HP 3000 Manuals HP 3000 Manuals
Expression Types
Table 5-1. Expression Types
---------------------------------------------------------------------------------------------
| | |
| SPL | HP C/XL Equivalent |
| | |
---------------------------------------------------------------------------------------------
| | |
| Variables on either side of an operator | Variable types may be mixed in expressions. |
| must be of the same type. Type transfer | Automatic type conversion is performed |
| functions are used to convert types. | prior to execution of an operator. See "HP |
| | C/XL Rules for Automatic Numeric Type |
| | Conversion" in this chapter. |
| | |
| | A "cast" operator may be used to force an |
| | expression to a desired data type, perhaps |
| | for an actual parameter to a function. |
| | |
---------------------------------------------------------------------------------------------
| The type transfer functions are the names | The corresponding cast operators are |
| of the simple variable types, plus two | similar to SPL, except that the type names |
| | |
| additions, in the function form: | are enclosed in parentheses: |
| | |
| BYTE ( double | integer | logical ) | (unsigned char) (expression) |
| DOUBLE ( byte | integer | logical ) | (long int) (expression) |
| FIXR ( real ) rounds to DOUBLE | (No equivalent) |
| FIXT ( real ) truncates to DOUBLE | (long int) (expression) |
| INTEGER ( byte | double | logical ) | (short int) (expression) |
| LOGICAL ( byte | double | integer ) | (unsigned short int) (expression) |
| LONG ( double | real ) | (double) (expression) |
| REAL ( byte | double | integer | | (float) (expression) |
| logical | long ) | |
| | The expression may have any appropriate |
| byte, double, integer, logical, long, and | character or numeric value. The |
| real are the types permitted in the | parentheses around expression may be |
| particular functions. | omitted if it is a single entity. |
| | Conversion from float or double to any char |
| Sometimes more than one function is | or int type is by truncation. There is no |
| required, as in the conversion from REAL to | rounding function. HP C/XL also allows |
| INTEGER, which requires either | other simple data and pointer types in cast |
| "INTEGER(FIXR(real))" or | operations. |
| "INTEGER(FIXT(real))". | |
---------------------------------------------------------------------------------------------
HP C/XL Rules for Automatic Numeric Type Conversion
As an expression is evaluated, HP C/XL performs automatic data type
conversions on the operands. First, each operand is evaluated and
converted, according to Table 5-2.
Table 5-2. Automatic Unary Type Conversions
---------------------------------------------------------------------------------------------
| | |
| Original Type | Converted to |
| | |
---------------------------------------------------------------------------------------------
| | |
| char | int |
| | |
---------------------------------------------------------------------------------------------
| | |
| short int | int |
| | |
---------------------------------------------------------------------------------------------
| | |
| unsigned char | unsigned int |
| | |
---------------------------------------------------------------------------------------------
| | |
| unsigned short int | unsigned int |
| | |
---------------------------------------------------------------------------------------------
| | |
| float | double 1 |
| | |
---------------------------------------------------------------------------------------------
| | |
| int | int 2 |
| | |
---------------------------------------------------------------------------------------------
| | |
| long int | long int 2 |
| | |
---------------------------------------------------------------------------------------------
| | |
| unsigned int | unsigned int 2 |
| | |
---------------------------------------------------------------------------------------------
| | |
| unsigned long int | unsigned long int 2 |
| | |
---------------------------------------------------------------------------------------------
| |
| 1 This conversion from float to double may be prevented with the HP C/XL compiler |
| option "-Wc,-r". See the HP C/XL Reference Manual Supplement for details. |
| |
| 2 These types are not converted. They are included here for completeness. |
| |
---------------------------------------------------------------------------------------------
Second, arithmetic operands in binary operations are converted. If the
two operands are the same type, the conversion is complete. Otherwise,
the process continues row-by-row through Table 5-3 until a conversion
makes the operand types equal.
Table 5-3. Automatic Binary Type Conversions
------------------------------------------------------------------------------------------------
| | | |
| One Operand | Other Operand | Conversion |
| | | |
------------------------------------------------------------------------------------------------
| | | |
| double | any type | Other becomes double |
| | | |
------------------------------------------------------------------------------------------------
| | | |
| float | any type | Other becomes float |
| | | |
------------------------------------------------------------------------------------------------
| | | |
| unsigned long int | any type | Other becomes unsigned long int |
| | | |
------------------------------------------------------------------------------------------------
| | | |
| long int | unsigned int | Both become unsigned long int |
| | | |
------------------------------------------------------------------------------------------------
| | | |
| long int | not unsigned int | Other becomes long int |
| | | |
------------------------------------------------------------------------------------------------
When a value is stored (as in an assignment), it is converted to the
destination type.