text_stemmer

routines (
  R1 R2 RV
  initial_morph
  mark_regions
  noun_sfx
  deriv
  verb_sfx
)

externals ( stem )

integers ( pV p1 p2 )

groupings ( v )

stringescapes {}

/* Accented characters */

stringdef a'   '{U+00E1}'  // a-acute
stringdef e'   '{U+00E9}'  // e-acute
stringdef i'   '{U+00ED}'  // i-acute
stringdef o'   '{U+00F3}'  // o-acute
stringdef u'   '{U+00FA}'  // u-acute

define v 'aeiou{a'}{e'}{i'}{o'}{u'}'

define mark_regions as (

    $pV = limit
    $p1 = limit
    $p2 = limit  // defaults

    do (
        gopast v setmark pV
        gopast non-v setmark p1
        gopast v gopast non-v setmark p2
    )
)

define initial_morph as (
  [substring] among (
    'h-' 'n-' 't-' //nAthair -> n-athair, but alone are problematic
    (delete)

    // verbs
    'd{'}'
    (delete)
    'd{'}fh'
    (<- 'f')
    // other contractions
    'm{'}' 'b{'}'
    (delete)

    'sh'
    (<- 's')

    'mb'
    (<- 'b')
    'gc'
    (<- 'c')
    'nd'
    (<- 'd')
    'bhf'
    (<- 'f')
    'ng'
    (<- 'g')
    'bp'
    (<- 'p')
    'ts'
    (<- 's')
    'dt'
    (<- 't')

    // Lenition
    'bh'
    (<- 'b')
    'ch'
    (<- 'c')
    'dh'
    (<- 'd')
    'fh'
    (<- 'f')
    'gh'
    (<- 'g')
    'mh'
    (<- 'm')
    'ph'
    (<- 'p')
    'th'
    (<- 't')
  )
)

backwardmode (

  define RV as $pV <= cursor
  define R1 as $p1 <= cursor
  define R2 as $p2 <= cursor

  define noun_sfx as (
    [substring] among (
      'amh' 'eamh' 'abh' 'eabh'
      'aibh' 'ibh' 'aimh' 'imh'
      'a{i'}ocht' '{i'}ocht' 'a{i'}ochta' '{i'}ochta'
      (R1 delete)
      'ire' 'ir{i'}' 'aire' 'air{i'}'
      (R2 delete)
    )
  )
  define deriv as (
    [substring] among (
      'acht' 'eacht' 'ach' 'each' 'eacht{u'}il' 'eachta' 'acht{u'}il' 'achta'
      (R2 delete)  //siopadóireacht -> siopadóir but not poblacht -> pobl
      'arcacht' 'arcachta{i'}' 'arcachta'
      (<- 'arc') // monarcacht -> monarc
      'gineach' 'gineas' 'ginis'
      (<- 'gin')
      'grafa{i'}och' 'grafa{i'}ocht' 'grafa{i'}ochta' 'grafa{i'}ochta{i'}'
      (<- 'graf')
      'paite' 'patach' 'pataigh' 'patacha'
      (<- 'paite')
      '{o'}ideach' '{o'}ideacha' '{o'}idigh'
      (<- '{o'}id')
    )
  )
  define verb_sfx as (
    [substring] among (
      'imid' 'aimid' '{i'}mid' 'a{i'}mid'
      'faidh' 'fidh'
      (RV delete)
      'ain'
      'eadh' 'adh'
      '{a'}il'
      'tear' 'tar'
      (R1 delete)
    )
  )
)

define stem as (
  do initial_morph
  do mark_regions
  backwards (
    do noun_sfx
    do deriv
    do verb_sfx
  )
)