Enter the fields of this wizard to create a Data Element.
- Project
- Select a project in the list or enter the name of the project
in which the instances are to be saved.
A project represents the
physical distribution of your instances in your workspace. It groups
directories and files. It is used to organize resources which can
be built, shared, or stored as a version.
A Pacbase project
contains a subset of files which correspond to the models and the
generated sources which make up an application.
There can be
various projects for each location (a location is the equivalent of
a database).
- Package
- Select a package in the list or enter the name of a new package.
A
package corresponds to a logical group of information.
The logical
resources contained in a package can be distributed into various projects.
The
packages represent the branches of the projects. Each package contains
its own instances.
For the resources transferred from Pacbase,
the naming convention of a package is: {location name}.{Library
path}.{Library name}. To reflect the Library hierarchy,
you can enter the Library names (starting from the highest-level Library),
and separate each with a period (example: COM.CIF).
This way, there will be a first-level branch (COM)
with its instances, and a second-level branch (CIF)
with its own instances.
- Name
- This name references the Data Element independently of any Data
Structure, Report, or Screen to which the Data Element might belong.
Including a Report, Screen or Segment name in the Data Element name
is not necessary since it is automatically performed.
This name
consists of alphabetic or numeric characters only. The following table
lists the Data Element names that cannot be created:
Table 1. List of reserved Data Element namesName |
Reserved for |
FILLER |
Field alignment |
ENPR
GRPR
ERUT
|
Error check fields on transaction files (options
of the Batch Applications function) Data Element error check
Segment
error check
User-defined errors
|
- Inherits from
- A Data Element is an elementary concept. In your applications,
you might want to use several technical variants of this concept.
So to guarantee the coherence of your data, you manage these variants
by creating parent and child Data Elements.
All Data Elements
from the same family are logically linked in the database. A child
Data Element cannot be a parent as well. The inheritance tree of the
current instance displays its parent, if any, and its children, if
any.
To specify a parent Data Element, click Change and
select a Data Element in the selection wizard.
- Internal format
- This format, displayed in the Overview tab,
is required, except if the Data Element is a Property.
The internal
format is used when the Data Element is called in:
- An application file (permanent file, database, temporary file),
- A Screen as a data-entry field, if no conversational format has
been specified (see the type of line field
in the -D Lines tab of the Data Element editor).
This format is automatically transferred into the Segments where
the Data Element is called. But you eventually select the appropriate
format in the -CD Lines tab of the Programs which
call these Segments.You must ensure the compatibility between the
input and internal formats.
The internal format must be coded
like a COBOL picture (without print characters).
Notes: - If the format of a numeric Data Element is more than 10 characters
long, you must omit the 9 that would normally
be entered after the V. For example, S9(10)V9(3) must
be entered as S9(10)V(3). This way of coding must
not be used when the format is shorter than 10 characters
- For Unicode-type Data Elements, you must indicate 'N(n)'
or a signed or unsigned numeric format.
The
INTERNAL USAGE clause is associated
with this format.
Note: If a Unicode-type Data Element has a signed
numeric internal format, its internal usage must be X or 1.
Data
Elements that represent a date can be assigned a symbolic format:
Table 2. Symbolic formats for datesSymbolic value |
Meaning |
D |
Without century (DDMMYY or MMDDYY) |
I |
Without century (YYMMDD) |
E |
Without century (DD/MM/YY or MM/DD/YY) |
S |
With century (CCYYMMDD) |
C |
With century (DDMMCCYY or MMDDCCYY) |
M |
With century (DD/MM/CCYY or MM/DD/CCYY) |
G |
Gregorian format (CCYY-MM-DD) |
T |
Time format (HH:MM:SS) |
TS |
Timestamp format |
Note: For the formats which include a separator (E, G, M,
and T), you can specify a separator after the
character which represents the format, if you do not want to use the
separator included by default in the format. For example, a G/ format
generates CCYY/MM/DD instead of CCYY-MM-DD,
which is the default Gregorian format.
For details on
the use of the formats with the various types of Database Blocks,
see the summary tables in chapter Columns: Data Elements of
the 'Relational SQL Database Description' manual at this URL:
http://www-01.ibm.com/support/docview.wss?rs=37&uid=swg27005477
- Generation parameters
- Generation parameters are specified in Libraries. These parameters
are related to the adaptation to the operating system in use.
You
select the Library by clicking Browse.
- Label
- Enter the label, which identifies the new instance. It constitutes
a search criterion for a subsequent search.
- Open editor on finish
- Check this box if you want the editor of the instance you are
creating to open up automatically. The editor opens up on the Overview tab.
From there, you can access the other tabs to specify all the characteristics
of the instance.