This is a flowchart showing inputs and outputs for the JPublisher publishing process, as described in the accompanying text. In addition to those descriptions, the graphic shows the following:
Input to JPublisher through the command line, properties file, or INPUT
file includes SQL object types, SQL collection types, PL/SQL packages, server-side Java classes, SELECT
/UPDATE
/INSERT
/DELETE
statements, or WSDL files.
The SQLJ translator, invoked by JPublisher, is depicted as an entity within the JPublisher product. The Java compiler, by contrast, is shown as an external entity. SQLJ source files are shown as being input to the SQLJ translator, with Java source files being output; Java source files (either from the SQLJ translator, or generated directly by JPublisher in some circumstances) is shown as being input to the Java compiler, with Java class files (.class
) being output.
JPublisher .class
output is shown as optionally going to the database JVM or the client-side or middle-tier JVM. (Classes for Web services call-outs are loaded to the database JVM by default.)
JPublisher.sql
output, consisting of PL/SQL wrappers and scripts, is shown as optionally going to the database PL/SQL engine. (This happens by default.)
The database is shown to include the PL/SQL engine and database JVM, which includes the SQLJ runtime and JDBC driver. The database is also shown to include the metadata that JPublisher checks for defintions of SQL types, PL/SQL packages, and shapes of query results.
User-written classes, including subclasses of JPublisher-generated classes, are shown as going to the client-side or middle-tier JVM.
The client-side or middle-tier JVM is shown to include the SQLJ runtime and JDBC driver.