Skip Headers
Oracle® Database SQLJ Developer's Guide and Reference
10g Release 2 (10.2)

Part Number B16018-01
Go to Documentation Home
Home
Go to Book List
Book List
Go to Table of Contents
Contents
Go to Index
Index
Go to Master Index
Master Index
Go to Feedback page
Feedback

Go to previous page
Previous
Go to next page
Next
View PDF

9 Translator and Run Time Functionality

This chapter discusses internal operations and functionality of the Oracle SQLJ translator and run time.

The following topics are covered:

9.1 Internal Translator Operations

The following topics summarize the operations executed by the SQLJ translator during a translation:

9.1.1 Java and SQLJ Code-Parsing and Syntax-Checking

In this first phase of SQLJ translation, a SQLJ parser and a Java parser are used to process all the source code and check syntax. As the SQLJ translator parses the .sqlj file, it invokes a Java parser to check the syntax of Java statements and a SQLJ parser to check the syntax of SQLJ constructs (anything preceded by #sql). The SQLJ parser also invokes the Java parser to check the syntax of Java host variables and expressions within SQLJ executable statements.

The SQLJ parser checks the grammar of SQLJ constructs according to the SQLJ language specification. However, it does not check the grammar of the embedded SQL operations. SQL syntax is not checked until the semantics-checking or offline parsing step.

This syntax-check will look for errors like missing semi-colons, mismatched curly braces, and obvious type mismatches, such as multiplying a number by a string. If the parsers find any syntax errors or type mismatches during this phase, then the translation is aborted and the errors are reported to the user.

9.1.2 SQL Semantics-Checking and Offline Parsing

Once the SQLJ and Java application source code is verified as syntactically correct, the translator enters into the semantics-checking phase and invokes a SQL semantics-checker or a SQL offline parser or both, according to SQLJ option settings.

Setting the -user option enables online checking, and setting the -password and -url options specifies the database connection, if the password and URL were not specified in the -user option. The -offline or -online option specifies which checker to use. The default, typically sufficient, is a checker front end called OracleChecker that chooses the most appropriate checker, according to whether you have enabled online checking and which Java Database Connectivity (JDBC) driver you are using.

The -parse option, true by default, is for enabling the offline parser, which offers a way to verify SQL and PL/SQL syntax (but not data types against database columns) without necessitating a database connection during translation. Note that some settings of the -parse option will override the -user option and disable online checking.


Note:

For ISO code generation, semantics-checking can also be performed on a profile that was produced during a previous execution of the SQLJ translator. Refer to "SQLCheckerCustomizer for Profile Semantics-Checking".

The following tasks are always performed during semantics-checking, regardless of the status of online checking or offline parsing:

  1. SQLJ analyzes the types of Java expressions in your SQLJ executable statements.

    This includes examining the SQLJ source files being translated, any .java files entered on the command line, and any imported Java classes whose .class files or .java files can be found through the classpath. SQLJ examines whether and how stream types are used in SELECT or CAST statements, what Java types are used in iterator columns or INTO-lists, what Java types are used as input host variables, and what Java types are used as output host variables.

    SQLJ also processes FETCH, CAST, CALL, SET TRANSACTION, VALUES, and SET statements syntactically.

    Any Java expression in a SQLJ executable statement must have a Java type valid for the given situation and usage. For example, consider the following statement:

    #sql [myCtx] { UPDATE ... };
    
    

    The myCtx variable, which might be used to specify a connection context instance or execution context instance for this statement, must actually resolve to a SQLJ connection context type or execution context type.

    Now consider the following example:

    #sql { UPDATE emp SET sal = :newSal };
    
    

    If newSal is a variable, as opposed to a field, then an error is generated if newSal was not previously declared. In any case, an error is generated if it cannot be assigned to a valid Java type or its Java type cannot be used in a SQL statement (for example, java.util.Vector).


    Note:

    Semantics-checking of Java types is performed only for Java expressions within SQLJ executable statements. Such errors in your standard Java statements will not be detected until compilation by the Java compiler.

  2. SQLJ tries to categorize your embedded SQL operations. Each operation must have a recognizable keyword, such as SELECT or INSERT, so that SQLJ knows what kind of operation it is. For example, the following statement will generate an error:

    #sql { foo };
    
    
  3. If either online checking or offline parsing (or both) is enabled, then SQLJ analyzes and verifies the syntax of embedded SQL and PL/SQL operations.

  4. If either online checking or offline parsing (or both) is enabled, then SQLJ checks the types of Java expressions in SQLJ executable statements against SQL types of corresponding columns in the database and SQL types of corresponding arguments and return variables of stored procedures and functions.

    In the process of doing this, SQLJ verifies that the SQL entities used in your SQLJ executable statements, such as tables, views, and stored procedures, actually exist in the database. SQLJ also checks nullability of database columns whose data is being selected into iterator columns of Java primitive types, which cannot process null data. However, nullability is not checked for stored procedure and function output parameters and return values.

9.1.3 Code Generation

For the .sqlj application source file, the SQLJ translator generates a .java file and, for ISO standard SQLJ code generation, at least one profile (either in .ser or .class files). The .java contains your translated application source code, class definitions for any private iterators and connection contexts you declared, and, for ISO code, a profile-keys class definition generated and used internally by SQLJ.


Note:

No profiles or profile-keys class are generated if you use the default Oracle-specific code generation mode.

With ISO code generation, there are no profiles or profile-keys class if you do not use any SQLJ executable statements in your code.


Generated Application Code in .java File

For the default Oracle-specific code generation, the generated .java file for your application contains direct calls to the Oracle JDBC driver in place of the original SQLJ executable statements. There are also calls to an Oracle-specific SQLJ run time. For ISO standard SQLJ code generation, SQLJ executable statements are replaced by calls to the SQLJ run time, which in turn contains calls to the JDBC driver.

For convenience, generated .java files also include a comment for each of your #sql statements, repeating the statement in its entirety for reference. The generated .java file will have the same base name as the input .sqlj file, which would be the name of the public class defined in the .sqlj file or the first class defined if there are no public classes. For example, Foo.sqlj defines the Foo class. The Foo.java source file will be generated by the translator.

The location of the generated .java file depends on whether and how you set the SQLJ -dir option. By default, the .java file will be placed in the directory of the .sqlj input file.

Generated Profile-Keys Class in .java File (ISO Code Generation)

If you use ISO standard SQLJ code generation, SQLJ generates a profile-keys class that it uses internally during run time to load and access the serialized profile. This class contains mapping information between the SQLJ run time calls in your translated application and the SQL operations placed in the serialized profile. It also contains methods to access the serialized profile.


Note:

If you use the default Oracle-specific code generation, no profiles or profile-keys classes are generated.

The profile-keys class is defined in the same .java output file that has your translated application source code, with a class name based on the base name of your .sqlj source file as follows:

Basename_SJProfileKeys

For example, translating Foo.sqlj defines the following profile-keys class in the generated .java file:

Foo_SJProfileKeys

If your application is in a package, this is reflected appropriately. For example, translating Foo.sqlj in the a.b package defines the following class:

a.b.Foo_SJProfileKeys

Generated Profiles in .ser or .class Files (ISO Code Generation)

If you use ISO standard SQLJ code generation, SQLJ generates profiles that it uses to store information about the SQL operations found in the input file. A profile is generated for each connection context class that you use in your application. It describes the operations to be performed using instances of the associated connection context class, such as SQL operations to execute, tables to access, and stored procedures and functions to call.


Note:

If you use the default Oracle-specific code generation, then information about the SQL operations is embedded in the generated code, which calls the Oracle JDBC driver directly. In this case, SQLJ does not generate profiles.

Profiles are generated in .ser serialized resource files. However, if you enable the SQLJ -ser2class option, then the profiles are automatically converted to .class files as part of the translation. In this case, no further customization of the profile is possible. You would have to delete the .class file and rerun the SQLJ translator to regenerate the profile.

Profile base names are generated similarly to the profile-keys class name. They are fully qualified with the package name, followed by the .sqlj file base name, followed by the string:

_SJProfilen

Where n is a unique number, starting with 0, for each profile generated for a particular .sqlj input file.

Again using the example of the Foo.sqlj input file, if two profiles are generated, then they will have the following base names (presuming no package):

Foo_SJProfile0
Foo_SJProfile1

If Foo.sqlj is in the a.b package, then the profile base names will be:

a.b.Foo_SJProfile0
a.b.Foo_SJProfile1

Physically, a profile exists as a Java serialized object contained within a resource file. Resource files containing profiles use the .ser extension and are named according to the base name of the profile (excluding package names). Resource files for the two previously mentioned profiles will be named as follows:

Foo_SJProfile0.ser
Foo_SJProfile1.ser

Or, they will be named Foo_SJProfile0.class and Foo_SJProfile1.class if you enable the -ser2class option. If you choose this option, then the conversion to .class takes place after the customization step.

The location of these files depends on how the SQLJ -d option is set, which determines where all generated .ser and .class files are placed.

More About Generated Calls to SQLJ Run Time

When #sql statements are replaced by calls to the JDBC driver (for Oracle-specific code generation) or to the SQLJ run time (for ISO standard SQLJ code generation), these calls implement the steps in Table 9-1.

Table 9-1 Steps for Generated Calls, ISO Standard Versus Oracle-Specific

Steps for ISO Standard Code Generation Steps for Oracle Code Generation
Get a SQLJ statement object, using information stored in the associated profile entry. Get an Oracle JDBC statement object.
Bind inputs into the statement, using setXXX() methods of the statement object. Bind inputs using Oracle JDBC statement methods and, if necessary, register output parameters.
Execute the statement, using the executeUpdate() or executeQuery() method of the statement object. Execute the Oracle statement.
Create iterator instances, if applicable. Create iterator instances, if applicable.
Retrieve outputs from the statement, using getXXX() methods of the statement object. Retrieve outputs from the statement using appropriate JDBC getter methods.
Close the SQLJ statement object (by default, recycling it through the SQLJ statement cache). Close the JDBC statement object (by default, recycling it through the JDBC statement cache).

A SQLJ run time uses SQLJ statement objects that are similar to JDBC statement objects, although a particular implementation of SQLJ might or might not use JDBC statement classes directly. SQLJ statement classes add functionality particular to SQLJ. For example:

  • Standard SQLJ statement objects raise a SQL exception if a null value from the database is to be output to a primitive Java type, such as int or float, which cannot take null values.

  • Oracle SQLJ statement objects allow user-defined object and collection types to be passed to or retrieved from an Oracle database.

9.1.4 Java Compilation

After code generation, SQLJ invokes the Java compiler to compile the generated .java file. This produces a .class file for each class you defined in your application, including iterator and connection context declarations, as well as a .class file for the generated profile-keys class if you use ISO code generation (and presuming your application uses SQLJ executable statements). Any .java files you specified directly on the SQLJ command line (for type-resolution, for example) are compiled at this time as well.

In the example used in "Code Generation", the following .class files would be produced in the appropriate directory (given package information in the source code):

  • Foo.class

  • Foo_SJProfileKeys.class (ISO code generation only)

  • A .class file for each additional class you defined in Foo.sqlj

  • A .class file for each iterator and connection context class you declared in Foo.sqlj (whether public or private)

To ensure that .class files and profiles (if any, whether .ser or .class) will be located in the same directory, SQLJ passes its -d option to the Java compiler. If the -d option is not set, then .class files and profiles are placed in the same directory as the generated .java file, which is placed according to the -dir option setting.

In addition, so that SQLJ and the Java compiler will use the same encoding, SQLJ passes its -encoding option to the Java compiler unless the SQLJ -compiler-encoding-flag is turned off. If the -encoding option is not set, then SQLJ and the compiler will use the setting in the Java virtual machine (JVM) file.encoding property.

By default, SQLJ invokes the standard javac compiler of the Sun Microsystems Java Development Kit (JDK), but other compilers can be used instead. You can request that an alternative Java compiler be used by setting the SQLJ -compiler-executable option.


Note:

If you are using the SQLJ -encoding option but using a compiler that does not have an -encoding option, then turn off the SQLJ -compiler-encoding-flag. Otherwise, SQLJ will attempt to pass the -encoding option to the compiler.

For information about SQLJ support for compiler options and compiler-related SQLJ options, refer to the following:

9.1.5 Profile Customization (ISO Code Generation)

After Java compilation, if you are using ISO standard code generation, then the generated profiles containing information about your embedded SQL instructions are customized, so that your application can work efficiently with your database and use vendor-specific extensions.


Note:

If you use the default Oracle-specific code generation, then SQLJ produces no profiles and skips the customization step. Your code will support Oracle-specific features through direct calls to Oracle JDBC application programming interfaces (APIs).

To accomplish customization, SQLJ invokes a front end called the customizer harness, which is a Java class that functions as a command-line utility. The harness, in turn, invokes a particular customizer, either the default Oracle customizer or a customizer that you specify through SQLJ option settings.

During customization, profiles are updated in the following ways:

  • To allow your application to use any vendor-specific database types or features, if applicable

  • To tailor the profiles so that your application is as efficient as possible in using features of the relevant database environment

Without customization, you can access and use only standard JDBC types.

For example, the Oracle customizer can update a profile to support a SQL PERSON type that you had defined. You could then use PERSON as you would any other supported data type.

You must also customize with the Oracle customizer to use any of the oracle.sql type extensions.


Note:

Be aware of the following regarding profile customization:
  • The Oracle SQLJ run time and an Oracle JDBC driver will be required by your application whenever you use the Oracle customizer during translation, even if you do not use Oracle extensions in your code.

  • The generic SQLJ run time will be used if your application has no customizations, or none suitable for the connection.

  • You can customize previously created profiles by specifying .ser files, or .jar files containing .ser files, on the command line. But you cannot do this in the same running of SQLJ where translations are taking place. You can specify .ser/.jar files to be customized or .sqlj/.java files to be translated, compiled, and customized, but not both categories. For more information about how .jar files are used, refer to "JAR Files for Profiles".


For more information about profile customization, refer to Appendix A, "Customization and Specialized Customizers".

Also see the following for information about SQLJ options related to profile customization:

9.2 Functionality of Translator Errors, Messages, and Exit Codes

This section provides an overview of SQLJ translator messages and exit codes. It covers the following topics:

9.2.1 Translator Error, Warning, and Information Messages

There are three major levels of SQLJ messages that you may encounter during the translation phase: error, warning, and information. Warning messages can be further broken down into two levels: non-suppressible and suppressible. Therefore, there are four message categories (in order of seriousness):

  1. Errors

  2. Non-suppressible warnings

  3. Suppressible warnings

  4. Information

You can control suppressible warnings and information by using the SQLJ -warn option.

Error messages, prefixed by Error:, indicate that one of the following has been encountered:

  • A condition that would prevent compilation (for example, the source file contains a public class whose name does not match the base file name)

  • A condition that would result in a run-time error if the code were executed (for example, the code attempts to fetch a VARCHAR into a java.util.Vector, using an Oracle JDBC driver)

If errors are encountered during SQLJ translation, then no output is produced and compilation and customization are not executed.

Non-suppressible warning messages, prefixed by Warning:, indicate that one of the following has been encountered:

  • A condition that would probably, but not necessarily, result in a run-time error if the code were executed (for example, a SELECT statement whose output is not assigned to anything)

  • A condition that compromises the ability of SQLJ to verify run-time aspects of your source code (for example, not being able to connect to the database you specify for online checking)

  • A condition that presumably resulted from a coding error or oversight

SQLJ translation will complete if a non-suppressible warning is encountered, but you should analyze the problem and determine if it should be fixed before running the application. If online checking is specified but cannot be completed, then offline checking is performed instead.


Note:

For logistical reasons, the parser that the SQLJ translator uses to analyze SQL operations is not the same top-level SQL parser that will be used at run time. Therefore, errors might occasionally be detected during translation that will not actually cause problems when your application runs. Accordingly, such errors are reported as non-suppressible warnings, rather than fatal errors.

Suppressible warning messages, also prefixed by Warning:, indicate that there is a problem with a particular aspect of your application, such as portability. An example of this is using an Oracle-specific type, such as oracle.sql.NUMBER, to read from or write to Oracle Database 10g.

Informational or status messages prefixed by Info: do not indicate an error condition. They merely provide additional information about what occurred during the translation phase.

Suppressible warning and status messages can be suppressed by using the various -warn option flags:

  • cast/nocast: The nocast setting suppresses warnings about possible run-time errors when trying to cast an object type instance to an instance of a subtype.

  • precision/noprecision: The noprecision setting suppresses warnings regarding possible loss of data precision during conversion.

  • nulls/nonulls: The nonulls setting suppresses warnings about possible run-time errors due to nullable columns or types.

  • portable/noportable: The noportable setting suppresses warnings regarding SQLJ code that uses Oracle-specific features or might otherwise be non-standard and, therefore, not portable to other environments.

  • strict/nostrict: The nostrict setting suppresses warnings issued if there are fewer columns in a named iterator than in the selected data that is to populate the iterator.

  • verbose/noverbose: The noverbose setting suppresses status messages that are merely informational and do not indicate error or warning conditions.

If you receive warnings during your SQLJ translation, then you can try running the translator again with -warn=none to see if any of the warnings are of the more serious (non-suppressible) variety.

Table 9-2 summarizes the categories of error and status messages generated by the SQLJ translator.

Table 9-2 SQLJ Translator Error Message Categories

Message Category Prefix Indicates Suppressed By
Error Error: Fatal error that will cause compilation failure or run-time failure (translation aborted) NA
Non-suppressible warning Warning: Condition that prevents proper translation or might cause run-time failure (translation completed) NA
Suppressible warning Warning: Problem regarding a particular aspect of your application (translation completed) -warn option flags: nocast noprecision nonulls noportable nostrict
Informational/status message Info: Information regarding the translation process -warn option flag: noverbose

9.2.2 Translator Status Messages

In addition to the error, warning, and information messages, SQLJ can produce status messages throughout all phases of SQLJ operation: translation, compilation, and customization. Status messages are produced as each file is processed and at each phase of SQLJ operation.

You can control status messages by using the SQLJ -status option.

9.2.3 Translator Exit Codes

The following exit codes are returned by the SQLJ translator to the operating system upon completion:

  • 0: No error in execution

  • 1: Error in SQLJ execution

  • 2: Error in Java compilation

  • 3: Error in profile customization

  • 4: Error in class instrumentation, the optional mapping of line numbers from your .sqlj source file to the resulting .class file

  • 5: Error in ser2class conversion, the optional conversion of profile files from .ser files to .class files


Notes:

  • If you issue the -help or -version option, then the SQLJ exit code is 0.

  • If you run SQLJ without specifying any files to process, then SQLJ issues help output and returns exit code 1.


9.3 SQLJ Run Time

This section presents information about the Oracle SQLJ run time, which is a thin layer of pure Java code that runs above the JDBC driver.

If you use the default Oracle-specific code generation, then the SQLJ run time layer becomes even thinner, with a run time subset being used in conjunction with an Oracle JDBC driver. Most of the run time functionality is compiled directly into Oracle JDBC calls. You cannot use a non-Oracle JDBC driver.

When SQLJ translates SQLJ source code using ISO standard code generation, embedded SQL commands in your Java application are replaced by calls to the SQLJ run time. Run time classes act as wrappers for equivalent JDBC classes, providing special SQLJ functionality. When the end user runs the application, the SQLJ run time acts as an intermediary, reading information about your SQL operations from your profile and passing instructions along to the JDBC driver.

Generally speaking, however, a SQLJ run time can be implemented to use any JDBC driver or vendor-proprietary means of accessing the database. The Oracle SQLJ run time requires a JDBC driver, but can use any standard JDBC driver. To use Oracle-specific data types and features, however, you must use an Oracle JDBC driver. For the purposes of this document, it is generally assumed that you are using an Oracle database and one of the Oracle JDBC drivers.


Note:

For ISO standard SQLJ code generation, the Oracle SQLJ run time and an Oracle JDBC driver will be required by your application whenever you use the Oracle customizer during translation, even if you do not use Oracle extensions in your code. The generic SQLJ run time will be used if your application has no customizations, or none suitable for the connection.

9.3.1 SQLJ Run Time Packages

The Oracle SQLJ run time includes packages you will likely import and use directly, and others that are used only indirectly.


Note:

These packages are included in the run time libraries runtime12, runtime12ee, and runtime.

Packages Used Directly

The packages containing classes that you can import and use directly in your application are:

  • sqlj.runtime

    This package includes the ExecutionContext class, ConnectionContext interface, ConnectionContextFactory interface, ResultSetIterator interface, ScrollableResultSetIterator interface, and wrapper classes for streams (BinaryStream and CharacterStream, as well as the deprecated AsciiStream and UnicodeStream).

    Interfaces and abstract classes in this package are implemented by classes in the sqlj.runtime.ref or oracle.sqlj.runtime package or by classes generated by the SQLJ translator.

  • sqlj.runtime.ref

    The classes in this package implement interfaces and abstract classes in the sqlj.runtime package. You will likely use the sqlj.runtime.ref.DefaultContext class, which is used to specify your default connection and create default connection context instances. The other classes in this package are used internally by SQLJ in defining classes during code generation, such as iterator classes and connection context classes that you declare in your SQLJ code.

  • oracle.sqlj.runtime

    This package contains the Oracle class that you can use to instantiate the DefaultContext class and establish your default connection. It also contains Oracle-specific run time classes used by the Oracle implementation of SQLJ, including functionality to convert to and from Oracle type extensions.


Note:

Packages whose names begin with oracle are for Oracle-specific SQLJ features.

Packages Used Indirectly

The packages containing classes that are for internal use by SQLJ are:

  • sqlj.runtime.profile

    This package contains interfaces and abstract classes that define what SQLJ profiles look like (applicable only for ISO standard code generation). This includes the EntryInfo and TypeInfo classes. Each entry in a profile is described by an EntryInfo object, where a profile entry corresponds to a SQL operation in your application. Each parameter in a profile entry is described by a TypeInfo object.

    The interfaces and classes in this package are implemented by classes in the sqlj.runtime.profile.ref package.

  • sqlj.runtime.profile.ref

    This package contains classes that implement the interfaces and abstract classes of the sqlj.runtime.profile package and are used internally by the SQLJ translator in defining profiles (for ISO standard code generation only). It also provides the default JDBC-based run time implementation.

  • sqlj.runtime.error

    This package, used internally by SQLJ, contains resource files for all generic (not Oracle-specific) error messages that can be generated by the SQLJ translator.

  • oracle.sqlj.runtime.error

    This package, used internally by SQLJ, contains resource files for all Oracle-specific error messages that can be generated by the SQLJ translator.

9.3.2 Categories of Run-Time Errors

Run-time errors can be generated by any of the following:

  • SQLJ run time

  • JDBC driver

  • RDBMS

In any of these cases, a SQL exception is generated as an instance of the java.sql.SQLException class, or as a subclass, such as sqlj.runtime.SQLNullException.

Depending on where the error came from, there might be meaningful information you can retrieve from an exception using the getSQLState(), getErrorCode(), and getMessage() methods. SQLJ errors, for example, include meaningful SQL states and messages.

If errors are generated by the Oracle JDBC driver or RDBMS at run time, look at the prefix and consult the appropriate documentation:

9.4 Globalization Support in the Translator and Run Time

The Oracle SQLJ implementation uses the Java built-in capabilities for globalization support. This section discusses the following:


Note:

Some prior knowledge of Oracle Globalization Support is assumed, particularly regarding character encoding and locales. For information, refer to the Oracle Database Globalization Support Guide.

This section covers the following topics:

9.4.1 Character Encoding and Language Support

There are two main areas of SQLJ globalization support:

  • Character encoding

    There are three parts to this:

    • Character encoding for reading and generating source files during SQLJ translation

    • Character encoding for generating error and status messages during SQLJ translation

    • Character encoding for generating error and status messages when the application runs

  • Language support

    This determines which translations of error and status message lists are used when SQLJ outputs messages to the user, either during SQLJ translation or at SQLJ run time.

Globalization support at run time is transparent to the user, presuming your SQLJ source code and SQL character data use only characters that are within the database character set. SQL character data is transparently mapped into and out of Unicode.

Note that for multi-language applications, it is advisable to use one of the following options:


Notes:

  • The SQLJ translator fully supports Unicode 2.0 and Java Unicode escape sequences. However, the SQLJ command-line utility does not support Unicode escape sequences. You can use only native characters supported by the operating system. Command-line options requiring Unicode escape sequences can be entered in a SQLJ properties file instead, because properties files do support Unicode escape sequences.

  • Encoding and conversion of characters in your embedded SQL operations and characters read or written to the database, are handled by JDBC directly. SQLJ does not play a role in this. If online semantics-checking is enabled during translation, however, then you will be warned if there are characters within the text of your SQL data manipulation language (DML) operations that might not be convertible to the database character set.

  • For information about JDBC globalization support functionality, refer to the Oracle Database JDBC Developer's Guide and Reference.


Overview of Character Encoding

The character encoding setting for source files tells SQLJ two things:

  • How source code is represented in .sqlj and .java input files that the SQLJ translator must read

  • How SQLJ should represent source code in .java output files that it generates

By default, SQLJ uses the encoding indicated by the JVM file.encoding property. If your source files use other encodings, then you must indicate this to SQLJ so that appropriate conversion can be performed.

Use the SQLJ -encoding option to accomplish this. SQLJ also passes the -encoding setting to the compiler for it to use in reading .java files, unless the SQLJ -compiler-encoding-flag is off.


Note:

Do not alter the file.encoding system property to specify encodings for source files. This might impact other aspects of your Java operation and might offer only a limited number of encodings, depending on platform or operating system considerations.

The system character-encoding setting also determines how SQLJ error and status messages are represented when output to the user, either during translation or during run time when the end user is running the application. This is set according to the file.encoding property and is unaffected by the SQLJ -encoding option.

For source file encoding, you can use the -encoding option to specify any character encoding supported by your Java environment. If you are using the Sun Microsystems JDK, then these are listed in the native2ascii documentation, which you can find at the following Web site:

http://java.sun.com/j2se/1.3/docs/tooldocs/solaris/native2ascii.html

Dozens of encodings are supported by the Sun Microsystems JDK. These include 8859_1 through 8859_9 (ISO Latin-1 through ISO Latin-9), JIS (Japanese), SJIS (shift-JIS, Japanese), and UTF8.

Character Encoding Notes

Be aware of the following:

  • A character that is not representable in the encoding used, for either messages or source files, can always be represented as a Java Unicode escape sequence. This is of the form \uHHHH, where each H is a hexadecimal digit.

  • As a .sqlj source file is read and processed during translation, error messages quote source locations based on character position (not byte position) in the input encoding.

  • Encoding settings, either set through the SQLJ -encoding option or the Java file.encoding setting, do not apply to Java properties files, such as sqlj.properties and connect.properties. Properties files always use the encoding 8859_1. This is a feature of Java in general, not SQLJ in particular. However, you can use Unicode escape sequences in a properties file. You can use the native2ascii utility to determine escape sequences.

Overview of Language Support

SQLJ error and status reporting, either during translation or during run time, uses the Java locale setting in the JVM user.language property. Users typically do not have to alter this setting.

Language support is implemented through message resources that use key/value pairs. For example, where an English-language resource has a key/value pair of "OkKey", "Okay", a German-language resource has a key/value pair of "OkKey", "Gut". The locale setting determines the message resources used.

SQLJ supports locale settings of en (English), de (German), fr (French), and ja (Japanese).


Note:

Java locale settings can support country and variant extensions in addition to language extensions. For example, consider ErrorMessages_de_CH_var1, where CH is the Swiss country extension of German and var1 is an additional variant. SQLJ, however, currently supports only language extensions (de in this example), ignoring country and variant extensions.

9.4.2 SQLJ and Java Settings for Character Encoding and Language Support

The Oracle SQLJ implementation provides syntax that enables you to set the following:

  • The character encoding used by the SQLJ translator and Java compiler in representing source code

    Use the SQLJ -encoding option.

  • The character encoding used by the SQLJ translator and run time in representing error and status messages

    Use the SQLJ -J prefix to set the Java file.encoding property.

  • The locale used by the SQLJ translator and run time for error and status messages

    Use the SQLJ -J prefix to set the Java user.language property.

Setting Character Encoding for Source Code

Use the SQLJ -encoding option to determine the character encoding used in representing .sqlj files read by the translator, .java files generated by the translator, and .java files read by the compiler. The option setting is passed by SQLJ to the compiler, unless the SQLJ -compiler-encoding-flag is off.

This option can be set on the command line or SQLJ_OPTIONS environment variable, as in the following example:

-encoding=SJIS

Alternatively, you can set it in a SQLJ properties file, as follows:

sqlj.encoding=SJIS

If the encoding option is not set, then both the translator and compiler will use the encoding specified in the JVM file.encoding property. This can also be set through the SQLJ command line.


Note:

If your -encoding is to be set routinely to the same value, then it is most convenient to specify it in a properties file, as in the second example.

Setting Character Encoding and Locale for SQLJ Messages

Character encoding and locale for SQLJ error and status messages produced, during both translation and run time, are determined by the Java file.encoding and user.language properties. Although it is typically not necessary, you can set these and other JVM properties in the SQLJ command line by using the SQLJ -J prefix. Options marked by this prefix are passed to the JVM.

Set the character encoding as in the following example, which specifies shift-JIS Japanese character encoding:

-J-Dfile.encoding=SJIS


Note:

Only a limited number of encodings might be available, depending on platform or operating system considerations.

Set the locale as in the following example (which specifies Japanese locale):

-J-Duser.language=ja

The -J prefix can be used on the command line or SQLJ_OPTIONS environment variable only. It cannot be used in a properties file, because properties files are read after the JVM is invoked.


Notes:

  • If your file.encoding, user.language, or any other Java property is to be set routinely to the same value, it is most convenient to specify -J settings in the SQLJ_OPTIONS environment variable. This way, you do not have to repeatedly specify them on the command line. The syntax is essentially the same as on the command line. For more information, refer to "SQLJ_OPTIONS Environment Variable for Option Settings".

  • Remember that if you do not set the SQLJ -encoding option, then setting file.encoding will affect encoding for source files as well as error and status messages.

  • Be aware that altering the file.encoding property might have unforeseen consequences on other aspects of your Java operations. Also, any new setting must be compatible with your operating system.


SQLJ Command-Line Example: Setting Encoding and Locale

Following is a complete SQLJ command line, including JVM file.encoding and user.language settings:

% sqlj -encoding=8859_1 -J-Dfile.encoding=SJIS -J-Duser.language=ja Foo.sqlj

This example uses the SQLJ -encoding option to specify 8859_1 (Latin-1) for source code representation during SQLJ translation. This encoding is used by the translator in reading the .sqlj input file and in generating the .java output file. The encoding is then passed to the Java compiler to be used in reading the generated .java file. The -encoding option, when specified, is always passed to the Java compiler unless the SQLJ -compiler-encoding-flag is disabled.

For error and status messages output during translation of Foo.sqlj, the SQLJ translator uses the SJIS encoding and the ja locale.

9.4.3 SQLJ Extended Globalization Support

The Oracle SQLJ implementation includes support for Java types (Unicode character types) derived from existing character and stream types that convey expected usage for globalization. In SQLJ it is not possible to use JDBC statement or result set methods directly that otherwise serve the purpose of globalization support. If you are interested in information about those methods, refer to the Oracle Database JDBC Developer's Guide and Reference

If the database natively supports Unicode, then the types described in "Java Types for Globalization Support" are unnecessary. In this case, globalization support will be handled transparently. It is when the database does not natively support Unicode, but has a national language character set that does support Unicode, that you will typically use these types (for columns that use the national language character set).

Java Types for Globalization Support

The Oracle SQLJ implementation provides a number of Java types for globalization support. Table 9-3 notes the correspondence between these globalization support types and general-use JDBC and SQLJ character and stream types. Each globalization support type, except for NString, is a subclass of its corresponding JDBC or SQLJ type.

Table 9-3 JDBC and SQLJ Types and Corresponding Globalization Types

JDBC and SQLJ Types Globalization Support Types
JDBC types:  
oracle.sql.CHAR oracle.sql.NCHAR
java.lang.String oracle.sql.NString
oracle.sql.CLOB oracle.sql.NCLOB
SQLJ types:  
sqlj.runtime.CharacterStream oracle.sqlj.runtime.NcharCharacterStream
sqlj.runtime.AsciiStream

(Deprecated; use CharacterStream.)

oracle.sqlj.runtime.NcharAsciiStream

(Deprecated; use NcharCharacterStream.)

sqlj.runtime.UnicodeStream

(Deprecated; use CharacterStream.)

oracle.sqlj.runtime.NcharUnicodeStream

(Deprecated; use NcharCharacterStream.)


In situations where your application must handle national language character strings, either inserting them into or selecting them from national language character set columns, use the globalization support types instead of the corresponding general-use types.


Notes:

  • All globalization support types add automatic registration of intended usage for IN and OUT parameters, but are otherwise identical in usage to the corresponding JDBC or SQLJ type (including constructors).

  • Use of globalization support types is unnecessary in iterator columns, because the underlying network protocol supports national language characters implicitly for the underlying result sets.


NString Class Usage and Notes

The oracle.sql.CHAR class, and therefore its NCHAR subclass, provides only constructors that require explicit knowledge of the database character set. Therefore, the oracle.sql.NString class, a wrapper for java.lang.String, is preferable in most circumstances. The NString class provides simpler constructors and ensures that the national language character form of use is registered with the JDBC driver.

Following are the key NString methods:

  • NString(String): This constructor creates an NString instance from an existing String instance.

  • String toString(): This method returns the underlying String instance.

  • String getString(): This method also returns the underlying String instance.

The toString() method enables you to use the NString instance in string concatenation expressions (such as "a"+b, where b is a string). The getString() method, provided in the CHAR superclass, is supported as well for uniformity. In addition, the member methods of the String class are carried over to the NString wrapper class to enable you to write more concise code.

In SQLJ applications, for versions prior to Oracle Database 10g, you must use host variables of the NString type to bind columns of the NCHAR type. For example, consider the following table:

CREATE TABLE Tbl1 (
       ColA  CHAR
       NColB NCHAR
)

To insert a row in this table through a SQLJ application, use a code similar to the following:

...
String v_a = "\uFF5E";
NString v_nb = "\uFF5E";
#sql {INSERT INTO Tbl1 (ColA, NColB) VALUES (:v_a, :v_nb)};
...

In Oracle Database 10g release 2 (10.2), SQLJ applications can use String variables to bind NCHAR columns. Therefore, the preceding example can be rewritten as follows:

...
String v_a = "\uFF5E";
String v_nb = "\uFF5E";
#sql {INSERT INTO Tbl1 (ColA, NColB) VALUES (:v_a, :v_nb)};
...

However, if you want to use String host variable to bind to NCHAR columns, then you must translate the SQLJ file with the -ncharconv SQLJ translator option, as follows:

sqlj -ncharconv [-options] app.sqlj

In the preceding command, options can be other SQLJ options and app.sqlj is the SQLJ file that contains the code.

When this option is used, the setFormOfUse method will be generated for all binds to character columns, that is, CHAR or NCHAR columns.


Note:

When the SQLJ file is compiled with the -ncharconv option, the setFormOfUse method is used in the generated code for codegen=oracle. For codegen=iso, this option information is passed to the Oracle SQLJ run time, which internally uses SetFormOfUse for bind at run time.

The SQLJ application can use String host variable to retrieve data from the server without using the -ncharconv option, because the information about the column type is fetched at the client side and JDBC internally sets the form for the appropriate columns.


Note:

There may be a small difference in the performance when using the -ncharconv option, depending on the database character set and national charster set and the number of character columns in the table.

Globalization Support Examples

The following examples show use of the NString class:

  • NString as IN argument

    This example uses an NString instance as an input parameter to the database.

    import oracle.sql.NString;
    ...
    NString nc_name = new NString("Name with strange characters");
    #sql { update PEOPLE
           set city = :(new NString("\ufff2")), name = :nc_name
           where num= :n };
    ...
    
    
  • NString as OUT argument

    This example uses an NString instance as an output parameter from the database.

    import oracle.sql.NString;
    ...
    NString nstr;
    #sql { call foo(:out nstr) };
    System.out.println("Result is: "+nstr);
    // or, explicitly: System.out.println("Result is: "+nstr.toString());
    ...
    
    
  • NString as Result Set column

    This example uses the NString type for an iterator column. Such usage is superfluous, given that the underlying network protocol supports national language characters implicitly, but harmless. This example also shows use of one of the String methods, substring(), that is carried over to NString.

    import oracle.sql.NString;
    import oracle.sql.NCLOB;
    ...
    #sql iterator NIter(NString title, NCLOB article);
    
    NIter nit;
    #sql nit = { SELECT article, title FROM page_table };
    while (nit.next())
    {
      System.out.println("<TITLE>"+nit.title()+"</TITLE>");
      ...
      nit.article().substring(0, 1000); ...
    }
    
    

Note:

Using the NCHAR type instead of the NString type for the preceding examples requires the following changes:
  • Use the appropriate NCHAR constructor. NCHAR constructors mirror CHAR constructors, such as the following:

    NCHAR(String str, oracle.sql.CharacterSet charset)
    
    
  • Although you have the option of using either toString() or getString() to retrieve the underlying String instance from an NString instance, for an NCHAR instance you must use the getString() method. When using the NString type, the toString() method is used automatically for string concatenation, such as in "NString as OUT argument".


9.4.4 Manipulation Outside of SQLJ for Globalization Support

This section discusses ways to manipulate your Oracle Globalization Support configuration outside of SQLJ.

Setting Encoding and Locale at Application Run Time

As with any end user running any Java application, those running your SQLJ application can specify JVM properties, such as file.encoding and user.language directly, as they invoke the JVM to run your application. This determines the encoding and locale used for message output as your application executes.

They can accomplish this as in the following example:

% java -Dfile.encoding=SJIS -Duser.language=ja Foo

This will use SJIS encoding and Japanese locale.

Using API to Determine Java Properties

In Java code, you can determine values of Java properties by using the java.lang.System.getProperty() method, specifying the appropriate property. For example:

public class Settings
{
   public static void main (String[] args)
   {
      System.out.println("Encoding: " + System.getProperty("file.encoding")
                       + ", Language: " + System.getProperty("user.language"));
   }
}

You can compile this and run it as a standalone utility.

There is also a getProperties() method that returns the values of all properties, but this will raise a security exception if you try to use it in code that runs in the server.

Using native2ascii for Source File Encoding

If you are using a Sun Microsystems JDK, then there is an alternative to having SQLJ do the character encoding for your source files. You can use the native2ascii utility to convert sources with native encoding to sources in 7-bit ASCII with Unicode escape sequences.


Note:

To use SQLJ to translate source created by native2ascii, ensure that the JVM that invokes SQLJ has a file.encoding setting that supports some superset of 7-bit ASCII. This is not the case with settings for EBCDIC or Unicode encoding.

Run native2ascii as follows:

% native2ascii <options> <inputfile> <outputfile>

Standard input or standard output are used if you omit the input file or output file. Two options are supported:

  • -reverse (Reverse the conversion. Convert from Latin-1 or Unicode to native encoding)

  • -encoding <encoding>

For example:

% native2ascii -encoding SJIS Foo.sqlj Temp.sqlj

For more information, see the following Web site:

http://java.sun.com/j2se/1.3/docs/tooldocs/solaris/native2ascii.html