Bitmap: Difference between revisions
| Line 4,121: | Line 4,121: | ||
<pre>cmd:> main.exe |
<pre>cmd:> main.exe |
||
(B:51,G:51,R:255)</pre> |
(B:51,G:51,R:255)</pre> |
||
=={{header|Smalltalk}}== |
|||
<lang Smalltalk> |
|||
</lang> |
|||
=={{header|Tcl}}== |
=={{header|Tcl}}== |
||
Revision as of 15:34, 29 May 2021
You are encouraged to solve this task according to the task description, using any language you may know.
Show a basic storage type to handle a simple RGB raster graphics image, and some primitive associated functions.
If possible provide a function to allocate an uninitialised image, given its width and height, and provide 3 additional functions:
- one to fill an image with a plain RGB color,
- one to set a given pixel with a color,
- one to get the color of a pixel.
(If there are specificities about the storage or the allocation, explain those.)
These functions are used as a base for the articles in the category raster graphics operations,
and a basic output function to check the results
is available in the article write ppm file.
ActionScript
ActionScript 3 has a BitmapData class (in the flash.display package) which can be used for storage and handling of bitmap images.
To display these images, the Bitmap class can be used.
<lang ActionScript3> // To import the BitmapData class: import flash.display.BitmapData;
// Creates a new BitmapData object with a width of 500 pixels and a height of 300 pixels. var bitmap:BitmapData = new BitmapData(500, 300);
// Create a BitmapData with transparency disallowed var opaqueBitmap:BitmapData = new BitmapData(500, 300, false);
// Bitmap with initial fill colour, as 0xAARRGGBB (default is white) var redFilledBitmap:BitmapData = new BitmapData(400, 300, true, 0xFFFF0000);
// Get the colour value of the pixel at point (200, 200) bitmap.getPixel(200, 200) // As 0xRRGGBB bitmap.getPixel32(200, 200) // As 0xAARRGGBB
// Set the colour value of the pixel at point (300, 200) to blue bitmap.setPixel(300, 200, 0x0000FF); // As 0xRRGGBB bitmap.setPixel32(300, 200, 0xFF0000FF); // As 0xAARRGGBB
// Fill the bitmap with a given colour (as 0xAARRGGBB) after construction bitmap.fillRect(bitmap.rect, 0xFF44FF44); </lang>
Ada
The package interface: <lang ada>package Bitmap_Store is
type Luminance is mod 2**8;
type Pixel is record
R, G, B : Luminance := Luminance'First;
end record;
Black : constant Pixel := (others => Luminance'First);
White : constant Pixel := (others => Luminance'Last);
type Image is array (Positive range <>, Positive range <>) of Pixel;
procedure Fill (Picture : in out Image; Color : Pixel);
procedure Print (Picture : Image);
type Point is record
X, Y : Positive;
end record;
end Bitmap_Store;</lang> The implementation of: <lang ada>with Ada.Text_IO; use Ada.Text_IO;
package body Bitmap_Store is
procedure Fill (Picture : in out Image; Color : Pixel) is
begin
for p of Picture loop x:= Color;end loop;
end Fill;
procedure Print (Picture : Image) is
begin
for I in Picture'Range (1) loop
for J in Picture'Range (2) loop
Put (if Picture (I, J) = White then ' ' else 'H');
end loop;
New_Line;
end loop;
end Print;
end Bitmap_Store;</lang> This can be used like: <lang ada>use Bitmap_Store; with Bitmap_Store;
... X : Image (1..64, 1..64);
begin
Fill (X, (255, 255, 255)); X (1, 2) := (R => 255, others => 0); X (3, 4) := X (1, 2);</lang>
ALGOL 68
Note: short and shorten need to be tuned (added or removed) to match the underlying graphic hardware colour depth.
File: prelude/Bitmap.a68<lang algol68># -*- coding: utf-8 -*- #
MODE PIXEL = STRUCT(#SHORT# BITS red,green,blue); MODE POINT = STRUCT(INT x,y);
MODE IMAGE = [0,0]PIXEL; # instance attributes #
MODE CLASSIMAGE = STRUCT ( # class attributes #
PIXEL black, red, green, blue, white, PROC (REF IMAGE)REF IMAGE init, PROC (REF IMAGE, PIXEL)VOID fill, PROC (REF IMAGE)VOID print,
- virtual: #
REF PROC (REF IMAGE, POINT, POINT, PIXEL)VOID line, REF PROC (REF IMAGE, POINT, INT, PIXEL)VOID circle, REF PROC (REF IMAGE, POINT, POINT, POINT, POINT, PIXEL, UNION(INT, VOID))VOID cubic bezier
);
CLASSIMAGE class image = (
# black = # (#SHORTEN# 16r00, #SHORTEN# 16r00, #SHORTEN# 16r00),
# red = # (#SHORTEN# 16rff, #SHORTEN# 16r00, #SHORTEN# 16r00),
# green = # (#SHORTEN# 16r00, #SHORTEN# 16rff, #SHORTEN# 16r00),
# blue = # (#SHORTEN# 16r00, #SHORTEN# 16r00, #SHORTEN# 16rff),
# white = # (#SHORTEN# 16rff, #SHORTEN# 16rff, #SHORTEN# 16rff),
# PROC init = # (REF IMAGE self)REF IMAGE:
BEGIN
(fill OF class image)(self, black OF class image);
self
END,
# PROC fill = # (REF IMAGE self, PIXEL color)VOID:
FOR x FROM 1 LWB self TO 1 UPB self DO
FOR y FROM 2 LWB self TO 2 UPB self DO
self[x,y] := color
OD
OD,
# PROC print = # (REF IMAGE self)VOID:
printf(($n(UPB self)(3(16r2d))l$, self)),
- virtual: #
# REF PROC line = # LOC PROC (REF IMAGE, POINT, POINT, PIXEL)VOID, # REF PROC circle = # LOC PROC (REF IMAGE, POINT, INT, PIXEL)VOID, # REF PROC cubic bezier = # LOC PROC (REF IMAGE, POINT, POINT, POINT, POINT, PIXEL, UNION(INT, VOID))VOID
);
OP CLASSOF = (IMAGE image)CLASSIMAGE: class image; OP INIT = (REF IMAGE image)REF IMAGE: (init OF (CLASSOF image))(image);
SKIP</lang>File: test/Bitmap.a68<lang algol68>#!/usr/bin/a68g --script #
- -*- coding: utf-8 -*- #
- The test program ###
PR READ "prelude/Bitmap.a68" PR;
test:(
REF IMAGE x := INIT LOC[1:16, 1:16]PIXEL; (fill OF class image) (x, white OF class image); (print OF class image) (x)
)</lang>
- Output:
(A 16x16 white block)
ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff
AutoHotkey
<lang AutoHotkey>test: blue := color(0,0,255) ; rgb cyan := color(0,255,255) blue_square := Bitmap(10, 10, blue) cyanppm := Bitmap(10, 10, cyan) x := blue_square[4,4] ; get pixel(4,4) msgbox % "blue: 4,4,R,G,B, RGB: " x.R ", " x.G ", " x.B ", " x.rgb() blue_square[4,4] := cyan ; set pixel(4,4) x := blue_square[4,4] ; get pixel(4,4) blue_square.write("blue.ppm") return
Bitmap(width = 1, height = 1, background = 0) { global black black := color(0,0,0) if !background background := black
static BitmapType
if !BitmapType
BitmapType
:= Object("fill", "Bitmap_Fill"
,"write", "Bitmap_write_ppm3")
img := Object("width", width
,"height", height
, "base" , BitmapType)
img._SetCapacity(height) ; an array of rows
img.fill(background)
Return img }
Bitmap_Fill(bitmap, color)
{
r := color.r
g := color.g
b := color.b
loop % bitmap.height
{
height := A_Index
loop % bitmap.width
{
width := A_Index
bitmap[height, width] := color(r, g, b)
}
}
return bitmap
}
Bitmap_write_ppm3(bitmap, filename) { file := FileOpen(filename, 0x11) ; utf-8, write file.seek(0,0) file.write("P3`n" . bitmap.width . " " . bitmap.height . "`n" . "255`n")
loop % bitmap.height
{
height := A_Index
loop % bitmap.width
{
width := A_Index
color := bitmap[height, width]
file.Write(color.R . " ")
file.Write(color.G . " ")
file.Write(color.B . " ")
}
file.write("`n")
}
file.close()
return 0
}
Color(r, g, b) {
static ColorType
if !ColorType
ColorType
:= Object("rgb" , "Color_rgb")
return Object("r" , r, "g", g, "b", b
, "base" , ColorType)
; return Object("r" , r, "g", g, "b", b, "rgb", "Color_rgb")
}
Color_rgb(clr) { return clr.R << 16 | clr.G << 8 | clr.B }</lang>
Axe
All of the functions specified in the task are built in to Axe. Note that bitmaps are always 96x64 black and white. Thus, since each pixel takes 1 bit, a complete bitmap is 768 bytes.
Two bitmaps can be masked together to create 3- and 4-color grayscale.
<lang axe>Buff(768)→Pic1 Fill(Pic1,768,255) Pxl-Off(45,30,Pic1)
.Display the bitmap to demonstrate Copy(Pic1) DispGraph Pause 4500
Disp pxl-Test(50,50,Pic1)▶Dec,i</lang>
BASIC256
<lang BASIC256>graphsize 30,30 call fill(rgb(255,0,0)) call setpixel(10,10,rgb(0,255,255)) print "pixel 10,10 is " + pixel(10,10) print "pixel 20,20 is " + pixel(20,10)
imgsave "BASIC256_bitmap.png" end
subroutine fill(c) color c rect 0,0,graphwidth, graphheight end subroutine
subroutine setpixel(x,y,c) color c plot x,y end subroutine</lang>
- Output:
pixel 10,10 is 4278255615 pixel 20,20 is 4294901760
BBC BASIC
BBC BASIC expects a bitmap always to be associated with a window; for simplicity this code uses the main output window. <lang bbcbasic> Width% = 200
Height% = 200
REM Set window size:
VDU 23,22,Width%;Height%;8,16,16,128
REM Fill with an RGB colour:
PROCfill(100,150,200)
REM Set a pixel:
PROCsetpixel(100,100,255,255,0)
REM Get a pixel:
rgb% = FNgetpixel(100,100)
PRINT RIGHT$("00000" + STR$~rgb%, 6)
END
DEF PROCfill(r%,g%,b%)
COLOUR 1,r%,g%,b%
GCOL 1+128
CLG
ENDPROC
DEF PROCsetpixel(x%,y%,r%,g%,b%)
COLOUR 1,r%,g%,b%
GCOL 1
LINE x%*2,y%*2,x%*2,y%*2
ENDPROC
DEF FNgetpixel(x%,y%)
LOCAL col%
col% = TINT(x%*2,y%*2)
SWAP ?^col%,?(^col%+2)
= col%</lang>
C
Working excerpt from imglib.h usable as "interface" (some includes are needed for other functions of the same category). This code uses functions from category Raster graphics operations. One must create files imglib.h and imglib.c using code from these pages. Start from bitmap page
<lang c>#ifndef _IMGLIB_0
- define _IMGLIB_0
- include <stdio.h>
- include <stdlib.h>
- include <sys/types.h>
- include <string.h>
- include <math.h>
- include <sys/queue.h>
typedef unsigned char color_component; typedef color_component pixel[3]; typedef struct {
unsigned int width; unsigned int height; pixel * buf;
} image_t; typedef image_t * image;
image alloc_img(unsigned int width, unsigned int height); void free_img(image); void fill_img(image img,
color_component r,
color_component g,
color_component b );
void put_pixel_unsafe(
image img,
unsigned int x,
unsigned int y,
color_component r,
color_component g,
color_component b );
void put_pixel_clip(
image img,
unsigned int x,
unsigned int y,
color_component r,
color_component g,
color_component b );
- define GET_PIXEL(IMG, X, Y) (IMG->buf[ ((Y) * IMG->width + (X)) ])
- endif</lang>
<lang c>image alloc_img(unsigned int width, unsigned int height) {
image img; img = malloc(sizeof(image_t)); img->buf = malloc(width * height * sizeof(pixel)); img->width = width; img->height = height; return img;
}
void free_img(image img) {
free(img->buf); free(img);
}
void fill_img(
image img,
color_component r,
color_component g,
color_component b )
{
unsigned int i, n;
n = img->width * img->height;
for (i=0; i < n; ++i)
{
img->buf[i][0] = r;
img->buf[i][1] = g;
img->buf[i][2] = b;
}
}
void put_pixel_unsafe(
image img,
unsigned int x,
unsigned int y,
color_component r,
color_component g,
color_component b )
{
unsigned int ofs; ofs = (y * img->width) + x; img->buf[ofs][0] = r; img->buf[ofs][1] = g; img->buf[ofs][2] = b;
}
void put_pixel_clip(
image img,
unsigned int x,
unsigned int y,
color_component r,
color_component g,
color_component b )
{
if (x < img->width && y < img->height)
put_pixel_unsafe(img, x, y, r, g, b);
}</lang>
C#
This implementation uses a multidemensional array to store the Color structure (which stores the RGB values). No exception catching for out-of-bounds errors if they occur, but provides Height and Width properties so a program using it can avoid them. <lang csharp>public class Bitmap {
public struct Color
{
public byte Red { get; set; }
public byte Blue { get; set; }
public byte Green { get; set; }
}
Color[,] _imagemap;
public int Width { get { return _imagemap.GetLength(0); } }
public int Height { get { return _imagemap.GetLength(1); } }
public Bitmap(int width, int height)
{
_imagemap = new Color[width, height];
}
public void Fill(Color color)
{
for (int y = 0; y < Height; y++)
for (int x = 0; x < Width; x++)
{
_imagemap[x, y] = color;
}
}
public Color GetPixel(int x, int y)
{
return _imagemap[x, y];
}
public void SetPixel(int x, int y, Color color)
{
_imagemap[x, y] = color;
}
}</lang>
C++
<lang cpp>#include <iostream>
- include <boost/gil/gil_all.hpp>
int main() {
using namespace boost::gil; // create 30x40 image rgb8_image_t img(30, 40);
// fill with red rgb8_pixel_t red(255, 0, 0); fill_pixels(view(img), red);
// set pixel at 10x20 to blue rgb8_pixel_t blue(0, 0, 255); view(img)(10, 20) = blue;
// read the value of pixel at 11x20 rgb8_pixel_t px = const_view(img)(11, 20); std::cout << "the pixel at 11, 20 is " << (unsigned)px[0] << ':' << (unsigned)px[1] << ':' << (unsigned)px[2] << '\n';
}</lang>
See also Basic bitmap storage/C++
Clojure
<lang Clojure>(import '[java.awt Color Graphics Image] '[java.awt.image BufferedImage])
(defn blank-bitmap [width height]
(BufferedImage. width height BufferedImage/TYPE_3BYTE_BGR))
(defn fill [image color]
(doto (.getGraphics image) (.setColor color) (.fillRect 0 0 (.getWidth image) (.getHeight image))))
(defn set-pixel [image x y color]
(.setRGB image x y (.getRGB color)))
(defn get-pixel [image x y]
(Color. (.getRGB image x y)))</lang>
Common Lisp
<lang lisp>(defpackage #:rgb-pixel-buffer
(:use #:common-lisp)
(:export #:rgb-pixel-component #:rgb-pixel #:rgb-pixel-buffer
#:+red+ #:+green+ #:+blue+ #:+black+ #:+white+
#:make-rgb-pixel #:make-rgb-pixel-buffer #:rgb-pixel-buffer-width
#:rgb-pixel-buffer-height #:rgb-pixel-red #:rgb-pixel-green
#:rgb-pixel-blue #:fill-rgb-pixel-buffer))</lang>
<lang lisp>(in-package #:rgb-pixel-buffer)
(deftype rgb-pixel-component ()
'(unsigned-byte 8))
(deftype rgb-pixel ()
'(unsigned-byte 24))
(deftype rgb-pixel-buffer (&optional (width '*) (height '*))
`(array rgb-pixel (,width ,height)))
(defconstant +black+ 0) (defconstant +white+ #xFFFFFF) (defconstant +red+ #xFF0000) (defconstant +green+ #x00FF00) (defconstant +blue+ #x0000FF)
(defun make-rgb-pixel (r g b)
(declare (type rgb-pixel-component r g b)) (logior (ash r 16) (ash g 8) b))
(defun rgb-pixel-red (rgb)
(declare (type rgb-pixel rgb)) (logand (ash rgb -16) #xFF))
(defun rgb-pixel-green (rgb)
(declare (type rgb-pixel rgb)) (logand (ash rgb -8) #xFF))
(defun rgb-pixel-blue (rgb)
(declare (type rgb-pixel rgb)) (logand rgb #xFF))
(defun make-rgb-pixel-buffer (width height &optional (initial-element +black+))
(declare (type (integer 1) width height)) (declare (type rgb-pixel initial-element)) (make-array (list width height) :element-type 'rgb-pixel :initial-element initial-element))
(defun rgb-pixel-buffer-width (buffer)
(first (array-dimensions buffer)))
(defun rgb-pixel-buffer-height (buffer)
(second (array-dimensions buffer)))
(defun rgb-pixel (buffer x y)
(declare (type rgb-pixel-buffer buffer)) (declare (type (integer 0) x y)) (aref buffer x y))
(defun (setf rgb-pixel) (value buffer x y)
(declare (type rgb-pixel-buffer buffer)) (declare (type rgb-pixel value)) (declare (type (integer 0) x y)) (setf (aref buffer x y) value))
(defun fill-rgb-pixel-buffer (buffer pixel)
(declare (type rgb-pixel-buffer buffer)) (declare (type rgb-pixel pixel)) (let* ((dimensions (array-dimensions buffer))
(width (first dimensions)) (height (second dimensions)))
(loop
:for y :of-type fixnum :upfrom 0 :below height
:do (loop
:for x :of-type fixnum :upfrom 0 :below width :do (setf (rgb-pixel buffer x y) pixel)))
buffer))</lang>
Example:
<lang lisp>(defvar *buffer* (make-rgb-pixel-buffer 10 10)) (fill-rgb-pixel-buffer *buffer* +white+) (setf (rgb-pixel *buffer* 0 0) +red+) (setf (rgb-pixel *buffer* 0 9) +red+) (setf (rgb-pixel *buffer* 9 0) +red+) (setf (rgb-pixel *buffer* 9 9) +red+)</lang>
Crystal
<lang ruby> class RGBColor
getter red, green, blue
def initialize(@red = 0_u8, @green = 0_u8, @blue = 0_u8) end RED = new(red: 255_u8) GREEN = new(green: 255_u8) BLUE = new(blue: 255_u8) BLACK = new WHITE = new(255_u8, 255_u8, 255_u8)
end
class Pixmap
getter width, height @data : Array(Array(RGBColor))
def initialize(@width : Int32, @height : Int32)
@data = Array.new(@width) { Array.new(@height, RGBColor::WHITE) }
end
def fill(color) @data.each &.fill(color) end
def [](x, y) @data[x][y] end
def []=(x, y, color) @data[x][y] = color end
end
bmap = Pixmap.new(5, 5) pp bmap </lang>
D
This code is a little complex because many Tasks use this module for various purposes. <lang d>module bitmap;
import std.stdio, std.array, std.exception, std.string, std.conv,
std.algorithm, std.ascii;
final class Image(T) {
static if (is(typeof({ auto x = T.black; })))
const static T black = T.black;
else
const static T black = T.init;
static if (is(typeof({ auto x = T.white; })))
const static T white = T.white;
T[] image; private size_t nx_, ny_;
this(in int nxx=0, in int nyy=0, in bool inizialize=true)
pure nothrow {
allocate(nxx, nyy, inizialize);
}
void allocate(in int nxx=0, in int nyy=0, in bool inizialize=true)
pure nothrow @safe in {
assert(nxx >= 0 && nyy >= 0);
} body {
this.nx_ = nxx;
this.ny_ = nyy;
if (nxx * nyy > 0) {
if (inizialize)
image.length = nxx * nyy;
else // Optimization.
image = minimallyInitializedArray!(typeof(image))
(nxx * nyy);
}
}
@property Image dup() const pure nothrow @safe {
auto result = new Image();
result.image = this.image.dup;
result.nx_ = this.nx;
result.ny_ = this.ny;
return result;
}
static Image fromData(T[] data, in size_t nxx=0, in size_t nyy=0)
pure nothrow @safe in {
assert(nxx >= 0 && nyy >= 0 && data.length == nxx * nyy);
} body {
auto result = new Image();
result.image = data;
result.nx_ = nxx;
result.ny_ = nyy;
return result;
}
@property size_t nx() const pure nothrow @safe @nogc { return nx_; }
@property size_t ny() const pure nothrow @safe @nogc { return ny_; }
ref T opIndex(in size_t x, in size_t y) pure nothrow @safe @nogc
in {
assert(x < nx_ && y < ny_);
//assert(x < nx_, format("opIndex, x=%d, nx=%d", x, nx));
//assert(y < ny_, format("opIndex, y=%d, ny=%d", y, ny));
} body {
return image[x + y * nx_];
}
T opIndex(in size_t x, in size_t y) const pure nothrow @safe @nogc
in {
assert(x < nx_ && y < ny_);
//assert(x < nx_, format("opIndex, x=%d, nx=%d", x, nx));
//assert(y < ny_, format("opIndex, y=%d, ny=%d", y, ny));
} body {
return image[x + y * nx_];
}
T opIndexAssign(in T color, in size_t x, in size_t y)
pure nothrow @safe @nogc
in {
assert(x < nx_ && y < ny_);
//assert(x < nx_, format("opIndex, x=%d, nx=%d", x, nx));
//assert(y < ny_, format("opIndex, y=%d, ny=%d", y, ny));
} body {
return image[x + y * nx_] = color;
}
void opIndexUnary(string op)(in size_t x, in size_t y)
pure nothrow @safe @nogc
if (op == "++" || op == "--") in {
assert(x < nx_ && y < ny_);
} body {
mixin("image[x + y * nx_] " ~ op ~ ";");
}
void clear(in T color=this.black) pure nothrow @safe @nogc {
image[] = color;
}
/// Convert a 2D array of chars to a binary Image.
static Image fromText(in string txt,
in char one='#', in char zero='.') pure {
auto M = txt
.strip
.split
.map!(row => row
.filter!(c => c == one || c == zero)
.map!(c => T(c == one))
.array)
.array;
assert(M.join.length > 0); // Not empty.
foreach (row; M)
assert(row.length == M[0].length); // Rectangular
return Image.fromData(M.join, M[0].length, M.length);
}
/// The axis origin is at the top left.
void textualShow(in char bl='#', in char wh='.') const nothrow {
size_t i = 0;
foreach (immutable y; 0 .. ny_) {
foreach (immutable x; 0 .. nx_)
putchar(image[i++] == black ? bl : wh);
putchar('\n');
}
}
}
struct RGB {
ubyte r, g, b; static immutable black = typeof(this)(); static immutable white = typeof(this)(255, 255, 255);
}
Image!RGB loadPPM6(ref Image!RGB img, in string fileName) {
if (img is null)
img = new Image!RGB;
auto f = File(fileName, "rb");
enforce(f.readln.strip == "P6");
string line;
do {
line = f.readln();
} while (line.length && line[0] == '#'); // Skip comments.
const size = line.split;
enforce(size.length == 2);
img.allocate(size[0].to!uint, size[1].to!uint);
enforce(f.readln().strip() == "255");
auto l = new ubyte[img.nx * 3];
size_t i = 0;
foreach (immutable y; 0 .. img.ny) {
f.rawRead!ubyte(l);
foreach (immutable x; 0 .. img.nx)
img.image[i++] = RGB(l[x * 3], l[x * 3 + 1], l[x * 3 + 2]);
}
return img;
}
void savePPM6(in Image!RGB img, in string fileName)
in {
assert(img !is null); assert(img.nx > 0 && img.nx > 0);
} body {
auto f = File(fileName, "wb");
f.writefln("P6\n%d %d\n255", img.nx, img.ny);
size_t i = 0;
foreach (immutable y; 0 .. img.ny)
foreach (immutable x; 0 .. img.nx) {
immutable p = img.image[i++];
f.write(cast(char)p.r, cast(char)p.g, cast(char)p.b);
}
}
version (bitmap_main) {
void main() {
auto img = new Image!RGB(30, 10);
img[4, 5] = RGB.white;
img.textualShow;
}
}</lang>
Compiling it with version=bitmap_main prints:
- Output:
############################## ############################## ############################## ############################## ############################## ####.######################### ############################## ############################## ############################## ##############################
Delphi
<lang Delphi> program BitmapTest;
{$APPTYPE CONSOLE}
type
TColor = record private function GetColor: Cardinal; procedure SetColor(const Value: Cardinal); public Red, Green, Blue, Alpha: Byte; property Color: Cardinal read GetColor write SetColor; end;
TBitmap = class private FPixels: array of array of TColor; FHeight: Integer; FWidth: Integer; function GetPixel(X, Y: integer): TColor; public procedure Fill(aColor: TColor); overload; procedure Fill(aColor: Cardinal); overload; procedure SetSize(w, h: Integer); constructor Create(); overload; constructor Create(w, h: Integer); overload; property Height: Integer read FHeight; property Width: Integer read FWidth; property Pixel[X, Y: integer]: TColor read GetPixel; end;
{ TColor }
function TColor.GetColor: Cardinal; begin
Result := (alpha shl 24) + (red shl 16) + (green shl 8) + blue;
end;
procedure TColor.SetColor(const Value: Cardinal); begin
blue := (Value and $FF); green := ((Value shr 8) and $FF); red := ((Value shr 16) and $FF); alpha := ((Value shr 24) and $FF);
end;
{ TBitmap }
constructor TBitmap.Create; begin
inherited; FHeight := 0; FWidth := 0;
end;
constructor TBitmap.Create(w, h: Integer); begin
Create; SetSize(w, h);
end;
procedure TBitmap.Fill(aColor: Cardinal); var
x, y: Integer;
begin
if (Width > 0) and (Height > 0) then
for x := 0 to width - 1 do
for y := 0 to height - 1 do
FPixels[x, y].Color := aColor;
end;
procedure TBitmap.Fill(aColor: TColor); begin
Fill(aColor.Color);
end;
function TBitmap.GetPixel(X, Y: integer): TColor; begin
Result := FPixels[X, Y];
end;
procedure TBitmap.SetSize(w, h: Integer); var
i: Integer;
begin
if (h = 0) or (w = 0) then begin h := 0; w := 0; end;
FHeight := h;
FWidth := w;
SetLength(FPixels, w);
if w > 0 then
for i := 0 to w - 1 do
SetLength(FPixels[i], h);
end;
var
bmp: TBitmap; x, y: Integer;
begin
bmp := TBitmap.Create(200, 200);
bmp.Fill($00FF0000);
for y := 0 to bmp.Height - 1 do
for x := 0 to bmp.Width - 1 do
begin
if x mod 2 = 1 then
bmp.Pixel[x, y].Color := $0000FF;
end;
bmp.Free;
end.</lang>
E
This example includes the write ppm file code, because it is most naturally written as a method on the image object.
<lang e>def makeFlexList := <elib:tables.makeFlexList> def format := <import:java.lang.makeString>.format
def CHANNELS := 3 def UByte := 0..255
def makeColor {
to fromFloat(r, g, b) {
return makeColor.fromByte((r * 255).round(),
(g * 255).round(),
(b * 255).round())
}
to fromByte(r :UByte, g :UByte, b :UByte) {
def color {
to __printOn(out) {
out.print(format("%02x%02x%02x", [color.rb(), color.gb(), color.bb()]))
}
to rf() { return r / 255 }
to gf() { return g / 255 }
to bf() { return b / 255 }
to rb() { return r }
to gb() { return g }
to bb() { return b }
}
return color
}
}
/** Convert 0..255 into 0..127 -128..-1 */ def sign(v) {
return v %% 256 - 2*(v & 128)
}
def makeImage(width, height) {
# NOTE: The primary E implementation is in Java and Java's fixed-size integers only # come in signed varieties. Therefore, there is a little bit of extra arithmetic. # # In an ideal E implementation we would specify the type 0..255, but this is not # currently possible everywhere, or efficient. def storage := makeFlexList.fromType(<type:java.lang.Byte>, width * height * CHANNELS) storage.setSize(width * height * CHANNELS)
def X := 0..!width def Y := 0..!height
def flexImage {
to __printOn(out) {
for y in Y {
out.print("[")
for x in X {
out.print(flexImage[x, y], " ")
}
out.println("]")
}
}
to width() { return width }
to height() { return height }
to fill(color) {
for x in X {
for y in Y {
flexImage[x, y] := color
}
}
}
to get(x :X, y :Y) {
def base := (y * width + x) * CHANNELS
return makeColor.fromByte(storage[base + 0] %% 256,
storage[base + 1] %% 256,
storage[base + 2] %% 256)
}
/** Provided to make Flood fill slightly less insanely slow. */
to test(x :X, y :Y, c) {
def base := (y * width + x) * CHANNELS
return storage[base + 0] <=> sign(c.rb()) &&
storage[base + 1] <=> sign(c.gb()) &&
storage[base + 2] <=> sign(c.bb())
}
to put(x :X, y :Y, c) {
def base := (y * width + x) * CHANNELS
storage[base + 0] := sign(c.rb())
storage[base + 1] := sign(c.gb())
storage[base + 2] := sign(c.bb())
}
to writePPM(outputStream) {
outputStream.write(`P6$\n$width $height$\n255$\n`.getBytes("US-ASCII"))
outputStream.write(storage.getArray())
}
/** Used for Read ppm file */
to replace(list :List) {
require(list.size() == width * height * CHANNELS)
storage(0) := list
}
}
return flexImage
}</lang>
Examples/tests:
<lang e>? def i := makeImage(3, 3)
- value: [000000 000000 000000 ]
- [000000 000000 000000 ]
- [000000 000000 000000 ]
? i.fill(makeColor.fromFloat(1, 0, 0)) ? i
- value: [ff0000 ff0000 ff0000 ]
- [ff0000 ff0000 ff0000 ]
- [ff0000 ff0000 ff0000 ]
? i[1, 1] := makeColor.fromFloat(0.5, 0.5, 0.5)
- value: 808080
? i
- value: [ff0000 ff0000 ff0000 ]
- [ff0000 808080 ff0000 ]
- [ff0000 ff0000 ff0000 ]
? i[0, 1]
- value: ff0000
? i[1, 1]
- value: 808080
? i.writePPM(<import:java.io.makeFileOutputStream>(<file:~/Desktop/Rosetta.ppm>))</lang>
EchoLisp
<lang scheme> (lib 'plot) (define width 600) (define height 400)
(plot-size width height) ;; set image size
(define (blue x y) (rgb 0.0 0.0 1.0)) ;; a constant function (plot-rgb blue 1 1) ;; blue everywhere
(lib 'types) ;; uint32 and uint8 vector types
- bit-map pixel access
(define bitmap (pixels->uint32-vector)) ;; screen to vector of int32
→ 240000
(define (pix-at x y) (vector-ref bitmap (+ x (* y width)))) (rgb->list (pix-at 100 200)) → (0 0 255 255) ;; rgb blue
- writing to bitmap
(define (set-color-xy x y col) (vector-set! bitmap (+ x (* y width)) col))
(for* ((x 100)(y 200)) (set-color-xy x y (rgb 1 1 0))) ;; to bitmap (vector->pixels bitmap) ;; bitmap to screen
- bit-map color components (r g b a) = index (0 1 2 3) access
(define bitmap (pixels->uint8-clamped-vector)) ;; screen to vector of uint8 (vector-length bitmap)
→ 960000
(define (blue-at-xy x y) (vector-ref bitmap (+ x 3 (* y width)))) ;; 3 = blue component (blue-at-xy 100 200)
→ 255
</lang>
Elixir
Translation of the erlang version of the code.
<lang elixir> defmodule RosBitmap do
defrecord Bitmap, pixels: nil, shape: {0, 0}
defp new(width, height, {:rgb, r, g, b}) do
Bitmap[
pixels: :array.new(width * height,
{:default, <<r::size(8), g::size(8), b::size(8)>>}),
shape: {width, height}]
end
def new(width, height), do: new(width, height, {:rgb, 0, 0, 0})
def fill(Bitmap[shape: {width, height}], {:rgb, _r, _g, _b}=color) do
new(width, height, color)
end
def set_pixel(Bitmap[pixels: pixels, shape: {width, _height}]=bitmap,
{:at, x, y}, {:rgb, r, g, b}) do
index = x + y * width
bitmap.pixels(:array.set(index, <<r::size(8), g::size(8), b::size(8)>>, pixels))
end
def get_pixel(Bitmap[pixels: pixels, shape: {width, _height}], {:at, x, y}) do
index = x + y * width
<<r::size(8), g::size(8), b::size(8)>> = :array.get(index, pixels)
{:rgb, r, g, b}
end
end </lang>
Erlang
Stores pixels as a 1d array and colors as binaries.
<lang erlang> -module(ros_bitmap).
-export([new/2, fill/2, set_pixel/3, get_pixel/2]).
-record(bitmap, {
pixels = nil,
shape = {0, 0}
}).
new(Width, Height) ->
#bitmap{pixels=array:new(Width * Height, {default, <<0:8, 0:8, 0:8>>}), shape={Width, Height}}.
fill(#bitmap{shape={Width, Height}}, {rgb, R, G, B}) ->
#bitmap{
pixels=array:new(Width * Height, {default, <<R:8, G:8, B:8>>}),
shape={Width, Height}}.
set_pixel(#bitmap{pixels=Pixels, shape={Width, _Height}}=Bitmap, {at, X, Y}, {rgb, R, G, B}) ->
Index = X + Y * Width,
Bitmap#bitmap{pixels=array:set(Index, <<R:8, G:8, B:8>>, Pixels)}.
get_pixel(#bitmap{pixels=Pixels, shape={Width, _Height}}, {at, X, Y}) ->
Index = X + Y * Width,
<<R:8, G:8, B:8>> = array:get(Index, Pixels),
{rgb, R, G, B}.
</lang>
Euphoria
<lang euphoria>-- Some color constants: constant
black = #000000, white = #FFFFFF, red = #FF0000, green = #00FF00, blue = #0000FF
-- Create new image filled with some color function new_image(integer width, integer height, atom fill_color)
return repeat(repeat(fill_color,height),width)
end function
-- Usage example: sequence image image = new_image(800,600,black)
-- Set pixel color: image[400][300] = red
-- Get pixel color atom color color = image[400][300] -- Now color is #FF0000</lang> ?color -- Should print out 16711680
F#
FSharp can accomplish this task in several ways. This version is purely functional. The bitmap data structure does not mutate. Set pixel, for example, simply transforms the input bitmap into a new bitmap with that pixel set to the input color. If you have Framework 4.5, you can use ImmutableArray to force this immutability.
Solution: <lang fsharp> //pure functional version ... changing a pixel color provides a new Bitmap type Color = {red: byte; green: byte; blue: byte} type Point = {x:uint32; y:uint32} type Bitmap = {color: Color array; maxX: uint32; maxY: uint32}
let colorBlack = {red = (byte) 0; green = (byte) 0; blue = (byte) 0} let emptyBitmap = {color = Array.empty; maxX = (uint32) 0; maxY = (uint32) 0} let bitmap (width: uint32) (height: uint32) =
match width, height with
| 0u,0u | 0u,_ | _, 0u -> emptyBitmap
| _,_ -> {color = Array.create ((int) (width * height)) colorBlack;
maxX = width;
maxY = height}
let getPixel point bitmap =
match bitmap.color with | c when c |> Array.isEmpty -> None | c when (uint32) c.Length <= (point.y * bitmap.maxY + point.x) -> None | c -> Some c.[(int) (point.y * bitmap.maxY + point.x)]
let setPixel point color bitmap =
{bitmap with color = bitmap.color |> Array.mapi (function
| i when i = (int) (point.y * bitmap.maxY + point.x) ->
(fun _ -> color)
| _ -> id)}
let fill color bitmap = {bitmap with color = bitmap.color |> Array.map (fun _ ->color)} </lang>
Tests: <lang fsharp> //setups //==check pixel for color function let check bitmap color (x,y) =
match (getPixel {x=x;y=y} bitmap) with
| Some(v) -> v = color
| _ -> false
let allPixels i j = [for x in [0u..(i-1u)] do for y in [0u..(j-1u)] -> (x,y)]
//create new empty bitmap let myBitmap = bitmap 0u 0u printfn "Is empty: %b" (myBitmap = emptyBitmap) let myBitmap2 = bitmap 1u 0u printfn "Is empty: %b" (myBitmap2 = emptyBitmap) let myBitmap3 = bitmap 0u 1u printfn "Is empty: %b" (myBitmap3 = emptyBitmap) //create normal bitmap let myBitmap4 = bitmap 14u 14u printfn "Is not empty: %b" (not (myBitmap4 = emptyBitmap)) //just check one color printfn "Is 1,1 black: %b" (check myBitmap4 colorBlack (1u,1u)) //check out of range color printfn "Is 100,100 nothing: %b" (not(check myBitmap4 colorBlack (100u,100u))) //make sure all pixels are black printfn "Is all black: %b" ((allPixels 14u 14u) |> List.forall (check myBitmap4 colorBlack)) //fill bitmap color let colorWhite = {red = (byte) 255; green = (byte) 255; blue = (byte) 255} let myBitmap5 = myBitmap4 |> fill colorWhite printfn "Is all white: %b" ((allPixels 14u 14u) |> List.forall (check myBitmap5 colorWhite)) //change just one pixel let myBitmap6 = myBitmap5 |> setPixel {x=5u;y=10u} colorBlack printfn "Is 5,10 black: %b" (check myBitmap4 colorBlack (5u,10u)) </lang>
- Output:
Is empty
Is empty: true
Is empty: true
Is not empty: true
Is 1,1 black: true
Is 100,100 nothing: true
Is all black: true
Is all white: true
Is 5,10 black: true Usage: <lang fsharp> bitmap 14u 14u |> fill {red = (byte) 200; green = (byte) 0; blue = (byte) 10} |> setPixel {x=5u;y=10u} {red = (byte) 0; green = (byte) 0; blue = (byte) 0} |> getPixel {x=5u;y=10u} |> printfn "%A" </lang>
- Output:
Some {red = 0uy;
green = 0uy;
blue = 0uy;}
Factor
The image is a matrix of triples {R,G,B}. The various utilities could be defined in another file, most of them are not used right now, but we need them for drawing so I put every thing here.. <lang factor>USING: arrays fry kernel math.matrices sequences ; IN: rosettacode.raster.storage
! Various utilities
- meach ( matrix quot -- ) [ each ] curry each ; inline
- meach-index ( matrix quot -- )
[ swap 2array ] prepose [ curry each-index ] curry each-index ; inline
- mmap ( matrix quot -- matrix' ) [ map ] curry map ; inline
- mmap! ( matrix quot -- matrix' ) [ map! ] curry map! ; inline
- mmap-index ( matrix quot -- matrix' )
[ swap 2array ] prepose [ curry map-index ] curry map-index ; inline
- matrix-dim ( matrix -- i j ) [ length ] [ first length ] bi ;
- set-Mi,j ( elt {i,j} matrix -- ) [ first2 swap ] dip nth set-nth ;
- Mi,j ( {i,j} matrix -- elt ) [ first2 swap ] dip nth nth ;
! The storage functions
- <raster-image> ( width height -- image )
zero-matrix [ drop { 0 0 0 } ] mmap ;
- fill-image ( {R,G,B} image -- image )
swap '[ drop _ ] mmap! ;
- set-pixel ( {R,G,B} {i,j} image -- ) set-Mi,j ; inline
- get-pixel ( {i,j} image -- pixel ) Mi,j ; inline
</lang>
FBSL
Volatility in FBSL is a feature uncommon to most other languages. It is the ability of its intrinsic functions as well as its user-defined functions, DynAsm and DynC blocks, and functions imported from 3rd-party DLL's to preserve their return values between function calls in FBSL Variants that have the same names as their respective functions but use neither the parentheses nor the arguments. These Variants belong to the global namespace and can be used throughout the entire script until another fully qualified function call to their respective functions is made, whereby they change their values accordingly. The feature minimizes the need for temporary variables and assignments.
This feature is a logical extension of VisualBasic way to formalize its function return value by assigning it to a Variant of the same name as that of the respective function. However, the VB Variant is only effective within the scope of its own function.
Using pure FBSL's built-in graphics functions: <lang qbasic>#DEFINE WM_LBUTTONDOWN 513
- DEFINE WM_RBUTTONDOWN 516
- DEFINE WM_CLOSE 16
FBSLSETFORMCOLOR(ME, RGB(0, 255, 255)) ' Cyan: set persistent background color DRAWWIDTH(5) ' Adjust point size FBSL.GETDC(ME) ' Use volatile FBSL.GETDC below to avoid extra assignments
RESIZE(ME, 0, 0, 300, 200) CENTER(ME) SHOW(ME)
BEGIN EVENTS SELECT CASE CBMSG CASE WM_LBUTTONDOWN ' Set color at current coords as hex literal PSET(FBSL.GETDC, LOWORD(CBLPARAM), HIWORD(CBLPARAM), &H0000FF) ' Red: Windows stores colors in BGR order CASE WM_RBUTTONDOWN ' Get color at current coords as hex literal FBSLSETTEXT(ME, "&H" & HEX(POINT(FBSL.GETDC, LOWORD(CBLPARAM), HIWORD(CBLPARAM)))) CASE WM_CLOSE ' Clean up FBSL.RELEASEDC(ME, FBSL.GETDC) END SELECT END EVENTS</lang>
- Output:
Forth
This creates bitmaps on the heap (they may be deallocated with "FREE"). 32-bit or greater cells are assumed, one pixel per cell. This automatically word-aligns rows, so a separate stride field is not required. <lang forth>hex 0000ff constant red 00ff00 constant green ff0000 constant blue decimal
1 cells constant pixel
- pixels cells ;
- bdim ( bmp -- w h ) 2@ ;
- bheight ( bmp -- h ) @ ;
- bwidth ( bmp -- w ) bdim drop ;
- bdata ( bmp -- addr ) 2 cells + ;
- bitmap ( w h -- bmp )
2dup * pixels bdata allocate throw dup >r 2! r> ;
- bfill ( pixel bmp -- )
dup bdata swap bdim * pixels bounds do dup i ! pixel +loop drop ;
- bxy ( x y bmp -- addr )
dup >r bwidth * + pixels r> bdata + ;
- b@ ( x y bmp -- pixel ) bxy @ ;
- b! ( pixel x y bmp -- ) bxy ! ;
- bshow ( bmp -- )
hex
dup bdim
0 do cr
dup 0 do
over i j rot b@ if [char] * else bl then emit \ 7 u.r
loop
loop
2drop decimal ;
4 3 bitmap value test red test bfill test bshow cr</lang>
Fortran
See Basic bitmap storage/Fortran
Go
Standard library
Go's standard library include image, color, and drawing packages and the source for them is easy to read.
There is also a Go Blog article on the image package
The image.NRGBA type supports everything this task requires (the 'A' is for alpha channel, each are 8 bits, if 16 bits each of RGBA is desired there is also the image.NRGBA64 type). The 'N' of NRGBA stands for Non-alpha-premultiplied, color values can trivially be converted to/from alpha-premultiplied RGBA values via a color.Model.
Here's how to use the standard packages to do what this task requires: <lang go>package main
import ( "bytes" "fmt" "image" "image/color" "image/draw" "image/png" )
func main() { // A rectangle from 0,0 to 300,240. r := image.Rect(0, 0, 300, 240)
// Create an image im := image.NewNRGBA(r)
// set some color variables for convience var ( red = color.RGBA{0xff, 0x00, 0x00, 0xff} blue = color.RGBA{0x00, 0x00, 0xff, 0xff} )
// Fill with a uniform color draw.Draw(im, r, &image.Uniform{red}, image.ZP, draw.Src)
// Set individual pixels im.Set(10, 20, blue) im.Set(20, 30, color.Black) im.Set(30, 40, color.RGBA{0x10, 0x20, 0x30, 0xff})
// Get the values of specific pixels as color.Color types. // The color will be in the color.Model of the image (in this // case color.NRGBA) but color models can convert their values // to other models. c1 := im.At(0, 0) c2 := im.At(10, 20)
// or directly as RGB components (scaled values) redc, greenc, bluec, _ := c1.RGBA() redc, greenc, bluec, _ = im.At(30, 40).RGBA()
// Images can be read and writen in various formats var buf bytes.Buffer err := png.Encode(&buf, im) if err != nil { fmt.Println(err) }
fmt.Println("Image size:", im.Bounds().Dx(), "×", im.Bounds().Dy()) fmt.Println(buf.Len(), "bytes when encoded as PNG.") fmt.Printf("Pixel at %7v is %v\n", image.Pt(0, 0), c1) fmt.Printf("Pixel at %7v is %#v\n", image.Pt(10, 20), c2) // %#v shows type details fmt.Printf("Pixel at %7v has R=%d, G=%d, B=%d\n", image.Pt(30, 40), redc, greenc, bluec) }</lang>
- Output:
Image size: 300 × 240
786 bytes when encoded as PNG.
Pixel at (0,0) is {255 0 0 255}
Pixel at (10,20) is color.NRGBA{R:0x0, G:0x0, B:0xff, A:0xff}
Pixel at (30,40) has R=4112, G=8224, B=12336
DIY
Not a complete working program. Presented here are just types and functions requested by the task. <lang go>// Raster package used with a number of RC tasks. // // For each task, documentation in package main source will list this // file and others that are necessary to build a raster package with // sufficient functionality for the task. To build a working program, // build a raster package from the files listed, install the package, // and then compile and link the package main that completes the task. // // Alternatively, files in the raster package can be combined as desired // to build a package that meets the needs of multiple tasks. package raster
// Rgb is a 24 bit color value represented with a 32 bit int // in the conventional way. This is expected to be convenient // for the programmer in many cases. type Rgb int32
// Pixel has r, g, and b as separate fields. This is used as // the in-memory representation of a bitmap. type Pixel struct {
R, G, B byte
}
// Pixel returns a new Pixel from a Rgb value func (c Rgb) Pixel() Pixel {
return Pixel{R: byte(c >> 16), G: byte(c >> 8), B: byte(c)}
}
// Rgb returns a single Rgb value computed from rgb fields of a Pixel // of a Pixel. func (p Pixel) Rgb() Rgb {
return Rgb(p.R)<<16 | Rgb(p.G)<<8 | Rgb(p.B)
}
// Bitmap is the in-memory representation, or image storage type of a bitmap. // Zero value for type is a valid zero-size bitmap. // The only exported field is Comments. Remaining fields have interdepencies // that are managed by package code and so should not be directly accessed // from outside the package. type Bitmap struct {
Comments []string rows, cols int px []Pixel // all pixels as a single slice, row major order pxRow [][]Pixel // rows of pixels as slices of px
}
const creator = "# Creator: Rosetta Code http://rosettacode.org/"
// New is a Bitmap "constructor." Parameters x and y are extents. // That is, the new bitmap will have x columns and y rows. func NewBitmap(x, y int) (b *Bitmap) {
b = &Bitmap{
Comments: []string{creator},
rows: y, // named fields here to prevent possible mix-ups.
cols: x,
px: make([]Pixel, x*y),
pxRow: make([][]Pixel, y),
}
// Note rows of pixels are not allocated separately.
// Rather the whole bitmap is allocted in one chunk as px.
// This simplifies allocation and maintains locality.
x0, x1 := 0, x
for i := range b.pxRow {
b.pxRow[i] = b.px[x0:x1] // slice operation. does no allocation.
x0, x1 = x1, x1+x
}
return b
}
// Extent returns bitmap dimensions. func (b *Bitmap) Extent() (cols, rows int) {
return b.cols, b.rows
}
// Fill entire bitmap with solid color. func (b *Bitmap) Fill(p Pixel) {
for i := range b.px {
b.px[i] = p
}
}
func (b *Bitmap) FillRgb(c Rgb) {
b.Fill(c.Pixel())
}
// Set a single pixel color value. // Clips to bitmap boundaries. // Returns true if pixel was set, false if clipped. func (b *Bitmap) SetPx(x, y int, p Pixel) bool {
defer func() { recover() }()
b.pxRow[y][x] = p
return true
}
func (b *Bitmap) SetPxRgb(x, y int, c Rgb) bool {
return b.SetPx(x, y, c.Pixel())
}
// Note: Clipping to bitmap boundaries is needed for program correctness // but is otherwise not required by the task. It is implemented with the // combination of pxRow and the deferred recover. SetPx, GetPx return the // clipping result as a way for higher level graphics functions to track // plotting and clipping status. As this is not required by tasks though, // it is generally not implemented.
// Get a single pixel color value. // Returns pixel and ok=true if coordinates are within bitmap boundaries. // Returns ok=false if coordinates are outside bitmap boundaries. func (b *Bitmap) GetPx(x, y int) (p Pixel, ok bool) {
defer func() { recover() }()
return b.pxRow[y][x], true
}
func (b *Bitmap) GetPxRgb(x, y int) (Rgb, bool) {
p, ok := b.GetPx(x, y)
if !ok {
return 0, false
}
return p.Rgb(), true
}</lang>
Haskell
We implement the Image type as an STArray so that we can use it in an imperative fashion in the ST monad.
<lang haskell>module Bitmap(module Bitmap) where
import Control.Monad import Control.Monad.ST import Data.Array.ST
newtype Pixel = Pixel (Int, Int) deriving Eq
instance Ord Pixel where
compare (Pixel (x1, y1)) (Pixel (x2, y2)) =
case compare y1 y2 of
EQ -> compare x1 x2
v -> v
instance Ix Pixel where {- This instance differs from the one for (Int, Int) in that the ordering of indices is
(0,0), (1,0), (2,0), (0,1), (1,1), (2,1)
instead of
(0,0), (0,1), (1,0), (1,1), (2,0), (2,1). -}
range (Pixel (xa, ya), Pixel (xz, yz)) =
[Pixel (x, y) | y <- [ya .. yz], x <- [xa .. xz]]
index (Pixel (xa, ya), Pixel (xz, _)) (Pixel (xi, yi)) =
(yi - ya)*(xz - xa + 1) + (xi - xa)
inRange (Pixel (xa, ya), Pixel (xz, yz)) (Pixel (xi, yi)) =
not $ xi < xa || xi > xz || yi < ya || yi > yz
rangeSize (Pixel (xa, ya), Pixel (xz, yz)) =
(xz - xa + 1) * (yz - ya + 1)
instance Show Pixel where
show (Pixel p) = show p
class Ord c => Color c where
luminance :: c -> Int
-- The Int should be in the range [0 .. 255].
black, white :: c
toNetpbm :: [c] -> String
fromNetpbm :: [Int] -> [c]
netpbmMagicNumber, netpbmMaxval :: c -> String
{- The argument to these two functions is ignored; the
parameter is only for typechecking. -}
newtype Color c => Image s c = Image (STArray s Pixel c)
image :: Color c => Int -> Int -> c -> ST s (Image s c) {- Creates a new image with the given width and height, filled with the given color. -} image w h = liftM Image .
newArray (Pixel (0, 0), Pixel (w - 1, h - 1))
listImage :: Color c => Int -> Int -> [c] -> ST s (Image s c) {- Creates a new image with the given width and height, with each pixel set to the corresponding element of the given list. -} listImage w h = liftM Image .
newListArray (Pixel (0, 0), Pixel (w - 1, h - 1))
dimensions :: Color c => Image s c -> ST s (Int, Int) dimensions (Image i) = do
(_, Pixel (x, y)) <- getBounds i return (x + 1, y + 1)
getPix :: Color c => Image s c -> Pixel -> ST s c getPix (Image i) = readArray i
getPixels :: Color c => Image s c -> ST s [c] getPixels (Image i) = getElems i
setPix :: Color c => Image s c -> Pixel -> c -> ST s () setPix (Image i) = writeArray i
fill :: Color c => Image s c -> c -> ST s () fill (Image i) c = getBounds i >>= mapM_ f . range
where f p = writeArray i p c
mapImage :: (Color c, Color c') =>
(c -> c') -> Image s c -> ST s (Image s c')
mapImage f (Image i) = liftM Image $ mapArray f i</lang>
This module provides an instance of Color. <lang haskell>module Bitmap.RGB(module Bitmap.RGB) where
import Bitmap import Control.Monad.ST
newtype RGB = RGB (Int, Int, Int) deriving (Eq, Ord)
instance Color RGB where
luminance (RGB (r, g, b)) = round x
where x = 0.2126*r' + 0.7152*g' + 0.0722*b'
(r', g', b') = (toEnum r, toEnum g, toEnum b)
black = RGB (0, 0, 0)
white = RGB (255, 255, 255)
toNetpbm = concatMap f
where f (RGB (r, g, b)) = [toEnum r, toEnum g, toEnum b]
fromNetpbm [] = []
fromNetpbm (r : g : b : rest) = RGB (r, g, b) : fromNetpbm rest
netpbmMagicNumber _ = "P6"
netpbmMaxval _ = "255"
toRGBImage :: Color c => Image s c -> ST s (Image s RGB) toRGBImage = mapImage $ f . luminance
where f x = RGB (x, x, x)</lang>
Icon and Unicon
The language has a built-in window data type with associated graphics primitives. A bitmap is just a window that isn't visible on-screen at the moment. <lang Icon>procedure makebitmap(width,height)
return open("bitmap", "g", "canvas=hidden",
"size="||width||","||height)
end procedure fillimage(w,color)
Fg(w,color) FillRectangle(w)
end procedure setpixel(w,x,y,color)
Fg(w,color) DrawPixel(x,y)
end procedure getpixel(w,x,y)
return Pixel(w,x,y)
end</lang>
J
A number of addon packages are available for J that work with common image formats (including PPM), but here we will show a basic bitmap storage type as per the task description.
The structure is a 3-dimensional array of numbers. The shape of the array is height by width by 3. Each 1-dimensional cell of size 3 contains R, G and B numbers, in that order. Indexing is zero based. (We could instead have encoded RGB in a single integer...)
No parameter validity checks are currently implemented.
In J, allocating an uninitialized image would not normally be separated from creating
the colored image, so makeRGB allows the specification of color during allocation. As a monad, makeRGB creates a black image with the specified height and width. It can also take a left argument (dyadic form) specifying the color(s) of the image. fillRGB requires a left argument specifying the color(s), but takes a bitmap (RGB) structure as the right argument.
Solution: <lang j>makeRGB=: 0&$: : (($,)~ ,&3) fillRGB=: makeRGB }:@$ setPixels=: (1&{::@[)`(<"1@(0&{::@[))`]} getPixels=: <"1@[ { ]</lang>
Examples:
<lang j> myimg=: makeRGB 5 8 NB. create a bitmap with height 5 and width 8 (black)
myimg=: 255 makeRGB 5 8 NB. create a white bitmap with height 5 and width 8
myimg=: 127 makeRGB 5 8 NB. create a gray bitmap with height 5 and width 8
myimg=: 0 255 0 makeRGB 5 8 NB. create a green bitmap with height 5 and width 8
myimg=: 0 0 255 fillRGB myimg NB. fill myimg with blue
colors=: 0 255 {~ #: i.8 NB. black,blue,green,cyan,red,magenta,yellow,white
myimg=: colors fillRGB myimg NB. fill myimg with vertical stripes of colors
2 4 getPixels myimg NB. get the pixel color from point (2, 4)
255 0 0
myimg=: (2 4 ; 255 255 255) setPixels myimg NB. set pixel at point (2, 4) to white 2 4 getPixels myimg NB. get the pixel color from point (2, 4)
255 255 255
}:$ myimg NB. get height and width of the image
5 8</lang>
getPixels and setPixels are generalized to set and get lists/arrays of pixels.
<lang j>pixellist=: ,"0/~ i. 10 NB. row and column indices for 10 by 10 block of pixels
NB. create 10 by 10 block of magenta pixels in the middle of a 300 by 300 green image myimg=: ((145 + pixellist) ; 255 0 255) setPixels 0 255 0 makeRGB 300 300
NB. get pixel color for 10x10 block offset from magenta block subimg=: (140 + pixellist) getPixels myimg</lang>
To display the image in a window at any point for verification:
<lang j>require 'viewmat' viewRGB=: [: viewrgb 256&#.
viewRGB myimg</lang>
Note that height comes before width here. This is inconsistent with marketing of display resolutions, but matches J's treatment of dimensions.
Java
Solution
<lang java>import java.awt.Color; import java.awt.Graphics; import java.awt.Image; import java.awt.image.BufferedImage;
public class BasicBitmapStorage {
private final BufferedImage image;
public BasicBitmapStorage(int width, int height) {
image = new BufferedImage(width, height, BufferedImage.TYPE_INT_RGB);
}
public void fill(Color c) {
Graphics g = image.getGraphics();
g.setColor(c);
g.fillRect(0, 0, image.getWidth(), image.getHeight());
}
public void setPixel(int x, int y, Color c) {
image.setRGB(x, y, c.getRGB());
}
public Color getPixel(int x, int y) {
return new Color(image.getRGB(x, y));
}
public Image getImage() {
return image;
}
}</lang>
Test Program
<lang java>import static org.junit.Assert.assertEquals;
import java.awt.Color; import org.junit.Test;
public class BasicBitmapStorageTest {
@Test
public void testHappy() {
int width = 640;
int height = 480;
BasicBitmapStorage bbs = new BasicBitmapStorage(width, height);
bbs.fill(Color.CYAN);
bbs.setPixel(width / 2, height / 2, Color.BLACK);
Color c1 = bbs.getPixel(width / 2, height / 2);
Color c2 = bbs.getPixel(20, 20);
assertEquals(Color.BLACK, c1);
assertEquals(Color.CYAN, c2);
}
}</lang>
JavaScript
JavaScript can interact with a drawing context using the HTML5 Canvas API.
<lang javascript> // Set up the canvas var canvas = document.createElement("canvas"),
ctx = canvas.getContext("2d"),
width = 400, height = 400;
ctx.canvas.width = width; ctx.canvas.height = height;
// Optionaly add it to the current page document.body.appendChild(canvas);
// Draw an image var img = document.createElement("img"); img.onload = function(){
// Draw the element into the top-left of the canvas ctx.drawImage(img, 0, 0);
}; img.src = "//placehold.it/400x400";
// Fill the canvas with a solid blue color ctx.fillStyle = "blue"; ctx.fillRect(0, 0, width, height);
// Place a black pixel in the middle // Note that a pixel is a 1 by 1 rectangle // This is the fastest method as of 2012 benchmarks ctx.fillStyle = "black"; ctx.fillRect(width / 2, height / 2, 1, 1); </lang>
Julia
Using packages (Images.jl, Colors.jl): <lang julia>using Images, Colors
Base.hex(p::RGB{T}) where T = join(hex(c(p), 2) for c in (red, green, blue)) function showhex(m::Matrix{RGB{T}}, pad::Integer=4) where T
for r in 1:size(m, 1)
println(" " ^ pad, join(hex.(m[r, :]), " "))
end
end
w, h = 5, 7 cback = RGB(1, 0, 1) cfore = RGB(0, 1, 0)
img = Array{RGB{N0f8}}(h, w); println("Uninitialized image:") showhex(img)
fill!(img, cback) println("\nImage filled with background color:") showhex(img)
img[2, 3] = cfore println("\nImage with a pixel set for foreground color:") showhex(img)</lang>
- Output:
Uninitialized image:
10DFF8 7F0000 F84A00 0030DA 007F00
4A007F B0DDF8 7F0000 F84A00 00D0DB
0000F0 4A007F D0D9F8 7F0000 F84A00
DFF84A 000050 4A007F 50DAF8 7F0000
007F00 D9F84A 000010 4A007F 30DCF8
0050E0 007F00 DAF84A 000050 4A007F
F84A00 00B0D9 007F00 DBF84A 000050
Image filled with background color:
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
Image with a pixel set for foreground color:
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF 00FF00 FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
FF00FF FF00FF FF00FF FF00FF FF00FF
KonsolScript
<lang KonsolScript>function main() {
Var:Number shape;
Image:New(50, 50, shape)
Draw:RectFill(0, 0, 50, 50, 0xFF0000, shape) //one to fill an image with a plain RED color
Draw:Pixel(30, 30, 0x0000FF, shape) //set a given pixel at (30,30) with a BLUE color
while (B1 == false) {
Image:Blit(10, 10, shape, screen)
Screen:Render()
}
}</lang>
Kotlin
<lang scala>// version 1.1.4-3
import java.awt.Color import java.awt.Graphics import java.awt.image.BufferedImage
class BasicBitmapStorage(width: Int, height: Int) {
val image = BufferedImage(width, height, BufferedImage.TYPE_3BYTE_BGR)
fun fill(c: Color) {
val g = image.graphics
g.color = c
g.fillRect(0, 0, image.width, image.height)
}
fun setPixel(x: Int, y: Int, c: Color) = image.setRGB(x, y, c.getRGB())
fun getPixel(x: Int, y: Int) = Color(image.getRGB(x, y))
}
fun main(args: Array<String>) {
val width = 640
val height = 480
val bbs = BasicBitmapStorage(width, height)
with (bbs) {
fill(Color.cyan)
setPixel(width / 2, height / 2, Color.black)
val c1 = getPixel(width / 2, height / 2)
val c2 = getPixel(20, 20)
print("The color of the pixel at (${width / 2}, ${height / 2}) is ")
println(if (c1 == Color.black) "black" else "unknown")
print("The color of the pixel at (120, 120) is ")
println(if (c2 == Color.cyan) "cyan" else "unknown")
}
}</lang>
- Output:
The color of the pixel at (320, 240) is black The color of the pixel at (120, 120) is cyan
Lingo
<lang lingo>-- Creates a new image object of size 640x480 pixel and 32-bit color depth img = image(640, 480, 32)
-- Fills image with plain red img.fill(img.rect, rgb(255,0,0))
-- Gets the color value of the pixel at point (320, 240) col = img.getPixel(320, 240)
-- Changes the color of the pixel at point (320, 240) to black img.setPixel(320, 240, rgb(0,0,0))</lang>
LiveCode
LiveCode has built in support for importing and exporting PBM, JPEG, GIF, BMP or PNG graphics formats
<lang LiveCode>
-- create an image container box at the center of the current stack window with default properties create image "test" -- programtically choose the paint bucket tool choose bucket tool -- LiveCode engine has built-in color keywords: set the brushColor to "dark green" -- programtically mouse click at the center of image container box to fill click at the loc of image "test" -- get the RGBA values of the first pixel in the image box put byteToNum(byte 1 of the imageData of image "test") into tRed put byteToNum(byte 2 of the imageData of image "test") into tGreen put byteToNum(byte 3 of the imageData of image "test") into tBlue put byteToNum(byte 4 of the imageData of image "test") into tAlpha -- log message the info in the message box put "First Pixel Color is Red:"& tRed &" Green:"& tGreen &" Blue:"& tBlue &" Transparency:"& tAlpha -- just for fun replace the contents with RosettaCode logo wait 2 seconds set the filename of image "test" to "http://rosettacode.org/mw/title.png" -- the next line is copy of the Write a PPM task: export image "test" to file "~/Test.PPM" as paint -- paint format is one of PBM, PGM, or PPM
</lang>
Lua
Original
<lang lua>function Allocate_Bitmap( width, height )
local bitmap = {}
for i = 1, width do
bitmap[i] = {}
for j = 1, height do
bitmap[i][j] = {}
end
end
return bitmap
end
function Fill_Bitmap( bitmap, color )
for i = 1, #bitmap do
for j = 1, #bitmap[1] do
bitmap[i][j] = color
end
end
end
function Get_Pixel( bitmap, x, y )
return bitmap[x][y]
end</lang> This can be used like: <lang lua>bitmap = Allocate_Bitmap( 100, 50 ) Fill_Bitmap( bitmap, { 15, 200, 80 } ) pixel = Get_Pixel( bitmap, 20, 25 ) print( pixel[1], pixel[2], pixel[3] )</lang>
Alternate
A more object-oriented and extensible approach for easier re-use elsewhere. <lang lua>local Bitmap = {
new = function(self, width, height)
local instance = setmetatable({ width=width, height=height }, self)
instance:alloc()
return instance
end,
alloc = function(self)
self.pixels = {}
for y = 1, self.height do
self.pixels[y] = {}
for x = 1, self.width do
self.pixels[y][x] = 0x00000000
end
end
end,
clear = function(self, c)
for y = 1, self.height do
for x = 1, self.width do
self.pixels[y][x] = c or 0x00000000
end
end
end,
get = function(self, x, y)
x, y = math.floor(x+1), math.floor(y+1) -- given 0-based indices, use 1-based indices
if ((x>=1) and (x<=self.width) and (y>=1) and (y<=self.height)) then
return self.pixels[y][x]
else
return nil
end
end,
set = function(self, x, y, c)
x, y = math.floor(x+1), math.floor(y+1) -- given 0-based indices, use 1-based indices
if ((x>=1) and (x<=self.width) and (y>=1) and (y<=self.height)) then
self.pixels[y][x] = c or 0x00000000
end
end,
} Bitmap.__index = Bitmap setmetatable(Bitmap, { __call = function (t, ...) return t:new(...) end })</lang> Usage: <lang lua>local bitmap = Bitmap(32,32)
-- default pixel representation is 32-bit packed ARGB on [0,255] bitmap:clear(0xFFFF0000) -- fill with red bitmap:set(1, 1, 0xFF00FF00) -- one green pixel bitmap:set(2, 2, 0xFF0000FF) -- one blue pixel print(string.format("pixel at 0,0 = %x", bitmap:get(0,0))) print(string.format("pixel at 1,1 = %x", bitmap:get(1,1))) print(string.format("pixel at 2,2 = %x", bitmap:get(2,2)))
-- but note that pixel representation is agnostic.. -- (it's just a wrapper around a 2d-array of any valid type)
-- want to switch to RGB-tuple on [0,1]?? bitmap:clear({1,0,0}) -- fill with red bitmap:set(1, 1, {0,1,0}) -- one green pixel bitmap:set(2, 2, {0,0,1}) -- one blue pixel print(string.format("pixel at 0,0 = %s", table.concat(bitmap:get(0,0),", "))) print(string.format("pixel at 1,1 = %s", table.concat(bitmap:get(1,1),", "))) print(string.format("pixel at 2,2 = %s", table.concat(bitmap:get(2,2),", ")))
-- or strings?? bitmap:clear("red") -- fill with red bitmap:set(1, 1, "green") -- one green pixel bitmap:set(2, 2, "blue") -- one blue pixel print(string.format("pixel at 0,0 = %s", bitmap:get(0,0))) print(string.format("pixel at 1,1 = %s", bitmap:get(1,1))) print(string.format("pixel at 2,2 = %s", bitmap:get(2,2)))</lang> Caveat: Just be aware that as currently written the storage of complex types are referenced rather than copied. So, for example, using the clear() method with a table representing an RGB-tuple, will store the same identical reference throughout the bitmap - so direct modification of any one pixel's internal components would affect all other pixels as well. You should override the clear() method as appropriate to better support your desired pixel representation if this is not the behavior you desire.
- Output:
pixel at 0,0 = ffff0000 pixel at 1,1 = ff00ff00 pixel at 2,2 = ff0000ff pixel at 0,0 = 1, 0, 0 pixel at 1,1 = 0, 1, 0 pixel at 2,2 = 0, 0, 1 pixel at 0,0 = red pixel at 1,1 = green pixel at 2,2 = blue
M2000 Interpreter
The easy way is to make a function to return an object with all functions on it, for specific image. We have to make the image in a way to render it to screen. The render statement get data in a string using a header of 12 characters (24 bytes). Raster lines are in down-top order. So last raster line is the top one. Also RGB is BGR in this data structure. Raster lines has to be aligned proper, so we may have add some bytes.
There are four functions (lambda functions) in the returned group, each of them has closures, an one of that closure is a pointer to a buffer object. We use this object to set and get pixels.
First two functions are for set and get pixel. Third return image as a string. Forth function copy an image as string to buffer, if they have the same width and height (else we get error)
P3 ppm
<lang M2000 Interpreter> \ Bitmap width in pixels, height in pixels \ Return a group object with some lambda as members: SetPixel, GetPixel, Image$ \ copyimage \ using Copy x, y Use Image$ we can display image$ to x, y as twips \ we can use x*twipsx, y*twipsy for x,y as pixels Function Bitmap (x as long, y as long) {
if x<1 or y<1 then Error "Wrong dimensions"
structure rgb {
red as byte
green as byte
blue as byte
}
m=len(rgb)*x mod 4
if m>0 then m=4-m ' add some bytes to raster line
m+=len(rgb) *x
Structure rasterline {
{
pad as byte*m
}
\\ union pad+hline
hline as rgb*x
}
Structure Raster {
magic as integer*4
w as integer*4
h as integer*4
lines as rasterline*y
}
Buffer Clear Image1 as Raster
\\ 24 chars as header to be used from bitmap render build in functions
Return Image1, 0!magic:="cDIB", 0!w:=Hex$(x,2), 0!h:=Hex$(y, 2)
\\ fill white (all 255)
\\ Str$(string) convert to ascii, so we get all characters from words width to byte width
Return Image1, 0!lines:=Str$(String$(chrcode$(255), Len(rasterline)*y))
Buffer Clear Pad as Byte*4
SetPixel=Lambda Image1, Pad,aLines=Len(Raster)-Len(Rasterline), blines=-Len(Rasterline) (x, y, c) ->{
where=alines+3*x+blines*y
if c>0 then c=color(c)
c-!
Return Pad, 0:=c as long
Return Image1, 0!where:=Eval(Pad, 2) as byte, 0!where+1:=Eval(Pad, 1) as byte, 0!where+2:=Eval(Pad, 0) as byte
}
GetPixel=Lambda Image1,aLines=Len(Raster)-Len(Rasterline), blines=-Len(Rasterline) (x,y) ->{
where=alines+3*x+blines*y
=color(Eval(image1, where+2 as byte), Eval(image1, where+1 as byte), Eval(image1, where as byte))
}
StrDib$=Lambda$ Image1, Raster -> {
=Eval$(Image1, 0, Len(Raster))
}
CopyImage=Lambda Image1 (image$) -> {
if left$(image$,12)=Eval$(Image1, 0, 24 ) Then {
Return Image1, 0:=Image$
} Else Error "Can't Copy Image"
}
Group Bitmap {
SetPixel=SetPixel
GetPixel=GetPixel
Image$=StrDib$
Copy=CopyImage
}
=Bitmap
} A=Bitmap(100,100) Call A.SetPixel(50,50, color(128,0,255)) Print A.GetPixel(50,50)=color(128,0,255) \\ display image to screen at 100, 50 pixel copy 100*twipsx,50*twipsy use A.Image$() A1=Bitmap(100,100) Call A1.copy(A.Image$()) copy 500*twipsx,50*twipsy use A1.Image$()
</lang>
P6 ppm
Need Version 9.4, Rev >=19 <lang M2000 Interpreter> Module P6 {
Function Bitmap {
def x as long, y as long, Import as boolean
If match("NN") then {
Read x, y
} else.if Match("N") Then {
\\ is a file?
Read f as long
buffer whitespace as byte
if not Eof(f) then {
get #f, whitespace :P6$=eval$(whitespace)
get #f, whitespace : P6$+=eval$(whitespace)
def boolean getW=true, getH=true, getV=true
def long v
\\ str$("P6") has 2 bytes. "P6" has 4 bytes
If p6$=str$("P6") Then {
do {
get #f, whitespace
if Eval$(whitespace)=str$("#") then {
do {get #f, whitespace} until eval(whitespace)=10
} else {
select case eval(whitespace)
case 32, 9, 13, 10
{ if getW and x<>0 then {
getW=false
} else.if getH and y<>0 then {
getH=false
} else.if getV and v<>0 then {
getV=false
}
}
case 48 to 57
{if getW then {
x*=10
x+=eval(whitespace, 0)-48
} else.if getH then {
y*=10
y+=eval(whitespace, 0)-48
} else.if getV then {
v*=10
v+=eval(whitespace, 0)-48
}
}
End Select
}
iF eof(f) then Error "Not a ppm file"
} until getV=false
} else Error "Not a P6 ppm"
Import=True
}
} else Error "No proper arguments"
if x<1 or y<1 then Error "Wrong dimensions"
structure rgb {
red as byte
green as byte
blue as byte
}
m=len(rgb)*x mod 4
if m>0 then m=4-m ' add some bytes to raster line
m+=len(rgb) *x
Structure rasterline {
{
pad as byte*m
}
hline as rgb*x
}
Structure Raster {
magic as integer*4
w as integer*4
h as integer*4
{
linesB as byte*len(rasterline)*y
}
lines as rasterline*y
}
Buffer Clear Image1 as Raster
Return Image1, 0!magic:="cDIB", 0!w:=Hex$(x,2), 0!h:=Hex$(y, 2)
if not Import then Return Image1, 0!lines:=Str$(String$(chrcode$(255), Len(rasterline)*y))
Buffer Clear Pad as Byte*4
SetPixel=Lambda Image1, Pad,aLines=Len(Raster)-Len(Rasterline), blines=-Len(Rasterline) (x, y, c) ->{
where=alines+3*x+blines*y
if c>0 then c=color(c)
c-!
Return Pad, 0:=c as long
Return Image1, 0!where:=Eval(Pad, 2) as byte, 0!where+1:=Eval(Pad, 1) as byte, 0!where+2:=Eval(Pad, 0) as byte
}
GetPixel=Lambda Image1,aLines=Len(Raster)-Len(Rasterline), blines=-Len(Rasterline) (x,y) ->{
where=alines+3*x+blines*y
=color(Eval(image1, where+2 as byte), Eval(image1, where+1 as byte), Eval(image1, where as byte))
}
StrDib$=Lambda$ Image1, Raster -> {
=Eval$(Image1, 0, Len(Raster))
}
CopyImage=Lambda Image1 (image$) -> {
if left$(image$,12)=Eval$(Image1, 0, 24 ) Then {
Return Image1, 0:=Image$
} Else Error "Can't Copy Image"
}
Export2File=Lambda Image1, x, y (f) -> {
Print #f, "P6";chr$(10);"# Created using M2000 Interpreter";chr$(10);
Print #f, x;" ";y;" 255";chr$(10);
x2=x-1 : where=0
Buffer pad as byte*3
For y1= 0 to y-1 {
For x1=0 to x2 {
Return pad, 0:=eval$(image1, 0!linesB!where, 3)
Push Eval(pad, 2) : Return pad, 2:=Eval(pad, 0), 0:=Number
Put #f, pad : where+=3
}
m=where mod 4 : if m<>0 then where+=4-m
}
}
if Import then {
x0=x-1 : where=0
Buffer Pad1 as byte*3
For y1=y-1 to 0 {
For x1=0 to x0 {Get #f, Pad1 : Push Eval(pad1, 2) : Return pad1, 2:=Eval(pad1, 0), 0:=Number
Return Image1, 0!linesB!where:=Eval$(Pad1) : where+=3}
m=where mod 4 : if m<>0 then where+=4-m}
}
Group Bitmap {
SetPixel=SetPixel
GetPixel=GetPixel
Image$=StrDib$
Copy=CopyImage
ToFile=Export2File
}
=Bitmap
}
A=Bitmap(150,100)
For i=0 to 98 {
Call A.SetPixel(i, i, 0)
Call A.SetPixel(99, i, 0)
}
Call A.SetPixel(i,i,0)
Copy 200*twipsx, 100*twipsy use A.Image$()
Profiler
Open "a.ppm" for output as #F
Call A.tofile(f)
Close #f
Print Filelen("a.ppm")
Print Timecount/1000;"sec"
Profiler
Image A.Image$() Export "a.jpg", 100 ' per cent quality
Print Filelen("a.jpg")
Image A.Image$() Export "a1.jpg", 10 ' per cent quality
Print Filelen("a1.jpg")
Image A.Image$() Export "a.bmp"
Print Filelen("a.bmp") ' no compression
Print Timecount/1000;"sec"
Move 5000,5000 ' twips
Image "a.jpg"
Move 5000,8000
Image "a1.jpg"
Move 8000, 5000
Image "a.bmp"
} P6 </lang> Export using M2000 code for ppm is slower than using internal jpg and bmp encoders. Jpg encoder has a 100% quality, and because this image is black and white we get the best compression. Time 0.304sec is for three exports, two jpg and one bmp.
- Output:
45049
47.3661341sec
1018
691
45254
0.3040944sec
Maple
<lang Maple>allocateImg := proc(width, height) return Array(1..width, 1..height, 1..3); end proc: fillColor := proc(img, rgb::list) local i; for i from 1 to 3 do img[..,..,i] := map(x->rgb[i], img[..,..,i]): end do: end proc: setColor := proc(x, y, img, rgb::list) local i: for i from 1 to 3 do img[x,y,i] := rgb[i]: end do: end proc: getColor := proc(x,y,img) local rgb,i: rgb := Array(1..3): for i from 1 to 3 do rgb(i) := img[x,y,i]: end do: return rgb: end proc:</lang>
- Use:
a := allocateImg(200,200); fillColor(a,[255,223,0]); setColor(150,150, a, [0,0,0]); getColor(150,150,a); #Output the image ImageTools:-Embed(ImageTools:-Create(a))
Mathematica / Wolfram Language
In Mathematica 7/8: <lang Mathematica>img = Image[ConstantArray[{1, 0, 0}, {1000, 1000}]]; img = ReplacePart[img, {1, 1, 1} -> {0, 0, 1}]; ImageValue[img, {1, 1}]</lang> In Mathematica 9: <lang Mathematica>img = Image[ConstantArray[{1, 0, 0}, {1000, 1000}]]; img = ReplacePixelValue[img, {1, 1} -> {0, 0, 1}]; ImageValue[img, {1, 1}]</lang>
MATLAB
Save this in a file named Bitmap.mat in a folder named @Bitmap in your MATLAB root directory. <lang MATLAB> %Bitmap class % %Implements a class to manage bitmap images without the need for the %various conversion and display functions % %Available functions: % %fill(obj,color) %setPixel(obj,pixel,color) %getPixel(obj,pixel,[optional: color channel]) %display(obj) %disp(obj) %plot(obj) %image(obj) %save(obj) %open(obj)
classdef Bitmap
%% Public Properties properties
%Channel arrays
red;
green;
blue;
end
%% Public Methods methods
%Creates image and defaults it to black
function obj = Bitmap(width,height)
obj.red = zeros(height,width,'uint8');
obj.green = zeros(height,width,'uint8');
obj.blue = zeros(height,width,'uint8');
end % End Bitmap Constructor
%Fill the image with a specified color
%color = [red green blue] max for each is 255
function fill(obj,color)
obj.red(:,:) = color(1);
obj.green(:,:) = color(2);
obj.blue(:,:) = color(3);
assignin('caller',inputname(1),obj); %saves the changes to the object
end
%Set a pixel to a specified color
%pixel = [x y]
%color = [red green blue]
function setPixel(obj,pixel,color)
obj.red(pixel(2),pixel(1)) = color(1);
obj.green(pixel(2),pixel(1)) = color(2);
obj.blue(pixel(2),pixel(1)) = color(3);
assignin('caller',inputname(1),obj); %saves the changes to the object
end
%Get pixel color
%pixel = [x y]
%varargin can be:
% no input for all channels
% 'r' or 'red' for red channel
% 'g' or 'green' for green channel
% 'b' or 'blue' for blue channel
function color = getPixel(obj,pixel,varargin)
if( ~isempty(varargin) )
switch (varargin{1})
case {'r','red'}
color = obj.red(pixel(2),pixel(1));
case {'g','green'}
color = obj.red(pixel(2),pixel(1));
case {'b','blue'}
color = obj.red(pixel(2),pixel(1));
end
else
color = [obj.red(pixel(2),pixel(1)) obj.green(pixel(2),pixel(1)) obj.blue(pixel(2),pixel(1))];
end
end
%Display the image
%varargin can be:
% no input for all channels
% 'r' or 'red' for red channel
% 'g' or 'green' for green channel
% 'b' or 'blue' for blue channel
function display(obj,varargin)
if( ~isempty(varargin) )
switch (varargin{1})
case {'r','red'}
image(obj.red)
case {'g','green'}
image(obj.green)
case {'b','blue'}
image(obj.blue)
end
colormap bone;
else
bitmap = cat(3,obj.red,obj.green,obj.blue);
image(bitmap);
end
end
%Overload several commonly used display functions
function disp(obj,varargin)
display(obj,varargin{:});
end
function plot(obj,varargin)
display(obj,varargin{:});
end
function image(obj,varargin)
display(obj,varargin{:});
end
%Saves the image
function save(obj)
%Open file dialogue
[fileName,pathName,success] = uiputfile({'*.bmp','Bitmap Image (*.bmp)'},'Save Bitmap As');
if( not(success == 0) )
imwrite(cat(3,obj.red,obj.green,obj.blue),[pathName fileName],'bmp'); %Write image file to disk
disp('Save Complete');
end
end
%Opens an image and overwrites what is currently stored
function success = open(obj)
%Open file dialogue
[fileName,pathName,success] = uigetfile({'*.bmp','Bitmap Image (*.bmp)'},'Open Bitmap ');
if( not(success == 0) )
channels = imread([pathName fileName], 'bmp'); %returns color indexed data
%Store each channel
obj.red = channels(:,:,1);
obj.green = channels(:,:,2);
obj.blue = channels(:,:,3);
assignin('caller',inputname(1),obj); %saves the changes to the object
success = true;
return
else
success = false;
return
end
end
end %methods
end %classdef </lang>
Sample Usage: <lang MATLAB> >> img = Bitmap(20,30); >> img.fill([30 30 150]); >> img.setPixel([10 15],[20 130 66]); >> disp(img) >> img.getPixel([10 15])
ans =
20 130 66
>> img.getPixel([10 15],'red')
ans =
20
>> img.save() Save Complete </lang>
MAXScript
MAXScript provides a built-in Bitmap class. <lang maxscript>local myBitmap = bitmap 512 512</lang>
Filling the image with a single colour can be accomplished at creation time by setting the color property. <lang maxscript>local myBitmap = bitmap 512 512 color:(color 128 128 128)</lang>
Use setPixels to set the colour of a pixel. This function takes an array of colours and is optimised to set the colours of a whole row of pixels. <lang maxscript>setPixels myBitmap [256, 256] #((color 255 255 255))</lang>
Use getPixels to retrieve the colour of a pixel. As with setPixels, this function is optimised to retrieve one row at a time as an array of colour values. <lang maxscript>local myPixel = getPixels myBitmap [256, 256] 1</lang>
Modula-3
Since this code is for use with other tasks, it uses an interface as well as the implementation module. <lang modula3>INTERFACE Bitmap;
TYPE UByte = BITS 8 FOR [0 .. 16_FF];
Pixel = RECORD R, G, B: UByte; END;
Point = RECORD x, y: UByte; END;
T = REF ARRAY OF ARRAY OF Pixel;
CONST
Black = Pixel{0, 0, 0};
White = Pixel{255, 255, 255};
Red = Pixel{255, 0, 0};
Green = Pixel{0, 255, 0};
Blue = Pixel{0, 0, 255};
Yellow = Pixel{255, 255, 0};
EXCEPTION BadImage;
BadColor;
PROCEDURE NewImage(height, width: UByte): T RAISES {BadImage}; PROCEDURE Fill(VAR pic: T; color: Pixel); PROCEDURE GetPixel(VAR pic: T; point: Point): Pixel RAISES {BadColor}; PROCEDURE SetPixel(VAR pic: T; point: Point; color: Pixel);
END Bitmap.</lang> <lang modula3>MODULE Bitmap;
PROCEDURE NewImage(height, width: UByte): T RAISES {BadImage} =
(* To make things easier, limit image size to also
be UByte (0 to 255), and to have equal dimensions. *)
BEGIN
IF height # width THEN
RAISE BadImage;
END;
RETURN NEW(T, height, width);
END NewImage;
PROCEDURE Fill(VAR pic: T; color: Pixel) =
BEGIN
FOR i := FIRST(pic^) TO LAST(pic^) DO
FOR j := FIRST(pic[0]) TO LAST(pic[0]) DO
pic[i,j] := color;
END;
END;
END Fill;
PROCEDURE GetPixel(VAR pic: T; point: Point): Pixel RAISES {BadColor} =
VAR pixel := pic[point.x, point.y];
BEGIN
IF pixel = White THEN
RETURN White;
ELSIF pixel = Black THEN
RETURN Black;
ELSIF pixel = Red THEN
RETURN Red;
ELSIF pixel = Green THEN
RETURN Green;
ELSIF pixel = Blue THEN
RETURN Blue;
ELSIF pixel = Yellow THEN
RETURN Yellow;
ELSE
RAISE BadColor;
END;
END GetPixel;
PROCEDURE SetPixel(VAR pic: T; point: Point; color: Pixel) =
BEGIN pic[point.x, point.y] := color; END SetPixel;
BEGIN END Bitmap.</lang>
Nim
<lang nim>type
Luminance* = uint8 Index* = int
Color* = tuple r, g, b: Luminance
Image* = ref object w*, h*: Index pixels*: seq[Color]
Point* = tuple x, y: Index
proc color*(r, g, b: SomeInteger): Color =
## Build a color value from R, G and B values. result.r = r.uint8 result.g = g.uint8 result.b = b.uint8
const
Black* = color( 0, 0, 0) White* = color(255, 255, 255)
proc newImage*(width, height: int): Image =
## Create an image with given width and height. new(result) result.w = width result.h = height result.pixels.setLen(width * height)
iterator indices*(img: Image): Point =
## Yield the pixels coordinates as tuples.
for y in 0 ..< img.h:
for x in 0 ..< img.w:
yield (x, y)
proc `[]`*(img: Image; x, y: int): Color =
## Get a pixel RGB value. img.pixels[y * img.w + x]
proc `[]=`*(img: Image; x, y: int; c: Color) =
## Set a pixel RGB value to given color. img.pixels[y * img.w + x] = c
proc fill*(img: Image; color: Color) =
## Fill the image with a color. for x, y in img.indices: img[x, y] = color
proc print*(img: Image) =
## Output an ASCII representation of the image.
for x, y in img.indices:
if x mod img.w == 0:
stdout.write '\n'
stdout.write if img[x, y] == White: '.' else: 'H'
stdout.write '\n'
when isMainModule:
var img = newImage(100, 20) img.fill color(255, 255, 255) img[1, 2] = color(255, 0, 0) img[3, 4] = img[1, 2] img.print</lang>
OCaml
<lang ocaml>let new_img ~width ~height =
let all_channels = let kind = Bigarray.int8_unsigned and layout = Bigarray.c_layout in Bigarray.Array3.create kind layout 3 width height in let r_channel = Bigarray.Array3.slice_left_2 all_channels 0 and g_channel = Bigarray.Array3.slice_left_2 all_channels 1 and b_channel = Bigarray.Array3.slice_left_2 all_channels 2 in (all_channels, r_channel, g_channel, b_channel)</lang>
and here is the type of the raster image this function returns:
type raster = (int, Bigarray.int8_unsigned_elt, Bigarray.c_layout) Bigarray.Array3.t * (int, Bigarray.int8_unsigned_elt, Bigarray.c_layout) Bigarray.Array2.t * (int, Bigarray.int8_unsigned_elt, Bigarray.c_layout) Bigarray.Array2.t * (int, Bigarray.int8_unsigned_elt, Bigarray.c_layout) Bigarray.Array2.t
What is particular with this allocation and its associated type is that there is not only one buffer for each RGB channel, but also an additionnal one that handles all the three channels, and what is important here is that it is not additionnal memory, the memory is shared, so there are 2 ways to access the raster buffer: through the separated RGB channels, or through the joint channel (all_channels).
This solution have a lot of advantages across a more naive one: this type is compatible to memory-map a file (a ppm file for instance, where the data is not compressed), the buffer can be shared/exchanged with C (for OpenGL textures for instance), etc.
A more naive form would be this one:
<lang ocaml>let new_img ~width ~height =
let r_channel, g_channel, b_channel =
let kind = Bigarray.int8_unsigned
and layout = Bigarray.c_layout
in
(Bigarray.Array2.create kind layout width height,
Bigarray.Array2.create kind layout width height,
Bigarray.Array2.create kind layout width height)
in
(r_channel,
g_channel,
b_channel)</lang>
Here are the functions to fill with a color and to set one given pixel:
<lang ocaml>let fill_img ~img:(_, r_channel, g_channel, b_channel) ~color:(r,g,b) =
Bigarray.Array2.fill r_channel r; Bigarray.Array2.fill g_channel g; Bigarray.Array2.fill b_channel b;
- </lang>
<lang ocaml>let put_pixel_unsafe (_, r_channel, g_channel, b_channel) (r,g,b) =
(fun x y ->
r_channel.{x,y} <- r;
g_channel.{x,y} <- g;
b_channel.{x,y} <- b;
)</lang>
<lang ocaml>let get_pixel_unsafe (_, r_channel, g_channel, b_channel) =
(fun x y ->
(r_channel.{x,y},
g_channel.{x,y},
b_channel.{x,y})
)</lang>
we can overload these functions to make some bound checks:
<lang ocaml>let put_pixel img color x y =
let _, r_channel,_,_ = img in let width = Bigarray.Array2.dim1 r_channel and height = Bigarray.Array2.dim2 r_channel in
if (x < 0) || (x >= width) then invalid_arg "x out of bounds"; if (y < 0) || (y >= height) then invalid_arg "y out of bounds";
let r, g, b = color in if (r < 0) || (r > 255) then invalid_arg "red out of bounds"; if (g < 0) || (g > 255) then invalid_arg "green out of bounds"; if (b < 0) || (b > 255) then invalid_arg "blue out of bounds";
put_pixel_unsafe img color x y;
let get_pixel ~img ~pt:(x, y) =
let _, r_channel,_,_ = img in let width = Bigarray.Array2.dim1 r_channel and height = Bigarray.Array2.dim2 r_channel in if (x < 0) || (x >= width) then invalid_arg "x out of bounds"; if (y < 0) || (y >= height) then invalid_arg "y out of bounds"; get_pixel_unsafe img x y;
- </lang>
and a function to get the dimensions:
<lang ocaml>let get_dims ~img:(_, r_channel, _, _) =
let width = Bigarray.Array2.dim1 r_channel and height = Bigarray.Array2.dim2 r_channel in (width, height)</lang>
Octave
In Octave, images are matrix. A grayscale W×H image is stored as a W×H matrix, and RGB W×H image is stored as a W×H×3 image. Possible levels depend on the class of the storage: if it is double, the intensity is a floating point double number between 0 and 1; if it is uint8, the intensity is from 0 to 255; if it is uint16, the intensity is between 0 and 65535.
<lang octave>im = zeros(W, H, 3, "uint8"); % create an RGB image of width W and height H
% and intensity from 0 to 255; black (all zeros)
im(:,:,1) = 255; % set R to 255 im(:,:,2) = 100; % set G to 100 im(:,:,3) = 155; % set B to 155 im(floor(W/2), floor(H/2), :) = 0; % pixel in the center made black disp(im(floor(W/3), floor(H/3), :)) % display intensities of the pixel
% at W/3, H/3
p = im(40,40,:); % or just store it in the vector p, so that
% p(1) is R, p(2) G and p(3) is B</lang>
We can hide this in helper functions like:
<lang octave>function im = create_rgb_image(w, h)
im = zeros(w, h, 3, "uint8");
endfunction
function set_background(im, colorvector)
im(:,:,1) = colorvector(1); im(:,:,2) = colorvector(2); im(:,:,3) = colorvector(3);
endfunction
function set_pixel(im, coord, colorvector)
im(coord(1), coord(2), 1) = colorvector(1); im(coord(1), coord(2), 2) = colorvector(2); im(coord(1), coord(2), 3) = colorvector(3);
endfunction
function [r, g, b] = get_pixel(im, coord)
r = im(coord(1), coord(2), 1) g = im(coord(1), coord(2), 2) b = im(coord(1), coord(2), 3)
endfunction</lang>
The only thing to note is that indexing start from 1.
<lang octave>%example im = create_rgb_image(200,200); for x = 1:128
im = set_pixel(im, [x, x], [200, 50, 220]);
endfor
% it seems like saveimage wants double class on [0,1] saveimage("image.ppm", double(im)./256, "ppm");</lang>
OxygenBasic
<lang oxygenbasic> 'GENERIC BITMAP
type pixel byte r,g,b
'=========== class BitMap '===========
% sp sizeof(pixel) sys wx,wy,px,py string buf sys pb method Constructor(sys x=640,y=480) { wx=x : wy=y : buf=nuls x*y*sp : pb=strptr buf} method Destructor() {buf="" : wx=0 : wy=0 : pb=0} method GetPixel(sys x,y,pixel*p) {copy @p,pb+(y*wx+x)*sp,sp} method SetPixel(sys x,y,pixel*p) {copy pb+(y*wx+x)*sp,@p,sp} ' method Fill(pixel*p)
sys i, e=wx*wy*sp+pb-1
for i=pb to e step sp {copy i,@p,sp}
end method
end class
pixel p,q
new BitMap m(400,400) 'width, height in pixels
p<=100,120,140 'red,green,blue
m.fill p
m.getPixel 200,100,q print "" q.r "," q.g "," q.b 'result 100,120,140 q<=10,20,40 m.setPixel 200,100,q m.getPixel 200,100,p print "" p.r "," p.g "," p.b 'result 10,20,40
del m
</lang>
Oz
We first create a 2D array data type as a functor in a file "Array2D.oz": <lang oz>functor export
New Get Set Transform
define
fun {New Width Height Init}
C = {Array.new 1 Height unit}
in
for Row in 1..Height do
C.Row := {Array.new 1 Width Init}
end
array2d(width:Width
height:Height contents:C)
end
fun {Get array2d(contents:C ...) X Y}
C.Y.X
end
proc {Set array2d(contents:C ...) X Y Val}
C.Y.X := Val
end
proc {Transform array2d(contents:C width:W height:H ...) Fun}
for Y in 1..H do
for X in 1..W do C.Y.X := {Fun C.Y.X} end
end end
%% omitted: Clone, Map, Fold, ForAll
end</lang>
Based on this, we create a functor "Bitmap.oz": <lang oz>%% For real task prefer QTk's images: %% http://www.mozart-oz.org/home/doc/mozart-stdlib/wp/qtk/html/node38.html
functor import
Array2D
export
New Fill GetPixel SetPixel
define
Black = color(0x00 0x00 0x00)
fun {New Width Height}
bitmap( {Array2D.new Width Height Black} )
end
proc {Fill bitmap(Arr) Color}
{Array2D.transform Arr fun {$ _} Color end}
end
fun {GetPixel bitmap(Arr) X Y}
{Array2D.get Arr X Y}
end
proc {SetPixel bitmap(Arr) X Y Color}
{Array2D.set Arr X Y Color}
end
%% Omitted: MaxValue, ForAllPixels, Transform
end</lang>
Some functions that are used in other tasks were omitted. See here for the complete module definitions: Basic bitmap storage/Oz
Pascal
<lang Pascal>Interface uses crt, { GetDir }
graph; { function GetPixel }
type { integer numbers }
{ from unit bitmaps XPERT software production Tamer Fakhoury }
_bit = $00000000..$00000001; {number 1 bit without sign = (0..1) }
_byte = $00000000..$000000FF; {number 1 byte without sign = (0..255)}
_word = $00000000..$0000FFFF; {number 2 bytes without sign = (0..65 535)}
_dWord = $00000000..$7FFFFFFF; {number 4 bytes without sign = (0..4 294 967 296)}
_longInt = $80000000..$7FFFFFFF; {number 4 bytes with sign
= (-2 147 483 648..2 147 483 648}
TbmpFileHeader =
record
ID: _word; { Must be 'BM' =19778=$424D for windows }
FileSize: _dWord; { Size of this file in bytes }
Reserved: _dWord; { ??? }
bmpDataOffset: _dword; { = 54 = $36 from begining of file to begining of bmp data }
end;
TbmpInfoHeader =
record
InfoHeaderSize: _dword; { Size of Info header
= 28h = 40 (decimal)
for windows }
Width,
Height: _longInt; { Width and Height of image in pixels }
Planes, { number of planes of bitmap }
BitsPerPixel: _word; { Bits can be 1, 4, 8, 24 or 32 }
Compression,
bmpDataSize: _dword; { in bytes rounded to the next 4 byte boundary }
XPixPerMeter, { horizontal resolution in pixels }
YPixPerMeter: _longInt; { vertical }
NumbColorsUsed,
NumbImportantColors: _dword; {= NumbColorUsed}
end; { TbmpHeader = Record ... }
T32Color =
record { 4 byte = 32 bit }
Blue: byte;
Green: byte;
Red: byte;
Alfa: byte
end;
var directory,
bmpFileName: string;
bmpFile: file; { untyped file }
bmpFileHeader: TbmpFileHeader;
bmpInfoHeader: TbmpInfoHeader;
color32: T32Color;
RowSizeInBytes: integer;
BytesPerPixel: integer;
const defaultBmpFileName = 'test';
DefaultDirectory = 'c:\bp\';
DefaultExtension = '.bmp';
bmpFileHeaderSize = 14;
{ compression specyfication }
bi_RGB = 0; { compression }
bi_RLE8 = 1;
bi_RLE4 = 2;
bi_BITFIELDS = 3;
bmp_OK = 0;
bmp_NotBMP = 1;
bmp_OpenError = 2;
bmp_ReadError = 3;
Procedure CreateBmpFile32(directory: string; FileName: string;
iWidth, iHeight: _LongInt);
{************************************************} Implementation {-----------------------------} {************************************************}
Procedure CreateBmpFile32(directory: string; FileName: string;
iWidth, iHeight: _LongInt);
var
x, y: integer;
begin
if directory = then
GetDir(0, directory);
if FileName = then
FileName: = DefaultBmpFileName;
{ create a new file on a disk in a given directory with given name }
Assign(bmpFile, directory + FileName + DefaultExtension);
ReWrite(bmpFile, 1);
{ fill the headers }
with bmpInfoHeader, bmpFileHeader do
begin
ID := 19778;
InfoheaderSize := 40;
width := iWidth;
height := iHeight;
BitsPerPixel := 32;
BytesPerPixel := BitsPerPixel div 8;
reserved := 0;
bmpDataOffset := InfoHeaderSize + bmpFileHeaderSize;
planes := 1;
compression := bi_RGB;
XPixPerMeter := 0;
YPixPerMeter := 0;
NumbColorsUsed := 0;
NumbImportantColors := 0;
RowSizeInBytes := (Width * BytesPerPixel); { only for >=8 bits per pixel }
bmpDataSize := height * RowSizeinBytes;
FileSize := InfoHeaderSize + bmpFileHeaderSize + bmpDataSize;
{ copy headers to disk file }
BlockWrite(bmpFile, bmpFileHeader, bmpFileHeaderSize);
BlockWrite(bmpFile, bmpInfoHeader, infoHeaderSize);
{ fill the pixel data area }
for y := (height - 1) downto 0 do
begin
for x := 0 to (width - 1) do begin { Pixel(x,y) } color32.Blue := 255; color32.Green := 0; color32.Red := 0; color32.Alfa := 0; BlockWrite(bmpFile, color32, 4); end; { for x ... }
end; { for y ... }
Close(bmpFile);
end; { with bmpInfoHeader, bmpFileHeader }
end; { procedure }
</lang>
Perl
<lang perl>#! /usr/bin/perl
use strict;
use Image::Imlib2;
- create the "canvas"
my $img = Image::Imlib2->new(200,200);
- fill with a plain RGB(A) color
$img->set_color(255, 0, 0, 255); $img->fill_rectangle(0,0, 200, 200);
- set a pixel to green (at 40,40)
$img->set_color(0, 255, 0, 255); $img->draw_point(40,40);
- "get" pixel rgb(a)
my ($red, $green, $blue, $alpha) = $img->query_pixel(40,40); undef $img;
- another way of creating a canvas with a bg colour (or from
- an existing "raw" data)
my $col = pack("CCCC", 255, 255, 0, 0); # a, r, g, b my $img = Image::Imlib2->new_using_data(200, 200, $col x (200 * 200));
exit 0;</lang>
Phix
Copy of Euphoria <lang Phix>-- Some colour constants: constant black = #000000, -- blue = #0000FF, -- green = #00FF00, -- red = #FF0000,
white = #FFFFFF
-- Create new image filled with some colour function new_image(integer width, integer height, integer fill_colour=black)
return repeat(repeat(fill_colour,height),width)
end function
-- Usage example: sequence image = new_image(800,600)
-- Set pixel color: image[400][300] = white
-- Get pixel color integer colour = image[400][300] -- Now colour is #FF0000</lang>
PHP
<lang PHP>class Bitmap {
public $data;
public $w;
public $h;
public function __construct($w = 16, $h = 16){
$white = array_fill(0, $w, array(255,255,255));
$this->data = array_fill(0, $h, $white);
$this->w = $w;
$this->h = $h;
}
//Fills a rectangle, or the whole image with black by default
public function fill($x = 0, $y = 0, $w = null, $h = null, $color = array(0,0,0)){
if (is_null($w)) $w = $this->w;
if (is_null($h)) $h = $this->h;
$w += $x;
$h += $y;
for ($i = $y; $i < $h; $i++){
for ($j = $x; $j < $w; $j++){
$this->setPixel($j, $i, $color);
}
}
}
public function setPixel($x, $y, $color = array(0,0,0)){
if ($x >= $this->w) return false;
if ($x < 0) return false;
if ($y >= $this->h) return false;
if ($y < 0) return false;
$this->data[$y][$x] = $color;
}
public function getPixel($x, $y){
return $this->data[$y][$x];
}
}
$b = new Bitmap(16,16); $b->fill(); $b->fill(2, 2, 18, 18, array(240,240,240)); $b->setPixel(0, 15, array(255,0,0)); print_r($b->getPixel(3,3)); //(240,240,240)</lang>
PicoLisp
For time critical applications this would be done with inline-C in PicoLisp, but especially for small bitmaps the following makes sense. <lang PicoLisp># Create an empty image of 120 x 90 pixels (setq *Ppm (make (do 90 (link (need 120)))))
- Fill an image with a given color
(de ppmFill (Ppm R G B)
(for Y Ppm
(map
'((X) (set X (list R G B)))
Y ) ) )
- Set pixel with a color
(de ppmSetPixel (Ppm X Y R G B)
(set (nth Ppm Y X) (list R G B)) )
- Get the color of a pixel
(de ppmGetPixel (Ppm X Y)
(get Ppm Y X) )</lang>
PL/I
<lang PL/I> /* Declaration for an image, suitable for BMP files. */ declare image(0:500, 0:500) bit (24) aligned;
image = '000000000000000011111111'b;
/* Sets the entire image to red. */
image(10,40) = '111111110000000000000000'b;
/* Sets one pixel to blue. */
declare color bit (24) aligned; color = image(20,50); /* Obtain the color of a pixel */
/* To allocate an image of size (x,y) */ allocate_image: procedure (image, x, y);
declare image (*, *) controlled bit (24) aligned; declare (x, y) fixed binary (31);
allocate image (0:x, 0:y);
end allocate_image;
/* To use the above procedure, it's necessary to define */ /* the image in the calling program thus, for BMP images: */
declare image(*,*) controlled bit (24) aligned; </lang>
Processing
<lang java>PGraphics bitmap = createGraphics(100,100); // Create the bitmap bitmap.beginDraw(); bitmap.background(255, 0, 0); // Fill bitmap with red rgb color bitmap.endDraw(); image(bitmap, 0, 0); // Place bitmap on screen. color b = color(0, 0, 255); // Define a blue rgb color set(50, 50, b); // Set blue colored pixel in the middle of the screen color c = get(50, 50); // Get the color of same pixel if(b == c) print("Color changed correctly"); // Verify </lang>
Prolog
<lang prolog>
- - module(bitmap, [
new_bitmap/3, fill_bitmap/3, get_pixel0/3, set_pixel0/4 ]).
- - use_module(library(lists)).
%-----------------------------------------------------------------------------% % Convenience Predicates replicate(Term,Times,L):- length(L,Times), maplist(=(Term),L).
replace0(N,OL,E,NL):- nth0(N,OL,_,TL), nth0(N,NL,E,TL). %-----------------------------------------------------------------------------% % Bitmap Utilities % % The Bitmap structure is a list with pixels kept in row major order: % [dimensions-[X,Y],pixels-[[n11,n12...],[n21,n22...]]]
% In this code what exactly an RGB value is doesn't matter however % in other bitmap tasks it is assumed to be a list [R,G,B] where % each is an int between 0 and 255, in code: rgb_pixel(RGB):- length(RGB,3), maplist(integer,RGB), maplist(between(0,255),RGB).
%new_bitmap(Bitmap,Dimensions,RGB) new_bitmap([[X,Y],Pixels],[X,Y],RGB) :- replicate(RGB,X,Row), replicate(Row,Y,Pixels).
%fill_bitmap(New_Bitmap,Bitmap,RGB) fill_bitmap(New_Bitmap,[[X,Y],_],RGB) :- new_bitmap(New_Bitmap,[X,Y],RGB).
%here get and set use 0 based indexing %get_pixel0(Bitmap,Coordinates,RGB) get_pixel0([[_DimX,_DimY],Pixels],[X,Y],RGB) :- nth0(Y,Pixels,Row), nth0(X,Row,RGB).
%set_pixel0(New Bitmap, Bitmap, Coordinates, RGB) set_pixel0([[DimX,DimY],New_Pixels],[[DimX,DimY],Pixels],[X,Y],RGB) :- nth0(Y,Pixels,Row), replace0(X,Row,RGB,New_Row), replace0(Y,Pixels,New_Row,New_Pixels). </lang>
PureBasic
<lang PureBasic>w=800 : h=600 CreateImage(1,w,h)
- 1 is internal id of image
StartDrawing(ImageOutput(1))
- fill with color red
Box(0,0,w,h,$ff)
- or using another (but slower) way in green
FillArea(0,0,-1,$ff00)
- a green Dot
Plot(10,10,$ff0000)
- check if we set it right (should be 255)
Debug Blue(Point(10,10)) </lang>
Python
See Basic bitmap storage/Python
R
R can write to most bitmap image formats by default (mostly for the purpose of saving graphs), however there is no built-in way of manipulating images. The pixmap package reads, writes and manipulates portable bitmap file types: PBM, PGM, PPM. See also, the image function, and the rimage and ReadImage packages, which use libjpeg to read JPEG and PNG files. <lang r># See the class definitions and constructors with, e.g. getClass("pixmapIndexed", package=pixmap) pixmapIndexed
- Image with all one colour
plot(p1 <- pixmapIndexed(matrix(0, nrow=3, ncol=4), col="red"))
- Image with one pixel specified
cols <- rep("blue", 12); cols[7] <- "red" plot(p2 <- pixmapIndexed(matrix(1:12, nrow=3, ncol=4), col=cols))
- Retrieve colour of a pixel
getcol <- function(pm, i, j) {
pmcol <- pm@col dim(pmcol) <- dim(pm@index) pmcol[i,j]
} getcol(p2, 3, 4) #red</lang>
Racket
<lang racket>
- lang racket
- The racket/draw libraries provide imperative drawing functions.
- http://docs.racket-lang.org/draw/index.html
(require racket/draw)
- To create an image with width and height, use the make-bitmap
- function.
- For example, let's make a small image here
(define bm (make-bitmap 640 480))
- We use a drawing context handle, a "dc", to operate on the bitmap.
(define dc (send bm make-dc))
- We can fill the bitmap with a color by using a combination of
- setting the background, and clearing.
(send dc set-background (make-object color% 0 0 0)) ;; Color it black. (send dc clear)
- Let's set a few pixels to a greenish color with set-pixel
(define aquamarine (send the-color-database find-color "aquamarine")) (for ([i 480])
(send dc set-pixel i i aquamarine))
- We can get at the color of a bitmap pixel by using the get-pixel
- method. However, it may be faster to use get-argb-pixels if we
- need a block of the pixels. Let's use get-argb-pixels and look
- at a row starting at (0, 42)
(define buffer (make-bytes (* 480 4))) ;; alpha, red, green, blue (send dc get-argb-pixels 0 42 480 1 buffer)
- We can inspect the buffer
(bytes-ref buffer 0) ;; and see that the first pixel's alpha is 255, (bytes-ref buffer 1) ;; and the red, green, and blue components are 0. (bytes-ref buffer 2) (bytes-ref buffer 3)
- If we are using DrRacket, we can just print the bm as a toplevel expression
- to view the final image
bm </lang>
Raku
(formerly Perl 6) <lang perl6>class Pixel { has UInt ($.R, $.G, $.B) } class Bitmap {
has UInt ($.width, $.height); has Pixel @!data;
method fill(Pixel $p) {
@!data = $p.clone xx ($!width*$!height)
}
method pixel(
$i where ^$!width, $j where ^$!height --> Pixel
) is rw { @!data[$i + $j * $!width] }
method set-pixel ($i, $j, Pixel $p) {
self.pixel($i, $j) = $p.clone;
}
method get-pixel ($i, $j) returns Pixel {
self.pixel($i, $j);
}
}
my Bitmap $b = Bitmap.new( width => 10, height => 10);
$b.fill( Pixel.new( R => 0, G => 0, B => 200) );
$b.set-pixel( 7, 5, Pixel.new( R => 100, G => 200, B => 0) );
say $b.perl;</lang>
Thanks to the rw trait on the pixel method, we don't actually need to define two separate methods, set-pixel and get-pixel, but that is an explicit requirement of the task. (Beware your presuppositions! In Raku, accessors only determine identity, not use. In particular, identity is considered orthogonal to lvalue/rvalue context.)
RapidQ
QCanvas is an empty image on which you can draw. QForm is the main window of the application. The commands to draw on the canvas are in the procedure PaintCanvas, which is executed each time the canvas need to be (re)painted.
<lang rapidq>DECLARE SUB PaintCanvas
CREATE form AS QForm
Width = 640
Height = 480
CREATE canvas AS QCanvas
Height = form.ClientHeight
Width = form.ClientWidth OnPaint = PaintCanvas
END CREATE
END CREATE
SUB PaintCanvas
' Fill background canvas.FillRect(0, 0, canvas.Width, canvas.Height, &H301000)
' Draw a pixel canvas.Pset(300, 200, &H00ddff)
' Read pixel color PRINT canvas.Pixel(300, 200)
END SUB
form.ShowModal</lang>
REXX
version 1
The REXX language has no need to declare the size of (stemmed) arrays.
Indeed, there is no way to declare array sizes (or any variable, for that matter).
The image (raster) created was also written to a file (image.PPM) to show verification of the image. <lang rexx>/*REXX program demonstrates how to process/display a simple RGB raster graphics image.*/ red = 'ff 00 00'x /*a method to define a red value. */ blue = '00 00 ff'x /*" " " " " blue " */ @. = /*define entire @. array to nulls. */ outFN = 'image' /*the filename of the output image PPM */ sWidth = 500; sHeight= 500 /*the screen width and height in pixels*/ call RGBfill red /*set the entire image to red. */
x= 10; y= 40 /*set pixel's coördinates. */
call RGBset x, y, blue /*set a pixel (at 10,40) to blue. */ color = RGBget(x, y) /*get the color of a pixel. */ hexV = c2x(color) /*get hex value of pixel's color. */ binV = x2b(hexV) /* " binary " " " " */ bin3V = left(binV, 8) substr(binV, 9, 8) right(binV, 8) hex3V = left(hexV, 2) substr(hexV, 3, 2) right(hexV, 2) xy= '(' || x","y')' /*create a handy─dandy literal for SAY.*/ say xy ' pixel in binary: ' binV /*show the binary value of 20,50 */ say xy ' pixel in binary: ' bin3V /*show again, but with spaces. */ say /*show a blank between binary and hex. */ say xy ' pixel in hex: ' hexV /*show again, but in hexadecimal. */ say xy ' pixel in hex: ' hex3V /*show again, but with spaces. */ call PPMwrite outFN, sWidth, sHeight /*create a PPM (output) file of image. */ /* ◄■■■■■■■■ not part of this task.*/ say /*show a blank. */ say 'The file ' outFN".PPM was created." /*inform user that a file was created. */ exit /*stick a fork in it, we're all done. */ /*──────────────────────────────────────────────────────────────────────────────────────*/ RGBfill: @.=arg(1); return /*fill image with a color.*/ RGBget: parse arg px,py; return @.px.py /*get a pixel's color. */ RGBset: parse arg px,py,p$; @.px.py=p$; return /*set " " " */ /*──────────────────────────────────────────────────────────────────────────────────────*/ PPMwrite: parse arg oFN, width, height /*obtain output filename, width, height*/
oFID= oFN'.PPM'; $='9'x; #=255 /*fileID; separator; max color value.*/
call charout oFID, , 1 /*set the position of the file's output*/
call charout oFID,'P6'width || $ || height || $ || # || $ /*write hdr info.*/
do j=1 for width
do k=1 for height; call charout oFID, @.j.k
end /*k*/ /* ↑ write the PPM file, ··· */
end /*j*/ /* └───────── ··· one pixel at a time.*/
call charout oFID; return /*close the output file just to be safe*/</lang>
- output:
(10,40) pixel in binary: 000000000000000011111111 (10,40) pixel in binary: 00000000 00000000 11111111 (10,40) pixel in hex: 0000FF (10,40) pixel in hex: 00 00 FF The file image.PPM was created.
version 2
This program actually creates a BMP file <lang rexx>/* REXX ***************************************************************
- Draw a picture from pixels
- 16.06.2014 Walter Pachl
- /
oid='pic.bmp'; 'erase' oid
blue ='FF0000'x; green='00FF00'x; red ='0000FF'x; white='ffffff'x; black='000000'x;
w=600 /* width */ h=300 /* height */ w3=w*3
bfType ='BM' bfSize ='46000000'x bfReserved ='00000000'x bfOffBits ='36000000'x biSize ='28000000'x biWidth =lend(w) biHeight =lend(h) biPlanes ='0100'x biBitCount ='1800'x biCompression ='00000000'x biSizeImage ='10000000'x biXPelsPerMeter='00000000'x biYPelsPerMeter='00000000'x biClrUsed ='00000000'x biClrImportant ='00000000'x
s=bfType||,
bfSize||, bfReserved||, bfOffBits||, biSize||, biWidth||, biHeight||, biPlanes||, biBitCount||, biCompression||, biSizeImage||, biXPelsPerMeter||, biYPelsPerMeter||, biClrUsed||, biClrImportant
pic=copies(red,w*h) /* fill the rectangle with color red */ Call rect 100,100,180,180,green /* draw a green rectangle */ Call rect 100,100,160,160,blue /* and a blue rectangle within that */ Call dot 120,120,white /* one pixel is hardly visible */ Do x=98 To 102 /* draw a square of 25 pixels */
Do y=98 To 102 Call dot x,y,white End End
Call charout oid,s||pic /* write the picture to file */ dmy=col(97,98) dmy=col(98,98) Exit
lend: Procedure /**********************************************************************
- compute the representation of a number (little endian)
- /
Parse Arg n res=reverse(d2c(n,4)) rev=reverse(res) say 'lend:' arg(1) '->' c2x(res) '=>' c2d(rev) Return res
rect: Procedure Expose pic w h w3 /**********************************************************************
- Fill a rectangle with center at x,y and width/height = wr/hr
- /
Parse Arg x,y,wr,hr,color Say x y wr hr c2x(color) i=w3*(y-1)+3*(x-1)+1 /* Pixel position of center */ ia=max(w3*(y-1)+1,i-3*(wr%2)) /* position of left border */ ib=min(i+3*wr%2,w3*y) /* position of right border */ lc=ib-ia /* length of horizontal line */ If lc>=0 Then Do
os=copies(color,lc%3) /* the horizontal line */
Do hi=-hr%2 to hr%2 /* loop from lower to upper border*/
i=trunc(ia+w3*hi) /* position of line's left border */
If i>1 Then Do
pic=overlay(os,pic,i) /* put the line into the picture */
j=i%w3
End
End
End
Return
dot: Procedure Expose pic w h w3 /**********************************************************************
- Put a dot at position x/y into the picture
- /
Parse Arg x,y,color i=w3*(y-1)+3*(x-1) pic=overlay(color,pic,i+1) Return
col: Procedure Expose pic w h w3 /**********************************************************************
- get the color at position x/y
- /
Parse Arg x,y,color i=w3*(y-1)+3*(x-1) say 'color at pixel' x'/'y'='c2x(substr(pic,i+1,3)) Return c2x(substr(pic,i+1,3))</lang>
- Output:
lend: 600 -> 58020000 => 600 lend: 300 -> 2C010000 => 300 100 100 180 180 00FF00 100 100 160 160 FF0000 color at pixel 97/98=FF0000 color at pixel 98/98=FFFFFF and have a look at the file pic.bmp created by this program
Ruby
I haven't been able to find any kind of package for manipulating bitmap images, so let's roll one <lang ruby>class RGBColour
def initialize(red, green, blue)
unless red.between?(0,255) and green.between?(0,255) and blue.between?(0,255)
raise ArgumentError, "invalid RGB parameters: #{[red, green, blue].inspect}"
end
@red, @green, @blue = red, green, blue
end
attr_reader :red, :green, :blue
alias_method :r, :red
alias_method :g, :green
alias_method :b, :blue
RED = RGBColour.new(255,0,0) GREEN = RGBColour.new(0,255,0) BLUE = RGBColour.new(0,0,255) BLACK = RGBColour.new(0,0,0) WHITE = RGBColour.new(255,255,255)
end
class Pixmap
def initialize(width, height) @width = width @height = height @data = fill(RGBColour::WHITE) end attr_reader :width, :height
def fill(colour)
@data = Array.new(@width) {Array.new(@height, colour)}
end
def validate_pixel(x,y)
unless x.between?(0, @width-1) and y.between?(0, @height-1)
raise ArgumentError, "requested pixel (#{x}, #{y}) is outside dimensions of this bitmap"
end
end
def [](x,y) validate_pixel(x,y) @data[x][y] end alias_method :get_pixel, :[]
def []=(x,y,colour) validate_pixel(x,y) @data[x][y] = colour end alias_method :set_pixel, :[]=
end</lang>
Rust
<lang Rust>#[derive(Copy, Clone, Debug, PartialEq, Eq)] pub struct Rgb {
pub r: u8, pub g: u8, pub b: u8,
}
impl Rgb {
pub fn new(r: u8, g: u8, b: u8) -> Self {
Rgb { r, g, b }
}
pub const BLACK: Rgb = Rgb { r: 0, g: 0, b: 0 };
pub const RED: Rgb = Rgb { r: 255, g: 0, b: 0 };
pub const GREEN: Rgb = Rgb { r: 0, g: 255, b: 0 };
pub const BLUE: Rgb = Rgb { r: 0, g: 0, b: 255 };
}
- [derive(Clone, Debug)]
pub struct Image {
width: usize, height: usize, pixels: Vec<Rgb>,
}
impl Image {
pub fn new(width: usize, height: usize) -> Self {
Image {
width,
height,
pixels: vec![Rgb::BLACK; width * height],
}
}
pub fn width(&self) -> usize {
self.width
}
pub fn height(&self) -> usize {
self.height
}
pub fn fill(&mut self, color: Rgb) {
for pixel in &mut self.pixels {
*pixel = color;
}
}
pub fn get(&self, row: usize, col: usize) -> Option<&Rgb> {
if row >= self.width {
return None;
}
self.pixels.get(row * self.width + col)
}
pub fn get_mut(&mut self, row: usize, col: usize) -> Option<&mut Rgb> {
if row >= self.width {
return None;
}
self.pixels.get_mut(row * self.width + col)
}
}
fn main() {
let mut image = Image::new(16, 9); assert_eq!(Some(&Rgb::BLACK), image.get(3, 4)); assert!(image.get(22, 3).is_none());
image.fill(Rgb::RED); assert_eq!(Some(&Rgb::RED), image.get(3, 4));
if let Some(pixel) = image.get_mut(3, 4) {
*pixel = Rgb::GREEN;
}
assert_eq!(Some(&Rgb::GREEN), image.get(3, 4));
if let Some(pixel) = image.get_mut(3, 4) {
pixel.g -= 100;
pixel.b = 20;
}
assert_eq!(Some(&Rgb::new(0, 155, 20)), image.get(3, 4));
}</lang>
Scala
Java translation
<lang scala>import java.awt.image.BufferedImage import java.awt.Color
class RgbBitmap(val width:Int, val height:Int) {
val image=new BufferedImage(width, height, BufferedImage.TYPE_3BYTE_BGR)
def fill(c:Color)={
val g=image.getGraphics()
g.setColor(c)
g.fillRect(0, 0, width, height)
}
def setPixel(x:Int, y:Int, c:Color)=image.setRGB(x, y, c.getRGB()) def getPixel(x:Int, y:Int)=new Color(image.getRGB(x, y))
}</lang> Usage: <lang scala>val img=new RgbBitmap(50, 50); img.fill(Color.CYAN) img.setPixel(5, 5, Color.BLUE)
assert(img.getPixel(1,1)==Color.CYAN) assert(img.getPixel(5,5)==Color.BLUE) assert(img.width==50) assert(img.height==50)</lang>
Scala idiom
A more Scalesque version could be with the use of its idiom:
- Output:
Best experienced in your browser with Scastie (remote JVM). <lang scala>import java.awt.image.BufferedImage import java.awt.Color
object RgbBitmap extends App {
class RgbBitmap(val dim: (Int, Int)) {
def width = dim._1
def height = dim._2
private val image = new BufferedImage(width, height, BufferedImage.TYPE_3BYTE_BGR)
def apply(x: Int, y: Int) = new Color(image.getRGB(x, y))
def update(x: Int, y: Int, c: Color) = image.setRGB(x, y, c.getRGB)
def fill(c: Color) = {
val g = image.getGraphics
g.setColor(c)
g.fillRect(0, 0, width, height)
}
}
object RgbBitmap {
def apply(width: Int, height: Int) = new RgbBitmap(width, height)
}
/** Even Javanese style testing is still possible.
*/
private val img0 = new RgbBitmap(50, 60) { // Wrappers to enable adhoc Javanese style
def getPixel(x: Int, y: Int) = this(x, y)
def setPixel(x: Int, y: Int, c: Color) = this(x, y) = c
}
img0.fill(Color.CYAN)
img0.setPixel(5, 6, Color.BLUE)
// Testing in Java style
assert(img0.getPixel(0, 1) == Color.CYAN)
assert(img0.getPixel(5, 6) == Color.BLUE)
assert(img0.width == 50)
assert(img0.height == 60)
println("Tests successfully completed with no errors found.")
}</lang>
Scheme
Definitions of list procedures: <lang scheme>(define (make-list length object)
(if (= length 0)
(list)
(cons object (make-list (- length 1) object))))
(define (list-fill! list object)
(if (not (null? list))
(begin (set-car! list object) (list-fill! (cdr list) object))))
(define (list-set! list element object)
(if (= element 1)
(set-car! list object)
(list-set! (cdr list) (- element 1) object)))
(define (list-get list element)
(if (= element 1)
(car list)
(list-get (cdr list) (- element 1))))</lang>
Definitions of image procedures: <lang scheme>(define (make-image columns rows)
(if (= rows 0)
(list)
(cons (make-list columns (list)) (make-image columns (- rows 1)))))
(define (image-fill! image colour)
(if (not (null? image))
(begin (list-fill! (car image) colour) (image-fill! (cdr image) colour))))
(define (image-set! image column row colour)
(list-set! (list-get image row) column colour))
(define (image-get image column row)
(list-get (list-get image row) column))</lang>
Definitions of some colours: <lang scheme>(define *black* (list 0 0 0)) (define *white* (list 255 255 255)) (define *red* (list 255 0 0)) (define *green* (list 0 255 0)) (define *blue* (list 0 0 255))</lang> This creates a small image with a black background and a single blue pixel: <lang scheme>(define image (make-image 3 2)) (image-fill! image *black*) (image-set! image 2 1 *blue*) (display image) (newline)</lang>
- Output:
<lang>(((0 0 0) (0 0 255) (0 0 0)) ((0 0 0) (0 0 0) (0 0 0)))</lang>
Seed7
The types and functions requested are predefined in the libraries graph.s7i and draw.s7i:
- The type to handle an RGB raster graphics image is PRIMITIVE_WINDOW.
- The function to create an image is newPixmap.
- An imaged can be filled with a color with clear.
- A given pixel can be set with point.
- The color of a pixel can be retrieved with getPixelColor.
<lang seed7>$ include "seed7_05.s7i";
include "draw.s7i";
const proc: main is func
local var PRIMITIVE_WINDOW: myPixmap is PRIMITIVE_WINDOW.value; var color: myColor is black; begin myPixmap := newPixmap(300, 200); clear(myPixmap, light_green); point(myPixmap, 20, 30, color(256, 512, 768)); myColor := getPixelColor(myPixmap, 20, 30); writeln(myColor.redLight <& " " <& myColor.greenLight <& " " <& myColor.blueLight); end func;</lang>
SequenceL
<lang seed7>RGB ::= (R: int(0), G: int(0), B: int(0));
newBitmap: int * int -> RGB(2); newBitmap(width, height)[y, x] := (R: 0, G: 0, B: 0) foreach y within 1 ... height, x within 1 ... width;
fill: RGB(2) * RGB -> RGB(2); fill(bitmap(2), color)[y, x] := color foreach y within 1 ... size(bitmap), x within 1 ... size(bitmap[y]);
setColorAt: RGB(2) * int * int * RGB -> RGB(2); setColorAt(bitmap(2), x, y, color)[Y, X] := color when Y = y and X = x else bitmap[Y, X];
getColorAt: RGB(2) * int * int -> RGB; getColorAt(bitmap(2), x, y) := bitmap[y, x];
lightGreen := (R: 51, G: 255, B: 51); lightRed := (R: 255, G: 51, B: 51);
main(args(2)) := let width := 1920; height := 1200;
cleanImage := newBitmap(width, height);
filledGreen := fill(cleanImage, lightGreen);
redCenter := setColorAt(filledGreen, width / 2, height / 2, lightRed); in getColorAt(redCenter, width / 2, height / 2);</lang>
- Output:
cmd:> main.exe (B:51,G:51,R:255)
Smalltalk
<lang Smalltalk> </lang>
Tcl
<lang tcl>package require Tcl 8.5 package require Tk namespace path ::tcl::mathfunc ;# for [max] function
proc newImage {width height} {
return [image create photo -width $width -height $height]
} proc fill {image colour} {
$image put $colour -to 0 0 [$image cget -width] [$image cget -height]
} proc setPixel {image colour point} {
lassign $point x y $image put $colour -to [max 0 $x] [max 0 $y]
} proc getPixel {image point} {
lassign $point x y
# [$img get] returns a list: {r g b}; this proc should return a colour value
format {#%02x%02x%02x} {*}[$image get $x $y]
}
- create the image and display it
set img [newImage 150 150] label .l -image $img pack .l
fill $img red
setPixel $img green {40 40}
set rbg [getPixel $img {40 40}]</lang>
TI-89 BASIC
TI-89 BASIC does not have user-defined data structures. The Rosetta Code tasks which use this image type have instead been implemented using the TI-89's graph screen.
UNIX Shell
<lang bash>typeset -T RGBColor_t=(
integer r g b
function to_s {
printf "%d %d %d" ${_.r} ${_.g} ${_.b}
}
function white { print "255 255 255"; }
function black { print "0 0 0"; }
function red { print "255 0 0"; }
function green { print "0 255 0"; }
function blue { print "0 0 255"; }
function yellow { print "255 255 0"; }
function magenta { print "255 0 255"; }
function cyan { print "0 255 255"; }
)
typeset -T Bitmap_t=(
integer height integer width typeset -a data
function fill {
typeset color=$1
if [[ -z ${color:+set} ]]; then
print -u2 "error: no fill color specified"
return 1
fi
integer x y
for ((y=0; y<_.height; y++)); do
for ((x=0; x<_.width; x++)); do
_.data[y][x]="$color"
done
done
}
function setpixel {
integer x=$1 y=$2
typeset color=$3
_.data[y][x]=$color
}
function getpixel {
integer x=$1 y=$2
print "${_.data[y][x]}"
}
function to_s {
typeset ppm=""
ppm+="P3"$'\n'
ppm+="${_.width} ${_.height}"$'\n'
ppm+="255"$'\n'
typeset sep
for ((y=0; y<_.height; y++)); do
sep=""
for ((x=0; x<_.width; x++)); do
ppm+="$sep${_.data[y][x]}"
sep=" "
done
ppm+=$'\n'
done
print -- "$ppm"
}
)
RGBColor_t color Bitmap_t b=( width=3 height=2 ) b.fill "$(color.white)" b.setpixel 0 0 "$(color.red)" b.setpixel 1 0 "$(color.green)" b.setpixel 2 0 "$(color.blue)" b.setpixel 0 1 "$(color.yellow)" b.setpixel 1 1 "$(color.white)" b.setpixel 2 1 "$(color.black)" echo "$(b.getpixel 0 0)" b.to_s</lang>
- Output:
255 0 0 P3 3 2 255 255 0 0 0 255 0 0 0 255 255 255 0 255 255 255 0 0 0
Vedit macro language
An edit buffer is used to store pixel data. In order to allow unlimited image size, a temporary file (here pixel.data) can be assosicated to the buffer. You could directly open the image file you are creating (as in the task Dragon_curve, but here we first create just the plain pixel data so that the required image file format can be decided later. <lang vedit>#11 = 400 // Width of the image
- 12 = 300 // Height of the image
// Create an empty RGB image and fill it with black color // File_Open("|(VEDIT_TEMP)\pixel.data", OVERWRITE+NOEVENT) BOF Del_Char(ALL)
- 10 = Buf_Num
Repeat(#11 * #12) {
Ins_Char(0, COUNT, 3)
}
// Fill the image with dark blue color //
- 5 = 0 // Red
- 6 = 0 // Green
- 7 = 64 // Blue
Call("FILL_IMAGE")
// Draw one pixel in orange color //
- 1 = 100 // x
- 2 = 50 // y
- 5 = 255 #6 = 128 #7 = 0 // Orange color
Call("DRAW_PIXEL")
// Get the color of a pixel //
- 1 = 10
- 2 = 3
Call("GET_COLOR")
Buf_Switch(#10) Buf_Quit(OK) Return
///////////////////////////////////////////////////////////////////// // // Fill image with given color: #5 = Red, #6 = Green, #7 = Blue //
- FILL_IMAGE:
BOF Repeat (File_Size/3) {
IC(#5,OVERWRITE) IC(#6,OVERWRITE) IC(#7,OVERWRITE)
} Return
///////////////////////////////////////////////////////////////////// // // Daw a pixel. #1 = x, #2 = y //
- DRAW_PIXEL:
Goto_Pos((#1 + #2*#11)*3) IC(#5,OVERWRITE) IC(#6,OVERWRITE) IC(#7,OVERWRITE) Return
///////////////////////////////////////////////////////////////////// // // Get color of a pixel. #1 = x, #2 = y // Return: #5 = Red, #6 = Green, #7 = Blue //
- GET_COLOR:
Goto_Pos((#1 + #2*#11)*3)
- 5 = Cur_Char
- 6 = Cur_Char(1)
- 7 = Cur_Char(2)
Return</lang>
Visual Basic .NET
<lang vbnet>' The StructLayout attribute allows fields to overlap in memory. <System.Runtime.InteropServices.StructLayout(LayoutKind.Explicit)> _ Public Structure Rgb
<FieldOffset(0)> _ Public Rgb As Integer
<FieldOffset(0)> _ Public B As Byte
<FieldOffset(1)> _ Public G As Byte
<FieldOffset(2)> _ Public R As Byte
Public Sub New(ByVal r As Byte, ByVal g As Byte, ByVal b As Byte)
Me.R = r
Me.G = g
Me.B = b
End Sub
End Structure</lang>
<lang vbnet>Public Class RasterBitmap
Private m_pixels() As Rgb
Private m_width As Integer
Public ReadOnly Property Width As Integer
Get
Return m_width
End Get
End Property
Private m_height As Integer
Public ReadOnly Property Height As Integer
Get
Return m_height
End Get
End Property
Public Sub New(ByVal width As Integer, ByVal height As Integer)
m_pixels = New Rgb(width * height - 1) {}
m_width = width
m_height = height
End Sub
Public Sub Clear(ByVal color As Rgb)
For i As Integer = 0 To m_pixels.Length - 1
m_pixels(i) = color
Next
End Sub
Public Sub SetPixel(ByVal x As Integer, ByVal y As Integer, ByVal color As Rgb)
m_pixels((y * m_width) + x) = color
End Sub
Public Function GetPixel(ByVal x As Integer, ByVal y As Integer) As Rgb
Return m_pixels((y * m_width) + x)
End Function
End Class</lang>
Wren
The above library's ImageData class fits the bill here (version 1.3.0 or later). <lang ecmascript>import "graphics" for Canvas, ImageData, Color import "dome" for Window
class Game {
static bmpCreate(name, w, h) { ImageData.create(name, w, h) }
static bmpFill(name, col) {
var image = ImageData[name]
for (x in 0...image.width) {
for (y in 0...image.height) image.pset(x, y, col)
}
}
static bmpPset(name, x, y, col) { ImageData[name].pset(x, y, col) }
static bmpPget(name, x, y) { ImageData[name].pget(x, y) }
static init() {
Window.title = "Bitmap"
var size = 600
Window.resize(size, size)
Canvas.resize(size, size)
var bmp = bmpCreate("rcbmp", size/2, size/2)
bmpFill("rcbmp", Color.yellow)
bmpPset("rcbmp", size/4, size/4, Color.blue) // 'blue' is #29ADFF on the default palette
var col = bmpPget("rcbmp", size/4, size/4)
System.print(col.toString) // check it's blue - alpha component (FF) will also be shown
bmp.draw(150, 150)
}
static update() {}
static draw(alpha) {}
}</lang>
- Output:
Color (#29ADFFFF)
Xojo
<lang vb>Function CreatePicture(width As Integer, height As Integer) As Picture
Return New Picture(width, height)
End Function
Sub FillPicture(ByRef p As Picture, FillColor As Color)
p.Graphics.ForeColor = FillColor p.Graphics.FillRect(0, 0, p.Width, p.Height)
End Sub
Function GetPixelColor(p As Picture, x As Integer, y As Integer) As Color
Return p.RGBSurface.Pixel(x, y)
End Function
Sub SetPixelColor(p As Picture, x As Integer, y As Integer, pColor As Color)
p.RGBSurface.Pixel(x, y) = pColor
End Sub </lang>
XPL0
<lang XPL0>include c:\cxpl\codes; \include 'code' declarations def Width=180, Height=135, Color=$123456; int X, Y; [SetVid($112); \set display for 640x480 graphics in 24-bit RGB color for Y:= 0 to Height-1 do \fill area with Color one pixel at a time
for X:= 0 to Width-1 do \(this takes 4.12 ms on a Duron 850)
Point(X, Y, Color);
Move(60, 60); HexOut(6, ReadPix(0,0)); \show color of pixel at 0,0 X:= ChIn(1); \wait for keystroke SetVid(3); \restore display to normal text mode ]</lang>
zkl
This solution creates a PPM image. No error checking or clipping. <lang zkl>class PPM{ // (0,0) is logically bottom left
fcn init(width,height){
sz:=width*height*3;
var [const]
data=sz.pump(Data(sz),0), // initialize to Black (RGB=000)
w=width, h=height;
}
fcn fill(rgb){
sz:=data.len()/3;
data.clear(); sz.pump(data,T(Void,rgb.toBigEndian(3)));
}
fcn __sGet(x,y) { data.toBigEndian(3*y*w + 3*x,3); } //ppm[x,y]
fcn __sSet(rbg,x,y){ data[3*y*w + 3*x,3]=rbg.toBigEndian(3); } //ppm[x,y]=rgb
fcn write(out){ // write bottom to top to move (0,0) from bottom left to bottom left
out.write("P6\n#rosettacode PPM\n%d %d\n255\n".fmt(w,h));
[h-1..0, -1].pump(out,'wrap(h){ data.seek(3*h*w); data.read(3*w) });
out.close();
}
}</lang> <lang zkl>ppm:=PPM(256,256); ppm.fill(0x00FF88); foreach x in ([50..200]){ ppm[x,50]=0xff|00|00; } // horizontal red line
ppm.write(File("foo.ppm","wb"));</lang>
- Output:
$ zkl hexDump foo.ppm | less 0: 50 36 0a 23 72 6f 73 65 | 74 74 61 63 6f 64 65 20 P6.#rosettacode 16: 50 50 4d 0a 32 35 36 20 | 32 35 36 0a 32 35 35 0a PPM.256 256.255. 32: 00 ff 88 00 ff 88 00 ff | 88 00 ff 88 00 ff 88 00 ................ 48: ff 88 00 ff 88 00 ff 88 | 00 ff 88 00 ff 88 00 ff ................ 64: 88 00 ff 88 00 ff 88 00 | ff 88 00 ff 88 00 ff 88 ................ 80: 00 ff 88 00 ff 88 00 ff | 88 00 ff 88 00 ff 88 00 ................ ...
- Programming Tasks
- Raster graphics operations
- AWK/Omit
- PARI/GP/Omit
- ActionScript
- Ada
- ALGOL 68
- AutoHotkey
- Axe
- BASIC256
- BBC BASIC
- C
- C sharp
- C++
- Boost
- Clojure
- Common Lisp
- Crystal
- D
- Delphi
- E
- EchoLisp
- Elixir
- Erlang
- Euphoria
- F Sharp
- Factor
- FBSL
- Forth
- Fortran
- Go
- Haskell
- Icon
- Unicon
- J
- Java
- AWT
- JUnit
- JavaScript
- Julia
- KonsolScript
- Kotlin
- Lingo
- LiveCode
- Lua
- M2000 Interpreter
- Maple
- Mathematica
- Wolfram Language
- MATLAB
- MAXScript
- Modula-3
- Nim
- OCaml
- Octave
- OxygenBasic
- Oz
- Pascal
- Perl
- Imlib2
- Phix
- PHP
- PicoLisp
- PL/I
- Processing
- Prolog
- PureBasic
- Python
- R
- Pixmap
- Racket
- Raku
- RapidQ
- REXX
- Ruby
- Rust
- Scala
- Scheme
- Seed7
- SequenceL
- Smalltalk
- Tcl
- Tk
- TI-89 BASIC
- UNIX Shell
- Vedit macro language
- Visual Basic .NET
- Wren
- DOME
- Xojo
- XPL0
- Zkl