Multi Language Program Support

cli/pom.xml

@@ -118,6 +118,11 @@
             <artifactId>llvmir</artifactId>
             <version>${revision}</version>
         </dependency>
+        <dependency>
+            <groupId>de.jplag</groupId>
+            <artifactId>multi-language</artifactId>
+            <version>${revision}</version>
+        </dependency>
         <!-- CLI -->
         <dependency>
             <groupId>org.kohsuke.metainf-services</groupId>

cli/src/main/java/de/jplag/cli/options/LanguageCandidates.java

@@ -2,6 +2,8 @@
 
 import java.util.ArrayList;
 
+import de.jplag.LanguageLoader;
+
 /**
  * Helper class for picocli to find all available languages.
  */

cli/src/main/java/de/jplag/cli/options/LanguageConverter.java

@@ -1,6 +1,7 @@
 package de.jplag.cli.options;
 
 import de.jplag.Language;
+import de.jplag.LanguageLoader;
 
 import picocli.CommandLine;
 

cli/src/main/java/de/jplag/cli/picocli/CliInputHandler.java

@@ -14,9 +14,9 @@
 import java.util.stream.Collectors;
 
 import de.jplag.Language;
+import de.jplag.LanguageLoader;
 import de.jplag.cli.CliException;
 import de.jplag.cli.options.CliOptions;
-import de.jplag.cli.options.LanguageLoader;
 import de.jplag.options.LanguageOption;
 import de.jplag.options.LanguageOptions;
 

cli/src/test/java/de/jplag/cli/LanguageTest.java

@@ -13,8 +13,8 @@
 import org.junit.jupiter.params.provider.MethodSource;
 
 import de.jplag.Language;
+import de.jplag.LanguageLoader;
 import de.jplag.cli.options.CliOptions;
-import de.jplag.cli.options.LanguageLoader;
 import de.jplag.cli.test.CliArgument;
 import de.jplag.cli.test.CliTest;
 import de.jplag.exceptions.ExitException;

endtoend-testing/src/main/java/de/jplag/endtoend/helper/LanguageDeserializer.java

@@ -3,7 +3,7 @@
 import java.io.IOException;
 
 import de.jplag.Language;
-import de.jplag.cli.options.LanguageLoader;
+import de.jplag.LanguageLoader;
 
 import com.fasterxml.jackson.core.JsonParser;
 import com.fasterxml.jackson.databind.DeserializationContext;

cli/src/main/java/de/jplag/cli/options/LanguageLoader.java → language-api/src/main/java/de/jplag/LanguageLoader.java

@@ -1,4 +1,4 @@
-package de.jplag.cli.options;
+package de.jplag;
 
 import java.util.Collections;
 import java.util.Map;
@@ -11,8 +11,6 @@
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
-import de.jplag.Language;
-
 /**
  * This class contains methods to load {@link Language Languages}.
  * @author Dominik Fuchss

language-api/src/main/java/de/jplag/options/LanguageOptions.java

@@ -56,7 +56,7 @@ protected <T> LanguageOption<T> createDefaultOption(OptionType<T> type, String n
      * @return The new option
      */
     protected <T> LanguageOption<T> createOption(OptionType<T> type, String name, String description) {
-        LanguageOption<T> option = new DefaultLanguageOption<>(type, name, description);
+        LanguageOption<T> option = new DefaultLanguageOption<>(type, description, name);
         this.options.add(option);
         return option;
     }

languages/multi-language/pom.xml

@@ -0,0 +1,26 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
+    <modelVersion>4.0.0</modelVersion>
+    <parent>
+        <groupId>de.jplag</groupId>
+        <artifactId>languages</artifactId>
+        <version>${revision}</version>
+    </parent>
+    <artifactId>multi-language</artifactId>
+
+    <dependencies>
+        <dependency>
+            <groupId>de.jplag</groupId>
+            <artifactId>java</artifactId>
+            <version>${revision}</version>
+            <scope>test</scope>
+        </dependency>
+        <dependency>
+            <groupId>de.jplag</groupId>
+            <artifactId>cpp</artifactId>
+            <version>${revision}</version>
+            <scope>test</scope>
+        </dependency>
+    </dependencies>
+
+</project>

languages/multi-language/src/main/java/de/jplag/multilang/MultiLanguage.java

@@ -0,0 +1,55 @@
+package de.jplag.multilang;
+
+import java.io.File;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Set;
+
+import org.kohsuke.MetaInfServices;
+
+import de.jplag.Language;
+import de.jplag.LanguageLoader;
+import de.jplag.ParsingException;
+import de.jplag.Token;
+import de.jplag.options.LanguageOptions;
+
+@MetaInfServices(Language.class)
+public class MultiLanguage implements Language {
+    private final MultiLanguageOptions options;
+
+    public MultiLanguage() {
+        this.options = new MultiLanguageOptions();
+    }
+
+    @Override
+    public String[] suffixes() {
+        return LanguageLoader.getAllAvailableLanguages().values().stream().filter(it -> !(it == this)).flatMap(it -> Arrays.stream(it.suffixes()))
+                .toArray(String[]::new);
+    }
+
+    @Override
+    public String getName() {
+        return "multi-language";
+    }
+
+    @Override
+    public String getIdentifier() {
+        return "multi";
+    }
+
+    @Override
+    public int minimumTokenMatch() {
+        return this.options.getLanguages().stream().mapToInt(Language::minimumTokenMatch).min().orElse(9);
+    }
+
+    @Override
+    public List<Token> parse(Set<File> files, boolean normalize) throws ParsingException {
+        MultiLanguageParser parser = new MultiLanguageParser(this.options);
+        return parser.parseFiles(files, normalize);
+    }
+
+    @Override
+    public LanguageOptions getOptions() {
+        return this.options;
+    }
+}

languages/multi-language/src/main/java/de/jplag/multilang/MultiLanguageOptions.java

@@ -0,0 +1,40 @@
+package de.jplag.multilang;
+
+import java.util.Arrays;
+import java.util.List;
+
+import de.jplag.Language;
+import de.jplag.LanguageLoader;
+import de.jplag.options.LanguageOption;
+import de.jplag.options.LanguageOptions;
+import de.jplag.options.OptionType;
+
+public class MultiLanguageOptions extends LanguageOptions {
+    private static final String ERROR_LANGUAGE_NOT_FOUND = "The selected language %s could not be found";
+    private static final String ERROR_NOT_ENOUGH_LANGUAGES = "To use multi language specify at least 1 language";
+
+    public LanguageOption<String> languageNames = createOption(OptionType.string(), "languages",
+            "The languages that should be used. This is a ',' separated list");
+    private List<Language> languages = null;
+
+    public List<Language> getLanguages() {
+        if (this.languages == null) {
+            if (languageNames.getValue() == null) {
+                throw new IllegalArgumentException(ERROR_NOT_ENOUGH_LANGUAGES);
+            }
+
+            List<Language> languages = Arrays.stream(languageNames.getValue().split(","))
+                    .map(name -> LanguageLoader.getLanguage(name)
+                            .orElseThrow(() -> new IllegalArgumentException(String.format(ERROR_LANGUAGE_NOT_FOUND, name))))
+                    .filter(language -> !language.getClass().equals(MultiLanguage.class)).toList();
+
+            if (languages.isEmpty()) {
+                throw new IllegalArgumentException(ERROR_NOT_ENOUGH_LANGUAGES);
+            }
+
+            this.languages = languages;
+        }
+
+        return this.languages;
+    }
+}

languages/multi-language/src/main/java/de/jplag/multilang/MultiLanguageParser.java

@@ -0,0 +1,36 @@
+package de.jplag.multilang;
+
+import java.io.File;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Optional;
+import java.util.Set;
+
+import de.jplag.Language;
+import de.jplag.ParsingException;
+import de.jplag.Token;
+
+public class MultiLanguageParser {
+    private final List<Language> languages;
+
+    public MultiLanguageParser(MultiLanguageOptions options) {
+        this.languages = options.getLanguages();
+    }
+
+    public List<Token> parseFiles(Set<File> files, boolean normalize) throws ParsingException {
+        List<Token> results = new ArrayList<>();
+        for (File file : files) {
+            Optional<Language> language = findLanguageForFile(file);
+            if (language.isPresent()) {
+                results.addAll(language.get().parse(Set.of(file), normalize));
+            }
+        }
+        return results;
+    }
+
+    private Optional<Language> findLanguageForFile(File file) {
+        return this.languages.stream().filter(language -> Arrays.stream(language.suffixes()).anyMatch(suffix -> file.getName().endsWith(suffix)))
+                .findFirst();
+    }
+}

languages/multi-language/src/test/java/de/java/multilang/MultilangTest.java

@@ -0,0 +1,79 @@
+package de.java.multilang;
+
+import static de.jplag.SharedTokenType.FILE_END;
+
+import java.io.File;
+import java.io.FileOutputStream;
+import java.io.IOException;
+import java.nio.file.Files;
+import java.util.List;
+import java.util.Set;
+import java.util.TreeSet;
+
+import org.junit.jupiter.api.AfterAll;
+import org.junit.jupiter.api.Assertions;
+import org.junit.jupiter.api.BeforeAll;
+import org.junit.jupiter.api.Test;
+
+import de.jplag.ParsingException;
+import de.jplag.Token;
+import de.jplag.TokenType;
+import de.jplag.cpp.CPPTokenType;
+import de.jplag.java.JavaTokenType;
+import de.jplag.multilang.MultiLanguage;
+import de.jplag.multilang.MultiLanguageOptions;
+
+public class MultilangTest {
+    private static File testDataDirectory;
+    private static File javaCode;
+    private static File cppCode;
+
+    private static List<TokenType> expectedTokens = List.of(CPPTokenType.FUNCTION_BEGIN, CPPTokenType.RETURN, CPPTokenType.FUNCTION_END, FILE_END,
+            JavaTokenType.J_CLASS_BEGIN, JavaTokenType.J_CLASS_END, FILE_END);
+
+    @BeforeAll
+    static void setUp() throws IOException {
+        testDataDirectory = Files.createTempDirectory("multiLanguageTestData").toFile();
+        cppCode = new File(testDataDirectory, "CppCode.cpp");
+        javaCode = new File(testDataDirectory, "JavaCode.java");
+
+        MultilangTest.class.getResourceAsStream("/de/jplag/multilang/testDataSet/CppCode.cpp").transferTo(new FileOutputStream(cppCode));
+        MultilangTest.class.getResourceAsStream("/de/jplag/multilang/testDataSet/JavaCode.java").transferTo(new FileOutputStream(javaCode));
+    }
+
+    @Test
+    void testMultiLanguageParsing() throws ParsingException {
+        MultiLanguage languageModule = new MultiLanguage();
+        ((MultiLanguageOptions) languageModule.getOptions()).languageNames.setValue("java,cpp");
+
+        Set<File> sources = new TreeSet<>(List.of(javaCode, cppCode)); // Using TreeSet to ensure order of entries
+        List<Token> tokens = languageModule.parse(sources, false);
+
+        Assertions.assertEquals(expectedTokens, tokens.stream().map(Token::getType).toList());
+    }
+
+    @Test
+    void testNoLanguagesConfigured() {
+        MultiLanguage languageModule = new MultiLanguage();
+        Assertions.assertThrowsExactly(IllegalArgumentException.class, () -> {
+            languageModule.parse(Set.of(javaCode, cppCode), false);
+        });
+    }
+
+    @Test
+    void testInvalidLanguage() {
+        MultiLanguage languageModule = new MultiLanguage();
+        ((MultiLanguageOptions) languageModule.getOptions()).languageNames.setValue("thisIsNotALanguage");
+
+        Assertions.assertThrowsExactly(IllegalArgumentException.class, () -> {
+            languageModule.parse(Set.of(javaCode, cppCode), false);
+        });
+    }
+
+    @AfterAll
+    static void cleanUp() {
+        javaCode.delete();
+        cppCode.delete();
+        testDataDirectory.delete();
+    }
+}

languages/multi-language/src/test/resources/de/jplag/multilang/testDataSet/CppCode.cpp

@@ -0,0 +1,3 @@
+int main() {
+    return 0;
+}

languages/multi-language/src/test/resources/de/jplag/multilang/testDataSet/JavaCode.java

@@ -0,0 +1,3 @@
+public class JavaCode {
+
+}

languages/pom.xml

@@ -29,6 +29,7 @@
         <module>typescript</module>
         <module>javascript</module>
         <module>llvmir</module>
+        <module>multi-language</module>
     </modules>
     <dependencies>
         <dependency>