Description of the Kannada Language

Kannada Unicode Design Guide

Abstract: This document provides general information about the Kannada language and conventions of its usage in computers. It provides information about the Input, Storage, Display and Printing of Kannada Characters. We strongly feel that this information gathered from various standards is necessary for the correct usage of the language in various applications of Kannada Language Computing. It also includes the sorting sequence for Kannada in Unicode.

Note 1: This document contains Unicode characters and can be viewed using MS Office XP on Windows XP or equivalent

Note 2: The Convention followed in Unicode (Version 3.0) Chapter 9 (South and Southeast Asian Scripts) is used in this document and might differ from the notation commonly used in the Kannada Script.

Contact Information:

Chief Investigator

Resource Centre for Indian Language Technology Solutions- Kannada

Department of Management Studies

Indian Institute of Science

Bangalore – 560 012

Phone : 91-80-346 6022 / 394 2377 (Dir)

91-80-394 2378 / 394 2567

Fax : 91-80-346 6022 / 3600683 / 3600085

Email : root@iltwebserver.mgmt.iisc.ernet.in

Table of Contents

1. History of Kannada Language 5

1.1 Description of Kannada Language 5

1.2 Brief introduction to Kannada language 6

1.2.1 Vowels 6

1.2.2 Anuswaras 6

1.2.3 Visarga 6

1.2.4 Avagraha 6

1.2.5 Consonants 6

1.2.6 Basic Language Rule in Kannada 7

2. Technical Characteristics 9

2.1 Kannada Alphabet Characteristic 9

2.1.1 Consonant Letters 9

2.1.2 Independent Vowel Letters 9

2.1.3 Dependent Vowel Signs 9

2.1.4 Virama (Halant) 11

2.1.5 Consonant Conjuncts 11

2.1.6 Visarg 12

2.1.7 Avagrah 12

2.1.8 Numerals 12

2.1.9 Punctuation Marks 12

2.1.10 Ancient Signs 12

2.2 Fonts 12

2.2.1 Font developing Tools 12

2.3 Keyboard 13

2.4 Presentation and Storage Considerations 14

2.5 Rendering Rules 14

2.5.1 Dead Consonant Rule 15

2.5.2 Consonant RA Rules 15

2.5.3 Ligature Rules 16

2.6 Sorting issues in Kannada 17

2.6.1 Sorting of Nukta characters 17

2.6.2 Sorting the data records containing anuswara and visarga 17

2.6.3 Sorting of words with dead consonants 18

2.6.4 Sorting of Conjuncts having two different display forms 19

2.6.5 Sorting of Diacritic characters 19

2.6.6 Conclusion 19

3. References 20

Appendix 1: Unicode chart and the Collation chart if deletion and relocation are not allowed 21

Appendix 2: Unicode chart and the Collation chart if deletion and relocation are allowed 24

Appendix 3: Output from FontLab displaying all glyphs in the glyph set standardised by KGP 27

1. History of Kannada Language

Kannada is a south Indian language spoken in Karnataka state of India.Kannada is originated from the Dravidian Language. Telugu, Tamil, Malayalam are the other South Indian Languages originated from Dravidian Language. Kannada and Telugu have almost the same script. Malayalam and Tamil have resemblance. Kannada as a language has undergone modifications since BCs. It can be classified into four types-

Purva Halegannada (from the beginning till 10^th Century)

Halegannada (from 10^th Century to 12^th Century)

Nadugannada (from 12^th Century to 15^th Century)

Hosagannada (from 15^th Century)

1.1 Description of Kannada Script

Kannada script is the visual form of Kannada language. It originated from southern Bramhi lipi of Ashoka period. It underwent modifications periodically in the reign of Sathavahanas, Kadambas, Gangas, Rastrakutas, and Hoysalas. Even before seventh-Century, the Telugu-Kannada script was used in the inscriptions of the Kadambas of Banavasi and the early Chalukya of Badami in the west. From the middle of the seventh century the archaic variety of the Telugu-Kannada script developed a middle variety. The modern Kannada and Telugu scripts emerged in the thirteenth Century. Kannada script is also used to write Tulu, Konkani and Kodava languages.

Kannada along with other Indian language scripts shares a large number of structural features. The writing system of Kannada script encompasses the principles governing the phonetics and a syllabic writing systems, and phonemic writing systems (alphabets). The effective unit of writing Kannada is the orthographic syllable consisting of a consonant and vowel (CV) core and optionally, one or more preceding consonants, with a canonical structure of ((C) C) CV. The orthographic syllable need not correspond exactly with a phonological syllable, especially when a consonant cluster is involved, but the writing system is built on phonological principles and tends to correspond quite closely to pronunciation. The orthographic syllable is built up of alphabetic pieces, the actual letters of Kannada script. These consist of distinct character types: Consonant letters, independent vowels and the corresponding dependent vowel signs. In a text sequence, these characters are stored in logical phonetic order.

The Kannada block of Unicode Standard (0C80 to 0CFF) is based on ISCII-1988 (Indian Standard Code for Information Interchange). The Unicode Standard (Version 3) encodes Kannada characters in the same relative positions as those coded in the ISCII-1988 standard.

1.2 Brief introduction to Kannada language

1.2.1 Vowels (Swaras) Vowels are the independently existing letters which are called Swaras. They are-

ಅ ಆ ಇ ಈ ಉ ಊ ಋ ಎ ಏ ಐ ಒ ಓ ಔ

There are two types of Swaras depending on the time used to pronounce. They are Hrasva Swara and Deerga Swara.

Hrasva Swara

A freely existing independent vowel which can be pronounced in a single matra time (matra kala) also called as a matra. They are-

ಅ ಇ ಉ ಋ ಎ ಐ ಒ ಔ

Deergha Swara A freely existing independent vowel which can be pronounced in two matras. They are-

ಆ ಈ ಊ ಏ ಓ

1.2.2 Anuswaras ಅಂ

1.2.3 Visarga ಅಃ

1.2.4 Avagraha Also called as Plutha, which is used for the third matra either in a consonant or a vowel.

1.2.5 Consonants (Vyanjanas) These are dependent on vowels to take a independent form of the Consonant. These can be divided into Vargeeya and Avargeeya.

Vargeeya Vyanjanas

ಕ್ ಖ್ ಗ್ ಘ್ ಙ್

ಚ್ ಛ್ ಜ್ ಝ್ ಞ್

ತ್ ಥ್ ದ್ ಧ್ ನ್

ಟ್ ಠ್ ಡ್ ಢ್ ಣ್

ಪ್ ಫ್ ಬ್ ಭ್ ಮ್

Avargeeya Vyanjanas

ಯ್ ರ್ ಲ್ ವ್ ಶ್ ಷ್ ಸ್ ಹ್ ಳ್

1.2.6 Basic Language Rule in Kannada

When a dependent consonant combines with an independent vowel, a Akshara is formed.

Consonant (Vyanjana) + Vowel (matra) ---> Letter (Akshara)

Example: ಕ್ + ಅ ---> ಕ

Based on this rule we can combine all the Consonants (Vyanjanas) with the existing Vowels (matra)

to form Kagunitha for Kannada alphabet.

ಕ ಕಾ ಕಿ ಕೀ ಕು ಕೂ ಕೃ ಕೆ ಕೇ ಕೈ ಕೊ ಕೋ ಕೌ ಕಂ ಕಃ

ಖ ಖಾ ಖಿ ಖೀ ಖು ಖೂ ಖೃ ಖೆ ಖೇ ಖೈ ಖೊ ಖೋ ಖೌ ಖಂ ಖಃ

ಗ ಗಾ ಗಿ ಗೀ ಗು ಗೂ ಗೃ ಗೆ ಗೇ ಗೈ ಗೊ ಗೋ ಗೌ ಗಂ ಗಃ

ಘ ಘಾ ಘಿ ಘೀ ಘೃ ಘೆ ಘೇ ಘೈ ಘೊ ಘೊ ಘೋ ಘೌ ಘಂ ಘಃ

ಙ ಙಾ ಙಿ ಙೀ ಙು ಙೂ ಙೃ ಙೆ ಙೇ ಙೈ ಙೊ ಙೋ ಙೌ ಙಂ ಙಃ

ಚ ಚಾ ಚಿ ಚೀ ಚು ಚೂ ಚೃ ಚೆ ಚೇ ಚೈ ಚೊ ಚೋ ಚೌ ಚಂ ಚಃ

ಛಾ ಛಿ ಛೀ ಛು ಛೂ ಛೃ ಛೆ ಛೇ ಛೈ ಛೊ ಛೋ ಛೌ ಛಂ ಛಃ

ಜ ಜಾ ಜಿ ಜೀ ಜು ಜೂ ಜೃ ಜೆ ಜೇ ಜೈ ಜೊ ಜೋ ಜೌ ಜಂ ಜಃ

ಝ ಝಾ ಝಿ ಝೀ ಝು ಝೂ ಝೃ ಝೆ ಝೇ ಝೈ ಝೊ ಝೋ ಝೌ ಝಂ ಝಃ

ಞ ಞಾ ಞಿ ಞೀ ಞು ಞೂ ಞೃ ಞೆ ಞೇ ಞೈ ಞೊ ಞೋ ಞೌ ಞಂ ಞಃ

ತ ತಾ ತಿ ತೀ ತು ತೂ ತೃ ತೆ ತೇ ತೈ ತೊ ತೋ ತೌ ತಂ ತಃ

ಥ ಥಾ ಥಿ ಥೀ ಥು ಥೂ ಥೃ ಥೆ ಥೇ ಥೈ ಥೊ ಥೋ ಥೌ ಥಂ ಥಃ

ದ ದಾ ದಿ ದೀ ದು ದೂ ದೃ ದೆ ದೇ ದೈ ದೊ ದೋ ದೌ ದಂ ದಃ

ಧ ಧಾ ಧಿ ಧೀ ಧು ಧೂ ಧೃ ಧೆ ಧೇ ಧೈ ಧೊ ಧೋ ಧೌ ಧಂ ಧಃ

ನ ನಾ ನಿ ನೀ ನು ನೂ ನೃ ನೆ ನೇ ನೈ ನೊ ನೋ ನೌ ನಂ ನಃ

ಟ ಟಾ ಟಿ ಟೀ ಟು ಟೂ ಟೃ ಟೆ ಟೇ ಟೈ ಟೊ ಟೋ ಟೌ ಟಂ ಟಃ

ಠ ಠಾ ಠಿ ಠೀ ಠು ಠೂ ಠೃ ಠೆ ಠೇ ಠೈ ಠೊ ಠೋ ಠೌ ಠಂ ಠಃ

ಡ ಡಾ ಡಿ ಡೀ ಡು ಡೂ ಡೃ ಡೆ ಡೇ ಡೈ ಡೊ ಡೋ ಡೌ ಡಂ ಡಃ

ಡ ಢಾ ಢಿ ಢೀ ಢು ಢೂ ಢೃ ಢೆ ಢೇ ಢೈ ಢೊ ಢೋ ಢೌ ಢಂ ಢಃ

ಣ ಣಾ ಣಿ ಣೀ ಣು ಣೂ ಣೃ ಣೆ ಣೇ ಣೈ ಣೊ ಣೋ ಣೌ ಣಂ ಣಃ

ಪ ಪಾ ಪಿ ಪೀ ಪು ಪೂ ಪೃ ಪೆ ಪೇ ಪೈ ಪೊ ಪೋ ಪೌ ಪಂ ಪಃ

ಫ ಫಾ ಫಿ ಫೀ ಫು ಫೂ ಫೃ ಫೆ ಫೇ ಫೈ ಫೊ ಫೋ ಫೌ ಫಂ ಫಃ

ಬ ಬಾ ಬಿ ಬೀ ಬು ಬೂ ಬೃ ಬೆ ಬೇ ಬೈ ಬೊ ಬೋ ಬೌ ಬಂ ಬಃ

ಭ ಭಾ ಭಿ ಭೀ ಭು ಭೂ ಭೃ ಭೆ ಭೇ ಭೈ ಭೊ ಭೋ ಭೌ ಭಂ ಭಃ

ಮ ಮಾ ಮಿ ಮೀ ಮು ಮೂ ಮೃ ಮೆ ಮೇ ಮೈ ಮೊ ಮೋ ಮೌ ಮಂ ಮಃ

ಯ ಯಾ ಯಿ ಯೀ ಯು ಯೂ ಯೃ ಯೆ ಯೇ ಯೈ ಯೊ ಯೋ ಯೌ ಯಂ ಯಃ

ರ ರಾ ರಿ ರೀ ರು ರೂ ರೃ ರೆ ರೇ ರೈ ರೊ ರೋ ರೌ ರಂ ರಃ

ಲ ಲಾ ಲಿ ಲೀ ಲು ಲೂ ಲೃ ಲೆ ಲೇ ಲೈ ಲೊ ಲೋ ಲೌ ಲಂ ಲಃ

ವ ವಾ ವಿ ವೀ ವು ವೂ ವೃ ವೆ ವೇ ವೈ ವೊ ವೋ ವೌ ವಂ ವಃ

ಶ ಶಾ ಶಿ ಶೀ ಶು ಶೂ ಶೃ ಶೆ ಶೇ ಶೈ ಶೊ ಶೋ ಶೌ ಶಂ ಶಃ

ಷ ಷಾ ಷಿ ಷೀ ಷು ಷೂ ಷೃ ಷೆ ಷೇ ಷೈ ಷೊ ಷೋ ಷೌ

ಸ ಸಾ ಸಿ ಸೀ ಸು ಸೂ ಸೃ ಸೆ ಸೇ ಸೈ ಸೊ ಸೋ ಸೌ ಸಂ ಸಃ

ಹ ಹಾ ಹಿ ಹೀ ಹು ಹೂ ಹೃ ಹೆ ಹೇ ಹೈ ಹೊ ಹೋ ಹೌ ಹಂ ಹಃ

ಳ ಳಾ ಳಿ ಳೀ ಳು ಳೂ ಳೃ ಳೆ ಳೇ ಳೈ ಳೊ ಳೋ ಳೌ ಳಂ ಳಃ

2. Technical Characteristics

Note: The Convention followed from this section of the document is same as the Unicode Chapter 9 (South and Southeast Asian Scripts) and might not be grammatically correct.

2.1 Kannada Alphabet Characteristic

2.1.2 Consonant Letters

Each of the consonant represents a single consonantal sound but also has the peculiarity of having inherent vowel, generally the short vowel ಅ (U+0C85). Thus, U+0C95 Kannada letter KA represents not just K (ಕ್) but KA (ಕ). In the presence of the dependent vowel, however, the inherent vowel associated with a consonant letter is overridden by the dependent vowel. The different Consonants in Kannada are:

ಕ ಖ ಗ ಘ ಙ

ಚ ಛ ಜ ಝ ಞ

ತ ಥ ದ ಧ ನ

ಟ ಠ ಡ ಢ ಣ

ಪ ಫ ಬ ಭ ಮ

ಯ ರ ಲ ವ

ಶ ಷ ಸ ಹ ಳ

2.1.3 Dependent Vowel Signs (Matras)

The dependent vowels, also known as Swaras in Kannada, serve as the common manner of writing non-inherent vowels and are generally referred to as Swara Chinhas in Kannada or Matras in Sanskrit. The dependent vowels do not appear stand-alone; rather, they are visibly depicted in combination with a base-letter form (generally a consonant). A single consonant or a consonant cluster may have a dependent vowel applied to it to indicate the vowel quality of the syllable, when it is different from the inherent vowel. Explicit appearance of a dependent vowel in a syllable overrides the inherent vowel ಆ (U +0C85) of a single consonant letter.

There are several variations with which the dependent vowels are applied to the base letterforms. Most of them appear as non-spacing dependent vowels signs when applied to base letterforms; above to the right side of a consonant letter or a consonant cluster. The following are the exceptions and variations for the above rule:

· The two dependent vowel signs (U+0CCC3 & U+0CC4) appear one level below and to the right of the consonant or the consonant cluster, separated by a small white space.

· Each of the five dependent vowels (U+0CC0, U+0CC7, U+0CC8, U+0CCA & U+0CCB) are depicted by two or three glyph components (two part or three part vowel signs ) with one component appearing with a space to the right of the consonant or the consonant cluster.

i) In the case o f three the above-mentioned two/three-part dependent vowels (at U+0CC0, U+0CC7, U+0CCB), the non-spacing components of each of them is (are) the same as the vowel sign(s) of the corresponding preceding short vowels. The spacing component for each of these dependent vowels is the same ‘length mark U+0CD5 given in Unicode version 3. The logic for this is that these dependent vowels are nothing but the long forms (independent and phonetically distinct) of the preceding short vowels.

ii) The first component of the dependent vowel (U+0CC8) mentioned above, is the same as the dependent vowel (ೆ, U+0CC6) with the second component (U+0CD6) defined independently in Unicode version 3. The second part appears slightly below and to the right of the consonant or the consonant clusters.

· In view of this, it is important to note that the two glyphs (the length mark and the second component of ೈ) represent with the codes at U+0CD5 and U+0CD6 in Unicode version 3 have no independent existence and do not play any part as independent codes in the collation algorithm.

· Unlike Devanagari, the Kannada script does not have any character with a left-side dependent vowel sign.

· A one-to-one correspondence exists between independent vowels and dependent vowel signs.

The Matras are-

ಾ ಿ ೀ ು ೂ ೃ ೆ ೇ ೈ ೊ ೋ ೌ ಂ ಃ

2.1.4 Virama (Halant)

Like Devanagari, Kannada script also employs a sign known as Halant or vowel omission sign. A halant sign (್, U+0CCD) nominally serves to cancel (or kill) the inherent vowel of the consonant to which it is applied. It functions as a combining character. When a consonant has lost its inherent vowel by the application of halant, it is known as a dead consonant. The dead consonants are the presentation forms used to depict the consonants without an inherent vowel. Their rendered forms in Kannada resemble the full consonant with vertical stem replaced by the halant sign, which marks a character core. The stem glyph (U+0CBB) is graphically and historically related to the sign denoting the inherent /a/ (ಅ) vowel (U+0C85). In contrast, a live consonant is a consonant that retains its inherent vowel or is written with an explicit dependent vowel sign. The dead consonant is defined as a sequence consisting of a consonant letter followed by a halant. The default rendering for a dead consonant is to position the halant as a combining mark bound to the consonant letterform. The Halant in Kannada is ್

2.1.5 Consonant Conjuncts

Like any other Indian script, Kannada is also noted for a large number of consonant conjunct forms that serve as orthographic abbreviations (ligatures) of two or more adjacent forms. This abbreviation takes place only in the context of a consonant cluster. An orthographic consonant cluster is defined as a sequence of characters that represent one or more dead consonants (denoted by C_d) followed by a normal live consonant (denoted by C_l).

Corresponding to each Kannada consonant, there exists a separate and unique glyph, which is specially used to represent the corresponding consonant in a consonant cluster. Most of these conjunct consonant glyphs resemble their original consonant forms (many without the implicit vowel sign, wherever applicable).

In Kannada, there is only one type of conjunct formation (consonant cluster) and it is depicted as follows:

· The first consonant of the consonant cluster is rendered with the implicit or a different dependant vowel appearing as the terminal element of the consonant cluster.

· The remaining consonants (consonants in between the first consonant and the terminal vowel element) appear in conjunct consonant glyph forms in the phonetic order. They are generally depicted directly below or sometimes below but to the right of the first consonant.

Thus, the systematically designed Kannada script font contains the conjunct glyph components, but they are not encoded as Unicode characters, because they are the resultant of ligation of distinct letters. Kannada script rendering software must be able to map appropriate combinations of characters in context to the appropriate conjunct glyphs in fonts.

2.1.6 Visarg

Comes after a vowel sound and represents a sound similar to /h/.

2.1.7 Avagraha

Avagraha sign is a spacing mark used while rendering Sanskrit text. This is located at U+0CBD.

2.1.8 Numerals

Kannada numerals are located from U+0CE6 to U+0CEF

2.1.9 Punctuation Marks

All Punctuation mark in Kannada is borrowed from English.

These characters are not included in the range for Kannada in the Unicode Character set.

2.1.10 Ancient Signs

Some of he Halegannada characters are placed at U+0C8C, U+0CB1, U+0CE1.

2.2 Fonts

There are a number of TrueType fonts available for Kannada among which some of them follow an encoding standard (like ISCII) and others do not follow any encoding standard and is tied to a proprietary encoding. The Kannada Ganaka Parishat has standardised the glyph set to be used by all the Software that support Kannada. Annexure-1 displays the glyphs standardised by KGP. Microsoft has released an OpenType font (with TrueType outlines) for Kannada – “Tunga.ttf” that follows Unicode as its encoding standard.

2.2.1 Font Developing Tools

The OpenType font format is an extension of the TrueType font format, adding support for PostScript font data. The following tools can be used for designing of the OpenType Font.

FontLab version 4.0 (A trial version for Windows is available for download at http://www.fontlab.com/html/fontlab.html#downloads)
Font Creator Program version 3.0 (A trial version is available for download at http://www.high-logic.com/download.htm)
Fontographer version 4.1 (No trial version is available. For more information - http://www.macromedia.com/software/fontographer/)
Microsoft Visual OpenType Layout Tool "VOLT" provides an easy-to-use graphical user interface to add OpenType layout tables to fonts with TrueType outlines. It is licensed free and can be downloaded from the online community http://communities.msn.com/MicrosoftVOLTuserscommunity

2.3 Keyboard

Inputting Kannada or any other Indian language needs Keyboard driver / Input Method, which is a Software Component that interprets user operations such as typing keys. There are many Keyboard Drivers/Input Methods available in the market for Windows Operating Systems like Baraha, Sreelipi, Akruthi, Kalitha etc. They follow different encoding methods (glyph codes) and support different keyboard layouts like INSCRIPT, English Phonetic, Typewriter 1, and Typewriter 2.

Microsoft supports Input Methods for nine Indian languages (including Kannada) in Office XP on Windows XP with INSCRIPT keyboard layout, which is common to all Indian languages and uses Unicode as the encoding Standard.

Government of Karnataka (Kannada Ganaka Parishat) has proposed a Standard Keyboard layout for Kannada. In this layout only 26 keys which are painted with English characters on a Keyboard can be used to represent 51 basic alphabets and special symbols in Kannada (13 swaras, 34 consonants and 4 special symbols). This is possible as each key has a dual function of representing the small case (normal key) and Capital case (Shift key) letters in English, as shown in the figure 2.1

Fig. 2.1: Keyboard Layout proposed by Kannada Ganaka Parishath.

Since the 51 keys have been used while the keyboard provides 52 possibilities, that option (key X is not assigned) can be used to represent foreign sounds such as combination of Nukta and ಫ. Besides, there is a need to represent the old ಱ. The combination of X with consonant ರ yields ಱ.

2.4 Presentation and Storage Considerations

The order for storage of plain text in Kannada generally follows the phonetic order, that is, a CV syllable with a dependant vowel is always encoded as a consonant letter C followed by a vowel sign V in the memory representation. This order is employed by the ISCII standard and corresponds with phonetic and keying order of textual data. Unlike Devanagari and some other Indian scripts, all the dependent vowels in Kannada are depicted to the right of their consonant letters. Hence there is no need to reorder the elements in mapping from the logical (character) store to the presentation (glyph) rendering and vice versa.

Character order Glyph order

KA + I KA + I

ಕ + ಇ ಕಿ

Further, Kannada script does not allow half-consonants, ligatures and half ligature forms.

2.5 Rendering Rules

(Based on Microsoft Uniscribe-OpenType implementation of the UNICODE Rendering Rules)

2.5.1 Notation. In the next set of rules, the following notation applies:

C_n Nominal glyph form of a consonant C as it appears in the code charts.

C_l A live consonant, depicted identically to C_n.

C_d Glyph depicting the dead consonant form of a consonant C.

C_h Glyph depicting the half-consonant form of a consonant C.

L_n Nominal glyph form of a conjunct ligature consisting of two or more component consonants. A conjunct ligature composed of two consonants X and Y is also denoted by X.Y_n.

RA_sub A non-spacing combining mark glyph form positioned below the base glyph form.

V_vs Glyph depicting the dependent vowel sign form of a vowel V.

Virama_n The nominal glyph form non-spacing combining mark depicting U+0CCD Kannada sign Virama.

A virama character is not always depicted; when it is depicted, it adopts this non-spacing mark form.

2.5.1 Dead Consonant Rule. The following rule logically precedes the application of any other rule to form a dead consonant. Once formed, a dead consonant may be subject to other rules described next.

R0: When a consonant C_n precedes a VIRAMA_n, it is considered to be a dead consonant C_d. A consonant C_n that does not precede VIRAMA_n is considered to be live consonant C_l.

TA_n+ VIRAMA_n TA_d

ತ + ್ ತ್

(Based on Microsoft Uniscribe-OpenType implementation of the UNICODE Rendering Rules)

2.5.2 Consonant RA Rules:

R1: If the dead consonant RA_d precedes either a consonant or an independent vowel, then it is replaced by the postscript ARKAVATTU, which is positioned so that it applies to the logically subsequent element in the memory representation.

RA_d + KA_l KA_l + ARKAVATTU Displayed Output

ರ್ + ಕ ಕ +ARKAVATTU ರ್ಕ

R2: Except for the dead consonant RA_d, when a dead consonant C_d precedes the live consonant RA_l, then C_d is replaced with its nominal form C_n , and RA is replaced by the subscript non-spacing mark RA_sub , which is positioned so that it applies to C_n.

THA_d + RA_l THA_n + RA_sub Displayed Output

ಠ್ + ರ ಠ + RA_sub ಠ್ರ

R3: If a dead consonant (other than RA_d) precedes RA_d, then the substitution of RA for RA_sub is performed as described above; however, the VIRAMA that formed RA_d remains so as to form a dead consonant conjunct form. A dead consonant conjunct form that contains an absorbed RA_d may subsequently combine to form a multipart conjunct form.

TA_d + RA_d TA_n + RA_sub+ VIRAMA_n T.RA_d

ತ್ + ರ್ ತ + RA_sub + ್ ತ್ರ್

A dead consonant conjunct form that contains an absorbed RA_d may subsequently combine to form a multipart conjunct form.

T.RA_d + YA_l T.R.YA_n

ತ್ರ್ + ಯ ತ್ರ್ಯ

Ligature Rules: Subsequent to the application of the rules just described, a set of rules governing ligature formation apply. The precise application of these rules depends on the availability of glyphs in the current font(s) being used to display text.

R4: If a dead consonant immediately precedes another dead consonant or a live consonant, then the first dead consonant may join the subsequent element to form a twopart conjunct ligature form.

JA_d + NYA_l J.NYA_n

ಜ್ + ಞ ಜ್ಞ

R5: A conjunct ligature form can itself behave as a dead consonant and enter into further, more complex ligatures.

SAd + TAd + KAl S.T.KAn

ಸ್ + ತ್ + ಕ ಸ್ತ್ಕ

R6: If a nominal consonant or conjunct ligature form precedes RA_sub as a result of the application of rule R2, then the consonant or ligature form may join with RA_sub to form a multipart conjunct ligature (see rule R2 for more information).

KA_l + RA_d K.RA_n

ಕ + ರ್ ಕ್ರ

R7: In some cases, other combining marks will also combine with a base consonant, either attaching at a nonstandard location or changing shape. In minimal rendering there are only two cases, RA_l with U_vs or UU_vs .

RA_l + U_vs RU_n RA_l + UU_vs RUU_n

ರ + ು ರು ರ + ೂ ರೂ

R8: When the dependent vowel I_vs is used to override the inherent vowel of a syllable, it is always written to the extreme left of the orthographic syllable. If the orthographic syllable contains a consonant cluster, then this vowel is always depicted to the left of that cluster.

THA_d + RA_l + I_vs T.RI_n

ತ್ + ರ + ಿ ತ್ರಿ

2.6 Sorting issues in Kannada

The sorting sequence for Kannada Unicode is as per the collation chart given as an annexure. However, the following are some important issues, which have to be addressed separately for proper sorting of data in Kannada.

ISCII – 91 provides direct sorting through its codes. It is the natural sorting method just based on code values. There are no special algorithms for language specific issues for sorting the data. This results in non-conventional sorting in some specific cases. The scholars in Kannada have specified the sorting standards in Kannada. These standards are being followed in all dictionaries and other documents in Kannada. With this in view, the following four special cases have been identified.

2.6.1 Sorting of Nukta characters

The modifying mark or Nukta located at U+0CBC and included in the collation table is enough to take care of the sorting issues of characters ಜ + (U+0CBC) (modified ಜ) and ಫ + (U+0CBC) (modified ಫ). It also takes care of any other consonant, which may be modified using Nukta.

2.6.2 Sorting the data records containing anuswara and visarga

In case of sorting a data set containing words terminating with anuswara, visarga together with other words, words without terminating dependent vowels are placed in wrong positions.

Sorting sequence as per the Unicode is according to the specified standards if the anuswara and visarga appear within a word.

2.6.3 Sorting of words with dead consonants

Sorting of words terminating with dead consonants

Sorting in this case also violates the sorting rules of Kannada. The Unicode sorting places the word terminating with the dead consonant at the end of the list. The following list compares the sorting of a sample data using Unicode table and the acceptable sorting for this case.

Sorted data as per Unicode	Acceptable sorting
ರಾಕ	ರಾಕ್
ರಾಕ್	ರಾಕ
ರಾಗ	ರಾಗ್
ರಾಗೋ	ರಾಗ
ರಾಗ್	ರಾಗೋ

Dead consonants within words

Proper sorting of data with such words can be achieved by using the invisible zero width consonant just after the dead consonant.

To circumvent unacceptable situations mentioned in 2.6.2 and 2.6.3 above, the Unicode Standard character U+200C (Zero width Non-Joiner) can be used appropriately in the preprocessor and collation algorithms.

2.6.4 Sorting of Conjuncts having two different display forms

Two such conjuncts are rendered in Kannada at present.

· Conjuncts with (U+0CB0) as the first consonant

This has been explained at an earlier section as Consonant RA rules (section 2.5.2)

Words containing both the display forms of the same consonant cluster with ರ (U+0CB0) as the first consonant of the cluster had to be sorted as follows. Even though the display rendering are different, both are identical in all respects. It is therefore natural that they should appear at consecutive positions. Even though a separate glyph and a corresponding glyph code are present in the display/storage codes such an arrangement in Unicode will not render for proper sorting.

The only alternative is to represent both the display forms by the same set of codes with a distinguishing code (U+0CF5) within the string for the second display form. In Unicode form, the distinguishing code value within the string of the consonant sorting. This can be achieved through preprocessing software, with specific functions to generate proper glyph codes, storage codes and Unicode at different levels. Such a situation–specific code representation guarantees proper sorting of data containing consonant clusters with two different display forms by ignoring the code U+0CF5 for . This condition has to be incorporated at the appropriate place in the sorting algorithm.

· The second case of rendering a same character in two different display forms is the dead consonant ನ್. It is also written in a second form as U+0CF5. Sorting issue in regards to this case is also dealt with the same way as in the previous case.

The zero width Non-Joiner at U+200C cannot be used instead of (U+0CF5),as the same sequence of characters appear both with Zero width Non-Joiner and with U+0CF5, the two sequences representing two different syllables (conjuncts).

2.6.5 Sorting of Diacritic characters

Diacritic characters formed using symbols located at 0CD1, 0CD2, 0CD3, 0CD4 and 0CF9 to render accents to consonants, are considered to be equivalent to the corresponding consonants for sorting purposes and hence the above procedure can be adopted in such cases also.

2.6.6 Conclusion

The sorting issues mentioned above may have multiple solutions. Similar issues might have been solved by different methods in respect of other Indian languages. Hence, it is desirable to evolve uniform procedures for issues common to all Indian languages. However, solutions for sorting problems mentioned here with respect Kannada have been obtained by considering all consonants from U+0C95 to U+0CB9 and the consonant U+0CDE when they appear independently in a data field as pure consonants (i.e. as two part coded[Ex:0C95=(0C95,0CBB)]). The sorting of a data field is achieved by the indexing method. All these can be elaborated to give the actual algorithms and floe charts, if need be.

References:

1. “South and South East Asian Scripts”, Chapter 9 of The Unicode Standard (Version 3.0), http://www.unicode.org/

2. “Creating and Supporting Open Type Fonts for Indic Scripts”, http://www.microsoft.com/typography

3. “Unicode for Kannada Script” (written by Dr. CV Srinatha Sastry), Directorate of Information Technology, Government of Karnataka

4. “User-Friendly Keyboard Layout for Kannada”, Kannada Ganaka Parishat

5. “Standards for Kannada in Computers prescribed by the Government of Karnataka”, Kannada Ganaka Parishat.

Appendix – 1: Unicode Chart and Collation if the suggested deletion and relocation of characters are not allowed

Below figure shows the Unicode Chart for Kannada if deletion and relocation of Characters are not allowed

	0C8	0C9	0CA	0CB	0CC	0CD	0CE	0CF
0	▓▓	ಐ	ಠ	ರ	ೀ	▓▓	ೠ	▓▓
1	▓▓	▓▓	ಡ	ಱ	ು		ೡ	▓▓
2	ಂ	ಒ	ಢ	ಲ	ೂ		▓▓	▓▓
3	ಃ	ಓ	ಣ	ಳ	ೃ		▓▓	▓▓
4	▓▓	ಔ	ತ	▓▓		▓▓	▓▓	▓▓
5	ಅ	ಕ	ಥ	ವ	▓▓	ೕ	▓▓
6	ಆ	ಖ	ದ	ಶ	ೆ	ೖ	0	▓▓
7	ಇ	ಗ	ಧ	ಷ	ೇ	▓▓	೧	▓▓
8	ಈ	ಘ	ನ	ಸ	ೈ	▓▓	೨	▓▓
9	ಉ	ಙ	▓▓	ಹ	▓▓	▓▓	೩
A	ಊ	ಚ	ಪ		ೊ	▓▓	೪	▓▓
B	ಋ	ಛ	ಫ		ೋ	▓▓	೫	▓▓
C		ಜ	ಬ		ೌ	▓▓	೬	▓▓
D	▓▓	ಝ	ಭ		್	▓▓	೭	▓▓
E	ಎ	ಞ	ಮ	ಾ	▓▓		೮	▓▓
F	ಏ	ಟ	ಯ	ಿ	▓▓	▓▓	೯	▓▓

Below figure shows the Collating sequence of Kannada Unicode characters, if additions and relocations are not allowed. The sequence is column wise, top to bottom

Column 1

Column 2

Column 3

Column 4

Column5

0C82

ಂ

0CCD

್

0C96

ಖ

0CA6

ದ

0CB9

ಹ

0C83

ಃ

0CBB

0C97

ಗ

0CA7

ಧ

0CB3

ಳ

0C85

ಅ

0CBE

ಾ

0C98

ಘ

0CA8

ನ

0CB4

0C86

ಆ

0CBF

ಿ

0C99

ಙ

0CAA

ಪ

0CBC

0C87

ಇ

0CC0

ೀ

0C9A

ಚ

0CAB

ಫ

0C88

ಈ

0CC1

ು

0C9B

ಛ

0CAC

ಬ

0C89

ಉ

0CC2

ೂ

0C9C

ಜ

0CAD

ಭ

0C8A

ಊ

0CC3

ೃ

0C9D

ಝ

0CAE

ಮ

0C8B

ಋ

0CC4

ೄ

0C9E

ಞ

0CAF

ಯ

0CE0

ೠ

0CC6

ೆ

0C9F

ಟ

0CB0

ರ

0C8E

ಎ

0CC7

ೇ

0CA0

ಠ

0CB1

ಱ

0C8F

ಏ

0CC8

ೈ

0CA1

ಡ

0CB2

ಲ

0C90

ಐ

0CCA

ೊ

0CA2

ಢ

0CB5

ವ

0C92

ಒ

0CCB

ೋ

0CA3

ಣ

0CB6

ಶ

0C93

ಓ

0CCC

ೌ

0CA4

ತ

0CB7

ಷ

0C94

ಔ

0C95

ಕ

0CA5

ಥ

0CB8

ಸ

Appendix – 2: Unicode Chart and Collation if the suggested deletion and relocation of characters are allowed

Below figure shows the Unicode Chart for Kannada if deletion and relocation of Characters are allowed

	0C8	0C9	0CA	0CB	0CC	0CD	0CE	0CF
0	▓▓	ಐ	ಠ	ರ	ೀ	▓▓	ೠ	▓▓
1	▓▓	▓▓	ಡ	ಱ	ು		▓▓	▓▓
2	ಂ	ಒ	ಢ	ಲ	ೂ		▓▓	▓▓
3	ಃ	ಓ	ಣ	ಳ	ೃ		▓▓	▓▓
4	▓▓	ಔ	ತ	▓▓		▓▓	▓▓	▓▓
5	ಅ	ಕ	ಥ	ವ	▓▓	▓▓	▓▓
6	ಆ	ಖ	ದ	ಶ	ೆ	▓▓	0	▓▓
7	ಇ	ಗ	ಧ	ಷ	ೇ	▓▓	೧	▓▓
8	ಈ	ಘ	ನ	ಸ	ೈ	▓▓	೨	▓▓
9	ಉ	ಙ	▓▓	ಹ	▓▓	▓▓	೩
A	ಊ	ಚ	ಪ		ೊ	▓▓	೪	▓▓
B	ಋ	ಛ	ಫ		ೋ	▓▓	೫	▓▓
C	▓▓	ಜ	ಬ		ೌ	▓▓	೬	▓▓
D	▓▓	ಝ	ಭ		್	▓▓	೭	▓▓
E	ಎ	ಞ	ಮ	ಾ	▓▓	▓▓	೮	▓▓
F	ಏ	ಟ	ಯ	ಿ	▓▓	▓▓	೯	▓▓

Below figure shows the Collating sequence of Kannada Unicode characters, if additions and relocations are allowed. The sequence is column wise, top to bottom.