rere 0.0.1

data/example/solr/collection1/conf/lang/stopwords_it.txt

@@ -0,0 +1,301 @@
+ | From svn.tartarus.org/snowball/trunk/website/algorithms/italian/stop.txt
+ | This file is distributed under the BSD License.
+ | See http://snowball.tartarus.org/license.php
+ | Also see http://www.opensource.org/licenses/bsd-license.html
+ |  - Encoding was converted to UTF-8.
+ |  - This notice was added.
+
+ | An Italian stop word list. Comments begin with vertical bar. Each stop
+ | word is at the start of a line.
+
+ad             |  a (to) before vowel
+al             |  a + il
+allo           |  a + lo
+ai             |  a + i
+agli           |  a + gli
+all            |  a + l'
+agl            |  a + gl'
+alla           |  a + la
+alle           |  a + le
+con            |  with
+col            |  con + il
+coi            |  con + i (forms collo, cogli etc are now very rare)
+da             |  from
+dal            |  da + il
+dallo          |  da + lo
+dai            |  da + i
+dagli          |  da + gli
+dall           |  da + l'
+dagl           |  da + gll'
+dalla          |  da + la
+dalle          |  da + le
+di             |  of
+del            |  di + il
+dello          |  di + lo
+dei            |  di + i
+degli          |  di + gli
+dell           |  di + l'
+degl           |  di + gl'
+della          |  di + la
+delle          |  di + le
+in             |  in
+nel            |  in + el
+nello          |  in + lo
+nei            |  in + i
+negli          |  in + gli
+nell           |  in + l'
+negl           |  in + gl'
+nella          |  in + la
+nelle          |  in + le
+su             |  on
+sul            |  su + il
+sullo          |  su + lo
+sui            |  su + i
+sugli          |  su + gli
+sull           |  su + l'
+sugl           |  su + gl'
+sulla          |  su + la
+sulle          |  su + le
+per            |  through, by
+tra            |  among
+contro         |  against
+io             |  I
+tu             |  thou
+lui            |  he
+lei            |  she
+noi            |  we
+voi            |  you
+loro           |  they
+mio            |  my
+mia            |
+miei           |
+mie            |
+tuo            |
+tua            |
+tuoi           |  thy
+tue            |
+suo            |
+sua            |
+suoi           |  his, her
+sue            |
+nostro         |  our
+nostra         |
+nostri         |
+nostre         |
+vostro         |  your
+vostra         |
+vostri         |
+vostre         |
+mi             |  me
+ti             |  thee
+ci             |  us, there
+vi             |  you, there
+lo             |  him, the
+la             |  her, the
+li             |  them
+le             |  them, the
+gli            |  to him, the
+ne             |  from there etc
+il             |  the
+un             |  a
+uno            |  a
+una            |  a
+ma             |  but
+ed             |  and
+se             |  if
+perché         |  why, because
+anche          |  also
+come           |  how
+dov            |  where (as dov')
+dove           |  where
+che            |  who, that
+chi            |  who
+cui            |  whom
+non            |  not
+più            |  more
+quale          |  who, that
+quanto         |  how much
+quanti         |
+quanta         |
+quante         |
+quello         |  that
+quelli         |
+quella         |
+quelle         |
+questo         |  this
+questi         |
+questa         |
+queste         |
+si             |  yes
+tutto          |  all
+tutti          |  all
+
+               |  single letter forms:
+
+a              |  at
+c              |  as c' for ce or ci
+e              |  and
+i              |  the
+l              |  as l'
+o              |  or
+
+               | forms of avere, to have (not including the infinitive):
+
+ho
+hai
+ha
+abbiamo
+avete
+hanno
+abbia
+abbiate
+abbiano
+avrò
+avrai
+avrà
+avremo
+avrete
+avranno
+avrei
+avresti
+avrebbe
+avremmo
+avreste
+avrebbero
+avevo
+avevi
+aveva
+avevamo
+avevate
+avevano
+ebbi
+avesti
+ebbe
+avemmo
+aveste
+ebbero
+avessi
+avesse
+avessimo
+avessero
+avendo
+avuto
+avuta
+avuti
+avute
+
+               | forms of essere, to be (not including the infinitive):
+sono
+sei
+è
+siamo
+siete
+sia
+siate
+siano
+sarò
+sarai
+sarà
+saremo
+sarete
+saranno
+sarei
+saresti
+sarebbe
+saremmo
+sareste
+sarebbero
+ero
+eri
+era
+eravamo
+eravate
+erano
+fui
+fosti
+fu
+fummo
+foste
+furono
+fossi
+fosse
+fossimo
+fossero
+essendo
+
+               | forms of fare, to do (not including the infinitive, fa, fat-):
+faccio
+fai
+facciamo
+fanno
+faccia
+facciate
+facciano
+farò
+farai
+farà
+faremo
+farete
+faranno
+farei
+faresti
+farebbe
+faremmo
+fareste
+farebbero
+facevo
+facevi
+faceva
+facevamo
+facevate
+facevano
+feci
+facesti
+fece
+facemmo
+faceste
+fecero
+facessi
+facesse
+facessimo
+facessero
+facendo
+
+               | forms of stare, to be (not including the infinitive):
+sto
+stai
+sta
+stiamo
+stanno
+stia
+stiate
+stiano
+starò
+starai
+starà
+staremo
+starete
+staranno
+starei
+staresti
+starebbe
+staremmo
+stareste
+starebbero
+stavo
+stavi
+stava
+stavamo
+stavate
+stavano
+stetti
+stesti
+stette
+stemmo
+steste
+stettero
+stessi
+stesse
+stessimo
+stessero
+stando

data/example/solr/collection1/conf/lang/stopwords_ja.txt

@@ -0,0 +1,127 @@
+#
+# This file defines a stopword set for Japanese.
+#
+# This set is made up of hand-picked frequent terms from segmented Japanese Wikipedia.
+# Punctuation characters and frequent kanji have mostly been left out.  See LUCENE-3745
+# for frequency lists, etc. that can be useful for making your own set (if desired)
+#
+# Note that there is an overlap between these stopwords and the terms stopped when used
+# in combination with the JapanesePartOfSpeechStopFilter.  When editing this file, note
+# that comments are not allowed on the same line as stopwords.
+#
+# Also note that stopping is done in a case-insensitive manner.  Change your StopFilter
+# configuration if you need case-sensitive stopping.  Lastly, note that stopping is done
+# using the same character width as the entries in this file.  Since this StopFilter is
+# normally done after a CJKWidthFilter in your chain, you would usually want your romaji
+# entries to be in half-width and your kana entries to be in full-width.
+#
+の
+に
+は
+を
+た
+が
+で
+て
+と
+し
+れ
+さ
+ある
+いる
+も
+する
+から
+な
+こと
+として
+い
+や
+れる
+など
+なっ
+ない
+この
+ため
+その
+あっ
+よう
+また
+もの
+という
+あり
+まで
+られ
+なる
+へ
+か
+だ
+これ
+によって
+により
+おり
+より
+による
+ず
+なり
+られる
+において
+ば
+なかっ
+なく
+しかし
+について
+せ
+だっ
+その後
+できる
+それ
+う
+ので
+なお
+のみ
+でき
+き
+つ
+における
+および
+いう
+さらに
+でも
+ら
+たり
+その他
+に関する
+たち
+ます
+ん
+なら
+に対して
+特に
+せる
+及び
+これら
+とき
+では
+にて
+ほか
+ながら
+うち
+そして
+とともに
+ただし
+かつて
+それぞれ
+または
+お
+ほど
+ものの
+に対する
+ほとんど
+と共に
+といった
+です
+とも
+ところ
+ここ
+##### End of file

data/example/solr/collection1/conf/lang/stopwords_lv.txt

@@ -0,0 +1,172 @@
+# Set of Latvian stopwords from A Stemming Algorithm for Latvian, Karlis Kreslins
+# the original list of over 800 forms was refined: 
+#   pronouns, adverbs, interjections were removed
+# 
+# prepositions
+aiz
+ap
+ar
+apakš
+ārpus
+augšpus
+bez
+caur
+dēļ
+gar
+iekš
+iz
+kopš
+labad
+lejpus
+līdz
+no
+otrpus
+pa
+par
+pār
+pēc
+pie
+pirms
+pret
+priekš
+starp
+šaipus
+uz
+viņpus
+virs
+virspus
+zem
+apakšpus
+# Conjunctions
+un
+bet
+jo
+ja
+ka
+lai
+tomēr
+tikko
+turpretī
+arī
+kaut
+gan
+tādēļ
+tā
+ne
+tikvien
+vien
+kā
+ir
+te
+vai
+kamēr
+# Particles
+ar
+diezin
+droši
+diemžēl
+nebūt
+ik
+it
+taču
+nu
+pat
+tiklab
+iekšpus
+nedz
+tik
+nevis
+turpretim
+jeb
+iekam
+iekām
+iekāms
+kolīdz
+līdzko
+tiklīdz
+jebšu
+tālab
+tāpēc
+nekā
+itin
+jā
+jau
+jel
+nē
+nezin
+tad
+tikai
+vis
+tak
+iekams
+vien
+# modal verbs
+būt  
+biju 
+biji
+bija
+bijām
+bijāt
+esmu
+esi
+esam
+esat 
+būšu     
+būsi
+būs
+būsim
+būsiet
+tikt
+tiku
+tiki
+tika
+tikām
+tikāt
+tieku
+tiec
+tiek
+tiekam
+tiekat
+tikšu
+tiks
+tiksim
+tiksiet
+tapt
+tapi
+tapāt
+topat
+tapšu
+tapsi
+taps
+tapsim
+tapsiet
+kļūt
+kļuvu
+kļuvi
+kļuva
+kļuvām
+kļuvāt
+kļūstu
+kļūsti
+kļūst
+kļūstam
+kļūstat
+kļūšu
+kļūsi
+kļūs
+kļūsim
+kļūsiet
+# verbs
+varēt
+varēju
+varējām
+varēšu
+varēsim
+var
+varēji
+varējāt
+varēsi
+varēsiet
+varat
+varēja
+varēs