Chapter Contents

Previous

Next

Data Set Model: _setKey

Argument	Type	Description
rc	N	returns whether the key was successfully applied
		0	key was successfully applied
		not 0	otherwise
keyname	C	specifies the name of the key or index to use on the table. The keyname may specify a single or a composite index.
condition	C	specifies the condition to use when comparing the key value:
		'EQ'	equal to
		'GE'	greater than or equal to
		'GT'	greater than
		'LE'	less than or equal to
		'LT'	less than
scroll	C	specifies whether rows can be retrieved in random order:
		'SCROLL'	rows can be retrieved in random order (default)
		'NOSCROLL'	rows can only be retrieved sequentially
val-list	N	specifies the identifier of an SCL list that contains values to use in the key. The item name should reflect the appropriate column name, and the item value should be the value that the key value is compared against.

Details

If you modify the current row in the table, you must use the _updateRow method before calling the _setKey method.

The _setKey method enables you to set an active key in an open table to a simple or composite key. It establishes a set of criteria for reading table rows by evaluating the value of the columns against the key value in the rows.

Using a composite key with _setKey operates the same way as the _where method only when the condition is EQ. The value returned when the condition is EQ is the same as if the columns specified in the composite key were connected by WHERE conditions that use AND or ALSO.

For all other conditions (GT, GE, LT, or LE) specified with _setKey for a composite key, the composite key columns are concatenated together to form the index key. The number returned by the _keyCount method is the number of rows in the table that satisfy the composite key. For example, if the composite index consists of columns SEX and AGE and the condition is GT (greater than), the values to search for are concatenated such that key values of F for SEX and 13 for AGE yield an index key of F13. Because the search is performed on the concatenated values, some values may meet the search condition that you did not expect, such as SEX of M and AGE of 11, because the string M11 is considered greater than the string F13.

Once an active key is set through the _setKey method, it remains active until the following conditions are met:

• the table is closed

• another key is set

• the current setting is cleared by passing the rc argument alone to the _setKey method.

The table is automatically positioned at the row that meets the specified criteria. Use the _getRow or _fetchRow method to read the row.

The _setKey method cannot be used in conjunction with a WHERE clause.

The _setKey method sets SYSRC for error, note, and warning conditions.

Example

In the following example, the Data Set Model class is being used as a stand-alone model, that is, the model class is not being used with a data form or data table object.

This example creates an index on the STATE variable in the table SASUSER.CRIME. It subsets on STATE values less than 20. In this example, COUNT returns 15.

INIT: dsid=open('sasuser.crime', 'v'); rc=icreate(dsid, 'state', 'state'); dsid=close(dsid); datcl=loadclass('sashelp.fsp.datast_m.class'); datid=instance(datcl);

call send(datid,'_setDataset', 'sasuser.crime');

list=makelist(); list=setnitemn(list, 20, 'state'); call send(datid, '_setKey', rc, 'state', 'LT', 'scroll', list); call send(datid, '_keyCount', rc, count); put count=; return;

TERM: call send(datid,'_term'); list=dellist(list); return;

Chapter Contents

Previous

Next

Top of Page

Copyright 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.