-- Copyright (c) 2016-present, Facebook, Inc. -- All rights reserved. -- -- This source code is licensed under the BSD-style license found in the -- LICENSE file in the root directory of this source tree. An additional grant -- of patent rights can be found in the PATENTS file in the same directory. {-# LANGUAGE GADTs #-} {-# LANGUAGE OverloadedStrings #-} {-# LANGUAGE NoRebindableSyntax #-} module Duckling.Numeral.HI.Rules ( rules ) where import Data.HashMap.Strict (HashMap) import Data.Maybe import Data.String import Data.Text (Text) import Prelude import qualified Data.HashMap.Strict as HashMap import qualified Data.Text as Text import Duckling.Dimensions.Types import Duckling.Numeral.Helpers import Duckling.Numeral.Types (NumeralData (..)) import Duckling.Regex.Types import Duckling.Types import qualified Duckling.Numeral.Types as TNumeral devanagariMap :: HashMap Char Char devanagariMap = HashMap.fromList [ ( '०', '0' ) , ( '१', '1' ) , ( '२', '2' ) , ( '३', '3' ) , ( '४', '4' ) , ( '५', '5' ) , ( '६', '6' ) , ( '७', '7' ) , ( '८', '8' ) , ( '९', '9' ) ] devanagariToArab :: Char -> Char devanagariToArab c = HashMap.lookupDefault c c devanagariMap ruleDevanagari :: Rule ruleDevanagari = Rule { name = "devanagari forms" , pattern = [ regex "([०१२३४५६७८९]{1,18})" ] , prod = \tokens -> case tokens of (Token RegexMatch (GroupMatch (match:_)):_) -> toInteger <$> parseInt (Text.map devanagariToArab match) >>= integer _ -> Nothing } ruleNumeralMap :: HashMap Text Integer ruleNumeralMap = HashMap.fromList [ ( "शून्य", 0 ) , ( "एक", 1 ) , ( "दो" , 2 ) , ( "तीन", 3 ) , ( "चार", 4 ) , ( "पाँच", 5 ) , ( "छह", 6 ) , ( "सात", 7 ) , ( "आठ", 8 ) , ( "नौ" , 9 ) , ( "दस", 10 ) ] ruleNumeral :: Rule ruleNumeral = Rule { name = "number (0..10)" , pattern = [ regex "(शून्य|एक|दो|तीन|चार|पाँच|छह|सात|आठ|नौ|दस)" ] , prod = \tokens -> case tokens of (Token RegexMatch (GroupMatch (match:_)):_) -> HashMap.lookup match ruleNumeralMap >>= integer _ -> Nothing } elevenToNineteenMap :: HashMap Text Integer elevenToNineteenMap = HashMap.fromList [ ( "ग्यारह", 11 ) , ( "बारह", 12 ) , ( "तेरह", 13 ) , ( "चौदह", 14 ) , ( "पन्द्रह", 15 ) , ( "सोलह", 16 ) , ( "सत्रह", 17 ) , ( "अठारह", 18 ) , ( "उन्नीस", 19 ) ] ruleElevenToNineteen :: Rule ruleElevenToNineteen = Rule { name = "number (11..19)" , pattern = [ regex "(ग्यारह|बारह|तेरह|चौदह|पन्द्रह|सोलह|सत्रह|अठारह|उन्नीस)" ] , prod = \tokens -> case tokens of (Token RegexMatch (GroupMatch (match:_)):_) -> HashMap.lookup match elevenToNineteenMap >>= integer _ -> Nothing } twentyoneToTwentynineMap :: HashMap Text Integer twentyoneToTwentynineMap = HashMap.fromList [ ( "इक्कीस", 21 ) , ( "बाईस", 22 ) , ( "तेईस", 23 ) , ( "चौबीस", 24 ) , ( "पच्चीस", 25 ) , ( "छब्बीस", 26 ) , ( "सत्ताईस", 27 ) , ( "अट्ठाईस", 28 ) , ( "उनतीस", 29 ) ] ruleTwentyoneToTwentynine :: Rule ruleTwentyoneToTwentynine = Rule { name = "number (21..29)" , pattern = [ regex "(इक्कीस|बाईस|तेईस|चौबीस|पच्चीस|छब्बीस|सत्ताईस|अट्ठाईस|उनतीस)" ] , prod = \tokens -> case tokens of (Token RegexMatch (GroupMatch (match:_)):_) -> HashMap.lookup match twentyoneToTwentynineMap >>= integer _ -> Nothing } tensMap :: HashMap Text Integer tensMap = HashMap.fromList [ ( "बीस", 20 ) , ( "तीस", 30 ) , ( "चालीस", 40 ) , ( "पचास", 50 ) , ( "साठ", 60 ) , ( "सत्तर", 70 ) , ( "अस्सी", 80 ) , ( "नब्बे", 90 ) ] ruleTens :: Rule ruleTens = Rule { name = "integer (20,30..90)" , pattern = [ regex "(बीस|तीस|चालीस|पचास|साठ|सत्तर|अस्सी|नब्बे)" ] , prod = \tokens -> case tokens of (Token RegexMatch (GroupMatch (match:_)):_) -> HashMap.lookup match tensMap >>= integer _ -> Nothing } rules :: [Rule] rules = [ ruleDevanagari , ruleNumeral , ruleElevenToNineteen , ruleTwentyoneToTwentynine , ruleTens ]