[Unicode]  Technical Reports

Working Draft Proposed Update Unicode Technical Standard #18

Unicode Regular Expressions

Version 14 draft 3
Editors Mark Davis, Andy Heninger
Date 2012-01-08
This Version http://www.unicode.org/reports/tr18/tr18-14.html
Previous Version http://www.unicode.org/reports/tr18/tr18-13.html
Latest Version http://www.unicode.org/reports/tr18/
Proposed Update, Current Draft http://www.unicode.org/reports/tr18/proposed.html
Revision 14


This document describes guidelines for how to adapt regular expression engines to use Unicode.


This is a draft document which may be updated, replaced, or superseded by other documents at any time. Publication does not imply endorsement by the Unicode Consortium. This is not a stable document; it is inappropriate to cite this document as other than a work in progress.

A Unicode Technical Standard (UTS) is an independent specification. Conformance to the Unicode Standard does not imply conformance to any UTS.

Please submit corrigenda and other comments with the online reporting form [Feedback]. Related information that is useful in understanding this document is found in [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].


0 Introduction

[Review Note: see also PRI #179 Changes to Unicode Regular Expression Guidelines.]

The following describes general guidelines for extending regular expression engines (Regex) to handle Unicode. The following issues are involved in such extensions.

There are three fundamental levels of Unicode support that can be offered by regular expression engines:

In particular:

  1. Level 1 is the minimally useful level of support for Unicode. All regex implementations dealing with Unicode should be at least at Level 1.
  2. Level 2 is recommended for implementations that need to handle additional Unicode features. This level is achievable without too much effort. However, some of the subitems in Level 2 are more important than others: see Level 2.
  3. Level 3 contains information about extensions only useful for specific applications. Features at this level may require further investigation for effective implementation.

One of the most important requirements for a regular expression engine is to document clearly what Unicode features are and are not supported. Even if higher-level support is not currently offered, provision should be made for the syntax to be extended in the future to encompass those features.

Note: Unicode is a constantly evolving standard: new characters will be added in the future. This means that a regular expression that tests for currency symbols, for example, has different results in Unicode 2.0 than in Unicode 2.1, where the Euro currency symbol was added.

At any level, efficiently handling properties or conditions based on a large character set can take a lot of memory. A common mechanism for reducing the memory requirements — while still maintaining performance — is the two-stage table, discussed in Chapter 5 of The Unicode Standard [Unicode]. For example, the Unicode character properties required in RL1.2 Properties can be stored in memory in a two-stage table with only 7 or 8 Kbytes. Accessing those properties only takes a small amount of bit-twiddling and two array accesses.

Note: For ease of reference, the section ordering for this document is intended to be as stable as possible over successive versions. That may lead, in some cases, to the ordering of the sections being less than optimal.

0.1 Notation

In order to describe regular expression syntax, an extended BNF form is used:

x y the sequence consisting of x then y
x* zero or more occurrences of x
x? zero or one occurrence of x
x | y either x or y
( x ) for grouping
"XYZ" terminal character(s)

The following syntax for character ranges will be used in successive examples.

Note: This is only a sample syntax for the purposes of examples in this document. Regular expression syntax varies widely: the issues discussed here would need to be adapted to the syntax of the particular implementation. However, it is important to have a concrete syntax to correctly illustrate the different issues. In general, the syntax here is similar to that of Perl Regular Expressions [Perl].) In some cases, this gives multiple syntactic constructs that provide for the same functionality.

     := CODE_POINT2 "-" CODE_POINT2 // range

            := CODE_POINT

SEP := ""   // no separator = union 
    := "||" // union
ESCAPE := "\"

CODE_POINT refers to any Unicode code point from U+0000 to U+10FFFF, although typically the only ones of interest will be those representing characters. Whitespace is allowed between any elements, but to simplify the presentation the many occurrences of " "* are omitted.

Code points that are syntax characters or whitespace are typically escaped. For more information see [UAX31]. In examples, the syntax \s to mean white space is sometimes used. See also Annex C. Compatibility Properties.


[a-z || A-Z || 0-9] Match ASCII alphanumerics
[a-z A-Z 0-9]
[^a-z A-Z 0-9] Match anything but ASCII alphanumerics
[\] \- \ ] Match the literal characters ], -, <space>

Where string offsets are used in examples, they are from zero to n (the length of the string), and indicate positions between characters. Thus in "abcde", the substring from 2 to 4 includes the two characters "cd".

The following notation is defined for use here and in other Unicode documents:

\n As used within regular expressions, expands to the text matching the nth parenthesized group in regular expression. (à la Perl). Note that most engines limit n to be [1-9]; thus \456 would be the reference to the 4th group followed by the literal '56'.
$n As used within replacement strings for regular expressions, expands to the text matching the nth parenthesized group in a corresponding regular expression. The value of $0 is the entire expression.( à la Perl)
$xyz As used within regular expressions or replacement strings, expands to an assigned variable value. The 'xyz' is of the form of an identifier. For example, given $greek_lower = [[:greek:]&&[:lowercase:]], the regular expression pattern "ab$greek_lower" is equivalent to "ab[[:greek:]&&[:lowercase:]]".

Note: Because any character could occur as a literal in a regular expression, when regular expression syntax is embedded within other syntax it can be difficult to determine where the end of the regex expression is. Common practice is to allow the user to choose a delimiter like '/' in /ab(c)*/. The user can then simply choose a delimiter that is not in the particular regular expression.

0.2 Conformance

The following describes the possible ways that an implementation can claim conformance to this technical standard.

All syntax and API presented in this document is only for the purpose of illustration; there is absolutely no requirement to follow such syntax or API. Regular expression syntax varies widely: the features discussed here would need to be adapted to the syntax of the particular implementation. In general, the syntax in examples is similar to that of Perl Regular Expressions [Perl], but it may not be exactly the same. While the API examples generally follow Java style, it is again only for illustration.

C0. An implementation claiming conformance to this specification at any Level shall identify the version of this specification and the version of the Unicode Standard.
C1. An implementation claiming conformance to Level 1 of this specification shall meet the requirements described in the following sections:
RL1.1 Hex Notation
RL1.2 Properties
RL1.2a Compatibility Properties
RL1.3 Subtraction and Intersection
RL1.4 Simple Word Boundaries
RL1.5 Simple Loose Matches
RL1.6 Line Boundaries
RL1.7 Supplementary Code Points
C2. An implementation claiming conformance to Level 2 of this specification shall satisfy C1, and meet the requirements described in the following sections:
RL2.1 Canonical Equivalents
RL2.2 Extended Grapheme Clusters
RL2.3 Default Word Boundaries
RL2.4 Default Case Conversion
RL2.5 Name Properties
RL2.6 Wildcards in Property Values
RL2.7 Full Properties
C3. An implementation claiming conformance to Level 3 of this specification shall satisfy C1 and C2, and meet the requirements described in the following sections:
RL3.1 Tailored Punctuation
RL3.2 Tailored Grapheme Clusters
RL3.3 Tailored Word Boundaries
RL3.4 Tailored Loose Matches
RL3.5 Tailored Ranges
RL3.6 Context Matching
RL3.7 Incremental Matches
RL3.9 Possible Match Sets
RL3.10 Folded Matching
RL3.11 Submatchers
C4. An implementation claiming partial conformance to this specification shall clearly indicate which levels are completely supported (C1-C3), plus any additional supported features from higher levels.

For example, an implementation may claim conformance to Level 1, plus Context Matching, and Incremental Matches. Another implementation may claim conformance to Level 1, except for Subtraction and Intersection.


1 Basic Unicode Support: Level 1

Regular expression syntax usually allows for an expression to denote a set of single characters, such as [a-z A-Z 0-9]. Because there are a very large number of characters in the Unicode standard, simple list expressions do not suffice.

1.1 Hex notation

The character set used by the regular expression writer may not be Unicode, or may not have the ability to input all Unicode code points from a keyboard.

RL1.1 Hex Notation
To meet this requirement, an implementation shall supply a mechanism for specifying any Unicode code point (from U+0000 to U+10FFFF), using the hexadecimal code point representation.

The syntax must use the code point in its hexadecimal representation. For example, syntax such as \uD834\uDD1E or \xF0\x9D\x84\x9E does not meet this requirement for expressing U+1D11E (𝄞) because "1D11E" does not appear in the syntax. Syntax such as \U0001D11E, \x{1D11E} or \u{1D11E} does satisfy the requirement for expressing U+1D11E.

A sample notation for listing hex Unicode characters within strings is by using "\u" followed by four hex digits or "\u{" followed by any number of hex digits and terminated by "}", with multiple characters indicating by separating the hex digits by spaces. This would provide for the following addition:

<codepoint> := <character>
<codepoint> := "\u{" HEX_CHAR+ "}"
<codepoints> := "\u{" HEX_CHAR+ (SEP HEX_CHAR+)* "}"
<sep>       := \s+

<codepoint> := ESCAPE U_LONG_MARK



[\u3040-\u309F \u30FC] Match Hiragana characters, plus prolonged sound sign
[\u00B2 \u2082] Match superscript and subscript 2
[a \u{10450}] Match "a" or U+10450 SHAVIAN LETTER PEEP
ab\u{63 64} Match "abcd"

More advanced regular expression engines can also offer the ability to use the Unicode character name for readability. See 2.5 Name Properties.

1.1.1 Hex Notation and Normalization

The Unicode Standard treats certain sequences of characters as equivalent, such as the following:

u + grave U+0075 ( u ) LATIN SMALL LETTER U +

Literal text in regular expressions may be normalized (converted to equivalent characters) in transmission, out of the control of the authors of of that text. For example, a regular expression may contain a sequence of literal characters 'u' and grave, such as the expression [aeiou ̀ ́ ̈̈] (the last three character being U+0300 ( ̀ ) COMBINING GRAVE ACCENT, U+0301 ( ́ ) COMBINING ACUTE ACCENT, and U+0308 ( ̈ ) COMBINING DIAERESIS. In transmission, the two adjacent characters in Row 1 might be changed to the different expression containing just one character in Row 2, thus changing the meaning of the regular expression. Hex notation can be used to avoid this problem. In the above example, the regular expression should be written as [aeiou\u0300\u0301\u0308] for safety.

A regular expression engine may also enforce a single, uniform interpretation of regular expressions by always normalizing input text to Normalization Form NFC before interpreting that text. For more information, see UAX #15: Unicode Normalization Forms [UAX15].

1.2 Properties

Because Unicode is a large character set, a regular expression engine needs to provide for the recognition of whole categories of characters as well as simply ranges of characters; otherwise the listing of characters becomes impractical and error-prone. This is done by providing syntax for sets of characters based on the Unicode character properties, and allowing them to be mixed with lists and ranges of individual code points.

There are a large number of Unicode Properties defined in the Unicode Character Database (UCD), which also provides the official data for mapping Unicode characters (and code points) to property values. See Section 2.7, Full Properties, also UAX #44, Unicode Character Database [UAX44] and Chapter 4 in The Unicode Standard [Unicode]. The defined Unicode string functions, such as isNFC() and isLowercase(), also apply to single code points and are useful to support in regular expressions.

The recommended names for UCD properties and property values are in PropertyAliases.txt [Prop] and PropertyValueAliases.txt [PropValue]. There are both abbreviated names and longer, more descriptive names. It is strongly recommended that both names be recognized, and that loose matching of property names be used, whereby the case distinctions, whitespace, hyphens, and underbar are ignored.

Note: it may be a useful implementation technique to load the Unicode tables that support properties and other features on demand, to avoid unnecessary memory overhead for simple regular expressions that do not use those properties.

Where a regular expression is expressed as much as possible in terms of higher-level semantic constructs such as Letter, it makes it practical to work with the different alphabets and languages in Unicode. Here is an example of a syntax addition that permits properties.

Notice that following Perl Syntax, the p is lowercase to indicate a positive match, and uppercase to indicate a negative match.

POSITIVE_SPEC := ("\p{" PROP_SPEC "}") | ("[:" PROP_SPEC ":]")
NEGATIVE_SPEC := ("\P{" PROP_SPEC "}") | ("[:^" PROP_SPEC ":]")
PROP_SPEC  := <binary_unicode_property>
PROP_SPEC  := <unicode_property> (":" | "=" | "≠" ) VALUE
PROP_SPEC  := <script_or_category_property_value>  ("|" <script_or_category_property_value>)*
PROP_VALUE :=     <unicode_property_value> ("|" <unicode_property_value>)*


[\p{L} \p{Nd}] Match all letters and decimal digits
[\p{letter} \p{decimal number}]
[\p{letter|decimal number}]
\P{script=greek} Match anything that does not have the Greek script
\p{East Asian Width:Narrow} Match anything that has the East Asian Width property value of Narrow
\p{Whitespace} Match anything that has the binary property Whitespace

Some properties are binary: they are either true or false for a given code point. In that case, only the property name is required. Others have multiple values, so for uniqueness both the property name and the property value need to be included. For example, Alphabetic is both a binary property and a value of the Line_Break enumeration, so \p{Alphabetic} would mean the binary property, and \p{Line Break:Alphabetic} or \p{Line_Break=Alphabetic} would mean the enumerated property. There are two exceptions to this: the properties Script and General Category commonly have the property name omitted. Thus \p{Not_Assigned} is equivalent to \p{General_Category = Not_Assigned}, and \p{Greek} is equivalent to \p{Script:Greek}.

RL1.2 Properties
To meet this requirement, an implementation shall provide at least a minimal list of properties, consisting of the following:

The values for these properties must match the Unicode definitions.

RL1.2a Compatibility Properties
  To meet this requirement, an implementation shall provide the properties listed in Annex C. Compatibility Properties, with the property values as listed there. Such an implementation shall document whether it is using the Standard Recommendation or POSIX-compatible properties.

In order to meet requirements RL1.2 and RL1.2a, the implementation must satisfy the Unicode definition of the properties for the supported version of Unicode, rather than other possible definitions. However, the names used by the implementation for these properties may differ from the formal Unicode names for the properties. For example, if a regex engine already has a property called "Alphabetic", for backwards compatibility it may need to use a distinct name, such as "Unicode_Alphabetic", for the corresponding property listed in RL1.2.

For more information on properties, see UAX #44, Unicode Character Database [UAX44].

Of the properties in RL1.2, only General Category and Script have multiple values; the rest are binary. An implementation that does not support non-binary enumerated properties can essentially "flatten" the enumerated type. Thus, for example, instead of \p{script=latin} the syntax could be \p{script_latin}.

1.2.1 General Category Property

The most basic overall character property is the General Category, which is a basic categorization of Unicode characters into: Letters, Punctuation, Symbols, Marks, Numbers, Separators, and Other. These property values each have a single letter abbreviation, which is the uppercase first character except for separators, which use Z. The official data mapping Unicode characters to the General Category value is in UnicodeData.txt [UData].

Each of these categories has different subcategories. For example, the subcategories for Letter are uppercase, lowercase, titlecase, modifier, and other (in this case, other includes uncased letters such as Chinese). By convention, the subcategory is abbreviated by the category letter (in uppercase), followed by the first character of the subcategory in lowercase. For example, Lu stands for Uppercase Letter.

Note: Because it is recommended that the property syntax be lenient as to spaces, casing, hyphens and underbars, any of the following should be equivalent: \p{Lu}, \p{lu}, \p{uppercase letter}, \p{uppercase letter}, \p{Uppercase_Letter}, and \p{uppercaseletter}

The General Category property values are listed below. For more information on the meaning of these values, see UAX #44, Unicode Character Database [UAX44].

Abb. Long form
L Letter
Lu Uppercase Letter
Ll Lowercase Letter
Lt Titlecase Letter
Lm Modifier Letter
Lo Other Letter
M Mark
Mn Non-Spacing Mark
Mc Spacing Combining Mark
Me Enclosing Mark
N Number
Nd Decimal Digit Number
Nl Letter Number
No Other Number
Abb. Long form
S Symbol
Sm Math Symbol
Sc Currency Symbol
Sk Modifier Symbol
So Other Symbol
P Punctuation
Pc Connector Punctuation
Pd Dash Punctuation
Ps Open Punctuation
Pe Close Punctuation
Pi Initial Punctuation
Pf Final Punctuation
Po Other Punctuation
Abb. Long form
Z Separator
Zs Space Separator
Zl Line Separator
Zp Paragraph Separator
C Other
Cc Control
Cf Format
Cs Surrogate
Co Private Use
Cn Not Assigned
- Any*
- Assigned*
* The last few properties are not part of the General Category.
  • Any matches all code points. This could also be captured with [\u0000-\u10FFFF], but with Tailored Ranges off. In some regular expression languages, \p{Any} may be expressed by a period, but that may exclude newline characters.
  • Assigned is equivalent to \P{Cn}, and matches all assigned characters (for the target version of Unicode). It also includes all private use characters. It is useful for avoiding confusing double negatives. Note that Cn includes noncharacters, so Assigned excludes them.
  • ASCII is equivalent to [\u0000-\u007F], but with Tailored Ranges off.

1.2.2 Script Property

A regular-expression mechanism may choose to offer the ability to identify characters on the basis of other Unicode properties besides the General Category. In particular, Unicode characters are also divided into scripts as described in UAX #24: Unicode Script Property [UAX24] (for the data file, see Scripts.txt [ScriptData]). Using a property such as \p{Greek} allows implementations to test whether letters are Greek or not.

Note, however, that the usage model for the script property normally requires that people construct somewhat more complex regular expressions, because a great many characters are shared between scripts. Documentation should point users to the description in UAX #24.

There are situations where characters are regularly used with multiple scripts, including common characters such as U+30FC ( ー ) KATAKANA-HIRAGANA PROLONGED SOUND MARK. To account for such cases, support of the Script_Extensions property (abbreviated as scx) is recommended. Note, however, that the values for such a property are likely be extended over time as new information is gathered on the use of characters with different scripts. For more information, see Multiple Script Values in UAX #24: Unicode Script Property [UAX24].

The Script_Extensions property maps a code point to a set of scripts. The expression \p{scx=Hira} includes all the characters whose Script_Extensions value contains Hira. For example, [\p{script extensions=Hira}] contains not only the characters in \p{script=Hira}, but but (in Unicode 6.1) 256 characters such as U+30FC ( ー ), which is either Hiragana and Katakana.

While in most cases, \p{scx=X} ⊇ \p{sc=X}, in some cases that is not true. For example, the Script property value for U+30FC ( ー ) is Common, but the Script_Extensions value for U+30FC ( ー ) does not contain Common. So \p{scx=Common} ⊉ \p{sc=Common}.

1.2.3 Other Properties

Other recommended properties are described in 2.7 Full Properties. See also 2.5 Name Properties and 2.6 Wildcards in Property Values.

For example:

String properties Description
[:toNFC=Å:] The set of all characters X such that toNFC(X) = "a"
[:toNFD=A\u0300:] The set of all characters X such that toNFD(X) = "A\u0300"
[:toNFKC=A:] The set of all characters X such that toNFKC(X) = "A"
[:toNFKD=A\u0300:] The set of all characters X such that toNFKD(X) = "a"
[:toLowercase=a:] The set of all characters X such that toLowercase(X) = "a"
[:toUppercase=A:] The set of all characters X such that toUppercase(X) = "A"
[:toTitlecase=A:] The set of all characters X such that toTitlecase(X) = "A"
[:toCaseFold=a:] The set of all characters X such that toCasefold(X) = "A"
Binary properties Description
[:isNFC:] The set of all characters X such that toNFC(X) = X
[:isNFD:] The set of all characters X such that toNFD(X) = X
[:isNFKC:] The set of all characters X such that toNFKC(X) = X
[:isNFKD:] The set of all characters X such that toNFKD(X) = X
[:isLowercase:] The set of all characters X such that toLowercase(X) = X
[:isUppercase:] The set of all characters X such that toUppercase(X) = X
[:isTitlecase:] The set of all characters X such that toTitlecase(X) = X
[:isCaseFolded:] The set of all characters X such that toCasefo(X) = X
[:isCased:] The set of all

1.2.4 Age

As defined in the Unicode Standard, the Age property (in the DerivedAge data file in the UCD) specifies the first version of the standard in which each character was assigned. It does not refer to how long it has been encoded, nor does it indicate the historic status of the character.

In regex expressions, the age property is used to indicate the characters that were in a particular version of Unicode. That is, a character has the Unicode Age property of that version or less. Thus \p{age=3.0} includes the letter a, which was included in Unicode 1.0. To get characters that are new in a particular version, subtract off the previous version as described in 1.3 Subtraction and Intersection. For example: [\p{age=3.1} -- \p{age=3.0}].

1.2.5 Blocks

Unicode blocks can sometimes also be a useful enumerated property. However, there are some very significant caveats to the use of Unicode blocks for the identification of characters: see Annex A. Character Blocks. If blocks are used, some of the names can collide with Script names, so they should be distinguished, with syntax such as \p{Greek Block} or \p{Block=Greek}.

1.3 Subtraction and Intersection

As discussed earlier, character properties are essential with a large character set. In addition, there needs to be a way to "subtract" characters from what is already in the list. For example, one may want to include all non-ASCII letters without having to list every character in \p{letter} that is not one of those 52.

RL1.3 Subtraction and Intersection
To meet this requirement, an implementation shall supply mechanisms for union, intersection and set-difference of Unicode sets.


ITEM     := "[" ITEM "]" // for grouping
OPERATOR := "" // no separator = union
         := "||" // union: A∪B
         := "&&" // intersection: A∩B
         := "--" // set difference: A-B
         := "~~" // symmetric difference: A⊖B = (A∪B)-(A∩B)

Implementations may also choose to offer other set operations. The symmetric difference of two sets is particularly useful. It is defined as being the union minus the intersection. Thus [\p{letter}~~\p{ascii}] is equivalent to [[\p{letter}\p{ascii}]--[\p{letter}&&\p{ascii}]].

For compatibility with industry practice, symbols are doubled in the above notation. This practice provides for better backwards compatibility with expressions using older syntax, because they are unlikely to contain doubled characters. It also allows the operators to appear adjacent to ranges without ambiguity, such as [\p{letter}--a-z].

Binding or precedence may vary by regular expression engine, so it is safest to always disambiguate using brackets to be sure. In particular, precedence may put all operators on the same level, or may take union as binding more closely. For example, where A..E stand for expressions, not characters:

Expression Equals When
[ABC--DE] [[AB]C]--[DE]] Union binds more closely. That is, it means:

Form the union of A, B, and C, and then subtract the union of D and E.

[[[[[AB]C]--D]E]] Operators are on the same level. That is, it means:

Form the union of A, B, and C, and then subtract D, and then add E.

Even where an expression is not ambiguous, extra grouping brackets may be useful for clarity.


[\p{L}--QW] Match all letters but Q and W
[\p{N}--[\p{Nd}--0-9]] Match all non-decimal numbers, plus 0-9
[\u0000-\u007F--\P{letter}] Match all letters in the ASCII range, by subtracting non-letters
[\p{Greek}--\N{GREEK SMALL LETTER ALPHA}] Match Greek letters except alpha
[\p{Assigned}--\p{Decimal Digit Number}--a-fA-Fa-fA-F] Match all assigned characters except for hex digits (using a broad definition)

1.4 Simple Word Boundaries

Most regular expression engines allow a test for word boundaries (such as by "\b" in Perl). They generally use a very simple mechanism for determining word boundaries: one example of that would be having word boundaries between any pair of characters where one is a <word_character> and the other is not, or at the start and end of a string. This is not adequate for Unicode regular expressions.

RL1.4 Simple Word Boundaries
To meet this requirement, an implementation shall extend the word boundary mechanism so that:
  1. The class of <word_character> includes all the Alphabetic values from the Unicode character database, from UnicodeData.txt [UData], plus the decimal numbers (General_Category=Decimal_Number), and the U+200C ZERO WIDTH NON-JOINER and U+200D ZERO WIDTH JOINER (Join_Control=True). See also Annex C: Compatibility Properties.
  2. Nonspacing marks are never divided from their base characters, and otherwise ignored in locating boundaries.

Level 2 provides more general support for word boundaries between arbitrary Unicode characters which may override this behavior.

1.5 Simple Loose Matches

Most regular expression engines offer caseless matching as the only loose matching. If the engine does offers this, then it needs to account for the large range of cased Unicode characters outside of ASCII.

RL1.5 Simple Loose Matches
To meet this requirement, if an implementation provides for case-insensitive matching, then it shall provide at least the simple, default Unicode case-insensitive matching.

To meet this requirement, if an implementation provides for case conversions, then it shall provide at least the simple, default Unicode case conversion.

In addition, because of the vagaries of natural language, there are situations where two different Unicode characters have the same uppercase or lowercase. To meet this requirement, implementations must implement these in accordance with the Unicode Standard. For example, the Greek U+03C3 "σ" small sigma, U+03C2 "ς" small final sigma, and U+03A3 "Σ" capital sigma all match.

Some caseless matches may match one character against two: for example, U+00DF "ß" matches the two characters "SS". And case matching may vary by locale. However, because many implementations are not set up to handle this, at Level 1 only simple case matches are necessary. To correctly implement a caseless match, see Chapter 3 of the Unicode Standard [Unicode]. The data file supporting caseless matching is CaseFolding.txt [CaseData].

To meet this requirement, where an implementation also offers case conversions, these must also follow Chapter 3 Conformance of [Unicode]. The relevant data files are SpecialCasing.txt [SpecialCasing] and UnicodeData.txt [UData].

Matching case-insensitively is one example of matching under an equivalence relation:

A regular expression R matches under an equivalence relation E whenever for all strings S and T:

If S is equivalent to T under E, then R matches S if and only if R matches T.

In the Unicode Standard, the relevant equivalence relation for case-insensitivity is established according to whether two strings case fold to the same value. The case folding can either be simple (a 1:1 mapping of code points) or full (with some 1:n mappings).

In practice, regex APIs are not set up to match parts of characters. For this reason, full case equivalence is difficult to handle with regular expressions. For more information, see Section 2.1 Canonical Equivalents.

A regular expression pattern R can be made to match case-insensitively (simple ), by making the following changes in the interpretation of R:

  1. Each string literal is matched case-insensitively. That is, it is logically expanded into a sequence of OR expressions, where each OR expression lists all of the characters that have a simple case-folding to the same value.
    • For example, /Dåb/ matches as if it were expanded into /(?:d|D)(?:å|Å|Å)(?:b|B)/
    • Back references are subject to this logical expansion, such as /(?i)(a.c)\1/, where \1 matches what is in the first grouping.
  2. Each character class is closed under case. That is, it is logically expanded into a set of code points, and then closed by adding all simple case equivalents of each of those code points.
    • For example, [\p{Block=Phonetic_Extensions} [A-E]] is a character class that matches 133 code points (under Unicode 6.0). Its case-closure adds 7 more code points: a-e, Ᵽ, and Ᵹ, for a total of 140 code points.

For both property character classes and explicit character classes, closing under simple case-insensitivity means including characters not in the set. For example:

[Review Note: Should we have options for different case folding for: literals vs .?, ranges, and properties?]

1.6 Line Boundaries

Most regular expression engines also allow a test for line boundaries: end-of-line or start-of-line. This presumes that lines of text are separated by line (or paragraph) separators.

RL1.6 Line Boundaries
To meet this requirement, if an implementation provides for line-boundary testing, it shall recognize not only CRLF, LF, CR, but also NEL (U+0085), PS (U+2029) and LS (U+2028).

Formfeed (U+000C) also normally indicates an end-of-line. For more information, see Chapter 3 of [Unicode].

These characters should be uniformly handled in determining logical line numbers, start-of-line, end-of-line, and arbitrary-character implementations. Logical line number is useful for compiler error messages and the like. Regular expressions often allow for SOL and EOL patterns, which match certain boundaries. Often there is also a "non-line-separator" arbitrary character pattern that excludes line separator characters.

The behavior of these characters may also differ depending on whether one is in a "multiline" mode or not. For more information, see Anchors and Other "Zero-Width Assertions" in Chapter 3 of [Friedl].

A newline sequence is defined to be any of the following:

\u000A | \u000B | \u000C | \u000D | \u0085 | \u2028 | \u2029 | \u000D\u000A

  1. Logical line number
    • The line number is increased by one for each occurrence of a newline sequence.
    • Note that different implementations may call the first line either line zero or line one.
  2. Logical beginning of line (often "^")
    • SOL is at the start of a file or string, and depending on matching options, also immediately following any occurrence of a newline sequence.
    • There is no empty line within the sequence \u000D\u000A, that is, between the first and second character.
    • Note that there may be a separate pattern for "beginning of text" for a multiline mode, one which matches only at the beginning of the first line, e.g., in Perl \A.
  3. Logical end of line (often "$")
    • EOL at the end of a file or string, and depending on matching options, also immediately preceding a final occurrence of a newline sequence.
    • There is no empty line within the sequence \u000D\u000A, that is, between the first and second character.
    • SOL and EOL are not symmetric because of multiline mode: EOL can be interpreted in at least three different ways:
      1. EOL matches at the end of the string
      2. EOL matches before final newline
      3. EOL matches before any newline
  4. Arbitrary character pattern (often ".")
    • Where the 'arbitrary character pattern' matches a newline sequence, it must match all of the newline sequences, and \u000D\u000A (CRLF) should match as if it were a single character. (The recommendation that CRLF match as a single character is, however, not required for conformance to RL1.6.)
    • Note that ^$ (an empty line pattern) should not match the empty string within the sequence \u000D\u000A, but should match the empty string within the reversed sequence \u000A\u000D.

It is strongly recommended that there be a regular expression meta-character, such as "\R", for matching all line ending characters and sequences listed above (e.g. in #1). It would thus be shorthand for:

( \u000D\u000A | [\u000A\u000B\u000C\u000D\u0085\u2028\u2029] )

Note: For some implementations, there may be a performance impact in recognizing CRLF as a single entity, such as with an arbitrary pattern character ("."). To account for that, an implementation may also satisfy R1.6 if there is a mechanism available for converting the sequence CRLF to a single line boundary character before regex processing.

For more information on line breaking, see [UAX14].

1.7 Code Points

A fundamental requirement is that Unicode text be interpreted semantically by code point, not code units.

RL1.7 Supplementary Code Points
To meet this requirement, an implementation shall handle the full range of Unicode code points, including values from U+FFFF to U+10FFFF. In particular, where UTF-16 is used, a sequence consisting of a leading surrogate followed by a trailing surrogate shall be handled as a single code point in matching.

UTF-16 uses pairs of Unicode code units to express code points above FFFF16. Surrogate pairs (or their equivalents in other encoding forms) are be handled internally as single code point values. In particular, [\u{0}-\u{10000}] will match all the following sequence of code units:

Code Point UTF-8 Code Units UTF-16 Code Units UTF-32 Code Units
7F 7F 007F 0000007F
80 C2 80 0080 00000080
7FF DF BF 07FF 000007FF
800 E0 A0 80 0800 00000800
10000 F0 90 80 80 D800 DC00 00010000

Note: It is permissible, but not required, to match an isolated surrogate code point (such as \u{D800}). These can only occur in Unicode 16-bit and 32-bit strings. See Unicode_String in the Unicode glossary.

2 Extended Unicode Support: Level 2

Level 1 support works well in many circumstances. However, it does not handle more complex languages or extensions to the Unicode Standard very well. Particularly important cases are canonical equivalence, word boundaries, extended grapheme cluster boundaries, and loose matches. (For more information about boundary conditions, see UAX #29: Unicode Text Segmentation [UAX29].)

Level 2 support matches much more what user expectations are for sequences of Unicode characters. It is still locale-independent and easily implementable. However, for compatibility with Level 1, it is useful to have some sort of syntax that will turn Level 2 support on and off.

[Review note: The UTC has received a request for UCA matching according to the DUCET rules, at level 2. Feedback on this issue would be appreciated. Note that such matching would run into the same issues as full case and canonical equivalence matching (characters splitting, combining, and reordering), and thus would only be able to produce an approximation of DUCET equivalence. It would probably also only be primary-weight equivalence.]

The features comprising Level 2 are not in order of importance. In particular, the most useful and highest priority features in practice are:

2.1 Canonical Equivalents

The equivalence relation for canonical equivalence is established by whether two strings are identical when normalized to NFD.

For most full-featured regular expression engines, it is quite difficult to match under canonical equivalence. This results from the fact that canonical equivalence may involve reordering, splitting, or merging of characters. For example, all of the following sequences are canonically equivalent:

The regular expression pattern /o\x{31B}/ matches the first two characters of #1, the first and third characters of #2, the first character of #3, part of the first character together with the third character of #4, and part of the character in #5.

In practice regex APIs are not set up to match parts of characters or handle discontiguous selections.

There are many other edge cases: A combining mark may come from some part of the pattern far removed from where the base character was, or may not explicitly be in the pattern at all. It is also unclear what /./ should match and how back references should work.

It is feasible, however, to construct patterns that will match against NFD (or NFKD) text. That can be done by:

  1. Putting the text to be matched into a defined normalization form (NFD or NFKD).
  2. Having the user design the regular expression pattern to match against that defined normalization form. For example, the pattern should contain no characters that would not occur in that normalization form, nor sequences that would not occur.

Matching proceeds on a code point by code point basis, as usual.

There are many instances where a character can be equivalently expressed by two different sequences of Unicode characters. For example, [ä] should match both "ä" and "a\u0308". (See UAX #15: Unicode Normalization [UAX15] and Sections 2.5 and 3.9 of The Unicode Standard [Unicode] for more information.)

RL2.1 Canonical Equivalents
To meet this requirement, an implementation shall provide a mechanism for ensuring that all canonically equivalent literal characters match.

One of the most effective ways to implement canonical equivalence is by having a special mode that only matches on extended grapheme cluster boundaries, since it avoids the reordering problems that can happen in normalization. See RL2.2 Extended Grapheme Clusters.

There are two other options for implementing this:

  1. Before (or during) processing, translate text (and pattern) into a normalized form. This is the simplest to implement, because there are available code libraries for doing normalization.
  1. Expand the regular expression internally into a more generalized regular expression that takes canonical equivalence into account. For example, the expression [a-z ä] can be internally turned into ([a-z ä] | (a \u0308)). While this can be faster, it may also be substantially more difficult to generate expressions capturing all of the possible equivalent sequences.

It may be useful to distinguish a regular-expression engine from the larger software package which uses it. For example, the requirements of this section can be met by requiring the package to normalize text before supplying it to the regular expression engine. However, where the regular expression engine returns offsets into the text, the package may need to map those back to what the offsets would be in the original, unnormalized text.

Note: Combining characters are required for many languages. Even when text is in Normalization Form C, there may be combining characters in the text.

2.2 Extended Grapheme Clusters

One or more Unicode characters may make up what the user thinks of as a character. To avoid ambiguity with the computer use of the term character, this is called a grapheme cluster. For example, "G" + acute-accent is a grapheme cluster: it is thought of as a single character by users, yet is actually represented by two Unicode characters. The Unicode Standard defines extended grapheme clusters that keep Hangul syllables together and do not break between base characters and combining marks. The precise definition is in UAX #29: Unicode Text Segmentation [UAX29]. These extended grapheme clusters are not the same as tailored grapheme clusters, which are covered in Level 3, Tailored Grapheme Clusters.

RL2.2 Extended Grapheme Clusters
To meet this requirement, an implementation shall provide a mechanism for matching against an arbitrary extended grapheme cluster, a literal cluster, and matching extended grapheme cluster boundaries.

For example, an implementation could interpret \X as matching any extended grapheme cluster, while interpreting "." as matching any single code point. It could interpret \b{g} as a zero-width match against any extended grapheme cluster boundary, and \B{g} as the negation of that.

More generally, it is useful to have zero width boundary detections for each of the different kinds of segment boundaries defined by Unicode ([UAX29] and [UAX14]). For example:

Syntax Description
\b{g} Zero-width match at a Unicode extended grapheme cluster boundary
\b{w} Zero-width match at a Unicode word boundary. Note that this is different than \b alone, which corresponds to \w and \W. See Annex C: Compatibility Properties.
\b{l} Zero-width match at a Unicode line break boundary
\b{s} Zero-width match at a Unicode sentence boundary

Thus \X is equivalent to .+?\b{g}; proceed the minimal number of characters (but at least one) to get to the next extended grapheme cluster boundary.

Regular expression engines should also provide some mechanism for easily matching against literal clusters, because they are more likely to match user expectations for many languages. One mechanism for doing that is to have explicit syntax for literal clusters, as in the following syntax:

ITEM := "\q{" CODE_POINT + "}"

This syntax can also be used for tailored grapheme clusters (see Tailored Grapheme Clusters).


[a-z\q{x\u0323}] Match a-z, and x with an under-dot (used in American Indian languages).
[a-z\q{aa}] Match a-z, and aa (treated as a single character in Danish).
[a-z ñ \q{ch} \q{ll} \q{rr}] Match some lowercase characters in traditional Spanish.

In implementing extended grapheme clusters, the expression /[a-m \q{ch} \q{rr}]/ should behave like /(?: ch | rr | [a-m])/. That is, the expression would:

Note that the strings need to be ordered as longest first to work correctly in arbitrary regex engines, because some regex engines try the leftmost matching alternative first. For example, the expression /[a-m {ch} {chh}]/ would need to behave like /(?: chh | ch | [a-m])/, with "chh" before "ch".

Matching a complemented set containing strings like \q{ch} may behave differently in the two different modes: the normal mode where code points are the unit of matching, or the mode where extended grapheme clusters are the unit of matching. That is, the expression [^ a-m \q{ch} \q{rr}] should behave in the following way:

Mode Behavior Description
normal (?! ch | rr | [a-m] )
failing with strings starting with a-m, ch, or rr, and otherwise advancing by one code point
grapheme cluster (?! ch | rr | [a-m] )
failing with strings starting with a-m, ch, or rr, and otherwise advancing by one extended grapheme cluster

A complex character set containing strings like \q{ch} plus embedded complement operations is interpreted as if the complement were pushed up to the top of the expression, using the following rewrites recursively:

Original Rewrite
^^x x
^x || ^y ^(x && y)
^x || y ^(x -- y)
x || ^y ^(y -- x)
^x && ^y ^(x || y)
^x -- y
^x && y y -- x
^x -- ^y
x && ^y x -- y
x -- ^y x && y
^x ~~ ^y x ~~ y
^x ~~ y ^(x ~~ y)
x ~~ ^y

Applying these rewrites results in a simplification of the regex expression. Either the complement operations will be completely eliminated, or a single remaining complement operation will remain at the top level of the expression. Logically, then, the rest of the expression consists of a flat list of characters and/or multi-character strings; matching strings can then can be handled as described above.

2.2.1 Grapheme Cluster Mode

A grapheme cluster mode behaves more like users' expectations for character boundaries, and is especially useful for handling canonically equivalent matching. In a grapheme cluster mode, matches are guaranteed to be on extended grapheme cluster boundaries. Each atomic literal of the pattern matches complete extended grapheme clusters, and thus behaves as if followed by \b{g}. Atomic literals include: a dot, a character class (like [a-m]), a sequence of characters (perhaps with some being escaped) that matches as a unit, or syntax that is equivalent to these. Note that in /abc?/, the "abc" is not matching as a unit; the ? modifier is only affecting the last character, and thus the ab and the c are separate atomic literals. To summarize:

Syntax Description
. Behaves like \X; that is, matches a full extended grapheme cluster going forward.
[abc{gh}] Behaves like [abc{gh}]\b{g}; that is, matches only if the end point of the match is at a grapheme cluster boundary
abcd Behaves like abcd\b{g}; that is, matches only if the end point of the match is at a grapheme cluster boundary

Note that subdivisions can modify the behavior in this mode. Normally /(xy)/ is equivalent to /(x)(y)/ in terms of matching (where x and y are arbitrary literal character strings); that is, only the grouping is different. That is not true in grapheme cluster mode, where each atomic literal acts like it is followed by \b{g}. For example, /(x\u0308)/ is not the same as /(x)(\u0308)/ in matching. The former behaves like /(x\u0308\b{g})/ while the latter behaves like /(x\b{g})(\u0308\b{g})/. The latter will never match in grapheme cluster mode, since it would only match if there were a grapheme cluster boundary after the x and if x is followed by \u0308, but that can never happen simultaneously.

2.3 Default Word Boundaries

RL2.3 Default Word Boundaries
To meet this requirement, an implementation shall provide a mechanism for matching Unicode default word boundaries.

The simple Level 1 support using simple <word_character> classes is only a very rough approximation of user word boundaries. A much better method takes into account more context than just a single pair of letters. A general algorithm can take care of character and word boundaries for most of the world's languages. For more information, see UAX #29: Unicode Text Segmentation [UAX29].

Note: Word boundaries and "soft" line-break boundaries (where one could break in line wrapping) are not generally the same; line breaking has a much more complex set of requirements to meet the typographic requirements of different languages. See UAX #14: Line Breaking Properties [UAX14] for more information. However, soft line breaks are not generally relevant to general regular expression engines.

A fine-grained approach to languages such as Chinese or Thai languages that do not use spaces requires information that is beyond the bounds of what a Level 2 algorithm can provide.

2.4 Default Case Conversion

RL2.4 Default Case Conversion
To meet this requirement:
  • if an implementation provides for case-insensitive matching, then it shall provide at least the full, default Unicode case-insensitive matching.
  • if an implementation provides for case conversions, then it shall provide at least the full, default Unicode case conversion.

[Review Note: the above restriction is from PRI #179 Changes to Unicode Regular Expression Guidelines.]

At Level 1, caseless matches do not need to handle cases where one character matches against two. Level 2 includes caseless matches where one character may match against two (or more) characters. For example, 00DF "ß" will match against the two characters "SS".

To correctly implement a caseless match and case conversions, see UAX #21: Case Mappings [Case]. For ease of implementation, a complete case folding file is supplied at CaseFolding.txt [CaseData]. Full case mappings use the data files If the implementation containing the regular expression engine also offers case conversions, then these should also be done in accordance with UAX #21, with the full mappings. The relevant data files are SpecialCasing.txt [SpecialCasing] and UnicodeData.txt [UData].

2.5 Name Properties

RL2.5 Name Properties
To meet this requirement, an implementation shall support individually named characters.

When using names in regular expressions, the main data is supplied in the Name and Name_Alias properties in the UCD, as described in UAX #44, Unicode Character Database [UAX44], or computed as in the case of CJK Ideographs or Hangul Syllables.


Syntax Description
\p{name=ZERO WIDTH NO-BREAK SPACE} [\u{FEFF}], using the Name property.
\p{name=BYTE ORDER MARK} [\u{FEFF}], using the Name_Alias property.
\p{name=BOM} [\u{FEFF}], using the Name_Alias property (a second value).
\p{HANGUL SYLLABLE GAG} [\u{AC01}], with a computed name.

Certain code points are not assigned names or name aliases in the standard. These should be given names based on the General_Category:

Control: The Unicode 1.0 name field (ISO control names).
Private Use: <no name>
\p{name=<private-use-E000>} [\u{E000}]
\p{name=<surrogate-D800>} [\u{D800}] Note: this would only apply to isolated surrogate code points.
\p{name=<noncharacter-FDD0>} [\u{FDD0}]
\p{name=<unassigned-30000>} [\u{30000}]

[Review Note: the above names have been aligned with what the text says and what is shown in the UCD files. The utility of these names is (over simply using an expression like [\u{FDD0}) for "documenting" the intent of the regex. Is it worth retaining?

One problematic area is unassigned code points: U+058F is new in 6.1; we don't want \p{unassigned-058F} to work for U6.0 but then break in U6.1.]

The ISO names for the control characters may be unfamiliar, especially because many people are not familiar with changes in the formal ISO names to make them more language neutral, so it is recommended that they be supplemented with other aliases. For example, for U+0009 the implementation could accept the official name CHARACTER TABULATION, and also the aliases HORIZONTAL TABULATION, HT, and TAB.

Implementers may add additional names to what is recognized according to the UCD, but must be aware that such names may cause problems if they collide with future UCD names.

2.5.1 Individually Named Characters

The following provides syntax for specifying a code point by supplying the precise name. This syntax specifies a single code point, which can thus be used in ranges.

<codepoint> := "\N{" <character_name> "}"

This is equivalent to using the property name, as in \p{name=WHITE SMILING FACE}. The only distinction between them is that \N should cause a syntax error if it fails to match a single character, and syntactically is equivalent to a sequence of one or more characters, not a set. The <character_name> also includes named character sequences (from the file NamedSequences.txt), such as \N{KHMER CONSONANT SIGN COENG KA}. Note that because this is a sequence it behaves as a single element, so \N{KHMER CONSONANT SIGN COENG KA}* should be treated as if it were the expression (\u{17D2|\u{1780})*.

As with other property values, names should use a loose match, disregarding case, spaces and hyphen (the underbar character "_" cannot occur in Unicode character names). An implementation may also choose to allow namespaces, where some prefix like "LATIN LETTER" is set globally and used if there is no match otherwise.

There are, however, three instances that require special-casing with loose matching, where an extra test shall be made for the presence or absence of a hyphen.


2.6 Wildcards in Property Values

RL2.6 Wildcards in Property Values
To meet this requirement, an implementation shall support wildcards in Unicode property values.

Instead of a single property value, this feature allows the use of a regular expression to pick out a set of characters based on whether the property values match the regular expression. The regular expression must support at least wildcards; other regular expressions features are recommended but optional.

PROP_VALUE := <value>
            | "/" <regex expression> "/"

Note: Where regular expressions are using in matching, case, spaces, hyphen, and underbar are significant; it is presumed that users will make use of regular-expression features to ignore these if desired.


Expression Description/Contents
\p{toNfd=/b/} Characters whose NFD form contains a "b" (U+0062) in the value.
\p{name=/^LATIN LETTER.*P$/} Characters with names starting with "LATIN LETTER" and ending with "P"
\p{name=/VARIA(TION|NT)/} Characters with names containing "VARIATION" or "VARIANT"

The above are all on the basis of Unicode 5.0; different versions of Unicode may produce different results.

Here are some additional samples, illustrating various sets. If you click on them, they will use the online Unicode utilities on the Unicode website to show the contents of the sets. Note that these online utilities curently use single-letter operations:

Expression Description
[[:name=/CJK/:]-[:ideographic:]] The set of all characters with names that contain CJK that are not Ideographic
[:name=/\bDOT$/:] The set of all characters with names that end with the word DOT
[:block=/(?i)arab/:] The set of all characters in blocks that contain the sequence of letters "arab" (case-insensitive)
[:toNFKC=/\./:] the set of all characters with toNFKC values that contain a literal period

2.7 Full Properties

RL2.7 Full Properties
To meet this requirement, an implementation shall support all of the properties listed below that are in the supported version of Unicode, with values that match the Unicode definitions for that version.

As in RL1.2 Properties, in order to meet requirement RL2.7, the implementation must satisfy the Unicode definition of the properties for the supported version of Unicode, rather than other possible definitions. However, the names used by the implementation for these properties may differ from the formal Unicode names for the properties. For example, if a regex engine already has a property called "Alphabetic", for backwards compatibility it may need to use a distinct name, such as "Unicode_Alphabetic", for the corresponding property listed in RL1.2.

The list excludes contributory properties, obsolete and deprecated properties, most provisional properties, and the Unicode 1 Name and Unicode Radical Stroke properties. The properties in gray are covered by RL1.2 Properties. For more information on properties, see UAX #44, Unicode Character Database [UAX44].

General Case Shaping and Rendering
Name Uppercase Join_Control
Name_Alias Lowercase Joining_Group
Block Lowercase_Mapping Joining_Type
Age Titlecase_Mapping Line_Break
General_Category Uppercase_Mapping Grapheme_Cluster_Break
Script Case_Folding Sentence_Break
Script_Extensions Simple_Lowercase_Mapping Word_Break
White_Space Simple_Titlecase_Mapping East_Asian_Width
Alphabetic Simple_Uppercase_Mapping
Hangul_Syllable_Type Simple_Case_Folding Bidirectional
Noncharacter_Code_Point Soft_Dotted Bidi_Class
Default_Ignorable_Code_Point Cased Bidi_Control
Deprecated Case_Ignorable Bidi_Mirrored
Logical_Order_Exception Changes_When_Lowercased Bidi_Mirroring_Glyph
Variation_Selector Changes_When_Uppercased
  Changes_When_Titlecased CJK
Numeric Changes_When_Casefolded Ideographic
Numeric_Value Changes_When_Casemapped Unified_Ideograph
Numeric_Type Radical
Hex_Digit Normalization IDS_Binary_Operator
ASCII_Hex_Digit Canonical_Combining_Class IDS_Trinary_Operator
Identifiers Composition_Exclusion Miscellaneous
ID_Continue Full_Composition_Exclusion Math
ID_Start Decomposition_Type Quotation_Mark
XID_Continue NFC_Quick_Check Dash
XID_Start NFKC_Quick_Check STerm
Pattern_Syntax NFD_Quick_Check Terminal_Punctuation
Pattern_White_Space NFKD_Quick_Check Diacritic
  NFKC_Casefold Extender
  Changes_When_NFKC_Casefolded Grapheme_Base

There are situations where characters are regularly used with multiple scripts, including common characters such as U+30FC ( ー ) KATAKANA-HIRAGANA PROLONGED SOUND MARK. To account for such cases, support of the Script_Extensions property (abbreviated as scx) is recommended. For more information, see Section 1.2.2 Script_Property.

3 Tailored Support: Level 3

All of the above deals with a default specification for a regular expression. However, a regular expression engine also may want to support tailored specifications, typically tailored for a particular language or locale. This may be important when the regular expression engine is being used by end-users instead of programmers, such as in a word-processor allowing some level of regular expressions in searching.

For example, the order of Unicode characters may differ substantially from the order expected by users of a particular language. The regular expression engine has to decide, for example, whether the list [a-ä] means:

If both tailored and default regular expressions are supported, then a number of different mechanism are affected. There are two main alternatives for control of tailored support:

For example, fine-grained support could use some syntax such as the following to indicate tailoring to a locale within a certain range. Locale (or language) IDs should use the syntax from locale identifier definition in [UTS35], Section 3. Identifiers. Note that the locale id of "root" or "und" indicates the root locale, such as for indicating the CLDR root collation.


There must be some sort of syntax that will allow Level 3 support to be turned on and off, for two reasons. Level 3 support may be considerably slower than Level 2, and most regular expressions may require Level 1 or Level 2 matches to work properly. The syntax should also specify the particular locale or other tailoring customization that the pattern was designed for, because tailored regular expression patterns are usually quite specific to the locale, and will generally not work across different locales.

Sections 3.6 and following describe some additional capabilities of regular expression engines that are very useful in a Unicode environment, especially in dealing with the complexities of the large number of writing systems and languages expressible in Unicode.

3.1 Tailored Punctuation

The Unicode character properties for punctuation may vary from language to language or from country to country. In most cases, the effects of such changes will be apparent in other operations, such as a determination of word breaks. But there are other circumstances where the effects should be apparent in the general APIs, such as when testing whether a curly quotation mark is opening or closing punctuation.

RL3.1 Tailored Punctuation
To meet this requirement, an implementation shall allow for punctuation properties to be tailored according to locale, using the locale identifier definition in [UTS35], Section 3. Identifiers.

As just described, there must be the capability of turning this support on or off.

3.2 Tailored Grapheme Clusters

RL3.2 Tailored Grapheme Clusters
To meet this requirement, an implementation shall provide for collation grapheme clusters matches based on a locale's collation order.

Tailored grapheme clusters may be somewhat different than the extended grapheme clusters discussed in Level 2. They are coordinated with the collation ordering for a given language in the following way. A collation ordering determines a collation grapheme cluster, which is a sequence of characters that is treated as a unit by the ordering. For example, ch is a collation grapheme cluster for a traditional Spanish ordering. More specifically, a collation grapheme cluster is the longest sequence of characters that maps to a sequence of one or more collation elements where the first collation element has a primary weight and subsequent elements do not, and no completely ignorable characters are included.

The tailored grapheme clusters for a particular locale are the collation grapheme clusters for the collation ordering for that locale. The determination of tailored grapheme clusters requires the regular expression engine to either draw upon the platform's collation data, or incorporate its own tailored data for each supported locale.

See UTS #10: Unicode Collation Algorithm [UTS10] for more information about collation, and Annex B. Sample Collation Grapheme Cluster Code for sample code.

3.3 Tailored Word Boundaries

RL3.3 Tailored Word Boundaries
To meet this requirement, an implementation shall allow for the ability to have word boundaries to be tailored according to locale.

Semantic analysis may be required for correct word boundary detection in languages that do not require spaces, such as Thai, Japanese, Chinese or Korean. This can require fairly sophisticated support if Level 3 word boundary detection is required, and usually requires drawing on platform OS services.

3.4 Tailored Loose Matches

RL3.4 Tailored Loose Matches
To meet this requirement, an implementation shall provide for loose matches based on a locale's collation order, with at least 3 levels.

In Level 1 and 2, caseless matches are described, but there are other interesting linguistic features that users may want to match. For example, V and W are considered equivalent in traditional Swedish collations, and so [V] should match W in Swedish. In line with the UTS #10: Unicode Collation Algorithm [UTS10], the following four levels of equivalences are recommended:

If users are to have control over these equivalence classes, here is an example of how the sample syntax could be modified to account for this. The syntax for switching the strength or type of matching varies widely. Note that these tags switch behavior on and off in the middle of a regular expression; they do not match a character.

ITEM := \v{PRIMARY}   // match primary only
ITEM := \v{SECONDARY} // match primary and secondary only
ITEM := \v{TERTIARY}  // match primary, secondary, and tertiary
ITEM := \v{EXACT}     // match all levels, normal state


[\v{SECONDARY}a-m] Match a-m, plus case variants A-M, plus compatibility variants

Basic information for these equivalence classes can be derived from the data tables referenced by UTS #10: Unicode Collation Algorithm [UTS10].

Another example of tailored loose matching is matching between Traditional and Simplified Chinese characters.

3.5 Tailored Ranges

RL3.5 Tailored Ranges
To meet this requirement, an implementation shall provide for ranges based on a locale's collation order.

Tailored character ranges will include tailored grapheme clusters, as discussed above. This broadens the set of grapheme clusters — in traditional Spanish, for example, [b-d] would match against "c" and against "ch". That is because in that collation, "ch" sorts as a single letter between "c" and "d".

Note: This is another reason why a property for all characters \p{Any} is needed—it is possible for a locale's collation to not have [\u{0}-\u{10FFFF}] encompass all characters.

Tailored ranges can be quite difficult to implement properly, and can have very unexpected results in practice. For example, languages may also vary whether they consider lowercase below uppercase or the reverse. This can have some surprising results: [a-Z] may not match anything if Z < a in that locale. Thus implementers should be cautious about implementing this feature.

3.6 Context Matching

RL3.6 Context Matching
To meet this requirement, an implementation shall provide for a restrictive match against input text, allowing for context before and after the match.

For parallel, filtered transformations, such as those involved in script transliteration, it is important to restrict the matching of a regular expression to a substring of a given string, and yet allow for context before and after the affected area. Here is a sample API that implements such functionality, where m is an extension of a Regex Matcher.

  if (m.matches(text, contextStart, targetStart, targetLimit, contextLimit)) {
    int end = p.getMatchEnd();

The range of characters between contextStart and targetStart define a precontext; the characters between targetStart and targetLimit define a target, and the offsets between targetLimit and contextLimit define a postcontext. Thus contextStarttargetStarttargetLimitcontextLimit. The meaning of this function is that:


In these examples, the text in the pre- and postcontext is italicized and the target is underlined. In the output column, the text in bold is the matched portion. The pattern syntax "(←x)" means a backwards match for x (without moving the cursor) This would be (?<=x) in Perl. The pattern "(→x)" means a forwards match for x (without moving the cursor). This would be (?=x) in Perl.

Pattern Input Output Comment
/(←a) (bc)* (→d)/ 1abcbcd2 1abcbcd2 matching with context
/(←a) (bc)* (→bcd)/ 1abcbcd2 1abcbcd2 stops early, because otherwise 'd' would not match.
/(bc)*d/ 1abcbcd2 no match 'd' ca not be matched in the target, only in the postcontext
/(←a) (bc)* (→d)/ 1abcbcd2 no match 'a' ca not be matched, because it is before the precontext (which is zero-length, in this case)

While it would be possible to simulate this API call with other regular expression calls, it would require subdividing the string and making multiple regular expression engine calls, significantly affecting performance.

There should also be pattern syntax for matches (like ^ and $) for the contextStart and contextLimit positions.

Internally, this can be implemented by modifying the regular expression engine so that all matches are limited to characters between contextStart and contextLimit, and so that all matches that are not zero-width look-arounds are limited to the characters between targetStart and targetLimit.

3.7 Incremental Matches

RL3.7 Incremental Matches
To meet this requirement, an implementation shall provide for incremental matching.

For buffered matching, one needs to be able to return whether there is a partial match; that is, whether there would be a match if additional characters were added after the targetLimit. This can be done with a separate method having an enumerated return value: match, no_match, or partial_match.

  if (m.incrementalmatches(text, cs, ts, tl, cl) == Matcher.MATCH) {

Thus performing an incremental match of /bcbce(→d)/ against "1abcbcd2" would return a partial_match because the addition of an e to the end of the target would allow it to match. Note that /(bc)*(→d)/ would also return a partial match, because if bc were added at the end of the target, it would match.

Here is the above table, when an incremental match method is called:

Pattern Input Output Comment
/(←a) (bc)* (→d)/ 1abcbcd2 partial match 'bc' could be inserted.
/(←a) (bc)* (→bcd)/ 1abcbcd2 partial match 'bc' could be inserted.
/(bc)*d/ 1abcbcd2 partial match 'd' could be inserted.
/(←a) (bc)* (→d)/ 1abcbcd2 no match as with the matches function; the backwards search for 'a' fails.

The typical usage of incremental matching is to make a series of incremental match calls, marching through a buffer with each successful match. At the end, if there is a partial match, one loads another buffer (or waits for other input). When the process terminates (no more buffers or input are available), then a regular match call is made.

Internally, incremental matching can be implemented in the regular expression engine by detecting whether the matching process ever fails when the current position is at or after targetLimit, and setting a flag if so. If the overall match fails, and this flag is set, then the return value is set to partial_match. Otherwise, either match or no_match is returned, as appropriate.

The return value partial_match indicates that there was a partial match: if further characters were added there could be a match to the resulting string. It may be useful to divide this return value into two, instead:

3.8 Unicode Set Sharing

For script transliteration and similar applications, there may be a hundreds of regular expressions, sharing a number of Unicode sets in common. These Unicode sets, such as [\p{Alphabetic} -- \p{Latin}], could take a fair amount of memory, because they would typically be expanded into an internal memory representation that allows for fast lookup. If these sets are separately stored, this means an excessive memory burden.

To reduce the storage requirements, an API may allow regular expressions to share storage of these and other constructs, by having a 'pool' of data associated with a set of compiled regular expressions.

rules.registerSet("$lglow", "[\p{lowercase}&&[\p{latin}\p{greek}]] ");
rules.registerSet("$mark", "[\p{Mark}]");
rules.add("θ", "th");
rules.add("Θ(→$mark*$lglow)", "Th");
rules.add("Θ", "TH");
rules.add("φ", "ph");
rules.add("Φ(→$mark*$lglow)", "Ph");
rules.add("Φ", "Ph");

3.9 Possible Match Sets

RL3.9 Possible Match Sets
To meet this requirement, an implementation shall provide for the generation of possible match sets from any regular expression pattern.

There are a number of circumstances where additional functions on regular expression patterns can be useful for performance or analysis of those patterns. These are functions that return information about the sets of characters that a regular expression can match.

When applying a list of regular expressions (with replacements) against a given piece of text, one can do that either serially or in parallel. With a serial application, each regular expression is applied the text, repeatedly from start to end. With parallel application, each position in the text is checked against the entire list, with the first match winning. After the replacement, the next position in the text is checked, and so on.

For such a parallel process to be efficient, one needs to be able to winnow out the regular expressions that simply could not match text starting with a given code point. For that, it is very useful to have a function on a regular expression pattern that returns a set of all the code points that the pattern would partially or fully match.

   myFirstMatchingSet = pattern.getFirstMatchSet(Regex.POSSIBLE_FIRST_CODEPOINT);

For example, the pattern /[[\u0000-\u00FF] && [:Latin:]] * [0-9]/ would return the set {0..9, A..Z, a..z}. Logically, this is the set of all code points that would be at least partial matches (if considered in isolation).

Note: An additional useful function would be one that returned the set of all code points that could be matched at any point. Thus a code point outside of this set cannot be in any part of a matching range.

The second useful case is the set of all code points that could be matched in any particular group, that is, that could be set in the standard $0, $1, $2, ... variables.

   myAllMatchingSet = pattern.getAllMatchSet(Regex.POSSIBLE_IN$0);

Internally, this can be implemented by analysing the regular expression (or parts of it) recursively to determine which characters match. For example, the first match set of an alternation (a | b) is the union of the first match sets of the terms a and b.

The set that is returned is only guaranteed to include all possible first characters; if an expression gets too complicated it could be a proper superset of all the possible characters.

3.10 Folded Matching

RL3.10 Folded Matching
To meet this requirement, an implementation shall provide for registration of folding functions for providing insensitive matching for linguistic features other than case.

Regular expressions typically provide for case-sensitive or case-insensitive matching. This accounts for the fact that in English and many other languages, users quite often want to disregard the differences between characters that are solely due to case. It would be quite awkward to do this manually: for example, to do a caseless match against the last name in /Mark\sDavis/, one would have to use the pattern /Mark\s[Dd][Aa][Vv][Ii][Ss]/, instead of some syntax that can indicate that the target text is to be matched after folding case, such as /Mark\s\CDavis\E/.

For many languages and writing systems, there are other differences besides case where users want to allow a loose match. Once such way to do this was discussed earlier, in the discussion of matching according to collation strength. There are others: for example, for Ethiopic one may need to match characters independent of their inherent vowel, or match certain types of vowels. It is difficult to tell exactly which ways users might want to match text for different languages, so the most flexible way to provide such support is by giving a general mechanism for overriding the way that regular expressions match literals.

One way to do this is to use folding functions. These are functions that map strings to strings, and are idempotent (applying a function more than once produces the same result: f(f(x)) = f(x). There are two parts to this: (a) allow folding functions to be registered, and (b) extend patterns so that registered folding functions can be activated. During the span of text in which a folding function is activated, both the pattern literals and the input text will be processed according to the folding before comparing. For example:

// Folds katakana and hiragana together
class KanaFolder implements RegExFolder {
// from RegExFolder, must be overridden in subclasses
String fold(String source) {...}

// from RegExFolder, may be overridden for efficiency
RegExFolder clone(String parameter, Locale locale) {...}
  int fold(int source) {...}
  UnicodeSet fold(UnicodeSet source) {...}

  RegExFolder.registerFolding("k_h", new KanaFolder());


  p = Pattern.compile("(\F{k_h=argument}マルク (\s)* ダ (ヸ | ビ) ス \E : \s+)*");

In the above example, the Kana folding is in force until terminated with \E. Within the scope of the folding, all text in the target would be folded before matching (the literal text in the pattern would also be folded). This only affects literals; regular expression syntax such as '(' or '*' are unaffected.

While it is sufficient to provide a folding function for strings, for efficiency one can also provide functions for folding single code points and Unicode sets (such as [a-z...]). For more information, see [UTR30].

3.11 Submatchers

RL3.11 Submatchers
To meet this requirement, an implementation shall provide for general registration of matching functions for providing matching for general linguistic features.

There are over 70 properties in the Unicode character database, yet there are many other sequences of characters that users may want to match, many of them specific to given languages. For example, characters that are used as vowels may vary by language. This goes beyond single-character properties, because certain sequences of characters may need to be matched; such sequences may not be easy themselves to express using regular expressions. Extending the regular expression syntax to provide for registration of arbitrary properties of characters allows these requirements to be handled.

The following provides an example of this. The actual function is just for illustration.

class MultipleMatcher implements RegExSubmatcher {
// from RegExFolder, must be overridden in subclasses
   * Returns -1 if there is no match; otherwise returns the endpoint;
   * an offset indicating how far the match got.
   * The endpoint is always between targetStart and targetLimit, inclusive.
   * Note that there may be zero-width matches.
int match(String text, int contextStart, int targetStart, int targetLimit, int contextLimit) {
// code for matching numbers according to numeric value.

// from RegExFolder, may be overridden for efficiency
   * The parameter is a number. The match will match any numeric value that is a multiple.
   * Example: for "2.3", it will match "0002.3000", "4.6", "11.5", and any non-Western
   * script variants, like Indic numbers.
RegExSubmatcher clone(String parameter, Locale locale) {...}

  RegExSubmatcher.registerMatcher("multiple", new MultipleMatcher());


  p = Pattern.compile("xxx\M{multiple=2.3}xxx");

In this example, the match function can be written to parse numbers according to the conventions of different locales, based on OS functions available for such parsing. If there are mechanisms for setting a locale for a portion of a regular expression, then that locale would be used; otherwise the default locale would be used.

Note: It might be advantageous to make the Submatcher API identical to the Matcher API; that is, only have one base class "Matcher", and have user extensions derive from the base class. The base class itself can allow for nested matchers.

Annex A. Character Blocks

The Block property from the Unicode Character Database can be a useful property for quickly describing a set of Unicode characters. It assigns a name to segments of the Unicode codepoint space; for example, [\u0370-\u03FF] is the Greek block.

However, block names need to be used with discretion; they are very easy to misuse because they only supply a very coarse view of the Unicode character allocation. For example:

The following table illustrates the mismatch between writing systems and blocks. These are only examples; this table is not a complete analysis. It also does not include common punctuation used with all of these writing systems.

Writing Systems versus Blocks
Writing Systems Blocks
Latin  Basic Latin, Latin-1 Supplement, Latin Extended-A, Latin Extended-B, Latin Extended Additional, Diacritics
Greek  Greek, Greek Extended, Diacritics
Arabic  Arabic, Arabic Presentation Forms-A, Arabic Presentation Forms-B
Korean Hangul Jamo, Hangul Compatibility Jamo, Hangul Syllables, CJK Unified Ideographs, CJK Unified Ideographs Extension A, CJK Compatibility Ideographs, CJK Compatibility Forms, Enclosed CJK Letters and Months, Small Form Variants
Yi  Yi Syllables, Yi Radicals
Chinese CJK Unified Ideographs, CJK Unified Ideographs Extension A, CJK Compatibility Ideographs, CJK Compatibility Forms, Enclosed CJK Letters and Months, Small Form Variants, Bopomofo, Bopomofo Extended

For the above reasons, Script values are generally preferred to Block values. Even there, they should be used in accordance with the guidelines in UAX #24: Unicode Script Property [UAX24].

Annex B: Sample Collation Grapheme Cluster Code

The following provides sample code for doing Level 3 collation grapheme cluster detection. This code is meant to be illustrative, and has not been optimized. Although written in Java, it could be easily expressed in any programming language that allows access to the Unicode Collation Algorithm mappings.

 * Return the end of a collation grapheme cluster.
 * @param s         the source string
 * @param start     the position in the string to search
 *                  forward from
 * @param collator  the collator used to produce collation elements.
 * This can either be a custom-built one, or produced from
 * the factory method Collator.getInstance(someLocale).
 * @return          the end position of the collation grapheme cluster

static int getLocaleCharacterEnd(String s,
  int start, RuleBasedCollator collator) {
    int lastPosition = start;
    CollationElementIterator it 
      = collator.getCollationElementIterator(
          s.substring(start, s.length()));
    it.next(); // discard first collation element
int primary;
// accumulate characters until we get to a non-zero primary
do {
        lastPosition = it.getOffset();
        int ce = it.next();
        if (ce == CollationElementIterator.NULLORDER) break;
        primary = CollationElementIterator.primaryOrder(ce);
    } while (primary == 0);
    return lastPosition;

Annex C: Compatibility Properties

The following are recommended assignments for compatibility property names, for use in Regular Expressions. There are two alternatives: the Standard Recommendation and the POSIX Compatible versions. Applications should use the former wherever possible. The latter is modified to meet the formal requirements of [POSIX], and also to maintain (as much as possible) compatibility with the POSIX usage in practice. That involves some compromises, because POSIX does not have as fine-grained a set of character properties as in the Unicode Standard, and also has some additional constraints. So, for example, POSIX does not allow more than 20 characters to be categorized as digits, whereas there are many more than 20 digit characters in Unicode.

Property Standard Recommendation POSIX Compatible
(where different)
alpha Alphabetic includes more than gc = Letter. Note that marks (Me, Mn, Mc) are required for words of many languages. While they could be applied to non-alphabetics, their principal use is on alphabetics. See DerivedCoreProperties in [UAX44] for Alphabetic, also DerivedGeneralCategory in [UAX44].
Alphabetic should not be used as an approximation for word boundaries: see word below.
lower Lowercase includes more than gc = Lowercase_Letter (Ll). See DerivedCoreProperties in [UAX44].
upper Uppercase includes more than gc = Uppercase_Letter (Lu).
punct POSIX adds symbols. Not recommended generally, due to the confusion of having punct include non-punctuation marks.
Non-decimal numbers (like Roman numerals) are normally excluded. In U4.0+, the recommended column is the same as gc = Decimal_Number (Nd). See DerivedNumericType in [UAX44].
Hex_Digit contains 0-9 A-F, fullwidth and halfwidth, upper and lowercase.
alnum Simple combination of other properties
See PropList in [UAX44] for the definition of Whitespace.
blank "horizontal" whitespace: space separators plus U+0009 tab. Engines implementing older versions of the Unicode Standard may need to use the longer formulation:
\p{Whitespace} --
[\N{LF} \N{VT} \N{FF} \N{CR} \N{NEL} \p{gc=Line_Separator} \p{gc=Paragraph_Separator}]
cntrl The characters in \p{gc=Format} share some, but not all aspects of control characters. Many format characters are required in the representation of plain text.
graph Warning: the set to the left is defined by excluding space, controls, and so on with ^.
print Includes graph and space-like characters.
This is only an approximation to Word Boundaries (see b below). The Connector Punctuation is added in for programming language identifiers, thus adding "_" and similar characters.
\X See [UAX29], also GraphemeClusterBreakTest. Other functions are used for programming language identifier boundaries.
\b If there is a requirement that \b align with \w, then it would use the approximation above instead. See [UAX29], also WordBreakTest.

Note that different functions are used for programming language identifier boundaries. See also [UAX31].


[Case] Section 3.13, Default Case Algorithms [Unicode]
[CaseData] http://www.unicode.org/Public/UNIDATA/CaseFolding.txt
[FAQ] Unicode Frequently Asked Questions
For answers to common questions on technical issues.
[Feedback] Reporting Form
For reporting errors and requesting information online.
[Friedl] Jeffrey Friedl, "Mastering Regular Expressions", 2nd Edition 2002, O'Reilly and Associates, ISBN 0-596-00289-0
[Glossary] Unicode Glossary

For explanations of terminology used in this and other documents.
[Online] http://www.unicode.org/onlinedat/online.html
[Perl] http://perldoc.perl.org/
See especially:
[POSIX] The Open Group Base Specifications Issue 6, IEEE Std 1003.1, 2004 Edition, "Locale" chapter
[Prop] http://www.unicode.org/Public/UNIDATA/PropertyAliases.txt
[PropValue] http://www.unicode.org/Public/UNIDATA/PropertyValueAliases.txt
[Reports] Unicode Technical Reports
For information on the status and development process for technical reports, and for a list of technical reports.
[ScriptData] http://www.unicode.org/Public/UNIDATA/Scripts.txt
[SpecialCasing] http://www.unicode.org/Public/UNIDATA/SpecialCasing.txt
[UAX14] UAX #14: Unicode Line Breaking Algorithm
[UAX15] UAX #15: Unicode Normalization Forms
[UAX24] UAX #24: Unicode Script Property
[UAX29] UAX #29: Unicode Text Segmentation
[UAX31] UAX #31: Unicode Identifier and Pattern Syntax
[UAX44] UAX #44: Unicode Character Database
[UCD] Unicode Character Database
For an overview of the Unicode Character Database and a list of its associated files, see:
[UCDDoc Unicode Character Database Documentation
[UData] http://www.unicode.org/Public/UNIDATA/UnicodeData.txt
[Unicode] The Unicode Standard
For the latest version, see:
For the 6.0.0 version, see:
[UTR30] UTR # 30: Character Foldings
[UTS10] UTS #10: Unicode Collation Algorithm (UCA)
[UTS35] UTS #35: Unicode Locale Data Markup Language (LDML)
[Versions] Versions of the Unicode Standard
For information on version numbering, and citing and referencing the Unicode Standard, the Unicode Character Database, and Unicode Technical Reports.



Mark Davis created the initial version of this annex and maintains the text, with significant contributions from Andy Heninger.

Thanks to Asmus Freytag, Tom Christiansen, Michael D’Errico Jeffrey Friedl, Peter Linsley, Alan Liu, Kent Karlsson, Jarkko Hietaniemi, Gurusamy Sarathy, Xueming Shen, Henry Spencer, Kento Tamura, Philippe Verdy, Tom Watson, and Karl Williamson for their feedback on the document.


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

Revision 14

Revision 13

Revision 12 being a Proposed Update, only changes between Revision 11 and 13 are listed here.

Revision 11

Revision 10

Revision 9

Revision 8

Revision 7

Revision 6