Ada Stemmer Library

The Ada Stemmer Library is a stemming processor for several natural languages. It is based on the Snowball compiler and stemming algorithms which has been adapted to generate Ada code (Snowball Ada). A stemming algorithm is used in natural language analysis to find the base or root form of a word. Such algorithm is specific to each natural language. The Porter Stemmer algorithm is specific to the English language and will not work for French, Greek or Russian.

The Ada Stemmer Library integrates stemming algorithms for: Arabic, Basque, Catalan, Danish, Dutch, English, Finnish, French, German, Greek, Hindi, Hungarian, Indonesian, Irish, Italian, Lithuanian, Serbian, Nepali, Norwegian, Portuguese, Romanian, Russian, Serbian, Spanish, Swedish, Tamil, Turkish.

Example of stemming:

Language	Word	Stem
French	chienne	chien
French	affectionnait	affection
English	zealously	zealous
English	transitional	transit
Greek	ποσοτητα	ποσοτητ
Greek	μνημειωδεσ	μνημειωδ
Russian	ячменный	ячмен
Russian	адом	ад

Version 1.2.0 - May 2022

• Update to use Snowball 2.2 (the Ada code generator has been integrated in Snowball 2.2!)
• Improvement to help in running the tests

Version 1.1.0 - Oct 2020

• Add support Arabic, Basque, Catalan, Finnish, Hindi, Hungarian, Indonesian, Irish, Lithuanian, Nepali, Norwegian, Porter, Portuguese, Romanian, Tamil, Turkish

Version 1.0.0 - May 2020

• First implementation of the Ada Stemmer Library

Build

Build with the following commands:

make

Unit test

To build the unit test, you will need the Ada Utility Library. The make test target will clone the git repository locally and it will configure the GNAT project accordingly to use and build the unit tests.

make build test HAVE_ADA_UTIL=yes ADA_PROJECT_PATH=./ada-util/.alire:./ada-util:./ada-util/.alire/unit

And unit tests are executed with:

make test

The unit tests contains several reference files in regtests/files that come from the Lucene search engine unit tests.

Examples

The samples can be built using:

gnatmake -Psamples

You will get two programs:

• bin/stemargs will give the stem of words given as program argument,
• bin/stemwords will read a file and stem the words to print the result.

The first argument is the language. For example:

bin/stemargs french chienne

or:

bin/stemwords english LICENSE.txt

Simple example

The Ada Stemmer library does not split words. You have to give them one word at a time to stem and it returns either the word itself or its stem. The Stemmer.Factory is the multi-language entry point. The stemmer algorithm is created for each call.

with Stemmer.Factory;

  Ada.Text_IO.Put_Line (Stem (L_FRENCH, "chienne"));

It is possible to instantiate a specific stemmer algorithm and then use it to stem words.

with Stemmer.English;

  Ctx : Stemmer.English.Context_Type;
  Result : Boolean;

  Ctx.Stem_Word ("zealously", Result);
  if Result then
     Ada.Text_IO.Put_Line (Ctx.Get_Result);
  end if;

References

• The Porter Stemming Algorithm
• Snowball Manual
• Lucene text analysis