This document uses a class generated from the Employees DDM as an example, providing detailed information about the class generation process. This document covers the following topics:
Jadabas provides two utilities for generating Java class objects:
The GENERATE utility is used to create Java
class objects from Natural DDMs on the mainframe.
The GENDDM utility is used with the Natural
SYSTRANS utility to create Java class objects on all
other platforms. You can also create input to GENDDM
manually; see the layout for a flat file description of a DDM in the section
Creating Input to
GENDDM.
The class that you create using these utilities plus the class Adabas comprise the Jadabas API to be used in your Jadabas applications.
Note:
GENDDM also supports Natural (version 6) object management.
The generation process creates an object that has properties and methods. With a few exceptions, the properties correspond to fields derived from the DDM from which the class was generated. To improve performance, all properties derived from field names are public. This means that any application can examine and modify the property directly (the field is declared final if it is read only and therefore cannot be modified). If desired, the Java source for the class can be manually modified to make these properties private and to add the corresponding accessor and modifier methods. Group fields in the DDM are never present in the Java class. The group field itself does not have data, and the fields contained within are always part of the Java class if selected.
The property names are of mixed case (upper and lower). When generating
a class, the GENERATE utility applies the following
rules:
Hyphens are removed.
The letter after the hyphen is upper case.
All other letters are lower case.
For example, the field "MIDDLE-NAME" in the DDM becomes property "middleName" in the Java class.
If this scheme generates duplicate names, you will need to edit the source file manually and correct the duplicates.
Every generated class has a public ISN property, that is, an "int" representing the ISN of the object.
If the Adabas field option of "NC" is used, Jadabas will generate a special field for it. For example, if a field named "BIRTH_DATE" uses the "NC" option, Jadabas generates a field called "BIRTH_DATE_NC".
Descriptors are class constants that are derived from the descriptor fields in the DDM. They are used in the find methods to indicate search keys and in the readFrom method to indicate sequential read order.
Descriptor property names are the original field names. When generating
the descriptor property, the GENERATE utility
applies the following rules:
Hyphens replaced by underscores.
All letters are upper case.
For example, the descriptor "PHONETIC-NAME" in the DDM becomes "PHONETIC_NAME" in the Java class.
Natural supports several field types, some of which can be adequately represented by the Java native types and some of which need the supporting classes provided in the Jadabas installation kit. When a record is read from the database, the properties of the class are set by converting the value from a byte stream in the record buffer to Java native types and objects. The following table shows the Natural Data Type (format) and the resulting Java Type or Object:
| Natural Data Type | Java Type or Object |
|---|---|
| A - alphanumeric | String |
| N - numeric | Decimal (class delivered with Jadabas) |
| P - packed numeric | Decimal (class delivered with Jadabas) |
| F - floating point | Double |
| B - binary | Binary (class delivered with Jadabas) |
| L - logical | boolean |
| I - integer | int |
| D - date | date |
| T - time | date |
MU and PE periodic group fields are stored as arrays, as follows:
MU field arrays have a variable number of occurrences; that is, the number of occurrences depends on the number of actual occurrences in the record. If the Java application needs to increase the number of elements, it should create a new array of the desired type and copy the old array into the new array.
PE field arrays have a fixed length that is specified when the class is generated. However, the Java application can extend or shrink the number of occurrences before updating or storing by creating a new array and copying the contents of the old array into it.
MU within PE field arrays are two-dimensional: the number of elements in the second dimension varies (the MU portion).
To use the GENERATE utility, follow the steps below:
Store in an Adabas file accessible to Jadabas a Natural DDM that corresponds to the view of the file you are using.
The Web Utility Server enables the
GENERATE utility. Start it as described in the
section Web Utility Server.
Start the Jadabas Proxy if you are connecting to any Adabas other than Adabas V7.
In Windows, start your browser and navigate to GENERATE as follows:
http://WebServerMachine:port/html/GENERATE.HTM
where WebServerMachine:port is the name and port number of the host where the WebServer is executing (the default port is "9980").
Example: "http://localhost:9980/html/GENERATE.HTM"
In UNIX, open a command prompt and enter
generate.sh.
The page that appears is similar to the following example:
Enter the DBID and FNR (database ID and file number) of the FDIC file (Predict dictionary file) in the text fields labeled FDIC dbid and FNR, respectively, and click the button.
The screen that appears contains a button for each DDM in the given FDIC:
Click the button with the desired DDM name.
The page that appears contains the DDM details:
In addition to the button, used to generate the Java class, this screen contains three other function buttons:
: Removes all check marks from the Select column.
: Selects the Readonly check box for all fields displayed.
: Resets to the default setting, all fields selected.
Clear the Select box for each field that is not to be included in the Java class object.
Radio buttons are used instead of check boxes if the field contains duplicate short Adabas names. The radio buttons ensure that only one of the fields can be selected. The first field of such a "duplicate group" has two radio buttons. Selecting the right-hand button excludes the field (the equivalent of unchecking a regular field).
Open the Encoding drop-down list box to select a character encoding for alpha fields. Note that an encoding class is required for execution at the view or field level.
Click the Readonly box to make a field "read only".
For fields within a PE group, set the PE occurrences field to a number between "1" and "199" if the field is to be included. The specified number of occurrences will be read from the file for each record.
When all selections have been made, click the button.
Save the result in a .java file, and compile
using the Java compiler (javac).
To use the GENDDM utility, follow the steps below:
Store in an Adabas file accessible to Jadabas a Natural
SYSTRANS or object manager that corresponds to the
view of the file you are using.
The Web Utility Server enables the GENDDM
utility. Start it as described in the section Web Utility Server.
Start the Jadabas Proxy.
In Windows, start your browser and navigate to
GENDDM as follows:
http://WebServerMachine:port/html/GENDDM.HTM
where WebServerMachine:port is the name and port number of the host where the WebServer is executing (the default port is "9980").
Example: "http://localhost:9980/html/GENDDM.HTM"
In UNIX, open a command prompt and enter
genddm.sh.
The screen that appears is similar to the following:
Enter the name of the SYSTRANS file, and click the button.
The screen that appears contains a button for each DDM in the given SYSTRANS file:
Click the button with the desired DDM name.
A sreen containing the DDM details appears:
In addition to the button, used to generate the Java class, this screen contains three other function buttons:
: Removes all check marks from the Select column.
: Selects the Readonly check box for all fields displayed.
: Resets to the default setting, all fields selected.
Clear the Select box for each field that is not to be included in the Java class object.
Radio buttons are used instead of checkboxes if the field contains duplicate short Adabas names. The radio buttons ensure that only one of the fields can be selected. The first field of such a "duplicate group" has two radio buttons. Selecting the right-hand button excludes the field (the equivalent of unchecking a regular field).
Open the Encoding drop-down list box to select a character encoding for alpha fields. Note that an encoding class is required for execution at the view or field level.
Click the Readonly box to make a field "read only".
For fields within a PE group, set the PE occurrences field to a number between "1" and "199" if the field is to be included. The specified number of occurrences will be read from the file for each record.
When all selections have been made, click the Generate button.
Save the result in a .java file, and compile using the Java compiler (javac).
The following layout for a flat file description of a DDM can be used to create input to GENDDM. The layout includes the following parts as well as an example:
| Columns | Contents | Number of Characters |
|---|---|---|
| 01 - 04 | Constant "****" | 4 |
| Columns | Contents | Number of Characters |
|---|---|---|
| 01 - 04 | Constant "*C**" | 4 |
| 45 - 76 | DDM Name | 32 |
| 77 | Constant "X" | 1 |
| Columns | Contents | Number of Characters |
|---|---|---|
| 01 - 04 | Constant "*D04" | 4 |
| 05 - 09 | Database ID | 5 |
| 10 - 14 | File Number | 5 |
Note that Jadabas supports the Adabas NC (null not counted) option used by Adabas SQL Server and other structured query languages. For example, the field 'BIRTH' in the EMPLOYEES file would be as follows:
*S**_1AHBIRTH D0060 D *S**_1AHN@BIRTH I0020 D
| Columns | Contents | Number of Characters |
|---|---|---|
| 01 - 04 | Constant "*S**" | 4 |
| 05 |
Group (G, M, P) G = group field (set Length/Decimal Places to 0000) M = multiple-value field P = periodic group |
1 |
| 06 | Level (1 - 7) | 1 |
| 07 - 08 | Short Name | 2 |
| 09 - 40 | Long Name | 32 |
| 41 |
Format (A, B, D, F, I, L, N, P, T) A - type = alpha B - type = binary D - type = date F - type = float I - type = integer L - type = logical N - type = decimal P - type = decimal T - type = time |
1 |
| 42 - 44 | Length | 3 |
| 45 | Decimal Places | 1 |
| 46 |
Option (' ', F, M, N) F = fixed storage (F) M = null not allowed (Adabas NN option indicating an SQL-significant null) N = null value suppression (Adabas NU option) \ |
1 |
| 47 |
Descriptor Flag (' ", D, H, P, S, U) D = normal descriptor H = hyper descriptor P = phonetic descriptor S = super descriptor U = sub descriptor |
1 |
| Columns | Contents | Number of Characters |
|---|---|---|
| 01 - 02 | Constant "*E" | 2 |
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 01 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 *C** EMPLOYEES X *D040019800011 *S** 1AAPERSONNEL-ID A0080 D *S**G1ABFULL-NAME 0000 *S** 2ACFIRST-NAME A0200N *S** 2ADMIDDLE-I A0010N *S** 2ADMIDDLE-NAME A0200N *S** 2AENAME A0200 D *S** 1AFMAR-STAT A0010F *S** 1AGSEX A0010F *S** 1AHBIRTH N0060 D *S**G1A1FULL-ADDRESS 0000 *S**M2AIADDRESS-LINE A0200N *S** 2AJCITY A0200ND *S** 2ADKIP A0100N *S** 2ADPOST-CODE A0100N *S** 2ALCOUNTRY A0030N *S**G1A2TELEPHONE 0000 *S** 2ANAREA-CODE A0060N *S** 2AMPHONE A0150N *S** 1AODEPT A0060 *S** 1APJOB-TITLE A0250ND *S**P1AQINCOME 0000 *S** 2ARCURR-CODE A0030N *S** 2ASSALARY P0090N *S**M2ATBONUS P0090N *S**G1A3LEAVE-DATA 0000 *S** 2AULEAVE-DUE N0020 *S** 2AVLEAVE-TAKEN N0020N *S**P1AWLEAVE-BOOKED 0000 *S** 2AXLEAVE-START N0060N *S** 2AYLEAVE-END N0060N *S**M1AZLANG A0030ND *S** 1PHPHON-NAME A0200 P *S** 1H1LEAVE-LEFT B0040 S *S** 1S1DEPARTMENT A0040 S *S** 1S2DEPT-PERSON A0260 S *S** 1S3CURRENCY-SALARY A0120 S *E |
