diff --git a/README.md b/README.md --- a/README.md +++ b/README.md @@ -1,31 +1,38 @@ # haskell-conll [![Build Status](https://travis-ci.com/mgajda/haskell-conll.svg?token=DtSUtnf37TUDy1p4x6fT&branch=master)](https://travis-ci.com/mgajda/haskell-conll) +![Hackage](https://img.shields.io/hackage/v/api-monobank.svg) -Set of Basic Types and Primitives that are used for NLP software such as -- CoreNLP -- SyntaxNet +This library package provides set of foundational types and primitives that can be used in NLP software. +Such as: -and can be extended for usage with other software. +- [CoreNLP](https://github.com/mgajda/syntaxnet-haskell) +- [SyntaxNet](https://ai.googleblog.com/2016/05/announcing-syntaxnet-worlds-most.html) +and can be extended for usage with other software. + If you're working on custom solutions you can use it also. -# TreeBanks +# Treebank -Library supports +Library provides support for -- [Penn Treebank Tag-set](http://www.comp.leeds.ac.uk/amalgam/tagsets/upenn.html) -- [Universal Dependencies](http://universaldependencies.org/en/dep/) +- English / [Penn Treebank Tag-set](http://www.comp.leeds.ac.uk/amalgam/tagsets/upenn.html) +- English / [Universal Dependencies](http://universaldependencies.org/en/dep/) # Used -This library is used in +This library is used in - [SyntaxNet Haskell bindings](https://github.com/mgajda/syntaxnet-haskell/) -- [CoreNLP Haskell bindings](https://github.com/mgajda/corenlp-haskell/) +- [CoreNLP Haskell bindings](https://github.com/mgajda/corenlp-haskell/) +# Maintainer + +[Sergey Bushnyak](https://github.com/sigrlami) + # Acknowledgements -- [Michał J. Gajda](https://github.com/mgajda) +- [Michał J. Gajda](https://github.com/mgajda) - initial idea +- [Sergey Bushnyak](https://github.com/sigrlami) - core maintainer - [Alejandro Duran Pallares](https://github.com/vwwv) -- [Sergey Bushnyak](https://github.com/sigrlami) diff --git a/haskell-conll.cabal b/haskell-conll.cabal --- a/haskell-conll.cabal +++ b/haskell-conll.cabal @@ -1,74 +1,55 @@ +cabal-version: 1.12 + +-- This file has been generated from package.yaml by hpack version 0.31.2. +-- +-- see: https://github.com/sol/hpack +-- +-- hash: b21428dd633a54dcad8df997e5dd388761faf2c2fb045e8f4e85db3fec48f6ca + name: haskell-conll -version: 0.1.0.1 +version: 0.1.0.2 synopsis: Core Types for NLP description: Provides core types to work with CoreNLP, SyntaxNet. Handling CoNLL format and Syntax Trees. -homepage: https://github.com/mgajda/haskell-conll#readme -license: BSD3 +homepage: https://github.com/sigrlami/haskell-conll#readme +bug-reports: https://github.com/sigrlami/haskell-conll/issues +license: MIT license-file: LICENSE -author: Sergey Bushnyak, Alejandro Duran-Pallares, Michal Gajda +author: Michal Gajda, + Sergey Bushnyak, + Alejandro Duran-Pallares maintainer: sergey.bushnyak@sigrlami.eu -copyright: Copyright: (c) 2017 Sergey Bushnyak, Michal Gajda, Alejandro Duran-Pallares +copyright: Copyright: (c) 2017-2019 Michal Gajda, + Sergey Bushnyak, + Alejandro Duran-Pallares category: Text build-type: Simple -extra-source-files: README.md -cabal-version: >=1.10 - -library - hs-source-dirs: src - exposed-modules: Data.ConllToken - , Data.ParsedSentence - , Data.SyntaxTree - , Data.TagLabel - , Model.NerCoreNlp - , Model.PennTreebank - , Model.UniversalTreebank - - build-depends: base >= 4.7 && < 5 - , containers - , lens - , pretty-show - , protolude - , split - , text - - default-language: Haskell2010 - - default-extensions: NoImplicitPrelude - , BangPatterns - , BinaryLiterals - , ConstraintKinds - , DataKinds - , DefaultSignatures - , DeriveFunctor - , DeriveGeneric - , DisambiguateRecordFields - , DuplicateRecordFields - , EmptyDataDecls - , FlexibleContexts - , FlexibleInstances - , FunctionalDependencies - , GADTs - , GeneralizedNewtypeDeriving - , LambdaCase - , MonadComprehensions - , MultiParamTypeClasses - , MultiWayIf - , OverloadedStrings - , PartialTypeSignatures - , PatternSynonyms - , RankNTypes - , RecordWildCards - , RecursiveDo - , ScopedTypeVariables - , StandaloneDeriving - , TemplateHaskell - , TupleSections - , TypeFamilies - , TypeInType - , TypeOperators - , TypeSynonymInstances - +extra-source-files: + README.md source-repository head - type: git + type: git location: https://github.com/sigrlami/haskell-conll + +library + hs-source-dirs: + src + exposed-modules: + Data.ConllToken + Data.ParsedSentence + Data.SyntaxTree + Data.TagLabel + Model.NerCoreNlp + Model.PennTreebank + Model.UniversalTreebank + other-modules: + Paths_haskell_conll + build-depends: + base >=4.7 && <5 + , containers + , lens + , pretty-show + , protolude + , split + , text + default-language: Haskell2010 + default-extensions: NoImplicitPrelude DefaultSignatures OverloadedStrings TemplateHaskell TypeOperators diff --git a/src/Data/ConllToken.hs b/src/Data/ConllToken.hs --- a/src/Data/ConllToken.hs +++ b/src/Data/ConllToken.hs @@ -1,3 +1,6 @@ +{-# LANGUAGE DeriveGeneric #-} +{-# LANGUAGE DeriveFunctor #-} + module Data.ConllToken where import Protolude @@ -6,31 +9,33 @@ -------------------------------------------------------------------------------- -- | Basic data type to work witn CoNLL data format --- +-- data ConllToken cpos fpos ger feats lemma = ConllToken - { _tnId :: Int -- ^ Index number - , _tnWord :: Text -- ^ Parsed word or punctuation symbol - , _tnLemma :: lemma -- ^ Lemma or stem - , _tnPosCG :: cpos -- ^ Part-of-Speech (POS) coarse-grained (PRON, VERB, DET, NOUN, etc) - , _tnPosFG :: fpos -- ^ Part-of-Speech (POS) fine-grained (PRP, VBD, DT, NN etc.) - , _tnFeats :: feats -- ^ Unordered set of syntactic and/or morphological features. - , _tnHead :: Int -- ^ Head of the current token, which is either a value of ID or '0'. - , _tnRel :: ger -- ^ grammatical relationships between different words in the sentence, alined with Head - , _tnHeadProj :: Text -- ^ Projective head of current token. - , _tnRelProj :: Text -- ^ Dependency relation to the PHEAD. + { _tnId :: Int -- ^ Index number + , _tnWord :: Text -- ^ Parsed word or punctuation symbol + , _tnLemma :: lemma -- ^ Lemma or stem + , _tnPosCG :: cpos -- ^ Part-of-Speech (POS) coarse-grained (PRON, VERB, DET, NOUN, etc) + , _tnPosFG :: fpos -- ^ Part-of-Speech (POS) fine-grained (PRP, VBD, DT, NN etc.) + , _tnFeats :: feats -- ^ Unordered set of syntactic and/or morphological features. + , _tnHead :: Int -- ^ Head of the current token, which is either a value of ID or '0'. + , _tnRel :: ger -- ^ grammatical relationships between different words in the sentence, alined with Head + , _tnHeadProj :: Text -- ^ Projective head of current token. + , _tnRelProj :: Text -- ^ Dependency relation to the PHEAD. } deriving (Show, Read, Eq, Ord, Generic, Functor) -- | Describes typical errors when parsing CoNLL from text data -- Contains next filelds: Reason, LineNumber, Culprit -data SyntaxErrorCoNLL = UnkonwnPosTag Int Text - | UnkwownRelTag Int Text - | CoulNotParseInteger Int Text - | InvalidNumberOfElementsOnLine Int Text - | TheresNoRoot - deriving(Show,Read,Eq,Ord) +data SyntaxErrorCoNLL = + UnkonwnPosTag Int Text + | UnkwownRelTag Int Text + | CoulNotParseInteger Int Text + | InvalidNumberOfElementsOnLine Int Text + | TheresNoRoot + deriving(Show, Read, Eq, Ord) -$(makeLenses ''ConllToken ) -$(makePrisms ''SyntaxErrorCoNLL) +-- There are multiple conll formats CoNLL, CoNLL-X, CoNLL-U, CoNLL-ST +$(makeLenses ''ConllToken ) +$(makePrisms ''SyntaxErrorCoNLL) diff --git a/src/Data/ParsedSentence.hs b/src/Data/ParsedSentence.hs --- a/src/Data/ParsedSentence.hs +++ b/src/Data/ParsedSentence.hs @@ -1,22 +1,24 @@ +{-# LANGUAGE DeriveGeneric #-} +{-# LANGUAGE RecordWildCards #-} + module Data.ParsedSentence ( module Data.ParsedSentence , module Data.SyntaxTree ) where -import Protolude -import Data.SyntaxTree import Data.Map +import Data.SyntaxTree +import Protolude -------------------------------------------------------------------------------- --- | +-- | All relations having as "head" the given index data ParsedSentence cpos fpos ger feats lemma = ParsedSentence { _rootNode :: SyntaxtTree cpos fpos ger feats lemma , _indexToNode :: Map Int (SyntaxtTree cpos fpos ger feats lemma) - , _headToNode :: Map Int [SyntaxtTree cpos fpos ger feats lemma] - -- ^ All relations having as "head" the given index - } deriving(Show,Read,Eq,Generic) + , _headToNode :: Map Int [SyntaxtTree cpos fpos ger feats lemma] + } deriving(Show, Read, Eq, Generic) -- | sentenceFromRootNode :: SyntaxtTree cpos fpos ger feats lemma @@ -24,7 +26,7 @@ sentenceFromRootNode _rootNode = ParsedSentence{..} where - _indexToNode = fromList [ (_tnId $ rootLabel node, node ) | node <- everyNode] + _indexToNode = fromList [ (_tnId $ rootLabel node, node ) | node <- everyNode] _headToNode = fromListWith (++) [ (_tnHead $ rootLabel node,[node]) | node <- everyNode] everyNode = everyNodeFrom _rootNode everyNodeFrom node@(Node _ children) = node:( everyNodeFrom =<< children) diff --git a/src/Data/SyntaxTree.hs b/src/Data/SyntaxTree.hs --- a/src/Data/SyntaxTree.hs +++ b/src/Data/SyntaxTree.hs @@ -11,27 +11,27 @@ -------------------------------------------------------------------------------- --- | +-- | Wrapper for the tree to describe syntax tree +-- as a structure of connected ConllTokens type SyntaxtTree cpos fpos ger feats lemma = Tree (ConllToken cpos fpos ger feats lemma) -- | -createSyntaxTree :: [ConllToken cpos fpos ger feats lemma] +createSyntaxTree :: [ConllToken cpos fpos ger feats lemma] -> Either SyntaxErrorCoNLL (SyntaxtTree cpos fpos ger feats lemma) createSyntaxTree conllLines = note TheresNoRoot $ listToMaybe (treeFrom 0) where - - indexToNode = fromList [ (_tnId l,l) | l <- conllLines] - - headToNode = fromListWith (++) [ (_tnHead node,[node]) - | node <- conllLines - ] - - treeFrom n = let relations = fromMaybe [] $ lookup n headToNode - in [ Node{ rootLabel = node - , subForest = treeFrom (_tnId node) - } - | node <- relations - ] + indexToNode = fromList [ (_tnId l,l) + | l <- conllLines + ] + headToNode = fromListWith (++) [ (_tnHead node,[node]) + | node <- conllLines + ] + treeFrom n = let relations = fromMaybe [] $ lookup n headToNode + in [ Node { rootLabel = node + , subForest = treeFrom (_tnId node) + } + | node <- relations + ] diff --git a/src/Data/TagLabel.hs b/src/Data/TagLabel.hs --- a/src/Data/TagLabel.hs +++ b/src/Data/TagLabel.hs @@ -1,15 +1,25 @@ -module Data.TagLabel ( module GHC.Generics - , TagLabel(..) - , SpelledAs(..) - , fromLabelText - , toLabelText - ) where +{-# LANGUAGE KindSignatures #-} +{-# LANGUAGE DataKinds #-} +{-# LANGUAGE TypeSynonymInstances #-} +{-# LANGUAGE FlexibleInstances #-} +{-# LANGUAGE FlexibleContexts #-} +{-# LANGUAGE DeriveGeneric #-} +{-# LANGUAGE ScopedTypeVariables #-} +module Data.TagLabel + ( module GHC.Generics + , TagLabel(..) + , SpelledAs(..) + , fromLabelText + , toLabelText + ) where + import Data.Char import Data.Map import GHC.Generics (Generic,Rep(..),(:+:)(..)) import GHC.TypeLits -import Protolude hiding(fromLabel) +import Protolude hiding (fromLabel) +import Protolude.Error import qualified Data.Text as T import qualified GHC.Generics as G @@ -29,10 +39,10 @@ fromLabelText :: (TagLabel label) => Text -> Maybe label fromLabelText t = lookup t labelMap --- | +-- | toLabelText :: (TagLabel label) => label -> Text toLabelText l = - fromMaybe (error "imposible branch on Data.Label") + fromMaybe (error "imposible branch on Data.Label") $ lookup l reverseLabelMap -------------------------------------------------------------------------------- @@ -46,31 +56,31 @@ gLabelMap _ = second M1 <$> gLabelMap (Proxy :: Proxy f) --- | The labels of sum types are the labels of its parts +-- | The labels of sum types are the labels of its parts instance (GLabel f, GLabel g) => GLabel ( f :+: g) where gLabelMap _ = ( second L1 <$> gLabelMap (Proxy :: Proxy f)) ++ ( second R1 <$> gLabelMap (Proxy :: Proxy g)) -- | The label of a constructor without arguments is the constructor's name itself instance (KnownSymbol constructor) => GLabel (C1 ('MetaCons constructor a b) U1) where - gLabelMap _ = [( convert . toSL $ symbolVal (Proxy :: Proxy constructor) , M1 U1)] + gLabelMap _ = [( convert . toSL $ symbolVal (Proxy :: Proxy constructor) , M1 U1)] where convert s = if T.any isLower s then T.toLower s else s --- | The labels with a constructor with a single argument, are the labels of its argument +-- | The labels with a constructor with a single argument, are the labels of its argument instance (TagLabel labelType) => GLabel (C1 meta1 (S1 meta2 (K1 meta3 labelType))) where gLabelMap _ = assocs $ (M1 . M1 . K1) <$> labelMap --------------------------------------------------------------------------------------- -- Standard labels --- | The labels of a sum type, is the cartesian sum of its labels -instance (TagLabel l1, TagLabel l2) => TagLabel (Either l1 l2) +-- | The labels of a sum type, is the cartesian sum of its labels +instance (TagLabel l1, TagLabel l2) => TagLabel (Either l1 l2) -- | The label of a `SpelledAs x` is x, where x is a type level literal string (a "Symbol"). instance (KnownSymbol symbol) => TagLabel (SpelledAs symbol) where labelMap = fromList [(toSL $ symbolVal (Proxy :: Proxy symbol), SymbolProxy)] -data SpelledAs (s :: Symbol) = SymbolProxy deriving(Show,Read,Eq,Ord,Generic) - - +data SpelledAs (s :: Symbol) = + SymbolProxy + deriving(Show, Read, Eq, Ord, Generic) diff --git a/src/Model/NerCoreNlp.hs b/src/Model/NerCoreNlp.hs --- a/src/Model/NerCoreNlp.hs +++ b/src/Model/NerCoreNlp.hs @@ -1,10 +1,12 @@ {-# LANGUAGE DeriveAnyClass #-} +{-# LANGUAGE DeriveGeneric #-} -module Model.NerCoreNlp where +module Model.NerCoreNlp where -import Protolude import Data.TagLabel +import Protolude +-------------------------------------------------------------------------------- {- | Named Entity Recognition (NER) tag set used by `CoreNLP`. @@ -14,24 +16,22 @@ -- | CoreNLP's NER tag set: -data NER = O - | CARDINAL - | DATE - | DURATION - | FACILITY - | GPE - | LOCATION - | MEASURE - | MISC - | MONEY - | NUMBER - | ORDINAL - | ORGANIZATION - | PERCENT - | PERSON - | SET - | TIME - deriving(Show,Eq,Read,Ord,Generic,TagLabel) - - - +data NER = + O + | CARDINAL + | DATE + | DURATION + | FACILITY + | GPE + | LOCATION + | MEASURE + | MISC + | MONEY + | NUMBER + | ORDINAL + | ORGANIZATION + | PERCENT + | PERSON + | SET + | TIME + deriving(Show, Eq, Read, Ord, Generic, TagLabel) diff --git a/src/Model/PennTreebank.hs b/src/Model/PennTreebank.hs --- a/src/Model/PennTreebank.hs +++ b/src/Model/PennTreebank.hs @@ -1,65 +1,63 @@ +{-# LANGUAGE DataKinds #-} {-# LANGUAGE DeriveAnyClass #-} +{-# LANGUAGE DeriveGeneric #-} -module Model.PennTreebank where +module Model.PennTreebank where -import Protolude import Data.TagLabel +import Protolude -------------------------------------------------------------------------------- --- | (Penn) Treebank Tag-set +-- | (Penn) Treebank Tag-set -- For documentation see http://www.comp.leeds.ac.uk/amalgam/tagsets/upenn.html -data POS = CC - | CD - | ClosePar (SpelledAs ")" ) - | Colon (SpelledAs ":" ) - | Coma (SpelledAs ",") - | Dash (SpelledAs "--" ) - | Dollar (SpelledAs "$" ) - | DT - | EX - | FW - | IN - | JJ - | JJR - | JJS - | LRB_ (SpelledAs "-LRB-") - | LS - | MD - | NN - | NNP - | NNPS - | NNS - | OpenPar (SpelledAs "(" ) - | PDT - | POS - | PRP - | PRP_Dollar (SpelledAs "PRP$") - | Quotes (SpelledAs "''") - | Quotes2 (SpelledAs "``") - | RB - | RBR - | RBS - | RP - | RRB_ (SpelledAs "-RRB-") - | SYM - | Terminator (SpelledAs "." ) - | TO - | UH - | VB - | VBD - | VBG - | VBN - | VBP - | VBZ - | WDT - | WP - | WP_Dollar (SpelledAs "WP$") - | WRB - | UNKNOWN -- when some unknown string used - deriving(Show,Read,Eq,Ord,Generic,TagLabel) - - - - - +data POS = + CC + | CD + | ClosePar (SpelledAs ")" ) + | Colon (SpelledAs ":" ) + | Coma (SpelledAs ",") + | Dash (SpelledAs "--" ) + | Dollar (SpelledAs "$" ) + | DT + | EX + | FW + | IN + | JJ + | JJR + | JJS + | LRB_ (SpelledAs "-LRB-") + | LS + | MD + | NN + | NNP + | NNPS + | NNS + | OpenPar (SpelledAs "(" ) + | PDT + | POS + | PRP + | PRP_Dollar (SpelledAs "PRP$") + | Quotes (SpelledAs "''") + | Quotes2 (SpelledAs "``") + | RB + | RBR + | RBS + | RP + | RRB_ (SpelledAs "-RRB-") + | SYM + | Terminator (SpelledAs "." ) + | TO + | UH + | VB + | VBD + | VBG + | VBN + | VBP + | VBZ + | WDT + | WP + | WP_Dollar (SpelledAs "WP$") + | WRB + | UNKNOWN -- when some unknown string used + deriving(Show,Read,Eq,Ord,Generic,TagLabel) diff --git a/src/Model/UniversalTreebank.hs b/src/Model/UniversalTreebank.hs --- a/src/Model/UniversalTreebank.hs +++ b/src/Model/UniversalTreebank.hs @@ -1,62 +1,64 @@ +{-# LANGUAGE DataKinds #-} {-# LANGUAGE DeriveAnyClass #-} +{-# LANGUAGE DeriveGeneric #-} -module Model.UniversalTreebank where +module Model.UniversalTreebank where -import Protolude import Data.TagLabel +import Protolude -------------------------------------------------------------------------------- -- | Tags from Universal Dependencies project -- see `http://universaldependencies.org/en/dep/` -data REL = Acl -- ^ clausal modifier of noun - | Acl_Relcl (SpelledAs "acl:relcl" ) -- ^ relative clause modifier - | Advcl -- ^ adverbial clause modifier - | Advmod -- ^ adverbial modifier - | Amod -- ^ adjectival modifier - | Appos -- ^ appositional modifier - | Aux -- ^ auxiliary - | Auxpass -- ^ passive auxiliary - | Case -- ^ case marking - | Cc -- ^ coordination - | Cc_Preconj (SpelledAs "cc:preconj" ) -- ^ preconjunct - | Ccomp -- ^ clausal complement - | Compound -- ^ compound - | Compound_Pr(SpelledAs "compound:prt") -- ^ phrasal verb particle - | Conj -- ^ conjunct - | Cop -- ^ copula - | Csubj -- ^ clausal subject - | Csubjpass -- ^ clausal passive subject - | Dep -- ^ dependent - | Det -- ^ determiner - | Det_Predet (SpelledAs "det:predet" ) -- ^ predeterminer - | Discourse -- ^ discourse element - | Dislocated -- ^ dislocated elements - | Dobj -- ^ direct object - | Expl -- ^ expletive - | Fixed (SpelledAs "mwe") -- ^ multi-word expression - | Flat -- ^ name - | Foreign -- ^ foreign words - | Goeswith -- ^ goes with - | Iobj -- ^ indirect object - | List -- ^ list - | Mark -- ^ marker - | Neg -- ^ negation modifier - | Nmod -- ^ nominal modifier - | Nmod_npmod (SpelledAs "nmod:npmod" ) -- ^ noun phrase as adverbial modifier - | Nmod_poss (SpelledAs "nmod:poss" ) -- ^ possessive nominal modifier - | Nmod_tmod (SpelledAs "nmod:tmod" ) -- ^ temporal modifier - | Nsubj -- ^ nominal subject - | Nsubjpass -- ^ passive nominal subject - | Nummod -- ^ numeric modifier - | Orphan -- ^ remnant in ellipsis - | Parataxis -- ^ parataxis - | Punct -- ^ punctuation - | Reparandum -- ^ overridden disfluency - | ROOT -- ^ root - | Root -- ^ root (not main) - | Vocative -- ^ vocative - | Xcomp -- ^ open clausal complement - | Unknown -- ^ unknown - deriving(Show,Read,Eq,Ord,Generic,TagLabel) - +data REL = + Acl -- ^ clausal modifier of noun + | Acl_Relcl (SpelledAs "acl:relcl" ) -- ^ relative clause modifier + | Advcl -- ^ adverbial clause modifier + | Advmod -- ^ adverbial modifier + | Amod -- ^ adjectival modifier + | Appos -- ^ appositional modifier + | Aux -- ^ auxiliary + | Auxpass -- ^ passive auxiliary + | Case -- ^ case marking + | Cc -- ^ coordination + | Cc_Preconj (SpelledAs "cc:preconj" ) -- ^ preconjunct + | Ccomp -- ^ clausal complement + | Compound -- ^ compound + | Compound_Pr(SpelledAs "compound:prt") -- ^ phrasal verb particle + | Conj -- ^ conjunct + | Cop -- ^ copula + | Csubj -- ^ clausal subject + | Csubjpass -- ^ clausal passive subject + | Dep -- ^ dependent + | Det -- ^ determiner + | Det_Predet (SpelledAs "det:predet" ) -- ^ predeterminer + | Discourse -- ^ discourse element + | Dislocated -- ^ dislocated elements + | Dobj -- ^ direct object + | Expl -- ^ expletive + | Fixed (SpelledAs "mwe") -- ^ multi-word expression + | Flat -- ^ name + | Foreign -- ^ foreign words + | Goeswith -- ^ goes with + | Iobj -- ^ indirect object + | List -- ^ list + | Mark -- ^ marker + | Neg -- ^ negation modifier + | Nmod -- ^ nominal modifier + | Nmod_npmod (SpelledAs "nmod:npmod" ) -- ^ noun phrase as adverbial modifier + | Nmod_poss (SpelledAs "nmod:poss" ) -- ^ possessive nominal modifier + | Nmod_tmod (SpelledAs "nmod:tmod" ) -- ^ temporal modifier + | Nsubj -- ^ nominal subject + | Nsubjpass -- ^ passive nominal subject + | Nummod -- ^ numeric modifier + | Orphan -- ^ remnant in ellipsis + | Parataxis -- ^ parataxis + | Punct -- ^ punctuation + | Reparandum -- ^ overridden disfluency + | ROOT -- ^ root + | Root -- ^ root (not main) + | Vocative -- ^ vocative + | Xcomp -- ^ open clausal complement + | Unknown -- ^ unknown + deriving(Show, Read, Eq, Ord, Generic, TagLabel)