1.1 Goals and Scope

The goal of the Character Model for the World WideWeb is to facilitate use of the Web by all people,regardless of their language, script, writing system, and cultural conventions,in accordance with theW3Cgoal of universal access. One basic prerequisite to achieve this goalis to be able to transmit and process the characters used around the world in awell-defined and well-understood way.

The main target audience of this specificationis W3C specification developers. This specificationand parts of it can be referenced from other W3C specifications. It defines conformance criteria for W3C specificationsas well as other specifications.

Other audiences of this specificationinclude software developers, contentdevelopers, and authors of specifications outside the W3C. Software developersand content developers implement and use W3C specifications. Thisspecificationdefines some conformance criteria for implementations (software) and contentthat implement and use W3C specifications. It also helps software developers andcontent developers to understand the character-related provisions in W3Cspecifications.

The character model described in this specificationprovides authors ofspecifications, software developers, and content developers with a commonreference for consistent, interoperable text manipulation on the World Wide Web.Working together, these three groups can build a more international Web.

Topics addressed in this part of the Character Model for the World Wide Webinclude use of the terms 'character', 'encoding' and'string', a reference processing model, choice and identificationof character encodings, character escaping, and string indexing.

Other parts of the Character Model address normalization and string identity matching ([CharNorm]) and Internationalized Resource Identifiers (IRI) conventions([CharIRI]).

Topics as yet not addressed or barely touched include fuzzymatching, and language tagging. Some of these topics may be addressed in afuture version of this specification.

At the core of the model is the Universal Character Set (UCS), definedjointly by the Unicode Standard[Unicode] and ISO/IEC 10646[ISO/IEC 10646]. In this document, Unicode is used as asynonym for the Universal Character Set. The model will allow Web documentsauthored in the world's scripts (and on different platforms) to be exchanged,read, and searched by Web users around the world.

1.2 Background

This section provides some historical background on the topicsaddressed in this specification.

Starting withInternationalization of the Hypertext Markup Language[RFC 2070], the Web community has recognized the needfor a character model for the World Wide Web. The first step towards buildingthis model was the adoption of Unicode as the document character set for HTML.

The choice of Unicode was motivated by the fact that Unicode:

• is the only universal character repertoire available,

• provides a way of referencing characters independent of the encoding of the text,

• is being updated/completed carefully,

• is widely accepted and implemented by industry.

W3C adopted Unicode as the document character set for HTML in[HTML 4.0]. The same approach was later used for specifications such as XML 1.0[XML 1.0] and CSS2[CSS2]. W3C specifications andapplications now use Unicode as the common reference character set.

When data transfer on the Web remained mostly unidirectional (from server tobrowser), and where the main purpose was to render documents, the use of Unicodewithout specifying additional details was sufficient. However, the Web hasgrown:

• Data transfers among servers, proxies, and clients, in all directions, have increased.

• Characters outside the US-ASCII[ISO/IEC 646][MIME-charset] repertoireare being used in more and more places.

• Data transfers between different protocol/format elements (such as element/attribute names, URI components, and textual content) have increased.

• More and more APIs are defined, not just protocols and formats.

In short, the Web may be seen as a single, very large application (see[Nicol]), rather than as a collection of small independentapplications.

While these developments strengthen the requirement that Unicode be the basisof a character model for the Web, they also create the need for additionalspecifications on the application of Unicode to the Web. Some aspects of Unicodethat require additional specification for the Web include:

• Choice of Unicode encoding forms (UTF-8, UTF-16, UTF-32).

• Counting characters, measuring string length in the presence of variable-length character encodings and combining characters.

• Duplicate encodings of characters (e.g. precomposed vs decomposed).

• Use of control codes for various purposes (e.g. bidirectionality control, symmetric swapping, etc.).

It should be noted that such aspects also exist in variousencodings, and in many cases have been inherited by Unicodein one way or another from these encodings.

The remainder of this specification presentsadditional requirements to ensure an interoperable character model for the Web, taking intoaccount earlier work (from W3C, ISO and IETF).

The first few chapters of the Unicode Standard[Unicode]provide very useful background reading. The policies adopted by the IETF for onthe use of character sets on the Internet are documented in[RFC 2277].

1.3 Terminology and Notation

Unicode code points are denoted as U+hhhh, where "hhhh" is asequence of at least four, and at most six hexadecimal digits.

Text has been used for examples to allow them to be cut and pasted by thereader. Characters used will not appear as intended unless you have theappropriate font, but care has been taken to annotate the examples so that theyremain understandable even if you do not. In some cases it is important to seethe result of an example, so images have been used; by clicking on the image itis possible to link to the text for these examples inC Example text.

NOTE:RFC 2119 makes it clear that requirements that useSHOULD are not optional and must be complied with unless there are specific reasons not to: "This word, or the adjective "RECOMMENDED", mean that there may exist valid reasons in particular circumstances to ignore a particular item, but the full implications must be understood and carefully weighed before choosing a different course."

NOTE:Requirements placed on specifications might indirectly cause requirements to be placed on implementations or content that claim to conform to those specifications. Likewise, requirements placed on content may affect implementations designed to produce such content, and so on.

3.1 Introduction

The glossary entry in the Unicode Standard[Unicode 4.0] gives:

"Character. (1) The smallest component of written languagethat has semantic value; refers to the abstract meaning and/or shape..."

The word 'character' is used in many contexts, withdifferent meanings. Human cultures have radically differing writing systems,leading to radically differing concepts of a character. Such wide variation inend user experience can, and often does, result in misunderstanding. Thisvariation is sometimes mistakenly seen as the consequence of imperfecttechnology. Instead, it derives from the great flexibility and creativity ofthe human mind and the long tradition of writing as an important part of thehuman cultural heritage. The alphabetic approach used by scripts such as Latin,Cyrillic and Greek is only one of several possibilities.

The developers of specifications, and the developers ofsoftware based on those specifications, are likely to be more familiar withusages of the term 'character' they have experienced and less familiar with the wide variety of usages in an international context. Furthermore, within a computing context, characters are often confused with related concepts, resulting in incomplete or inappropriate specifications and software.

This section examines some of these contexts, meanings andconfusions.

3.2 Units of aural rendering

In some scripts, characters have a close relationship to phonemes(aphoneme is a minimally distinct sound in the context of aparticular spoken language), while in others they are closely related tomeanings. Even when characters (loosely) correspond to phonemes, thisrelationship may not be simple, and there is rarely a one-to-one correspondencebetween character and phoneme.

C001 [S] [I] [C] Specifications,software and contentMUST NOT require or depend on a one-to-onecorrespondence between characters and the sounds of alanguage.

3.3 Units of visualrendering

Visual rendering introduces the notion of aglyph.Glyphs are defined by ISO/IEC 9541-1[ISO/IEC 9541-1] as "a recognizable abstract graphic symbol whichis independent of a specific design". There isnot aone-to-one correspondence between characters and glyphs:

• A single character can be represented by multiple glyphs(each glyph is then part of the representation of that character). These glyphsmay be physically separated from one another.

• A single glyph may represent a sequence of characters (thisis the case with ligatures, among others).

• A character may be rendered with very different glyphsdepending on the context.

• A single glyph may represent different characters (e.g.capital Latin A, capital Greek A and capital Cyrillic A).

A set of glyphs makes up afont. Glyphs can beconstrued as the basic units of organization of the visual rendering of text,just as characters are the basic unit of organization of encoded text.

C002 [S] [I] [C] Specifications,software and contentMUST NOT require or depend on a one-to-one mapping betweencharacters and units of displayed text.

See the appendixB Examples of Characters, Keystrokes and Glyphs for examples ofthe complexities of character to glyph mapping.

3.3.1 Visual Rendering and Logical Order

Some scripts, in particular Arabic and Hebrew, are written fromright to left. Text including characters from these scripts can run in bothdirections and is therefore called bidirectional text. The Unicode Standard[Unicode] requires that characters be stored and interchanged inlogical order, i.e. roughly corresponding to the order in which textis typed in via the keyboard or spoken (for a more detailed definition see[Unicode 4.0], Section 2.2). Logical ordering isimportant to ensure interoperability of data, and also benefits accessibility,searching, and collation.

C003 [S] [I] [C] Protocols,data formats and APIsMUST store, interchange or processtext data in logical order.

In the presence of bidirectional text, two possible selectionmodes can be considered. The first islogical selection mode,which selects all the characterslogically located between theend-points of the user's mouse gesture. Here the user selects from between thefirst and second letters of the second word to the middle of the number.Logical selection looks like this:

Visual display
Logical order

Logical selection resulting in discontiguous visual ranges

It is a consequence of the bidirectionality of the text that asingle, continuous logical selection in memory results in adiscontinuousselection appearing on the screen. This discontinuity makes some users prefer avisual selection mode, which selects all the charactersvisually located between the end-points of the user's mousegesture. With the same mouse gesture as before, we now obtain:

Visual display
Logical order

Visual selection resulting in discontiguous logical ranges

In visual selection mode, as seen in the example above, a single visual selection range may result intwo or more logical ranges, which may have to be accommodated by protocols,APIs and implementations. Other, related aspects of a user interface for bidirectional text include caret movement, behavior of backspace/delete keys, and so on.

Currently, most implementations provide logical selection, while only very few provide visual selection.

C075 [I] Independent of whether some implementation uses logical selection or visual selection, characters selectedMUST be kept in logical order in storage.

C004 [S] Specifications of protocolsand APIs that involve selection of rangesSHOULD provide fordiscontiguous logical selections, at least to the extent necessary to supportimplementation of visual selection on screen on top of those protocols andAPIs.

3.4 Units of input

In keyboard input, it isnot always the case thatkeystrokes and input characters correspond one-to-one. A limited number of keyscan fit on a keyboard. Some keyboards will generate multiple characters from asingle keypress. In other cases ('dead keys') a key will generateno characters, but affect the results of subsequent keypresses. Many writingsystems have far too many characters to fit on a keyboard and must rely on morecomplexinput methods, which transform keystroke sequences intocharacter sequences. Other languages may make it necessary to input somecharacters with special modifier keys. SeeB Examples of Characters, Keystrokes and Glyphsfor examples of non-trivial input.

C005 [S] [I] Specificationsand softwareMUST NOT require nor depend on a single keystroke resultingin a single character, nor that a single character be input with a singlekeystroke (even with modifiers), nor that keyboards are the same all over theworld.

3.5 Units of collation

String comparison as used in sorting and searching is based onunits which do not in general have a one-to-one relationship to encodedcharacters. Such string comparison can aggregate a character sequence into asinglecollation unit with its own position in the sorting order,can separate a single character into multiple collation units, and candistinguish various aspects of a character (case, presence of diacritics, etc.)to be sorted separately (multi-level sorting).

In addition, a certain amount of pre-processing may also berequired, and in some languages (such as Japanese and Arabic) sort order may begoverned by higher order factors such as phonetics or word roots. Collationmethods may also vary by application.

C006 [S] [I] Softwarethat sorts or searches text for usersSHOULD do so onthe basis of appropriate collation units and ordering rules for the relevantlanguage and/or application.

C007 [S] [I] Where searching or sorting is done dynamically,particularly in a multilingual environment, the 'relevant language'SHOULD be determined to be that of the current user, and maythus differ from user to user.

C066 [S] [I] Software that allows users to sort or search textSHOULD allow the user to select alternative rules for collation units and ordering.

C008 [S] [I] Specifications and implementations of sorting and searching algorithmsSHOULD accommodate text that contains any character in Unicode.

Note that this requires, as a minimum, that a collation algorithm does not break down if the text contains Unicode characters that are not covered by its rules. It does not necessarily require full implementation of complex algorithms for all scripts. One useful way of satisfying the requirement is to apply a default collation algorithm that covers all Unicode characters.

ISO/IEC 14651[ISO/IEC 14651] and Unicode Technical Report #10, the Unicode Collation Algorithm[UTR #10], describe a model for collation that accommodates most languages and provide a default collation order. They are appropriate references for collation and provide implementation guidelines.The default collation order can be used in conjunction with rules tailored for a particular localeto ensure a predictable ordering and comparison of strings, whatever charactersthey include.

3.6 Units of storage

Computer storage and communication rely on units of physicalstorage and information interchange, such as bits and bytes (8-bit units, also called octets). A frequent error in specifications and implementations isthe equating of characters with units of physical storage. The mapping betweencharacters and such units of storage is actually quite complex, and isdiscussed in the next section,4.1 Character Encoding.

C009 [S] [I] Specifications,software and contentMUST NOT require or depend on a one-to-one relationshipbetween characters and units of physical storage.

3.7 Summary

The term character is used differently in a varietyof contexts and often leads to confusion when used outside of these contexts.In the context of the digital representations of text, acharacter can bedefined as a small logical unit of text.Text is thendefined as sequences of characters. While such an informal definition issufficient to create or capture a common understanding in many cases, it isalso sufficiently open to create misunderstandings as soon as details start tomatter. In order to write effective specifications, protocol implementations,and software for end users, it is very important to understand that thesemisunderstandings can occur.

This section,3 Perceptions of Characters, has discussed terms for units that do not necessarily overlap with the term 'character', such as phoneme, glyph, and collation unit. The next section,4.1 Character Encoding, lists terms that should be used rather than 'character' to precisely defineunits of encoding (code point, code unit, and byte).

C010 [S] When specifications use theterm 'character' the specificationsMUSTdefine which meaning they intend.

C067 [S] SpecificationsSHOULD use specific terms, when available, instead of the general term 'character'.

4.1 Character Encoding

On the WWW, as in any computing environment, characters must be encoded to be of any use. To achieve text encoding, a large variety of character encodings have been devised. Character encodings can loosely be explained as mappings between the character sequences that users manipulate and the sequences of bits that computers manipulate.

Given the complexity of text encoding and the large variety of mechanisms for character encoding invented throughout the computer age, a more formal description of the encoding process is useful. The process of defining a text encoding can be described as follows (see Unicode Technical Report #17: Character Encoding Model[UTR #17] for a more detailed description):

1. A set of characters to be encoded is identified. The characters are pragmatically chosen to express text and to efficiently allow various text processes in one or more target languages. They may not correspond precisely to what users perceive as letters and other characters. The set of characters is called arepertoire.

2. Each character in the repertoire is then associated with a (mathematical, abstract) non-negative integer, thecode point (also known as acharacter number orcode position). The result, a mapping from the repertoire to the set of non-negative integers, is called acoded character set (CCS).

3. To enable use in computers, a suitable base datatype is identified (such as a byte, a 16-bit unit of storage or other) and acharacter encoding form (CEF) is used, which encodes the abstract integers of a coded character set (CCS) into sequences of thecode units of the base datatype. The character encoding form can be extremely simple (for instance, one which encodes the integers of theCCS into the natural representation of integers of the chosen datatype of the computing platform) or arbitrarily complex (a variable number of code units, where the value of each unit is a non-trivial function of the encoded integer).

4. To enable transmission or storage using byte-oriented devices, aserialization scheme orcharacter encoding scheme (CES) is next used. A character encoding scheme is a mapping of the code units of a character encoding form (CEF) into well-defined sequences of bytes, taking into account the necessary specification of byte-order for multi-byte base datatypes and including in some cases switching schemes between the code units of multiple character encoding schemes (an example is ISO 2022). A character encoding scheme, together with the coded character sets it is used with, is called acharacter encoding, and is identified by a unique identifier, such as anIANA charset identifier. Given a sequence of bytes representing text and a character encoding identified by acharset identifier, one can in principle unambiguously recover the sequence of characters of the text.

In very simple cases, the whole encoding process can be collapsed to a single step, a trivial one-to-one mapping from characters to bytes; this is the case, for instance, for US-ASCII[ISO/IEC 646] and ISO-8859-1.

Text is said to be in aUnicode encoding form if it is encoded in UTF-8, UTF-16 or UTF-32.

4.2 Transcoding

Transcoding is the process of converting text from onecharacter encoding to another. Transcoders work only at the level of character encoding and do not parse the text; consequently, they do not deal withcharacter escapes such as numeric character references (see4.6 Character Escaping) and do not adjust embedded character encoding information (for instance in an XML declaration or in an HTMLmeta element).

4.3 Reference Processing Model

Many Internet protocols and data formats, most notably the very important Web formats HTML, CSS and XML, are based on text. In those formats, everything is text but the relevant specifications impose a structure on the text, giving meaning to certain constructs so as to obtain functionality in addition to that provided byplain text (text that is not in the context of markup or a programming language). HTML and XML aremarkup languages, defining documents entirely composed of text but with conventions allowing the separation of this text intomarkup andcharacter data. Citing from the XML 1.0 specification[XML 1.0],section 2.4:

"Text consists of intermingled character data and markup. [...] All text that is not markup constitutes the character data of the document."

For the purposes of this section, the important aspect is that everything istext, that is, a sequence of characters.

Atextual data object is a whole text protocol message or a whole text document, or a part of it that is treated separately for purposes of external storage and retrieval. Examples include external parsed entities in XML and textual MIME entity bodies[MIME-entity].

C013 [S] [C] Textual data objects defined byprotocol or format specificationsMUST be in asingle character encoding.

Note that this does notimply that character set switching schemes such as ISO 2022 cannot beused, since such schemes perform character set switching within a singlecharacter encoding.

Since its early days, the Web has seen the development of aReference Processing Model, first described for HTML in RFC 2070[RFC 2070]. This model was later embraced by XML and CSS. It is applicable to any data format or protocol that is text-based as described above. The essence of the Reference Processing Model is the use of Unicode as a common reference. Use of the Reference Processing Model by a specification does not, however, require that implementations actually use Unicode. The requirement is only that the implementations behave as if the processing took place as described by the Model. Also, while this document uses the term ReferenceProcessing Model and describes its properties in terms of processing, the model also applies to specifications that do not explicitly define a processing model.

C014[S] All specifications that involve processing of textMUST specify the processing of text according to theReference Processing Model, namely:

C070 [S] SpecificationsSHOULD NOTarbitrarily exclude code points from the full rangeof Unicodecode points from U+0000 to U+10FFFF inclusive.

C077 [S] SpecificationsMUST NOT allow codepoints above U+10FFFF.

Unicode contains some code points for internal use (such as noncharacters) orspecial functions (such as surrogate code points).

C079 [S] SpecificationsSHOULD NOT allow the use of codepoints reserved by Unicode for internal use.

C078 [S] SpecificationsMUST NOT allow theuse of surrogate code points.

Excluding code points without good reason conflicts with the W3C goal ofuniversal accessibility. Excluding code points would prevent some scripts frombeing used which may be important to a user community or communities. Forexample, without strong reasons to do so, decisions to exclude code points abovethe Basic Multilingual Plane or to limit code points to the US-ASCII or Latin-1repertoire are inappropriate. Also, please note that the Unicode Standard requires software to not corrupt any code points.

Other examples of legitimate and non-arbitrary reasons to exclude characters canbe seen inUnicode in XML and other Markup Languages[UXML], where theuse of certain characters is discouraged for reasons such as:

• They are deprecated in the Unicode Standard.

• They cannot be supported without additional data.

• They are better handled by markup.

• They conflict with equivalent markup.

4.4 Choice and Identification of Character Encodings

Because encoded textcannot be interpreted and processed without knowing the encoding, it is vitally important that the character encoding (see4.1 Character Encoding) is known at all times and places where text is exchanged, stored or processed. In what follows we use 'character encoding' to mean eithercharacter encoding form (CEF) orcharacter encoding scheme (CES) depending on the context. When text is transmitted or stored as a byte stream, for instance in a protocol or file system, specification of aCES is required to ensure proper interpretation. In contexts such as an API, where the environment (typically the processor architecture) specifies the byte order of multibyte quantities, specification of aCEF suffices.

C015 [S] SpecificationsMUST either specify a unique character encoding, or provide character encoding identification mechanisms such that the encoding of text can be reliably identified.

C016 [S] When designing a new protocol, format or API, specificationsSHOULD require a unique character encoding.

C017 [S] When basing a protocol, format, or API on a protocol, format, or API that already has rules for character encoding, specificationsSHOULD use rather than change these rules.

4.5 Private use code points

Certain ranges of Unicodecode points are designated for private use: the Private Use Area (PUA) (U+E000-F8FF) and planes 15 and 16 (U+F0000-FFFFD and U+100000-10FFFD). These code points are guaranteed to never be allocated to standard characters, and are available for use by private agreement. However, private agreements do not scale on the Web. Code points from different private agreements may collide. Also, a private agreement, and therefore the meaning of the code points, can quickly become lost.

C073 [C] Publicly interchanged contentSHOULD NOT usecodepoints in the private use area.

C076 [C] ContentMUST NOT use a code point for any purpose other than that defined by its coded character set.

This prohibits, for example, the construction of fonts that misuse the codepoints in the ISO Latin 1 character set to represent different scripts, characters, or symbols than those actually encoded in iso-8859-1.

C038 [S] SpecificationsMUST NOT require the use of private use areacharacters with particular assignments.

C039 [S] SpecificationsMUSTNOT require the use of mechanisms for defining agreements of privateuse code points.

C040 [S] [I] Specifications andimplementationsSHOULD NOT disallow the use of private use code points by privateagreement.

As an example, XML does not disallow the use ofprivate use code points.

C041 [S] SpecificationsMAY definemarkup toallow the transmission of symbols not in Unicode or to identify specificvariants of Unicode characters.

C068 [S] SpecificationsSHOULD allow the inclusion of or reference to pictures and graphics where appropriate, to eliminate the need to (mis)use character-oriented mechanisms for pictures or graphics.

4.6 Character Escaping

Markup languages or programming languages often designate certain characters assyntax-significant, giving them specific functions within the language (e.g. '<' and '&' serve as markup delimiters in HTML and XML). As a consequence, these syntax-significant characters cannot be used to represent themselves in text in the same way as all other characters do, creating the need for a mechanism to "escape" their syntax-significance. There is also a need, often satisfied by the same or similar mechanisms, to express characters not directly representable in the character encoding chosen for a particular document or program (an instance of the markup or programming language).

Formally, acharacter escape is a syntactic device defined in a markup or programming language that allows one or more of:

1. expressing syntax-significant characters while disregarding their significance in the syntax of the language, or

2. expressing characters not representable in the character encoding chosen for an instance of the language, or

3. expressing characters in general, without use of the corresponding encoded characters.

Escaping a character means expressing it using such a syntactic device, appropriate to the format or protocol in which the character appears;expanding a character escape (orunescaping) means replacing it with the character that it represents.

The following guidelines apply to the way specifications define character escapes.

• C042 [S] SpecificationsSHOULD NOT invent a new escaping mechanism if an appropriate one already exists.

• C043 [S] The number of different ways to escape a characterSHOULD be minimized (ideally to one).

  A well-known counter-example is that for historical reasons, both HTML and XML have redundant decimal (&#ddddd;) and hexadecimal (&#xhhhh;) character escapes.

• C044 [S] Escape syntaxSHOULD require either explicit end delimiters or a fixed number of characters in each character escape. Escape syntaxes where the end is determined by any character outside the set of characters admissible in the character escape itselfSHOULD be avoided.

  These character escapes are not clear visually, and can cause an editor to insert spurious line-breaks when word-wrapping on spaces. Forms like SPREAD's &UABCD;[SPREAD] or XML's &#xhhhh;, where the character escape is explicitly terminated by a semicolon, are much better.

• C045 [S] Whenever specifications define character escapes that allow the representation of characters using a number, the numberMUST represent the Unicode code point of the character andSHOULD be in hexadecimal notation.

• C046 [S] Escaped charactersSHOULD be acceptable wherever their unescaped forms are; this does not preclude thatsyntax-significant characters, when escaped, lose their significance in the syntax. In particular, if a character is acceptable in identifiers and comments, then its escaped form should also be acceptable.

The following guidelines apply to content developers, as well as to software that generates content:

• C047 [I] [C] EscapesSHOULD only be used when the characters to be expressed are not directly representable in the format or the character encoding of the document, or when the visual representation of the character is unclear.

  NOTE:An example of when the visual representation of the character is unclear is the use of &nbsp; to distinguish a non-breaking space from a normal space.

• C048 [I] [C] ContentSHOULD use the hexadecimal form of character escapes rather than the decimal form when there are both.

  NOTE:The hexadecimal form is preferred because character encoding standards (in particular Unicode) usually list character numbers as hexadecimal, making lookup easier.

6.1 String concepts

Various specifications use the notion of a 'string', sometimes without defining precisely what is meant and sometimes defining it differently from other specifications. The reason for this variability is that there are in fact multiple reasonable definitions for a string, depending on one's intended use of the notion; the term 'string' is used for all these different notions because these are actually just different views of the same reality: a piece of text stored inside a computer.

Byte string: A string viewed as a sequence of bytes representing characters ina particular character encoding. This corresponds to acharacter encoding scheme (CES).Text processing of a byte string is dependent on the particularencoding used. When the encoding changes the processing must also be changed to reflect the stucture of the new encoding. Such a change could require significant redesign of the functions or APIused to process the byte strings as text. Therefore, this definition is only useful in specifications when the textual nature of a string is unimportant and the string is considered only as a piece of opaque data with a length in bytes (such as when copying a buffer).

C011 [S] SpecificationsSHOULD NOT define a string as a 'byte string'.

Code unit string: A string viewed as a sequence ofcode units representing characters in a particular character encoding. This corresponds to acharacter encoding form (CEF). A definition of a code unit string needs to include the size of the code units (e.g. 16 bits) and the character encoding used (e.g. UTF-16). Code unit strings are useful in APIs that expose a physical representation of string data basedon reliable knowledge of the encoding forms that are likely candidates forimplementation. Example: For the DOM[DOM Level 1], UTF-16 was chosen based on widespread implementation practice. In general, 'code unit string' is onlyuseful if the implementation candidates are likely to be either UTF-16 or UTF-32.

Character string: A string viewed as a sequence of characters, each represented by acode point in Unicode[Unicode]. This is usually what programmers consider to be a string, although it may not match exactly what most users perceive as characters. This is the highest layer of abstraction that ensures interoperability with very low implementation effort. The 'character string' definition of astring is generally the most useful. Good examples using this definition include the Production [2] of XML 1.0[XML 1.0], the SGML declaration of HTML 4.0[HTML 4.01], and the character model of RFC 2070[RFC 2070].

C012 [S] The 'character string' definitionSHOULD be used by most specifications.

EXAMPLE: Consider the string comprising the characters U+233B4 (a Chinese character meaning 'stumpof tree'), U+2260NOT EQUAL TO, U+0071LATIN SMALL LETTER Q and U+030CCOMBINING CARON,encoded in UTF-16 in big-endian byte order. The rows of the following table show thestring viewed as acharacter string,code unit string andbyte string, respectively:

Glyphs
Character string	U+233B4				U+2260		U+0071		U+030C
Code unit string	D84C		DFB4		2260		0071		030C
Byte string	D8	4C	DF	B4	22	60	00	71	03	0C

6.2 String indexing

There are many situations where a software process needs to access a substring or to point within a string and does so by the use ofindices, i.e. numeric "positions" within a string. Where such indices are exchanged between components of the Web, there is a need for an agreed-upon definition of string indexing in order to ensure consistent behavior. The requirements for string indexing are discussed inRequirements for String Identity Matching[CharReq],section 4. The two main questions that arise are: "What is the unit of counting?" and "Do we start counting at 0 or 1?".

The example in the previous section,6.1 String concepts, shows astring viewed as acharacter string,code unit string andbyte string, respectively, each of which involves different units for indexing.

Depending on the particular requirements of a process, the unit of counting may correspond to definitions of a string provided in section6.1 String concepts. In particular:

• C051 [S] [I] Thecharacter string isRECOMMENDED as a basis for string indexing.

  (Example: the XML Path Language[XPath]).

• C052 [S] [I] Acode unit stringMAY be used as a basis for string indexing if this results in a significant improvement in the efficiency of internal operations when compared to the use ofcharacter string.

  (Example: the use of UTF-16 in[DOM Level 1]).

• C071 [S] [I] Grapheme clustersMAY be used as a basis for string indexing in applications where user interaction is the primary concern.

  See Unicode Standard Annex #29, Text Boundaries[UTR #29].

  C074 [S] Specifications that define indexing in terms of grapheme clustersMUST either: a) define grapheme clusters in terms of default grapheme clusters as defined in Unicode Standard Annex #29, Text Boundaries[UTR #29], or b) define specifically how tailoring is applied to the indexing operation.

• C072 [S] [I] The use ofbyte strings for indexing isNOT RECOMMENDED.

It is noteworthy that there exist other, non-numeric ways of identifying substrings which have favorable properties. For instance, substrings based on string matching are quite robust against small edits; substrings based on document structure (in structured formats such as XML) are even more robust against edits and even against translation of a document from one human language to another.

C053 [S] Specifications that need a way to identify substrings or point within a stringSHOULD provide ways other than string indexing to perform this operation.

C054 [I] [C] Users of specifications (software developers, content developers)SHOULD whenever possible prefer ways other than string indexing to identify substrings or point within a string.

Experience shows that more general, flexible and robust specifications result when individual characters are understood and processed as substrings, identified by a position before and a position after the substring. Understanding indices as boundary positionsbetween the counting units also makes it easier to relate the indices resulting from the different string definitions.

C055 [S] SpecificationsSHOULD understand and process single characters as substrings, and treat indices as boundary positionsbetween counting units, regardless of the choice of counting units.

C056 [S] Specifications of APIsSHOULD NOT specify single characters or single 'units of encoding' as argument or return types.

The issue of index origin, i.e. whether we count from 0 or 1, actually arises only after a decision has been made on whether it is the units themselves that are counted or the positions between the units.

C057 [S] When the positions between the units are counted for string indexing, starting with an index of 0 for the position at the start of the string is theRECOMMENDED solution, with the last index then being equal to the number of counting units in the string.

NOTE:Seehttp://www.unicode.org/unicode/standard/versions/#Citations for guidance on referring to specific versions of the Unicode Standard.

A.1 Normative References

IANA

Internet Assigned Numbers Authority,Official Namesfor Character Sets. (Seehttp://www.iana.org/assignments/character-sets.)

ISO/IEC 10646

ISO/IEC 10646:2003,Informationtechnology -- Universal Multiple-Octet Coded Character Set (UCS), as, from time to time, amended, replaced by anew edition or expanded by the addition of new parts. (Seehttp://www.iso.org/iso/en/ISOOnline.openerpage for thelatest version.)

MIME-entity

N.Freed, N. Borenstein,Multipurpose Internet MailExtensions (MIME). Part One: Format of Internet Message Bodies, RFC 2045, November 1996,http://www.ietf.org/rfc/rfc2045.txt.

MIME-charset

Multipurpose Internet MailExtensions (MIME). Part Two: Media Types, N. Freed, N. Borenstein, RFC 2046,November 1996,http://www.ietf.org/rfc/rfc2046.txt.

RFC 2119

S. Bradner,Key words for use in RFCsto Indicate Requirement Levels, IETF RFC 2119. (Seehttp://www.ietf.org/rfc/rfc2119.txt.)

Unicode

The Unicode Consortium,The Unicode Standard, Version 4, ISBN 0-321-18578-1, asupdated from time to time by the publication of new versions. (Seehttp://www.unicode.org/unicode/standard/versionsfor the latest version and additional information on versions of the standardand of the Unicode Character Database).

Unicode 3.2

The Unicode Consortium,The Unicode Standard, Version 3.2.0 is defined byThe Unicode Standard, Version 3.0 (Reading, MA,Addison-Wesley, 2000. ISBN 0-201-61633-5), as amended by theUnicodeStandard Annex #27: Unicode 3.1 (seehttp://www.unicode.org/reports/tr27)and by theUnicode Standard Annex #28: Unicode 3.2 (seehttp://www.unicode.org/reports/tr28).

Unicode 4.0

The Unicode Consortium.The Unicode Standard, Version 4.0, Reading, MA, Addison-Wesley, 2003. ISBN 0-321-18578-1. (Seehttp://www.unicode.org/versions/Unicode4.0.0/)

A.2 Other References

CharNorm

Martin J. Dürst,François Yergeau, Richard Ishida, Misha Wolf, Tex Texin, Addison PhillipsCharacter Model for the World Wide Web 1.0: Normalization,W3C Working Draft. (Seehttp://www.w3.org/TR/charmod-norm.)

CharIRI

Martin J. Dürst, François Yergeau, Richard Ishida, Misha Wolf, Tex Texin,Character Model for the World Wide Web 1.0: Resource Identifiers, W3C Candidate Recommendation. (Seehttp://www.w3.org/TR/charmod-resid.)

CharReq

Martin J. Dürst,Requirements for StringIdentity Matching and String Indexing, W3C Working Draft. (Seehttp://www.w3.org/TR/WD-charreq.)

Connolly

D. Connolly,CharacterSet Considered Harmful, W3C Note. (Seehttp://www.w3.org/MarkUp/html-spec/charset-harmful.)

CSS2

Bert Bos, Håkon Wium Lie, Chris Lilley,Ian Jacobs, Eds.,CascadingStyle Sheets, level 2 (CSS2 Specification), W3C Recommendation. (Seehttp://www.w3.org/TR/REC-CSS2.)

DOM Level 1

Vidur Apparao et al.,Document Object Model(DOM) Level 1 Specification, W3C Recommendation. (Seehttp://www.w3.org/TR/REC-DOM-Level-1.)

HTML 4.0

Dave Raggett, Arnaud Le Hors, IanJacobs, Eds.,HTML 4.0Specification, W3C Recommendation, 18-Dec-1997 (Seehttp://www.w3.org/TR/REC-html40-971218.)

HTML 4.01

Dave Raggett, Arnaud Le Hors, IanJacobs, Eds.,HTML 4.01Specification, W3C Recommendation. (Seehttp://www.w3.org/TR/html401.)

ISO/IEC 646

ISO/IEC 646:1991,Information technology -- ISO 7-bit coded character set for information interchange. This standard defines an International Reference Version (IRV) which corresponds exactly to what is widely known as ASCII or US-ASCII. ISO/IEC 646 was based on the earlier standard ECMA-6. ECMA has maintained its standard up to date with respect to ISO/IEC 646 and makes an electronic copy available athttp://www.ecma-international.org/publications/standards/Ecma-006.htm

ISO/IEC 9541-1

ISO/IEC 9541-1:1991,Information technology -- Font information interchange -- Part 1: Architecture. (Seehttp://www.iso.ch/iso/en/CatalogueDetailPage.CatalogueDetail?CSNUMBER=17277for the latest version.)

ISO/IEC 14651

ISO/IEC 14651:2000,Information technology -- International string ordering and comparison -- Method for comparing character strings and description of the common template tailorable ordering as,from time to time, amended, replaced by a new edition or expanded by theaddition of new parts. (Seehttp://www.iso.org/iso/en/ISOOnline.openerpage for thelatest version.)

MathML2

David Carlisle, Patrick Ion, RobertMiner, Nico Poppelier, Eds.,Mathematical Markup Language (MathML)Version 2.0, W3C Recommendation. (Seehttp://www.w3.org/TR/MathML2.)

Nicol

Gavin Nicol,TheMultilingual World Wide Web, Chapter 2: The WWW As A MultilingualApplication. (Seehttp://www.mind-to-mind.com/library/papers/multilingual/multilingual-www.html.)

RFC 2070

F. Yergeau, G. Nicol, G. Adams, M.Dürst,Internationalization of theHypertext Markup Language, IETF RFC 2070, January 1997. (Seehttp://www.ietf.org/rfc/rfc2070.txt.)

RFC 2277

H. Alvestrand,IETF Policy on CharacterSets and Languages, IETF RFC 2277, BCP 18, January 1998. (Seehttp://www.ietf.org/rfc/rfc2277.txt.)

RFC 2978

N. Freed, J. Postel,IANA Charset Registration Procedures, IETF RFC 2978, BCP 19, October 2000. (Seehttp://www.ietf.org/rfc/rfc2978.txt.)

RFC 3629

F. Yergeau,UTF-8, a transformationformat of ISO 10646, IETF RFC 3629, STD 63, November 2003. (Seehttp://www.ietf.org/rfc/rfc3629.txt.)

RFC 2781

P. Hoffman, F. Yergeau,UTF-16, an encoding of ISO10646, IETF RFC 2781, February 2000. (Seehttp://www.ietf.org/rfc/rfc2781.txt.)

SPREAD

SPREAD -Standardization Project for East Asian Documents Universal Public EntitySet. (Seehttp://www.ascc.net/xml/resource/entities/index.html)

SVG

Jon Ferraiolo, 藤沢 淳 (FUJISAWA Jun), Dean Jackson, Ed.,Scalable Vector Graphics (SVG) 1.1Specification, W3C Recommendation. (Seehttp://www.w3.org/TR/SVG.)

UTR #10

Mark Davis, Ken Whistler,Unicode CollationAlgorithm, Unicode Technical Report #10. (Seehttp://www.unicode.org/unicode/reports/tr10.)

UTR #17

Ken Whistler, Mark Davis,CharacterEncoding Model, Unicode Technical Report #17. (Seehttp://www.unicode.org/unicode/reports/tr17.)

UTR #29

Mark Davis,Text Boundaries, Unicode Standard Annex #29. (Seehttp://www.unicode.org/unicode/reports/tr29for the latest version).

UXML

Martin Dürst and Asmus Freytag,Unicode in XML and otherMarkup Languages, Unicode Technical Report #20 and W3C Note. (Seehttp://www.w3.org/TR/unicode-xml.)

XML 1.0

Tim Bray, Jean Paoli, C. M.Sperberg-McQueen, Eve Maler, François Yergeau, Eds.,Extensible Markup Language (XML)1.0, W3C Recommendation. (Seehttp://www.w3.org/TR/REC-xml.)

XML Japanese Profile

MURATAMakoto Ed.,XML JapaneseProfile, W3C Note. (Seehttp://www.w3.org/TR/japanese-xml.)

XPath

James Clark, Steve DeRose, Eds,XML Path Language (XPath) Version1.0, W3C Recommendation. (Seehttp://www.w3.org/TR/xpath.)