├── src ├── main │ ├── resources │ │ └── models │ │ │ └── sentencepiece.bpe.model │ └── java │ │ └── com │ │ └── sentencepiece │ │ ├── Scoring.java │ │ ├── TokenType.java │ │ ├── ResultBuilderImpl.java │ │ ├── Piece.java │ │ ├── TrieNode.java │ │ ├── SegmentEnd.java │ │ ├── SentencePieceProcessor.java │ │ ├── SentencePieceAlgorithm.java │ │ └── Model.java └── test │ └── java │ ├── Tests2.java │ ├── Tests.java │ └── Main.java ├── .idea ├── vcs.xml ├── .gitignore ├── encodings.xml ├── misc.xml └── uiDesigner.xml ├── .gitignore ├── .github └── workflows │ └── maven.yml ├── README.md ├── pom.xml └── LICENSE /src/main/resources/models/sentencepiece.bpe.model: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/eix128/SentencePiece4J/HEAD/src/main/resources/models/sentencepiece.bpe.model -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/Scoring.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | public enum Scoring { 4 | FEWEST_SEGMENTS, 5 | HIGHEST_SCORE 6 | } 7 | -------------------------------------------------------------------------------- /.idea/vcs.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | -------------------------------------------------------------------------------- /.idea/.gitignore: -------------------------------------------------------------------------------- 1 | # Default ignored files 2 | /shelf/ 3 | /workspace.xml 4 | # Editor-based HTTP Client requests 5 | /httpRequests/ 6 | # Datasource local storage ignored files 7 | /dataSources/ 8 | /dataSources.local.xml 9 | -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/TokenType.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | 4 | public enum TokenType { 5 | UNKNOWN, 6 | TEXT, // NORMAL 7 | USER_DEFINED, 8 | CONTROL, 9 | BYTE, 10 | UNUSED 11 | } -------------------------------------------------------------------------------- /.idea/encodings.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | -------------------------------------------------------------------------------- /.idea/misc.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 10 | 11 | 12 | 13 | 14 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | target/ 2 | !.mvn/wrapper/maven-wrapper.jar 3 | !**/src/main/**/target/ 4 | !**/src/test/**/target/ 5 | 6 | ### IntelliJ IDEA ### 7 | .idea/modules.xml 8 | .idea/jarRepositories.xml 9 | .idea/compiler.xml 10 | .idea/libraries/ 11 | *.iws 12 | *.iml 13 | *.ipr 14 | 15 | ### Eclipse ### 16 | .apt_generated 17 | .classpath 18 | .factorypath 19 | .project 20 | .settings 21 | .springBeans 22 | .sts4-cache 23 | 24 | ### NetBeans ### 25 | /nbproject/private/ 26 | /nbbuild/ 27 | /dist/ 28 | /nbdist/ 29 | /.nb-gradle/ 30 | build/ 31 | !**/src/main/**/build/ 32 | !**/src/test/**/build/ 33 | 34 | ### VS Code ### 35 | .vscode/ 36 | 37 | ### Mac OS ### 38 | .DS_Store -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/ResultBuilderImpl.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | import java.util.ArrayList; 4 | import java.util.List; 5 | 6 | public class ResultBuilderImpl implements SentencePieceAlgorithm.ResultBuilder { 7 | private final List tokenIds = new ArrayList<>(); 8 | 9 | @Override 10 | public void build(String input, SegmentEnd[] segmentEnds, boolean collapseUnknowns) { 11 | int i = segmentEnds.length - 1; 12 | while (i > 0) { 13 | SegmentEnd end = segmentEnds[i]; 14 | tokenIds.add(0, end.getId()); 15 | i = end.getSegmentStart(); 16 | } 17 | } 18 | 19 | public List getTokenIds() { 20 | return tokenIds; 21 | } 22 | } -------------------------------------------------------------------------------- /src/test/java/Tests2.java: -------------------------------------------------------------------------------- 1 | import com.sentencepiece.SentencePieceProcessor; 2 | 3 | import java.io.IOException; 4 | import java.nio.file.Path; 5 | import java.nio.file.Paths; 6 | import java.util.List; 7 | 8 | public class Tests2 { 9 | public static void main(String[] args) throws IOException { 10 | Path modelPath = Paths.get("models/sentencepiece.bpe.model"); 11 | SentencePieceProcessor processor = new SentencePieceProcessor(modelPath); 12 | 13 | String raw = "Akşam eve gidince yağlı ballı ekmek yemek istiyorum."; 14 | List ids = processor.encode(raw); 15 | 16 | System.out.println("IDs: " + ids); 17 | System.out.println("Decoded: " + processor.decode(ids)); 18 | System.out.println("Escaped: " + processor.decodeSmart(ids)); 19 | 20 | } 21 | } 22 | -------------------------------------------------------------------------------- /.github/workflows/maven.yml: -------------------------------------------------------------------------------- 1 | name: Java CI with Maven (Java 8) 2 | 3 | on: 4 | push: 5 | branches: [ "main" ] 6 | pull_request: 7 | branches: [ "main" ] 8 | 9 | jobs: 10 | build: 11 | 12 | runs-on: ubuntu-latest 13 | 14 | steps: 15 | - uses: actions/checkout@v4 16 | 17 | - name: Set up JDK 8 18 | uses: actions/setup-java@v4 19 | with: 20 | java-version: '8' 21 | distribution: 'temurin' 22 | cache: maven 23 | 24 | - name: Build with Maven 25 | run: mvn -B package --file pom.xml 26 | 27 | # Optional: Uploads the full dependency graph to GitHub to improve the quality of Dependabot alerts this repository can receive 28 | - name: Update dependency graph 29 | uses: advanced-security/maven-dependency-submission-action@571e99aab1055c2e71a1e2309b9691de18d6b7d6 30 | -------------------------------------------------------------------------------- /src/test/java/Tests.java: -------------------------------------------------------------------------------- 1 | import com.sentencepiece.Model; 2 | 3 | import java.io.IOException; 4 | import java.nio.file.Path; 5 | import java.nio.file.Paths; 6 | import java.util.List; 7 | 8 | 9 | public class Tests { 10 | public static void main(String[] args) throws IOException { 11 | Path modelPath = Paths.get("models/sentencepiece.bpe.model"); 12 | 13 | // Load the model 14 | System.out.println("Loading model..."); 15 | Model model = Model.parseFrom(modelPath); 16 | System.out.println("Model loaded with maxScore = " + model.getMaxScore()); 17 | 18 | // Test encode 19 | String text = "▁this ▁is ▁a ▁test"; 20 | List ids = model.encode(text); 21 | System.out.println("Encoded: " + ids); 22 | 23 | // Test decode 24 | String decoded = model.decode(ids); 25 | System.out.println("Decoded: " + decoded); 26 | 27 | // Extra: reverse check 28 | for (int id : ids) { 29 | System.out.printf("id: %d → token: '%s'%n", id, model.getTokenById(id)); 30 | } 31 | } 32 | } -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/Piece.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | 4 | import sentencepiece.SentencepieceModel; 5 | 6 | public class Piece { 7 | private final String token; 8 | private final int id; 9 | private final float score; 10 | private final TokenType type; // NEW 11 | 12 | public Piece(String token, int id, float score, TokenType type) { 13 | this.token = token; 14 | this.id = id; 15 | this.score = score; 16 | this.type = type; 17 | } 18 | 19 | public String getToken() { return token; } 20 | public int getId() { return id; } 21 | public float getScore() { return score; } 22 | public TokenType getType() { return type; } // NEW 23 | 24 | public static TokenType mapType(SentencepieceModel.ModelProto.SentencePiece.Type t) { 25 | switch (t) { 26 | case NORMAL: return TokenType.TEXT; 27 | case UNKNOWN: return TokenType.UNKNOWN; 28 | case USER_DEFINED: return TokenType.USER_DEFINED; 29 | case CONTROL: return TokenType.CONTROL; 30 | case BYTE: return TokenType.BYTE; 31 | default: return TokenType.TEXT; 32 | } 33 | } 34 | } 35 | -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/TrieNode.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | import it.unimi.dsi.fastutil.ints.Int2ObjectOpenHashMap; 4 | 5 | public class TrieNode { 6 | private final it.unimi.dsi.fastutil.ints.Int2ObjectOpenHashMap children = new Int2ObjectOpenHashMap<>(); 7 | private boolean isToken; 8 | private TokenType type = TokenType.TEXT; 9 | private int id = -1; 10 | private float score; 11 | 12 | 13 | public TrieNode getOrCreate(int codePoint) { 14 | TrieNode n = children.get(codePoint); 15 | if (n == null) { n = new TrieNode(); children.put(codePoint, n); } 16 | return n; 17 | } 18 | 19 | public TrieNode child(int codePoint) { 20 | return children.get(codePoint); 21 | } 22 | 23 | public void mark(int id, float score, TokenType type) { 24 | this.isToken = true; this.id = id; this.score = score; this.type = type; 25 | } 26 | 27 | public boolean isToken() { 28 | return isToken; 29 | } 30 | 31 | public TokenType getType() { 32 | return type; 33 | } 34 | 35 | public int getId() { 36 | return id; 37 | } 38 | 39 | public float getScore() { 40 | return score; 41 | } 42 | 43 | } -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/SegmentEnd.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | public class SegmentEnd { 4 | private final TokenType type; 5 | private final int id; 6 | private final float pathScoreSum; 7 | private final int pathSegmentCount; 8 | private final int segmentStart; 9 | 10 | public SegmentEnd(TokenType type, int id, float pathScoreSum, int pathSegmentCount, int segmentStart) { 11 | this.type = type; 12 | this.id = id; 13 | this.pathScoreSum = pathScoreSum; 14 | this.pathSegmentCount = pathSegmentCount; 15 | this.segmentStart = segmentStart; 16 | } 17 | 18 | public float score(Scoring scoring) { 19 | switch (scoring) { 20 | case FEWEST_SEGMENTS: 21 | return 1f / pathSegmentCount * 10_000_000 + pathScoreSum; 22 | case HIGHEST_SCORE: 23 | return pathScoreSum; 24 | default: 25 | throw new IllegalArgumentException("Unknown scoring " + scoring); 26 | } 27 | } 28 | 29 | public float scoreWith(Scoring scoring, float additionalSegmentScore) { 30 | switch (scoring) { 31 | case FEWEST_SEGMENTS: 32 | return 1f / (pathSegmentCount + 1) * 10_000_000 + (pathScoreSum + additionalSegmentScore); 33 | case HIGHEST_SCORE: 34 | return pathScoreSum + additionalSegmentScore; 35 | default: 36 | throw new IllegalArgumentException("Unknown scoring " + scoring); 37 | } 38 | } 39 | 40 | public int getId() { 41 | return id; 42 | } 43 | 44 | public float getPathScoreSum() { 45 | return pathScoreSum; 46 | } 47 | 48 | public int getPathSegmentCount() { 49 | return pathSegmentCount; 50 | } 51 | 52 | public int getSegmentStart() { 53 | return segmentStart; 54 | } 55 | } -------------------------------------------------------------------------------- /src/test/java/Main.java: -------------------------------------------------------------------------------- 1 | import com.sentencepiece.Model; 2 | import com.sentencepiece.Scoring; 3 | import com.sentencepiece.SentencePieceAlgorithm; 4 | 5 | import java.io.IOException; 6 | import java.nio.file.Paths; 7 | import java.util.List; 8 | 9 | public class Main { 10 | public static void main(String[] args) throws IOException { 11 | Model model = Model.getInstance(); 12 | SentencePieceAlgorithm algorithm = new SentencePieceAlgorithm( 13 | true, Scoring.HIGHEST_SCORE 14 | ); 15 | 16 | { 17 | String raw = "o captain! my captain! our fearful trip is done," + 18 | "the ship has weather’d every rack, the prize we sought is won," + 19 | "the port is near, the bells i hear, the people all exulting,"; 20 | List ids = model.encodeNormalized(raw, algorithm); 21 | 22 | System.out.println("Token IDs: " + ids); 23 | System.out.println("Decoded text: " + model.decodeSmart(ids)); 24 | assert (raw.hashCode() == model.decodeSmart(ids).hashCode()); 25 | System.out.println("✔ Success"); 26 | 27 | } 28 | 29 | //Test Hebrew 30 | { 31 | String raw = "השתיקה יפה לחכמים"; 32 | List ids = model.encodeNormalized(raw, algorithm); 33 | 34 | System.out.println("Token IDs: " + ids); 35 | System.out.println("Decoded text: " + model.normalizeHebrew(model.decodeSmart(ids) , true )); 36 | assert (raw.hashCode() == model.decodeSmart(ids).hashCode()); 37 | System.out.println("✔ Success"); 38 | } 39 | 40 | 41 | //Test Turkish 42 | { 43 | String raw = "kusur bulmak için bakma birine! bakmak için bakarsan bulursun.kusursuz olmayı marifet edin kendine , işte asıl o zaman kusursuz olursun!..."; 44 | List ids = model.encodeNormalized(raw, algorithm); 45 | 46 | System.out.println("Token IDs: " + ids); 47 | System.out.println("Decoded text: " + model.decodeSmart(ids)); 48 | assert (raw.hashCode() == model.decodeSmart(ids).hashCode()); 49 | System.out.println("✔ Success"); 50 | } 51 | 52 | } 53 | } 54 | -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/SentencePieceProcessor.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | 4 | 5 | import java.io.IOException; 6 | import java.nio.file.Path; 7 | import java.text.Normalizer; 8 | import java.util.*; 9 | import java.util.regex.Pattern; 10 | import java.util.stream.Collectors; 11 | 12 | public class SentencePieceProcessor { 13 | 14 | private final Model model; 15 | private final SentencePieceAlgorithm algorithm; 16 | private final boolean lowercase; 17 | 18 | // Model fields 19 | 20 | public SentencePieceProcessor(Path modelPath) throws IOException { 21 | this.model = Model.parseFrom(modelPath); 22 | this.algorithm = new SentencePieceAlgorithm(true, Scoring.HIGHEST_SCORE); 23 | this.lowercase = false; 24 | } 25 | 26 | 27 | public List encode(String rawText) { 28 | String normalized = normalize(rawText); 29 | 30 | // Add SentencePiece boundary markers (▁) 31 | StringBuilder sb = new StringBuilder(); 32 | for (String word : normalized.split(" ")) { 33 | sb.append('▁').append(word); 34 | } 35 | String prepared = sb.toString(); 36 | 37 | ResultBuilderImpl builder = new ResultBuilderImpl(); 38 | algorithm.segment(prepared, builder, model); 39 | return builder.getTokenIds(); 40 | } 41 | 42 | public String decode(List ids) { 43 | return ids.stream() 44 | .map(model::tokenById) 45 | .map(t -> t.startsWith("▁") ? t.substring(1) : t) 46 | .collect(Collectors.joining(" ")) 47 | .replaceAll(" +", " ") 48 | .trim(); 49 | } 50 | 51 | public String decodeSmart(List ids) { 52 | StringBuilder sb = new StringBuilder(); 53 | 54 | for (int id : ids) { 55 | String token = model.tokenById(id); 56 | if (token.startsWith("▁")) { 57 | if (sb.length() > 0) sb.append(' '); 58 | sb.append(token.substring(1)); 59 | } else { 60 | sb.append(token); 61 | } 62 | } 63 | 64 | return sb.toString().trim(); 65 | } 66 | 67 | 68 | private static final Pattern WS = Pattern.compile("\\p{Z}+"); 69 | private String normalize(String input) { 70 | String s = lowercase ? input.toLowerCase() : input; 71 | s = Normalizer.normalize(s, Normalizer.Form.NFKC); 72 | s = s.replaceAll("\\p{Cc}+", ""); // drop control chars 73 | s = WS.matcher(s.trim()).replaceAll(" "); 74 | return s; 75 | } 76 | } 77 | -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/SentencePieceAlgorithm.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | public class SentencePieceAlgorithm { 4 | 5 | 6 | private final boolean collapseUnknowns; 7 | private final Scoring scoring; 8 | 9 | public SentencePieceAlgorithm(boolean collapseUnknowns, Scoring scoring) { 10 | this.collapseUnknowns = collapseUnknowns; 11 | this.scoring = scoring; 12 | } 13 | 14 | public void segment(String input, ResultBuilder resultBuilder, Model model) { 15 | SegmentEnd[] segmentEnds = new SegmentEnd[input.length() + 1]; 16 | segmentEnds[0] = new SegmentEnd(TokenType.UNKNOWN, 0, 0, 0, 0); 17 | 18 | int start = 0; 19 | while (start < input.length()) { 20 | TrieNode node = model.getRoot(); 21 | int pos = start; 22 | while (node != null && pos < input.length()) { 23 | int cp = input.codePointAt(pos); 24 | pos += Character.charCount(cp); 25 | node = node.child(cp); 26 | int length = pos - start; 27 | if (node != null && node.isToken() && node.getType() != TokenType.UNUSED) { 28 | float score = (node.getType() == TokenType.USER_DEFINED) 29 | ? (length * model.getMaxScore() - 0.1f) 30 | : node.getScore(); 31 | addSegment(node.getType(), node.getId(), start, pos, score, segmentEnds); 32 | } else if (length == Character.charCount(cp)) { 33 | addSegment(TokenType.UNKNOWN, model.getUnkId() >= 0 ? model.getUnkId() : 0, 34 | start, start + length, model.getMinScore() - 10.0f, segmentEnds); 35 | } 36 | } 37 | start += Character.charCount(input.codePointAt(start)); 38 | } 39 | resultBuilder.build(input, segmentEnds, collapseUnknowns); 40 | } 41 | 42 | private void addSegment(TokenType type, int id, int start, int end, float score, SegmentEnd[] segmentEnds) { 43 | if (segmentEnds[end] == null || 44 | segmentEnds[start].scoreWith(scoring, score) > segmentEnds[end].score(scoring)) { 45 | 46 | segmentEnds[end] = new SegmentEnd( 47 | type, id, 48 | segmentEnds[start].getPathScoreSum() + score, 49 | segmentEnds[start].getPathSegmentCount() + 1, 50 | start 51 | ); 52 | } 53 | } 54 | 55 | public interface ResultBuilder { 56 | void build(String input, SegmentEnd[] segmentEnds, boolean collapseUnknowns); 57 | } 58 | } 59 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # SentencePiece4J 2 | SentencePiece in pure Java , No JNI Required. 3 | Works Cross Platform for all OS 4 | 5 | use with : 6 | 7 | # download sentencepiece 8 | wget https://github.com/eix128/SentencePiece4J/raw/refs/heads/main/src/main/resources/sentencepiece.bpe.model 9 | 10 | # mvn package 11 | 12 | io.github.eix128 13 | sentencepiece4j 14 | 1.0.2 15 | 16 | 17 | # Usage 18 | import com.sentencepiece.Model; 19 | import com.sentencepiece.Scoring; 20 | import com.sentencepiece.SentencePieceAlgorithm; 21 | 22 | import java.io.IOException; 23 | import java.nio.file.Paths; 24 | import java.util.List; 25 | 26 | public class Main { 27 | public static void main(String[] args) throws IOException { 28 | Model model = Model.parseFrom(Paths.get("sentencepiece.bpe.model")); 29 | SentencePieceAlgorithm algorithm = new SentencePieceAlgorithm( true, Scoring.HIGHEST_SCORE ); 30 | 31 | { 32 | String raw = "o captain! my captain! our fearful trip is done," + 33 | "the ship has weather’d every rack, the prize we sought is won," + 34 | "the port is near, the bells i hear, the people all exulting,"; 35 | List ids = model.encodeNormalized(raw, algorithm); 36 | 37 | System.out.println("Token IDs: " + ids); 38 | System.out.println("Decoded text: " + model.decodeSmart(ids)); 39 | assert (raw.hashCode() == model.decodeSmart(ids).hashCode()); 40 | System.out.println("✔ Success"); 41 | 42 | } 43 | 44 | //Test Hebrew 45 | { 46 | String raw = "השתיקה יפה לחכמים"; 47 | List ids = model.encodeNormalized(raw, algorithm); 48 | 49 | System.out.println("Token IDs: " + ids); 50 | System.out.println("Decoded text: " + model.normalizeHebrew(model.decodeSmart(ids) , true )); 51 | assert (raw.hashCode() == model.decodeSmart(ids).hashCode()); 52 | System.out.println("✔ Success"); 53 | } 54 | 55 | 56 | //Test Turkish 57 | { 58 | String raw = "kusur bulmak için bakma birine! kakmak için bakarsan bulursun.kusursuz olmayı marifet edin kendine , işte asıl o zaman kusursuz olursun!..."; 59 | List ids = model.encodeNormalized(raw, algorithm); 60 | 61 | System.out.println("Token IDs: " + ids); 62 | System.out.println("Decoded text: " + model.decodeSmart(ids)); 63 | assert (raw.hashCode() == model.decodeSmart(ids).hashCode()); 64 | System.out.println("✔ Success"); 65 | } 66 | } 67 | } 68 | 69 | -------------------------------------------------------------------------------- /pom.xml: -------------------------------------------------------------------------------- 1 | 2 | 5 | 4.0.0 6 | 7 | io.github.eix128 8 | sentencepiece4j 9 | 1.0.2 10 | 11 | 12 | ${project.groupId}:${project.artifactId} 13 | Java utilities/bindings for SentencePiece (tokenize/encode/decode, BPE/SPM helpers). 14 | https://github.com/eix128/SentencePiece4J 15 | 16 | 17 | 18 | Apache License 2.0 19 | https://www.apache.org/licenses/LICENSE-2.0.txt 20 | 21 | 22 | 23 | 24 | 25 | Kadir BASOL 26 | kadir.bayner@gmail.com 27 | https://github.com/eix128 28 | 29 | 30 | 31 | 32 | https://github.com/eix128/SentencePiece4J 33 | scm:git:https://github.com/eix128/SentencePiece4J.git 34 | scm:git:ssh://github.com/eix128/SentencePiece4J.git 35 | 36 | 37 | 38 | 39 | 8 40 | 8 41 | UTF-8 42 | 43 | 44 | 5.9.3 45 | 1.9.3 46 | 47 | 85D81515741EDCFE6BF86ED41A342E30F64EAA40 48 | 49 | 50 | 3.1.2 51 | 52 | 53 | 54 | 55 | com.google.protobuf 56 | protobuf-java 57 | 4.32.1 58 | 59 | 60 | it.unimi.dsi 61 | fastutil 62 | 8.5.12 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 | maven-source-plugin 73 | 3.3.1 74 | 75 | 76 | attach-sources 77 | jar-no-fork 78 | 79 | 80 | 81 | 82 | 83 | 84 | 85 | org.apache.maven.plugins 86 | maven-gpg-plugin 87 | 3.2.4 88 | 89 | signverifysign 90 | 91 | 92 | 93 | --pinentry-modeloopback 94 | 95 | 96 | 97 | 98 | 99 | 100 | maven-javadoc-plugin 101 | 3.6.3 102 | 103 | 104 | attach-javadocs 105 | jar 106 | 107 | 8 108 | false 109 | 110 | 111 | 112 | 113 | 114 | 115 | 116 | 117 | org.sonatype.central 118 | central-publishing-maven-plugin 119 | 0.9.0 120 | true 121 | 122 | central 123 | 127 | 128 | 129 | 130 | 131 | -------------------------------------------------------------------------------- /src/main/java/com/sentencepiece/Model.java: -------------------------------------------------------------------------------- 1 | package com.sentencepiece; 2 | 3 | 4 | import com.google.protobuf.InvalidProtocolBufferException; 5 | import sentencepiece.SentencepieceModel; 6 | 7 | import java.io.ByteArrayOutputStream; 8 | import java.io.IOException; 9 | import java.io.InputStream; 10 | import java.nio.file.Files; 11 | import java.nio.file.Path; 12 | import java.text.Normalizer; 13 | import java.util.ArrayList; 14 | import java.util.HashMap; 15 | import java.util.List; 16 | import java.util.Map; 17 | import java.util.regex.Pattern; 18 | import java.util.stream.Collectors; 19 | 20 | import static com.sentencepiece.Piece.mapType; 21 | 22 | 23 | /** 24 | * Wraps the SentencePiece protobuf model into a Java-friendly structure. 25 | */ 26 | public class Model { 27 | 28 | private final Map vocabulary = new HashMap<>(); 29 | private final TrieNode root = new TrieNode(); 30 | private float maxScore = Float.NEGATIVE_INFINITY; 31 | private float minScore = Float.POSITIVE_INFINITY; 32 | 33 | private final List piecesById = new ArrayList<>(); 34 | // NEW: store id (or -1 if absent) 35 | private int unkId = -1; 36 | 37 | 38 | public Model() { 39 | 40 | } 41 | 42 | 43 | private static byte[] toByteArray(InputStream in) throws IOException { 44 | ByteArrayOutputStream buffer = new ByteArrayOutputStream(); 45 | byte[] data = new byte[8192]; 46 | int nRead; 47 | while ((nRead = in.read(data, 0, data.length)) != -1) { 48 | buffer.write(data, 0, nRead); 49 | } 50 | return buffer.toByteArray(); 51 | } 52 | 53 | 54 | 55 | 56 | private static Model init( byte[] bytes ) throws InvalidProtocolBufferException { 57 | SentencepieceModel.ModelProto modelProto = SentencepieceModel.ModelProto.parseFrom(bytes); 58 | Model model = new Model(); 59 | 60 | int id = 0; 61 | for (SentencepieceModel.ModelProto.SentencePiece sp : modelProto.getPiecesList()) { 62 | String token = sp.getPiece(); 63 | float score = sp.getScore(); 64 | model.addPiece(token, id++, score, mapType(sp.getType())); 65 | } 66 | 67 | // determine unkId 68 | if (modelProto.hasTrainerSpec() && modelProto.getTrainerSpec().hasUnkId()) { 69 | model.unkId = modelProto.getTrainerSpec().getUnkId(); 70 | } else { 71 | for (Piece p : model.piecesById) { 72 | if (p != null && "".equals(p.getToken())) { 73 | model.unkId = p.getId(); 74 | break; 75 | } 76 | } 77 | } 78 | return model; 79 | } 80 | 81 | /** 82 | * @return returns sentencepiece.bpe.model model which is stored inside library 83 | * @throws IOException : when model is not found 84 | */ 85 | public static Model getInstance( ) throws IOException { 86 | return parseFromResource("models/sentencepiece.bpe.model"); 87 | } 88 | 89 | 90 | public static Model parseFromResource(String resourceName) throws IOException { 91 | try (InputStream in = Model.class.getClassLoader().getResourceAsStream(resourceName)) { 92 | if (in == null) { 93 | throw new IOException("Resource not found: " + resourceName); 94 | } 95 | byte[] bytes = toByteArray(in); 96 | return init(bytes); 97 | } 98 | } 99 | 100 | 101 | 102 | public static Model parseFrom(Path modelPath) throws IOException { 103 | byte[] bytes = Files.readAllBytes(modelPath); 104 | return init(bytes); 105 | } 106 | 107 | public void addPiece(String token, int id, float score, TokenType type) { 108 | Piece piece = new Piece(token, id, score, type); 109 | vocabulary.put(token, piece); 110 | 111 | // ensure O(1) id lookup 112 | while (piecesById.size() <= id) piecesById.add(null); 113 | piecesById.set(id, piece); 114 | 115 | maxScore = Math.max(maxScore, score); 116 | minScore = Math.min(minScore, score); 117 | insertIntoTrie(token, piece); 118 | } 119 | 120 | private void insertIntoTrie(String token, Piece piece) { 121 | TrieNode node = root; 122 | for (int i = 0; i < token.length(); ) { 123 | int cp = token.codePointAt(i); 124 | i += Character.charCount(cp); 125 | node = node.getOrCreate(cp); 126 | } 127 | // If Piece doesn't have getType(), use a default TokenType.TEXT here 128 | node.mark(piece.getId(), piece.getScore(), piece.getType()); 129 | } 130 | 131 | public TrieNode getRoot() { 132 | return root; 133 | } 134 | 135 | public float getMaxScore() { 136 | return maxScore; 137 | } 138 | 139 | public float getMinScore() { 140 | return minScore; 141 | } 142 | 143 | public String getTokenById(int id) { 144 | for (Piece p : vocabulary.values()) { 145 | if (p.getId() == id) return p.getToken(); 146 | } 147 | return ""; 148 | } 149 | 150 | public String tokenById(int id) { 151 | if (id < 0 || id >= piecesById.size()) return ""; 152 | Piece piece = piecesById.get(id); 153 | return piece != null ? piece.getToken() : ""; 154 | } 155 | 156 | public int getUnkId() { return unkId; } 157 | 158 | public int getIdForToken(String token) { 159 | Piece piece = vocabulary.get(token); 160 | return (piece != null) ? piece.getId() : -1; 161 | } 162 | 163 | 164 | public List encode(String input) { 165 | // Naive space-based token matching for test (improve later with BPE segmenter) 166 | String trimmedInput = input.trim(); 167 | String[] tokens = trimmedInput.split("\\s+"); 168 | List ids = new ArrayList<>(); 169 | for (String t : tokens) { 170 | int id = getIdForToken(t); 171 | if (id != -1) { 172 | ids.add(id); 173 | } else { 174 | System.err.println("Unknown token: " + t); 175 | } 176 | } 177 | return ids; 178 | } 179 | 180 | private static final Pattern pattern = Pattern.compile("\\s+"); 181 | 182 | public List encodeNormalized(String rawInput, SentencePieceAlgorithm algorithm) { 183 | // Step 1: Unicode normalization 184 | String normalized = Normalizer.normalize(rawInput, Normalizer.Form.NFKC).toLowerCase(); 185 | 186 | // Step 2: Collapse whitespace 187 | normalized = pattern.matcher(normalized.trim()).replaceAll(" "); 188 | 189 | // Step 3: Add '▁' marker before each word 190 | StringBuilder sb = new StringBuilder(); 191 | for (String word : normalized.split(" ")) { 192 | sb.append('▁').append(word); 193 | } 194 | 195 | String prepared = sb.toString(); 196 | 197 | // Step 4: Segment using algorithm 198 | ResultBuilderImpl builder = new ResultBuilderImpl(); 199 | algorithm.segment(prepared, builder, this); 200 | return builder.getTokenIds(); 201 | } 202 | 203 | 204 | public String decodeSmart(List ids) { 205 | StringBuilder sb = new StringBuilder(); 206 | 207 | for (int id : ids) { 208 | String token = tokenById(id); 209 | 210 | if (token.equals("")) { 211 | sb.append("�"); // or some fallback 212 | continue; 213 | } 214 | 215 | if (token.startsWith("▁")) { 216 | // Word boundary → insert a space if not at the very start 217 | if (sb.length() > 0) sb.append(' '); 218 | sb.append(token.substring(1)); 219 | } else { 220 | // Continuation of the current word → append directly 221 | sb.append(token); 222 | } 223 | } 224 | 225 | return sb.toString().trim(); 226 | } 227 | 228 | // Choose ONE consistently with training data 229 | private static final Pattern ZS = Pattern.compile("\\p{Z}+"); 230 | 231 | // Strip Hebrew diacritics if model expects undiacritized text 232 | private static final Pattern HEBREW_DIACRITICS = 233 | Pattern.compile("[\\u0591-\\u05BD\\u05BF\\u05C1-\\u05C2\\u05C4-\\u05C5\\u05C7]"); 234 | 235 | public String normalizeHebrew(String input, boolean stripDiacritics) { 236 | String s = Normalizer.normalize(input, Normalizer.Form.NFC); 237 | if (stripDiacritics) s = HEBREW_DIACRITICS.matcher(s).replaceAll(""); 238 | s = ZS.matcher(s.trim()).replaceAll(" "); 239 | return s; 240 | } 241 | 242 | public String decode(List ids) { 243 | return ids.stream() 244 | .map(this::tokenById) 245 | .collect(Collectors.joining(" ")); 246 | } 247 | } 248 | -------------------------------------------------------------------------------- /.idea/uiDesigner.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 | 73 | 74 | 75 | 76 | 77 | 78 | 79 | 80 | 81 | 82 | 83 | 84 | 85 | 86 | 87 | 88 | 89 | 90 | 91 | 92 | 93 | 94 | 95 | 96 | 97 | 98 | 99 | 100 | 101 | 102 | 103 | 104 | 105 | 106 | 107 | 108 | 109 | 110 | 111 | 112 | 113 | 114 | 115 | 116 | 117 | 118 | 119 | 120 | 121 | 122 | 123 | 124 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Apache License 2 | Version 2.0, January 2004 3 | http://www.apache.org/licenses/ 4 | 5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 6 | 7 | 1. Definitions. 8 | 9 | "License" shall mean the terms and conditions for use, reproduction, 10 | and distribution as defined by Sections 1 through 9 of this document. 11 | 12 | "Licensor" shall mean the copyright owner or entity authorized by 13 | the copyright owner that is granting the License. 14 | 15 | "Legal Entity" shall mean the union of the acting entity and all 16 | other entities that control, are controlled by, or are under common 17 | control with that entity. For the purposes of this definition, 18 | "control" means (i) the power, direct or indirect, to cause the 19 | direction or management of such entity, whether by contract or 20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the 21 | outstanding shares, or (iii) beneficial ownership of such entity. 22 | 23 | "You" (or "Your") shall mean an individual or Legal Entity 24 | exercising permissions granted by this License. 25 | 26 | "Source" form shall mean the preferred form for making modifications, 27 | including but not limited to software source code, documentation 28 | source, and configuration files. 29 | 30 | "Object" form shall mean any form resulting from mechanical 31 | transformation or translation of a Source form, including but 32 | not limited to compiled object code, generated documentation, 33 | and conversions to other media types. 34 | 35 | "Work" shall mean the work of authorship, whether in Source or 36 | Object form, made available under the License, as indicated by a 37 | copyright notice that is included in or attached to the work 38 | (an example is provided in the Appendix below). 39 | 40 | "Derivative Works" shall mean any work, whether in Source or Object 41 | form, that is based on (or derived from) the Work and for which the 42 | editorial revisions, annotations, elaborations, or other modifications 43 | represent, as a whole, an original work of authorship. For the purposes 44 | of this License, Derivative Works shall not include works that remain 45 | separable from, or merely link (or bind by name) to the interfaces of, 46 | the Work and Derivative Works thereof. 47 | 48 | "Contribution" shall mean any work of authorship, including 49 | the original version of the Work and any modifications or additions 50 | to that Work or Derivative Works thereof, that is intentionally 51 | submitted to Licensor for inclusion in the Work by the copyright owner 52 | or by an individual or Legal Entity authorized to submit on behalf of 53 | the copyright owner. For the purposes of this definition, "submitted" 54 | means any form of electronic, verbal, or written communication sent 55 | to the Licensor or its representatives, including but not limited to 56 | communication on electronic mailing lists, source code control systems, 57 | and issue tracking systems that are managed by, or on behalf of, the 58 | Licensor for the purpose of discussing and improving the Work, but 59 | excluding communication that is conspicuously marked or otherwise 60 | designated in writing by the copyright owner as "Not a Contribution." 61 | 62 | "Contributor" shall mean Licensor and any individual or Legal Entity 63 | on whose behalf a Contribution has been received by Licensor and 64 | subsequently incorporated within the Work. 65 | 66 | 2. Grant of Copyright License. Subject to the terms and conditions of 67 | this License, each Contributor hereby grants to You a perpetual, 68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 69 | copyright license to reproduce, prepare Derivative Works of, 70 | publicly display, publicly perform, sublicense, and distribute the 71 | Work and such Derivative Works in Source or Object form. 72 | 73 | 3. Grant of Patent License. Subject to the terms and conditions of 74 | this License, each Contributor hereby grants to You a perpetual, 75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 76 | (except as stated in this section) patent license to make, have made, 77 | use, offer to sell, sell, import, and otherwise transfer the Work, 78 | where such license applies only to those patent claims licensable 79 | by such Contributor that are necessarily infringed by their 80 | Contribution(s) alone or by combination of their Contribution(s) 81 | with the Work to which such Contribution(s) was submitted. If You 82 | institute patent litigation against any entity (including a 83 | cross-claim or counterclaim in a lawsuit) alleging that the Work 84 | or a Contribution incorporated within the Work constitutes direct 85 | or contributory patent infringement, then any patent licenses 86 | granted to You under this License for that Work shall terminate 87 | as of the date such litigation is filed. 88 | 89 | 4. Redistribution. You may reproduce and distribute copies of the 90 | Work or Derivative Works thereof in any medium, with or without 91 | modifications, and in Source or Object form, provided that You 92 | meet the following conditions: 93 | 94 | (a) You must give any other recipients of the Work or 95 | Derivative Works a copy of this License; and 96 | 97 | (b) You must cause any modified files to carry prominent notices 98 | stating that You changed the files; and 99 | 100 | (c) You must retain, in the Source form of any Derivative Works 101 | that You distribute, all copyright, patent, trademark, and 102 | attribution notices from the Source form of the Work, 103 | excluding those notices that do not pertain to any part of 104 | the Derivative Works; and 105 | 106 | (d) If the Work includes a "NOTICE" text file as part of its 107 | distribution, then any Derivative Works that You distribute must 108 | include a readable copy of the attribution notices contained 109 | within such NOTICE file, excluding those notices that do not 110 | pertain to any part of the Derivative Works, in at least one 111 | of the following places: within a NOTICE text file distributed 112 | as part of the Derivative Works; within the Source form or 113 | documentation, if provided along with the Derivative Works; or, 114 | within a display generated by the Derivative Works, if and 115 | wherever such third-party notices normally appear. The contents 116 | of the NOTICE file are for informational purposes only and 117 | do not modify the License. You may add Your own attribution 118 | notices within Derivative Works that You distribute, alongside 119 | or as an addendum to the NOTICE text from the Work, provided 120 | that such additional attribution notices cannot be construed 121 | as modifying the License. 122 | 123 | You may add Your own copyright statement to Your modifications and 124 | may provide additional or different license terms and conditions 125 | for use, reproduction, or distribution of Your modifications, or 126 | for any such Derivative Works as a whole, provided Your use, 127 | reproduction, and distribution of the Work otherwise complies with 128 | the conditions stated in this License. 129 | 130 | 5. Submission of Contributions. Unless You explicitly state otherwise, 131 | any Contribution intentionally submitted for inclusion in the Work 132 | by You to the Licensor shall be under the terms and conditions of 133 | this License, without any additional terms or conditions. 134 | Notwithstanding the above, nothing herein shall supersede or modify 135 | the terms of any separate license agreement you may have executed 136 | with Licensor regarding such Contributions. 137 | 138 | 6. Trademarks. This License does not grant permission to use the trade 139 | names, trademarks, service marks, or product names of the Licensor, 140 | except as required for reasonable and customary use in describing the 141 | origin of the Work and reproducing the content of the NOTICE file. 142 | 143 | 7. Disclaimer of Warranty. Unless required by applicable law or 144 | agreed to in writing, Licensor provides the Work (and each 145 | Contributor provides its Contributions) on an "AS IS" BASIS, 146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or 147 | implied, including, without limitation, any warranties or conditions 148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A 149 | PARTICULAR PURPOSE. You are solely responsible for determining the 150 | appropriateness of using or redistributing the Work and assume any 151 | risks associated with Your exercise of permissions under this License. 152 | 153 | 8. Limitation of Liability. In no event and under no legal theory, 154 | whether in tort (including negligence), contract, or otherwise, 155 | unless required by applicable law (such as deliberate and grossly 156 | negligent acts) or agreed to in writing, shall any Contributor be 157 | liable to You for damages, including any direct, indirect, special, 158 | incidental, or consequential damages of any character arising as a 159 | result of this License or out of the use or inability to use the 160 | Work (including but not limited to damages for loss of goodwill, 161 | work stoppage, computer failure or malfunction, or any and all 162 | other commercial damages or losses), even if such Contributor 163 | has been advised of the possibility of such damages. 164 | 165 | 9. Accepting Warranty or Additional Liability. While redistributing 166 | the Work or Derivative Works thereof, You may choose to offer, 167 | and charge a fee for, acceptance of support, warranty, indemnity, 168 | or other liability obligations and/or rights consistent with this 169 | License. However, in accepting such obligations, You may act only 170 | on Your own behalf and on Your sole responsibility, not on behalf 171 | of any other Contributor, and only if You agree to indemnify, 172 | defend, and hold each Contributor harmless for any liability 173 | incurred by, or claims asserted against, such Contributor by reason 174 | of your accepting any such warranty or additional liability. 175 | 176 | END OF TERMS AND CONDITIONS 177 | 178 | APPENDIX: How to apply the Apache License to your work. 179 | 180 | To apply the Apache License to your work, attach the following 181 | boilerplate notice, with the fields enclosed by brackets "[]" 182 | replaced with your own identifying information. (Don't include 183 | the brackets!) The text should be enclosed in the appropriate 184 | comment syntax for the file format. We also recommend that a 185 | file or class name and description of purpose be included on the 186 | same "printed page" as the copyright notice for easier 187 | identification within third-party archives. 188 | 189 | Copyright 2025 [Kadir BASOL] 190 | 191 | Licensed under the Apache License, Version 2.0 (the "License"); 192 | you may not use this file except in compliance with the License. 193 | You may obtain a copy of the License at 194 | 195 | http://www.apache.org/licenses/LICENSE-2.0 196 | 197 | Unless required by applicable law or agreed to in writing, software 198 | distributed under the License is distributed on an "AS IS" BASIS, 199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 200 | See the License for the specific language governing permissions and 201 | limitations under the License. 202 | 203 | Copyright 2016 Luca Martino. 204 | 205 | Licensed under the Apache License, Version 2.0 (the "License"); 206 | you may not use this file except in compliance with the License. 207 | You may obtain a copyFile of the License at 208 | 209 | http://www.apache.org/licenses/LICENSE-2.0 210 | 211 | Unless required by applicable law or agreed to in writing, software 212 | distributed under the License is distributed on an "AS IS" BASIS, 213 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 214 | See the License for the specific language governing permissions and 215 | limitations under the License. --------------------------------------------------------------------------------