Fetches data from the database to be updated

int sm_tm_command ("SELECT [ tableViewName [ tableViewScope ] ]");

tableViewName

The name of a server view in the current transaction. This parameter is case sensitive.

If tableViewName is specified, the command is applied according to the tableViewScope parameter. (Since the entire table view tree might not be included, this is known as a partial command, and sm_tm_command sets TM_FULL to 0.) The specified table view must either be a server view or be the server view to which the desired table view belongs.

If tableViewName is not specified, the command is applied for each table/server view, starting with the root table view. This is known as a full command, and sm_tm_command sets TM_FULL to 1.

tableViewScope

One of the following parameters, which must be preceded by a table view name.

• TV_AND_BELOW which applies the command to the specified table view and all table views below it on the tree. If no parameter is specified, the transaction manager acts as though TV_AND_BELOW was supplied.

• BELOW_TV which applies the command to the table views below the specified table view.

• TV_ONLY which applies the command to the specified table view only.

• SV_ONLY which applies the command only to the table views of the specified server view.

SELECT fetches data from the database so it can be modified. In order to success fully update data or insert new data, all the fields in a server view which are included in the select list need to have the same number of occurrences.

After you choose SELECT, the following steps occur:

1. If you have made changes in the table views on which this command operates in a previous NEW, COPY, COPY_FOR_UPDATE, or SELECT, you are prompted to discard your changes. If you choose OK, changes are discarded and fields in the specified table views are cleared. If you choose Cancel, you return to the screen so you can save your changes.
2. The transaction mode is set to update unless a table view is specified and the mode is not initial mode. By default, update mode protects the primary key fields from data entry and sets the display attributes differently for key and non-key fields.
3. If the Count Select property is set to Yes, the transaction manager issues a SELECT statement using COUNT(*) to find the number of rows in the select set. If this number exceeds the amount set in the Count Threshold property, a message box offers the user the choice of discontinuing data selection.
4. The screen displays the first set of data for all linked table views. When you choose SELECT, the standard transaction models have the SQL generator execute a SELECT statement for the database table named in the root table view and any table views connected to it via a server link. Then, recursively, SELECT statements are issued for the child table views having sequential links, and any table views connected to those child table views by server links.
5. The before image, or snapshot, of the screen is taken for the screen\Qs updatable table views. An updatable table view must have its primary key fields on screen. Any changes made to the screen following this step can then be processed using a SAVE command.

Push buttons and menu selections for the SELECT command can choose to set the class property to view_button. By default, view_button is active in initial or view modes.

If you want to select a specific record or group of records, set the widget's use_in_where property to PV_YES and the type of operator (where_operator) to be used in the WHERE clause. Then, in the transaction manager, choose CLEAR to clear the fields, enter a value in your query field, and then choose SELECT. The screen displays the specified information.

If the server view's Count Select and Count Warning properties are set to Yes, the application will warn users about large select sets; however, the SELECT statement is performed twice, and the tables must remain locked for the result to be the same for both statements.

To save the changes or additions made to the selected data, choose SAVE as the next transaction command.

To display the next row of information, choose CONTINUE as the next transaction command. If you have updated the data on the screen, you are prompted to discard your changes. If you choose OK, changes are discarded. If you choose Cancel, you return to the screen so you can save your changes.

To discard any changes you have made to the screen, choose CLOSE or FORCE_CLOSE. For some database engines, such as SYBASE CT-Lib, the CLOSE command does not release the database locks when a SELECT command is not followed by a SAVE command. In this case, follow the CLOSE with the RELEASE command which gives up the locks on the database.

The following request events can be generated by the SELECT command to ascertain whether the changes from the previous command have been saved and, if desired, discard those changes:

• TM_PRE_CLOSE (described under CLOSE)

• TM_CLOSE (described under CLOSE)

• TM_QUERY (described under CLOSE)

• TM_DISCARD (described under CLOSE)

• TM_POST_CLOSE (described under CLOSE)

The SELECT command generates TM_CLEAR requests if TM_SELECT for a parent table view returns no data. In that case, TM_CLEAR is generated for all subordinate table views, but not for table views at the same level of the tree. TM_CLEAR requests are described under CLEAR.

Table 8-31 Table 29. the SELECT command.

If TM_SELECT for a parent table view returns no data, TM_CLEAR requests are generated for all subordinate table views, but not for table views at the same level of the tree. TM_CLEAR requests are described under CLEAR.