[Unicode] Technical Reports
 

Unicode Technical Report #23

The Unicode Character Property Model

Editors Ken Whistler (ken@unicode.org), Asmus Freytag (asmus@unicode.org)
Date 2021-09-15
This Version https://www.unicode.org/reports/tr23/tr23-13.html
Previous Version https://www.unicode.org/reports/tr23/tr23-11.html
Latest Version https://www.unicode.org/reports/tr23/
Revision 13

Summary

This document presents a conceptual model of character properties defined in the Unicode Standard. The model also covers properties for enumerated character sequences as well as string functions.

Status

This document has been reviewed by Unicode members and other interested parties, and has been approved for publication by the Unicode Consortium. This is a stable document and may be used as reference material or cited as a normative reference by other specifications.

A Unicode Technical Report (UTR) contains informative material. Conformance to the Unicode Standard does not imply conformance to any UTR. Other specifications, however, are free to make normative references to a UTR.

Please submit corrigenda and other comments with the online reporting form [Feedback]. Related information that is useful in understanding this document is found in the References. For the latest version of the Unicode Standard, see [Unicode]. For a list of current Unicode Technical Reports, see [Reports]. For more information about versions of the Unicode Standard, see [Versions].

Contents

  1. Scope
  2. Overview
  3. Definitions
  4. Conformance-related Considerations
  5. Updating Character Properties and Extending the Standard
  6. Special Property Values

1. Scope

This report presents a general overview and typology of character properties and property values, as well as those of properties of enumerated character sequences and string functions. This description of the Unicode character property model is not intended to supersede the normative information on properties in The Unicode Standard [Unicode], nor the existing body of technical reports and documentation files in the Unicode Character Database [UCDDoc] that provide detailed descriptions for particular character properties or properties of enumerated character sequences and string functions. Instead it focuses on the overall model behind and common aspects of all of these.

This report specifically covers formal character properties, which are those attributes of characters specified according to the definitions set forth in this report. Such formal character properties are only a subset of character properties in the generic sense, and they further subdivide into those properties defined in the Unicode Standard or Unicode Character Database, and those defined by related standards. Also included in the scope are formal properties of enumerated character sequences and string functions.

2. Overview

At its most basic, a character property relates a character to a value. Thus, a property can be considered a function that maps from code points to specific property values. These concepts can be readily extended to mapping a specific sequence of characters to a property value, or to generic string functions that algorithmically map arbitrary strings or substrings to property values. To keep the discussion simple, the basic concepts are introduced in the context of properties of individual characters or code points.

2.1 Origin of Character Properties

The Unicode Standard views character semantics as inherent to the definition of a character, and conformant processes are required to take these into account when interpreting characters. 

D3 Character semantics: The semantics of a character are determined by its identity, normative properties, and behavior.

Note: Quotations from the core specification of the Unicode Standard are cited in this indented boxed style for clarity. Definition numbers or conformance clause numbers in those citations are as in the core specification.

The assignment of character semantics in the Unicode Standard is based on character behavior. Other character set standards leave it to the implementer, or to unrelated secondary standards, to assign character semantics to characters. In contrast, the Unicode Standard supplies a rich set of character attributes, called properties, for each character contained in it. Many properties are specified in relation to processes or algorithms that interpret them, in order to implement the character behavior. There are character behaviors that are specific to a particular text process and that have not been formally defined in the Unicode Standard. Implementations often provide internal definitions of character properties to achieve the desired behavior. Implementers may find many of the concepts discussed here applicable to such cases.

2.2 Character Behavior in Context

The interpretation of some properties (such as whether a character is a digit or not) is largely independent of context, whereas the interpretation of others (such as directionality) is applicable to a character sequence as a whole, rather than to the individual characters that compose the sequence.

Other examples that require context include title casing, and the classification of punctuation or symbols for script assignments. The line breaking rules of UAX#14 Unicode Line Breaking Algorithm [LineBreak] involve character pairs and triples, and in certain cases, longer sequences. The glyph(s) defined by a combining character sequence are the result of contextual analysis in the display shaping engine. Isolated character properties typically only tell part of the story. Characters that are constituent elements of an enumerated list of character sequences obviously exist in the context of such sequences. However, the property defined for specific, enumerated lists of sequences discussed below is different from the kind of algorithmic context discussed here. In fact, algorithms may be defined to evaluate the contexts surrounding not only individual characters or code points, but also the context surrounding certain enumerated character sequences.

In some cases, the expected character behavior depends on external context, such as the type and nature of the document, the language of the text, or the cultural expectations of the user. Properties modeling such behaviors may be specified in separate standards, as is the case for the UTS#10 Unicode Collation Algorithm [UCA]. Where a reasonably generic set of property values can be assigned, for example for [LineBreak], such properties may be defined as part of [Unicode]. Such properties and any algorithms related to them define useful default behavior, which can be further customized or tailored to meet more specific requirements.

2.3 Relation of Character Properties to Algorithms

When modeling character behavior with computer processes, formal character properties are assigned to achieve the expected results. Such modeling depends heavily on the algorithms used to produce these results. In some cases, a given character property is specified in close conjunction with a detailed specification of an algorithm. In other cases, algorithms are implied but not specified, or there are several algorithms that can make use of the same general character property, such as the classification of characters by General_Category or Indic_Syllabic_Type. Such general properties may require occasional implementation-specific adjustments in character property assignment to make all algorithms work correctly. This can usually be achieved by overriding specific properties for specific algorithms. (See also Section 4.3 "Overriding Properties via Higher-level Protocols")

When assigning character properties for use with a given algorithm, it may be tempting to assign somewhat arbitrary values to some characters, as long as the algorithm happens to produce the expected results.Proceeding in this way hides the nature of the character and limits the re-use of character properties by related processes. Therefore, instead of tweaking the properties to simply make a particular algorithm easier, the Unicode Standard pays careful attention to the essential underlying linguistic identity of the character. However, not all aspects of a character’s identity are relevant in all circumstances, and some characters can be used in many different ways, depending on context or circumstance. This means the formal character properties alone are not sufficient to describe the complete range of desirable or acceptable character behaviors.

Note: In some cases, the relevant algorithm is not defined in the Unicode standard. For example, the algorithm that converts strings of digits into numerical values is not defined in the Unicode Standard, but implementations will nevertheless refer to the numeric_value property.

2.4 Code Point And Abstract Character Properties

Code point properties are properties of code points per se: in a character encoding standard these are independent of any assignment of actual abstract characters to those code points. In most character encoding standards, these are trivial, but in the Unicode Standard they are not.

Examples of code point properties include:

These statements remain true of a code point whether or not there is a particular abstract character assigned to them. For example, they track status of the code points: whether any abstract character is assigned to them or can be assigned to them, and so on. Essentially, whenever code points are designated or ranges are reserved in some way, code point properties are assigned.

Character properties are those properties that abstract characters have independent of any consideration of their encoding.

Examples of character properties, not limited to formal properties, include:

By virtue of encoding the abstract character LATIN CAPITAL LETTER G at the code point U+0047, this universe of character properties, some known and obvious, others obscure or even undiscovered, are associated with that code point.

Some of those character properties are generic and systematic enough to be useful or even necessary in the implementation of general text processing algorithms — those are the ones that the Unicode Standard formalizes as properties in the Unicode Character Database.

General text processing algorithms and the programming APIs through which they are accessed, must be prepared to deal with any code point, even one that is unassigned to any characters at the time the implementation was created. As a result, they nearly always need to properly handle each and every code point for any character property, even if they only associate a property value of 'unknown' or 'inapplicable' to unassigned or unsupported code points.

This requirement leads to the use of the unifying concept of Encoded Character Property in the Unicode character property model. An encoded character property combines the concept of a code point property associating ranges of code points with default values of a property, with the concept of a character property associating specific values to the assigned characters. This unified model correlates well with the reality of Unicode-based implementations, which must supply some value for each and every code point. In addition, this unified concept simplifies most of the definitions that are built on top of it, since it is no longer necessary to separately account for definitions applying to character properties vs. code point properties.

2.5 Properties Applied to Strings

Character and code point properties are defined such that all assigned characters and all code points have a defined property value, even if that value is "N/A" ("does not apply"). Assigned characters and code points each form a finite set. This is generally not true for strings. Because there is no inherent, fixed limit to the length of a string, the number of possible sequences is in principle not bounded. Some properties for strings can be described algorithmically, via String Functions, and such properties can be said to apply to every possible string. Other properties apply only to a specific set of strings which is listed explicitly. In this latter case, the properties are referred to as properties of an enumerated set of strings. These concepts are elaborated below in Section 3.6, Strings, and Section 3.7, Properties of Strings.

2.6 Normative Properties

In Chapter 3, Conformance, The Unicode Standard [Unicode] defines a Normative Property as "a Unicode character property used in the specification of the standard" (definition D33) and provides the following explanation:

Specification that a character property is normative means that implementations which claim conformance to a particular version of the Unicode Standard and which make use of that particular property must follow the specifications of the standard for that property for the implementation to be conformant. For example, the Bidi_Class property is required for conformance whenever rendering text that requires bidirectional layout, such as Arabic or Hebrew.

Whenever a normative process depends on a property in a specified way, that property is designated as normative.

The fact that a given Unicode character property is normative does not mean that the values of the property will never change for particular characters. Corrections and extensions to the standard in the future may require minor changes to normative values, even though the Unicode Technical Committee strives to minimize such changes...

Some of the normative Unicode algorithms depend critically on particular property values for their behavior. Normalization, for example, defines an aspect of textual interoperability that many applications rely on to be absolutely stable. As a result, some of the normative properties disallow any kind of overriding by higher-level protocols. Thus the decomposition of Unicode characters is both normative and not overridable; no higher-level protocol may override these values, because to do so would result in non-interoperable results for the normalization of Unicode text. Other normative properties, such as case mapping, are overridable by higher-level protocols, because their intent is to provide a common basis for behavior. Nevertheless, they may require tailoring for particular local cultural conventions or particular implementations.

By making a property normative and non-overridable, the Unicode Standard guarantees that conformant implementations can rely on other conformant implementations to interpret the character in the same way. This is most useful for those properties where the Unicode Standard provides precise rules for the interpretation of characters based on their properties, such as the decompositions and their use by the Normalization forms [Normal].

Note: One trivial, but important example of conformant implementation is runtime access to information from the Unicode Character Database [UCD]. For normative properties exposed by a conformant implementation, conformance requires the returned values to match the values defined by the Unicode Consortium.

For some character properties, such as the general category, the Unicode standard does not define what model of processing the property is intended to support, nor does it specify the required consequences of a character being defined as "Letter Other" as opposed to "Symbol Other", for example. In the absence of such definition, the only effect of conformance that can be rigorously tested is whether a conformant implementation of a character property function returns the correct value to its caller. However, many implementations use such normative properties for their own purposes and guaranteed access to this information helps interoperability.

For information on which properties are normative, see the documentation file for the Unicode Character Database [UCDDoc].

For more information on overriding normative properties, see Section 4.3 Overriding properties via Higher-level Protocols.

2.7 Informative Properties

The Unicode Standard [Unicode] defines an Informative Property as "a Unicode character property whose values are provided for information only" (definition D35) and provides the following explanation:

A conformant implementation is free to use or change informative property values as it may require, while remaining conformant to the standard. An implementer has the option of establishing a protocol to convey that particular informative properties are being used in distinct ways.

Informative properties capture expert implementation experience. When an informative property is explicitly specified in the Unicode Character Database, its use is strongly recommended for implementations to encourage comparable behavior between implementations. Note that it is possible for an informative property in one version of the Unicode Standard to become a normative property in a subsequent version of the standard if its use starts to acquire conformance implications in some part of the standard. [emphasis added].

Properties may be informative for two main reasons:

  1. The exact nature or applicability of the property may be unclear. In some cases, the precise set of characters to which it applies may also not be well-determined.
  2. Existing implementations show a range of behaviors for the same character, many or all of which may be equally useful choices on the part of their designers.

In some cases, properties are too tentative to be published as informative properties. In that case they may be explicitly designated as provisional.

2.8 Referring to Properties

The Property Aliases [Alias] and Property Value Aliases [ValueAlias] define a set of names and abbreviations, called aliases, that are used to refer to properties and property values. These names can be used for XML formats of data in the Unicode Character Database [UCD], for regular-expression property tests, and other programmatic textual descriptions of Unicode data. The names themselves are not normative, except where they correspond to normative properties in the UCD. However, other standards may make normative references to both normative and informative aliases. For more information, see UTS #18: Unicode Regular Expressions [RegEx].

There is one abbreviated name and one long name for most of the properties. Additional aliases may be added at any time. The property value names are not unique across properties. For example, AL means Arabic Letter for the Bidi_Class property, and AL means Alpha_Left for the Combining_Class property, and AL means Alphabetic for the Line_Break property. In addition, some property names may be the same as some property value names. For example, cc means Combining_Class property, and cc means the General_Category property value Control. The combination of property value and property name is, however, unique.

The aliases may be translated in appropriate environments, and additional aliases may be used. The case distinctions, whitespace, and '_' in the property names are not normative. Unless a specific form is required in a particular application, all forms are equivalent. For further information see section 5.9 Matching Rules in UAX#44 Unicode Character Database [UCDDoc].

[Unicode] Section 3.1 gives a prescription for referencing properties: 

References to Unicode Character Properties

Properties and property values have defined names and abbreviations, such as

Property: General_Category (gc)
Property Value: Uppercase_Letter (Lu)

To reference a given property and property value, these aliases are used, as in this example:

The property value Uppercase_Letter from the General_Category property, as specified in Version 14.0.0 of the Unicode Standard.

Then cite that version of the standard, using the standard citation format that is provided for each version of the Unicode Standard.

Additional reference examples are available online.

2.9 The Unicode Character Database

The Unicode Character Database [UCD] is the main repository for machine-readable character properties. It consists of a number of files containing property data along with a documentation file explaining the organization of the database and the format and meaning of the property data. The main file, "The Unicode Character Database" [UCDDoc] explains the overall organization of the current version of the UCD and tells which files contain which properties.

While the Unicode Consortium strives to minimize changes to character property data, occasionally the character properties for already encoded characters must be updated. When this situation occurs, the relevant data files of the Unicode Character Database are revised. The revised data files are posted on the Unicode Web site as an update version of the standard.

A visual documentation of character code point, character name and reference glyph, together with excerpts from some of the character properties and augmented by additional annotations can be found in the Character Code [Charts].

3. Definitions

The following presents a consistent set of definitions related to character properties. Where possible, these definitions match the formal definitions in Chapter 3, Conformance, in [Unicode]. In those cases, the original number of the definition is given at the end of each definition in square brackets. As much as possible, the definition numbers in this document will be retained as new definitions are added. When referring to these definitions in other contexts, it is customary to prefix the term 'Unicode' to the defined term to indicate the context. For example 'Character Property', becomes 'Unicode Character Property', etc.

3.1 Properties and Property Values

PD1. Property
A named attribute of an entity in the Unicode Standard, associated with a defined set of values. [D19]
PD2. Code Point Property
A property of code points. [D20]
A code point property defines a set of values and a mapping from each Unicode code point to one of the values of the set.
PD3. Abstract Character Property
A property of abstract characters. [D21]
PD4. Encoded character property.
A property of encoded characters in the Unicode Standard. [D22]

An encoded character property defines a set of values and a mapping from each Unicode code point to one of the values of the set.
Encoded character properties typically map a default value to any code point not assigned to a character.

In the rest of this document, as in the Unicode Standard, the term 'character property', or the term 'property' without qualifier includes both character and code point properties and their combined form, the encoded character properties.

PD5. Property Value
One of the set of values associated with a property. [D23 - but there limited to 'encoded character property']

For example, the East Asian Width [EAW] property has the possible values "Narrow", "Neutral", "Wide", "Ambiguous" and "Unassigned". See [Alias] and [ValueAlias] for a list of labels for properties and their values respectively.

3.2 Types of Property Values

PD6. Explicit Property Value
A value for an encoded character property which is explicitly associated with a code point in one of the data files of the Unicode Character Database. [D24]
PD7. Implicit Property Value
A value for an encoded character property which is given by a generic rule or by an "otherwise" clause in one of the data files of the Unicode Character Database. [D25]
PD8. Default Property Value
The value (or in some cases small set of values) of a property associated with unassigned code points or with encoded characters for which the property is irrelevant. [D26]
Note: There may be more than one default value per property, with different values for different ranges, as in the Bidi property.

3.3 Types of Properties

PD9. Enumerated Property
A property with a small set of named values. [D27]
As characters are added to the Unicode Standard, the set of values may need to be extended in the future, but enumerated properties, such as the LineBreak property have a relatively fixed set of possible values.
PD10. Closed Enumeration
An enumerated property for which the set of values is closed and will not be extended for future versions of the Unicode Standard. [D28]
Note: Currently, the General_Category and Bidi_Class properties are the only closed enumerations, other than Boolean properties.
PD11. Boolean Property
A closed enumerated property whose set of values is limited to 'true' and 'false'. [D29]
The presence or absence of the property is the essential information.

A Boolean property is sometimes called a 'single valued' property since 'false' often has the meaning of 'this property does not apply'.
PD12. Numeric Property
A numeric property is a property whose value is a number that can take on any integer, or real value. [D30]
An example is the Numeric_Value property. There is no implied limit to the number of possible distinct values for the property, except the limitations on representing integers or real numbers in computers.
PD13. String-Valued Property
A property whose value is a string. [D31]
A string-valued property is one for which the co-domain, or set of values, consists of strings. (See PD32.)
The Canonical_Decomposition property is a string-valued property.

Note: Properties classed in [UCDDoc] as type "String-valued" are string-valued properties. However, some properties classed as "Miscellaneous" are also string-valued properties.

PD13a: Identifier Property
A string-valued property that represents a member of a namespace, with additional rules defining identifier well-formedness, uniqueness and comparison.
For example, a Unicode character name is part of a namespace that also includes name aliases and named sequences. Special rules are defined for comparsion of names and for determination of uniqueness, as well as for which characters are permissible. See Section 4.8 Name in [Unicode].
PD14. Catalog property
A property that is an enumerated property, typically unrelated to an algorithm, that may be extended in each successive version of the Unicode Standard. [D32]

Examples are the Age, Block, and Script properties. Additional new values may be added to the set of enumerated values each time the standard is revised. Each new Unicode version adds a new value for Age. When a new block is added to the standard, a corresponding new value is added to the Block property. Likewise, when a new script is added, a corresponding new value of the Script property is also added.
PD15. Miscellaneous property
A property not of the type Boolean, Enumerated, Numeric, String-valued, Identifier, or Catalog.
The Script_Extensions property is a miscellaneous property.

Note: Actually, some properties classed in [UCDDoc] as type "Miscellaneous" can also be considered string-valued properties. The Jamo_Short_Name property is such an example. The distinction is that most properties currently designated to be of type "String-valued" are conceived of as mapping from some Unicode character to some other Unicode character (or sequence of characters) for the purposes of such operations as case mapping, case folding, or normalization of strings, whereas the string values of Miscellaneous properties tend to be just arbitrary strings.

3.4 Conformance Status of Properties

PD16. Normative Property
A [Unicode character] property used in the specification of the Unicode standard. [D33]
Note: A normative process that depends on a property in a normative and testable way, is usually sufficient reason to designate a property as normative. For example, the interpretation of the bidirectional class is precisely defined in [Bidi].
If a process does not interpret a given character, it may remain unaware of its properties. However, it is recommended that processes use carefully-chosen default values for characters that they do not handle.
See also Section 2.6, Normative Properties.
PD17. Overridable Property
A normative property whose values may be overridden by conformant higher-level protocols. [D34]
See Section 4.3 Overriding properties via Higher-level Protocols.
PD18. Informative Property
A [Unicode character] property whose values are provided for information only. [D35]
Note: Informative properties capture expert implementation experience and their use is strongly recommended by the Consortium, but there are no requirements on implementations of the Unicode Standard.
See also Section 2.7, Informative Properties.
PD19. Provisional Property
A [Unicode character] property whose values are unapproved and tentative, and which may be incomplete or otherwise not in a usable state. [D36]

Provisional properties may be removed from future versions of the standard, without prior notice.
See also Section 5.4, Provisional Properties.

3.5 Classification of Properties

The following definitions do not define character or code point properties, but properties of such properties. In the definitions in this section, the term 'code point' is used inclusively to mean code point for a code point property and character for a character property, respectively.

PD20. Context-dependent Property
A property that applies to a code point in the context of a longer code point sequence. [D37]

For example, the lower case mapping of Greek sigma depends on the surrounding characters. See also PD33: Context-dependent String Function.
PD21. Context-independent Property
A property that is not context-dependent: it applies to a code point in isolation. [D38]
PD22. Stable Transformation
A transformation T on a property P is stable with respect to an algorithm A, if the result of the algorithm on the transformed property A(T(P)) is the same as the original result A(P) for all code points. [D39]
PD23. Stable Property
A property is stable with respect to a particular algorithm or process, as long as possible changes in the assignment of property values are restricted in such a manner that the result of the algorithm on the property continues to be the same as the original result for all previously assigned code points. [D40]
For example, while the absolute values of the canonical combining classes are not guaranteed to be the same between versions of the Unicode Standard, their relative values will be maintained. As a result, the Canonical Combining Class, while not immutable, is a stable property with respect to the Normalization Forms as defined in [Normal].
Note: As new characters are assigned to previously unassigned code points, replacing any default values for these code points with actual property values must maintain stability.
PD24. Fixed Property
A property whose values, (other than the default value) once associated with a character or other designated code point, are fixed and will not be changed, except to correct obvious or clerical errors. [D41]
For a fixed property, any default values can be replaced without restriction by actual property values, as new characters are assigned to previously unassigned code points. Examples of fixed properties are Age or Hangul Syllable Type.
Note: Designating a property as fixed does not imply stability or immutability, see below. While the age of a character, for example, is established by the version of the Unicode Standard at which it was added, errors in the published listing of the property value could be corrected. For some other properties, there are explicit stability guarantees that prohibit the correction even of such errors. See Section 5.2 Stability Guarantees.
PD25. Immutable Property
A fixed property that is also subject to a stability guarantee preventing any change in the published listing of property values other than assignment of new values to formerly unassigned code points. [D42]
An immutable property is trivially stable with respect to all algorithms. An example of an immutable property is the Unicode character name. See Section 5.2 Stability Guarantees.
Note: Because character names are values of an immutable property, misspellings and incorrect names will never be corrected. Any errata will be noted in a comment in the names list, and, where needed, an informative character name alias will be provided.
PD26. Stabilized Property
A property which is neither extended to new characters, nor maintained in any other manner, but which is retained in the Unicode Character Database. [D43]
A stabilized property is also a fixed property.
PD27. Deprecated Property
A property whose use by implementations is discouraged. [D44]

One of the reasons a property may be deprecated is because a different combination of properties better expresses the intended semantics.
Where sufficiently widespread legacy support exists for the deprecated property, not all implementation may be able to discontinue the use of the deprecated property. In such a case, a deprecated property may be extended to new characters, so as to maintain it in a usable and consistent state.
PD28. Simple Property
A property whose values are specified directly in the Unicode Character Database (or elsewhere in the Unicode Standard) and whose values cannot be derived from other simple properties. [D45]
PD29. Derived Property
A property whose values are algorithmically derived from some combination of simple properties. [D46]
PD30. Property Alias
A unique identifier for a particular [Unicode character] property. [D47]
The set of property aliases forms a namespace. See Section 2.8 Referring to Properties.
PD31. Property Value Alias
A unique identifier for a particular enumerated value for a particular [Unicode character] property. [D48]

The set of property value aliases for each property form a separate namespace. Values from different properties may have non-unique names. As a trivial example, the property value aliases for all Boolean properties are 'true' and 'false'.

See also Section 2.8 Referring to Properties.

3.6 Strings

This section introduces definitions for strings, which are needed for the discussion of properties of strings and the role of string functions in the character property model.

PD32. String
An ordered sequence of zero or more code points.
At its most general, a string is any coded character sequence but extending the concept to encompass the empty sequence. Character mappings are common examples of properties for which the values are strings but not necessarily Unicode strings.
All code points in a string are from the same character encoding.
PD32a. Empty String
A string consisting of exactly zero code points.
Note that in principle any empty string is equivalent to any other empty string, so in many contexts, an instance of an empty string is simply referred to as the empty string.

The following three string-related definitions are specified in Chapter 3, Conformance, of the Unicode Standard [Unicode].

PD32b. Code Unit Sequence
An ordered sequence of one or more code units. [D78]
A code unit sequence may consist of a single code unit.
PD32c. Unicode String
An code unit sequence containing code units of a particular Unicode encoding form. [D80]
A single Unicode string must contain only code units from a single Unicode encoding form. It is not permissible to mix forms within a string.
PD32d. Coded Character Sequence
An ordered sequence of one or more code points [D12].
A coded character sequence is also known as a coded character representation.
Normally a coded character sequence consists of a sequence of encoded characters, but it may also include noncharacters or reserved code points.

Those definitions were originally developed to focus on the identity of encoded characters and of sequences of encoded characters, in the context of specifying Unicode encoding forms and other concepts of the Unicode Standard. As such, the formal definitions do not include zero-length sequences as part of their definitions. Where these definitions are used in Chapter 3, the absence of a character is generally not pertinent to the explication.

In programming contexts, however, strings are almost always defined to include the empty string as part of the class or type definition. This is more elegant for implementations of strings and for the design of string-based APIs, including those supporting the implementation of character properties. This distinction is important for the discussion of the Unicode character property model. When the concept of character properties is extended to deal with the properties of Unicode strings, as well as single characters, implementations need to take the empty string into account.

In the Unicode character property model, the primary concern is with properties of characters (or code points), rather than the very limited concept of properties which might apply directly to code units. To avoid clumsiness of terminology, instead of using the formal definition, "coded character sequence," the term Unicode string is simply stipulated, in this context, to also refer to a coded character sequence, instead of only to a code unit sequence.

Furthermore, in the subsequent discussion of properties of strings, for simplicity of presentation, any mention of a Unicode string is also stipulated to extend to include the empty string.

3.7 Properties of Strings

None of the following definitions are found in the Unicode Standard at this point; they extend the existing definitions to cover properties for character sequences.

PD32e. Enumerated Set of Strings
A set of Unicode strings enumerated by an explicit, finite list of members.
This definition is specified as a set, rather than as a list, because typically it is not meaningful to implementations for the same sequences to be included multiple times.
Note that an empty string may explicitly be listed as a member of the set, as appropriate for certain edge cases.
This definition contrasts with sets of strings defined by a rule or definition, such as Combining Character Sequences [D56].
PD32f. Property of Strings
A character property whose domain extends to Unicode strings, as opposed to single code points.
The same categorizations of property types, values, and statuses applies as for encoded character properties.
PD32g. Explicit Property of Strings
A Property of Strings for which each value is specified explicitly for each member of a particular Enumerated Set of Strings.
An example of an Explicit Property of Strings is RGI_Emoji_Flag_Sequence. That is a simple Boolean property, but its domain is the set of emoji flag sequences explicitly listed in the data file emoji-sequences.txt. If a particular sequence is listed in that file, then the value of the RGI_Emoji_Flag_Sequence property for that sequence is True. Otherwise, it is False for any other Unicode string, including the empty string.
RGI_Emoji_Flag_Sequence is also an example of a normative Property of Strings: it is formally defined in a Unicode specification, is maintained in a data file updated with each release, and has conformance implications for implementations of emoji.
Typically, an Explicit Property of Strings will be of type Boolean: either a given sequence is a member of the set or not. However, in principle, properties of more complex types could also be defined to apply to members of an Enumerated Set of Strings.
PD32h. Algorithmic Property of Strings
A Property of Strings whose values are determined by a String Function applied to the entire string (offsets 0 and n).
An example of an Algorithmic Property of Strings is isLowercase. That property is defined in Section 3.13, Default Case Algorithms of the Unicode Standard [Unicode]. It has type Boolean, and is either True or False for any Unicode string, but its value is determined by an algorithm that involves casing the Unicode string and examining the result of that operation.
Another example of an Algorithmic Property of Strings would be isEmojiFlagSequence. That property is not formally defined in Unicode Technical Standard #51, Unicode Emoji [UTS51], but it is implied by definition ED-14 in that specification, for emoji flag sequence. The BNF which defines an emoji flag sequence can be applied algorithmically to any given Unicode string to determine whether that sequence meets the formal syntactic definition or not. That determination does not require checking against an explicit, enumerated character sequence set. And in fact, the entire point of the Explicit Property of Strings, RGI_Emoji_Flag_Sequence, by contrast, is to allow for picking out of the domain of all possible syntactically correct emoji flag sequences, just the precise set listed in emoji-sequences.txt as recommended for general interchange [RGI]. The RGI status is not algorithmically derivable, and can only be specified by providing an Enumerated Set of Strings to test against.

3.8 String Functions

None of the following definitions is found in the Unicode Standard at this point, however, they are useful in the context of discussing Unicode algorithms and their relation to properties.

PD33. Offset
An offset into a string is a number from 0 to n where n is the length of the string in code points. It indicates a position that is logically between code points. An offset of 0 indicates the position before the first code point in the string, and an offset of n indicates the position after the last code point in the string.

Dealing with offsets at the level of code units is the concern of lower-level implementation processes, which must deal with the details of character encoding forms. For the purposes of the character property model, strings are simply defined abstractly in terms of encoded character sequences and code points.

PD34. [Definition removed]
 
PD35. String Function
A string function is a function whose input is a string S and two offsets a and b, with ab.
PD36. Text Boundary Property
A string function whose value is defined for a particular offset.

Text boundary functions are also called segmentation functions, because they are commonly used to return segments of text between boundaries. A simple text boundary function, like IsBreak(S,a,b) minimally returns a Boolean value. However, other text boundary functions may return additional information. For example, a word-selection boundary function may return whether the previous segment contained a letter, or a linebreak function may return information on the relative priority of the break.

3.9 Classification of String Functions

PD37. Context-independent String Function
Given a string S, and offsets a and b, a context-independent string function is any string function F for which F(S,a,b) is independent of the content of S before a and after b.

In other words, the input to a context-independent function is fully defined by the code points between the given offsets.
PD38. Context-dependent String Function
A context-dependent string function is a string function that is not context-independent.

In other words, the input to a context-dependent string function requires additional information, such as information about the code points surrounding the code point range defined by the offsets as well as the code points in the range. Any text boundary function of the form B (S,x,x) is by definition context dependent.
PD39. String Transform
A string-valued string function.
PD40. Idempotent String Function (Folding)
A string transform F, with the property that repeated applications of the same function F produce the same output: F(F(S)) = F(S) for all input strings S.
Such a string function is also called a folding.
A folding establishes an equivalence relation, whereby X ≡ Y if and only if F(X) = F(Y). This equivalence relation partitions the set of all strings into the set of equivalence classes for the relation. Conversely, any partition of strings can be used to generate a folding, by choosing one element of each partition to be the "target member" that the members of that partition map to.

The notation toX(s) may be used for the folding, and isX(s) for the corresponding binary function, defined such that isX(s) if and only if toX(s) = s. For example, toNFC() is the folding that converts to NFC format, while isNFC() is the test for whether a string is in that format.

A well known example of a folding function is case folding. For case folding, the equivalence class consists of all case variations, including upper, lower, title case and mixed case. In the case of Unicode case folding, the target member is chosen to be the lowercase character.
Folding functions may be context dependent. Normalization is an example of a context dependent folding.
PD41. Code Point Count Preserving String Function
A string function whose result is a string containing the same number of code points as its input, is a count preserving string function.
PD42. Buffer Length Preserving String Function
A string function whose result is a string containing the same number of code units as its input, is a buffer length preserving string function.

3.10 Other Definitions

PD43. Higher-level Protocol
Any agreement on the interpretation of Unicode characters that extends beyond the scope of the Unicode Standard. [D16]

4. Conformance-related Considerations

This technical report does not define conformance requirements, but the following subsections discuss and summarize the conformance requirements related to character properties stated in the Unicode Standard. Where applicable, the number of the corresponding conformance clause or definition is given in square brackets.

4.1 Conformance Requirements

In Chapter 3, Conformance, The Unicode Standard [Unicode] states that "A process shall interpret a coded character sequence according to the character semantics established by this standard, if that process does interpret that coded character sequence." [C4] The semantics of a character are established by taking its coded representation, character name and representative glyph in context and are further defined by its normative properties and behavior. Neither character name nor representative glyphs can be relied upon absolutely; a character may have a broader range of use than the most literal interpretation of its character name, and the representative glyph is only indicative of one of a range of typical glyphs representing the same character.

4.2 Algorithms and Character Properties

Unicode algorithms are specified as an idealized series of steps (rules) performed on an input of character codes and their associated properties. [Unicode] states:

As long as the same results are achieved, the implementation is also not required to use the actual properties published in the [UCD]. Overriding a property value therefore does not necessarily imply an actual change in property assignments, merely that the conformant implementation of an algorithm now produces the same results as if the property values had been changed in the description of the ideal algorithm.

4.3 Overriding Properties via Higher-level Protocols

In discussing character semantics, the Unicode Standard [Unicode] makes this statement about overriding properties and character behavior:

Some normative behavior is default behavior; this behavior can be overridden by higher-level protocols. However, in the absence of such protocols, the behavior must be observed so as to follow the character semantics. See [D3].

Overrides by a higher-level protocol can conceptually take many forms, including, but not limited to:

Where overrides involve normative properties, specific restrictions apply, for example:

• The character combination properties and the canonical ordering behavior cannot be overridden by higher-level protocols. See [D3].

For additional examples of higher-level protocols as well as restrictions on them see section 4.3 in UAX #9: Unicode Bidirectional Algorithm [Bidi]. There are some normative properties that are fully overridable, for example General Category.

On the other hand, any and all informative properties may be overridden. However, if doing so changes the result of a Unicode Algorithm, any implementation wishing to conform to that algorithm must indicate that overrides have been applied.

5. Updating Properties and Extending the Standard

5.1 Updating Properties

Updates to properties the Unicode Character Database can be required for three reasons:

  1. To cover new characters added to the Unicode Standard
  2. To add new properties
  3. To change the assigned values for a property for some characters

While the Unicode Consortium endeavors to keep the values of all character properties as stable as possible, some circumstances may arise that require changing them. Changing a character's property assignment may impact existing implementations and is therefore done judiciously and with great care, only when there is no better alternative.

In particular, as Unicode encodes less well-documented scripts, such as those for minority languages, the exact character properties and behavior may not be known when the script is first encoded. The properties for such characters are expected to be changed as information becomes available.

As implementation experience grows, it may become necessary to readjust property values. As much as possible, such readjustments are compatible with established practice. Occasionally, a character property is changed to prevent incorrect generalizations of a character's use based on its nominal property values. For example, U+200B ZERO WIDTH SPACE was originally classified as a space character (General Category=Zs), but is now classified as a Formal Control (gc=Cf) to distinguish this line break control from space characters.

In other cases, there may have been unintentional mistakes in the original information that require corrections.

The [UTC] carefully weighs the costs of a change against the benefit of the correction. In addition, all updates to properties are subject to the stability guarantees described in the next section.

5.2 Stability Guarantees

Unicode guarantees the stability of character assignments; that is, the identity of a character encoded at a given location will remain the same. Once a character is encoded, its properties may still be changed, but not in such a way as to change the fundamental identity of the character.

For example, the representative glyph for U+0041 "A" could not be changed to "B"; the general category for U+0041 "A" could not be changed to Ll (lowercase letter); and the decomposition mapping for U+00C1 (Á) could not be changed to <U+0042, U+0301> (B, ´).

In addition, for some properties, one or more of the following aspects are guaranteed to be invariant:

For the most up-to-date specification of all stability guarantees in effect see the Unicode Character Encoding Stability Policy [Stability]. Note that the status of a property as normative does not imply a stability guarantee.

5.2.1 Stability of Assignment

Stability of assignment is the characteristic of an immutable property. For example, once a character is encoded, its code point and name are immutable properties. An immutable property allows software and documents to refer to its values without needing to track future updates to the Standard. One side effect of an immutable property is that errors in property values cannot be fixed. For example, mistakes in naming are annotated in the Unicode character names list in a note or by using an alias, but the formal name remains unchanged, even in cases of clear-cut typographical errors.

Because Code_Point is an immutable property, if a character is ever found to be unnecessary, or a mistaken duplicate of an existing character, it will not be removed. Instead, it can be given an additional property, deprecated, and its use strongly discouraged. However, the interpretation of all existing documents containing the character remains the same.

5.2.2 Stability of Result when Applying the Property

Stability of result is the characteristic of a stable property. For example, once a character is encoded, its canonical combining class and decomposition (canonical or compatibility) are stable with respect to normalization. Stability with respect to normalization is defined in such a way that if a string contains only characters from a given version of the Unicode Standard (say Unicode 3.2), and it is put into a normalized form in accordance with that version of Unicode, then it will be in normalized form when normalized according to any future version of Unicode.

However, unlike character code and character name, some properties that are guaranteed to be stable may be corrected in exceptional circumstances that are clearly defined by the Unicode Character Encoding Stability Policy [Stability]. In addition to other requirements, the correction must be of an obvious mistake, such as a typographical error, and any alternative would have to violate the stability of the identity of the character in question. Allowing such carefully restricted exceptions obviates the need for encoding duplicate characters simply to correct clerical or other clear-cut errors in property assignments.

5.2.3 Stability of Set of Values for a Property

For most properties, additional property values may be created and assigned to both new and existing characters. For example additional line breaking classes will be assigned if characters are discovered to require line breaking behavior that cannot be expressed with the existing set of classes. For other properties the set of values is guaranteed to be fixed, or their range is limited. For example, the set of values for the General_Category or Bidirectional_Class is fixed, while combining classes are limited to the values 0 to 254.

5.2.4 Stability of Relation to Another Property

In many cases, once a character has a certain value for one property, it is likely to have a particular value for a given other property. These relations are used by the Unicode Consortium in assigning properties to new characters, and in evaluating properties for internal consistency. In some cases, such dependencies are explicitly guaranteed and stable.

For example, all characters other than those of General Category M* have the combining class 0.

5.2.5 Stability of File Formats

In principle, the way the property information is presented in the Unicode Character Database is independent of the way this information is defined. However, as the Unicode Standard gets updated, it becomes easier for implementations to track updates if file formats remain unchanged and other aspects of the way the data are organized can remain stable. For the majority of properties, such stability is an informal goal of the development process, but in a few cases, some aspects of the data organization are covered by formal stability guarantees.

For example, Canonical and Compatibility mappings are always in canonical order, and the resulting recursive decomposition will also be in canonical order. Canonical mappings are also always limited either to a single value or to a pair. The second character in the pair cannot itself have a canonical mapping.

As an alternative to the legacy conventions of semicolon-separated text files, the Unicode Character Database is now also available as a single XML file. See UAX#42 Unicode Character Database in XML [XML].

5.3 Consistency of Properties

In an ideal world, all character properties would be perfectly self-consistent, and related properties would be consistent with each other over the entire range of code points. However, The Unicode Standard is the product of many compromises. It has to strike a balance between uniformity of treatment for similar characters, and compatibility with existing practice for characters inherited from legacy encodings. Because of this balancing act, one can expect a certain number of anomalies in character properties.

Sometimes it may be advantageous for an implementation to purposefully override some of the anomalous property values, increasing the efficiency and uniformity of algorithms—as long as the results they produce do not conflict with those specified by the normative properties of this standard. See Chapter 4, Character Properties in [Unicode] for some examples.

Property values assigned to new characters added to the Unicode Standard are generally defined so that related characters are given consistent values, unless deliberate exceptions are needed. For some properties, definite links between that property and one or more other properties are defined. For example, for the LineBreak property, many line break classes are defined in relation to General Category values.

There are some properties that are interrelated or that are derived from a combination of other properties, with or without a list of explicit exceptions. When properties are assigned to newly assigned characters, or when properties are adjusted, it is necessary to take into account all existing relevant properties, any derivational relations to derived properties, and all property stability guarantees.

5.4 Provisional Properties

Some of the information provided about characters in the Unicode Character Database constitutes provisional data. Provisional property data may capture partial or preliminary information. Such data may contain errors or omissions, or otherwise not be ready for systematic use; however, provisional property data are included in the data files for distribution partly to encourage review and improvement of the information. For example, a number of the tags in the Unihan database provide provisional property values of various sorts about Han characters.

5.5 Stabilized Properties

Occasionally, as the standard matures, and new characters, properties or algorithms are defined, the information presented in an existing property may be better represented via other properties, or it may no longer make sense to extend the property to new characters. Such a property may then no longer be maintained in future versions of the Unicode Standard. In that case, it will be designated as stabilized. For backwards compatibility, a stabilized property will remain part of the Unicode Character database, but will not be updated or corrected.

An example of a stabilized property is Hyphen.

6. Special Property Values

6.1 Not Applicable Value

Limited properties apply to only a subset of characters. Where these properties are implemented as a partition of the Unicode code space, the characters to which the property does not apply are given a special value denoting that the property does not apply. The "not applicable" value may be the explicit value "NA" or, for some properties, take other values such as "XX".

6.2 Default Values

Implementations often need specific properties for all code points, including those that are unassigned. To meet this need, the Unicode standard assigns default properties to ranges of unassigned code points.

All implementations of the Unicode Standard should endeavor to handle additions to the character repertoire gracefully. In some cases this may require that an implementation attempts to 'anticipate' likely property values for code points for which characters have not yet been defined, but where surrounding characters exist that make it probable that similar characters will be assigned to the code point in question.

There are three strategies:

  1. Rely on the recommendation from the Unicode Consortium. For example, for the Bidirectional Class, the Unicode Consortium has published recommended default values for all code points. For details of these recommendations for various properties see [UCDDoc].
  2. Treat the unassigned areas of a given character block as if they had property values common to other characters of the block. A variation of this scheme bridges small gaps in the allocation inside a block by using the property values for the characters bracketing the hole.
  3. Give an unassigned code point an implementation defined default property that will result in graceful if not completely correct behavior, if an encoded characters is later assigned at that code point.

Each of these strategies has advantages and drawbacks, and none can guarantee that the behavior of an implementation that is conformant to a prior version of the Unicode Standard will support characters added in a later version of the Unicode Standard in precisely the same way as an implementation that is conformant to the later version. The most that can be hoped for, is that the earlier implementation will behave more gracefully in such circumstances.

In principle, default values are temporary: they are superseded by final assignments once characters are assigned to a given code point.

For noncharacter code points, a character property function would return the same value as the default value for unassigned characters.

6.3 Preliminary Property Assignments

Sometimes, a determination and assignment of property values can be made, but the information on which it was based may be incomplete or preliminary. In such cases, the property value may be changed when better information becomes available. Currently, there is no machine readable way to provide information about the confidence of a property assignment; however, the text of the Standard or a Technical Report defining the property may provide general indications of preliminary status of property assignments where they are known.

This is distinct from provisional properties, where the entire property is preliminary.

References

[Alias] Property Aliases
https://www.unicode.org/unicode/Public/UCD/latest/ucd/PropertyAliases.txt
[Bidi] Unicode Standard Annex #9: The Unicode Bidirectional Algorithm
https://www.unicode.org/reports/tr9/
[Charts] The online code charts can be found at https://www.unicode.org/charts/ An index to characters names with links to the corresponding chart is found at https://www.unicode.org/charts/charindex.html
[EAW] Unicode Standard Annex #11: East Asian Width
https://www.unicode.org/reports/tr11/
[FAQ] Unicode Frequently Asked Questions
https://www.unicode.org/faq/
For answers to common questions on technical issues.
[Glossary] Unicode Glossary
https://www.unicode.org/glossary/
For explanations of terminology used in this and other documents.
[LineBreak] Unicode Standard Annex #14: Unicode Line Breaking Algorithm
https://www.unicode.org/reports/tr14/
[Normal] Unicode Standard Annex #15: Unicode Normalization Forms
https://www.unicode.org/unicode/reports/tr15/
[RegEx] Unicode Technical Standard #18: Unicode Regular Expressions
https://www.unicode.org/unicode/reports/tr18/
[Stability] Unicode Character Encoding Stability Policy
https://www.unicode.org/policies/stability_policy.html
[UCA] Unicode Technical Standard #10: Unicode Collation Algorithm
https://www.unicode.org/reports/tr10/
[UCD] About the Unicode Character Database
https://www.unicode.org/ucd/
For an overview of the Unicode Character Database
[UCDDoc] Unicode Standard Annex #44: Unicode Character Database
https://www.unicode.org/reports/tr44/
For documentation of the contents of the Unicode Character Database and its associated files
[Unicode] The Unicode Standard
For the latest version see:
https://www.unicode.org/versions/latest/
For Version 14.0 see: The Unicode Consortium. The Unicode Standard, Version 14.0.0 (Mountain View, CA: The Unicode Consortium, 2021. ISBN 978-1-936213-29-0).
https://www.unicode.org/versions/Unicode14.0.0/
[Unihan] Unicode Standard Annex #38: Unicode Han Database (Unihan)
https://www.unicode.org/reports/tr38/
The database itself is available online at
https://www.unicode.org/Public/UCD/latest/ucd/Unihan.zip (large download)
[UTC] The Unicode Technical Committee
For more information see https://www.unicode.org/consortium/utc.html
[UTS51] Unicode Technical Standard #51: Unicode Emoji
https://www.unicode.org/reports/tr51/
[ValueAlias] Property Value Aliases
https://www.unicode.org/Public/UCD/latest/ucd/PropertyValueAliases.txt
[XML] Unicode Standard Annex #42: Unicode Character Database in XML
https://www.unicode.org/reports/tr42/
The XML version of the database is available online at
https://www.unicode.org/Public/UCD/latest/ucdxml/

Acknowledgements

Asmus Freytag was the initial author of this report, with additional content provided by Ken Whistler.

The editors wish to thank Mark Davis for his extensive contributions and insightful comments, and Dr. Julie Allen for extensive copy-editing.

Modifications

The following summarizes modifications from the previous version of this document.

Revision 13 [AF, KW]

Revision 12 being a proposed update, only modifications between revisions 11 and 13 are noted here.

Revision 11 [AF, KW]

Revision 10 being a proposed update, only modifications between revisions 9 and 11 are noted here.

Revision 9 [KW]

Revision 8 being a proposed update, only modifications between revisions 7 and 9 are noted here.

Revision 7 [AF]

Revision 6 being a proposed update, only modifications between revisions 5 and 7 are noted here.

Revision 5 [AF]