Text.NKJP.Named

Parsing the NKJP named entity layer.

Data types

data Cert

A certainty of an annotator.

Constructors

High

Medium

Low

Instances

Show Cert

data Ptr t

A pointer.

Constructors

Local

Of #id form.

Fields

target :: t

Global

Of loc#id form.

Fields

target :: t

location :: t

Instances

Functor Ptr

Show t => Show (Ptr t)

data Deriv t

A derivation structure.

Constructors

Deriv

Fields

derivType :: t

derivFrom :: t

Instances

Functor Deriv

Show t => Show (Deriv t)

data Para t

A paragraph.

Constructors

Para

Fields

paraID :: t

sentences :: [Sent t]

Instances

Functor Para

Show t => Show (Para t)

data Sent t

A sentence.

Constructors

Sent

Fields

sentID :: t

names :: [NE t]

Instances

Functor Sent

Show t => Show (Sent t)

data NE t

A segment element in a file.

Constructors

NE

Fields

neID :: t

derived :: Maybe (Deriv t)

neType :: t

subType :: Maybe t

orth :: t

base :: Maybe (Either t t)

Left base or Right when.

cert :: Maybe Cert

certComment :: Maybe t

ptrs :: [Ptr t]

Instances

Functor NE

Show t => Show (NE t)

Parsing

parseNamed :: Text -> [Para Text]

Parse textual contents of the ann_named.xml file.

readNamed :: FilePath -> IO [Para Text]

Parse the stand-alone ann_named.xml file.

readCorpus :: FilePath -> IO [(FilePath, Maybe [Para Text])]

Parse all ann_named.xml files from the NCP .tar.gz file.

readTrees :: FilePath -> IO [Forest (Either (NE Text) (Seg Text))]

Parse the NCP .tar.gz corpus, extract all NEs and translate them to the tree form using the mkForest function.

Utilities

mkForest :: Ord t => [Seg t] -> [NE t] -> Forest (Either (NE t) (Seg t))

Make NE forest from a segment list and a list of NEs, both lists corresponding to the same sentence.