The following lexical conventions are used in the layout definition:

NAME

A string of characters that consists of printable ASCII characters. It includes DECIMAL and HEXADECIMAL also. Examples: test, a1_src, b32, 123.

HEXADECIMAL_BYTE

Two-digit hexadecimal number. The number starts with a hexadecimal digit followed by another hexadecimal digit. Examples: e0, E1, a7, fe.

HEXADECIMAL

A hexadecimal number. The hexadecimal representation consists of an escape character, '0' followed by the constant 'x' or 'X' and one or more hexadecimal digits. Examples: 0x0, 0x1, 0x1a, 0xA, 0x1b3.

DECIMAL

A decimal number, represented by one or more decimal digits. Examples: 0, 123, 2165.

Each comment must start with '#'. The comment ends at the end of the line.

The following keywords are reserved:

active_directional, active_shape_editing, AL, ALGORITHM_BASIC, ALGORITHM_IMPLICIT, AN, BN, check_mode, context, CONTEXT_LTR, CONTEXT_RTL, CS, EN, END, ES, ET, FALSE, FILE_CODE_REPRESENTATION, implicit_algorithm, keep, L, LAYOUT_ATTRIBUTES, LAYOUT_BIDI_CHAR_TYPE_DATA, LAYOUT_SHAPE_DATA, LRE, LRO, MODE_EDIT, MODE_STREAM, NSM, national_numerals, numerals, NUMERALS_CONTEXTUAL, NUMERALS_NATIONAL, NUMERALS_NOMINAL, ON, orientation, ORIENTATION_CONTEXTUAL, ORIENTATION_LTR, ORIENTATION_RTL, ORIENTATION_TTBLR, ORIENTATION_TTBRL, PDF, PROCESS_CODE_REPRESENTATION, PS, R, repeat*, repeat+, RLE, RLO, S, shape_charset, shape_charset_size, shape_context_size, swapping, SWAPPING_NO, swapping_pairs, SWAPPING_YES, TEXT_EXPLICIT, TEXT_IMPLICIT, TEXT_NOMINAL, TEXT_SHAPED, text_shaping, TEXT_VISUAL, TRUE, type_of_text, WS

Additionally, the following symbols are also reserved as tokens:

( ) [ ] , : ; ... = -> +

The layout attribute definition section defines the layout attributes and their associated values.

The definition starts with a keyword, LAYOUT_ATTRIBUTES, and ends with END LAYOUT_ATTRIBUTES:

LAYOUT_ATTRIBUTES

    # Layout attributes here.
    :
    :

END LAYOUT_ATTRIBUTES

There are a total of eight layout attribute value trios that can be defined in this section:

• orientation
• context
• type_of_text
• implicit_algorithm
• swapping
• numerals
• text_shaping
• shape_context_size

Additionally, there are five layout attribute value pairs that also can be defined in this section:

• active_directional
• active_shape_editing
• shape_charset
• shape_charset_size
• check_mode

Each attribute value trio will have an attribute name, an attribute value for the input buffer, and an attribute value for the output buffer, as in the following example:

# Orientation layout attribute value trio. The input and output # attribute values are separated by a colon and the left one # is the input attribute value: orientation ORIENTATION_LTR:ORIENTATION_LTR

Each attribute value pair will have an attribute name and an associated attribute value, as in the following example:

# Shape charset attribute value pair: shape_charset ISO8859-6

The orientation value trio defines the global directional text orientation. The possible values are:

ORIENTATION_LTR

Left-to-right horizontal rows that progress from top to bottom.

ORIENTATION_RTL

Right-to-left horizontal rows that progress from top to bottom.

ORIENTATION_TTBRL

Top-to-bottom vertical columns that progress from right to left.

ORIENTATION_TTBLR

Top-to-bottom vertical columns that progress from left to right.

ORIENTATION_CONTEXTUAL

The global orientation is set according to the direction of the first significant (strong) character. If there are no strong characters in the text and the attribute is set to this value, the global orientation of the text is set according to the value of the attribute context. This value is meaningful only for bidirectional text.

If no value or value trio is defined, the default is ORIENTATION_LTR.

The context value trio is meaningful only if the attribute orientation is set to ORIENTATION_CONTEXTUAL. It defines what orientation is assumed when no strong character appears in the text. The possible values are:

CONTEXT_LTR

In the absence of characters with strong directionality in the text, orientation is assumed to be left-to-right rows progressing from top to bottom.

CONTEXT_RTL

In the absence of characters with strong directionality in the text, orientation is assumed to be right-to-left rows progressing from top to bottom.

If no value or value trio is specified, the default is CONTEXT_LTR.

The type_of_text value trio specifies the ordering of the directional text. The possible values are:

TEXT_VISUAL

Code elements are provided in visually ordered segments, which can be rendered without any segment inversion.

TEXT_IMPLICIT

Code elements are provided in logically ordered segments. Logically ordered means that the order in which the characters are provided is the same as the order in which the characters are pronounced when reading the presented text or the order in which characters would be entered from a keyboard.

TEXT_EXPLICIT

Code elements are provided in logically ordered segments with a set of embedded controls. Some examples of such embedded controls from ISO/IEC 10646-1 are:

LEFT-TO-RIGHT EMBEDDING (LRE) RIGHT-TO-LEFT EMBEDDING (RLE) RIGHT-TO-LEFT OVERRIDE (RLO) LEFT-TO-RIGHT OVERRIDE (LRO) POP DIRECTIONAL FORMAT (PDF)

If no value or value trio is specified, the default is TEXT_IMPLICIT.

The implicit_algorithm value trio specifies the type of bidirectional implicit algorithm used in reordering and shaping of directional or context-dependent text. The possible values are:

ALGORITHM_IMPLICIT

Directional code elements will be reordered using an implementation-defined implicit algorithm.

ALGORITHM_BASIC

Directional code elements will be reordered using a basic implicit algorithm defined in the Unicode standard.

Even though we allow two different values for the implicit_algorithm, since the Solaris implementation-defined implicit algorithm is based on the Unicode standard, there is no difference in behavior whether you choose ALGORITHM_IMPLICIT or ALGORITHM_BASIC for this attribute.

The default value is ALGORITHM_IMPLICIT.

The swapping value trio specifies whether symmetric swapping is applied to the text. The possible values are:

SWAPPING_YES

The text conforms to symmetric swapping.

SWAPPING_NO

The text does not conform to symmetric swapping.

If no value or value trio is specified, the default is SWAPPING_NO.

The numerals value trio specifies the shaping of numerals. The possible values are:

NUMERALS_NOMINAL

Nominal shaping of numerals using the Arabic numbers of the portable character set (in Solaris, ASCII digits).

NUMERALS_NATIONAL

National shaping of numerals based on the script of the locale. For instance, Thai digits in the Thai locale.

NUMERALS_CONTEXTUAL

Contextual shaping of numerals depending on the context script of surrounding text, such as Hindi numbers in Arabic text and Arabic numbers otherwise.

If no value or value trio is specified, the default is NUMERALS_NOMINAL.

The text_shaping value trio specifies the shaping; that is, choosing (or composing) the correct shape of the input or output text. The possible values are:

TEXT_SHAPED

The text has presentation form shapes.

TEXT_NOMINAL

The text is in basic form.

If no value or value trio is specified, the default is TEXT_NOMINAL for input and TEXT_SHAPED for output.

The shape_context_size value trio specifies the size of the context (surrounding code elements) that must be accounted for when performing active shape editing. If not defined, the default value 0 is used for the number of surrounding code elements at both front and rear:

    # The shape_context_size for both front and rear surrounding code
    # elements are all zero:
    shape_context_size  0:0

The front and rear attribute values are separated by a colon, with the front value to the left of the colon.

The active_directional value pair specifies whether the current locale requires (bi-)directional processing. The possible values are:

TRUE

Requires (bi-)directional processing.

FALSE

Does not require (bi-)directional processing.

The active_shape_editing value pair specifies whether the current locale requires context-dependent shaping for presentation. The possible values are:

TRUE

Requires context-dependent shaping.

FALSE

Does not require context-dependent shaping.

The shape_charset value pair specifies the current locale's shape charset on which the complex text layout is based. There are two different kinds of shape charset values that can be specified:

• A single shape charset
• Multiple shape charsets

For a single shape charset, it can be defined by using NAME as defined in the Lexical Convention section above. For multiple shape charsets, however, it should follow the syntax given below in extended BNF form:

            multiple_shape_charset
                                : charset_list
                                ;

            charset_list        : charset
                                | charset_list ';' charset
                                ;

            charset             : charset_name '=' charset_id
                                ;

            charset_name        : NAME
                                ;

            charset_id          : HEXADECIMAL_BYTE
                                ;

For instance, the following is a valid multiple shape charsets value for the shape_charset attribute:

    # Multi-shape charsets:
    shape_charset       tis620.2533=e4;iso8859-8=e5;iso8859-6=e6

The shape_charset must be specified.

The shape_charset_size value pair specifies the encoding size of the current shape_charset. The valid value is a positive integer from 1 to 4. If the multiple shape charsets value is defined for the shape_charset attribute, the shape_charset_size must be 4.

The shape_charset_size must be specified.

The check_mode value pair specifies the level of checking of the elements in the input buffer for shaping and reordering purposes. The possible values are:

MODE_STREAM

The string in the input buffer is expected to have valid combinations of characters or character elements.

MODE_EDIT

The shaping of input text may vary depending on locale-specific validation or assumption.

When no value or value pair is not specified, the default value is MODE_STREAM.

This section defines the bidirectional and other character types that will be used in the Unicode Bidirectional Algorithm and the shaping algorithm part of the UMLE.

The definition starts with a keyword LAYOUT_BIDI_CHAR_TYPE_DATA and ends with END LAYOUT_BIDI_CHAR_TYPE_DATA:

    LAYOUT_BIDI_CHAR_TYPE_DATA

            # Layout bidi definitions here.
            :
            :

    END LAYOUT_BIDI_CHAR_TYPE_DATA

The bidirectional data and character type data definition should be defined for the two different kinds of text shape forms, TEXT_SHAPED and TEXT_NOMINAL, depending on the text_shaping attribute value and also for the two different kinds of text representations, file code representation and process code representation (that is, wide character representation):

    LAYOUT_BIDI_CHAR_TYPE_DATA

        FILE_CODE_REPRESENTATION
            TEXT_SHAPED

                # TEXT_SHAPED bidi and character type data
                # definition in file code representation here.
                :
                :

            END TEXT_SHAPED

            TEXT_NOMINAL

                # TEXT_NOMINAL bidi and character type data
                # definition in file code representation here.
                :
                :

             END TEXT_NOMINAL
         END FILE_CODE_REPRESENTATION
              
         PROCESS_CODE_REPRESENTATION
             TEXT_SHAPED

                 # TEXT_SHAPED bidi and character type data
                 # definition in process code representation here.
                 :
                 :

             END TEXT_SHAPED

             TEXT_NOMINAL

                 # TEXT_NOMINAL bidi and character type data
                 # definition in process code representation here.
                 :
                 :

             END TEXT_NOMINAL
         END PROCESS_CODE_REPRESENTATION

     END LAYOUT_BIDI_CHAR_TYPE_DATA

Each bidi and character type data definition can have the following definitions:

• Bidirectional data type definition
• swapping_pairs character type definition
• national_numerals character type definition

There are nineteen different bidirectional data types that can be defined, as in the following table:

If not defined in this section, the characters belong to the other neutrals type, ON.

Each keyword list above will be accompanied by one or more HEXADECIMAL ranges of characters that belong to the bidirectional character type. The syntax is as follows:

    bidi_char_type      : bidi_keyword ':' range_list
                        ;

    bidi_keyword        : 'L'
                        | 'LRE'
                        | 'LRO'
                        | 'R'
                        | 'AL'
                        | 'RLE'
                        | 'RLO'
                        | 'PDF'
                        | 'EN'
                        | 'ES'
                        | 'ET'
                        | 'AN'
                        | 'CS'
                        | 'PS'
                        | 'S'
                        | 'WS'
                        | 'ON'
                        | 'NSM'
                        | 'BN'
                        ;

    range_list          : range
                        | range_list ',' range
                        ;

    range               : HEXADECIMAL
                        | HEXADECIMAL '...' HEXADECIMAL
                        ;

For example:

# Bidi character type definitions: L: 0x26, 0x41...0x5a, 0xc380...0xc396, 0xe285a0...0xe28682 WS: 0x20, 0xc2a0, 0xe28080...0xe28086

The swapping_pairs specifies the list of swappable characters if SWAPPING_YES is specified as a value at the swapping value trio. The syntax of the swapping_pairs is as follows:

    swapping_pair_list  : swapping_keyword ':' swap_pair_list
                        ;

    swapping_keyword    : 'swapping_pairs'
                        ;

    swap_pair_list      : swap_pair
                        | swap_pair_list ',' swap_pair
                        ;

    swap_pair           : '(' HEXADECIMAL ',' HEXADECIMAL ')'

For example:

    # Swapping pair definitions:
    swapping_pairs:     (0x28, 0x29), (0x7b, 0x7d)

The national_numerals specifies the list of national digits that can be converted as the numerals value trio specifies. The syntax of the national_numerals is as follows:

    numerals_list       : numerals_keyword ':'
                         numerals_list ';' contextual_range_list
                        ;

    numerals_keyword    : 'national_numerals'
                        ;

    numerals_list       : '(' zero ',' one ',' two ',' three ','
                          four ',' five ',' six ',' seven ','
                          eight ',' nine ')'

    zero                : HEXADECIMAL
                        ;

    one                 : HEXADECIMAL
                        ;

    two                 : HEXADECIMAL
                        ;

    three               : HEXADECIMAL
                        ;

    four                : HEXADECIMAL
                        ;

    five                : HEXADECIMAL
                        ;

    six                 : HEXADECIMAL
                        ;

    seven               : HEXADECIMAL
                        ;

    eight               : HEXADECIMAL
                        ;

    nine                : HEXADECIMAL
                        ;

    contextual_range_list
                        : contextual_range
                        | contextual_range_list ',' contextual_range
                        ;

    contextual_range    : HEXADECIMAL
                        | HEXADECIMAL '...' HEXADECIMAL
                        :

For instance:

    # National numerals definition. The national number that will
    # replace Arabic number 0 to 9 is 0, 0x41, 0x42, and so on.
    # The contextual surrounding characters are 0x20 to 0x40 and
    # 0x50 to 0x7f:
    national_numerals:
        (0x0, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48, 0x49)
        ; 0x20...0x40, 0x50...0x7f

Unless NUMERALS_CONTEXTUAL is the value of the numerals attribute, the contextual range list definition is meaningless.

The shaping data definition section defines the context-dependent shaping rules that will be used in the shaping algorithm of the UMLE.

The definition starts with a keyword, LAYOUT_SHAPE_DATA, and ends with END LAYOUT_SHAPE_DATA:

    LAYOUT_SHAPE_DATA

        # Layout shaping data definitions here.
        :
        :

    END LAYOUT_SHAPE_DATA

The shaping data definition should be defined for the two different kinds of text shape forms, TEXT_SHAPED and TEXT_NOMINAL, depending on the text_shaping attribute value and also for the two different kinds of text representations, file code representation and process code representation (that is, wide character representation:

    LAYOUT_SHAPE_DATA

        FILE_CODE_REPRESENTATION
            TEXT_SHAPED

                # TEXT_SHAPED shaping data definition in file code
                # representation here.
                :
                :

            END TEXT_SHAPED

            TEXT_NOMINAL

                # TEXT_NOMINAL shaping data definition in file code
                # representation here.
                :
                :

            END TEXT_NOMINAL
        END FILE_CODE_REPRESENTATION

        PROCESS_CODE_REPRESENTATION
            TEXT_SHAPED

                # TEXT_SHAPED shaping data definition in process code
                # representation here.
                :
                :

            END TEXT_SHAPED

            TEXT_NOMINAL

                # TEXT_NOMINAL shaping data definition in process
                # code representation here.
                :
                :

            END TEXT_NOMINAL
        END PROCESS_CODE_REPRESENTATION

    END LAYOUT_SHAPE_DATA

Each shaping data definition consists of one or more of the shaping sequence definitions. Each shaping sequence definition is a representation of a series of state transitions triggered by an input character and the current state at each transition.

The syntax of the shaping sequence definition is as follows:

    shaping_sequence    : initial_state '+' input '->' next_state_list
                        ;

    initial_state       : '()'
                        ;

    input               : HEXADECIMAL
                        ;

    next_state_list     : next_state
                        | next_state_list '+' input '->' next_state
                        | '(' next_state_list '+' input ')' 'repeat+'
                        | '(' next_state_list '+' input ')' 'repeat*'
                        ;

    next_state          : '(' out_buffer ',' in2out ',' out2in ','
                                property ')'
                        ;

    out_buffer          : '[' out_char_list ']'
                        ;

    out_char_list       : HEXADECIMAL
                        | '(' HEXADECIMAL ')' 'repeat+'
                        | out_char_list ';' HEXADECIMAL
                        ;

    in2out              : '[' i2o_list ']'
                        ;

    i2o_list            : DECIMAL
                        | '(' DECIMAL ')' 'repeat+'
                        | i2o_list ';' DECIMAL
                        ;

    out2in              : '[' o2i_list ']'
                        ;

    o2i_list            : DECIMAL
                        | '(' DECIMAL ')' 'repeat+'
                        | o2i_list ';' DECIMAL
                        ;

    property            : '[' prop_list ']'
                        ;

    prop_list           : HEXADECIMAL
                        | '(' HEXADECIMAL ')' 'repeat+'
                        | prop_list ';' HEXADECIMAL
                        ;

For example, the following shaping sequences can be defined:

    # A simple shaping sequence:
    () + 0x21 ->
        ( [0x0021], [0], [0;0], [0x80] ) + 0x22 ->
        ( [0x0021;0x0022], [0;1], [0;0;1;1], [0x80;0x80] ) + 0xc2a0 ->
        ( [0x0021;0x0022;0xe030], [0;1;2], [0;0;1;1;2;2],
          [0x80;0x80;0x80] )

    # A repeating shaping sequence:
    () + 0x21 ->
        (
            ( [0x0021], [0], [0;0], [0x80] ) + 0x22 ->
            ( [0x0021;0x0022], [0;1], [0;0;1;1], [0x80;0x80] ) + 0xc2a2
        ) repeat+

The first example shows a shaping sequence such that if 0x21, 0x22, and 0xc2a0 are the input buffer contents, it will be converted into an output buffer containing 0x0021, 0x0022, and 0xe030; an input to the output buffer containing 0, 1, and 2; an output to the input buffer containing 0, 0, 1, 1, 2, and 2; and a property buffer containing 0x80, 0x80, and 0x80.

The second example shows a repeating shaping sequence where, if the first input code element is 0x21, then the second and third input code elements are 0x22 and 0xc2a2, respectively.