Fields of the Interface control block

Use the fields of the interface control block to pass information to and from your exit routine. Although there are separate interface control blocks that work with Assembly, PL/I, or COBOL, the fields of the interface control block are standard regardless of the programming language your edit exit routine is written in. These fields are shown in Table 69. Unless otherwise stated, each field relates to all formatting calls.

These same fields appear in each sample program (one for each programming language supported) shipped with QMF. You can include these field names in your own source program. The QMF production disk contains the sample programs.

Table 69. Fields of the QMF interface control block
Name	Contents
ECSDECPT	Contains the current decimal point symbol as determined by the DECOPT option of PROFILE (period or comma).
ECSECODE	Contains the user edit code.
ECSERRET	Contains a zero at the point of call. Set this to a nonzero return code to record an error. Use one of the values on the following list for an error of the indicated type: Number Error 99101 Unrecognized edit code 99102 Improper input data type for edit code 99103 Invalid input value for item to be formatted 99104 Item to be formatted is too short 99105 Not enough room for result in ECSRSLT (result is too wide for the space allotted) The error codes listed (and their associated messages and help panels) are specific to the error. For any other code, a general error message, with a general backup help panel, is displayed.
ECSFREQ	Holds E for a formatting call, T for a termination call.
ECSINLEN	Contains the length, in bytes, of the value to be formatted.
ECSINNUL	Holds an N if the value to be formatted is null.
ECSINPRC	Contains the precision of the value to be formatted. Applies only to U-type codes when the data type is DECIMAL, or to V-type codes when the character string to be formatted was derived from numeric data.
ECSINSCL	Contains the scale of the value to be formatted. Applies only to U-type codes when the data type is DECIMAL, or to V-type codes when the character string to be formatted was derived from numeric data.
ECSINSGN	Holds the sign of a converted numeric value (blank or - ). Applies only to V codes when the character string to be formatted was derived from numeric data.
ECSINTYP	Indicates, in database terms, how the value to be formatted is represented. Applies to edit codes of every type. Values can be: 384 DATE data type 388 TIME data type 392 TIMESTAMP data type 448 VARCHAR data type 452 CHAR data type 456 LONG VARCHAR data type 464 VARGRAPHIC data type 468 GRAPHIC data type 472 LONG VARGRAPHIC data type 480 FLOAT data type 484 DECIMAL data type 496 INTEGER data type 500 SMALLINT data type 940 Extended floating point data type The extended floating point data type is not supported by the database (or by COBOL); it is limited to functions such as AVERAGE and STDEV. Extended floating point values are precise to more than 30 digits.
ECSNAME	Contains the name of the control block, which is DXEECS. Serves as an eye catcher in storage dumps.
ECSRQMF	Set this to T to request a termination call.
ECSRSLEN	Contains the length of the output area, in bytes. (Value is taken from the column WIDTH in the FORM)
ECSTHSEP	Contains the thousands separator as determined by the DECOPT option of PROFILE (blank or a comma).
ECSUSERS	A 256-byte scratchpad area where your exit routine can record information that persists from one call to the next. On the first call after the edit routine is loaded, this field contains binary zeros.

[ Previous Page | Next Page | Contents | Index ]