Inc and Dec

From Lazarus wiki
Revision as of 15:06, 26 May 2018 by Kai Burghardt (talk | contribs) (unify Inc and Dec)
Jump to navigationJump to search

Template:Translate The procedures inc and dec increment or decrement a given variable by default by one.

usage

The first parameter specifies an ordinal value variable (e.g. an integer or enumeration type) and the second optional parameter may specify a different addend/subtrahend.

 1program incDecDemo(input, output, stderr);
 2
 3type
 4	primaryColor = (red, green, blue);
 5
 6var
 7	phase: primaryColor;
 8	x: longint;
 9
10begin
11	// enumeration type
12	phase := red;
13	inc(phase); // phase becomes green
14	writeLn(phase);
15	
16	// integer
17	x := 1;
18	inc(x, -1); // x becomes zero
19	writeLn(x);
20	
21	x := 1;
22	dec(x); // same as above: x becomes zero
23	writeLn(x);
24end.

background

In FPC's system unit the inc and dec procedures are compiler procedures. They exist in order to optimize for certain architectures where faster inc and dec assembler instructions are available.

special behaviors

If {$rangeChecks} are turned on, those procedures may generate a run-time error (RTE 201). Also {$overflowChecks} may be generated (with a small difference in TP mode).

If pointer arithmetics are allowed by the {$pointerMath} compiler switch, inc and dec work on pointers, too. In the case of typed pointers, e.g. a pointer to a record, the target's type size is considered automatically. For instance

 1program pointerIncDemo(input, output, stderr);
 2
 3{$pointerMath on}
 4
 5var
 6	p: PQWord;
 7
 8begin
 9	inc(p);
10	inc(p, 3);
11end.

will generate (excerpt):

13; [pointerIncDemo.pas]
14; [8] begin
15	leaq	-8(%rsp),%rsp
16.Lc3:
17; Var p located in register rax
18	call	FPC_INITIALIZEUNITS
19	movq	$0,%rax
20; [9] inc(p);
21	addq	$8,%rax
22; [10] inc(p, 3);
23	addq	$24,%rax
24; [11] end.
25	call	FPC_DO_EXIT
26	leaq	8(%rsp),%rsp
27	ret

see also