Sanscript is a transliteration library for Indian languages. It supports the most popular Indian scripts and several different romanization schemes. Although Sanscript focuses on Sanskrit transliteration, it has partial support for other languages and is easy to extend.
The package is officially distributed at npm here, whereas a variant due to Vikram Iyer is separately available here. So one can use commands such as:
npm install @indic-transliteration/sanscript
Sanscript is simple to use:
var output = Sanscript.t(input, from, to);
Here, from
and to
are the names of different schemes. In Sanscript, the word "scheme" refers to both scripts and romanizations. These schemes are of two types:
- Brahmic schemes, which are abugidas. All Indian scripts are Brahmic schemes.
- Roman schemes, which are alphabets. All romanizations are Roman schemes.
For a full list of schemes, see schemes
in https://github.com/indic-transliteration/common_maps . A possibly outdated listing of supported schemes:
ahom, assamese, avestan, balinese, bengali, bhaisuki, brahmi, brahmi_tamil, burmese, chakma, cham, cyrillic, devanagari, dogra, gondi_gunjala, gondi_masaram, grantha, grantha_pandya, gujarati, gurmukhi, hk, iast, itrans, itrans_dravidian, javanese, kannada, khamti_shan, kharoshti, khmer, khom_thai, khudawadi, kolkata, lao, lao_pali, lepcha, limbu, mahajani, malayalam, manipuri, marchen, modi, mon, mro, multani, newa, ol_chiki, oriya, persian_old, phags_pa, ranjana, rejang, rohingya, sanskritOCR, shan, sharada, siddham, sinhala, slp1, sora_sompeng, sundanese, syloti_nagari, tagalog, tagbanwa, tai_laing, takri, tamil, tamil_extended, tamil_superscripted, telugu, thai, tibetan, tirhuta_maithili, urdu, vattelutu, velthuis, wancho, warang_citi, wx, zanbazar_square
of which the following are Roman schemes:
hk
(Harvard-Kyoto)iast
(International Alphabet of Sanskrit Transliteration)iso
itrans
(ITRANS)itrans_dravidian
(ITRANS with support for Dravidian short "e" and "o")kolkata
(National Library at Kolkata)slp1
(Sanskrit Library Phonetic Basic)velthuis
(Velthuis)wx
(WX)cyrillic
When Sanscript sees the token ##
, it toggles the transliteration state:
Sanscript.t('ga##Na##pa##te', 'hk', 'devanagari'); // गNaपte
Sanscript.t('ध##र्म##क्षेत्रे', 'devanagari', 'hk'); // dhaर्मkSetre
When Sanscript sees the token \
, it disables transliteration on the character that immediately follows. \
is used for ITRANS compatibility; we recommend always using ##
instead.
Sanscript.t('a \\a', 'itrans', 'devanagari'); // अ a
Sanscript.t('\\##aham', 'itrans', 'devanagari'); // ##अहम्
A lossy scheme does not have the letters needed to support lossless translation. For example, Bengali is a lossy scheme because it uses ব
for both ba
and va
. In future releases, Sanscript might let you choose how to handle lossiness. For the time being, it makes some fairly bad hard-coded assumptions. Corrections and advice are always welcome.
You can tweak the transliteration function by passing an options
object:
<script src="node_modules/@indic-transliteration/sanscript/sanscript.js"></script>
<script>
var output = Sanscript.t(input, from, to, options);
</script>
options
maps options to values. Currently, these options are supported:
skip_sgml
- If true, transliterate SGML tags as if they were ordinary words (<b>iti</b>
→<ब्>इति</ब्>
). Defaults tofalse
.syncope
- If true, use Hindi-style transliteration (ajay
→अजय
). In linguistics, this behavior is known as schwa syncope. Defaults tofalse
.preferred_alternates
- Thisobject
map can define which alternates should be used during transliteration. I.e. in case you transliterate to itrans and you preferaa
instead ofA
,ii
instead ofI
, etc., you can set a map like this:
{ itrans : { "A" : "aa", "I" : "ii", "U" : "uu", "j~n" : "GY" } }
- Check out repo from github.
- Install package dependencies with
npm install
Please note that schemes are in a different repo as a separate package dependency, which you can find in common_maps repo.
If you want to efficiently work locally editing schemes try out npm link
. Check out usage here.
In a nutshell, the steps:
- Check out in a separate folder the common_maps repo.
- Navigate into the folder where you checked out the common_maps
- Type (sudo)
npm link
in the shell - Navigate to the
sanscript.js
repo and typenpm link @indic-transliteration/common_maps
This way the node_modules/@indic-transliteration/common_maps
folder will become a link, pointing to your local checked out folder. So you can simply make there your changes.
When you want to revert to the real downloaded package simply type npm unlink @indic-transliteration/common_maps
Adding a new scheme is simple:
Sanscript.addBrahmicScheme(schemeName, schemeData);
Sanscript.addRomanScheme(schemeName, schemeData);
For help in creating schemeData
, see the comments on the addBrahmicScheme
and addRomanScheme
functions.
Prior to testing, run npm run dist
so that the distribution file sanscript.js is generated at the root folder.
We use qunit
for testing.
After installing dependencies, you can either:
- run
npm run test
to run tests from the command line - open test/index.html to run tests in the browser
npm publish --access public