Lets the user choose a value from a dropdown filled with predefined values.
The values can be combined with friendly labels, and in that case
the labels will be displayed to the user instead of the value.
In its closed state, the user can still select an item in the list using the keyboard. When the dropdown control has keyboard focus, typing a key will choose the entry that starts with that character (rapidly typing a whole sequense also works). The user can also navigate in the list with the Up/Down arrows, PgUp/PgDown and Home/End buttons.
The user can choose among the list's items using mouse or keyboard, but cannot enter any value not in the list.
SQL-statement that fills the control with possible values.
Retrieves the default value for the control.
This call is only made if there is a field validation set for the field info and the field has any content.
Performs field validation when the user leaves the field or one of its dependencies is changed, initial values set by default value and initial values in edit-mode are not validated.
When saving the validation runs server side if the field value has changed. A field value is considered changed if in new mode the value is anything other than
NULL. In edit mode it is considered changed if it has a value that was not returned by the GetEditFields procedure.
For stable and predictable behaviour, always use an explicit ORDER BY clause.
The name depends on the usage. If the control is used to retrieve values from a lookup table then the procedure should be named "<Schema>.<LookupTable>Dropdown". If it is not and the values are not directly linked to rows in a table or the functionality is specific for the current field (say for instance a choice between each of the last six years) then the procedure should be named "<Schema>.<Table><Field>_Dropdown".
The use of "HideUnlessCurrent" is meant to quickly provide a description for the current value in a dropdown, even when the value otherwise is not selectable in the stored procedure. Use freely when you only have small number of rows (less than 100). But if you have large amounts of inactive rows it will have an impact on performance. If you do, it's preferable to handle the missing id separately instead.