Developing fonts > Specifications

Developing OpenType Fonts
for Devanagari Script (3 of 3):
Features

The features listed below have been defined to create the basic forms for the languages that are supported on Devanagari systems. Regardless of the model an application chooses for supporting layout of complex scripts, the shaping engine requires a fixed order for executing features within a run of text to consistently obtain the proper basic form.

The features of the basic shaping forms are applied one at a time to the cluster or portion of the cluster. The result impacts the analysis in terms of the conjoining behavior and final reordering. The features of the presentation forms are applied next, to the entire cluster simultaneously. Mandatory features must always be applied; the discretionary presentation-forms features listed should be applied by default, but can be suppressed by a client (normally at the discretion of the user).

The order of the lookups within each feature is also very important. For more information on lookups and defining features in OpenType fonts, see the Encoding section of the OpenType Font Development document.

OpenType features used for Devanagari scripts, applied in the following order:

Feature Feature function Layout operation
Localized forms:
locl Localized form substitution GSUB
Basic shaping forms:
nukt Nukta form substitution GSUB
akhn Akhand ligature substitution GSUB
rphf Reph form substitution GSUB
pref Rakaar form substitution GSUB
blwf Below-base form substitution GSUB
half Half-form substitution GSUB
vatu Vattu variants GSUB
cjct Conjunct form substitution GSUB
Mandatory presentation forms:
pres Pre-base substitution GSUB
abvs Above-base substitution GSUB
blws Below-base substitution GSUB
psts Post-base substitution GSUB
haln Halant form substitution GSUB
Discretionary presentation forms:
calt Contextual alternates GSUB
Positioning features:
kern Kerning GPOS
dist Distances GPOS
abvm Above-base mark positioning GPOS
blwm Below-base mark positioning GPOS
       
[GSUB = glyph substitution, GPOS = glyph positioning]


Feature examples

Many of the registered features described and illustrated in this document are based on the Microsoft OpenType font Mangal (Mangal.ttf). Mangal contains layout information and glyphs to support all of the required features for the Devanagari script and language systems supported. Instructions for obtaining the Mangal font are given in the Appendix of this document.

The illustrations in the following examples show the result of that particular feature being applied. Features must be written to match glyph sequences after re-ordering has occurred. Note that the input context for a feature may be the result of a previous feature having already been applied.


Localized forms

Feature Tag: "locl"

This feature is used in association with OpenType language system tags to trigger lookups that will select alternate glyphs needed for language-specific typographic conventions. The ‘locl’ should not be used in association with the default language system, but only used with other language system tags. See the Appendix of this document for language system tags associated with the Devanagari script.


Basic shaping forms


Nukta

Feature Tag: "nukt"

The nukta alters the way a preceding consonant or vowel is pronounced. The most common nukta forms have been defined as separate characters in Unicode with their own code points. All consonants, as well as akhand forms should have an associated nukta form.

Note - Rather than using substitution, nukta forms can also be created by positioning the nukta as a below-base mark on the base glyph using the ‘blwm’ positioning feature

The input context for the nukt feature always consists of the full form of the consonant. The half form of nukta consonants will be substituted using the half feature.

Nukta feature applied substitutes Ka-nukta pre-composed glyph:


Akhand

Feature Tag: "akhn"

An akhand is a required consonant ligatures that may appear anywhere in the syllable, and may or may not involve the base glyph. Akhand ligatures have the highest priority and are formed first; some languages include them in their alphabets. There are 2 Akhand ligatures in Devanagari.

The input context for the akhand feature always consists of the full form of the consonant. The half forms of Akhand ligatures will be called later in the half feature.

Because the akhand feature is applied early in the sequence of features and is applied over the entire cluster, it can also be used to create certain forms that must take priority in particular contexts over forms that would be created during subsequent feature application.

Using the ‘akhn’ feature, Ka + halant + Ssa is substituted with the KaSsa ligature:

Ja + halant + Nya is substituted with the JaNya Ligature:


Reph

Feature Tag: "rphf"

Applying this feature substitutes the Reph glyph. If the first consonant of the cluster consists of the full form of Ra + Halant, this feature substitutes the combining-mark form of Reph. In addition, the position of the Reph glyph is adjusted with the ‘abvm’ GPOS feature.

The input context for the Reph feature always consists of the full form of Ra + Halant.

Reph feature substitutes the mark glyph form of Ra. After final reordering, positioning is adjusted in the ‘abvm’ GPOS feature:

Reph feature applied with multiple consonants. Note: reph is re-ordered to position on the 1st main consonant:


Rakaar

Feature Tag: "rkrf"

Applying this feature substitutes a consonant-rakaar (below-base Ra) ligature or an akhand-rakaar ligature. For consonants that do not form a ligature with the rakaar, the ‘rkrf’ feature may also be used to substitute a pre-composed glyph for the consonant plus rakaar.

Note: the rakaar forms feature is required specifically for those scripts in which a half-form of a consonant-rakaar ligature can occur. This includes Devanagari and Gujarati scripts. It is not used for the other Indic scripts.

The input context for the rakaar feature always consists of the full form consonant + halant + Ra. The half forms of rakaar ligatures should be substituted using the half feature.

The ‘rkrf’ feature applied to substitute the rakaar form of Ka:

Rakaar variants for ‘nukta’ glyphs as well as ‘akhand’ ligatures should also be created:

For consonants that do not form a ligature with the rakaar (like the Cha) the ‘rkrf’ feature may also be used to substitute a pre-composed glyph for the consonant plus rakaar combination:



Below-base form of consonant

Feature Tag: "blwf"

This feature substitutes the below-base forms of Consonants like the Ra in Devanagari (aka ‘rakaar’) when the rakaar does not form a ligature with the preceding consonant. If the rakaar and preceding consonant do form a ligature, it should be substituted in the previous feature ‘rkrf’.

Halant plus Ra (preceded by a consonant which does not form a ‘rkrf’ ligature) substitutes the rakaar form:


Half form of consonant

Feature Tag: "half"

Applying this feature substitutes half forms - forms of consonants used in the pre-base position. Consonants that have a half form should be listed in the ‘half’ feature. Devanagari has distinctly shaped half forms for most of the consonants as well as nukta and Akhand glyphs. If a consonant does not have a distinct shape for the half form and does not form any ligature, it will be displayed with an explicit Virama (same shape as the halant form).

Note – the result of listing a consonant in the half feature (whether it has a true half form or not) will affect the re-ordering (and positioning) of the reph and pre-pended matras. See illustration in the Introduction section of this document.

This feature is applied to all consonants preceding the ‘main’ consonant.

Example 1 - Half feature substitutes half form of Ka:

Example 2 - Half feature applied to multiple consonants:

Example 3 - Half feature used to substitute half form of Ta-Nukta:

Example 4 - Half form of Akhand ligature (KaSsa):

Example 5 - Half feature applied to Da, produces ‘halant’ form of Da since it has no distinct half form shape: (remember listing the Da in the half feature will cause it to behave like a half form for purposes of re-ordering)


Vattu variants

Feature Tag: "vatu"

The ‘vatu’ feature can be used to substitute a ligature of a full (or half) form consonant plus a below-base vattu (rakaar) mark. This feature was needed in earlier implementations to combine rakaar with both full and half forms It is not useful, however, for newer implementations that distinguish consonants that do not have half forms from other consonants, for purposes of re-ordering the reph and pre-pended matras. In Devanagari, this feature is not required if the ‘rkrf’ feature is used, however it is still supported for backward compatibility with existing fonts.

Example; the ‘vatu’ feature used in earlier implementations of the shaping engine to substitute a ligature of consonant (full or half form) plus vattu (rakaar), which now should be created using the ‘rkrf’ and ‘half’ features.


Conjunct forms

Feature Tag: "cjct"

Apply feature 'cjct' to substitute conjunct forms where the first consonant in the consonant-cluster pair does not have a half form. This feature allows for control over re-ordering of reph and pre-pended matras in case of consonants that do not take half forms yet do form conjunct ligatures in combination with certain following consonants.

The ‘cjct’ feature applied to substitute conjunct form of Da + Ga:

The ‘cjct’ feature applied to substitute conjunct form of Da + Ma:



Presentation forms

After the glyphs have been reordered, the presentation lookups are applied to provide the best typographic rendering of the text. The features of the presentation forms are applied to the entire cluster simultaneously, executing lookups within each feature in the order that they are specified in the font.

The pres, abvs, blws, psts and haln features are all mandatory for software implementations: they are required for correct script behaviour and none should ever be treated as discretionary. Because of this and because they are all applied simultaneously over entire clusters, they are not functionally different: a set of lookups could be divided between these features or grouped together under one of them with no difference in effect. These multiple features are provided, however, as an aid to the font developer for organizing lookups based on the combinations of glyphs they apply to. There are no specific requirements on how each should be used; the examples provided below illustrate typical usage, however.


Pre-base substitutions

Feature Tag: "pres"

This feature is used to substitute pre-base consonant conjuncts made with half forms, the type most common in Devanagari. The resulting conjunct can be in full or half form. See examples 1-4.

This feature is also used to select typographically correct forms of the I-Matra. For example, a font can have several versions of the I-Matra to be used in context with different consonant bases or clusters. See example 5 + 6. In addition the ‘pres’ feature can contain pre-composed ligatures of the I-matra with certain bases.

Example 1 - half Ka + full Ta is substituted by the full KaTa conjunct:

Example 2 - half Sha + full Na is substituted by the full ShaNa conjunct:

Example 3 - half Kha + half Na is substituted by the half KhaNa conjunct:

Example 4 - half Sha + half La is substituted by the half ShaLa conjunct:

Example 5 - The ‘pres’ feature is also used to substitute variations of the I-Matra based on context:

Example 6 - Using MS Volt, different shapes of the I-matra are selected based on the context. Note; in this example ‘glyph groups’ listing consonants with similar widths, have been used for substitution context.


Above-base substitutions

Feature Tag: "abvs"

This feature is used for glyph substitutions involving above-base marks. Such substitutions might be used to select contextual forms of marks, to create mark-mark ligatures, or to create mark-base ligatures. Specific context-dependent forms or below-base consonants are handled by this lookup as well.

Example 1- contextual ‘abvs’ substitution; used to select smaller candrabindu, when preceded by vowel short E.

Example 2- mark to mark ‘abvs’ substitution; matre E + reph substituted with matraE reph ligature. Note- the positioning of the ligature matra is done in the ‘abvm’ feature:

Example 3- mark to base ‘abvs’ substitution; vowel Ii + candrabindu substituted with pre-composed ligature:


Below-base substitutions

Feature Tag: "blws"

This feature is used for glyph substitutions involving below-base marks or consonants. Such substitutions can be used to create conjuncts of base glyphs with below-base consonants, below mark ligatures or below mark-base ligatures. Specific context-dependent forms are handled by this lookup as well.

Example 1- ‘blws’ substitution to create base + below base conjunct:

Example 2- ‘blws’ substitution to create base + below mark conjunct:

Example 3- ‘blws’ substitution used to create below-mark to base conjunct:


Post-base substitutions

Feature Tag: "psts"

This feature is used to substitute post-base consonants or matras. Such substitutions can be used to create conjuncts of base glyphs with post-base consonants or post-base matra ligatures. It can also be used to specify contextual alternates of post-base forms.

Example 1- contextual ‘psts’ substitution; used to select alternate form of vowel Ii, when preceded by a Ka.

Example 2 - contextual ‘psts’ substitution; Using MS Volt, different shapes of the vowel Ii are substituted based on the context. Note; in this example ‘glyph groups’ with similar widths are used for the context:


Halant form of consonants

Feature Tag: "haln"

This feature is used to substitute a pre-composed halant form of a base (or conjunct base) glyph in syllables ending with a halant. (Rather than using substitution, halant forms can also be created by positioning the halant as a below-base mark on the base glyph using the ‘blwm’ positioning feature.)

This feature is applied only on the base glyph if the syllable ends with a halant, or in the case of non-final consonants that do not take a half form and do not form a conjunct ligature with the following consonant.

Example 1 - ‘haln’ feature used to substitute halant form of base glyph:

Example 2 - ‘haln’ feature used to substitute halant form of conjunct base glyph:


Contextual Alternates

Feature Tag: "calt"

Unlike the previous presentation lookups, the ‘calt’ feature is optional and is used to substitute discretionary contextual alternates. It is important to note that an application may allow users to turn off this feature, therefore should not be used for any obligatory Devanagari typography.


Positioning features

Distance

Feature Tag: "dist"

This feature covers positioning lookups that adjust distances between glyphs, such as kerning between pre- and post-base elements and the base glyph. Note; the feature ‘dist’ can be used in the same way as the ‘kern’ feature. The advantage of using the ‘dist’ feature is that it does not rely on the application to enable kerning.


Above-base marks

Feature Tag: "abvm"

This feature positions all above-base marks on the base glyph or the post-base matra. The best method for encoding this feature in an OpenType font is to use a chaining context positioning lookup that triggers mark-to-base and mark-to-mark attachments for above-base marks.

The ‘abvm’ lookup in MS Volt using ‘Anchor Attachment’ for adjusting positions of above-marks over bases:

The ‘abvm’ lookup in MS Volt using ‘Pair Adjustment’ for adjusting positions of above-marks with post-base vowel:


Below-base marks

Feature Tag: "blwm"

This feature positions all below-base marks on the base glyph. The best method for encoding this feature in an OpenType font is to use a chaining context positioning lookup that triggers mark-to-base and mark-to-mark attachments for below-base marks.

The ‘blwm’ lookup in MS Volt using ‘Anchor Attachment’ for adjusting positions of below-marks with bases:

The ‘blwm’ lookup in MS Volt using ‘Anchor Attachment’ can also be used for adjusting positions of below-marks with other below-marks.


Examples Devanagari syllables

Complex Devanagari syllable formation is possible using the wide range of features available in OpenType. The following examples show how the shaping engine applies the OpenType features, one at a time to the input string. These combinations do not necessarily represent actual syllables or words, but are meant to illustrate the various OpenType features in a Devanagari font.

Example #1: Input text string and resulting display as OT features are applied.

Example #2: Input text string and resulting display as OT features are applied.

Example #3: Input text string and resulting display as OT features are applied.

Example #4: Input text string and resulting display as OT features are applied.

Example #5: Input text string and resulting display as OT features are applied.


Last updated: May 22, 2008

introduction | shaping engine | features | appendices

Top of page