Commit 5186af6c authored by Manuel  Segimon's avatar Manuel Segimon
Browse files

work in progress

parent a652f5a9
Loading
Loading
Loading
Loading
+35 −0
Original line number Diff line number Diff line
@@ -14,6 +14,23 @@
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    </properties>

    <build>
        <plugins>
            <plugin>
                <artifactId>maven-jar-plugin</artifactId>
                <version>3.2.0</version>
                <configuration>
                    <archive>
                        <manifest>
                            <addClasspath>true</addClasspath>
                            <mainClass>edu.bu.LanguageCorrection.MainApp</mainClass>  <!-- Replace with your main class -->
                        </manifest>
                    </archive>
                </configuration>
            </plugin>
        </plugins>
    </build>

    <dependencies>
        <dependency>
            <groupId>org.jsoup</groupId>
@@ -28,4 +45,22 @@
      </dependency>
    </dependencies>

  <profiles>
    <profile>
        <id>large-memory</id>
        <build>
            <plugins>
                <plugin>
                    <groupId>org.apache.maven.plugins</groupId>
                    <artifactId>maven-surefire-plugin</artifactId>
                    <version>3.0.0-M5</version>
                    <configuration>
                        <argLine>-Xmx100g</argLine>
                    </configuration>
                </plugin>
            </plugins>
        </build>
    </profile>
</profiles>

</project>
 No newline at end of file
+4 −4
Original line number Diff line number Diff line
@@ -146,12 +146,12 @@ public class crawler {
        String corpus = "";
        if (language.equals("en")) {
            corpus = "brown.txt";
        } else if (language.equals("es")) {
            corpus = "es.txt";
        } else if (language.equals("gr")) {
            corpus = "german.txt";
        } else if (language.equals("it")) {
            corpus = "it.txt";
            corpus = "italian.txt";
        } else if (language.equals("pt")) {
            corpus = "pt.txt";
            corpus = "portuguese.txt";
        } else {
            System.err.println("Unsupported language: " + language);
            return;