IBM OS/390 manual Open Damfile Read DAMFILE,KEY, Damfile Addition Write DAMFILE,AFTER

To create a file using this method under MVS, you would normally initialize each track by writing a capacity record (R0) and erasing the 2.sp of the track. In VSE, you would do this by using the WRITE RZERO macro; in MVS you use the WRITE SZ macro. However, in MVS, you need not update the track address because this is done automatically by the WRITE SZ macro. By testing register 15 for a non-zero value after each WRITE, you can determine when MVS has initialized all the tracks. Also, you need a second sequential DCB (DSORG=PS) for the WRITE SZ macro. An example of this procedure is shown in Figure 49 on

page 321. The example also shows how secondary space allocation can be obtained if an out-of-extent condition occurs while you are creating the data set.

Processing a DAM File under VSE

Figure 50 illustrates how a DA file that has been loaded sequentially under VSE may be processed. Records are retrieved for updating purposes by key and the relative track number. When the record-not-found condition occurs, the transaction record whose key was used for the search is added to the disk file by a WRITE AFTER.

	OPEN	DAMFILE
	.
	READ	DAMFILE,KEY
	WAITF	DAMFILE
	.
	.
	WRITE	DAMFILE,KEY
	WAITF	DAMFILE
	.
	.
ADDITION WRITE		DAMFILE,AFTER
	WAITF	DAMFILE
	.
	CLOSE	DAMFILE
	.
DAMFILE	DTFDA	BLKSIZE=58,ERRBYTE=ERROR,	C
		IOAREA1=OUTPUT,SEEKADR=ADDR,	C
		TYPEFLE=INPUT,AFTER=YES,DSKXTNT=3,	C
		KEYARG=KEY,KEYLEN=3,VERIFY=YES,	C
		READKEY=YES,RELTYPE=HEX,	C
		WRITEKY=YES,DEVICE=3340
	DAMOD	AFTER=YES,ERREXT=YES,RELTRK=YES
Figure 50.	Processing a DAM file under VSE

To process a randomly loaded file, use a similar process, but use READ ID for retrieving records and WRITE ID for updating and adding records.

Processing a DAM File under MVS

The procedure for adding records to a BDAM data set is similar to the one illustrated in Figure 50. The computation of the block address field varies according to the reference method used. For example, if the data set had been created sequentially, as in Figure 47 on page 320, record reference by block address only can be used. In this case, the coding might be as illustrated in Figure 42 on page 317.

324VSE to OS/390 Migration Workbook