BashaFusion (PhoneticMap)

A package used to

• convert indic language like hinid, kannada, malayalam,telugu,Odia, Bengali,Gujarathi to iast and
• iast to Other inidc langauge(hinid, kannada, malayalam,telugu,Odia, Bengali,Gujarathi) viceversa.
• Phonetic Word Search: Searching word which sound similar or Phoneticly Similar in different languages like hinid, kannada, malayalam,telugu,Odia,Bengali,Gujarathi...etc
• Not Started: addon for bash regual expression function to include Phonetic Word Search for Indic Language
• Inprogress: Mapping all Indic alphabets to International Phonetic Alphabet (IPA)
• Not Started: Generated Phonetic Word given new indic word using Phonetic Mapped Alphabet
• Not Started: Generating audio file for indic word
• Not Started: Pull request of Postgress to include Phonetic Alphabet Search in IPA and IAST format

Installation

pip install BashaFusion  

Usage

import sqlite3
import os
import sys

from BashaFusion import BashaFusion

#create a IAST object
bshf = BashaFusion()

# customization
# bshf = BashaFusion(db_path='iast-token.db', table_name_alpha='IndianAlphabet',table_name_barakadi='Barakhadi')

Converting All Indic language(hinid, telugu, kannada, Malayalam, Odia, Bengali&Assamese, Gujarati, tamil) to iast

InProgress Research and Analysis is going on in Tamil Script, Nastaliq Script, Sinhala Script.

bshf.to_iast('''ଧୃତରାଷ୍ଟ୍ର ଉଵାଚ |\tধৃতরাষ্ট্র উবাচ |\tધૃતરાષ્ટ્ર ઉવાચ |\tத்றுதராஷ்ட்ர உவாச |''')
# >>> 
# dhr̥tarāṣṭra uvāca |	dhr̥tarāṣṭra ubāca |	dhr̥tarāṣṭra uvāca |	ta்ṟutarāṣa்ṭa்ra uvāca |

Convert iast to Indic Language

Currently this can convert IAST to kannada, hindi, telugu, malyalam

word = 'kaṁ  itāḥ kiṁ  yuyutsavaḥ kl̥̄ kl̥ pāṇḍavānīkaṁ itāḥ kiṁ āṁ  īṁ   yuyutsuṁ  kiṁ rānsakhīṁstathā'
print(BashaFusion.iast2tokens( word) )
# >>> ['k', 'a', 'ṁ', '  ', 'i', 't', 'ā', 'ḥ', ' ', 'k', 'i', 'ṁ', '  ', 'y', 'u', 'y', 'u', 't', 's', 'a', 'v', 'aḥ', ' ', 'k', 'l̥̄', ' ', 'k', 'l̥', ' ', 'p', 'ā', 'ṇ', 'ḍ', 'a', 'v', 'ā', 'n', 'ī', 'k', 'a', 'ṁ', ' ', 'i', 't', 'ā', 'ḥ', ' ', 'k', 'i', 'ṁ', ' ', 'ā', 'ṁ', '  ', 'ī', 'ṁ', '   ', 'y', 'u', 'y', 'u', 't', 's', 'u', 'ṁ', '  ', 'k', 'i', 'ṁ', ' ', 'r', 'ā', 'n', 's', 'a', 'kh', 'ī', 'ṁ', 's', 't', 'a', 'th', 'ā']
indic_lang = 'Telugu' # 'Kannada' # 'Telugu', 'Odia', 'Gujarati', 'Bengali-Assamsese', 
# indic_lang='Devanagari'
# indic_lang='Kannada'
# indic_lang='Telugu'
# indic_lang='Odia'
# indic_lang='Bengali–Assamese'
# indic_lang='Tamil' # In development state

print(bshf.iast2indic(word,indic_lang))



BashaFusion.dict_tokens2indic(dict_tokene_list,halant)
# >>> కం  ఇతాః కిం  యుయుత్సవః  పాణ్డవానీకం ఇతాః కిం ఆం  ఈం  కిం యుయుత్సుం రాన్సఖీంస్తథా

Phonetic Hash for Phonetic Search

Presently there are two type of Phonetic Hash

• BASIC HASHING :
  

  • all vowel are removed and
    
  • consonats are truncated
    eg: kh is mapped to k,
    श, ष, स are mapped to स (s)
    ङ,ञ,ण,न are mapped न (n)
    ..etc

• NORMAL HASHING

  • vowel are truncated and
    
  • consonats are truncated

zero_vowels={ '':['a', "ā", "â","i", "ī","u", "ū",chr(805),chr(803),
                    "l̥", "l̥̄","e", "ē", "ê","o", "ō", "ô",
                    "ṁ", "m̐", "ṃ", "ṃ","n̆", "n̆", "n̆","ḥ" , "ḫ", "ẖ", "ḥ"],
                'r': ["r̥", "r̥̄"]
            } # replacing with r is not working for 'r̥' so we replace with chr(805) above
truncated_vowels = { '':[chr(805), chr(803), chr(772),chr(784),chr(774)],
                    'a':["ā", "â"], 
                    'i':["i", "ī"], 
                    'u':["u", "ū"], 
                    'r':["r̥", "r̥̄"],
                    'l':["l̥", "l̥̄"],
                    "e":["e", "ē", "ê"],
                            # "ai", 
                    "o": ["o", "ō", "ô"], 
                                            # "au",
                    'm' :["ṁ", "m̐", "ṃ", "ṃ"], 
                    'n': ["n̆", "n̆", "n̆"], 
                    'h' :["ḥ" , "ḫ", "ẖ", "ḥ"],
                    }
basic_truncated_consonat = {
                    'k' : ['ḵ', 'k', 'kh','k͟ha'],
                    'g' :['g','g̈','gh','ġ'],
                    'n' : ['ṅ','n̆','ñ','ṇ','ṉ'],
                    'c' : ['c', 'ĉ','ch','ĉh'],
                    'j' : ['j','ǰ', 'ĵ', 'jh'],
                    't': ['ṭ','ṭa','t','th','ṯ','t̤'],
                    'd' : ['ḍ', 'd̤','ḍ','ḍh','d','dh','ḏ'],
                    'p' : ['p', 'ph'],
                    'b' : [ 'b', 'b̤', 'bh'],
                    'm' : ['m̆' ],
                    'r' : ['ṟ', 'r̆'],
                    'l' :['ḻ', 'ḷ', 'l̤'],
                    'y' : ['y', 'ẏ', 'y̌'],
                    's': ['ś', 'ṣ', 's','s̱', 's̤','sh' ],
                    "z": ["z","ž","ž","ž",'ẕ','ẕ','ẓ','ż'],
                    'h' : ['h','h̤']
                    }

search_word = 'dhr̥tarāṣṭra uvāca'
search_word = search_word.strip().lower()



# to_iast
search_iast = bshf.to_iast(search_word) # similar to idempotent matrx no loss of info if ':' not present
# >>> dhr̥tarāṣṭra uvāca
print(search_iast)

print("# Original Text:", search_word)
print('BASIC HASHING: ',BashaFusion.basic_hash(search_iast))
print('NORMAL HASHING',BashaFusion.normal_hash(search_iast))

# >>> Original Test: dhr̥tarāṣṭra uvāca
# >>> BASIC HASHING: drtrstr vc
# >>> NORMAL HASHING: drtarastra uvaca

Contribute

InProgress Research and Analysis is going on in Tamil Script, Nastaliq Script(Urdu, Arabic), Sinhala(Sri Lanka) Script.
Not Strated Marathi can easily added into this package

Issue

Please open an issue
here in case any bug was encountered. Mail id : dankarthik25@gmail.com