appears, to the character encoding used
2058by the system-wide IRI invocation mechanism, or to a URI (according to
2059<xref target="mapping"/>) if the system-wide invocation mechanism only
2060accepts URIs.</t>
2061
2062<t>The clipboard is another frequently used way to transfer URIs and
2063IRIs from one application to another. On most platforms, the clipboard
2064is able to store and transfer text in many languages and scripts.
2065Correctly used, the clipboard transfers characters, not octets, which
2066will do the right thing with IRIs.</t>
2067</section>
2068
2069<section title="URI/IRI Generation">
2070
2071<t>Systems that offer resources through the Internet, where those
2072resources have logical names, sometimes automatically generate URIs
2073for the resources they offer. For example, some HTTP servers can
2074generate a directory listing for a file directory and then respond to
2075the generated URIs with the files.</t>
2076
2077<t>Many legacy character encodings are in use in various file systems.
2078Many currently deployed systems do not transform the local character
2079representation of the underlying system before generating URIs.</t>
2080
2081<t>For maximum interoperability, systems that generate resource
2082identifiers should make the appropriate transformations. For example,
2083if a file system contains a file named
2084"r&amp;#xE9;sum&amp;#xE9;.html", a server should expose this as
2085"r%C3%A9sum%C3%A9.html" in a URI, which allows use of
2086"r&amp;#xE9;sum&amp;#xE9;.html" in an IRI, even if locally the file
2087name is kept in a character encoding other than UTF-8.
2088</t>
2089
2090<t>This recommendation particularly applies to HTTP servers. For FTP
2091servers, similar considerations apply; see <xref target="RFC2640"/>.</t>
2092</section>
2093
2094<section title="URI/IRI Selection" anchor="selection">
2095<t>In some cases, resource owners and publishers have control over the
2096IRIs used to identify their resources. This control is mostly
2097executed by controlling the resource names, such as file names,
2098directly.</t>
2099
2100<t>In these cases, it is recommended to avoid choosing IRIs that are
2101easily confused. For example, for US-ASCII, the lower-case ell ("l") is
2102easily confused with the digit one ("1"), and the upper-case oh ("O") is
2103easily confused with the digit zero ("0"). Publishers should avoid
2104confusing users with "br0ken" or "1ame" identifiers.</t>
2105
2106<t>Outside the US-ASCII repertoire, there are many more opportunities for
2107confusion; a complete set of guidelines is too lengthy to include
2108here. As long as names are limited to characters from a single script,
2109native writers of a given script or language will know best when
2110ambiguities can appear, and how they can be avoided. What may look
2111ambiguous to a stranger may be completely obvious to the average
2112native user. On the other hand, in some cases, the UCS contains
2113variants for compatibility reasons; for example, for typographic purposes.
2114These should be avoided wherever possible. Although there may be exceptions,
2115newly created resource names should generally be in NFKC
2116<xref target="UTR15"></xref> (which means that they are also in NFC).</t>
2117
2118<t>As an example, the UCS contains the "fi" ligature at U+FB01
2119for compatibility reasons.
2120Wherever possible, IRIs should use the two letters "f" and "i" rather
2121than the "fi" ligature. An example where the latter may be used is
2122in the query part of an IRI for an explicit search for a word written
2123containing the "fi" ligature.</t>
2124
2125<t>In certain cases, there is a chance that characters from different
2126scripts look the same. The best known example is the similarity of the
2127Latin "A", the Greek "Alpha", and the Cyrillic "A". To avoid such
2128cases, IRIs should only be created where all the characters in a
2129single component are used together in a given language. This usually
2130means that all of these characters will be from the same script, but
2131there are languages that mix characters from different scripts (such
2132as Japanese).  This is similar to the heuristics used to distinguish
2133between letters and numbers in the examples above. Also, for Latin,
2134Greek, and Cyrillic, using lowercase letters results in fewer
2135ambiguities than using uppercase letters would.</t>
2136</section>
2137
2138<section title="Display of URIs/IRIs" anchor="display">
2139<t>
2140In situations where the rendering software is not expected to display
2141non-ASCII parts of the IRI correctly using the available layout and font
2142resources, these parts should be percent-encoded before being displayed.</t>
2143
2144<t>For display of Bidi IRIs, please see <xref target="visual"/>.</t>
2145</section>
2146
2147<section title="Interpretation of URIs and IRIs">
2148<t>Software that interprets IRIs as the names of local resources should
2149accept IRIs in multiple forms and convert and match them with the
2150appropriate local resource names.</t>
2151
2152<t>First, multiple representations include both IRIs in the native
2153character encoding of the protocol and also their URI counterparts.</t>
2154
2155<t>Second, it may include URIs constructed based on character
2156encodings other than UTF-8. These URIs may be produced by user agents that do
2157not conform to this specification and that use legacy character encodings to
2158convert non-ASCII characters to URIs. Whether this is necessary, and what
2159character encodings to cover, depends on a number of factors, such as
2160the legacy character encodings used locally and the distribution of
2161various versions of user agents. For example, software for Japanese
2162may accept URIs in Shift_JIS and/or EUC-JP in addition to UTF-8.</t>
2163
2164<t>Third, it may include additional mappings to be more user-friendly
2165and robust against transmission errors. These would be similar to how
2166some servers currently treat URIs as case insensitive or perform
2167additional matching to account for spelling errors. For characters
2168beyond the US-ASCII repertoire, this may, for example, include
2169ignoring the accents on received IRIs or resource names. Please note
2170that such mappings, including case mappings, are language
2171dependent.</t>
2172
2173<t>It can be difficult to identify a resource unambiguously if too
2174many mappings are taken into consideration. However, percent-encoded
2175and not percent-encoded parts of IRIs can always be clearly distinguished.
2176Also, the regularity of UTF-8 (see <xref target="Duerst97"/>) makes the
2177potential for collisions lower than it may seem at first.</t>
2178</section>
2179
2180<section title="Upgrading Strategy">
2181<t>Where this recommendation places further constraints on software
2182for which many instances are already deployed, it is important to
2183introduce upgrades carefully and to be aware of the various
2184interdependencies.</t>
2185
2186<t>If IRIs cannot be interpreted correctly, they should not be created,
2187generated, or transported. This suggests that upgrading URI interpreting
2188software to accept IRIs should have highest priority.</t>
2189
2190<t>On the other hand, a single IRI is interpreted only by a single or
2191very few interpreters that are known in advance, although it may be
2192entered and transported very widely.</t>
2193
2194<t>Therefore, IRIs benefit most from a broad upgrade of software to be
2195able to enter and transport IRIs. However, before an
2196individual IRI is published, care should be taken to upgrade the corresponding
2197interpreting software in order to cover the forms expected to be
2198received by various versions of entry and transport software.</t>
2199
2200<t>The upgrade of generating software to generate IRIs instead of using a
2201local character encoding should happen only after the service is upgraded
2202to accept IRIs. Similarly, IRIs should only be generated when the service
2203accepts IRIs and the intervening infrastructure and protocol is known
2204to transport them safely.</t>
2205
2206<t>Software converting from URIs to IRIs for display should be upgraded
2207only after upgraded entry software has been widely deployed to the
2208population that will see the displayed result.</t>
2209
2210
2211<t>Where there is a free choice of character encodings, it is often
2212possible to reduce the effort and dependencies for upgrading to IRIs
2213by using UTF-8 rather than another encoding. For example, when a new
2214file-based Web server is set up, using UTF-8 as the character encoding
2215for file names will make the transition to IRIs easier. Likewise, when
2216a new Web form is set up using UTF-8 as the character encoding of the
2217form page, the returned query URIs will use UTF-8 as the character
2218encoding (unless the user, for whatever reason, changes the character
2219encoding) and will therefore be compatible with IRIs.</t>
2220
2221
2222<t>These recommendations, when taken together, will allow for the
2223extension from URIs to IRIs in order to handle characters other than
2224US-ASCII while minimizing interoperability problems. For
2225considerations regarding the upgrade of URI scheme definitions, see
2226<xref target="UTF8use"/>.</t>
2227
2228</section>
2229</section> <!-- guidelines -->
2230
2231<section title="IANA Considerations" anchor="iana">
2232
2233<t>RFC Editor and IANA note: Please Replace RFC XXXX with the
2234number of this document when it issues as an RFC. </t>
2235
2236<t>IANA maintains a registry of "URI schemes". A "URI scheme" also
2237serves an "IRI scheme". </t>
2238
2239<t>To clarify that the URI scheme registration process also applies to
2240IRIs, change the description of the "URI schemes" registry
2241header to say "[RFC4395] defines an IANA-maintained registry of URI
2242Schemes. These registries include the Permanent and Provisional URI
2243Schemes.  RFC XXXX updates this registry to designate that schemes may
2244also indicate their usability as IRI schemes.</t>
2245
2246<t> Update "per RFC 4395" to "per RFC 4395 and RFC XXXX".
2247</t>
2248
2249</section> <!-- IANA -->
2250   
2251<section title="Security Considerations" anchor="security">
2252<t>The security considerations discussed in <xref target="RFC3986"/>
2253also apply to IRIs. In addition, the following issues require
2254particular care for IRIs.</t>
2255<t>Incorrect encoding or decoding can lead to security problems.
2256In particular, some UTF-8 decoders do not check against overlong
2257byte sequences. As an example, a "/" is encoded with the byte 0x2F
2258both in UTF-8 and in US-ASCII, but some UTF-8 decoders also wrongly
2259interpret the sequence 0xC0 0xAF as a "/". A sequence such as "%C0%AF.."
2260may pass some security tests and then be interpreted
2261as "/.." in a path if UTF-8 decoders are fault-tolerant, if conversion
2262and checking are not done in the right order, and/or if reserved
2263characters and unreserved characters are not clearly distinguished.</t>
2264
2265<t>There are various ways in which "spoofing" can occur with IRIs.
2266"Spoofing" means that somebody may add a resource name that looks the
2267same or similar to the user, but that points to a different resource.
2268The added resource may pretend to be the real resource by looking
2269very similar but may contain all kinds of changes that may be
2270difficult to spot and that can cause all kinds of problems.
2271Most spoofing possibilities for IRIs are extensions of those for URIs.</t>
2272
2273<t>Spoofing can occur for various reasons. First, a user's normalization expectations or actual normalization
2274when entering an IRI or  transcoding an IRI from a legacy character
2275encoding do not match the normalization used on the
2276server side. Conceptually, this is no different from the problems
2277surrounding the use of case-insensitive web servers. For example,
2278a popular web page with a mixed-case name ("http://big.example.com/PopularPage.html")
2279might be "spoofed" by someone who is able to create "http://big.example.com/popularpage.html".
2280However, the use of unnormalized character sequences, and of additional
2281mappings for user convenience, may increase the chance for spoofing.
2282Protocols and servers that allow the creation of resources with
2283names that are not normalized are particularly vulnerable to such
2284attacks. This is an inherent
2285security problem of the relevant protocol, server, or resource
2286and is not specific to IRIs, but it is mentioned here for completeness.</t>
2287
2288<t>Spoofing can occur in various IRI components, such as the
2289domain name part or a path part. For considerations specific
2290to the domain name part, see <xref target="RFC3491"/>.
2291For the path part, administrators of sites that allow independent
2292users to create resources in the same sub area may have to be careful
2293to check for spoofing.</t>
2294
2295<t>Spoofing can occur because in the UCS many characters look very similar. Details are discussed in <xref target="selection"/>.
2296Again, this is very similar to spoofing possibilities on US-ASCII,
2297e.g., using "br0ken" or "1ame" URIs.</t>
2298
2299<t>Spoofing can occur when URIs with percent-encodings based on various
2300character encodings are accepted to deal with older user agents. In some
2301cases, particularly for Latin-based resource names, this is usually easy to
2302detect because UTF-8-encoded names, when interpreted and viewed as
2303legacy character encodings, produce mostly garbage.</t><t>When
2304concurrently used character encodings have a similar structure but there
2305are no characters that have exactly the same encoding, detection is more
2306difficult.</t>
2307
2308<t>Spoofing can occur with bidirectional IRIs, if the restrictions
2309in <xref target="bidi-structure"/> are not followed. The same visual
2310representation may be interpreted as different logical representations,
2311and vice versa. It is also very important that a correct Unicode bidirectional
2312implementation be used.</t><t>The use of Legacy Extended IRIs introduces additional security issues.</t>
2313</section><!-- security -->
2314
2315<section title="Acknowledgements">
2316<t>This document was derived from <xref target="RFC3987"/>; the acknowledgments from
2317that specification still apply.</t>
2318<t>We would like to thank Ian Hickson, Michael Sperberg-McQueen,
2319  and Dan Connolly for their work on HyperText References, and Norman Walsh, Richard Tobin,
2320  Henry S. Thomson, John Cowan, Paul Grosso, and the XML Core Working Group of the W3C for their work on LEIRIs.</t>
2321<t>In addition, this document was influenced by contributions from (in no particular order) Chris
2322  Lilley, Bjoern Hoehrmann,
2323Felix Sasaki, Jeremy Carroll, Frank Ellermann, Michael Everson, Cary Karp, Matitiahu Allouche,
2324Richard Ishida, Addison Phillips, Jonathan Rosenne, Najib Tounsi, Debbie Garside, Mark Davis, Sarmad
2325Hussain, Ted Hardie, Konrad Lanz, Thomas Roessler, Lisa Dusseault, Julian Reschke, Giovanni
2326Campagna, Anne van Kesteren, Mark Nottingham, Erik van der Poel, Marcin Hanclik, Marcos Caceres, Roy
2327Fielding, Greg Wilkins, Pieter Hintjens, Daniel R. Tobias, Marko Martin, Maciej Stanchowiak, Wil
2328Tan, Yui Naruse, Michael A. Puls II, Dave Thaler, Tom Perch, John Klensin, Shawn Steele, Peter
2329Saint-Andre, Geoffrey Sneddon, Chris Weber, Alex Melnikov, Slim Amamou, SM, Tim Berners-Lee, Yaron
2330Goland, Sam Ruby, Adam Barth, Abdulrahman I. ALGhadir, Aharon Lanin, Thomas Milo, Murray Sargent,
2331Marc Blanchet, and Mykyta Yevstifeyev.</t>
2332</section>
2333
2334<section title="Main Changes Since RFC 3987">
2335  <t>This section describes the main changes since <xref target="RFC3987"></xref>.</t>
2336  <section title="Major restructuring of IRI processing model" anchor="forkChanges">
2337    <t>Major restructuring of IRI processing model to make scheme-specific translation
2338      necessary to handle IDNA requirements and for consistency with web implementations. </t>
2339    <t>Starting with IRI, you want one of:
2340      <list style="hanging">
2341        <t hangText="a"> IRI components (IRI parsed into UTF8 pieces)</t>
2342        <t hangText="b"> URI components (URI parsed into ASCII pieces, encoded correctly) </t>
2343        <t hangText="c"> whole URI  (for passing on to some other system that wants whole URIs) </t>
2344      </list></t>
2345   
2346    <section title="OLD WAY">
2347      <t><list style="numbers">
2348       
2349        <t>Pct-encoding on the whole thing to a URI.
2350          (c1) If you want a (maybe broken) whole URI, you might
2351          stop here.</t>
2352       
2353        <t>Parsing the URI into URI components.
2354          (b1) If you want (maybe broken) URI components, stop here.</t>
2355       
2356        <t> Decode the components (undoing the pct-encoding).
2357          (a) if you want IRI components, stop here.</t>
2358       
2359        <t> reencode:  Either using a different encoding some components
2360          (for domain names, and query components in web pages, which
2361          depends on the component, scheme and context), and otherwise
2362          using pct-encoding.
2363          (b2) if you want (good) URI components, stop here.</t>
2364       
2365        <t> reassemble the reencoded components.
2366          (c2) if you want a (*good*) whole URI stop here.</t>
2367      </list>
2368       
2369      </t>
2370     
2371    </section>
2372   
2373    <section title="NEW WAY">
2374      <t>
2375        <list style="numbers">
2376         
2377          <t> Parse the IRI into IRI components using the generic syntax.
2378            (a) if you want IRI components, stop here.</t>
2379         
2380          <t> Encode each components, using pct-encoding, IDN encoding, or
2381            special query part encoding depending on the component
2382            scheme or context. (b) If you want URI components, stop here.</t>
2383          <t> reassemble the a whole URI from URI components.
2384            (c) if you want a whole URI stop here.</t>
2385        </list></t>
2386    </section>
2387    <section title="Extension of Syntax">
2388      <t>Added the tag range (U+E0000-E0FFF) to the iprivate production.
2389        Some IRIs generated with the new syntax may fail to pass very strict checks
2390        relying on the old syntax. But characters in this range should be extremely infrequent
2391        anyway.</t>
2392    </section>
2393    <section title="More to be added"><t>TODO: There are more main changes that need to be
2394      documented in this section.</t></section>
2395</section>
2396
2397<section title="Change Log">
2398
2399<t>Note to RFC Editor: Please completely remove this section before publication.</t>
2400
2401<section title='Changes after draft-ietf-iri-3987bis-01'>
2402    <t>Changes from draft-ietf-iri-3987bis-01 onwards are available as changesets
2403      in the IETF tools subversion repository at
2404      http://trac.tools.ietf.org/wg/iri/trac/log/draft-ietf-iri-3987bis/draft-ietf-iri-3987bis.xml.</t>
2405</section>
2406 
2407<section title='Changes from draft-duerst-iri-bis-07 to draft-ietf-iri-3987bis-00'>
2408     <t>Changed draft name, date, last paragraph of abstract, and titles in change log, and added this section
2409     in moving from draft-duerst-iri-bis-07 (personal submission) to draft-ietf-iri-3987bis-00 (WG document).</t>
2410</section>
2411
2412<section title="Changes from -06 to -07 of draft-duerst-iri-bis">
2413  <t>Major restructuring of the processing model, see <xref target="forkChanges"></xref>.</t>
2414</section>
2415</section>
2416
2417<section title='Changes from -00 to -01'><t><list style="symbols">
2418  <t>Removed 'mailto:' before mail addresses of authors.</t>
2419  <t>Added "&lt;to be done&gt;" as right side of 'href-strip' rule. Fixed '|' to '/' for
2420    alternatives.</t>
2421</list></t>
2422</section>
2423
2424<section title="Changes from -05 to -06 of draft-duerst-iri-bis-00"><t><list style="symbols">
2425<t>Add HyperText Reference, change abstract, acks and references for it</t>
2426<t>Add Masinter back as another editor.</t>
2427<t>Masinter integrates HRef material from HTML5 spec.</t>
2428<t>Rewrite introduction sections to modernize.</t>
2429</list></t>
2430</section>
2431
2432<section title="Changes from -04 to -05 of draft-duerst-iri-bis">
2433  <t><list style="symbols">
2434    <t>Updated references.</t>
2435    <t>Changed IPR text to pre5378Trust200902.</t></list></t>
2436</section>
2437
2438<section title="Changes from -03 to -04 of draft-duerst-iri-bis">
2439  <t><list style="symbols">
2440    <t>Added explicit abbreviation for LEIRIs.</t>
2441    <t>Mentioned LEIRI references.</t>
2442    <t>Completed text in LEIRI section about tag characters and about specials.</t></list></t>
2443</section>
2444
2445<section title="Changes from -02 to -03 of draft-duerst-iri-bis">
2446  <t><list style="symbols">
2447    <t>Updated some references.</t>
2448    <t>Updated Michel Suginard's coordinates.</t></list></t>
2449</section>
2450
2451<section title="Changes from -01 to -02 of draft-duerst-iri-bis">
2452  <t><list style="symbols">
2453    <t>Added tag range to iprivate (issue private-include-tags-115).</t>
2454    <t>Added Specials (U+FFF0-FFFD) to Legacy Extended IRIs.</t></list></t>
2455</section>
2456<section title="Changes from -00 to -01 of draft-duerst-iri-bis">
2457  <t><list style="symbols">
2458    <t>Changed from "IRIs with Spaces/Controls" to "Legacy Extended IRI"
2459      based on input from the W3C XML Core WG.
2460      Moved the relevant subsections to the back and promoted them to a section.</t>
2461    <t>Added some text re. Legacy Extended IRIs to the security section.</t>
2462    <t>Added a IANA Consideration Section.</t>
2463    <t>Added this Change Log Section.</t>
2464    <t>Added a section about "IRIs with Spaces/Controls" (converting from a Note in RFC 3987).</t></list></t>
2465</section>
2466<section title="Changes from RFC 3987 to -00 of draft-duerst-iri-bis">
2467  <t><list>
2468    <t>Fixed errata (see http://www.rfc-editor.org/cgi-bin/errataSearch.pl?rfc=3987).</t></list></t>
2469</section>
2470</section>
2471</middle>
2472
2473<back>
2474<references title="Normative References">
2475
2476<reference anchor="ASCII">
2477<front>
2478<title>Coded Character Set -- 7-bit American Standard Code for Information
2479Interchange</title>
2480<author>
2481<organization>American National Standards Institute</organization>
2482</author>
2483<date year="1986"/>
2484</front>
2485<seriesInfo name="ANSI" value="X3.4"/>
2486</reference>
2487
2488<reference anchor="ISO10646">
2489<front>
2490<title>ISO/IEC 10646:2003: Information Technology -
2491Universal Multiple-Octet Coded Character Set (UCS)</title>
2492<author>
2493<organization>International Organization for Standardization</organization>
2494</author>
2495<date month="December" year="2003"/>
2496</front>
2497<seriesInfo name="ISO" value="Standard 10646"/>
2498</reference>
2499
2500&rfc2119;
2501&rfc3490;
2502&rfc3491;
2503&rfc3629;
2504&rfc3986;
2505
2506<reference anchor="STD68">
2507<front>
2508<title abbrev="ABNF">Augmented BNF for Syntax Specifications: ABNF</title>
2509<author initials="D." surname="Crocker" fullname="Dave Crocker"><organization/></author>
2510<author initials="P." surname="Overell" fullname="Paul Overell"><organization/></author>
2511<date month="January" year="2008"/></front>
2512<seriesInfo name="STD" value="68"/><seriesInfo name="RFC" value="5234"/>
2513</reference>
2514 
2515&rfc5890;
2516&rfc5891;
2517
2518<reference anchor="UNIV6">
2519<front>
2520<title>The Unicode Standard, Version 6.0.0 (Mountain View, CA, The Unicode Consortium, 2011, ISBN 978-1-936213-01-6)</title>
2521<author><organization>The Unicode Consortium</organization></author>
2522<date year="2010" month="October"/>
2523</front>
2524</reference>
2525
2526<reference anchor="UNI9" target="http://www.unicode.org/reports/tr9/tr9-13.html">
2527<front>
2528<title>The Bidirectional Algorithm</title>
2529<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2530<date year="2004" month="March"/>
2531</front>
2532<seriesInfo name="Unicode Standard Annex" value="#9"/>
2533</reference>
2534
2535<reference anchor="UTR15" target="http://www.unicode.org/unicode/reports/tr15/tr15-23.html">
2536<front>
2537<title>Unicode Normalization Forms</title>
2538<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2539<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2540<date year="2008" month="March"/>
2541</front>
2542<seriesInfo name="Unicode Standard Annex" value="#15"/>
2543</reference>
2544
2545</references>
2546
2547<references title="Informative References">
2548
2549<reference anchor="BidiEx" target="http://www.w3.org/International/iri-edit/BidiExamples">
2550<front>
2551<title>Examples of bidirectional IRIs</title>
2552<author><organization/></author>
2553<date year="" month=""/>
2554</front>
2555</reference>
2556
2557<reference anchor="CharMod" target="http://www.w3.org/TR/charmod-resid">
2558<front>
2559<title>Character Model for the World Wide Web: Resource Identifiers</title>
2560<author initials="M." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2561<author initials="F." surname="Yergeau" fullname="Francois Yergeau"><organization/></author>
2562<author initials="R." surname="Ishida" fullname="Richard Ishida"><organization/></author>
2563<author initials="M." surname="Wolf" fullname="Misha Wolf"><organization/></author>
2564<author initials="T." surname="Texin" fullname="Tex Texin"><organization/></author>
2565<date year="2004" month="November" day="25"/>
2566</front>
2567<seriesInfo name="World Wide Web Consortium" value="Candidate Recommendation"/>
2568</reference>
2569
2570<reference anchor="Duerst97" target="http://www.ifi.unizh.ch/mml/mduerst/papers/PDF/IUC11-UTF-8.pdf">
2571<front>
2572<title>The Properties and Promises of UTF-8</title>
2573<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2574<date year="1997" month="September"/>
2575</front>
2576<seriesInfo name="Proc. 11th International Unicode Conference, San Jose" value=""/>
2577</reference>
2578
2579<reference anchor="Gettys" target="http://www.w3.org/DesignIssues/ModelConsequences">
2580<front>
2581<title>URI Model Consequences</title>
2582<author initials="J." surname="Gettys" fullname="Jim Gettys"><organization/></author>
2583<date month="" year=""/>
2584</front>
2585</reference>
2586
2587<reference anchor="HTML4" target="http://www.w3.org/TR/html401/appendix/notes.html#h-B.2">
2588<front>
2589<title>HTML 4.01 Specification</title>
2590<author initials="D." surname="Raggett" fullname="Dave Raggett"><organization/></author>
2591<author initials="A." surname="Le Hors" fullname="Arnaud Le Hors"><organization/></author>
2592<author initials="I." surname="Jacobs" fullname="Ian Jacobs"><organization/></author>
2593<date year="1999" month="December" day="24"/>
2594</front>
2595<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2596</reference>
2597
2598<reference anchor="LEIRI" target="http://www.w3.org/TR/leiri/">
2599<front>
2600<title>Legacy extended IRIs for XML resource identification</title>
2601<author initials="H." surname="Thompson" fullname="Henry Thompson"><organization/></author>
2602<author initials="R." surname="Tobin"    fullname="Richard Tobin"><organization/></author>
2603<author initials="N." surname="Walsh" fullname="Norman Walsh"><organization/></author>
2604  <date year="2008" month="November" day="3"/>
2605
2606</front>
2607<seriesInfo name="World Wide Web Consortium" value="Note"/>
2608</reference>
2609
2610
2611&rfc2045;
2612&rfc2130;
2613&rfc2141;
2614&rfc2192;
2615&rfc2277;
2616&rfc2368;
2617&rfc2384;
2618&rfc2396;
2619&rfc2397;
2620&rfc2616;
2621&rfc1738;
2622&rfc2640;
2623&rfc3987;
2624<reference anchor='RFC4395bis'>
2625  <front>
2626    <title>Guidelines and Registration Procedures for New URI/IRI Schemes</title>
2627    <author initials='T.' surname='Hansen' fullname="Tony Hansen"><organization/></author>
2628    <author initials='T.' surname='Hardie' fullname="Ted Hardie"><organization/></author>
2629    <author initials='L.' surname='Masinter' fullname="Larry Masinter"><organization/></author>
2630    <date year="2010" month='September' day="30"/>
2631    <workgroup>IRI</workgroup>
2632  </front>
2633  <seriesInfo name="Internet-Draft" value="draft-hansen-iri-4395bis-irireg-00"/>
2634</reference>
2635 
2636 
2637<reference anchor="UNIXML" target="http://www.w3.org/TR/unicode-xml/">
2638<front>
2639<title>Unicode in XML and other Markup Languages</title>
2640<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2641<author initials="A." surname="Freytag" fullname="Asmus Freytag"><organization/></author>
2642<date year="2003" month="June" day="18"/>
2643</front>
2644<seriesInfo name="Unicode Technical Report" value="#20"/>
2645<seriesInfo name="World Wide Web Consortium" value="Note"/>
2646</reference>
2647 
2648<reference anchor="UTR36" target="http://unicode.org/reports/tr36/">
2649<front>
2650<title>Unicode Security Considerations</title>
2651<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2652<author initials="M." surname="Suignard" fullname="Michel Suignard"><organization/></author>
2653<date year="2010" month="August" day="4"/>
2654</front>
2655<seriesInfo name="Unicode Technical Report" value="#36"/>
2656</reference>
2657
2658<reference anchor="XLink" target="http://www.w3.org/TR/xlink/#link-locators">
2659<front>
2660<title>XML Linking Language (XLink) Version 1.0</title>
2661<author initials="S." surname="DeRose" fullname="Steve DeRose"><organization/></author>
2662<author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2663<author initials="D." surname="Orchard" fullname="David Orchard"><organization/></author>
2664<date year="2001" month="June" day="27"/>
2665</front>
2666<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2667</reference>
2668
2669<reference anchor="XML1" target="http://www.w3.org/TR/REC-xml">
2670  <front>
2671    <title>Extensible Markup Language (XML) 1.0 (Forth Edition)</title>
2672    <author initials="T." surname="Bray" fullname="Tim Bray"><organization/></author>
2673    <author initials="J." surname="Paoli" fullname="Jean Paoli"><organization/></author>
2674    <author initials="C.M." surname="Sperberg-McQueen" fullname="C. M. Sperberg-McQueen">
2675      <organization/></author>
2676    <author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2677    <author initials="F." surname="Yergeau" fullname="Francois Yergeau"><organization/></author>
2678    <date day="16" month="August" year="2006"/>
2679  </front>
2680  <seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2681</reference>
2682
2683<reference anchor="XMLNamespace" target="http://www.w3.org/TR/REC-xml-names">
2684  <front>
2685    <title>Namespaces in XML (Second Edition)</title>
2686    <author initials="T." surname="Bray" fullname="Tim Bray"><organization/></author>
2687    <author initials="D." surname="Hollander" fullname="Dave Hollander"><organization/></author>
2688    <author initials="A." surname="Layman" fullname="Andrew Layman"><organization/></author>
2689    <author initials="R." surname="Tobin" fullname="Richard Tobin"><organization></organization></author>
2690    <date day="16" month="August" year="2006"/>
2691  </front>
2692  <seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2693</reference>
2694
2695<reference anchor="XMLSchema" target="http://www.w3.org/TR/xmlschema-2/#anyURI">
2696<front>
2697<title>XML Schema Part 2: Datatypes</title>
2698<author initials="P." surname="Biron" fullname="Paul Biron"><organization/></author>
2699<author initials="A." surname="Malhotra" fullname="Ashok Malhotra"><organization/></author>
2700<date year="2001" month="May" day="2"/>
2701</front>
2702<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2703</reference>
2704
2705<reference anchor="XPointer" target="http://www.w3.org/TR/xptr-framework/#escaping">
2706<front>
2707<title>XPointer Framework</title>
2708<author initials="P." surname="Grosso" fullname="Paul Grosso"><organization/></author>
2709<author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2710<author initials="J." surname="Marsh" fullname="Jonathan Marsh"><organization/></author>
2711<author initials="N." surname="Walsh" fullname="Norman Walsh"><organization/></author>
2712<date year="2003" month="March" day="25"/>
2713</front>
2714<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2715</reference>
2716
2717<reference anchor="HTML5" target="http://www.w3.org/TR/2009/WD-html5-20090423/">
2718<front>
2719<title>A vocabulary and associated APIs for HTML and XHTML</title>
2720<author initials="I." surname="Hickson" fullname="Ian Hickson"><organization>Google, Inc.</organization></author>
2721<author initials="D." surname="Hyatt" fullname="David Hyatt"><organization>Apple, Inc.</organization></author>
2722<date year="2009"  month="April" day="23"/>
2723</front>
2724<seriesInfo name="World Wide Web Consortium" value="Working Draft"/>
2725</reference>
2726
2727</references>
2728
2729</back>
2730</rfc>
Note: See TracBrowser for help on using the repository browser.