The signature of an Integration Server service specifies the
data types for the parameters of the service. There are three data types:
String, Record and Object. Parameters of
type String are mapped to an IDL alphanumeric data type,
parameters of type Record to IDL groups. Parameters of type
Object cannot be mapped to an IDL data type and thus result in an
error. If such an error occurs, the IDL program which contains the invalid data
type mapping is written to the IDL file, however, all lines are comments.
Starting with version 9.7, a parameter of an EntireX Adapter
service of type Object is mapped to an IDL binary data type.
Depending on the target language, it is either mapped to BV (for
Natural and Other) or to BV256 (for
COBOL and PL/I).
All parameters of type String are mapped to an IDL
alphanumeric data type, available as variable (AV,
AVn) or fixed (An) length. Which alphanumeric type is
used is specified in the wizard page of the IDL Extractor for Integration Server or as a
parameter of the generation service.
Parameters of type String may have an associated context
type which is specified in the constraints of the parameter's properties. This
content type influences the data type mapping. The content type was already
considered in previous versions of the EntireX Adapter (version 9.6 or
lower). However, its usage was not documented and the mapping was not yet
complete. The handling of the content type was reworked with version 9.7.
The following table shows both the new and the old mapping. The new mapping is
different from the old one. If you want to use the old mapping with version 9.7
or higher, set the extended setting
watt.com.softwareag.entirex.wmadapter.extractor.contenttype.handling.pre97=true
(in the Integration Server administration page under
Settings > Extended).
| Integration Server Content Type | Software AG IDL Data Type | |||
|---|---|---|---|---|
| Adapter Version 9.7 or higher | Adapter Version 9.6 or lower | |||
boolean |
L |
Logical | L |
Logical |
decimal |
Nx, Nx.y or An, AV, AVn |
Unpacked decimal or Alphanumeric (1) | N10.2 |
Unpacked decimal |
float |
F4 |
Floating point (small) | F4 |
Floating point (small) |
double |
F8 |
Floating point (large) | F8 |
Floating point (large) |
Date |
An, AV, AVn |
Alphanumeric (2) | D |
Date |
dateTime |
An, AV, AVn |
Alphanumeric (2) | T |
Time |
base64Binary |
An, AV, AVn |
Alphanumeric (2) | BV |
Binary variable length |
byte |
I1 |
Integer (small) | I1 |
Integer (small) |
unsignedByte |
NU3 |
Unpacked decimal unsigned | I1 |
Integer (small) |
short |
I2 |
Integer (medium) | I2 |
Integer (medium) |
unsignedShort |
NU5 |
Unpacked decimal unsigned | I2 |
Integer (medium) |
int |
I4 |
Integer (large) | I4 |
Integer (large) |
unsignedInt |
NU10 |
Unpacked decimal unsigned | I4 |
Integer (large) |
long |
N19 |
Unpacked decimal | An, AV, AVn |
Alphanumeric (2) |
unsignedLong |
NU20 |
Unpacked decimal unsigned | An, AV, AVn |
Alphanumeric (2) |
integer |
N29 |
Unpacked decimal | An, AV, AVn |
Alphanumeric (2) |
positiveInteger |
NU29 |
Unpacked decimal unsigned | An, AV, AVn |
Alphanumeric (2) |
nonPositiveInteger |
N29 |
Unpacked decimal | An, AV, AVn |
Alphanumeric (2) |
negativeInteger |
N29 |
Unpacked decimal | An, AV, AVn |
Alphanumeric (2) |
nonNegativeInteger |
NU29 |
Unpacked decimal unsigned | An, AV, AVn |
Alphanumeric (2) |
string |
An, AV, AVn |
Alphanumeric (3) | An, AV, AVn |
Alphanumeric (2) |
all others |
An, AV, AVn |
Alphanumeric (2) | An, AV, AVn |
Alphanumeric (2) |
Notes:
Nx.y is used. If only a total
number of digits is specified, then Nx is used. If no total number
of digits is specified, or the total number exceeds the maximum allowed value,
or the number of fraction digits exceeds the maximum allowed value, then a
string according to (2) is used. The maximum allowed value for the total number
of digits is 29, for COBOL it is 31. The maximum allowed value for the number
of fraction digits is 7, for COBOL it is 31 and for Natural it is 29.
AV, AVn or An is used.
len, then
Alen is used. If the content type specifies a maximum length
max, then AVmax or Amax is used, see (2)
for the choice between A versus Av. If neither length
nor maximum length is specified, the choice is the same as in (2).