python3 --version && java -version 2>&1 | head -1
python3 -c "from jcc import config; print('JCC', config.VERSION)"
python3 -c "import lucene; print(lucene.__file__); print('PyLucene', lucene.VERSION)"

Python 3.10.12
openjdk version "17.0.20.1" 2026-08-18
JCC 3.14
/usr/local/lib/python3.10/dist-packages/lucene/__init__.py
PyLucene 9.12.0

python3 -m jcc --help | head -6

  JCC - C++/Python Java Native Interface Code Generator

  Usage: python -m jcc [options] [actions] classes...

cd /tmp && rm -rf jcc_demo && mkdir jcc_demo && cd jcc_demo
L=/root/build/pylucene-9.12.0/lucene-java-9.12.0/lucene
python3 -m jcc --vmarg -Danalysis.data.dir=/root/analysis-data \
  --jar $L/analysis/smartcn/build/runtimeJars/lucene-analysis-smartcn-9.12.0-SNAPSHOT.jar \
  --jar $L/core/build/runtimeJars/lucene-core-9.12.0-SNAPSHOT.jar \
  --jar $L/analysis/common/build/runtimeJars/lucene-analysis-common-9.12.0-SNAPSHOT.jar \
  --exclude module-info --use_full_names \
  --python cnlib --version 1.0 --generate 2>&1 | tail -2
ls
head -2 build/cnlib/__init__.py

  Warning: renaming static method 'toString' on class java.lang.Byte to 'toString_' since it is shadowed by non-static method of same name.
  Warning: renaming static method 'rescore' on class org.apache.lucene.search.QueryRescorer to 'rescore_' since it is shadowed by non-static method of same name.
build
setup.py
import os

python3 -c "from jcc import config; print('SHARED:', config.SHARED)"

SHARED: True

python3 -m jcc --vmarg -Danalysis.data.dir=/root/analysis-data --jar .../lucene-analysis-smartcn-9.12.0-SNAPSHOT.jar --python cnlib --version 1.0 --generate

（不传时）jcc.cpp.JavaError: java.lang.ExceptionInInitializerError
        Caused by: java.lang.RuntimeException: WARNING: Can not find lexical dictionary directory! ...
（传入后）无异常，正常生成 cnlib/

python3 <<'EOF'
import lucene
lucene.initVM(vmargs=['-Xmx256m', '-Danalysis.data.dir=/root/analysis-data'])
from java.lang import System
System.out.println('REPL试验 OK, java ' + System.getProperty('java.version'))
from org.apache.lucene.util import Version
print('lucene module dir sample:', [n for n in dir(lucene) if not n.startswith('_')][:4])
EOF

REPL试验 OK, java 17.0.20.1
lucene module dir sample: ['CLASSPATH', 'ConstVariableDescriptor', 'FinalizerClass', 'FinalizerProxy']

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# PyLucene 9.12.0 建索引 + 搜索完整示例
import lucene
from java.nio.file import Paths
from org.apache.lucene.analysis.standard import StandardAnalyzer
from org.apache.lucene.analysis.cn.smart import SmartChineseAnalyzer
from org.apache.lucene.document import Document, Field, StringField, TextField
from org.apache.lucene.index import (IndexWriter, IndexWriterConfig,
                                     DirectoryReader, Term)
from org.apache.lucene.search import IndexSearcher, TermQuery, BooleanQuery,\
    BooleanClause
from org.apache.lucene.queryparser.classic import QueryParser
from org.apache.lucene.store import FSDirectory

lucene.initVM(vmargs=['-Xmx256m', '-Danalysis.data.dir=/root/analysis-data'])

INDEX_DIR = Paths.get('/root/pylucene_demo_index').toAbsolutePath()

# ---------- 1. 建索引 ----------
store = FSDirectory.open(INDEX_DIR)
analyzer = StandardAnalyzer()
config = IndexWriterConfig(analyzer)
writer = IndexWriter(store, config)

docs = [
    ('lucene-intro', 'Apache Lucene is a high-performance full-featured text search engine library'),
    ('pylucene-intro', 'PyLucene is a Python binding for Apache Lucene generated by JCC'),
    ('es-intro', 'Elasticsearch is a distributed search engine built on top of Apache Lucene'),
]
for name, body in docs:
    doc = Document()
    doc.add(StringField('name', name, Field.Store.YES))
    doc.add(TextField('body', body, Field.Store.YES))
    writer.addDocument(doc)
writer.commit()
print('index docs:', DirectoryReader.open(store).numDocs())
writer.close()

# ---------- 2. 关键词检索（QueryParser）----------
searcher = IndexSearcher(DirectoryReader.open(store))
query = QueryParser('body', analyzer).parse('search engine')
hits = searcher.search(query, 10)
print('query="search engine" totalHits:', hits.totalHits.value)
for hit in hits.scoreDocs:
    d = searcher.storedFields().document(hit.doc)
    print('  score=%.4f name=%s' % (hit.score, d.get('name')))

# ---------- 3. TermQuery + BooleanQuery 组合 ----------
q1 = TermQuery(Term('body', 'python'))
q2 = TermQuery(Term('name', 'es-intro'))
bq = BooleanQuery.Builder()
bq.add(q1, BooleanClause.Occur.SHOULD)
bq.add(q2, BooleanClause.Occur.SHOULD)
hits2 = searcher.search(bq.build(), 10)
print('boolean query totalHits:', hits2.totalHits.value)
for hit in hits2.scoreDocs:
    d = searcher.storedFields().document(hit.doc)
    print('  score=%.4f name=%s' % (hit.score, d.get('name')))
searcher.indexReader.close()

# ---------- 4. 中文分词检索（SmartChineseAnalyzer）----------
store2 = FSDirectory.open(Paths.get('/root/pylucene_demo_index_cn').toAbsolutePath())
cn_writer = IndexWriter(store2, IndexWriterConfig(SmartChineseAnalyzer()))
cn_docs = ['Lucene 是一个高性能的全文检索引擎库', 'PyLucene 让 Python 直接调用 Lucene 的全部能力']
for body in cn_docs:
    doc = Document()
    doc.add(TextField('body', body, Field.Store.YES))
    cn_writer.addDocument(doc)
cn_writer.commit()
cn_writer.close()

cn_searcher = IndexSearcher(DirectoryReader.open(store2))
cn_query = QueryParser('body', SmartChineseAnalyzer()).parse('全文检索')
cn_hits = cn_searcher.search(cn_query, 10)
print('chinese query totalHits:', cn_hits.totalHits.value)
for hit in cn_hits.scoreDocs:
    d = cn_searcher.storedFields().document(hit.doc)
    print('  score=%.4f body=%s' % (hit.score, d.get('body')))
cn_searcher.indexReader.close()
print('ALL DONE')

cd /root && rm -rf pylucene_demo_index pylucene_demo_index_cn
python3 demo_pylucene.py

index docs: 3
query="search engine" totalHits: 2
  score=0.4223 name=lucene-intro
  score=0.4223 name=es-intro
boolean query totalHits: 2
  score=0.4565 name=pylucene-intro
  score=0.4458 name=es-intro
chinese query totalHits: 1
  score=0.6301 body=Lucene 是一个高性能的全文检索引擎库
ALL DONE

cd /root/build/pylucene-9.12.0/samples
sed -i 's|lucene.initVM(vmargs=\["-Djava.awt.headless=true"\])|lucene.initVM(vmargs=["-Djava.awt.headless=true", "-Danalysis.data.dir=/root/analysis-data"])|' IndexFiles.py SearchFiles.py
mkdir -p /root/sample_docs
printf 'Apache Lucene is a high-performance full-featured text search engine library written in Java.\n' > /root/sample_docs/doc1.txt
printf 'PyLucene is a Python binding around a Java Lucene core, generated by JCC.\n' > /root/sample_docs/doc2.txt
printf 'Elasticsearch and Solr are both built on top of the Lucene core library.\n' > /root/sample_docs/doc3.txt
python3 IndexFiles.py /root/sample_docs

lucene 9.12.0
adding doc3.txt
adding doc2.txt
adding doc1.txt
commit index .done
0:00:00.085558

printf 'lucene\n\n' | python3 SearchFiles.py

lucene 9.12.0

Hit enter with no input to quit.
Query:
Searching for: lucene
3 total matching documents.
path: /root/sample_docs name: doc3.txt
path: /root/sample_docs name: doc2.txt
path: /root/sample_docs name: doc1.txt

Hit enter with no input to quit.

with lucene.attachThread():
    searcher = IndexSearcher(DirectoryReader.open(store))
    hits = searcher.search(query, 10)