// license that can be found in the LICENSE file.

//go:build icu

package cases

// Ideally these functions would be defined in a test file, but go test doesn't
// allow CGO in tests. The build tag should ensure either way that these
// functions will not end up in the package.

// TODO: Ensure that the correct ICU version is set.

/*
#cgo LDFLAGS: -licui18n.57 -licuuc.57
#include <stdlib.h>

 * either express or implied. See the License for the specific language
 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.suggest.normalizer;

import com.ibm.icu.text.Transliterator;

public class ICUNormalizer implements Normalizer {
    protected Transliterator transliterator;

    public ICUNormalizer(final String transliteratorId) {

// only makes sense, though, if the performance and/or space penalty of using
// the generic breaker is big. Extra data will only be needed for non-cased
// runes, which means there are sufficient bits left in the caseType.
// ICU prohibits breaking in such cases as well.

// For the purpose of title casing we use an approximation of the Unicode Word
// Breaking algorithm defined in Annex #29:

        "@FXdiv//:__subpackages__",
        "@gemmlowp//:__subpackages__",
        "@gif//:__subpackages__",
        "@highwayhash//:__subpackages__",
        "@hwloc//:__subpackages__",
        "@icu//:__subpackages__",
        "@jsoncpp_git//:__subpackages__",
        "@libjpeg_turbo//:__subpackages__",
        "@llvm_openmp//:__subpackages__",
        "@llvm-project//:__subpackages__",

squareup-moshi-kotlin = { module = "com.squareup.moshi:moshi-kotlin", version.ref = "com-squareup-moshi" }
squareup-okhttp-icu = "com.squareup.okhttpicu:okhttp-icu:0.2.0"
squareup-kotlinPoet = "com.squareup:kotlinpoet:1.16.0"
squareup-okio = { module = "com.squareup.okio:okio", version.ref = "com-squareup-okio" }

			<releases>
				<enabled>false</enabled>
			</releases>
			<snapshots>
				<enabled>true</enabled>
			</snapshots>
		</repository>
	</repositories>
	<dependencies>
		<dependency>
			<groupId>com.ibm.icu</groupId>
			<artifactId>icu4j</artifactId>
			<version>${icu4j.version}</version>
		</dependency>
		<dependency>
			<groupId>org.codelibs</groupId>
			<artifactId>corelib</artifactId>

import org.apache.lucene.analysis.TokenStream;
import org.apache.lucene.analysis.TokenizerFactory;
import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
import org.opensearch.core.common.Strings;

import com.ibm.icu.text.Transliterator;

public class KatakanaConverter implements ReadingConverter {

    protected final Transliterator transliterator = Transliterator.getInstance("Hiragana-Katakana");

import org.opensearch.action.admin.indices.analyze.AnalyzeAction.AnalyzeToken;
import org.opensearch.client.Client;
import org.opensearch.core.common.Strings;

import com.ibm.icu.text.Transliterator;

public class AnalyzerConverter implements ReadingConverter {
    protected final Client client;
    private final SuggestSettings settings;
    protected final AnalyzerSettings analyzerSettings;

// - https://www.unicode.org/Public/6.3.0/ucd/auxiliary/WordBreakProperty.txt
// - https://www.unicode.org/Public/6.3.0/ucd/auxiliary/WordBreakTest.txt
// - http://userguide.icu-project.org/transforms/casemappings

// TODO:
// - Case folding
// - Wide and Narrow?
// - Segmenter option for title casing.
// - ASCII fast paths
// - Encode Soft-Dotted property within trie somehow.

func (t Tag) minimize() (Tag, error) {
	t, err := minimizeTags(t)
	if err != nil {
		return t, err
	}
	t.RemakeString()
	return t, nil
}

// minimizeTags mimics the behavior of the ICU 51 C implementation.
func minimizeTags(t Tag) (Tag, error) {
	if t.equalTags(Und) {
		return t, nil
	}
	max, err := addTags(t)
	if err != nil {
		return t, err
	}
	for _, id := range [...]Tag{