Reading pixel color / rebuilding an image

Asthalis

Hello everyone ;)

I was thinking about a way to use an image as a savestate for one application I'm working on. The main goal would be to avoid using a huge 2D array (that would need a lot of memory to save and reload).

But how can you "read" each pixel of an image using GDScript or build one pixel by pixel from 2-state simple data ?

Thanks if you can give me a clue :)

duane

Here's an example of manipulating pixels from my spritesheet helper. I'm taking sprite images and recoloring them, pixel-by-pixel, returning the new image.

func image(pos:Vector2, in_color=null, in_bgcolor=null, \
		rotate=false, size:=Vector2.ZERO) -> Image:
	var spout := Image.new()
	var spin:Image = images[pos]
	var dat:PoolByteArray
	var height = spin.get_height()
	var width = spin.get_width()
	var color
	var bgcolor

	if in_color != null:
		color = Color(in_color)

	if in_bgcolor != null:
		bgcolor = Color(in_bgcolor)

	# simple transpose -- width and height must be the same
	if rotate and width == height:
		var dat2:PoolByteArray = spin.get_data()
		dat = dat2

		for y in height:
			for x in width:
				for i in 4:
					dat[(x * width + y) * 4 + i] = dat2[(y * width + x) * 4 + i]

	if color != null or bgcolor != null:
		if len(dat) == 0:
			dat = spin.get_data()

		for i in range(0, len(dat), 4):
			if color != null and dat[i+3] == 255:
				dat[i+0] = color.r * 255
				dat[i+1] = color.g * 255
				dat[i+2] = color.b * 255
			elif bgcolor != null and dat[i+3] != 255:
				dat[i+0] = bgcolor.r * 255
				dat[i+1] = bgcolor.g * 255
				dat[i+2] = bgcolor.b * 255
				dat[i+3] = 255

		spout.create_from_data(spin.get_width(), spin.get_height(),
			false, spin.get_format(), dat)
	else:
		spout = spin

	if size:
		var sz:Vector2 = spout.get_size() * size
		spout.resize(int(sz.x), int(sz.y), 0)

	return spout

Asthalis

Thank for your listing. But I'm still a bit confused... I succeeded in making the "saving" part in reading the content or a tilemap and building an image from it :

var img = Image.new()
img.create(100, 100, false, Image.FORMAT_RGBA8)
img.lock()
for v in n: for u in n:
	if get_cell(u,v) == 1: img.set_pixel(u,v,Color(1,1,1,1))
	else: img.set_pixel(u,v,Color(0,0,0,1))
img.unlock()
img.save_png("res://image.png")

But I just can't get the "load" part working, when I want to "read" the pixels of the saved image to assign the correct values to my tilemap. Here's my (not working) code for now :

	var img = Image.new()
	img = load("res://image.png")
	for v in n: for u in n:
		if img.get_pixel(u,v) == 1: set_cell(u,v,1)
		else: set_cell(u,v,0)