B
Datatype Support

The Oracle Application Server Integration Adapter for IMS/DB supports a number of datatypes that are used when defining metadata for both the IMS/DB data source and IMS/DB backend adapter in Oracle Studio. The datatypes are mapped from the COBOL datatypes during the import procedure.

Note: The mapping of datatypes between Oracle Application Server Integration Adapter for IMS/DB and Oracle Application Server ProcessConnect is performed internally by Oracle Connect for IMS/DB.

This appendix contains the following sections:

• Datatype Mapping for the IMS/DB Data Source

• IMS/DB Data Source Metadata Datatypes

• Datatype Mapping for the Oracle Connect for IMS/DB

Datatype Mapping for the IMS/DB Data Source

The following table shows the mapping performed to generate metadata for a IMS/DB data source from COBOL copybooks.

The COBOL datatype COMP, in the table is an abbreviation for, and synonymous with, COMPUTATIONAL. Square brackets ([ ]) denote optional qualifiers for some COBOL compilers, which may not be allowed for other COBOL compilers.

Table B-1  Datatype Mapping: COBOL and IMS/DB Data Source

COBOL Datatype	Oracle Connect Datatype	Details
BINARY (3 <= size < 5 with fractional data)	scaled_int2	Signed word integer Size = 2 bytes Range: -32768 to +32767
BINARY (3 <= size < 5 without fractional data)	int2	Signed word integer Size = 2 bytes
BINARY (5 <= size < 10 with fractional data)	scaled_int4	Signed longword Size = 4 bytes Range: -2147483648 to [phone]
BINARY (5 <= size < 10 without fractional data)	int4	Signed longword Size = 4 bytes
BINARY (size < 3 with fractional data)	scaled_int1	Signed byte integer Size = 1 byte Range: -128 to +127
BINARY (size < 3 without fractional data)	int1	Signed byte integer Size = 1 byte
BINARY (size >= 10 with fractional data)	scaled_int8	Signed quadword Size = 8 bytes Range: -9,223,372,036,854,775,808 to +9,223,372,036,854,775,807
BINARY (size >= 10 without fractional data)	int8	Signed quadword Size = 8 bytes
COMP (3 <= size < 5 with fractional data)	scaled_int2	Signed word integer Size = 2 bytes Range: -32768 to +32767
COMP (3 <= size < 5 without fractional data)	int2	Signed word integer Size = 2 bytes
COMP (5 <= size < 10 with fractional data)	scaled_int4	Signed longword Size = 4 bytes Range: -2147483648 to [phone]
COMP (5 <= size < 10 without fractional data)	int4	Signed longword Size = 4 bytes
COMP (size < 3 with fractional data)	scaled_int1	Signed byte integer Size = 1 byte Range: -128 to +127
COMP (size < 3 without fractional data)	int1	Signed byte integer Size = 1 byte
COMP-2	dfloat	Double floating point number Size = 8 bytes Range: 0.29E-38 to 1.7E38 Precision: 16 digits
COMP-3	decimal	Packed decimal Maximum number of digits = 31 Maximum fractions = 11 Length = int (number of digits/2) + 1 bytes
COMP-4 (5 <= size < 10 with fractional data)	scaled_int4	Signed longword Size = 4 bytes Range: -2147483648 to [phone]
COMP-4 (5 <= size < 10 without fractional data)	int4	Signed longword Size = 4 bytes
COMP-4 (size < 5 with fractional data)	scaled_int2	Signed word integer Size = 2 bytes Range: -32768 to +32767
COMP-4 (size < 5 without fractional data)	int2	Signed word integer Size = 2 bytes
COMP-4 (size >= 10 with fractional data)	scaled_int8	Signed quadword Size = 8 bytes Range: -9,223,372,036,854,775,808 to +9,223,372,036,854,775,807
COMP-4 (size >= 10 without fractional data)	int8	Signed quadword Size = 8 bytes
COMP-5 (5 <= size < 10 with fractional data)	scaled_int4	Signed long integer Size = 4 bytes Range: -2147483648 to [phone]
COMP-5 (5 <= size < 10 without fractional data)	int4	Signed long integer Size = 4 bytes
COMP-5 (size < 5 with fractional data)	scaled_int2	Signed word integer Size = 2 bytes Range: -32768 to +32767
COMP-5 (size < 5 without fractional data)	int2	Signed word integer Size = 2 bytes
COMP-5 (size >= 10 with fractional data)	scaled_int8	Signed quadword Size = 8 bytes Range: -9,223,372,036,854,775,808 to +9,223,372,036,854,775,807
COMP-5 (size >= 10 without fractional data)	int8	Signed quadword Size = 8 bytes
COMP-X (3 <= size < 5 with fractional data)	scaled_int2	Signed word integer Size = 2 bytes Range: -32768 to +32767
COMP-X (3 <= size < 5 without fractional data)	int2	Signed word integer Size = 2 bytes
COMP-X (5 <= size < 10 with fractional data)	scaled_int4	Signed long integer Size = 4 bytes Range: -2147483648 to [phone]
COMP-X (5 <= size < 10 without fractional data)	int4	Signed long integer Size = 4 bytes
COMP-X (size < 3 with fractional data)	scaled_int1	Signed byte integer Size = 1 byte Range: -128 to +127
COMP-X (size < 3 without fractional data)	int1	Signed byte integer Size = 1 byte
COMP-X (size >= 10 with fractional data)	scaled_int8	Signed quadword Size = 8 bytes Range: -9,223,372,036,854,775,808 to +9,223,372,036,854,775,807
COMP-X (size >= 10 without fractional data)	int8	Signed quadword Size = 8 bytes
INDEX	int4	Signed long integer Size = 4 bytes
[SIGN [IS]] LEADING	numstr_nl	Signed numeric string. Sign is the first character of the string Maximum number of digits = 31 Maximum fractions = 11
[SIGN [IS]] LEADING SEPARATE [CHARACTER]	numstr_nl	Signed numeric string. Sign is the first character of the string Maximum number of digits = 31 Maximum fractions = 11
NATIVE-2	int2	Signed word integer Size = 2 bytes
NATIVE-4	int4	Signed long integer Size = 4 bytes
NATIVE-8	scaled_int8	Signed quadword Size = 8 bytes Range: -9,223,372,036,854,775,808 to +9,223,372,036,854,775,807
PACKED-DECIMAL	decimal	Packed decimal Maximum number of digits = 31 Maximum fractions = 11 Length = int (number of digits/2) + 1 bytes
POINTER	int4	Signed long integer Size = 4 bytes
[SIGN [IS]] TRAILING	numstr_nr	Signed numeric string. Sign is the last character of the string. Maximum number of digits = 31 Maximum fractions = 11
[SIGN [IS]] TRAILING SEPARATE [CHARACTER]	numstr_nr	Signed numeric string. Sign is the last character of the string Maximum number of digits = 31 Maximum fractions = 11

IMS/DB Data Source Metadata Datatypes

The following table shows all the valid datatypes that can be used to describe the metadata for a IMS/DB data source.

Note: This section is for reference only. The only datatypes that should be used are those generated for the COBOL copybook, as described in "Datatype Mapping for the IMS/DB Data Source".

Table B-2  IMS/DB Metadata Datatypes

Oracle Connect Datatype	Details
bit	Single bit within a byte Size = 1 byte Format: <field name="name" datatype="bits" onBit="n" ... />, where n specifies the bit (within a byte) the field uses
bits	A signed number of bits within a byte Size = 1 bit to 1 byte Format: <field name="name" datatype="bits" onBit="n" size="m"... />, where n specifies the bit (within a byte) to start from and m is the number of bits. If n is not specified then n defaults to 1 for the first occurrence of the field and is contiguous thereafter. The maximum number of bits you can map is 32
cstring	Null-terminated string of alphanumeric characters; maximum length must be specified. An extra byte is required for the null flag
date6	Date in a string having the form YYMMDD
date8	Date in a string having the form YYYYMMDD
decimal	Packed decimal Maximum number of digits = 31 Maximum fractions = 11 Length = int (number of digits/2) + 1 bytes
dfloat	Double floating point number Size = 8 bytes Range: 0.29E-38 to 1.7E38 Precision: 16 digits
double	Double floating number Size = 8 bytes Range: 0.56E-308 to 0.90E308 Precision: 15 digits
filler	Allocation for future use, string type; length must be specified
image	Image data (Binary large object)
int_date	Date in a four byte integer having the form YYMMDD or YYYYMMDD Example, 23-Jul-1998 has the form: 980723 or 19980723
int1	Signed byte integer Size = 1 byte
int2	Signed word integer Size = 2 bytes
int3	Signed integer Size = 3 bytes
int4	Signed longword Size = 4 bytes
int6	Signed integer Size = 6 bytes
int8	Signed quadword Size = 8 bytes
jdate	Julian date Size = 2 bytes Bits 0-6: (non-century) year Bits 7-15: day of the year
logical	Signed long integer Values: 1 for true, 0 for false.
numstr_bdn	Signed numeric string. Sign is the first character of the string. Maximum number of digits = 31 Maximum fractions = 11 Note: the number of fractions includes the decimal point
numstr_nl	Signed numeric string. Sign is the first character of the string Maximum number of digits = 31 Maximum fractions = 11
numstr_nlo	Signed numeric string. A left overpunched sign is implemented. Maximum number of digits = 31 Maximum fractions = 11
numstr_nr	Signed numeric string. Sign is the last character of the string. Maximum number of digits = 31 Maximum fractions = 11
numstr_s	Signed numeric string. A right overpunched sign is implemented Maximum number of digits = 31 Maximum fractions = 11
numstr_u	Unsigned numeric string Maximum number of digits = 31 Maximum fractions = 11
numstr_zoned	Signed numeric string Maximum number of digits = 31 Maximum fractions = 11
phdate	Size = 2 bytes Bits 0-6: (non-century) year Bits 7-10: number of month Bits 11-15: day of month
scaled_int1	Signed byte integer Size = 1 byte Range: -128 to +127
scaled_int2	Signed word integer Size = 2 bytes Range: -32768 to +32767
scaled_int3	Signed integer Size = 3 bytes
scaled_int4	Signed long integer Size = 4 bytes Range: -2147483648 to [phone]
scaled_int6	Signed integer Size = 6 bytes
scaled_int8	Signed quadword Size = 8 bytes Range: -9,223,372,036,854,775,808 to +9,223,372,036,854,775,807
scaled_uint1	Unsigned byte integer Size = 1 byte Range: 0 to 254
scaled_uint2	Unsigned word integer Size = 2 bytes Range: 0 to 65534
scaled_uint4	Unsigned long integer Size = 4 bytes Range: 0 to +4,294,967,294.
scaled_uint4	Unsigned long integer Size = 4 bytes Range: 0 to 4294967294
str_date	Atomic date string Size = 10 characters Format = YYYY-MM-DD
str_time	Atomic date string Size = 8 characters Format = HH:MM:SS
string	String of alphanumeric characters; length must be specified
text	Text data (Binary large object)
ubits	An unsigned number of bits within a byte Size = 1 bit to 1 byte Format: <field name="name" datatype="bits" onBit="n" size="m" ... />, where n specifies the bit (within a byte) to start from and m is the number of bits. If n is not specified then n defaults to 1 for the first occurrence of the field and is contiguous thereafter The maximum number of bits you can map is 31
uint2	Unsigned word integer Size = 2 bytes Range: 0 to +65534.
uint6	Unsigned integer Size = 6 bytes
unicode	Null-terminated alphanumeric unicode string; maximum length must be specified
unspecified	Unknown datatype, string type, length must be specified
varstring	16 bit count, followed by a string
varstring4	32 bit count, followed by a string

Datatype Mapping for the Oracle Connect for IMS/DB

The following table shows the mapping performed to generate metadata for the IMS/DB backend adapter from COBOL copybooks.

The COBOL datatype COMP, in the table is an abbreviation for, and synonymous with, COMPUTATIONAL. Square brackets ([ ]) denote optional qualifiers for some COBOL compilers, which may not be allowed for other COBOL compilers.

Table B-3  Datatype Mapping: COBOL and Oracle Connect for IMS/DB

COBOL Datatype	Xsd Datatype
BINARY (with fractional data)	string
BINARY (without fractional data)	int
COMP (with fractional data)	string
COMP (without fractional data)	int
COMP-2	double
COMP-3	string
COMP-4 (with fractional data)	string
COMP-4 (without fractional data)	int
COMP-5 (with fractional data)	string
COMP-5 (without fractional data)	int
COMP-X (with fractional data)	string
COMP-X (without fractional data)	int
INDEX	int
[SIGN [IS]] LEADING	string
[SIGN [IS]] LEADING SEPARATE [CHARACTER]	string
NATIVE-2	int
NATIVE-4	int
NATIVE-8	string
PACKED-DECIMAL	string
POINTER	int
[SIGN [IS]] TRAILING	string
[SIGN [IS]] TRAILING SEPARATE [CHARACTER]	string