Indic Scripts

TODO: this page is under construction (See: To-Do List)

Fonts

Fonts are complicated. Moreover, those for indic languages have to provide for rules for the formation of several complicated conjuncts. Each of these conjuncts can consist of several forms occurring simultaneously, which have to be positioned correctly relative to the base glyph. OTF documentation notwithstanding, font designers have their own interpretations of the specifications leading to a variety of implementations of the font features. Most of the available fonts are tested against Harfbuzz and/or ICU (sometimes only the former). Since ConTeXt uses its own OTF loading system, many indic fonts do not just work right away.

In early 2022, Hans made some nice improvements to the indic font system in ConTeXt. This was accompanied by some testing with various available fonts for some indic languages. As a result, there is an improved support for Indic fonts in ConTeXt. Various typescripts were then bundled into the ConTeXt distribution for easy use in documents.

The following table lists fonts that have been tested for use with ConTeXt. The list is by no means extensive: most of the fonts are relatively new and some (ubiquitous) old fonts are absent. If you find a font missing in this list and that works well in ConTeXt, please add it to the list. To use indic fonts place

\usetypescriptfile[indic]

in the document head and use \definetypeface and/or \setupbodyfont with the following typescripts:

Sans	Serif	Notes	Typescript(s)
Devanagari
	Adishila	4 different designs; many weights and styles; good conjunct coverage; IAST support	adishila adishila-semibold adishila-heavy adishila-dev adishila-dev-guru adishila-san adishila-san-letterpress
	Shobhika	two weights; good conjunct coverage; IAST support; some maths support	shobhika
Baloo		five weights	baloo baloo-semibold baloo-extrabold
	Amita	calligraphic style	amita
	Jaini, Jaini Purva	fonts with calligraphic style commonly found in Jaina kalpasūtra manuscripts	jaini jaini-purva
Malayalam
	RIT Rachana	an elegant font good conjunct coverage; two weights; italic style; one of the most complete fonts available	rit-rachana
	Panmana	good conjunct coverage; single weight; body-text font	panmana
Ezhuthu		handwriting font; single weight	ezhuthu
	RIT Sundar	Single weight	rit-sundar
TN Joy		three weights	tn-joy
Manjari		curvy handwriting font; suitable for body and titles; three weights	manjari
Gayathri		three weights	gayathri
Anjali Old Lipi		legible font intended for body text; glyphs for common Malayalam ligatures & Latin charset	anjali-old-lipi
Chilanka		handwriting style font with most of the unique Malayalam conjuncts; uniform width glyph strikes with round ends	chilanka
Dyuthi		an ornamental typeface; Latin and Malayalam; Malayalam glyphs are based on popular 'bulged ended' type designs; single size – thicker than usual Malayalam fonts; suited for titles	dyuthi
Karumbi		handwriting traditional script font; casual style	karumbi
Baloo Chettan 2		five weights	baloo-chettan baloo-chettan-semibold baloo-chettan-extrabold
Telugu
Baloo Tammudu 2		five weights	baloo-tammudu baloo-tammudu-semibold baloo-tammudu-extrabold
	Murty Telugu	good conjunct coverage; single-weight; consult license for terms of use	murty-telugu
Kannada
	Murty Kannada	good conjunct coverage; single-weight; consult license for terms of use	murty-kannada
Tamil
Baloo Thambi 2		five weights	baloo-thambi
Bengali
Baloo Da 2		five weights	baloo-da baloo-da-semibold baloo-da-extrabold
	Murty Bangla	good conjunct coverage; single-weight; consult license for terms of use	murty-bangla
Gujarati
Baloo Bhai		five weights	baloo-bhai baloo-bhai-semibold baloo-bhai-extrabold
	Noto Serif Gujarati	font from Google	noto-serif-gujarati

Supported Scripts and Font Features

One can of course use fonts not listed above. This section provides some details to this end.

Font feaures

The OTF specification has two shaping implementations for the indic scripts: the 'older' v1 and the 'newer' v2. See here for further details.

Script tags from the OpenType specification contains second versions for some Indic scripts.

Why are those second versions available? From their own explanation:

The OpenType script tags can also correlate with a particular OpenType Layout implementation, with the result that more than one script tag may be registered for a given Unicode script (e.g. 'deva' and 'dev2').

Features ending in -one use the older OpenType implementation, while the ones ending in -two deploy the newer implementation.

These are specified in ConTeXt by the following font features:

Script	OTF v1 script tag	OTF v2 script tag
Devanagari	devanagari-one	devanagari-two
Malayalam	malayalam-one	malayalam-two
Telugu	telugu-one	telugu-two
Kannada	kannada-one	kannada-two
Bengali	bengali-one	bengali-two

Please note that these font features also activate other font features as mandated in the OTF specification. These features can then be used to write typescripts for a font or to use the font directly in documents \definefontfamily and/or \definedfont.

One of the common problems one might encounter with indic fonts is that of incorrect rendering of conjuncts involving the rakaar. In case any problems are encountered, one can try setting the indic feature (in addition to relevant -one or -two features above) appropriately as follows:

\definefontfeature
    […]
    […]
    [indic={matra=auto,conjuncts=quit}]

Sanitizer

Sometimes, some fonts might still have issues with certain conjuncts. To overcome this a sanitizer option may be used in defining font features. A goodies file accompanies this option. An example is illustrated below:

return {
    name = "myfont",
    version = "1.00",
    comment = "Goodies that complement myfont.",
    sanitizers = {
        dev2rkrf  = { 
            mapping = {
                ["के्र"] = "क्रे",
                ["कै्र"] = "क्रै",
                ["खे्र"] = "ख्रे",
                ["खै्र"] = "ख्रै",
                ["गे्र"] = "ग्रे",
                ["गै्र"] = "ग्रै",
                ["घे्र"] = "घ्रे",
                ["घै्र"] = "घ्रै",
                ["चे्र"] = "च्रे",
                ["चै्र"] = "च्रै",
                ["छे्र"] = "छ्रे",
                ["छै्र"] = "छ्रै",
                ["जे्र"] = "ज्रे",
                ["जै्र"] = "ज्रै",
                ["झे्र"] = "झ्रे",
                ["झै्र"] = "झ्रै",
                ["ञे्र"] = "ञ्रे",
                ["ञै्र"] = "ञ्रै",
                ["णे्र"] = "ण्रे",
                ["णै्र"] = "ण्रै",
                ["ते्र"] = "त्रे",
                ["तै्र"] = "त्रै",
                ["थे्र"] = "थ्रे",
                ["थै्र"] = "थ्रै",
                ["दे्र"] = "द्रे",
                ["दै्र"] = "द्रै",
                ["धे्र"] = "ध्रे",
                ["धै्र"] = "ध्रै",
                ["ने्र"] = "न्रे",
                ["नै्र"] = "न्रै",
                ["पे्र"] = "प्रे",
                ["पै्र"] = "प्रै",
                ["फे्र"] = "फ्रे",
                ["फै्र"] = "फ्रै",
                ["बे्र"] = "ब्रे",
                ["बै्र"] = "ब्रै",
                ["भे्र"] = "भ्रे",
                ["भै्र"] = "भ्रै",
                ["मे्र"] = "म्रे",
                ["मै्र"] = "म्रै",
                ["ये्र"] = "य्रे",
                ["यै्र"] = "य्रै",
                ["वे्र"] = "व्रे",
                ["वै्र"] = "व्रै",
                ["से्र"] = "स्रे",
                ["सै्र"] = "स्रै",
                ["शे्र"] = "श्रे",
                ["शै्र"] = "श्रै",
                ["षे्र"] = "ष्रे",
                ["षै्र"] = "ष्रै",
                ["हे्र"] = "ह्रे",
                ["है्र"] = "ह्रै",
            }
        }
    }
}

Suppose that for a certain font (say myfont), using devanagari-two features, the above listed ra + consonant + vowel forms are not rendered properly. The above goodies file is then saved as myfont.lfg and used while defining the features thus:

\definefontfeature
    [myfontfeatures]
    [devanagari-two]
    [goodies=myfont.lfg,
     sanitizer=dev2rkrf,
     indic={movematra=auto,conjuncts=quit}]

Now, myfontfeatures can be used with \definedfont and/or while writing typescripts for the font.

Script and language features

The hyphenation patterns for the following languages are included in ConTeXt:

Script/Language	Conversion set	Hyphenation
Devanagari	devanagarinumerals
Malayalam	malayalamnumerals	\language[ml]
Kannada	kannadanumerals	\language[kn]
Telugu	telugunumerals	\language[te]
Bengali	bengalinumerals	\language[bn]
Tamil	tamilnumerals	\language[ta]
Gujarati	gujaratinumerals	\language[gu]
Gurmukhi	gurmukhinumerals
Hindi	devanagarinumerals	\language[hi]
Sanskrit	devanagarinumerals	\language[sa]

A pattern is activated with \language. The Sanskrit hyphenation patterns sa support hyphenation of Sanskrit written using the Malayalam, Telugu, Kannada, Bengali and Latin with IAST. Conversion sets are used as values of the keys numberconversion, conversion (wherever applicable) and with \convertnumber.

Sanskrit Transliteration

Transliteration of Sanskrit from IAST to Devanagari and vice-versa as well as from and to other Indic languages is available in ConTeXt. The following transliteration schemes are supported with more planned:

Transliteration Scheme	Vector
Devanagari to IAST	deva to iast
IAST to Devanagari	iast to deva
ITrans to Devanagari	itrans to deva
Devanagari to Malayalam	deva to mlym
Devanagari to Kannada	deva to knda
Devanagari to Telugu	deva to tlgu
Devanagari to Gujarati	deva to gujr
Devanagari to Bengali	deva to bngl

The main macro to set up a transliteration is \definetransliteration.

A simple example

A transliteration instance is defined using \definetransliteration:

\definetransliteration
    [TrDevaToIAST]
    [vector={deva to iast}]

This defines macros for both inline and block transliteration:

\starttext

% display mode
\starttransliteration[TrDevaToIAST]
महाजनस्य संसर्गः कस्य नोन्नतिकारकः।
पद्मपत्रस्थितं तोयं धत्ते मुक्ताफलश्रियम्॥
\stoptransliteration

% or inline
The company of great people (\transliteration[TrDevaToIAST] महाजनस्य संसर्गः} is there one for whom it is not beneficial {\transliteration[TrDevaToIAST] कस्य नोन्नतिकारकः}? (No!) Situated on the leaf of the lotus, (the mere) droplet of water {\transliteration[TrDevaToIAST] पद्मपत्रस्थितं तोयं} shines forth like a pearl {\transliteration[TrDevaToIAST] धत्ते मुक्ताफलश्रियम्}.

\stoptext

Or, more conveniently:

\starttext

% display mode
\startTrDevaToIAST
दानं प्रियवाक्सहितं 
ज्ञानमगर्वं क्षमान्वितं शौर्यम्। 
रूपं शीलसुयुक्तं
दुर्लभमेतच्चतुर्भद्रम्॥
\stopTrDevaToIAST

% or inline
Charity accompanied by sweet words {\TrDevaToIAST दानं प्रियवाक्सहितं}, knowledge devoid of arrogance, valour accompanied by forgiveness (pity) {\TrDevaToIAST ज्ञानमगर्वं क्षमान्वितं शौर्यम्}, beauty accompanied by virtue (grace) {\TrDevaToIAST रूपं शीलसुयुक्तं} – these four are scarce {\TrDevaToIAST दुर्लभमेतच्चतुर्भद्रम्}.

\stoptext

Please note that there is also \resettransliteration which can be used in stream to (temporarily) prevent any transliteration.

Exceptions

Sometimes, one might want to leave retain words or phrases in the original script and avoid transliteration. To this end, \transliterationexception may be used to define such exceptions

• at the level of a transliteration scheme, i.e., per vector as in:

\transliterationexception[deva to iast]{शरीरं}{देहं}

• and/or at the instance level, i.e., per transliteration block define with \definetransliteration as in:

\transliterationexception[TrDevaToIAST]{शरीरं}{देहं}

When both are defined, the latter overrides the former. Moreover, any derived transliteration instances also inherit the exceptions defined for the parent. So, if any exceptions are to be avoided/changed, they must be redefined for the derived instances.

Source with transliterated version

Very often, one wants to typeset paragraphs in the original script accompanied by a transliterated version. This can be achieved using the before key of \definetransliteration. A simple example is as follows:

\usetransliteration[indic]
\setuplines[indenting={yes,small,even}]
\definebuffer
    [padya]
\definetransliteration
    [padyaPair]
    [color=blue,
     vector={deva to iast},
     before={\startlines\getbuffer[padya]\par},
     after=\stoplines]

\starttext
\startbuffer[padya]
कोऽतिभारः समर्थानां किं दूरं व्यवसायिनाम्।
को विदेशस्तु विदुषां कः परः प्रियवादिनाम्॥
\stopbuffer

\startpadyaPair
\getbuffer[padya]
\stoppadyaPair

What is a burden for the abled, and what is faraway (beyond reach) to the
perservering? What is a foreign land to the learned and who are strangers to
the sweet-spoken? (None!)

\stoptext

In summary: one defines a buffer which contains the paragraph (a verse in this example) in the original script. Then, using the before and after keys of \definetransliteration, a pair of verses may be easily typeset into lines.

IAST to Devanagari

A sample with IAST transliterated Sanskrit (or so I hope) to Devanagari script may read:

\definefontfamily [nagari] [rm] [Adishila]
  [features=devanagari-one]

\setupbodyfont [nagari]

\switchtobodyfont[shobhika]

\usetransliteration[indic]

\definetransliteration
  [MyDeva]
  [lang=sa,
   vector={iast to deva}]

\starttext

Is this Sanskrit?

\starttransliteration[MyDeva]
idaṁ adbhutam kauśika tisraḥ garuḍavāhanan
\stoptransliteration

I really don’t know:
\transliteration[MyDeva]{idaṁ adbhutam kauśika tisraḥ garuḍavāhanan}.

\stoptext

Output not shown, because it requires fonts not installed on the wiki.