In database applications, the set of fields or columns, their types, their names, and any other descriptive information is often called a Schema. Fields can be one of five types:
Float/Double: The field is a floating point number. Float and Double are similar, except that double precision fields have more precision than float fields. The actual storage precision will depend on the format.
Integer: The field is an integral number.
String: The field is a variable length string. Many formats (i.e., .dbf) only have fixed length string fields, in which case, the formatting information will indicate the length of the field.
CountedInt: The field contains a list of numbers. This is typically used to maintain lists of related shapes (e.g., ArcIdList in a topological area layer, or RingStart for multi-ring polygons).
Many databases have additional field types that are either mapped onto GeoGateway field types or ignored (dates are usually treated as an integer or a string).
Each field has a field name associated with it. There are no fixed length restrictions on the name, but it is intended to be short, and will be used as a column header in many applications.
Each field has a format specification, which indicates the number of characters that should be used for the field when it is displayed in reports or dialogues. It also indicates whether the field should be left or right justified. For the cases of Float and Double variables, it indicates how many decimal places of precision should be shown. This value is a clue to applications, and does not have to be used. For some formats the length and precision fields imply actual limitations of the underlying format, while for others it is just a convenience for display purposes.
Each field also has a default value. When not otherwise established, it is zero for numeric values, an empty string for string values, and an empty list for CountedInt lists.