Datasets

The first field of each dataset must always be null. The name of this field determines how to interpret this dataset.

Field naming conventions

Any name in the system can consist of several (from one to three) items. The name items are distinguished by the symbol ! (exclamation point). Note that in SQL-Server, names that contain wildcard characters, spaces, or language keywords must be enclosed in square brackets. All parts of the name are optional, but delimiters are always required. In other words, if, for example, a field name begins with !!, then this means that the name is not specified (the field will not be in the model), there is no type, but only a modifier.

• The first item of the name is always the name of the property (possibly compound) in the model.
• The second item is the type name (for read/write) or the path to a property in the model (for metadata).
• The third item is a modifier that describes either the type of the field or its special purpose.

Recordset types

The set type is described by the third item (suffix) of the first field of each recordset.

Type	Description
Object	Single item.
Array	An array of items. If this array is a child, then the set must contain a property that defines the parent item (!ParentId). Read more...
LazyArray	"Lazy" array of child items. Loaded the first time a property is accessed. Read more...
Map	References. Always a child. May not have a name. The set must always contain an identifier (property with the !Id modifier). It is by this field that the binding with the source is carried out.
Tree	Item tree (static or dynamic). The set must contain properties with modifiers !Id, !ParentId, !Items. Read more...
Group	The item tree that is built by means of the flat table grouping. The set must contain properties of !GroupMarker and !Items types. Read more...
CrossArray	The cross array of items. Always a child. The set must contain a property that defines a parent item (!ParentId) and the !Key key property. Read more...