This is an automated email from the ASF dual-hosted git repository. krickert pushed a commit to branch light-stemmers in repository https://gitbox.apache.org/repos/asf/opennlp.git
commit 360d4d16e771329c7811443f8cd5c8586b22b1f6 Author: Kristian Rickert <[email protected]> AuthorDate: Mon Jul 13 04:06:31 2026 -0400 OPENNLP-1886: Keep the fixture regeneration procedure in the parity README --- .../src/test/resources/opennlp/tools/stemmer/light/README.txt | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/opennlp-core/opennlp-runtime/src/test/resources/opennlp/tools/stemmer/light/README.txt b/opennlp-core/opennlp-runtime/src/test/resources/opennlp/tools/stemmer/light/README.txt index 34bb91bd6..6021835d2 100644 --- a/opennlp-core/opennlp-runtime/src/test/resources/opennlp/tools/stemmer/light/README.txt +++ b/opennlp-core/opennlp-runtime/src/test/resources/opennlp/tools/stemmer/light/README.txt @@ -8,3 +8,8 @@ complete copies of Lucene's small vocabulary lists, split by written standard. The vocabulary word lists trace back to Jacques Savoy's CLEF resources (http://members.unine.ch/jacques.savoy/clef/index.html), BSD-licensed; see the notice carried in the stemmer sources and the distribution LICENSE. + +To regenerate: compile the Lucene stemmer classes standalone together with a small reflection +driver that reads a word list and writes word<TAB>stem lines, run it per algorithm (Norwegian +with flags 1 for Bokmaal and 2 for Nynorsk), and re-sample. Any behavior difference against the +originals then shows up as a parity test failure.
