Cagen CSV JDBC Driver v4.2 Documentation

Index:

1. Setting the CLASSPATH
2. Loading the Driver
3. Connecting to the Database
4. Supported SQL Syntax
5. Supported SQL Function

Setting the CLASSPATH

To put csvdriver.jar into your class path, you should use "export CLASSPATH=/usr/share/lib/csvdriver.jar:$CLASSPATH" on Solaris and Linux, and "SET CLASSPATH=\javalib\csvdriver.jar;%classpath%" on Windows.

Loading the Driver

Any source that uses JDBC needs to import the java.sql package by using " import java.sql.*;".

Caigen CSV driver' name is com.caigen.sql.text.CSVDriver, and you can uses it without involving hard coding the driver into your code. You do this by setting the jdbc.drivers system property. For example, for command line apps you can use:
java -Djdbc.drivers=com.caigen.sql.text.CSVDriver yourApp
Then, the JVM upon startup will load the drivers automatically. Some applications (JBoss, Tomcat etc) support a .properties file which they use to save putting this on the command line.

The second method is the most common and involves you loading the driver yourself. It's simple:
Class.forName("com.caigen.sql.text.CSVDriver");
From then on you can get connections from DriverManager.
Note: If Class.forName() throws ClassNotFoundException, you should check your classpath.

Data Source class is com.caigen.sql.CaigenDataSource, and Connection Pool Data Source class is com.caigen.sql.CaigenConnectionPoolDataSource.

Connecting to the Database

After the driver has been registered with the DriverManager, you can obtain a Connection instance that is connected to a particular database by calling DriverManager.getConnection(). With JDBC, a database is represented by a URL (Uniform Resource Locator).

        Embedded:
                jdbc:csv:[//]/[DatabasePath][?prop1=value1[;prop2=value2]] (You can omit that "//" characters sometimes)
            or
                jdbc:%csv:[//]/[DatabasePath][?prop1=value1[;prop2=value2]] (You can omit that "//" characters sometimes)
            The only difference for two kinds of url format is: with jdbc:csv prefix, the default table type for CREATE TABLE is CSV.
                        For example:
                                "jdbc:csv:/."
                                "jdbc:csv:/c:/data" for Windows driver
                                "jdbc:csv:///c:/data" for Windows driver
                                "jdbc:csv:////usr/data" for unix or linux
                                "jdbc:csv://///192.168.10.2/sharedir" for UNC path
                                "jdbc:csv:/./data"
                                "jdbc:csv:/."
                                "jdbc:csv:/c:/data"
                                "jdbc:csv:////usr/data" 
        Compressed Database:(.ZIP, .JAR, .GZ, .TAR, .BZ2, .TGZ, .TAR.GZ, .TAR.BZ2) 
                jdbc url format is the same as embedded url and remote url.
                        For example:
                                "jdbc:csv:/c:/test/testcsv.zip
        Memory-only Database:
                jdbc:csv:/_memory_/
        URL Database:(http protocol, https protocol, ftp protocol)
                jdbc:csv:http://httpURL
                jdbc:csv:https://httpsURL
                jdbc:csv:ftp://ftpURL
                        For example:
                                "jdbc:csv:http://www.caigen.com/test" //Note: FTP site's user/password should be set in ftpURL, and cannot be set in JDBC connection property because user/password JDBC connection property belongs to server/client connection.
        SAMBA Database:(smb protocol)
                                jdbc:csv:smb://[[[domain;]username[:password]@]server[:port]/[[share/[dir/]file]]][?[param=value]]
                        For example:
                                "jdbc:csv:smb://test1:123@100.100.13.94/csvfiles/zone" //Note: SAMBA user/password should be set in SMB url, and cannot be set in JDBC connection property because user/password JDBC connection property belongs to server/client connection.
	UNC path JDBC url:
                jdbc:csv:/uncpath
                jdbc:csv:///uncpath
                        For example:
                                "jdbc:csv:/\\PC17\c$\values" 
                                "jdbc:csv:/\\PC17\val"
	Free JDBC url:(Warning: only use it for special project)
                jdbc:csv:/" or "jdbc:csv:///". Then you can use some full UNC path names in SQL to visit anywhere where your Java VM has right to access.
                        For instance:
                                select * from \\amd2500\e$\csvfiles\test;
                                elect * from "\\amd2500\d$\csviles".test;
                                select * from ".".test;

         Caigen CSV supports seamlessly data mining on memory-only table, physical table, url table, compressed table, SAMBA table in a sql.

To connect, you need to get a Connection instance from JDBC. To do this, you use the DriverManager.getConnection() method:

Connection con = DriverManager.getConnection(url, properties);

There are a few different signatures for the getConnection() method. You should see the API documentation that comes with your JDK for more specific information on how to use them. You can specify additional properties to the JDBC driver by placing them in a java.util.Properties instance and passing that instance to the DriverManager when you connect.

            SQLWarning warnings=con.getWarnings();
            if(warnings!=null){
                do{//SQState: 2A000	Direct SQL syntax error or access rule violation
                    System.out.println(warnings.getMessage());
                    System.out.println("Error Code:"+warnings.getErrorCode());
                    System.out.println("SQL State:"+warnings.getSQLState());                   
                }while((warnings=warnings.getNextWarning())!=null);
                con.clearWarnings();                
            }
For CSV/TSV file, the preferable solution is using maxScanRows to detect automatically table information, not using CREATE TABLE sql.

                        warnings=rs.getWarnings();
                        if(warnings!=null){
                            do{//SQState: C0106	Convert dirty data into null value
                                System.out.println(warnings.getMessage());
                                System.out.println("Error Code:"+warnings.getErrorCode());
                                System.out.println("SQL State:"+warnings.getSQLState());
                            }while((warnings=warnings.getNextWarning())!=null);
                            rs.clearWarnings();
                        }

                        warnings=rs.getWarnings();
                        if(warnings!=null){
                            do{//SQState: C0106	Convert dirty data into null value
                                System.out.println(warnings.getMessage());
                                System.out.println("Error Code:"+warnings.getErrorCode());
                                System.out.println("SQL State:"+warnings.getSQLState());
                            }while((warnings=warnings.getNextWarning())!=null);
                            rs.clearWarnings();
                        }

When your code then tries to open a Connection, and you get a No driver available SQLException being thrown, this is probably caused by the driver not being in the class path, or the JDBC url not being correct.

To close the database connection, simply call the close() method to the Connection:

con.close();