diff --git a/LICENSE b/LICENSE --- a/LICENSE +++ b/LICENSE @@ -1,4 +1,5 @@ -Copyright (c) 2012-2017, Computational Linguistics Ltd. +Copyright (c) 2012-2015, Computational Linguistics Ltd. + 2012-2019, Mikolaj Konarski All rights reserved. diff --git a/NLP/Minimorph/English.hs b/NLP/Minimorph/English.hs --- a/NLP/Minimorph/English.hs +++ b/NLP/Minimorph/English.hs @@ -1,22 +1,13 @@ -{-# LANGUAGE OverloadedStrings #-} -{-# LANGUAGE ViewPatterns #-} +{-# LANGUAGE OverloadedStrings, ViewPatterns #-} -- TODO : learn how to use Functional Morphology instead - --- | Module : NLP.Minimorph.English --- Copyright : 2012 Eric Kow (Computational Linguistics Ltd.) --- License : BSD3 --- Maintainer : eric.kow@gmail.com --- Stability : experimental --- Portability : portable --- --- Simple default rules for English morphology +-- |Simple default rules for English morphology module NLP.Minimorph.English where -import Data.Char (toLower, isSpace, isUpper) -import Data.Text (Text) -import qualified Data.Text as T +import Data.Char (isSpace, isUpper, toLower) +import Data.Text (Text) +import qualified Data.Text as T -import NLP.Minimorph.Util +import NLP.Minimorph.Util -- --------------------------------------------------------------------- -- ** Punctuation @@ -35,12 +26,14 @@ -- ** Numbers -- --------------------------------------------------------------------- --- | > cardinal 1 == "one" +-- | > cardinal 0 == "zero" +-- > cardinal 1 == "one" -- > cardinal 2 == "two" --- > cardinal 3 == "three" +-- > cardinal 10 == "ten" -- > cardinal 11 == "11" cardinal :: Int -> Text cardinal n = case n of + 0 -> "zero" 1 -> "one" 2 -> "two" 3 -> "three" @@ -73,6 +66,7 @@ -- > ordinal 42 == "42nd" ordinal :: Int -> Text ordinal n = case n of + 0 -> "zeroth" 1 -> "first" 2 -> "second" 3 -> "third" @@ -98,6 +92,7 @@ -- > defaultNounPlural "thesis" == "theses" -- -- http://www.paulnoll.com/Books/Clear-English/English-plurals-1.html +-- -- http://en.wikipedia.org/wiki/English_plural defaultNounPlural :: Text -> Text defaultNounPlural x @@ -116,7 +111,7 @@ -- | Heuristics for 3rd person singular and past participle -- for an unknown regular verb. Doubling of final consonants --- can be handled via a table of (partially) irrefular verbs. +-- can be handled via a table of (partially) irregular verbs. -- -- > defaultVerbStuff "walk" == ("walks", "walked") -- > defaultVerbStuff "push" == ("pushes", "pushed") @@ -165,7 +160,7 @@ indefiniteDet t = if wantsAn t then "an" else "a" -- | True if the indefinite determiner for a word would normally be --- 'an' as opposed to 'a'. +-- \'an\' as opposed to \'a\'. wantsAn :: Text -> Bool wantsAn t_ = if startsWithAcronym t_ @@ -189,7 +184,7 @@ -- > wantsAn "x-ray" == False -- > acronymWantsAn "x-ray" == True -- --- Note that this won't do the right thing for words like @"SCUBA"@. +-- Note that this won't do the right thing for words like \"SCUBA\". -- You really have to reserve it for those separate-letter acronyms. acronymWantsAn :: Text -> Bool acronymWantsAn (T.toLower -> t) = @@ -215,7 +210,7 @@ looksLikeAcronym "" = False looksLikeAcronym x = T.all isUpper (if T.length x > 1 then T.drop 1 x else x) --- | True if the first word (separating on either - or space) +-- | True if the first word (separating on either hyphen or space) -- looks like an acronym. startsWithAcronym :: Text -> Bool startsWithAcronym = @@ -228,7 +223,7 @@ -- ** Sounds -- --------------------------------------------------------------------- --- | Ends with a sh sound. +-- | Ends with a \'sh\' sound. hasSibilantSuffix :: Text -> Bool hasSibilantSuffix x = any (`T.isSuffixOf` x) ["x","s","ch","sh","z","j"] @@ -236,7 +231,7 @@ hasSemivowelPrefix :: Text -> Bool hasSemivowelPrefix ls = any (`T.isPrefixOf` ls) ["y","w","eu","ewe"] --- | Starts with a vowel-y U sound +-- | Starts with a vowel-y \'U\' sound hasVowel_U_Prefix :: Text -> Bool hasVowel_U_Prefix t = case T.unpack t of @@ -245,12 +240,12 @@ ('u':c:v:_) -> not (isConsonant c && isVowel v) _ -> False --- | Last two letters are a consonant and 'y'. +-- | Last two letters are a consonant and \'y\'. hasCySuffix :: Text -> Bool hasCySuffix (T.unpack . tTakeEnd 2 -> [x, 'y']) = isConsonant x hasCySuffix _ = False --- | Last two letters are a consonant and 'o'. +-- | Last two letters are a consonant and \'o\'. hasCoSuffix :: Text -> Bool hasCoSuffix (T.unpack . tTakeEnd 2 -> [x, 'o']) = isConsonant x hasCoSuffix _ = False @@ -265,8 +260,8 @@ -- > isLetterWithInitialVowelSound 'r' == True -- > isLetterWithInitialVowelSound 'k' == False -- --- (In the above, @'r'@ is pronounced @"are"@, but @'k'@ is pronounced --- @"kay"@.) +-- (In the above, \'r\' is pronounced \"are\", but \'k\' is pronounced +-- \"kay\".) isLetterWithInitialVowelSound :: Char -> Bool isLetterWithInitialVowelSound = (`elem` ("aeiofhlmnrsx" :: String)) . toLower diff --git a/NLP/Minimorph/Number.hs b/NLP/Minimorph/Number.hs deleted file mode 100644 --- a/NLP/Minimorph/Number.hs +++ /dev/null @@ -1,17 +0,0 @@ -module NLP.Minimorph.Number where - -import Data.Text ( Text ) - --- | Singular and Plural. -data SingPlu a = SP - { sg :: a - , pl :: a - } - deriving (Show, Eq) - -data Number = Singular | Plural - deriving (Eq, Show) - -fromSP :: Number -> SingPlu a -> a -fromSP Singular = sg -fromSP Plural = pl diff --git a/NLP/Minimorph/Util.hs b/NLP/Minimorph/Util.hs --- a/NLP/Minimorph/Util.hs +++ b/NLP/Minimorph/Util.hs @@ -1,12 +1,5 @@ {-# LANGUAGE OverloadedStrings #-} --- | Module : NLP.Minimorph.Util --- Copyright : 2012 Eric Kow (Computational Linguistics Ltd.) --- License : BSD3 --- Maintainer : eric.kow@gmail.com --- Stability : experimental --- Portability : portable --- --- Text utility functions. +-- | Text utility functions. module NLP.Minimorph.Util ( tTakeEnd, tDropEnd, (<>), (<+>), tshow ) where @@ -31,6 +24,6 @@ | T.null t2 = t1 | otherwise = t1 <> " " <> t2 --- | Show a value in Text format. +-- | Show a value in `Text` format. tshow :: Show a => a -> Text tshow = T.pack . show diff --git a/minimorph.cabal b/minimorph.cabal --- a/minimorph.cabal +++ b/minimorph.cabal @@ -1,8 +1,11 @@ --- Initial minimorph.cabal generated by cabal init. For further --- documentation, see http://haskell.org/cabal/users-guide/ - name: minimorph -version: 0.1.6.1 +-- The package version. See the Haskell package versioning policy (PVP) +-- for standards guiding when and how versions should be incremented. +-- http://www.haskell.org/haskellwiki/Package_versioning_policy +-- PVP summary: +-+------- breaking API changes +-- | | +----- minor or non-breaking API additions +-- | | | +--- code changes with no API change +version: 0.2.0.0 synopsis: English spelling functions with an emphasis on simplicity. description: A set of simplistic functions capturing the more regular parts of English spelling (for generation, not parsing). @@ -21,10 +24,9 @@ license-file: LICENSE author: Eric Kow maintainer: Mikolaj Konarski --- copyright: category: Natural Language Processing build-type: Simple -cabal-version: >=1.8 +cabal-version: >=1.10 source-repository head type: git @@ -32,12 +34,23 @@ library exposed-modules: NLP.Minimorph.English - NLP.Minimorph.Number NLP.Minimorph.Util - -- other-modules: + build-depends: base < 5 , text + default-language: Haskell2010 + if impl(ghc >= 8.0) + { + default-extensions: MonoLocalBinds, ScopedTypeVariables, OverloadedStrings + BangPatterns, RecordWildCards, NamedFieldPuns, MultiWayIf, + LambdaCase, DefaultSignatures, InstanceSigs, + MonadFailDesugaring, StrictData, CPP + ghc-options: -Wall -Wcompat -Worphans -Wincomplete-uni-patterns -Wincomplete-record-updates -Wmissing-home-modules -Widentities -Wredundant-constraints + ghc-options: -Wall-missed-specialisations + ghc-options: -fno-ignore-asserts -fexpose-all-unfoldings -fspecialise-aggressively + } + test-suite test-minimorph type: exitcode-stdio-1.0 main-is: test-minimorph.hs @@ -49,3 +62,14 @@ , test-framework , test-framework-hunit , text + default-language: Haskell2010 + if impl(ghc >= 8.0) + { + default-extensions: MonoLocalBinds, ScopedTypeVariables, OverloadedStrings + BangPatterns, RecordWildCards, NamedFieldPuns, MultiWayIf, + LambdaCase, DefaultSignatures, InstanceSigs, + MonadFailDesugaring, StrictData, CPP + ghc-options: -Wall -Wcompat -Worphans -Wincomplete-uni-patterns -Wincomplete-record-updates -Wmissing-home-modules -Widentities -Wredundant-constraints + ghc-options: -Wall-missed-specialisations + ghc-options: -fno-ignore-asserts -fexpose-all-unfoldings -fspecialise-aggressively + } diff --git a/test/NLP/Minimorph/EnglishTest.hs b/test/NLP/Minimorph/EnglishTest.hs --- a/test/NLP/Minimorph/EnglishTest.hs +++ b/test/NLP/Minimorph/EnglishTest.hs @@ -2,12 +2,12 @@ module NLP.Minimorph.EnglishTest where -import Data.Text ( Text ) +import Data.Text (Text) import qualified Data.Text as T -import Test.HUnit -import Test.Framework.Providers.HUnit import Test.Framework +import Test.Framework.Providers.HUnit +import Test.HUnit import NLP.Minimorph.English @@ -29,11 +29,10 @@ t_defaultVerbStuff = testGroup "defaultVerbStuff" (map tc verbs) where - tc v@(pl, sg3, pastP) = testCase summary $ + tc (pl, sg3, pastP) = testCase summary $ assertEqual summary (sg3, pastP) (defaultVerbStuff pl) where summary = T.unpack $ T.concat [ pl, " (", sg3, ", ", pastP, ")" ] - v2 = defaultVerbStuff pl t_defaultNounPlural :: Test.Framework.Test t_defaultNounPlural = diff --git a/test/test-minimorph.hs b/test/test-minimorph.hs --- a/test/test-minimorph.hs +++ b/test/test-minimorph.hs @@ -1,10 +1,8 @@ -import Test.HUnit -import Test.Framework.Providers.HUnit import Test.Framework import qualified NLP.Minimorph.EnglishTest main :: IO () -main = defaultMain +main = defaultMain [ NLP.Minimorph.EnglishTest.suite ]