source: draft-ietf-iri-3987bis/draft-ietf-iri-3987bis.xml @ 32

Last change on this file since 32 was 32, checked in by duerst@…, 9 years ago

Added a sentence referring to Appendix A in RFC 3987 to the end of Overview subsection of Introduction section.
Removed Appendix A.
This closes issue #53.

  • Property svn:executable set to *
File size: 124.2 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-03" 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="2010" month="October" day="25"/>
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.</t>
129</note>
130</front>
131<middle>
132
133<section title="Introduction">
134
135<section title="Overview and Motivation" anchor="overview">
136
137<t>A Uniform Resource Identifier (URI) is defined in <xref
138target="RFC3986"/> as a sequence of characters chosen from a limited
139subset of the repertoire of US-ASCII <xref target="ASCII"/>
140characters.</t>
141
142<t>The characters in URIs are frequently used for representing words
143of natural languages.  This usage has many advantages: Such URIs are
144easier to memorize, easier to interpret, easier to transcribe, easier
145to create, and easier to guess. For most languages other than English,
146however, the natural script uses characters other than A - Z. For many
147people, handling Latin characters is as difficult as handling the
148characters of other scripts is for those who use only the Latin
149alphabet. Many languages with non-Latin scripts are transcribed with
150Latin letters. These transcriptions are now often used in URIs, but
151they introduce additional difficulties.</t>
152
153<t>The infrastructure for the appropriate handling of characters from
154additional scripts is now widely deployed in operating system and
155application software. Software that can handle a wide variety of
156scripts and languages at the same time is increasingly common. Also,
157an increasing number of protocols and formats can carry a wide range of
158characters.</t>
159
160<t>URIs are used both as a protocol element (for transmission and
161processing by software) and also a presentation element (for display
162and handling by people who read, interpret, coin, or guess them). The
163transition between these roles is more difficult and complex when
164dealing with the larger set of characters than allowed for URIs in
165<xref target="RFC3986"/>. </t>
166
167<t>This document defines the protocol element called Internationalized
168Resource Identifier (IRI), which allow applications of URIs to be
169extended to use resource identifiers that have a much wider repertoire
170of characters. It also provides corresponding "internationalized"
171versions of other constructs from <xref target="RFC3986"/>, such as
172URI references. The syntax of IRIs is defined in <xref
173target="syntax"/>.
174</t>
175
176<t>Using characters outside of A - Z in IRIs adds a number of
177difficulties. <xref target="Bidi"/> discusses the special case of
178bidirectional IRIs using characters from scripts written
179right-to-left.  <xref target="equivalence"/> discusses various forms
180of equivalence between IRIs. <xref target="IRIuse"/> discusses the use
181of IRIs in different situations.  <xref target="guidelines"/> gives
182additional informative guidelines.  <xref target="security"/>
183discusses IRI-specific security considerations.</t>
184
185  <t>When defining IRIs originally, several design alternatives were considered.
186    Historically interested readers can find an overview in Appendix A of <xref target="RFC3987"/>.</t>
187</section> <!-- overview -->
188
189<section title="Applicability" anchor="Applicability">
190
191<t>IRIs are designed to allow protocols and software that deal with
192URIs to be updated to handle IRIs. A "URI scheme" (as defined by <xref
193target="RFC3986"/> and registered through the IANA process defined in
194<xref target="RFC4395bis"/> also serves as an "IRI scheme". Processing of
195IRIs is accomplished by extending the URI syntax while retaining (and
196not expanding) the set of "reserved" characters, such that the syntax
197for any URI scheme may be uniformly extended to allow non-ASCII
198characters. In addition, following parsing of an IRI, it is possible
199to construct a corresponding URI by first encoding characters outside
200of the allowed URI range and then reassembling the components.
201</t>
202
203<t>Practical use of IRIs forms in place of URIs forms depends on the
204following conditions being met:</t>
205
206<t><list style="hanging">
207   
208<t hangText="a.">A protocol or format element MUST be explicitly designated to be
209  able to carry IRIs. The intent is to avoid introducing IRIs into
210  contexts that are not defined to accept them.  For example, XML
211  schema <xref target="XMLSchema"/> has an explicit type "anyURI" that
212  includes IRIs and IRI references. Therefore, IRIs and IRI references
213  can be in attributes and elements of type "anyURI".  On the other
214  hand, in the <xref target="RFC2616"/> definition of HTTP/1.1, the
215  Request URI is defined as a URI, which means that direct use of IRIs
216  is not allowed in HTTP requests.</t>
217
218<t hangText="b.">The protocol or format carrying the IRIs MUST have a
219  mechanism to represent the wide range of characters used in IRIs,
220  either natively or by some protocol- or format-specific escaping
221  mechanism (for example, numeric character references in <xref
222  target="XML1"/>).</t>
223
224<t hangText="c.">The URI scheme definition, if it explicitly allows a
225  percent sign ("%") in any syntactic component, SHOULD define the
226  interpretation of sequences of percent-encoded octets (using "%XX"
227  hex octets) as octet from sequences of UTF-8 encoded strings; this
228  is recommended in the guidelines for registering new schemes, <xref
229  target="RFC4395bis"/>.  For example, this is the practice for IMAP URLs
230  <xref target="RFC2192"/>, POP URLs <xref target="RFC2384"/> and the
231  URN syntax <xref target="RFC2141"/>). Note that use of
232  percent-encoding may also be restricted in some situations, for
233  example, URI schemes that disallow percent-encoding might still be
234  used with a fragment identifier which is percent-encoded (e.g.,
235  <xref target="XPointer"/>). See <xref target="UTF8use"/> for further
236  discussion.</t>
237</list></t>
238
239</section> <!-- applicability -->
240
241<section title="Definitions" anchor="sec-Definitions">
242 
243<t>The following definitions are used in this document; they follow the
244terms in <xref target="RFC2130"/>, <xref target="RFC2277"/>, and
245<xref target="ISO10646"/>.</t>
246<t><list style="hanging">
247   
248<t hangText="character:">A member of a set of elements used for the
249    organization, control, or representation of data. For example,
250    "LATIN CAPITAL LETTER A" names a character.</t>
251   
252<t hangText="octet:">An ordered sequence of eight bits considered as a
253    unit.</t>
254   
255<t hangText="character repertoire:">A set of characters (set in the
256    mathematical sense).</t>
257   
258<t hangText="sequence of characters:">A sequence of characters (one
259    after another).</t>
260   
261<t hangText="sequence of octets:">A sequence of octets (one after
262    another).</t>
263   
264<t hangText="character encoding:">A method of representing a sequence
265    of characters as a sequence of octets (maybe with variants). Also,
266    a method of (unambiguously) converting a sequence of octets into a
267    sequence of characters.</t>
268   
269<t hangText="charset:">The name of a parameter or attribute used to
270    identify a character encoding.</t>
271   
272<t hangText="UCS:">Universal Character Set. The coded character set
273    defined by ISO/IEC 10646 <xref target="ISO10646"/> and the Unicode
274    Standard <xref target="UNIV6"/>.</t>
275   
276<t hangText="IRI reference:">Denotes the common usage of an
277    Internationalized Resource Identifier. An IRI reference may be
278    absolute or relative.  However, the "IRI" that results from such a
279    reference only includes absolute IRIs; any relative IRI references
280    are resolved to their absolute form.  Note that in <xref
281    target="RFC2396"/> URIs did not include fragment identifiers, but
282    in <xref target="RFC3986"/> fragment identifiers are part of
283    URIs.</t>
284   
285<t hangText="URL:">The term "URL" was originally used <xref
286   target="RFC1738"/> for roughly what is now called a "URI".  Books,
287   software and documentation often refers to URIs and IRIs using the
288   "URL" term. Some usages restrict "URL" to those URIs which are not
289   URNs. Because of the ambiguity of the term using the term "URL" is
290   NOT RECOMMENDED in formal documents.</t>
291
292<t hangText="LEIRI (Legacy Extended IRI) processing:">  This term was used in
293   various XML specifications to refer
294   to strings that, although not valid IRIs, were acceptable input to
295   the processing rules in <xref target="LEIRIspec" />.</t>
296
297<t hangText="(Web Address, Hypertext Reference, HREF):"> These terms have been
298   added in this document for convenience, to allow other
299   specifications to refer to those strings that, although not valid
300   IRIs, are acceptable input to the processing rules in <xref
301   target="webaddress"/>. This usage corresponds to the parsing rules
302   of some popular web browsing applications.
303   ISSUE: Need to find a good name/abbreviation for these.</t>
304   
305<t hangText="running text:">Human text (paragraphs, sentences,
306   phrases) with syntax according to orthographic conventions of a
307   natural language, as opposed to syntax defined for ease of
308   processing by machines (e.g., markup, programming languages).</t>
309   
310<t hangText="protocol element:">Any portion of a message that affects
311    processing of that message by the protocol in question.</t>
312   
313<t hangText="presentation element:">A presentation form corresponding
314    to a protocol element; for example, using a wider range of
315    characters.</t>
316   
317<t hangText="create (a URI or IRI):">With respect to URIs and IRIs,
318     the term is used for the initial creation. This may be the
319     initial creation of a resource with a certain identifier, or the
320     initial exposition of a resource under a particular
321     identifier.</t>
322   
323<t hangText="generate (a URI or IRI):">With respect to URIs and IRIs,
324     the term is used when the identifier is generated by derivation
325     from other information.</t>
326
327<t hangText="parsed URI component:">When a URI processor parses a URI
328   (following the generic syntax or a scheme-specific syntax, the result
329   is a set of parsed URI components, each of which has a type
330   (corresponding to the syntactic definition) and a sequence of URI
331   characters.  </t>
332
333<t hangText="parsed IRI component:">When an IRI processor parses
334   an IRI directly, following the general syntax or a scheme-specific
335   syntax, the result is a set of parsed IRI components, each of
336   which has a type (corresponding to the syntactice definition)
337   and a sequence of IRI characters. (This definition is analogous
338   to "parsed URI component".)</t>
339
340<t hangText="IRI scheme:">A URI scheme may also be known as
341   an "IRI scheme" if the scheme's syntax has been extended to
342   allow non-US-ASCII characters according to the rules in this
343   document.</t>
344
345</list></t>
346</section> <!-- definitions -->
347<section title="Notation" anchor="sec-Notation">
348     
349<t>RFCs and Internet Drafts currently do not allow any characters
350outside the US-ASCII repertoire. Therefore, this document uses various
351special notations to denote such characters in examples.</t>
352     
353<t>In text, characters outside US-ASCII are sometimes referenced by
354using a prefix of 'U+', followed by four to six hexadecimal
355digits.</t>
356
357<t>To represent characters outside US-ASCII in examples, this document
358uses two notations: 'XML Notation' and 'Bidi Notation'.</t>
359
360<t>XML Notation uses a leading '&amp;#x', a trailing ';', and the
361hexadecimal number of the character in the UCS in between. For
362example, &amp;#x44F; stands for CYRILLIC CAPITAL LETTER YA. In this
363notation, an actual '&amp;' is denoted by '&amp;amp;'.</t>
364
365<t>Bidi Notation is used for bidirectional examples: Lower case
366letters stand for Latin letters or other letters that are written left
367to right, whereas upper case letters represent Arabic or Hebrew
368letters that are written right to left.</t>
369
370<t>To denote actual octets in examples (as opposed to percent-encoded
371octets), the two hex digits denoting the octet are enclosed in "&lt;"
372and "&gt;".  For example, the octet often denoted as 0xc9 is denoted
373here as &lt;c9&gt;.</t>
374
375<t> In this document, the key words "MUST", "MUST NOT", "REQUIRED",
376"SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY",
377and "OPTIONAL" are to be interpreted as described in <xref
378target="RFC2119"/>.</t>
379
380</section> <!-- notation -->
381</section> <!-- introduction -->
382
383<section title="IRI Syntax" anchor="syntax">
384<t>This section defines the syntax of Internationalized Resource
385Identifiers (IRIs).</t>
386
387<t>As with URIs, an IRI is defined as a sequence of characters, not as
388a sequence of octets. This definition accommodates the fact that IRIs
389may be written on paper or read over the radio as well as stored or
390transmitted digitally.  The same IRI might be represented as different
391sequences of octets in different protocols or documents if these
392protocols or documents use different character encodings (and/or
393transfer encodings).  Using the same character encoding as the
394containing protocol or document ensures that the characters in the IRI
395can be handled (e.g., searched, converted, displayed) in the same way
396as the rest of the protocol or document.</t>
397
398<section title="Summary of IRI Syntax" anchor="summary">
399
400<t>IRIs are defined by extending the URI syntax in <xref
401target="RFC3986"/>, but extending the class of unreserved characters
402by adding the characters of the UCS (Universal Character Set, <xref
403target="ISO10646"/>) beyond U+007F, subject to the limitations given
404in the syntax rules below and in <xref target="limitations"/>.</t>
405
406<t>The syntax and use of components and reserved characters is the
407same as that in <xref target="RFC3986"/>. Each "URI scheme" thus also
408functions as an "IRI scheme", in that scheme-specific parsing rules
409for URIs of a scheme are be extended to allow parsing of IRIs using
410the same parsing rules.</t>
411
412<t>All the operations defined in <xref target="RFC3986"/>, such as the
413resolution of relative references, can be applied to IRIs by
414IRI-processing software in exactly the same way as they are for URIs
415by URI-processing software.</t>
416
417<t>Characters outside the US-ASCII repertoire MUST NOT be reserved and
418therefore MUST NOT be used for syntactical purposes, such as to
419delimit components in newly defined schemes. For example, U+00A2, CENT
420SIGN, is not allowed as a delimiter in IRIs, because it is in the
421'iunreserved' category. This is similar to the fact that it is not
422possible to use '-' as a delimiter in URIs, because it is in the
423'unreserved' category.</t>
424
425</section> <!-- summary -->
426<section title="ABNF for IRI References and IRIs" anchor="abnf">
427
428<t>An ABNF definition for IRI references (which are the most general
429concept and the start of the grammar) and IRIs is given here. The
430syntax of this ABNF is described in <xref target="STD68"/>. Character
431numbers are taken from the UCS, without implying any actual binary
432encoding. Terminals in the ABNF are characters, not octets.</t>
433
434<t>The following grammar closely follows the URI grammar in <xref
435target="RFC3986"/>, except that the range of unreserved characters is
436expanded to include UCS characters, with the restriction that private
437UCS characters can occur only in query parts. The grammar is split
438into two parts: Rules that differ from <xref target="RFC3986"/>
439because of the above-mentioned expansion, and rules that are the same
440as those in <xref target="RFC3986"/>. For rules that are different
441than those in <xref target="RFC3986"/>, the names of the non-terminals
442have been changed as follows. If the non-terminal contains 'URI', this
443has been changed to 'IRI'. Otherwise, an 'i' has been prefixed.</t>
444
445<!--
446for line length measuring in artwork (max 72 chars, three chars at start):
447      1         2         3         4         5         6         7
448456789012345678901234567890123456789012345678901234567890123456789012
449-->
450<figure>
451<preamble>The following rules are different from those in <xref target="RFC3986"/>:</preamble>
452<artwork>
453IRI            = scheme ":" ihier-part [ "?" iquery ]
454                 [ "#" ifragment ]
455
456ihier-part     = "//" iauthority ipath-abempty
457               / ipath-absolute
458               / ipath-rootless
459               / ipath-empty
460
461IRI-reference  = IRI / irelative-ref
462
463absolute-IRI   = scheme ":" ihier-part [ "?" iquery ]
464
465irelative-ref  = irelative-part [ "?" iquery ] [ "#" ifragment ]
466
467irelative-part = "//" iauthority ipath-abempty
468               / ipath-absolute
469               / ipath-noscheme
470               / ipath-empty
471
472iauthority     = [ iuserinfo "@" ] ihost [ ":" port ]
473iuserinfo      = *( iunreserved / pct-form / sub-delims / ":" )
474ihost          = IP-literal / IPv4address / ireg-name
475
476pct-form       = pct-encoded
477
478ireg-name      = *( iunreserved / sub-delims )
479
480ipath          = ipath-abempty   ; begins with "/" or is empty
481               / ipath-absolute  ; begins with "/" but not "//"
482               / ipath-noscheme  ; begins with a non-colon segment
483               / ipath-rootless  ; begins with a segment
484               / ipath-empty     ; zero characters
485
486ipath-abempty  = *( path-sep isegment )
487ipath-absolute = path-sep [ isegment-nz *( path-sep isegment ) ]
488ipath-noscheme = isegment-nz-nc *( path-sep isegment )
489ipath-rootless = isegment-nz *( path-sep isegment )
490ipath-empty    = 0&lt;ipchar&gt;
491path-sep       = "/"
492
493isegment       = *ipchar
494isegment-nz    = 1*ipchar
495isegment-nz-nc = 1*( iunreserved / pct-form / sub-delims
496                     / "@" )
497               ; non-zero-length segment without any colon ":"                     
498
499ipchar         = iunreserved / pct-form / sub-delims / ":"
500               / "@"
501 
502iquery         = *( ipchar / iprivate / "/" / "?" )
503
504ifragment      = *( ipchar / "/" / "?" / "#" )
505
506iunreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~" / ucschar
507
508ucschar        = %xA0-D7FF / %xF900-FDCF / %xFDF0-FFEF
509               / %x10000-1FFFD / %x20000-2FFFD / %x30000-3FFFD
510               / %x40000-4FFFD / %x50000-5FFFD / %x60000-6FFFD
511               / %x70000-7FFFD / %x80000-8FFFD / %x90000-9FFFD
512               / %xA0000-AFFFD / %xB0000-BFFFD / %xC0000-CFFFD
513               / %xD0000-DFFFD / %xE1000-EFFFD
514
515iprivate       = %xE000-F8FF / %xE0000-E0FFF / %xF0000-FFFFD
516               / %x100000-10FFFD
517</artwork>
518</figure>
519
520<t>Some productions are ambiguous. The "first-match-wins" (a.k.a. "greedy")
521algorithm applies. For details, see <xref target="RFC3986"/>.</t>
522
523<figure>
524<preamble>The following rules are the same as those in <xref target="RFC3986"/>:</preamble>
525<artwork>
526scheme         = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
527 
528port           = *DIGIT
529 
530IP-literal     = "[" ( IPv6address / IPvFuture  ) "]"
531 
532IPvFuture      = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" )
533 
534IPv6address    =                            6( h16 ":" ) ls32
535               /                       "::" 5( h16 ":" ) ls32
536               / [               h16 ] "::" 4( h16 ":" ) ls32
537               / [ *1( h16 ":" ) h16 ] "::" 3( h16 ":" ) ls32
538               / [ *2( h16 ":" ) h16 ] "::" 2( h16 ":" ) ls32
539               / [ *3( h16 ":" ) h16 ] "::"    h16 ":"   ls32
540               / [ *4( h16 ":" ) h16 ] "::"              ls32
541               / [ *5( h16 ":" ) h16 ] "::"              h16
542               / [ *6( h16 ":" ) h16 ] "::"
543               
544h16            = 1*4HEXDIG
545ls32           = ( h16 ":" h16 ) / IPv4address
546
547IPv4address    = dec-octet "." dec-octet "." dec-octet "." dec-octet
548
549dec-octet      = DIGIT                 ; 0-9
550               / %x31-39 DIGIT         ; 10-99
551               / "1" 2DIGIT            ; 100-199
552               / "2" %x30-34 DIGIT     ; 200-249
553               / "25" %x30-35          ; 250-255
554           
555pct-encoded    = "%" HEXDIG HEXDIG
556
557unreserved     = ALPHA / DIGIT / "-" / "." / "_" / "~"
558reserved       = gen-delims / sub-delims
559gen-delims     = ":" / "/" / "?" / "#" / "[" / "]" / "@"
560sub-delims     = "!" / "$" / "&amp;" / "'" / "(" / ")"
561               / "*" / "+" / "," / ";" / "="
562</artwork></figure>
563
564<t>This syntax does not support IPv6 scoped addressing zone identifiers.</t>
565
566</section> <!-- abnf -->
567
568</section> <!-- syntax -->
569
570<section title="Processing IRIs and related protocol elements" anchor="processing">
571
572<t>IRIs are meant to replace URIs in identifying resources within new
573versions of protocols, formats, and software components that use a
574UCS-based character repertoire.  Protocols and components may use and
575process IRIs directly. However, there are still numerous systems and
576protocols which only accept URIs or components of parsed URIs; that is,
577they only accept sequences of characters within the subset of US-ASCII
578characters allowed in URIs. </t>
579
580<t>This section defines specific processing steps for IRI consumers
581which establish the relationship between the string given and the
582interpreted derivatives. These
583processing steps apply to both IRIs and IRI references (i.e., absolute
584or relative forms); for IRIs, some steps are scheme specific. </t>
585
586<section title="Converting to UCS" anchor="ucsconv"> 
587 
588<t>Input that is already in a Unicode form (i.e., a sequence of Unicode
589 characters or an octet-stream representing a Unicode-based character
590 encoding such as UTF-8 or UTF-16) should be left as is and not
591 normalized (see (see <xref target="normalization"/>).</t>
592
593  <t>An IRI or IRI reference is a sequence of characters from the UCS.
594    For IRIs that are not already in a Unicode form
595    (as when written on paper, read aloud, or represented in a text stream
596    using a legacy character encoding), convert the IRI to Unicode.
597    Note that some character encodings or transcriptions can be converted
598    to or represented by more than one sequence of Unicode characters.
599    Ideally the resulting IRI would use a normalized form,
600    such as Unicode Normalization Form C <xref target="UTR15"/>
601    (see <xref target='ladder'/> Normalization and Comparison),
602    since that ensures a stable, consistent representation
603    that is most likely to produce the intended results.
604    Implementers and users are cautioned that, while denormalized character sequences are valid,
605    they might be difficult for other users or processes to reproduce
606    and might lead to unexpected results.
607  </t>
608
609<t> In other cases (written on paper, read aloud, or otherwise
610 represented independent of any character encoding) represent the IRI
611 as a sequence of characters from the UCS normalized according to
612 Unicode Normalization Form C (NFC, <xref target="UTR15"/>).</t>
613</section> <!-- ucsconv -->
614
615<section title="Parse the IRI into IRI components">
616
617<t>Parse the IRI, either as a relative reference (no scheme)
618or using scheme specific processing (according to the scheme
619given); the result resulting in a set of parsed IRI components.
620(NOTE: FIX BEFORE RELEASE: INTENT IS THAT ALL IRI SCHEMES
621THAT USE GENERIC SYNTAX AND ALLOW NON-ASCII AUTHORITY CAN
622ONLY USE AUTHORITY FOR NAMES THAT FOLLOW PUNICODE.)
623 </t>
624
625<t>NOTE: The result of parsing into components will correspond result
626in a correspondence of subtrings of the IRI according to the part
627matched.  For example, in <xref target="HTML5"/>, the protocol
628components of interest are SCHEME (scheme), HOST (ireg-name), PORT
629(port), the PATH (ipath after the initial "/"), QUERY (iquery),
630FRAGMENT (ifragment), and AUTHORITY (iauthority).
631</t>
632
633<t>Subsequent processing rules are sometimes used to define other
634syntactic components. For example, <xref target="HTML5"/> defines APIs
635for IRI processing; in these APIs:
636
637<list style="hanging">
638<t hangText="HOSTSPECIFIC"> the substring that follows
639the substring matched by the iauthority production, or the whole
640string if the iauthority production wasn't matched.</t>
641<t hangText="HOSTPORT"> if there is a scheme component and a port
642component and the port given by the port component is different than
643the default port defined for the protocol given by the scheme
644component, then HOSTPORT is the substring that starts with the
645substring matched by the host production and ends with the substring
646matched by the port production, and includes the colon in between the
647two. Otherwise, it is the same as the host component.
648</t>
649</list>
650</t>
651</section> <!-- parse -->
652
653<section title="General percent-encoding of IRI components" anchor="compmapping">
654   
655<t>For most IRI components, it is possible to map the IRI component
656to an equivalent URI component by percent-encoding those characters
657not allowed in URIs. Previous processing steps will have removed
658some characters, and the interpretation of reserved characters will
659have already been done (with the syntactic reserved characters outside
660of the IRI component). This mapping is defined for all sequences
661of Unicode characters, whether or not they are valid for the component
662in question. </t>
663   
664<t>For each character which is not allowed in a valid URI (NOTE: WHAT
665IS THE RIGHT REFERENCE HERE), apply the following steps. </t>
666
667<t><list style="hanging">
668
669<t hangText="Convert to UTF-8">Convert the character to a sequence of
670  one or more octets using UTF-8 <xref target="RFC3629"/>.</t>
671
672<t hangText="Percent encode">Convert each octet of this sequence to %HH,
673   where HH is the hexadecimal notation of the octet value. The
674   hexadecimal notation SHOULD use uppercase letters. (This is the
675   general URI percent-encoding mechanism in Section 2.1 of <xref
676   target="RFC3986"/>.)</t>
677   
678</list></t>
679
680<t>Note that the mapping is an identity transformation for parsed URI
681components of valid URIs, and is idempotent: applying the mapping a
682second time will not change anything.</t>
683</section> <!-- general conversion -->
684
685<section title="Mapping ireg-name" anchor="dnsmapping">
686
687<t>Schemes that allow non-ASCII based characters
688in the reg-name (ireg-name) position MUST convert the ireg-name
689component of an IRI as follows:</t>
690
691<t>Replace the ireg-name part of the IRI by the part converted using
692the ToASCII operation specified in Section 4.1 of <xref
693target="RFC3490"/> on each dot-separated label, and by using U+002E
694(FULL STOP) as a label separator, with the flag UseSTD3ASCIIRules set
695to FALSE, and with the flag AllowUnassigned set to FALSE.
696The ToASCII operation may
697fail, but this would mean that the IRI cannot be resolved.
698In such cases, if the domain name conversion fails, then the
699entire IRI conversion fails. Processors that have no mechanism for
700signalling a failure MAY instead substitute an otherwise
701invalid host name, although such processing SHOULD be avoided.
702 </t>
703
704<t>For example, the IRI
705<vspace/>"http://r&amp;#xE9;sum&amp;#xE9;.example.org"<vspace/> MAY be
706converted to <vspace/>"http://xn--rsum-bad.example.org"<vspace/>;
707conversion to percent-encoded form, e.g.,
708 <vspace/>"http://r%C3%A9sum%C3%A9.example.org", MUST NOT be performed. </t>
709
710<t><list style="hanging"> 
711
712<t hangText="Note:">Domain Names may appear in parts of an IRI other
713than the ireg-name part.  It is the responsibility of scheme-specific
714implementations (if the Internationalized Domain Name is part of the
715scheme syntax) or of server-side implementations (if the
716Internationalized Domain Name is part of 'iquery') to apply the
717necessary conversions at the appropriate point. Example: Trying to
718validate the Web page at<vspace/>
719http://r&amp;#xE9;sum&amp;#xE9;.example.org would lead to an IRI of
720<vspace/>http://validator.w3.org/check?uri=http%3A%2F%2Fr&amp;#xE9;sum&amp;#xE9;.<vspace/>example.org,
721which would convert to a URI
722of<vspace/>http://validator.w3.org/check?uri=http%3A%2F%2Fr%C3%A9sum%C3%A9.<vspace/>example.org.
723The server-side implementation is responsible for making the
724necessary conversions to be able to retrieve the Web page.</t>
725
726<t hangText="Note:">In this process, characters allowed in URI
727references and existing percent-encoded sequences are not encoded further.
728(This mapping is similar to, but different from, the encoding applied
729when arbitrary content is included in some part of a URI.)
730
731For example, an IRI of
732<vspace/>"http://www.example.org/red%09ros&amp;#xE9;#red"
733(in XML notation) is converted to
734<vspace/>"http://www.example.org/red%09ros%C3%A9#red", not to
735something like
736<vspace/>"http%3A%2F%2Fwww.example.org%2Fred%2509ros%C3%A9%23red".
737((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?))
738
739</t>
740
741</list></t>
742</section> <!-- dnsmapping -->
743
744<section title="Mapping query components" anchor="querymapping">
745
746<t>((NOTE: SEE ISSUES LIST))
747
748For compatibility with existing deployed HTTP infrastructure,
749the following special case applies for schemes "http" and "https"
750and IRIs whose origin has a document charset other than one which
751is UCS-based (e.g., UTF-8 or UTF-16). In such a case, the "query"
752component of an IRI is mapped into a URI by using the document
753charset rather than UTF-8 as the binary representation before
754pct-encoding. This mapping is not applied for any other scheme
755or component.</t>
756
757</section> <!-- querymapping -->
758
759<section title="Mapping IRIs to URIs" anchor="mapping">
760
761<t>The canonical mapping from a IRI to URI is defined by applying the
762mapping above (from IRI to URI components) and then reassembling a URI
763from the parsed URI components using the original punctuation that
764delimited the IRI components. </t>
765
766</section> <!-- mapping -->
767
768<section title="Converting URIs to IRIs" anchor="URItoIRI">
769
770<t>In some situations, for presentation and further processing,
771it is desirable to convert a URI into an equivalent IRI in which
772natural characters are represented directly rather than
773percent encoded. Of course, every URI is already an IRI in
774its own right without any conversion, and in general there
775This section gives one such procedure for this conversion.
776</t>
777
778<t>
779The conversion described in this section, if given a valid URI, will
780result in an IRI that maps back to the URI used as an input for the
781conversion (except for potential case differences in percent-encoding
782and for potential percent-encoded unreserved characters).
783
784However, the IRI resulting from this conversion may differ
785from the original IRI (if there ever was one).</t> 
786
787<t>URI-to-IRI conversion removes percent-encodings, but not all
788percent-encodings can be eliminated. There are several reasons for
789this:</t>
790
791<t><list style="hanging">
792
793<t hangText="1.">Some percent-encodings are necessary to distinguish
794    percent-encoded and unencoded uses of reserved characters.</t>
795
796<t hangText="2.">Some percent-encodings cannot be interpreted as sequences
797    of UTF-8 octets.<vspace blankLines="1"/>
798    (Note: The octet patterns of UTF-8 are highly regular.
799    Therefore, there is a very high probability, but no guarantee,
800    that percent-encodings that can be interpreted as sequences of UTF-8
801    octets actually originated from UTF-8. For a detailed discussion,
802    see <xref target="Duerst97"/>.)</t>
803
804<t hangText="3.">The conversion may result in a character that is not
805    appropriate in an IRI. See <xref target="abnf"/>, <xref target="visual"/>,
806      and <xref target="limitations"/> for further details.</t>
807
808<t hangText="4.">IRI to URI conversion has different rules for
809    dealing with domain names and query parameters.</t>
810
811</list></t>
812
813<t>Conversion from a URI to an IRI MAY be done by using the following
814steps:
815
816<list style="hanging">
817<t hangText="1.">Represent the URI as a sequence of octets in
818       US-ASCII.</t>
819
820<t hangText="2.">Convert all percent-encodings ("%" followed by two
821      hexadecimal digits) to the corresponding octets, except those
822      corresponding to "%", characters in "reserved", and characters
823      in US-ASCII not allowed in URIs.</t> 
824
825<t hangText="3.">Re-percent-encode any octet produced in step 2 that
826      is not part of a strictly legal UTF-8 octet sequence.</t>
827
828
829<t hangText="4.">Re-percent-encode all octets produced in step 3 that
830      in UTF-8 represent characters that are not appropriate according
831      to <xref target="abnf"/>, <xref target="visual"/>, and <xref
832      target="limitations"/>.</t> 
833
834<t hangText="5.">Interpret the resulting octet sequence as a sequence
835      of characters encoded in UTF-8.</t>
836
837<t hangText="6.">URIs known to contain domain names in the reg-name
838      component SHOULD convert punycode-encoded domain name labels to
839      the corresponding characters using the ToUnicode procedure. </t>
840</list></t>
841
842<t>This procedure will convert as many percent-encoded characters as
843possible to characters in an IRI. Because there are some choices when
844step 4 is applied (see <xref target="limitations"/>), results may
845vary.</t>
846
847<t>Conversions from URIs to IRIs MUST NOT use any character
848encoding other than UTF-8 in steps 3 and 4, even if it might be
849possible to guess from the context that another character encoding
850than UTF-8 was used in the URI.  For example, the URI
851"http://www.example.org/r%E9sum%E9.html" might with some guessing be
852interpreted to contain two e-acute characters encoded as
853iso-8859-1. It must not be converted to an IRI containing these
854e-acute characters. Otherwise, in the future the IRI will be mapped to
855"http://www.example.org/r%C3%A9sum%C3%A9.html", which is a different
856URI from "http://www.example.org/r%E9sum%E9.html".</t>
857
858<section title="Examples">
859
860<t>This section shows various examples of converting URIs to IRIs.
861Each example shows the result after each of the steps 1 through 6 is
862applied. XML Notation is used for the final result.  Octets are
863denoted by "&lt;" followed by two hexadecimal digits followed by
864"&gt;".</t>
865
866<t>The following example contains the sequence "%C3%BC", which is a
867strictly legal UTF-8 sequence, and which is converted into the actual
868character U+00FC, LATIN SMALL LETTER U WITH DIAERESIS (also known as
869u-umlaut).
870
871<list style="hanging">
872<t hangText="1.">http://www.example.org/D%C3%BCrst</t>
873<t hangText="2.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
874<t hangText="3.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
875<t hangText="4.">http://www.example.org/D&lt;c3&gt;&lt;bc&gt;rst</t>
876<t hangText="5.">http://www.example.org/D&amp;#xFC;rst</t>
877<t hangText="6.">http://www.example.org/D&amp;#xFC;rst</t>
878</list>
879</t>
880
881<t>The following example contains the sequence "%FC", which might
882represent U+00FC, LATIN SMALL LETTER U WITH DIAERESIS, in
883the<vspace/>iso-8859-1 character encoding.  (It might represent other
884characters in other character encodings. For example, the octet
885&lt;fc&gt; in iso-8859-5 represents U+045C, CYRILLIC SMALL LETTER
886KJE.)  Because &lt;fc&gt; is not part of a strictly legal UTF-8
887sequence, it is re-percent-encoded in step 3.
888
889
890<list style="hanging">
891<t hangText="1.">http://www.example.org/D%FCrst</t>
892<t hangText="2.">http://www.example.org/D&lt;fc&gt;rst</t>
893<t hangText="3.">http://www.example.org/D%FCrst</t>
894<t hangText="4.">http://www.example.org/D%FCrst</t>
895<t hangText="5.">http://www.example.org/D%FCrst</t>
896<t hangText="6.">http://www.example.org/D%FCrst</t>
897</list>
898</t>
899
900<t>The following example contains "%e2%80%ae", which is the percent-encoded<vspace/>UTF-8
901character encoding of U+202E, RIGHT-TO-LEFT OVERRIDE. <xref target="visual"/>
902forbids the direct use of this character in an IRI. Therefore, the
903corresponding octets are re-percent-encoded in step 4. This example shows
904that the case (upper- or lowercase) of letters used in percent-encodings may not be preserved.
905The example also contains a punycode-encoded domain name label (xn--99zt52a),
906which is not converted.
907
908<list style="hanging">
909<t hangText="1.">http://xn--99zt52a.example.org/%e2%80%ae</t>
910<t hangText="2.">http://xn--99zt52a.example.org/&lt;e2&gt;&lt;80&gt;&lt;ae&gt;</t>
911<t hangText="3.">http://xn--99zt52a.example.org/&lt;e2&gt;&lt;80&gt;&lt;ae&gt;</t>
912<t hangText="4.">http://xn--99zt52a.example.org/%E2%80%AE</t>
913<t hangText="5.">http://xn--99zt52a.example.org/%E2%80%AE</t>
914<t hangText="6.">http://&amp;#x7D0D;&amp;#x8C46;.example.org/%E2%80%AE</t>
915</list></t>
916
917<t>Note that the label "xn--99zt52a" is converted to U+7D0D U+8C46
918(Japanese Natto). ((EDITOR NOTE: There is some inconsistency in this note.))</t>
919
920</section> <!-- examples -->
921</section> <!-- URItoIRI -->
922</section> <!-- processing -->
923<section title="Bidirectional IRIs for Right-to-Left Languages" anchor="Bidi">
924
925<t>Some UCS characters, such as those used in the Arabic and Hebrew
926scripts, have an inherent right-to-left (rtl) writing direction. IRIs
927containing these characters (called bidirectional IRIs or Bidi IRIs)
928require additional attention because of the non-trivial relation
929between logical representation (used for digital representation and
930for reading/spelling) and visual representation (used for
931display/printing).</t>
932
933<t>Because of the complex interaction between the logical representation,
934the visual representation, and the syntax of a Bidi IRI, a balance is
935needed between various requirements.
936The main requirements are<list style="hanging">
937<t hangText="1.">user-predictable conversion between visual and
938    logical representation;</t>
939<t hangText="2.">the ability to include a wide range of characters
940    in various parts of the IRI; and</t>
941<t hangText="3.">minor or no changes or restrictions for
942      implementations.</t>
943</list></t>
944
945<section title="Logical Storage and Visual Presentation" anchor="visual">
946
947<t>When stored or transmitted in digital representation, bidirectional
948IRIs MUST be in full logical order and MUST conform to the IRI syntax
949rules (which includes the rules relevant to their scheme). This
950ensures that bidirectional IRIs can be processed in the same way as
951other IRIs.</t> <t>Bidirectional IRIs MUST be rendered by using the
952Unicode Bidirectional Algorithm <xref target="UNIV6"/>, <xref
953target="UNI9"/>.  Bidirectional IRIs MUST be rendered in the same way
954as they would be if they were in a left-to-right embedding; i.e., as
955if they were preceded by U+202A, LEFT-TO-RIGHT EMBEDDING (LRE), and
956followed by U+202C, POP DIRECTIONAL FORMATTING (PDF).  Setting the
957embedding direction can also be done in a higher-level protocol (e.g.,
958the dir='ltr' attribute in HTML).</t> 
959
960<t>There is no requirement to use the above embedding if the display
961is still the same without the embedding. For example, a bidirectional
962IRI in a text with left-to-right base directionality (such as used for
963English or Cyrillic) that is preceded and followed by whitespace and
964strong left-to-right characters does not need an embedding.  Also, a
965bidirectional relative IRI reference that only contains strong
966right-to-left characters and weak characters and that starts and ends
967with a strong right-to-left character and appears in a text with
968right-to-left base directionality (such as used for Arabic or Hebrew)
969and is preceded and followed by whitespace and strong characters does
970not need an embedding.</t>
971
972<t>In some other cases, using U+200E, LEFT-TO-RIGHT MARK (LRM), may be
973sufficient to force the correct display behavior.  However, the
974details of the Unicode Bidirectional algorithm are not always easy to
975understand. Implementers are strongly advised to err on the side of
976caution and to use embedding in all cases where they are not
977completely sure that the display behavior is unaffected without the
978embedding.</t>
979
980<t>The Unicode Bidirectional Algorithm (<xref target="UNI9"/>, section
9814.3) permits higher-level protocols to influence bidirectional
982rendering. Such changes by higher-level protocols MUST NOT be used if
983they change the rendering of IRIs.</t> 
984
985<t>The bidirectional formatting characters that may be used before or
986after the IRI to ensure correct display are not themselves part of the
987IRI.  IRIs MUST NOT contain bidirectional formatting characters (LRM,
988RLM, LRE, RLE, LRO, RLO, and PDF). They affect the visual rendering of
989the IRI but do not appear themselves. It would therefore not be
990possible to input an IRI with such characters correctly.</t>
991
992</section> <!-- visual -->
993<section title="Bidi IRI Structure" anchor="bidi-structure">
994
995<t>The Unicode Bidirectional Algorithm is designed mainly for running
996text.  To make sure that it does not affect the rendering of
997bidirectional IRIs too much, some restrictions on bidirectional IRIs
998are necessary. These restrictions are given in terms of delimiters
999(structural characters, mostly punctuation such as "@", ".", ":",
1000and<vspace/>"/") and components (usually consisting mostly of letters
1001and digits).</t>
1002
1003<t>The following syntax rules from <xref target="abnf"/> correspond to
1004components for the purpose of Bidi behavior: iuserinfo, ireg-name,
1005isegment, isegment-nz, isegment-nz-nc, ireg-name, iquery, and
1006ifragment.</t>
1007
1008<t>Specifications that define the syntax of any of the above
1009components MAY divide them further and define smaller parts to be
1010components according to this document. As an example, the restrictions
1011of <xref target="RFC3490"/> on bidirectional domain names correspond
1012to treating each label of a domain name as a component for schemes
1013with ireg-name as a domain name.  Even where the components are not
1014defined formally, it may be helpful to think about some syntax in
1015terms of components and to apply the relevant restrictions.  For
1016example, for the usual name/value syntax in query parts, it is
1017convenient to treat each name and each value as a component. As
1018another example, the extensions in a resource name can be treated as
1019separate components.</t>
1020
1021<t>For each component, the following restrictions apply:</t>
1022<t>
1023<list style="hanging">
1024
1025<t hangText="1.">A component SHOULD NOT use both right-to-left and
1026  left-to-right characters.</t>
1027
1028<t hangText="2.">A component using right-to-left characters SHOULD
1029  start and end with right-to-left characters.</t>
1030
1031</list></t>
1032
1033<t>The above restrictions are given as "SHOULD"s, rather than as
1034"MUST"s.  For IRIs that are never presented visually, they are not
1035relevant.  However, for IRIs in general, they are very important to
1036ensure consistent conversion between visual presentation and logical
1037representation, in both directions.</t>
1038
1039<t><list style="hanging">
1040
1041<t hangText="Note:">In some components, the above restrictions may
1042  actually be strictly enforced.  For example, <xref
1043  target="RFC3490"></xref> requires that these restrictions apply to
1044  the labels of a host name for those schemes where ireg-name is a
1045  host name.  In some other components (for example, path components)
1046  following these restrictions may not be too difficult.  For other
1047  components, such as parts of the query part, it may be very
1048  difficult to enforce the restrictions because the values of query
1049  parameters may be arbitrary character sequences.</t>
1050
1051</list></t>
1052
1053<t>If the above restrictions cannot be satisfied otherwise, the
1054affected component can always be mapped to URI notation as described
1055in <xref target="compmapping"/>. Please note that the whole component
1056has to be mapped (see also Example 9 below).</t>
1057
1058</section> <!-- bidi-structure -->
1059
1060<section title="Input of Bidi IRIs" anchor="bidiInput">
1061
1062<t>Bidi input methods MUST generate Bidi IRIs in logical order while
1063rendering them according to <xref target="visual"/>.  During input,
1064rendering SHOULD be updated after every new character is input to
1065avoid end-user confusion.</t>
1066
1067</section> <!-- bidiInput -->
1068
1069<section title="Examples">
1070
1071<t>This section gives examples of bidirectional IRIs, in Bidi
1072Notation.  It shows legal IRIs with the relationship between logical
1073and visual representation and explains how certain phenomena in this
1074relationship may look strange to somebody not familiar with
1075bidirectional behavior, but familiar to users of Arabic and Hebrew. It
1076also shows what happens if the restrictions given in <xref
1077target="bidi-structure"/> are not followed. The examples below can be
1078seen at <xref target="BidiEx"/>, in Arabic, Hebrew, and Bidi Notation
1079variants.</t>
1080
1081<t>To read the bidi text in the examples, read the visual
1082representation from left to right until you encounter a block of rtl
1083text. Read the rtl block (including slashes and other special
1084characters) from right to left, then continue at the next unread ltr
1085character.</t>
1086
1087<t>Example 1: A single component with rtl characters is inverted:
1088<vspace/>Logical representation:
1089"http://ab.CDEFGH.ij/kl/mn/op.html"<vspace/>Visual representation:
1090"http://ab.HGFEDC.ij/kl/mn/op.html"<vspace/> Components can be read
1091one by one, and each component can be read in its natural
1092direction.</t>
1093
1094<t>Example 2: More than one consecutive component with rtl characters
1095is inverted as a whole: <vspace/>Logical representation:
1096"http://ab.CDE.FGH/ij/kl/mn/op.html"<vspace/>Visual representation:
1097"http://ab.HGF.EDC/ij/kl/mn/op.html"<vspace/> A sequence of rtl
1098components is read rtl, in the same way as a sequence of rtl words is
1099read rtl in a bidi text.</t>
1100
1101<t>Example 3: All components of an IRI (except for the scheme) are
1102rtl.  All rtl components are inverted overall: <vspace/>Logical
1103representation:
1104"http://AB.CD.EF/GH/IJ/KL?MN=OP;QR=ST#UV"<vspace/>Visual
1105representation: "http://VU#TS=RQ;PO=NM?LK/JI/HG/FE.DC.BA"<vspace/> The
1106whole IRI (except the scheme) is read rtl. Delimiters between rtl
1107components stay between the respective components; delimiters between
1108ltr and rtl components don't move.</t>
1109
1110<t>Example 4: Each of several sequences of rtl components is inverted
1111on its own: <vspace/>Logical representation:
1112"http://AB.CD.ef/gh/IJ/KL.html"<vspace/>Visual representation:
1113"http://DC.BA.ef/gh/LK/JI.html"<vspace/> Each sequence of rtl
1114components is read rtl, in the same way as each sequence of rtl words
1115in an ltr text is read rtl.</t>
1116
1117<t>Example 5: Example 2, applied to components of different kinds:
1118<vspace/>Logical representation: "http://ab.cd.EF/GH/ij/kl.html"
1119<vspace/>Visual representation:
1120"http://ab.cd.HG/FE/ij/kl.html"<vspace/> The inversion of the domain
1121name label and the path component may be unexpected, but it is
1122consistent with other bidi behavior.  For reassurance that the domain
1123component really is "ab.cd.EF", it may be helpful to read aloud the
1124visual representation following the bidi algorithm. After
1125"http://ab.cd." one reads the RTL block "E-F-slash-G-H", which
1126corresponds to the logical representation.
1127</t>
1128
1129<t>Example 6: Same as Example 5, with more rtl components:
1130<vspace/>Logical representation:
1131"http://ab.CD.EF/GH/IJ/kl.html"<vspace/>Visual representation:
1132"http://ab.JI/HG/FE.DC/kl.html"<vspace/> The inversion of the domain
1133name labels and the path components may be easier to identify because
1134the delimiters also move.</t>
1135
1136<t>Example 7: A single rtl component includes digits: <vspace/>Logical
1137representation: "http://ab.CDE123FGH.ij/kl/mn/op.html"<vspace/>Visual
1138representation: "http://ab.HGF123EDC.ij/kl/mn/op.html"<vspace/>
1139Numbers are written ltr in all cases but are treated as an additional
1140embedding inside a run of rtl characters. This is completely
1141consistent with usual bidirectional text.</t>
1142
1143<t>Example 8 (not allowed): Numbers are at the start or end of an rtl
1144component:<vspace/>Logical representation:
1145"http://ab.cd.ef/GH1/2IJ/KL.html"<vspace/>Visual representation:
1146"http://ab.cd.ef/LK/JI1/2HG.html"<vspace/> The sequence "1/2" is
1147interpreted by the bidi algorithm as a fraction, fragmenting the
1148components and leading to confusion. There are other characters that
1149are interpreted in a special way close to numbers; in particular, "+",
1150"-", "#", "$", "%", ",", ".", and ":".</t>
1151
1152<t>Example 9 (not allowed): The numbers in the previous example are
1153percent-encoded: <vspace/>Logical representation:
1154"http://ab.cd.ef/GH%31/%32IJ/KL.html",<vspace/>Visual representation:
1155"http://ab.cd.ef/LK/JI%32/%31HG.html"</t>
1156
1157<t>Example 10 (allowed but not recommended): <vspace/>Logical
1158representation: "http://ab.CDEFGH.123/kl/mn/op.html"<vspace/>Visual
1159representation: "http://ab.123.HGFEDC/kl/mn/op.html"<vspace/>
1160Components consisting of only numbers are allowed (it would be rather
1161difficult to prohibit them), but these may interact with adjacent RTL
1162components in ways that are not easy to predict.</t>
1163
1164<t>Example 11 (allowed but not recommended): <vspace/>Logical
1165representation: "http://ab.CDEFGH.123ij/kl/mn/op.html"<vspace/>Visual
1166representation: "http://ab.123.HGFEDCij/kl/mn/op.html"<vspace/>
1167Components consisting of numbers and left-to-right characters are
1168allowed, but these may interact with adjacent RTL components in ways
1169that are not easy to predict.</t>
1170</section><!-- examples -->
1171</section><!-- bidi -->
1172
1173<section title="Normalization and Comparison" anchor="equivalence">
1174
1175<t><list style="hanging"><t hangText="Note:">The structure and much of
1176  the material for this section is taken from section 6 of <xref
1177  target="RFC3986"></xref>; the differences are due to the specifics
1178  of IRIs.</t></list></t>
1179
1180<t>One of the most common operations on IRIs is simple comparison:
1181Determining whether two IRIs are equivalent, without using the IRIs to
1182access their respective resource(s). A comparison is performed
1183whenever a response cache is accessed, a browser checks its history to
1184color a link, or an XML parser processes tags within a
1185namespace. Extensive normalization prior to comparison of IRIs may be
1186used by spiders and indexing engines to prune a search space or reduce
1187duplication of request actions and response storage.</t>
1188
1189<t>IRI comparison is performed for some particular purpose. Protocols
1190or implementations that compare IRIs for different purposes will often
1191be subject to differing design trade-offs in regards to how much
1192effort should be spent in reducing aliased identifiers. This section
1193describes various methods that may be used to compare IRIs, the
1194trade-offs between them, and the types of applications that might use
1195them.</t>
1196
1197<section title="Equivalence">
1198
1199<t>Because IRIs exist to identify resources, presumably they should be
1200considered equivalent when they identify the same resource. However,
1201this definition of equivalence is not of much practical use, as there
1202is no way for an implementation to compare two resources to determine
1203if they are "the same" unless it has full knowledge or control of
1204them. For this reason, determination of equivalence or difference of
1205IRIs is based on string comparison, perhaps augmented by reference to
1206additional rules provided by URI scheme definitions.  We use the terms
1207"different" and "equivalent" to describe the possible outcomes of such
1208comparisons, but there are many application-dependent versions of
1209equivalence.</t>
1210
1211<t>Even when it is possible to determine that two IRIs are equivalent,
1212IRI comparison is not sufficient to determine whether two IRIs
1213identify different resources. For example, an owner of two different
1214domain names could decide to serve the same resource from both,
1215resulting in two different IRIs. Therefore, comparison methods are
1216designed to minimize false negatives while strictly avoiding false
1217positives.</t>
1218
1219<t>In testing for equivalence, applications should not directly
1220compare relative references; the references should be converted to
1221their respective target IRIs before comparison. When IRIs are compared
1222to select (or avoid) a network action, such as retrieval of a
1223representation, fragment components (if any) should be excluded from
1224the comparison.</t>
1225
1226<t>Applications using IRIs as identity tokens with no relationship to
1227a protocol MUST use the Simple String Comparison (see <xref
1228target="stringcomp"></xref>).  All other applications MUST select one
1229of the comparison practices from the Comparison Ladder (see <xref
1230target="ladder"></xref>.</t>
1231</section> <!-- equivalence -->
1232
1233
1234<section title="Preparation for Comparison">
1235<t>Any kind of IRI comparison REQUIRES that any additional contextual
1236processing is first performed, including undoing higher-level
1237escapings or encodings in the protocol or format that carries an
1238IRI. This preprocessing is usually done when the protocol or format is
1239parsed.</t>
1240
1241<t>Examples of contextual preprocessing steps are described in <xref
1242target="LEIRIHREF"/>. </t>
1243
1244<t>Examples of such escapings or encodings are entities and
1245numeric character references in <xref target="HTML4"></xref> and <xref
1246target="XML1"></xref>. As an example,
1247"http://example.org/ros&amp;eacute;" (in HTML),
1248"http://example.org/ros&amp;#233;" (in HTML or XML), and
1249<vspace/>"http://example.org/ros&amp;#xE9;" (in HTML or XML) are all
1250resolved into what is denoted in this document (see <xref
1251target="sec-Notation"></xref>) as "http://example.org/ros&amp;#xE9;"
1252(the "&amp;#xE9;" here standing for the actual e-acute character, to
1253compensate for the fact that this document cannot contain non-ASCII
1254characters).</t>
1255
1256<t>Similar considerations apply to encodings such as Transfer Codings
1257in HTTP (see <xref target="RFC2616"></xref>) and Content Transfer
1258Encodings in MIME (<xref target="RFC2045"></xref>), although in these
1259cases, the encoding is based not on characters but on octets, and
1260additional care is required to make sure that characters, and not just
1261arbitrary octets, are compared (see <xref
1262target="stringcomp"></xref>).</t>
1263
1264</section> <!-- preparation -->
1265
1266<section title="Comparison Ladder" anchor="ladder">
1267
1268<t>In practice, a variety of methods are used to test IRI
1269equivalence. These methods fall into a range distinguished by the
1270amount of processing required and the degree to which the probability
1271of false negatives is reduced. As noted above, false negatives cannot
1272be eliminated. In practice, their probability can be reduced, but this
1273reduction requires more processing and is not cost-effective for all
1274applications.</t>
1275
1276
1277<t>If this range of comparison practices is considered as a ladder,
1278the following discussion will climb the ladder, starting with
1279practices that are cheap but have a relatively higher chance of
1280producing false negatives, and proceeding to those that have higher
1281computational cost and lower risk of false negatives.</t>
1282
1283<section title="Simple String Comparison" anchor="stringcomp">
1284
1285<t>If two IRIs, when considered as character strings, are identical,
1286then it is safe to conclude that they are equivalent.  This type of
1287equivalence test has very low computational cost and is in wide use in
1288a variety of applications, particularly in the domain of parsing. It
1289is also used when a definitive answer to the question of IRI
1290equivalence is needed that is independent of the scheme used and that
1291can be calculated quickly and without accessing a network. An example
1292of such a case is XML Namespaces (<xref
1293target="XMLNamespace"></xref>).</t>
1294
1295
1296<t>Testing strings for equivalence requires some basic precautions.
1297This procedure is often referred to as "bit-for-bit" or
1298"byte-for-byte" comparison, which is potentially misleading. Testing
1299strings for equality is normally based on pair comparison of the
1300characters that make up the strings, starting from the first and
1301proceeding until both strings are exhausted and all characters are
1302found to be equal, until a pair of characters compares unequal, or
1303until one of the strings is exhausted before the other.</t>
1304
1305<t>This character comparison requires that each pair of characters be
1306put in comparable encoding form. For example, should one IRI be stored
1307in a byte array in UTF-8 encoding form and the second in a UTF-16
1308encoding form, bit-for-bit comparisons applied naively will produce
1309errors. It is better to speak of equality on a character-for-character
1310rather than on a byte-for-byte or bit-for-bit basis.  In practical
1311terms, character-by-character comparisons should be done codepoint by
1312codepoint after conversion to a common character encoding form.
1313
1314When comparing character by character, the comparison function MUST
1315NOT map IRIs to URIs, because such a mapping would create additional
1316spurious equivalences. It follows that an IRI SHOULD NOT be modified
1317when being transported if there is any chance that this IRI might be
1318used in a context that uses Simple String Comparison.</t>
1319
1320
1321<t>False negatives are caused by the production and use of IRI
1322aliases. Unnecessary aliases can be reduced, regardless of the
1323comparison method, by consistently providing IRI references in an
1324already normalized form (i.e., a form identical to what would be
1325produced after normalization is applied, as described below).
1326Protocols and data formats often limit some IRI comparisons to simple
1327string comparison, based on the theory that people and implementations
1328will, in their own best interest, be consistent in providing IRI
1329references, or at least be consistent enough to negate any efficiency
1330that might be obtained from further normalization.</t>
1331</section> <!-- stringcomp -->
1332
1333<section title="Syntax-Based Normalization">
1334
1335<figure><preamble>Implementations may use logic based on the
1336definitions provided by this specification to reduce the probability
1337of false negatives. This processing is moderately higher in cost than
1338character-for-character string comparison. For example, an application
1339using this approach could reasonably consider the following two IRIs
1340equivalent:</preamble>
1341
1342<artwork>
1343   example://a/b/c/%7Bfoo%7D/ros&amp;#xE9;
1344   eXAMPLE://a/./b/../b/%63/%7bfoo%7d/ros%C3%A9
1345</artwork></figure>
1346
1347<t>Web user agents, such as browsers, typically apply this type of IRI
1348normalization when determining whether a cached response is
1349available. Syntax-based normalization includes such techniques as case
1350normalization, character normalization, percent-encoding
1351normalization, and removal of dot-segments.</t>
1352
1353<section title="Case Normalization">
1354
1355<t>For all IRIs, the hexadecimal digits within a percent-encoding
1356triplet (e.g., "%3a" versus "%3A") are case-insensitive and therefore
1357should be normalized to use uppercase letters for the digits A-F.</t>
1358
1359<t>When an IRI uses components of the generic syntax, the component
1360syntax equivalence rules always apply; namely, that the scheme and
1361US-ASCII only host are case insensitive and therefore should be
1362normalized to lowercase. For example, the URI
1363"HTTP://www.EXAMPLE.com/" is equivalent to
1364"http://www.example.com/". Case equivalence for non-ASCII characters
1365in IRI components that are IDNs are discussed in <xref
1366target="schemecomp"></xref>.  The other generic syntax components are
1367assumed to be case sensitive unless specifically defined otherwise by
1368the scheme.</t>
1369
1370<t>Creating schemes that allow case-insensitive syntax components
1371containing non-ASCII characters should be avoided. Case normalization
1372of non-ASCII characters can be culturally dependent and is always a
1373complex operation. The only exception concerns non-ASCII host names
1374for which the character normalization includes a mapping step derived
1375from case folding.</t>
1376
1377</section> <!-- casenorm -->
1378
1379<section title="Character Normalization" anchor="normalization">
1380
1381<t>The Unicode Standard <xref target="UNIV6"></xref> defines various
1382equivalences between sequences of characters for various
1383purposes. Unicode Standard Annex #15 <xref target="UTR15"></xref>
1384defines various Normalization Forms for these equivalences, in
1385particular Normalization Form C (NFC, Canonical Decomposition,
1386followed by Canonical Composition) and Normalization Form KC (NFKC,
1387Compatibility Decomposition, followed by Canonical Composition).</t>
1388
1389<t> IRIs already in Unicode MUST NOT be normalized before parsing or
1390interpreting. In many non-Unicode character encodings, some text
1391cannot be represented directly. For example, the word "Vietnam" is
1392natively written "Vi&amp;#x1EC7;t Nam" (containing a LATIN SMALL
1393LETTER E WITH CIRCUMFLEX AND DOT BELOW) in NFC, but a direct
1394transcoding from the windows-1258 character encoding leads to
1395"Vi&amp;#xEA;&amp;#x323;t Nam" (containing a LATIN SMALL LETTER E WITH
1396CIRCUMFLEX followed by a COMBINING DOT BELOW). Direct transcoding of
1397other 8-bit encodings of Vietnamese may lead to other
1398representations.</t>
1399
1400<t>Equivalence of IRIs MUST rely on the assumption that IRIs are
1401appropriately pre-character-normalized rather than apply character
1402normalization when comparing two IRIs. The exceptions are conversion
1403from a non-digital form, and conversion from a non-UCS-based character
1404encoding to a UCS-based character encoding. In these cases, NFC or a
1405normalizing transcoder using NFC MUST be used for interoperability. To
1406avoid false negatives and problems with transcoding, IRIs SHOULD be
1407created by using NFC. Using NFKC may avoid even more problems; for
1408example, by choosing half-width Latin letters instead of full-width
1409ones, and full-width instead of half-width Katakana.</t>
1410
1411
1412<t>As an example,
1413"http://www.example.org/r&amp;#xE9;sum&amp;#xE9;.html" (in XML
1414Notation) is in NFC. On the other hand,
1415"http://www.example.org/re&amp;#x301;sume&amp;#x301;.html" is not in
1416NFC.</t>
1417
1418<t>The former uses precombined e-acute characters, and the latter uses
1419"e" characters followed by combining acute accents. Both usages are
1420defined as canonically equivalent in <xref target="UNIV6"></xref>.</t>
1421
1422<t><list style="hanging">
1423
1424<t hangText="Note:">
1425Because it is unknown how a particular sequence of characters is being
1426treated with respect to character normalization, it would be
1427inappropriate to allow third parties to normalize an IRI
1428arbitrarily. This does not contradict the recommendation that when a
1429resource is created, its IRI should be as character normalized as
1430possible (i.e., NFC or even NFKC). This is similar to the
1431uppercase/lowercase problems.  Some parts of a URI are case
1432insensitive (for example, the domain name). For others, it is unclear
1433whether they are case sensitive, case insensitive, or something in
1434between (e.g., case sensitive, but with a multiple choice selection if
1435the wrong case is used, instead of a direct negative result).  The
1436best recipe is that the creator use a reasonable capitalization and,
1437when transferring the URI, capitalization never be
1438changed.</t></list></t>
1439
1440<t>Various IRI schemes may allow the usage of Internationalized Domain
1441Names (IDN) <xref target="RFC3490"></xref> either in the ireg-name
1442part or elsewhere. Character Normalization also applies to IDNs, as
1443discussed in <xref target="schemecomp"></xref>.</t>
1444</section> <!-- charnorm -->
1445
1446<section title="Percent-Encoding Normalization">
1447
1448<t>The percent-encoding mechanism (Section 2.1 of <xref
1449target="RFC3986"></xref>) is a frequent source of variance among
1450otherwise identical IRIs. In addition to the case normalization issue
1451noted above, some IRI producers percent-encode octets that do not
1452require percent-encoding, resulting in IRIs that are equivalent to
1453their nonencoded counterparts. These IRIs should be normalized by
1454decoding any percent-encoded octet sequence that corresponds to an
1455unreserved character, as described in section 2.3 of <xref
1456target="RFC3986"></xref>.</t>
1457
1458<t>For actual resolution, differences in percent-encoding (except for
1459the percent-encoding of reserved characters) MUST always result in the
1460same resource.  For example, "http://example.org/~user",
1461"http://example.org/%7euser", and "http://example.org/%7Euser", must
1462resolve to the same resource.</t>
1463
1464<t>If this kind of equivalence is to be tested, the percent-encoding
1465of both IRIs to be compared has to be aligned; for example, by
1466converting both IRIs to URIs (see Section 3.1), eliminating escape
1467differences in the resulting URIs, and making sure that the case of
1468the hexadecimal characters in the percent-encoding is always the same
1469(preferably upper case). If the IRI is to be passed to another
1470application or used further in some other way, its original form MUST
1471be preserved.  The conversion described here should be performed only
1472for local comparison.</t>
1473
1474</section> <!-- pctnorm -->
1475
1476<section title="Path Segment Normalization">
1477
1478<t>The complete path segments "." and ".." are intended only for use
1479within relative references (Section 4.1 of <xref
1480target="RFC3986"></xref>) and are removed as part of the reference
1481resolution process (Section 5.2 of <xref target="RFC3986"></xref>).
1482However, some implementations may incorrectly assume that reference
1483resolution is not necessary when the reference is already an IRI, and
1484thus fail to remove dot-segments when they occur in non-relative
1485paths.  IRI normalizers should remove dot-segments by applying the
1486remove_dot_segments algorithm to the path, as described in Section
14875.2.4 of <xref target="RFC3986"></xref>.</t>
1488
1489</section> <!-- pathnorm -->
1490</section> <!-- ladder -->
1491
1492<section title="Scheme-Based Normalization" anchor="schemecomp">
1493
1494<t>The syntax and semantics of IRIs vary from scheme to scheme, as
1495described by the defining specification for each
1496scheme. Implementations may use scheme-specific rules, at further
1497processing cost, to reduce the probability of false negatives. For
1498example, because the "http" scheme makes use of an authority
1499component, has a default port of "80", and defines an empty path to be
1500equivalent to "/", the following four IRIs are equivalent:</t>
1501
1502<figure><artwork>
1503   http://example.com
1504   http://example.com/
1505   http://example.com:/
1506   http://example.com:80/</artwork></figure>
1507
1508<t>In general, an IRI that uses the generic syntax for authority with
1509an empty path should be normalized to a path of "/". Likewise, an
1510explicit ":port", for which the port is empty or the default for the
1511scheme, is equivalent to one where the port and its ":" delimiter are
1512elided and thus should be removed by scheme-based normalization. For
1513example, the second IRI above is the normal form for the "http"
1514scheme.</t>
1515
1516<t>Another case where normalization varies by scheme is in the
1517handling of an empty authority component or empty host
1518subcomponent. For many scheme specifications, an empty authority or
1519host is considered an error; for others, it is considered equivalent
1520to "localhost" or the end-user's host. When a scheme defines a default
1521for authority and an IRI reference to that default is desired, the
1522reference should be normalized to an empty authority for the sake of
1523uniformity, brevity, and internationalization. If, however, either the
1524userinfo or port subcomponents are non-empty, then the host should be
1525given explicitly even if it matches the default.</t>
1526
1527<t>Normalization should not remove delimiters when their associated
1528component is empty unless it is licensed to do so by the scheme
1529specification. For example, the IRI "http://example.com/?" cannot be
1530assumed to be equivalent to any of the examples above. Likewise, the
1531presence or absence of delimiters within a userinfo subcomponent is
1532usually significant to its interpretation.  The fragment component is
1533not subject to any scheme-based normalization; thus, two IRIs that
1534differ only by the suffix "#" are considered different regardless of
1535the scheme.</t>
1536 
1537<t>Some IRI schemes allow the usage of Internationalized Domain
1538Names (IDN) <xref target='RFC5890'></xref> either in their ireg-name
1539part or elswhere. When in use in IRIs, those names SHOULD
1540conform to the definition of U-Label in <xref
1541target='RFC5890'></xref>. An IRI containing an invalid IDN cannot
1542successfully be resolved. For legibility purposes, they
1543SHOULD NOT be converted into ASCII Compatible Encoding (ACE).</t>
1544
1545<t>Scheme-based normalization may also consider IDN
1546components and their conversions to punycode as equivalent. As an
1547example, "http://r&amp;#xE9;sum&amp;#xE9;.example.org" may be
1548considered equivalent to
1549"http://xn--rsum-bpad.example.org".</t><t>Other scheme-specific
1550normalizations are possible.</t>
1551
1552</section> <!-- schemenorm -->
1553
1554<section title="Protocol-Based Normalization">
1555
1556<t>Substantial effort to reduce the incidence of false negatives is
1557often cost-effective for web spiders. Consequently, they implement
1558even more aggressive techniques in IRI comparison. For example, if
1559they observe that an IRI such as</t>
1560
1561<figure><artwork>
1562   http://example.com/data</artwork></figure>
1563<t>redirects to an IRI differing only in the trailing slash</t>
1564<figure><artwork>
1565   http://example.com/data/</artwork></figure>
1566
1567<t>they will likely regard the two as equivalent in the future.  This
1568kind of technique is only appropriate when equivalence is clearly
1569indicated by both the result of accessing the resources and the common
1570conventions of their scheme's dereference algorithm (in this case, use
1571of redirection by HTTP origin servers to avoid problems with relative
1572references).</t>
1573
1574</section> <!-- protonorm -->
1575</section> <!-- equivalence -->
1576</section> 
1577
1578<section title="Use of IRIs" anchor="IRIuse">
1579
1580<section title="Limitations on UCS Characters Allowed in IRIs" anchor="limitations">
1581
1582<t>This section discusses limitations on characters and character
1583sequences usable for IRIs beyond those given in <xref target="abnf"/>
1584and <xref target="visual"/>. The considerations in this section are
1585relevant when IRIs are created and when URIs are converted to
1586IRIs.</t>
1587
1588<t>
1589
1590<list style="hanging"><t hangText="a.">The repertoire of characters allowed
1591    in each IRI component is limited by the definition of that component.
1592    For example, the definition of the scheme component does not allow
1593    characters beyond US-ASCII.
1594    <vspace blankLines="1"/>
1595    (Note: In accordance with URI practice, generic IRI
1596    software cannot and should not check for such limitations.)</t>
1597
1598<t hangText="b.">The UCS contains many areas of characters for which
1599    there are strong visual look-alikes. Because of the likelihood of
1600    transcription errors, these also should be avoided. This includes
1601    the full-width equivalents of Latin characters, half-width
1602    Katakana characters for Japanese, and many others. It also
1603    includes many look-alikes of "space", "delims", and "unwise",
1604    characters excluded in <xref target="RFC3491"/>.</t>
1605   
1606</list>
1607</t>
1608
1609<t>Additional information is available from <xref target="UNIXML"/>.
1610    <xref target="UNIXML"/> is written in the context of running text
1611    rather than in that of identifiers. Nevertheless, it discusses
1612    many of the categories of characters not appropriate for IRIs.</t>
1613</section> <!-- limitations -->
1614
1615<section title="Software Interfaces and Protocols">
1616
1617<t>Although an IRI is defined as a sequence of characters, software
1618interfaces for URIs typically function on sequences of octets or other
1619kinds of code units. Thus, software interfaces and protocols MUST
1620define which character encoding is used.</t>
1621
1622<t>Intermediate software interfaces between IRI-capable components and
1623URI-only components MUST map the IRIs per <xref target="mapping"/>,
1624when transferring from IRI-capable to URI-only components.
1625
1626This mapping SHOULD be applied as late as possible. It SHOULD NOT be
1627applied between components that are known to be able to handle IRIs.</t>
1628</section> <!-- software -->
1629
1630<section title="Format of URIs and IRIs in Documents and Protocols">
1631
1632<t>Document formats that transport URIs may have to be upgraded to allow
1633the transport of IRIs. In cases where the document as a whole
1634has a native character encoding, IRIs MUST also be encoded in this
1635character encoding and converted accordingly by a parser or interpreter.
1636
1637IRI characters not expressible in the native character encoding SHOULD
1638be escaped by using the escaping conventions of the document format if
1639such conventions are available. Alternatively, they MAY be
1640percent-encoded according to <xref target="mapping"/>. For example, in
1641HTML or XML, numeric character references SHOULD be used. If a
1642document as a whole has a native character encoding and that character
1643encoding is not UTF-8, then IRIs MUST NOT be placed into the document
1644in the UTF-8 character encoding.</t>
1645
1646<t>((UPDATE THIS NOTE)) Note: Some formats already accommodate IRIs,
1647although they use different terminology. HTML 4.0 <xref
1648target="HTML4"/> defines the conversion from IRIs to URIs as
1649error-avoiding behavior. XML 1.0 <xref target="XML1"/>, XLink <xref
1650target="XLink"/>, XML Schema <xref target="XMLSchema"/>, and
1651specifications based upon them allow IRIs. Also, it is expected that
1652all relevant new W3C formats and protocols will be required to handle
1653IRIs <xref target="CharMod"/>.</t>
1654
1655</section> <!-- format -->
1656
1657<section title="Use of UTF-8 for Encoding Original Characters" anchor="UTF8use">
1658
1659<t>This section discusses details and gives examples for point c) in
1660<xref target="Applicability"/>. To be able to use IRIs, the URI
1661corresponding to the IRI in question has to encode original characters
1662into octets by using UTF-8.  This can be specified for all URIs of a
1663URI scheme or can apply to individual URIs for schemes that do not
1664specify how to encode original characters.  It can apply to the whole
1665URI, or only to some part. For background information on encoding
1666characters into URIs, see also Section 2.5 of <xref
1667target="RFC3986"/>.</t>
1668
1669<t>For new URI schemes, using UTF-8 is recommended in <xref
1670target="RFC4395bis"/>.  Examples where UTF-8 is already used are the URN
1671syntax <xref target="RFC2141"/>, IMAP URLs <xref target="RFC2192"/>,
1672and POP URLs <xref target="RFC2384"/>.  On the other hand, because the
1673HTTP URI scheme does not specify how to encode original characters,
1674only some HTTP URLs can have corresponding but different IRIs.</t>
1675
1676<t>For example, for a document with a URI
1677of<vspace/>"http://www.example.org/r%C3%A9sum%C3%A9.html", it is
1678possible to construct a corresponding IRI (in XML notation, see <xref
1679target="sec-Notation"/>):
1680"http://www.example.org/r&amp;#xE9;sum&amp;#xE9;.html" ("&amp;#xE9;"
1681stands for the e-acute character, and "%C3%A9" is the UTF-8 encoded
1682and percent-encoded representation of that character). On the other
1683hand, for a document with a URI of
1684"http://www.example.org/r%E9sum%E9.html", the percent-encoding octets
1685cannot be converted to actual characters in an IRI, as the
1686percent-encoding is not based on UTF-8.</t>
1687
1688<t>For most URI schemes, there is no need to upgrade their scheme
1689definition in order for them to work with IRIs.  The main case where
1690upgrading makes sense is when a scheme definition, or a particular
1691component of a scheme, is strictly limited to the use of US-ASCII
1692characters with no provision to include non-ASCII characters/octets
1693via percent-encoding, or if a scheme definition currently uses highly
1694scheme-specific provisions for the encoding of non-ASCII characters.
1695An example of this is the mailto: scheme <xref target="RFC2368"/>.</t>
1696
1697<t>This specification updates the IANA registry of URI schemes to note
1698their applicability to IRIs, see <xref target="iana"/>.  All IRIs use
1699URI schemes, and all URIs with URI schemes can be used as IRIs, even
1700though in some cases only by using URIs directly as IRIs, without any
1701conversion.</t>
1702
1703<t>Scheme definitions can impose restrictions on the syntax of
1704scheme-specific URIs; i.e., URIs that are admissible under the generic
1705URI syntax <xref target="RFC3986"/> may not be admissible due to
1706narrower syntactic constraints imposed by a URI scheme
1707specification. URI scheme definitions cannot broaden the syntactic
1708restrictions of the generic URI syntax; otherwise, it would be
1709possible to generate URIs that satisfied the scheme-specific syntactic
1710constraints without satisfying the syntactic constraints of the
1711generic URI syntax. However, additional syntactic constraints imposed
1712by URI scheme specifications are applicable to IRI, as the
1713corresponding URI resulting from the mapping defined in <xref
1714target="mapping"/> MUST be a valid URI under the syntactic
1715restrictions of generic URI syntax and any narrower restrictions
1716imposed by the corresponding URI scheme specification.</t>
1717
1718<t>The requirement for the use of UTF-8 generally applies to all parts
1719of a URI.  However, it is possible that the capability of IRIs to
1720represent a wide range of characters directly is used just in some
1721parts of the IRI (or IRI reference). The other parts of the IRI may
1722only contain US-ASCII characters, or they may not be based on
1723UTF-8. They may be based on another character encoding, or they may
1724directly encode raw binary data (see also <xref
1725target="RFC2397"/>). </t>
1726
1727<t>For example, it is possible to have a URI reference
1728of<vspace/>"http://www.example.org/r%E9sum%E9.xml#r%C3%A9sum%C3%A9",
1729where the document name is encoded in iso-8859-1 based on server
1730settings, but where the fragment identifier is encoded in UTF-8 according
1731to <xref target="XPointer"/>. The IRI corresponding to the above
1732URI would be (in XML notation)<vspace/>"http://www.example.org/r%E9sum%E9.xml#r&amp;#xE9;sum&amp;#xE9;".</t>
1733
1734<t>Similar considerations apply to query parts. The functionality
1735of IRIs (namely, to be able to include non-ASCII characters) can
1736only be used if the query part is encoded in UTF-8.</t>
1737
1738</section> <!-- utf8 -->
1739
1740<section title="Relative IRI References">
1741<t>Processing of relative IRI references against a base is handled
1742straightforwardly; the algorithms of <xref target="RFC3986"/> can
1743be applied directly, treating the characters additionally allowed
1744in IRI references in the same way that unreserved characters are in URI
1745references.</t>
1746
1747</section> <!-- relative -->
1748</section> <!-- IRIuse -->
1749
1750<section title="Liberal handling of otherwise invalid IRIs" anchor="LEIRIHREF">
1751
1752<t>(EDITOR NOTE: This Section may move to an appendix.)
1753 
1754Some technical specifications and widely-deployed software have
1755allowed additional variations and extensions of IRIs to be used in
1756syntactic components. This section describes two widely-used
1757preprocessing agreements. Other technical specifications may wish to
1758reference a syntactic component which is "a valid IRI or a string that
1759will map to a valid IRI after this preprocessing algorithm". These two
1760variants are known as <xref target="LEIRI">Legacy Extended IRI or
1761LEIRI</xref>, and <xref target="HTML5">Web Address</xref>).
1762</t>
1763
1764<t>Future technical specifications SHOULD NOT allow conforming
1765producers to produce, or conforming content to contain, such forms,
1766as they are not interoperable with other IRI consuming software.</t>
1767
1768<section title="LEIRI processing"  anchor="LEIRIspec">
1769  <t>This section defines Legacy Extended IRIs (LEIRIs).
1770    The syntax of Legacy Extended IRIs is the same as that for IRIs,
1771    except that the ucschar production is replaced by the leiri-ucschar production:</t>
1772<figure>
1773
1774<artwork>
1775  leiri-ucschar  = " " / "&lt;" / "&gt;" / '"' / "{" / "}" / "|"
1776                   / "\" / "^" / "`" / %x0-1F / %x7F-D7FF
1777                   / %xE000-FFFD / %x10000-10FFFF
1778</artwork>
1779
1780<postamble>
1781  Among other extensions, processors based on this specification also
1782  did not enforce the restriction on bidirectional formatting
1783  characters in <xref target="visual"></xref>, and the iprivate
1784  production becomes redundant.</postamble>
1785</figure>
1786
1787<t>To convert a string allowed as a LEIRI to an IRI, each character
1788allowed in leiri-ucschar but not in ucschar must be percent-encoded
1789using <xref target="compmapping"/>.</t>
1790</section> <!-- leiriproc -->
1791
1792<section title="Web Address processing" anchor="webaddress">
1793
1794<t>Many popular web browsers have taken the approach of being quite
1795liberal in what is accepted as a "URL" or its relative
1796forms. This section describes their behavior in terms of a preprocessor
1797which maps strings into the IRI space for subsequent parsing and
1798interpretation as an IRI.</t>
1799
1800<t>In some situations, it might be appropriate to describe the syntax
1801that a liberal consumer implementation might accept as a "Web
1802Address" or "Hypertext Reference" or "HREF". However,
1803technical specifications SHOULD restrict the syntactic form allowed by compliant producers
1804to the IRI or IRI reference syntax defined in this document
1805even if they want to mandate this processing.</t>
1806
1807<t>
1808Summary:
1809<list style="symbols">
1810   <t>Leading and trailing whitespace is removed.</t>
1811   <t>Some additional characters are removed.</t>
1812   <t>Some additional characters are allowed and escaped (as with LEIRI).</t>
1813   <t>If interpreting an IRI as a URI, the pct-encoding of the query
1814   component of the parsed URI component depends on operational
1815   context.</t>
1816</list>
1817</t>
1818
1819<t>Each string provided may have an associated charset (called
1820the HREF-charset here); this defaults to UTF-8.
1821For web browsers interpreting HTML, the document
1822charset of a string is determined:
1823
1824<list style="hanging">
1825<t hangText="If the string came from a script (e.g. as an argument to
1826 a method)">The HRef-charset is the script's charset.</t>
1827
1828<t hangText="If the string came from a DOM node (e.g. from an
1829  element)">The node has a Document, and the HRef-charset is the
1830  Document's character encoding.</t>
1831
1832<t hangText="If the string had a HRef-charset defined when the string was
1833created or defined">The HRef-charset is as defined.</t>
1834
1835</list></t>
1836
1837<t>If the resulting HRef-charset is a unicode based character encoding
1838(e.g., UTF-16), then use UTF-8 instead.</t>
1839
1840
1841<figure>
1842<preamble>The syntax for Web Addresses is obtained by replacing the 'ucschar',
1843  pct-form, and path-sep rules with the href-ucschar, href-pct-form, and href-path-sep
1844  rules below. In addition, some characters are stripped.</preamble>
1845
1846<artwork type='abnf'>
1847  href-ucschar  = " " / "&lt;" / "&gt;" / DQUOTE / "{" / "}" / "|"
1848                   / "\" / "^" / "`" / %x0-1F / %x7F-D7FF
1849                   / %xE000-FFFD / %x10000-10FFFF
1850  href-pct-form = pct-encoded / "%"
1851  href-path-sep = "/" / "\"
1852  href-strip    = &lt;to be done&gt;
1853</artwork>
1854
1855<postamble>
1856(NOTE: NEED TO FIX THESE SETS TO MATCH HTML5; NOT SURE ABOUT NEXT SENTENCE)
1857browsers did not enforce the restriction on bidirectional formatting
1858  characters in <xref target="visual"></xref>, and the iprivate
1859  production becomes redundant.</postamble>
1860</figure>
1861
1862<t>'Web Address processing' requires the following additional
1863preprocessing steps:
1864
1865<list style="numbers">
1866
1867<t>Leading and trailing instances of space (U+0020),
1868CR (U+000A), LF (U+000D), and TAB (U+0009) characters are removed.</t>
1869
1870<t>strip all characters in href-strip.</t>
1871  <t>Percent-encode all characters in href-ucschar not in ucschar.</t>
1872  <t>Replace occurrences of "%" not followed by two hexadecimal digits by "%25".</t>
1873  <t>Convert backslashes ('\') matching href-path-sep to forward slashes ('/').</t>
1874</list></t>
1875</section> <!-- webaddress -->
1876
1877<section title="Characters not allowed in IRIs" anchor="notAllowed">
1878
1879<t>This section provides a list of the groups of characters and code
1880points that are allowed by LEIRI or HREF but are not allowed in IRIs or are
1881allowed in IRIs only in the query part. For each group of characters,
1882advice on the usage of these characters is also given, concentrating
1883on the reasons for why they are excluded from IRI use.</t>
1884
1885<t>
1886
1887<list><t>Space (U+0020): Some formats and applications use space as a
1888delimiter, e.g. for items in a list. Appendix C of <xref
1889target="RFC3986"></xref> also mentions that white space may have to be
1890added when displaying or printing long URIs; the same applies to long
1891IRIs. This means that spaces can disappear, or can make the what is
1892intended as a single IRI or IRI reference to be treated as two or more
1893separate IRIs.</t>
1894
1895<t>Delimiters "&lt;" (U+003C), "&gt;" (U+003E), and '"' (U+0022):
1896Appendix C of <xref target="RFC3986"></xref> suggests the use of
1897double-quotes ("http://example.com/") and angle brackets
1898(&lt;http://example.com/&gt;) as delimiters for URIs in plain
1899text. These conventions are often used, and also apply to IRIs.  Using
1900these characters in strings intended to be IRIs would result in the
1901IRIs being cut off at the wrong place.</t>
1902
1903<t>Unwise characters "\" (U+005C), "^" (U+005E), "`"
1904(U+0060), "{" (U+007B), "|" (U+007C), and "}" (U+007D): These
1905characters originally have been excluded from URIs because the
1906respective codepoints are assigned to different graphic characters in
1907some 7-bit or 8-bit encoding. Despite the move to Unicode, some of
1908these characters are still occasionally displayed differently on some
1909systems, e.g. U+005C may appear as a Japanese Yen symbol on some
1910systems. Also, the fact that these characters are not used in URIs or
1911IRIs has encouraged their use outside URIs or IRIs in contexts that
1912may include URIs or IRIs. If a string with such a character were used
1913as an IRI in such a context, it would likely be interpreted
1914piecemeal.</t>
1915
1916<t>The controls (C0 controls, DEL, and C1 controls, #x0 - #x1F #x7F -
1917#x9F): There is generally no way to transmit these characters reliably
1918as text outside of a charset encoding.  Even when in encoded form,
1919many software components silently filter out some of these characters,
1920or may stop processing alltogether when encountering some of
1921them. These characters may affect text display in subtle, unnoticable
1922ways or in drastic, global, and irreversible ways depending on the
1923hardware and software involved. The use of some of these characters
1924would allow malicious users to manipulate the display of an IRI and
1925its context in many situations.</t>
1926
1927<t>Bidi formatting characters (U+200E, U+200F, U+202A-202E): These
1928characters affect the display ordering of characters. If IRIs were
1929allowed to contain these characters and the resulting visual display
1930transcribed. they could not be converted back to electronic form
1931(logical order) unambiguously. These characters, if allowed in IRIs,
1932might allow malicious users to manipulate the display of IRI and its
1933context.</t>
1934
1935<t>Specials (U+FFF0-FFFD): These code points provide functionality
1936beyond that useful in an IRI, for example byte order identification,
1937annotation, and replacements for unknown characters and objects. Their
1938use and interpretation in an IRI would serve no purpose and might lead
1939to confusing display variations.</t>
1940
1941<t>Private use code points (U+E000-F8FF, U+F0000-FFFFD,
1942U+100000-10FFFD): Display and interpretation of these code points is
1943by definition undefined without private agreement. Therefore, these
1944code points are not suited for use on the Internet. They are not
1945interoperable and may have unpredictable effects.</t>
1946
1947<t>Tags (U+E0000-E0FFF): These characters provide a way to language
1948tag in Unicode plain text. They are not appropriate for IRIs because
1949language information in identifiers cannot reliably be input,
1950transmitted (e.g. on a visual medium such as paper), or
1951recognized.</t>
1952
1953<t>Non-characters (U+FDD0-FDEF, U+1FFFE-1FFFF, U+2FFFE-2FFFF,
1954U+3FFFE-3FFFF, U+4FFFE-4FFFF, U+5FFFE-5FFFF, U+6FFFE-6FFFF,
1955U+7FFFE-7FFFF, U+8FFFE-8FFFF, U+9FFFE-9FFFF, U+AFFFE-AFFFF,
1956U+BFFFE-BFFFF, U+CFFFE-CFFFF, U+DFFFE-DFFFF, U+EFFFE-EFFFF,
1957U+FFFFE-FFFFF, U+10FFFE-10FFFF): These code points are defined as
1958non-characters. Applications may use some of them internally, but are
1959not prepared to interchange them.</t>
1960
1961</list></t>
1962
1963<t>LEIRI preprocessing disallowed some code points and
1964code units:
1965
1966<list><t>Surrogate code units (D800-DFFF): These do not represent
1967Unicode codepoints.</t></list></t>
1968</section> <!-- notallowed -->
1969</section> <!-- lieirihref -->
1970 
1971<section title="URI/IRI Processing Guidelines (Informative)" anchor="guidelines">
1972
1973<t>This informative section provides guidelines for supporting IRIs in
1974the same software components and operations that currently process
1975URIs: Software interfaces that handle URIs, software that allows users
1976to enter URIs, software that creates or generates URIs, software that
1977displays URIs, formats and protocols that transport URIs, and software
1978that interprets URIs. These may all require modification before
1979functioning properly with IRIs. The considerations in this section
1980also apply to URI references and IRI references.</t>
1981
1982<section title="URI/IRI Software Interfaces">
1983<t>Software interfaces that handle URIs, such as URI-handling APIs and
1984protocols transferring URIs, need interfaces and protocol elements
1985that are designed to carry IRIs.</t>
1986
1987<t>In case the current handling in an API or protocol is based on
1988US-ASCII, UTF-8 is recommended as the character encoding for IRIs, as
1989it is compatible with US-ASCII, is in accordance with the
1990recommendations of <xref target="RFC2277"/>, and makes converting to
1991URIs easy. In any case, the API or protocol definition must clearly
1992define the character encoding to be used.</t>
1993
1994<t>The transfer from URI-only to IRI-capable components requires no
1995mapping, although the conversion described in <xref
1996target="URItoIRI"/> above may be performed. It is preferable not to
1997perform this inverse conversion unless it is certain this can be done
1998correctly.</t>
1999</section>
2000
2001<section title="URI/IRI Entry">
2002
2003<t>Some components allow users to enter URIs into the system
2004by typing or dictation, for example. This software must be updated to allow
2005for IRI entry.</t>
2006
2007<t>A person viewing a visual representation of an IRI (as a sequence
2008of glyphs, in some order, in some visual display) or hearing an IRI
2009will use an entry method for characters in the user's language to
2010input the IRI. Depending on the script and the input method used, this
2011may be a more or less complicated process.</t>
2012
2013<t>The process of IRI entry must ensure, as much as possible, that the
2014restrictions defined in <xref target="abnf"/> are met. This may be
2015done by choosing appropriate input methods or variants/settings
2016thereof, by appropriately converting the characters being input, by
2017eliminating characters that cannot be converted, and/or by issuing a
2018warning or error message to the user.</t>
2019
2020<t>As an example of variant settings, input method editors for East
2021Asian Languages usually allow the input of Latin letters and related
2022characters in full-width or half-width versions. For IRI input, the
2023input method editor should be set so that it produces half-width Latin
2024letters and punctuation and full-width Katakana.</t>
2025
2026<t>An input field primarily or solely used for the input of URIs/IRIs
2027might allow the user to view an IRI as it is mapped to a URI.  Places
2028where the input of IRIs is frequent may provide the possibility for
2029viewing an IRI as mapped to a URI. This will help users when some of
2030the software they use does not yet accept IRIs.</t>
2031
2032<t>An IRI input component interfacing to components that handle URIs,
2033but not IRIs, must map the IRI to a URI before passing it to these
2034components.</t>
2035
2036<t>For the input of IRIs with right-to-left characters, please see
2037<xref target="bidiInput"></xref>.</t>
2038</section>
2039
2040<section title="URI/IRI Transfer between Applications">
2041
2042<t>Many applications (for example, mail user agents) try to detect
2043URIs appearing in plain text. For this, they use some heuristics based
2044on URI syntax. They then allow the user to click on such URIs and
2045retrieve the corresponding resource in an appropriate (usually
2046scheme-dependent) application.</t>
2047
2048<t>Such applications would need to be upgraded, in order to use the
2049IRI syntax as a base for heuristics. In particular, a non-ASCII
2050character should not be taken as the indication of the end of an IRI.
2051Such applications also would need to make sure that they correctly
2052convert the detected IRI from the character encoding of the document
2053or application where the IRI appears, to the character encoding used
2054by the system-wide IRI invocation mechanism, or to a URI (according to
2055<xref target="mapping"/>) if the system-wide invocation mechanism only
2056accepts URIs.</t>
2057
2058<t>The clipboard is another frequently used way to transfer URIs and
2059IRIs from one application to another. On most platforms, the clipboard
2060is able to store and transfer text in many languages and scripts.
2061Correctly used, the clipboard transfers characters, not octets, which
2062will do the right thing with IRIs.</t>
2063</section>
2064
2065<section title="URI/IRI Generation">
2066
2067<t>Systems that offer resources through the Internet, where those
2068resources have logical names, sometimes automatically generate URIs
2069for the resources they offer. For example, some HTTP servers can
2070generate a directory listing for a file directory and then respond to
2071the generated URIs with the files.</t>
2072
2073<t>Many legacy character encodings are in use in various file systems.
2074Many currently deployed systems do not transform the local character
2075representation of the underlying system before generating URIs.</t>
2076
2077<t>For maximum interoperability, systems that generate resource
2078identifiers should make the appropriate transformations. For example,
2079if a file system contains a file named
2080"r&amp;#xE9;sum&amp;#xE9;.html", a server should expose this as
2081"r%C3%A9sum%C3%A9.html" in a URI, which allows use of
2082"r&amp;#xE9;sum&amp;#xE9;.html" in an IRI, even if locally the file
2083name is kept in a character encoding other than UTF-8.
2084</t>
2085
2086<t>This recommendation particularly applies to HTTP servers. For FTP
2087servers, similar considerations apply; see <xref target="RFC2640"/>.</t>
2088</section>
2089
2090<section title="URI/IRI Selection" anchor="selection">
2091<t>In some cases, resource owners and publishers have control over the
2092IRIs used to identify their resources. This control is mostly
2093executed by controlling the resource names, such as file names,
2094directly.</t>
2095
2096<t>In these cases, it is recommended to avoid choosing IRIs that are
2097easily confused. For example, for US-ASCII, the lower-case ell ("l") is
2098easily confused with the digit one ("1"), and the upper-case oh ("O") is
2099easily confused with the digit zero ("0"). Publishers should avoid
2100confusing users with "br0ken" or "1ame" identifiers.</t>
2101
2102<t>Outside the US-ASCII repertoire, there are many more opportunities for
2103confusion; a complete set of guidelines is too lengthy to include
2104here. As long as names are limited to characters from a single script,
2105native writers of a given script or language will know best when
2106ambiguities can appear, and how they can be avoided. What may look
2107ambiguous to a stranger may be completely obvious to the average
2108native user. On the other hand, in some cases, the UCS contains
2109variants for compatibility reasons; for example, for typographic purposes.
2110These should be avoided wherever possible. Although there may be exceptions,
2111newly created resource names should generally be in NFKC
2112<xref target="UTR15"></xref> (which means that they are also in NFC).</t>
2113
2114<t>As an example, the UCS contains the "fi" ligature at U+FB01
2115for compatibility reasons.
2116Wherever possible, IRIs should use the two letters "f" and "i" rather
2117than the "fi" ligature. An example where the latter may be used is
2118in the query part of an IRI for an explicit search for a word written
2119containing the "fi" ligature.</t>
2120
2121<t>In certain cases, there is a chance that characters from different
2122scripts look the same. The best known example is the similarity of the
2123Latin "A", the Greek "Alpha", and the Cyrillic "A". To avoid such
2124cases, IRIs should only be created where all the characters in a
2125single component are used together in a given language. This usually
2126means that all of these characters will be from the same script, but
2127there are languages that mix characters from different scripts (such
2128as Japanese).  This is similar to the heuristics used to distinguish
2129between letters and numbers in the examples above. Also, for Latin,
2130Greek, and Cyrillic, using lowercase letters results in fewer
2131ambiguities than using uppercase letters would.</t>
2132</section>
2133
2134<section title="Display of URIs/IRIs" anchor="display">
2135<t>
2136In situations where the rendering software is not expected to display
2137non-ASCII parts of the IRI correctly using the available layout and font
2138resources, these parts should be percent-encoded before being displayed.</t>
2139
2140<t>For display of Bidi IRIs, please see <xref target="visual"/>.</t>
2141</section>
2142
2143<section title="Interpretation of URIs and IRIs">
2144<t>Software that interprets IRIs as the names of local resources should
2145accept IRIs in multiple forms and convert and match them with the
2146appropriate local resource names.</t>
2147
2148<t>First, multiple representations include both IRIs in the native
2149character encoding of the protocol and also their URI counterparts.</t>
2150
2151<t>Second, it may include URIs constructed based on character
2152encodings other than UTF-8. These URIs may be produced by user agents that do
2153not conform to this specification and that use legacy character encodings to
2154convert non-ASCII characters to URIs. Whether this is necessary, and what
2155character encodings to cover, depends on a number of factors, such as
2156the legacy character encodings used locally and the distribution of
2157various versions of user agents. For example, software for Japanese
2158may accept URIs in Shift_JIS and/or EUC-JP in addition to UTF-8.</t>
2159
2160<t>Third, it may include additional mappings to be more user-friendly
2161and robust against transmission errors. These would be similar to how
2162some servers currently treat URIs as case insensitive or perform
2163additional matching to account for spelling errors. For characters
2164beyond the US-ASCII repertoire, this may, for example, include
2165ignoring the accents on received IRIs or resource names. Please note
2166that such mappings, including case mappings, are language
2167dependent.</t>
2168
2169<t>It can be difficult to identify a resource unambiguously if too
2170many mappings are taken into consideration. However, percent-encoded
2171and not percent-encoded parts of IRIs can always be clearly distinguished.
2172Also, the regularity of UTF-8 (see <xref target="Duerst97"/>) makes the
2173potential for collisions lower than it may seem at first.</t>
2174</section>
2175
2176<section title="Upgrading Strategy">
2177<t>Where this recommendation places further constraints on software
2178for which many instances are already deployed, it is important to
2179introduce upgrades carefully and to be aware of the various
2180interdependencies.</t>
2181
2182<t>If IRIs cannot be interpreted correctly, they should not be created,
2183generated, or transported. This suggests that upgrading URI interpreting
2184software to accept IRIs should have highest priority.</t>
2185
2186<t>On the other hand, a single IRI is interpreted only by a single or
2187very few interpreters that are known in advance, although it may be
2188entered and transported very widely.</t>
2189
2190<t>Therefore, IRIs benefit most from a broad upgrade of software to be
2191able to enter and transport IRIs. However, before an
2192individual IRI is published, care should be taken to upgrade the corresponding
2193interpreting software in order to cover the forms expected to be
2194received by various versions of entry and transport software.</t>
2195
2196<t>The upgrade of generating software to generate IRIs instead of using a
2197local character encoding should happen only after the service is upgraded
2198to accept IRIs. Similarly, IRIs should only be generated when the service
2199accepts IRIs and the intervening infrastructure and protocol is known
2200to transport them safely.</t>
2201
2202<t>Software converting from URIs to IRIs for display should be upgraded
2203only after upgraded entry software has been widely deployed to the
2204population that will see the displayed result.</t>
2205
2206
2207<t>Where there is a free choice of character encodings, it is often
2208possible to reduce the effort and dependencies for upgrading to IRIs
2209by using UTF-8 rather than another encoding. For example, when a new
2210file-based Web server is set up, using UTF-8 as the character encoding
2211for file names will make the transition to IRIs easier. Likewise, when
2212a new Web form is set up using UTF-8 as the character encoding of the
2213form page, the returned query URIs will use UTF-8 as the character
2214encoding (unless the user, for whatever reason, changes the character
2215encoding) and will therefore be compatible with IRIs.</t>
2216
2217
2218<t>These recommendations, when taken together, will allow for the
2219extension from URIs to IRIs in order to handle characters other than
2220US-ASCII while minimizing interoperability problems. For
2221considerations regarding the upgrade of URI scheme definitions, see
2222<xref target="UTF8use"/>.</t>
2223
2224</section>
2225</section> <!-- guidelines -->
2226
2227<section title="IANA Considerations" anchor="iana">
2228
2229<t>RFC Editor and IANA note: Please Replace RFC XXXX with the
2230number of this document when it issues as an RFC. </t>
2231
2232<t>IANA maintains a registry of "URI schemes". A "URI scheme" also
2233serves an "IRI scheme". </t>
2234
2235<t>To clarify that the URI scheme registration process also applies to
2236IRIs, change the description of the "URI schemes" registry
2237header to say "[RFC4395] defines an IANA-maintained registry of URI
2238Schemes. These registries include the Permanent and Provisional URI
2239Schemes.  RFC XXXX updates this registry to designate that schemes may
2240also indicate their usability as IRI schemes.</t>
2241
2242<t> Update "per RFC 4395" to "per RFC 4395 and RFC XXXX".
2243</t>
2244
2245</section> <!-- IANA -->
2246   
2247<section title="Security Considerations" anchor="security">
2248<t>The security considerations discussed in <xref target="RFC3986"/>
2249also apply to IRIs. In addition, the following issues require
2250particular care for IRIs.</t>
2251<t>Incorrect encoding or decoding can lead to security problems.
2252In particular, some UTF-8 decoders do not check against overlong
2253byte sequences. As an example, a "/" is encoded with the byte 0x2F
2254both in UTF-8 and in US-ASCII, but some UTF-8 decoders also wrongly
2255interpret the sequence 0xC0 0xAF as a "/". A sequence such as "%C0%AF.."
2256may pass some security tests and then be interpreted
2257as "/.." in a path if UTF-8 decoders are fault-tolerant, if conversion
2258and checking are not done in the right order, and/or if reserved
2259characters and unreserved characters are not clearly distinguished.</t>
2260
2261<t>There are various ways in which "spoofing" can occur with IRIs.
2262"Spoofing" means that somebody may add a resource name that looks the
2263same or similar to the user, but that points to a different resource.
2264The added resource may pretend to be the real resource by looking
2265very similar but may contain all kinds of changes that may be
2266difficult to spot and that can cause all kinds of problems.
2267Most spoofing possibilities for IRIs are extensions of those for URIs.</t>
2268
2269<t>Spoofing can occur for various reasons. First, a user's normalization expectations or actual normalization
2270when entering an IRI or  transcoding an IRI from a legacy character
2271encoding do not match the normalization used on the
2272server side. Conceptually, this is no different from the problems
2273surrounding the use of case-insensitive web servers. For example,
2274a popular web page with a mixed-case name ("http://big.example.com/PopularPage.html")
2275might be "spoofed" by someone who is able to create "http://big.example.com/popularpage.html".
2276However, the use of unnormalized character sequences, and of additional
2277mappings for user convenience, may increase the chance for spoofing.
2278Protocols and servers that allow the creation of resources with
2279names that are not normalized are particularly vulnerable to such
2280attacks. This is an inherent
2281security problem of the relevant protocol, server, or resource
2282and is not specific to IRIs, but it is mentioned here for completeness.</t>
2283
2284<t>Spoofing can occur in various IRI components, such as the
2285domain name part or a path part. For considerations specific
2286to the domain name part, see <xref target="RFC3491"/>.
2287For the path part, administrators of sites that allow independent
2288users to create resources in the same sub area may have to be careful
2289to check for spoofing.</t>
2290
2291<t>Spoofing can occur because in the UCS many characters look very similar. Details are discussed in <xref target="selection"/>.
2292Again, this is very similar to spoofing possibilities on US-ASCII,
2293e.g., using "br0ken" or "1ame" URIs.</t>
2294
2295<t>Spoofing can occur when URIs with percent-encodings based on various
2296character encodings are accepted to deal with older user agents. In some
2297cases, particularly for Latin-based resource names, this is usually easy to
2298detect because UTF-8-encoded names, when interpreted and viewed as
2299legacy character encodings, produce mostly garbage.</t><t>When
2300concurrently used character encodings have a similar structure but there
2301are no characters that have exactly the same encoding, detection is more
2302difficult.</t>
2303
2304<t>Spoofing can occur with bidirectional IRIs, if the restrictions
2305in <xref target="bidi-structure"/> are not followed. The same visual
2306representation may be interpreted as different logical representations,
2307and vice versa. It is also very important that a correct Unicode bidirectional
2308implementation be used.</t><t>The use of Legacy Extended IRIs introduces additional security issues.</t>
2309</section><!-- security -->
2310
2311<section title="Acknowledgements">
2312<t>For contributions to this update, we would like to thank Ian Hickson, Michael Sperberg-McQueen,
2313  Dan Connolly, Norman Walsh, Richard Tobin, Henry S. Thomson, and the XML Core Working Group of the W3C.</t>
2314
2315<t>The discussion on the issue addressed here started a long time
2316ago. There was a thread in the HTML working
2317group in August 1995 (under the topic of "Globalizing URIs") and in the
2318www-international mailing list in July 1996 (under the topic of
2319"Internationalization and URLs"), and there were ad-hoc meetings at the Unicode
2320conferences in September 1995 and September 1997.</t>
2321
2322<t>For contributions to the previous version of this document, RFC 3987, many thanks go to
2323Francois Yergeau, Matitiahu Allouche,
2324Roy Fielding, Tim Berners-Lee, Mark Davis,
2325M.T. Carrasco Benitez, James Clark, Tim Bray, Chris Wendt, Yaron Goland,
2326Andrea Vine, Misha Wolf, Leslie Daigle, Ted Hardie, Bill Fenner, Margaret Wasserman,
2327Russ Housley, Makoto MURATA, Steven Atkin,
2328Ryan Stansifer, Tex Texin, Graham Klyne, Bjoern Hoehrmann, Chris Lilley, Ian Jacobs,
2329Adam Costello, Dan Oscarson, Elliotte Rusty Harold, Mike J. Brown,
2330Roy Badami, Jonathan Rosenne, Asmus Freytag, Simon Josefsson, Carlos Viegas Damasio,
2331Chris Haynes, Walter Underwood, and many others.</t>
2332<t>A definition of HyperText Reference was initially produced by Ian Hixson,
2333and further edited by Dan Connolly and C. M. Spergerg-McQueen.</t>
2334<t>Thanks to the Internationalization Working
2335Group (I18N WG) of the World Wide Web Consortium (W3C),
2336and the members of the W3C
2337I18N Working Group and Interest Group for their contributions and their
2338work on <xref target="CharMod"/>. Thanks also go
2339to the members of many other W3C Working Groups for adopting IRIs, and to
2340the members of the Montreal IAB Workshop on Internationalization and
2341Localization for their review.</t>
2342</section>
2343
2344<section title="Main Changes Since RFC 3987">
2345  <t>This section describes the main changes since <xref target="RFC3987"></xref>.</t>
2346  <section title="Major restructuring of IRI processing model" anchor="forkChanges">
2347    <t>Major restructuring of IRI processing model to make scheme-specific translation
2348      necessary to handle IDNA requirements and for consistency with web implementations. </t>
2349    <t>Starting with IRI, you want one of:
2350      <list style="hanging">
2351        <t hangText="a"> IRI components (IRI parsed into UTF8 pieces)</t>
2352        <t hangText="b"> URI components (URI parsed into ASCII pieces, encoded correctly) </t>
2353        <t hangText="c"> whole URI  (for passing on to some other system that wants whole URIs) </t>
2354      </list></t>
2355   
2356    <section title="OLD WAY">
2357      <t><list style="numbers">
2358       
2359        <t>Pct-encoding on the whole thing to a URI.
2360          (c1) If you want a (maybe broken) whole URI, you might
2361          stop here.</t>
2362       
2363        <t>Parsing the URI into URI components.
2364          (b1) If you want (maybe broken) URI components, stop here.</t>
2365       
2366        <t> Decode the components (undoing the pct-encoding).
2367          (a) if you want IRI components, stop here.</t>
2368       
2369        <t> reencode:  Either using a different encoding some components
2370          (for domain names, and query components in web pages, which
2371          depends on the component, scheme and context), and otherwise
2372          using pct-encoding.
2373          (b2) if you want (good) URI components, stop here.</t>
2374       
2375        <t> reassemble the reencoded components.
2376          (c2) if you want a (*good*) whole URI stop here.</t>
2377      </list>
2378       
2379      </t>
2380     
2381    </section>
2382   
2383    <section title="NEW WAY">
2384      <t>
2385        <list style="numbers">
2386         
2387          <t> Parse the IRI into IRI components using the generic syntax.
2388            (a) if you want IRI components, stop here.</t>
2389         
2390          <t> Encode each components, using pct-encoding, IDN encoding, or
2391            special query part encoding depending on the component
2392            scheme or context. (b) If you want URI components, stop here.</t>
2393          <t> reassemble the a whole URI from URI components.
2394            (c) if you want a whole URI stop here.</t>
2395        </list></t>
2396    </section>
2397    <section title="Extension of Syntax">
2398      <t>Added the tag range (U+E0000-E0FFF) to the iprivate production.
2399        Some IRIs generated with the new syntax may fail to pass very strict checks
2400        relying on the old syntax. But characters in this range should be extremely infrequent
2401        anyway.</t>
2402    </section>
2403    <section title="More to be added"><t>TODO: There are more main changes that need to be
2404      documented in this section.</t></section>
2405</section>
2406
2407<section title="Change Log">
2408
2409<t>Note to RFC Editor: Please completely remove this section before publication.</t>
2410
2411<section title='Changes after draft-ietf-iri-3987bis-01'>
2412    <t>Changes from draft-ietf-iri-3987bis-01 onwards are available as changesets
2413      in the IETF tools subversion repository at
2414      http://trac.tools.ietf.org/wg/iri/trac/log/draft-ietf-iri-3987bis/draft-ietf-iri-3987bis.xml.</t>
2415</section>
2416 
2417<section title='Changes from draft-duerst-iri-bis-07 to draft-ietf-iri-3987bis-00'>
2418     <t>Changed draft name, date, last paragraph of abstract, and titles in change log, and added this section
2419     in moving from draft-duerst-iri-bis-07 (personal submission) to draft-ietf-iri-3987bis-00 (WG document).</t>
2420</section>
2421
2422<section title="Changes from -06 to -07 of draft-duerst-iri-bis">
2423  <t>Major restructuring of the processing model, see <xref target="forkChanges"></xref>.</t>
2424</section>
2425</section>
2426
2427<section title='Changes from -00 to -01'><t><list style="symbols">
2428  <t>Removed 'mailto:' before mail addresses of authors.</t>
2429  <t>Added "&lt;to be done&gt;" as right side of 'href-strip' rule. Fixed '|' to '/' for
2430    alternatives.</t>
2431</list></t>
2432</section>
2433
2434<section title="Changes from -05 to -06 of draft-duerst-iri-bis-00"><t><list style="symbols">
2435<t>Add HyperText Reference, change abstract, acks and references for it</t>
2436<t>Add Masinter back as another editor.</t>
2437<t>Masinter integrates HRef material from HTML5 spec.</t>
2438<t>Rewrite introduction sections to modernize.</t>
2439</list></t>
2440</section>
2441
2442<section title="Changes from -04 to -05 of draft-duerst-iri-bis">
2443  <t><list style="symbols">
2444    <t>Updated references.</t>
2445    <t>Changed IPR text to pre5378Trust200902.</t></list></t>
2446</section>
2447
2448<section title="Changes from -03 to -04 of draft-duerst-iri-bis">
2449  <t><list style="symbols">
2450    <t>Added explicit abbreviation for LEIRIs.</t>
2451    <t>Mentioned LEIRI references.</t>
2452    <t>Completed text in LEIRI section about tag characters and about specials.</t></list></t>
2453</section>
2454
2455<section title="Changes from -02 to -03 of draft-duerst-iri-bis">
2456  <t><list style="symbols">
2457    <t>Updated some references.</t>
2458    <t>Updated Michel Suginard's coordinates.</t></list></t>
2459</section>
2460
2461<section title="Changes from -01 to -02 of draft-duerst-iri-bis">
2462  <t><list style="symbols">
2463    <t>Added tag range to iprivate (issue private-include-tags-115).</t>
2464    <t>Added Specials (U+FFF0-FFFD) to Legacy Extended IRIs.</t></list></t>
2465</section>
2466<section title="Changes from -00 to -01 of draft-duerst-iri-bis">
2467  <t><list style="symbols">
2468    <t>Changed from "IRIs with Spaces/Controls" to "Legacy Extended IRI"
2469      based on input from the W3C XML Core WG.
2470      Moved the relevant subsections to the back and promoted them to a section.</t>
2471    <t>Added some text re. Legacy Extended IRIs to the security section.</t>
2472    <t>Added a IANA Consideration Section.</t>
2473    <t>Added this Change Log Section.</t>
2474    <t>Added a section about "IRIs with Spaces/Controls" (converting from a Note in RFC 3987).</t></list></t>
2475</section>
2476<section title="Changes from RFC 3987 to -00 of draft-duerst-iri-bis">
2477  <t><list>
2478    <t>Fixed errata (see http://www.rfc-editor.org/cgi-bin/errataSearch.pl?rfc=3987).</t></list></t>
2479</section>
2480</section>
2481</middle>
2482
2483<back>
2484<references title="Normative References">
2485
2486<reference anchor="ASCII">
2487<front>
2488<title>Coded Character Set -- 7-bit American Standard Code for Information
2489Interchange</title>
2490<author>
2491<organization>American National Standards Institute</organization>
2492</author>
2493<date year="1986"/>
2494</front>
2495<seriesInfo name="ANSI" value="X3.4"/>
2496</reference>
2497
2498<reference anchor="ISO10646">
2499<front>
2500<title>ISO/IEC 10646:2003: Information Technology -
2501Universal Multiple-Octet Coded Character Set (UCS)</title>
2502<author>
2503<organization>International Organization for Standardization</organization>
2504</author>
2505<date month="December" year="2003"/>
2506</front>
2507<seriesInfo name="ISO" value="Standard 10646"/>
2508</reference>
2509
2510&rfc2119;
2511&rfc3490;
2512&rfc3491;
2513&rfc3629;
2514&rfc3986;
2515
2516<reference anchor="STD68">
2517<front>
2518<title abbrev="ABNF">Augmented BNF for Syntax Specifications: ABNF</title>
2519<author initials="D." surname="Crocker" fullname="Dave Crocker"><organization/></author>
2520<author initials="P." surname="Overell" fullname="Paul Overell"><organization/></author>
2521<date month="January" year="2008"/></front>
2522<seriesInfo name="STD" value="68"/><seriesInfo name="RFC" value="5234"/>
2523</reference>
2524 
2525&rfc5890;
2526&rfc5891;
2527
2528<reference anchor="UNIV6">
2529<front>
2530<title>The Unicode Standard, Version 6.0.0 (Mountain View, CA, The Unicode Consortium, 2011, ISBN 978-1-936213-01-6)</title>
2531<author><organization>The Unicode Consortium</organization></author>
2532<date year="2010" month="October"/>
2533</front>
2534</reference>
2535
2536<reference anchor="UNI9" target="http://www.unicode.org/reports/tr9/tr9-13.html">
2537<front>
2538<title>The Bidirectional Algorithm</title>
2539<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2540<date year="2004" month="March"/>
2541</front>
2542<seriesInfo name="Unicode Standard Annex" value="#9"/>
2543</reference>
2544
2545<reference anchor="UTR15" target="http://www.unicode.org/unicode/reports/tr15/tr15-23.html">
2546<front>
2547<title>Unicode Normalization Forms</title>
2548<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2549<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2550<date year="2008" month="March"/>
2551</front>
2552<seriesInfo name="Unicode Standard Annex" value="#15"/>
2553</reference>
2554
2555</references>
2556
2557<references title="Informative References">
2558
2559<reference anchor="BidiEx" target="http://www.w3.org/International/iri-edit/BidiExamples">
2560<front>
2561<title>Examples of bidirectional IRIs</title>
2562<author><organization/></author>
2563<date year="" month=""/>
2564</front>
2565</reference>
2566
2567<reference anchor="CharMod" target="http://www.w3.org/TR/charmod-resid">
2568<front>
2569<title>Character Model for the World Wide Web: Resource Identifiers</title>
2570<author initials="M." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2571<author initials="F." surname="Yergeau" fullname="Francois Yergeau"><organization/></author>
2572<author initials="R." surname="Ishida" fullname="Richard Ishida"><organization/></author>
2573<author initials="M." surname="Wolf" fullname="Misha Wolf"><organization/></author>
2574<author initials="T." surname="Texin" fullname="Tex Texin"><organization/></author>
2575<date year="2004" month="November" day="25"/>
2576</front>
2577<seriesInfo name="World Wide Web Consortium" value="Candidate Recommendation"/>
2578</reference>
2579
2580<reference anchor="Duerst97" target="http://www.ifi.unizh.ch/mml/mduerst/papers/PDF/IUC11-UTF-8.pdf">
2581<front>
2582<title>The Properties and Promises of UTF-8</title>
2583<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2584<date year="1997" month="September"/>
2585</front>
2586<seriesInfo name="Proc. 11th International Unicode Conference, San Jose" value=""/>
2587</reference>
2588
2589<reference anchor="Gettys" target="http://www.w3.org/DesignIssues/ModelConsequences">
2590<front>
2591<title>URI Model Consequences</title>
2592<author initials="J." surname="Gettys" fullname="Jim Gettys"><organization/></author>
2593<date month="" year=""/>
2594</front>
2595</reference>
2596
2597<reference anchor="HTML4" target="http://www.w3.org/TR/html401/appendix/notes.html#h-B.2">
2598<front>
2599<title>HTML 4.01 Specification</title>
2600<author initials="D." surname="Raggett" fullname="Dave Raggett"><organization/></author>
2601<author initials="A." surname="Le Hors" fullname="Arnaud Le Hors"><organization/></author>
2602<author initials="I." surname="Jacobs" fullname="Ian Jacobs"><organization/></author>
2603<date year="1999" month="December" day="24"/>
2604</front>
2605<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2606</reference>
2607
2608<reference anchor="LEIRI" target="http://www.w3.org/TR/leiri/">
2609<front>
2610<title>Legacy extended IRIs for XML resource identification</title>
2611<author initials="H." surname="Thompson" fullname="Henry Thompson"><organization/></author>
2612<author initials="R." surname="Tobin"    fullname="Richard Tobin"><organization/></author>
2613<author initials="N." surname="Walsh" fullname="Norman Walsh"><organization/></author>
2614  <date year="2008" month="November" day="3"/>
2615
2616</front>
2617<seriesInfo name="World Wide Web Consortium" value="Note"/>
2618</reference>
2619
2620
2621&rfc2045;
2622&rfc2130;
2623&rfc2141;
2624&rfc2192;
2625&rfc2277;
2626&rfc2368;
2627&rfc2384;
2628&rfc2396;
2629&rfc2397;
2630&rfc2616;
2631&rfc1738;
2632&rfc2640;
2633&rfc3987;
2634<reference anchor='RFC4395bis'>
2635  <front>
2636    <title>Guidelines and Registration Procedures for New URI/IRI Schemes</title>
2637    <author initials='T.' surname='Hansen' fullname="Tony Hansen"><organization/></author>
2638    <author initials='T.' surname='Hardie' fullname="Ted Hardie"><organization/></author>
2639    <author initials='L.' surname='Masinter' fullname="Larry Masinter"><organization/></author>
2640    <date year="2010" month='September' day="30"/>
2641    <workgroup>IRI</workgroup>
2642  </front>
2643  <seriesInfo name="Internet-Draft" value="draft-hansen-iri-4395bis-irireg-00"/>
2644</reference>
2645 
2646 
2647<reference anchor="UNIXML" target="http://www.w3.org/TR/unicode-xml/">
2648<front>
2649<title>Unicode in XML and other Markup Languages</title>
2650<author initials="M.J." surname="Duerst" fullname="Martin Duerst"><organization/></author>
2651<author initials="A." surname="Freytag" fullname="Asmus Freytag"><organization/></author>
2652<date year="2003" month="June" day="18"/>
2653</front>
2654<seriesInfo name="Unicode Technical Report" value="#20"/>
2655<seriesInfo name="World Wide Web Consortium" value="Note"/>
2656</reference>
2657 
2658<reference anchor="UTR36" target="http://unicode.org/reports/tr36/">
2659<front>
2660<title>Unicode Security Considerations</title>
2661<author initials="M." surname="Davis" fullname="Mark Davis"><organization/></author>
2662<author initials="M." surname="Suignard" fullname="Michel Suignard"><organization/></author>
2663<date year="2010" month="August" day="4"/>
2664</front>
2665<seriesInfo name="Unicode Technical Report" value="#36"/>
2666</reference>
2667
2668<reference anchor="XLink" target="http://www.w3.org/TR/xlink/#link-locators">
2669<front>
2670<title>XML Linking Language (XLink) Version 1.0</title>
2671<author initials="S." surname="DeRose" fullname="Steve DeRose"><organization/></author>
2672<author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2673<author initials="D." surname="Orchard" fullname="David Orchard"><organization/></author>
2674<date year="2001" month="June" day="27"/>
2675</front>
2676<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2677</reference>
2678
2679<reference anchor="XML1" target="http://www.w3.org/TR/REC-xml">
2680  <front>
2681    <title>Extensible Markup Language (XML) 1.0 (Forth Edition)</title>
2682    <author initials="T." surname="Bray" fullname="Tim Bray"><organization/></author>
2683    <author initials="J." surname="Paoli" fullname="Jean Paoli"><organization/></author>
2684    <author initials="C.M." surname="Sperberg-McQueen" fullname="C. M. Sperberg-McQueen">
2685      <organization/></author>
2686    <author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2687    <author initials="F." surname="Yergeau" fullname="Francois Yergeau"><organization/></author>
2688    <date day="16" month="August" year="2006"/>
2689  </front>
2690  <seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2691</reference>
2692
2693<reference anchor="XMLNamespace" target="http://www.w3.org/TR/REC-xml-names">
2694  <front>
2695    <title>Namespaces in XML (Second Edition)</title>
2696    <author initials="T." surname="Bray" fullname="Tim Bray"><organization/></author>
2697    <author initials="D." surname="Hollander" fullname="Dave Hollander"><organization/></author>
2698    <author initials="A." surname="Layman" fullname="Andrew Layman"><organization/></author>
2699    <author initials="R." surname="Tobin" fullname="Richard Tobin"><organization></organization></author>
2700    <date day="16" month="August" year="2006"/>
2701  </front>
2702  <seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2703</reference>
2704
2705<reference anchor="XMLSchema" target="http://www.w3.org/TR/xmlschema-2/#anyURI">
2706<front>
2707<title>XML Schema Part 2: Datatypes</title>
2708<author initials="P." surname="Biron" fullname="Paul Biron"><organization/></author>
2709<author initials="A." surname="Malhotra" fullname="Ashok Malhotra"><organization/></author>
2710<date year="2001" month="May" day="2"/>
2711</front>
2712<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2713</reference>
2714
2715<reference anchor="XPointer" target="http://www.w3.org/TR/xptr-framework/#escaping">
2716<front>
2717<title>XPointer Framework</title>
2718<author initials="P." surname="Grosso" fullname="Paul Grosso"><organization/></author>
2719<author initials="E." surname="Maler" fullname="Eve Maler"><organization/></author>
2720<author initials="J." surname="Marsh" fullname="Jonathan Marsh"><organization/></author>
2721<author initials="N." surname="Walsh" fullname="Norman Walsh"><organization/></author>
2722<date year="2003" month="March" day="25"/>
2723</front>
2724<seriesInfo name="World Wide Web Consortium" value="Recommendation"/>
2725</reference>
2726
2727<reference anchor="HTML5" target="http://www.w3.org/TR/2009/WD-html5-20090423/">
2728<front>
2729<title>A vocabulary and associated APIs for HTML and XHTML</title>
2730<author initials="I." surname="Hickson" fullname="Ian Hickson"><organization>Google, Inc.</organization></author>
2731<author initials="D." surname="Hyatt" fullname="David Hyatt"><organization>Apple, Inc.</organization></author>
2732<date year="2009"  month="April" day="23"/>
2733</front>
2734<seriesInfo name="World Wide Web Consortium" value="Working Draft"/>
2735</reference>
2736
2737</references>
2738
2739</back>
2740</rfc>
Note: See TracBrowser for help on using the repository browser.