Difference: HfstNorthernSami (1 vs. 15)

Revision 152014-02-10 - ErikAxelson

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Line: 8 to 8
 

Introduction

Changed:
<
<
We've tested the performance of HfstLexc and HfstTwolC (of the version 2 branch) by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
>
>
We've tested the performance of HfstLexcWrapper and HfstTwolC (of the version 2 branch) by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
  The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, common nouns, proper nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).

Revision 142011-01-03 - TommiPirinen

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Added:
>
>
Tip, idea Current versions of giellatekno projects support HFST compiling out of the box. Just say make GTLANG=sme hfst to build Northern Smi with HFST tools. To understand the details under the hood, you can still read this documentation
 

Introduction

Changed:
<
<
We tested the performance of HfstLexc2Fst and HfstTwolC by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
>
>
We've tested the performance of HfstLexc and HfstTwolC (of the version 2 branch) by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
  The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, common nouns, proper nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).
Line: 22 to 24
  We need only make one modification, to make Xerox lexc-files compile undex hfst-lexc.
Changed:
<
<
Xerox lexc by default adds a word-boundary # at the beginning and end of all words. This is not done by hfst-lexc, so we need to do it explicitly. All lexical forms given by the Northern Smi lexicon end in the lexicon ENDLEX
>
>
Xerox lexc by default adds a word-boundary # at the beginning and end of all words. This was not done by hfst-lexc at that time, so we needed to do it explicitly. All lexical forms given by the Northern Smi lexicon end in the lexicon ENDLEX
 
LEXICON ENDLEX
 @D.NeedNoun.ON@%# ;
Changed:
<
<
We just add word-boundaries
>
>
We just added the word-boundaries
 
LEXICON ENDLEX
 @D.NeedNoun.ON@%# # ; ! This line differs. The first # is the word-boundary symbol.
Line: 40 to 42
  We used the command
Changed:
<
<
time ~/SFST/src/hlexc/src/hfst-lexc --output=sami.hlexc sme-lex.txt abbr-sme-lex.txt acro-sme-lex.txt adj-sme-lex.txt adv-sme-lex.txt conjunction-sme-lex.txt interjection-sme-lex.txt noun-sme-lex.txt numeral-sme-lex.txt particle-sme-lex.txt pp-sme-lex.txt pronoun-sme-lex.txt punct-sme-lex.txt subjunction-sme-lex.txt verb-sme-lex.txt
>
>
time hfst-lexc --output=sami.hlexc sme-lex.txt abbr-sme-lex.txt acro-sme-lex.txt adj-sme-lex.txt adv-sme-lex.txt conjunction-sme-lex.txt interjection-sme-lex.txt noun-sme-lex.txt numeral-sme-lex.txt particle-sme-lex.txt pp-sme-lex.txt pronoun-sme-lex.txt punct-sme-lex.txt subjunction-sme-lex.txt verb-sme-lex.txt
  The size of the binary produced was about 2,0Mb. The timing produced
Line: 153 to 155
 

Downloading

Changed:
<
<
The Sami transducer can be fetched from the Sourceforge download page.
>
>
The Sami transducer can be fetched from the Sourceforge download page, or for more up-to-date versions, directly from giellatekno SVN.
 
<--  
-->
Changed:
<
<
-- MiikkaSilfverberg - 05 Oct 2008
>
>
-- MiikkaSilfverberg
 
META TOPICMOVED by="KristerLinden" date="1212070604" from="KitWiki.HFSTTopicTemplate" to="KitWiki.HfstTopicTemplate"

Revision 132010-12-31 - TommiPirinen

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Line: 6 to 6
 

Introduction

Changed:
<
<
We tested the performance of HfstLexcCompiler and HfstTwolC by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
>
>
We tested the performance of HfstLexc2Fst and HfstTwolC by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
  The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, common nouns, proper nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).

Revision 122010-12-21 - TommiPirinen

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Line: 6 to 6
 

Introduction

Changed:
<
<
We tested the performance of HfstLexC and HfstTwolC by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
>
>
We tested the performance of HfstLexcCompiler and HfstTwolC by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
  The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, common nouns, proper nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).

Revision 112009-11-03 - ErikAxelson

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Line: 14 to 14
  The hfst-lexc and hfst-twolc input-file formats differ from corresponding Xerox formats. Hence lexicon and rules files, which have been developped for Xerox compilers need to be modified for use with hfst-tools. The necessary changes are explained below.
Changed:
<
<
All files were compiled on the CSC server corpus, which has a Intel(R) Xeon(R) 3.00GHz processor and 3652568 kB of memory. The lexicon and grammar were combined into a leical transducer using HfstIntersectingComposition.
>
>
All files were compiled on the CSC server corpus, which has a Intel(R) Xeon(R) 3.00GHz processor and 3652568 kB of memory. The lexicon and grammar were combined into a leical transducer using HfstComposeIntersect.
 

hfst-lexc

Revision 102009-10-13 - ErikAxelson

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Line: 151 to 151
 sys 24.45s
Added:
>
>

Downloading

The Sami transducer can be fetched from the Sourceforge download page.

 

Revision 92009-01-26 - KristerLinden

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Line: 67 to 67
  e7:e e9:e i7:i o7:o o9:o u7:u 7: ! Morphophonemes
Changed:
<
<
č đ ŋ ŧ ! Smi letters.
>
>
&269; đ ŋ ŧ ! Smi letters.
  Č Đ Ŋ Ŧ !

#:0 %^:0

Line: 112 to 112
  č:0 ':č ': h8:0 G8:0 M8:0 N8:0 H8:0
Changed:
<
<
Xerox twolc recognizes flag-diacritics from the lexicon and doesnt intrfere with them in any way. There is no similar automated mechanism in hfst-twolc (at least not yet). Hence flag-diacritics need to be declared as regular diacritics in the grammar. We add a Diacritics section to the grammar
>
>
Xerox twolc recognizes flag-diacritics from the lexicon and does not interfere with them in any way. There is no similar automated mechanism in hfst-twolc (at least not yet). Hence flag-diacritics need to be declared as regular diacritics in the grammar. We add a Diacritics section to the grammar
 
Diacritics

%@C.NeedNoun%@ %@D.NeedNoun.ON%@ %@P.NeedNoun.ON%@ %@U.Cap.Obl%@ %@U.Cap.Opt%@

Revision 82009-01-26 - MiikkaSilfverberg

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Added:
>
>
 

Introduction

Changed:
<
<
We tested the performance of hfst-lexc and hfst-twolc by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
>
>
We tested the performance of HfstLexC and HfstTwolC by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
  The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, common nouns, proper nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).

The twol-grammar consisted of 105 rules.

Changed:
<
<
All files were compiled on the CSC server corpus, which has a Intel(R) Xeon(R) 3.00GHz processor and 3652568 kB of memory.
>
>
The hfst-lexc and hfst-twolc input-file formats differ from corresponding Xerox formats. Hence lexicon and rules files, which have been developped for Xerox compilers need to be modified for use with hfst-tools. The necessary changes are explained below.

All files were compiled on the CSC server corpus, which has a Intel(R) Xeon(R) 3.00GHz processor and 3652568 kB of memory. The lexicon and grammar were combined into a leical transducer using HfstIntersectingComposition.

 

hfst-lexc

Added:
>
>

Modifying the Northern Smi lexicon

We need only make one modification, to make Xerox lexc-files compile undex hfst-lexc.

Xerox lexc by default adds a word-boundary # at the beginning and end of all words. This is not done by hfst-lexc, so we need to do it explicitly. All lexical forms given by the Northern Smi lexicon end in the lexicon ENDLEX

LEXICON ENDLEX
 @D.NeedNoun.ON@%# ;

We just add word-boundaries

LEXICON ENDLEX
 @D.NeedNoun.ON@%# # ; ! This line differs. The first # is the word-boundary symbol.
                       ! The one signifies, that there are no continuation-classes for this
                       ! lexicon.

Compiling the lexicon using hfst-lexc

  We used the command
time ~/SFST/src/hlexc/src/hfst-lexc --output=sami.hlexc sme-lex.txt abbr-sme-lex.txt acro-sme-lex.txt adj-sme-lex.txt adv-sme-lex.txt conjunction-sme-lex.txt interjection-sme-lex.txt noun-sme-lex.txt numeral-sme-lex.txt particle-sme-lex.txt pp-sme-lex.txt pronoun-sme-lex.txt punct-sme-lex.txt subjunction-sme-lex.txt verb-sme-lex.txt
Line: 25 to 50
 

hfst-twolc

Added:
>
>

Modifying the rule-file

More modifications are needed for the two-level rules, than for the lexicon.

The alphabet of the original two-level grammar contains all of the symbols used, but only a few of all pairs

Alphabet
 a b c d e f g h i j k l m n o p q   ! small
 r s t u v w x y z      %-
                           

 A B C D E F G H I J K L M N O P Q   ! capital
 R S T U V W X Y Z     
                       

 e7:e e9:e i7:i o7:o o9:o u7:u 7:  ! Morphophonemes

 &#269; &#273; &#331;  &#359;           ! Smi letters.
 &#268; &#272; &#330;  &#358;           !

 #:0 %^:0


 ':0 %/ &#164;:0 '7:'
                            ! ' is for CnsGrad of the lg:lgg and l'l:ll type
                            ! &#164;:0 prevents ConsGrad in certain words
                            ! '7 is the real apostroph
                            ! # is used to mark both lexicalised and
                            ! derived compounds



 h7:h h8:h g8:g m8:m n8:n   ! the x8 ones are consonants that alternate in
 H7:H H8:H G8:G M8:M N8:N   ! stem-final positions.

 j9:j b9:b d9:d g9:g h9:h k9:k m9:m n9:n p9:p s9:s t9:t z9:z 9: r9:r
 J9:J B9:B D9:D G9:G H9:H K9:K M9:M N9:N P9:P S9:S T9:T Z9:Z 9: R9:R

                            ! The x9 ones are consonants that never alternate.
                            ! The capital J9 etc. do not work.
 X1:0 X2:0 X3:0 X4:0 X5:0 X6:0 X7:0 X8:0 X9:0   ! diacritics
 Q1:0 Q2:0 Q3:0 Q4:0 Q5:0 Q6:0 Q7:0 Q8:0 Q9:0   ! They trigger morphophono-
 Y1:0 Y2:0 Y3:0 Y4:0 Y5:0 Y6:0 Y7:0 Y8:0 Y9:0   ! logical rules
 W1:0 W2:0 W3:0 W4:0 W5:0 W6:0 W7:0 W8:0 W9:0

 %>:0 %>7:%> :0        ! stem-suffix border mark, unvisible for normal use

;
In a hfst-twolc grammar-file, all pairs need to be added in the grammar. Hence we need to add the following pairs to the alphabet.
 
      ! Miikka added, because they are needed 
      ! by hfst-twolc, which has to know all pairs.

        j:i # k:0 t:0 h:0 d:0 ':t ':n ':l 0:s ':j ':v d:n ':k k:v d:t t:d ':d i:e u:o
        i: a:i :0 u:0 e:0 o:0 a:0 i:0 :0 n:0 z:s m:n h:t g:t b:t h:t : :0
        i: s:0 j:0 k:g p:0 f:0 l:0 ':b ':g ':r ':m b:m b:p ':s a:u p:b r:0 t:&#273;
        o:u e:i :&#269; a: a:o a:e i:o i:u g8:0 n8:0 m8:0 p:t g:0 m:0 :0 &#273;:0 &#331;:0 v:0 c:0
        ':f ':p ':z ':c ': ':z g:&#331; g:k z:c &#269;: c:z :a e7: g8:t h8:t m8:n &#359;:0 
        &#269;:0 ':&#269; ': h8:0 G8:0 M8:0 N8:0 H8:0

Xerox twolc recognizes flag-diacritics from the lexicon and doesnt intrfere with them in any way. There is no similar automated mechanism in hfst-twolc (at least not yet). Hence flag-diacritics need to be declared as regular diacritics in the grammar. We add a Diacritics section to the grammar

Diacritics

%@C.NeedNoun%@ %@D.NeedNoun.ON%@ %@P.NeedNoun.ON%@ %@U.Cap.Obl%@ %@U.Cap.Opt%@

;
For now, we also need to add a section Rule-variables, which contains every rule-variable used in the grammar. Omitting it doesn't really cause anything to be compiled incorrectly, but there will be numerous irritating warnings about symbols, which haven't been declared. We might fix this in future versions of hfst-twolc. We add the following between the Diacritics and Sets sections.
Rule-variables

Cx Cy Cz Vx Vy Vz ;

Compiling the rules using hfst-twolc

 We used the command
time htwolc --input sami.twol --output sami.twolc --resolve

Revision 72008-11-25 - MiikkaSilfverberg

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

Revision 62008-11-24 - KristerLinden

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"
Changed:
<
<

HFST: Hfst Tools Used to Compile a Northern Sami Lexical Transducer

>
>

HFST: Utilities for Compiling a Northern Sami Lexical Transducer

 

Introduction

Revision 52008-11-24 - MiikkaSilfverberg

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Hfst Tools Used to Compile a Northern Sami Lexical Transducer

Line: 10 to 10
  The twol-grammar consisted of 105 rules.
Added:
>
>
All files were compiled on the CSC server corpus, which has a Intel(R) Xeon(R) 3.00GHz processor and 3652568 kB of memory.
 

hfst-lexc

We used the command
Line: 29 to 31
  The size of the binary produced was about 7,7Mb. The timing produced
Changed:
<
<
real 7m5.831s user 6m24.951s sys 0m22.362s
>
>
real 3m10.985s user 2m56.345s sys 0m12.346s
 

hfst-compose-intersect

Line: 40 to 42
 
cat sami.twolc | time hfst-compose-intersect --lexicon sami.hlexc
Changed:
<
<
The size of the binary produced was about 2,2Mb. The timing produced
>
>
The size of the binary produced was about 2,7Mb. The timing produced
 
Changed:
<
<
real 710.09s user 649.69s sys 38.53s
>
>
real 276.56s (i.e. ~ 4min36s) user 249.37s (i.e. ~ 4min9s) sys 24.45s
 


Revision 42008-10-07 - MiikkaSilfverberg

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Hfst Tools Used to Compile a Northern Sami Lexical Transducer

Line: 42 to 42
  The size of the binary produced was about 2,2Mb. The timing produced
Changed:
<
<
real 710.09 user 649.69 sys 38.53
>
>
real 710.09s user 649.69s sys 38.53s
 


Revision 32008-10-06 - MiikkaSilfverberg

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Hfst Tools Used to Compile a Northern Sami Lexical Transducer

Line: 6 to 6
  We tested the performance of hfst-lexc and hfst-twolc by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.
Changed:
<
<
The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).
>
>
The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, common nouns, proper nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).
  The twol-grammar consisted of 105 rules.
Line: 15 to 15
 
time ~/SFST/src/hlexc/src/hfst-lexc --output=sami.hlexc sme-lex.txt abbr-sme-lex.txt acro-sme-lex.txt adj-sme-lex.txt adv-sme-lex.txt conjunction-sme-lex.txt interjection-sme-lex.txt noun-sme-lex.txt numeral-sme-lex.txt particle-sme-lex.txt pp-sme-lex.txt pronoun-sme-lex.txt punct-sme-lex.txt subjunction-sme-lex.txt verb-sme-lex.txt
Changed:
<
<
The timing produced
>
>
The size of the binary produced was about 2,0Mb. The timing produced
 
real    2m16.015s
user    1m51.960s
Line: 27 to 27
 
time htwolc --input sami.twol --output sami.twolc --resolve
Changed:
<
<
The timing produced
>
>
The size of the binary produced was about 7,7Mb. The timing produced
 
real    7m5.831s
user    6m24.951s
Line: 40 to 40
 
cat sami.twolc | time hfst-compose-intersect --lexicon sami.hlexc
Changed:
<
<
The timing produced
>
>
The size of the binary produced was about 2,2Mb. The timing produced
 
Changed:
<
<
440.29user 31.89system 8:18.28elapsed
>
>
real 710.09 user 649.69 sys 38.53
 


Revision 22008-10-05 - MiikkaSilfverberg

Line: 1 to 1
 
META TOPICPARENT name="HfstHome"

HFST: Hfst Tools Used to Compile a Northern Sami Lexical Transducer

Line: 15 to 15
 
time ~/SFST/src/hlexc/src/hfst-lexc --output=sami.hlexc sme-lex.txt abbr-sme-lex.txt acro-sme-lex.txt adj-sme-lex.txt adv-sme-lex.txt conjunction-sme-lex.txt interjection-sme-lex.txt noun-sme-lex.txt numeral-sme-lex.txt particle-sme-lex.txt pp-sme-lex.txt pronoun-sme-lex.txt punct-sme-lex.txt subjunction-sme-lex.txt verb-sme-lex.txt
Changed:
<
<
The resulting lexicon-binary was about 2Mb. The timing produced
>
>
The timing produced
 
real    2m16.015s
user    1m51.960s
Line: 27 to 27
 
time htwolc --input sami.twol --output sami.twolc --resolve
Changed:
<
<
The resulting binary was about 7,5Mb. The timing produced
>
>
The timing produced
 
real    7m5.831s
user    6m24.951s
Line: 38 to 38
  We used the command
Changed:
<
<
cat sami.twolc | hfst-compose-intersect --lexicon sami.hlexc
>
>
cat sami.twolc | time hfst-compose-intersect --lexicon sami.hlexc
 
Changed:
<
<
The size of the binary produced was about . The timing produced
>
>
The timing produced
 
Added:
>
>
440.29user 31.89system 8:18.28elapsed
 


Revision 12008-10-05 - MiikkaSilfverberg

Line: 1 to 1
Added:
>
>
META TOPICPARENT name="HfstHome"

HFST: Hfst Tools Used to Compile a Northern Sami Lexical Transducer

Introduction

We tested the performance of hfst-lexc and hfst-twolc by compiling a lexical transducer for northern smi. The open-source lexc-lexicon and twol-grammar for northern smi have been developped in the Smi language technology project and are available for everyone through anonymous SVN. They are distributed under the GPL-license.

The lexicon contained 61295 entries (i.e. stems and different kinds of affixes). The stems represented the word-classes verbs, nouns, adjectives, adverbs, pronouns, numerals, particles, conjunctions, subjunctions and interjections (+ some minor classes like abbreviations).

The twol-grammar consisted of 105 rules.

hfst-lexc

We used the command
time ~/SFST/src/hlexc/src/hfst-lexc --output=sami.hlexc sme-lex.txt abbr-sme-lex.txt acro-sme-lex.txt adj-sme-lex.txt adv-sme-lex.txt conjunction-sme-lex.txt interjection-sme-lex.txt noun-sme-lex.txt numeral-sme-lex.txt particle-sme-lex.txt pp-sme-lex.txt pronoun-sme-lex.txt punct-sme-lex.txt subjunction-sme-lex.txt verb-sme-lex.txt
The resulting lexicon-binary was about 2Mb. The timing produced
real    2m16.015s
user    1m51.960s
sys     0m8.727s

hfst-twolc

We used the command
time htwolc --input sami.twol --output sami.twolc --resolve
The resulting binary was about 7,5Mb. The timing produced
real    7m5.831s
user    6m24.951s
sys     0m22.362s

hfst-compose-intersect

We used the command

cat sami.twolc | hfst-compose-intersect --lexicon sami.hlexc
The size of the binary produced was about . The timing produced


<--  
-->
-- MiikkaSilfverberg - 05 Oct 2008

META TOPICMOVED by="KristerLinden" date="1212070604" from="KitWiki.HFSTTopicTemplate" to="KitWiki.HfstTopicTemplate"
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback