db2arff.java

weka机器学习系统（本站可下载）的拓展

package com.prcomps.cahitarf;

import java.io.File;
import java.io.FilenameFilter;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.PrintStream;
import java.io.FileOutputStream;
import java.io.ByteArrayOutputStream;
import java.io.FileNotFoundException;
import java.net.URL;
import java.net.URLClassLoader;
import java.net.MalformedURLException;
import java.util.Properties;
import java.util.ResourceBundle;
import java.sql.ResultSet;
import java.sql.SQLException;
import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.Statement;
import java.sql.Driver;
import java.text.MessageFormat;

/**
 */
public class Db2Arff
{

    public static final String PROP_JDBC_DRIVER   = "jdbc.driver";
    public static final String PROP_JDBC_USER     = "jdbc.user";
    public static final String PROP_JDBC_PASSWORD = "jdbc.password";
    public static final String PROP_JDBC_URL      = "jdbc.url";
    public static final String PROP_JDBC_SELECT   = "jdbc.select";

    public static final String PROP_FIELD_        = "attr.";
    public static final String PROP_RELATION      = "relation";

    private static ResourceBundle bundle = ResourceBundle.getBundle("messages");

    private Properties properties;
    private Configuration conf = new Configuration();
    private Connection conn = null;
    private int bufferSize = 2 * 1024 * 1024;
    private int maxRows = 0;

    public Db2Arff()
    {
    }


    /**
     * Scans ./lib directory and loads JDBC drivers to context classpath
     * @throws FileNotFoundException
     */
    public void loadDrivers()
        throws FileNotFoundException
    {
        // absolute file system path of ./lib directory
        String libPath = getAppPath();
        if ( libPath == null )
            libPath = System.getProperty( "user.dir" );
        libPath += "/lib";

        // check if './lib' exists
        File libDir = new File( libPath );
        if ( ! libDir.exists() || ! libDir.isDirectory() )
            throw new FileNotFoundException(
                    MessageFormat.format( bundle.getString( "db2arff.libdir.notfound" ), new Object[] { libPath } ) );

        // Get file list of .jar and .zip archives located in
        File libFiles[] = libDir.listFiles( new FilenameFilter() {

            public boolean accept( File f, String name )
            {
                // TODO: themes.jar, skinlf.jar, easylayout.jar ?
                return name.endsWith(".jar") || name.endsWith( ".zip" );
            }

        });

        // convert file list to URL list
        URL libURLs[] = new URL[ libFiles.length ];
        for ( int i=0; i<libFiles.length; i++ )
            try
            {
                libURLs[i] = libFiles[i].toURL();
            }
            catch ( MalformedURLException e )
            {
                throw new FileNotFoundException( e.getMessage() );
            }

        // Load JDBC classes
        URLClassLoader loader = new URLClassLoader( libURLs, getClass().getClassLoader() );
        Thread.currentThread().setContextClassLoader( loader );

    }

    /**
     *
     * @param path
     * @throws IOException
     */
    public void loadProperties( String path )
            throws IOException
    {
        File confFile = new File( path );

        properties = new Properties();
        FileInputStream in = new FileInputStream( confFile );
        properties.load( in );
        in.close();

        conf = new Configuration();

    }

    /**
     *
     * @param out
     * @param fields
     */
    private void createArffHeader( PrintStream out, Configuration.Field[] fields )
    {
        java.text.SimpleDateFormat df = new java.text.SimpleDateFormat( "dd MMM yyyy HH:mm:ss" );
        out.print( "% Created by Cahit Arf at " );
        out.print( df.format( new java.util.Date() ) );
        out.print( "\n\n" );

        out.print( "@RELATION " );
        String relation = properties.getProperty( "relation", "CahitArf" ).trim();
        if ( relation.indexOf( ' ' ) >= 0
             && (
                    ( !relation.startsWith( "\"" ) && !relation.endsWith( "\"" ) ) ||
                    ( !relation.endsWith( "'") && !relation.startsWith( "'" ) )
                )
        )
            relation = "'" + relation + "'";
        out.print( relation );
        out.print( "\n\n" );
        for ( int i=0; i<fields.length; i++ )
            out.print( fields[i].toDefinitionString() );
        out.print( "\n\n@DATA\n" );

    }

    /**
     * A wrapper to toArff( PrintSream out ).  Creates a new file pointed by the path parameter, creates
     * a PrintStream to the file, and then calls the main toArff() method.
     *
     * @param path Output file path.
     */
    public void toArff( String path )
    {
        File arff = new File( path );
        try
        {
            arff.createNewFile();
            PrintStream out = new PrintStream( new FileOutputStream( arff ) );
            toArff( out );
            out.close();
        }
        catch ( IOException e )
        {
            e.printStackTrace();  //To change body of catch statement use Options | File Templates.
        }
    }

    /**
     *
     * @param out
     */
    public void toArff( PrintStream out )
    {
        PrintStream tmp = null;
        ByteArrayOutputStream bout = null;
        Configuration.Field[] fields = null;

        try
        {
            fields = conf.createFields( properties );
            ResultSet rs = getQueryResult();

            if ( fields.length == 0 )
                fields = conf.createFields( rs );
            if ( fields.length != rs.getMetaData().getColumnCount() )
            {
                // TODO
            }

            if ( conf.isIncludesClass() )
            {
                tmp = out;
                bout = new ByteArrayOutputStream( bufferSize );
                out = new PrintStream( bout );
            }
            else
                createArffHeader( out, fields );

            conf.setOut( out );
            int len = fields.length - 1;
            while ( rs.next() )
            {
                for ( int i=0; i<len; i++ )
                {
                    fields[i].toData( rs.getString( i+1 ) );
                    out.print( ',' );
                }
                fields[len].toData( rs.getString( len+1 ) );
                out.print( '\n' );
            }
            rs.close();
            release();
        }
        catch ( SQLException e )
        {
            e.printStackTrace();  //To change body of catch statement use Options | File Templates.
        }

        if ( conf.isIncludesClass() )
        {
            try
            {
                createArffHeader( tmp, fields );
                bout.writeTo( tmp );
                out = null;
            }
            catch ( IOException e )
            {
                e.printStackTrace();  //To change body of catch statement use Options | File Templates.
            }
        }
    }

    /**
     *
     * @return
     * @throws SQLException
     * @throws ClassNotFoundException
     */
    public Connection connect()
        throws SQLException, ClassNotFoundException
    {
        if ( conn == null )
            try
            {
                Driver driver = (Driver)Class.forName(
                                 properties.getProperty( PROP_JDBC_DRIVER ),
                                 false,
                                 Thread.currentThread().getContextClassLoader() ).newInstance();
                //DriverManager.registerDriver( driver );
                Properties connInfo = new Properties();
                if ( properties.getProperty( PROP_JDBC_USER ) != null )
                    connInfo.setProperty( "user", properties.getProperty( PROP_JDBC_USER ) );
                if ( properties.getProperty( PROP_JDBC_PASSWORD ) != null )
                    connInfo.setProperty( "password", properties.getProperty( PROP_JDBC_PASSWORD ) );
                conn = driver.connect( properties.getProperty( PROP_JDBC_URL ), connInfo );
            }
            catch ( InstantiationException e )
            {
                throw new ClassNotFoundException( e.getMessage() );
            }
            catch ( IllegalAccessException e )
            {
                throw new ClassNotFoundException( e.getMessage() );
            }
        return conn;
    }

    /**
     *
     * @throws SQLException
     */
    public void release()
        throws SQLException
    {
        if ( conn == null )
            return;
        conn.close();
    }

    /**
     *
     * @return
     * @throws SQLException
     */
    public ResultSet getQueryResult()
        throws SQLException
    {
        if ( conn == null )
            try
            {
                connect();
            }
            catch ( ClassNotFoundException e )
            {
                throw new SQLException( e.getMessage() );
            }
        String select = properties.getProperty( PROP_JDBC_SELECT, "" );
        // TODO: put variables / V1.1
        Statement stmt = conn.createStatement();
        stmt.setMaxRows( maxRows );

        return stmt.executeQuery( select );
    }

    /**
     *
     * @return
     */
    public String getAppPath()
    {
        URL url = getClass().getClassLoader().getResource( "com/prcomps/cahitarf" );
        if ( url == null )
            return null;
        String surl = url.toString();
        if ( surl.startsWith( "jar:" ) )
        {
            surl = surl.substring( "jar:file:".length(), surl.indexOf('!') );
            surl = surl.substring( 0, surl.lastIndexOf( '/' ) );
        }
        else if ( surl.startsWith( "file:" ) )
        {
            surl = surl.substring( "file:".length(), surl.lastIndexOf( "/com/prcomps/cahitarf") );
        }
        else
            return null;

        int spc = surl.indexOf( "%20" );
        if ( spc > -1 )
        {
            StringBuffer str = new StringBuffer( surl.length() );
            int pos = 0;
            while ( spc > -1 )
            {
                str.append( surl.substring( pos, spc ) ).append( ' ' );
                pos = spc + 3;
                spc = surl.indexOf( "%20", pos );
            }
            str.append( surl.substring( pos ) );
            surl = str.toString();
        }

        return surl;
    }

    /**
     *
     * @return
     */
    public Properties getProperties()                   { return properties; }

    /**
     *
     * @param properties
     */
    public void setProperties( Properties properties )
    {
        this.properties = properties;
    }

    /**
     *
     * @return
     */
    public int getMaxRows()                             { return maxRows; }

    /**
     *
     * @param maxRows
     */
    public void setMaxRows( int maxRows )               { this.maxRows = maxRows; }

    /**
     *
     * @param args
     */
    public static void main( String args[] )
    {
        if ( args.length < 2 )
        {
            System.err.println( bundle.getString( "db2arff.usage" ) );
            System.exit( 1 );
        }

        Db2Arff app = new Db2Arff();
        try
        {
            app.loadProperties( args[0] );
        }
        catch ( IOException e )
        {
            System.err.println( MessageFormat.format( bundle.getString("db2arff.conf.notfound"), args ) );
            System.exit( 1 );
        }

        try
        {
            app.loadDrivers();
        }
        catch ( FileNotFoundException e )
        {
            System.err.println( e.getMessage() );
            System.exit( 1 );
        }
        app.toArff( args[1] );

    }
}