| Oracle Procedural Gateway for APPC User's Guide Release 9.0.1.0.1 for UNIX Part Number A90397-01 |
|
This appendix contains the following sections:
The Procedural Gateway Data Dictionary (PG DD) is maintained in a conventional Oracle database. It is installed by a SQL*Plus installation script (pgddcr8.sql in the $ORACLE_HOME/pg4appc/admin directory on the gateway machine) and manipulated by PGAU statements and standard SQL statements.
The dictionary is divided into two sections:
The environment dictionary is static and should not be changed. The contents of the environment dictionary support proper translation from the remote transaction's environment to the integrating Oracle integrating server's environment, and is platform-specific. The active dictionary is updated at the user's location by the PGAU in response to definitions supplied by the user.
The PGAU uses some dictionary tables strictly as input. These dictionary tables define environmental parameters for PGAU. Both table and values are installed by a SQL*Plus script at gateway installation time and are not to be modified by the installation.
The environment dictionary does not reference the active dictionary, but the active dictionary does reference environment dictionary entries.
The environment dictionary requires unique identifying numbers in some columns to join environment dictionary entries together. Oracle sequence objects are therefore created by the Oracle Procedural Gateway for APPC to support this requirement. The following table presents the Oracle sequence objects and their descriptions.
The environment dictionary tables contain constants that describe these components of the operating environment:
The pga_maint table stores the PG DD version number and change history, as presented in the following table.
The pga_usage table performs a referential integrity check of pga_data and pga_field column "usage" as presented in the following table.
| Column | Type | Content |
|---|---|---|
|
name |
varchar2(6) |
Value for the "usage" field of data dictionary tables. For example: Max length => 4-char string length |
The pga_modes table performs a referential integrity check of pga_call_parm column "mode", as presented in the following table.
.
| Column | Type | Content |
|---|---|---|
|
name |
varchar2(6) |
Name of valid parameter call modes. For example: Max length => 'IN OUT' string length |
The PG DD active data dictionary is created by pgddcr8.sql at installation, but maintained using PGAU. The active dictionary can refer to items (by ID number) in the environment dictionary.
The PG DD active dictionary tables contain the descriptions of transactions and data structures. There might be more than one version of a definition. Old versions are retained indefinitely.
In all PGAU dictionary operations, a definition is referred to by its "name". That name can be qualified by specific version number. If omitted, the most recent version is assumed.
Because the active dictionary is constantly changing, the identifying numbers needed to join active dictionary entries together must also change. To support this requirement, PG DD installation creates the following Oracle sequence objects. The table lists the Oracle sequence objects and their descriptions:
Following is a list of active dictionary tables:
One row exists in the pga_trans table for each user transaction. The row is created by a PGAU DEFINE TRANSACTION statement and used by a PGAU GENERATE statement to create the PL/SQL package (TIP).
The following table presents the column, type and content information for pga_trans:
The pga_trans_attr table relates a character string defining the transaction attributes supported by PGA to pga_trans_values entries through an attribute id number and type.
The pga_trans_attr table is also used for integrity checks of transaction attributes when new transactions are being defined.
There is an entry in the pga_trans_attr table for each transaction attribute name. All possible transaction attribute names supported by PGA on any defined transaction are specified. There is one row for each attribute, and no duplicates are allowed.
The following table presents the column, type and content information for pga_trans_attr:
The pga_trans_values table describes the values of transaction attributes.
A row exists to specify the value of each attribute of each transaction defined in the data dictionary. The column, type and content information for pga_trans_values is presented in the following table:
.
The pga_trans_calls table relates all calls available with any single transaction to each specific call definition through a call ID number.
An entry exists in the pga_trans_calls table for each PL/SQL call referenced in a transaction definition through the CALL(cname,...) operand. One row per transaction call; no duplicates.
The column, type and content information for pga_trans_calls is presented in the following table:
.
The pga_call table relates all calls that are available for all defined transactions, to a unique call id number and PL/SQL remote procedural call (RPC) name. One entry exists in this table for each PL/SQL call (defined in a DEFINE CALL statement).
One row per call, duplicates are possible when multiple transactions make identical calls. The plsrpc specification must be unique within the Oracle server which makes the calls, and rows are uniquely distinguished by call#.
The column, type and content information for pga_call are presented in the following table:
The pga_call_parm table relates all parameters of any single transaction call to the data definitions describing each parameter.
One entry exists in the pga_call_parm table for each parameter on a call in the PARMS() operand of the PGAU DEFINE CALL statement. One row per parameter, duplicates allowed when multiple calls (in the pga_call table) refer to the same parameters.
The following table presents the column, type and content information for pga_call_parm:
.
The pga_data table defines each data item used as a parameter in a call and relates the remote host data name to its PL/SQL variables and any component subfields or clauses within each data item (if the data item is an aggregate, such as a record). Each data item might have attributes related to it through its corresponding field definition. Even atomic data items have a single row in the pga_field table.
One row exists in the pga_data table for each data item defined by a PGAU DEFINE DATA or REDEFINE DATA statement.
The following table presents the column, type and content information for pga_data:
.
The pga_fields table defines each field within a data item and relates the remote host data field to its PL/SQL variables or nested records. Each field item might have attributes related to it (by field#) in the pga_data_attr and pga_data_values tables.
One row exists in the pga_fields table for each atomic item, field, clause, or nested record defined by a PGAU DEFINE DATA statement. Several rows would exist (related by a single data# and incrementing fld#) to define an aggregate data item, one row per field or group.
The following table presents the column, type and content information for pga_fields:
The pga_data_attr table defines all possible data attribute names allowed by PGA and relates each attribute name to a number and type, by which the value of this attribute for a specific data item can be selected from pga_data_values.
The pga_data_attr table is also used for integrity checks of data attributes when new data items are defined.
There is one entry in the pga_data_attr table for every possible attribute name to which any PGA supported data item might relate.
The following table presents the column, type and content information for pga_data_attr:
A row exists in the pga_data_values table for each attribute of each data item defined by each data definition.
The following table presents the column, type and content information for pga_data_values:
|
|
![]() Copyright © 2001 Oracle Corporation. All Rights Reserved. |
|