replace image tasks myvote whiteboard indah with four corners image t…

…ask loader

admin_bot/admin_bot/utilities/image_tasks.py

@@ -3,22 +3,98 @@
 import numpy
 
 
-@register_image_task("myvote")
-def myvote_task(input_image: Image.Image) -> Image.Image:
-    frame = Image.open("admin_bot/resources/vote.png")
-    input_image = input_image.resize((375, 250))
-    frame.paste(input_image, (40, 240))
-    return frame
+# You can add new image tasks here
+
+image_tasks = {
+    "indah": {
+        "image": "admin_bot/resources/indah_sad.png",
+        "type": "overlay",
+        "corners": [
+            (67, 446),
+            (406, 290),
+            (513, 557),
+            (195, 736),
+        ],
+    },
+    "vote": {
+        "image": "admin_bot/resources/vote.png",
+        "type": "background",
+        "corners": [
+            (40, 240),
+            (415, 240),
+            (415, 490),
+            (40, 490),
+        ],
+    },
+    "whiteboard": {
+        "image": "admin_bot/resources/look_at_this/foreground.png",
+        "type": "overlay",
+        "corners": [
+            (1500, 75),
+            (2500, 75),
+            (2500, 2075),
+            (1500, 2075),
+        ],
+    },
+}
+
+for name, task in image_tasks.items():
+
+    def image_task(input_image: Image.Image):
+        template = Image.open(task["image"]).convert("RGBA")
+        transformed_image = four_corners(input_image, task["corners"])
+        result_image = Image.new(
+            "RGBA", template.size, (0, 0, 0, 0)
+        )  # Fully transparent
+        if task["type"] == "overlay": # put the input_image on bottom
+            result_image.paste(transformed_image, (0, 0), transformed_image)
+            result_image.paste(template, (0, 0), template)
+        else:  # "background" put the input_image on top # also fallback
+            result_image.paste(template, (0, 0), template)
+            result_image.paste(transformed_image, (0, 0), transformed_image)
+        return result_image
+
+    register_image_task(name)(image_task)
 
 
-@register_image_task("whiteboard")
-def whiteboard_task(input_image: Image.Image) -> Image.Image:
-    background = Image.open("admin_bot/resources/look_at_this/background.png")
-    foreground = Image.open("admin_bot/resources/look_at_this/foreground.png")
-    input_image = input_image.resize((1000, 2000))
-    background.paste(input_image, (1500, 75))
-    background.paste(foreground, (0, 0), foreground)
-    return background
+def find_coeffs(pa, pb):  # source: stack overflow
+    matrix = []
+    for p1, p2 in zip(pa, pb):
+        matrix.append([p1[0], p1[1], 1, 0, 0, 0, -p2[0] * p1[0], -p2[0] * p1[1]])
+        matrix.append([0, 0, 0, p1[0], p1[1], 1, -p2[1] * p1[0], -p2[1] * p1[1]])
+
+    A = numpy.matrix(matrix, dtype=float)
+    B = numpy.array(pb).reshape(8)
+
+    res = numpy.dot(numpy.linalg.inv(A.T * A) * A.T, B)
+    return numpy.array(res).reshape(8)
+
+
+def four_corners(image: Image.Image, output_corners):
+    """Transform an image by its corners
+
+    input_image: Image.Image -- Image that will be transformed
+    output_corners           -- Where the four corners of the input will go
+                                Array of Tuples (x, y). Top left corner first then clockwise
+    """
+
+    # input rectangular image corners
+    input_corners = [(0, 0), (image.width, 0), image.size, (0, image.height)]
+
+    coeffs = find_coeffs(output_corners, input_corners)  # calculate transformation
+
+    size = tuple(map(max, zip(*output_corners)))  # find max size of image
+
+    # apply transformation
+    transformed_image = image.convert("RGBA").transform(
+        size, Image.PERSPECTIVE, coeffs, Image.BICUBIC
+    )
+
+    return transformed_image
+
+
+# Extra image tasks. You can still make more commands if
+# you want to do something other than four corners transform
 
 
 @register_image_task("dylan")
@@ -48,53 +124,3 @@ def dylan_task(input_image: Image.Image) -> Image.Image:
     result_image.paste(foreground, (0, 0), foreground)
 
     return result_image
-
-
-@register_image_task("indah")
-def indah_task(input_image: Image.Image) -> Image.Image:
-    """
-    You've upset Indah
-    """
-
-    # Load the foreground overlay (ensure it has an alpha channel)
-    foreground = Image.open("admin_bot/resources/indah_sad.png").convert("RGBA")
-
-    # Perspective Transform, transform based on moving the four corners
-    input_corners = [  # input rectangular image corners
-        (0, 0),
-        (input_image.width, 0),
-        (input_image.width, input_image.height),
-        (0, input_image.height),
-    ]
-    output_corners = [  # four corners of the screen in foreground image
-        (67, 446),
-        (406, 290),
-        (513, 557),
-        (195, 736),
-    ]
-
-    coeffs = find_coeffs(output_corners, input_corners)  # calculate transformation
-
-    transformed_image = input_image.transform(
-        foreground.size, Image.PERSPECTIVE, coeffs, Image.BICUBIC
-    )  # apply transformation
-
-    # Create a new blank RGBA image with the same size as the foreground
-    result_image = Image.new("RGBA", foreground.size, (0, 0, 0, 0))  # Fully transparent
-    result_image.paste(transformed_image, (0, 0))
-    result_image.paste(foreground, (0, 0), foreground)
-
-    return result_image
-
-
-def find_coeffs(pa, pb):  # source: stack overflow
-    matrix = []
-    for p1, p2 in zip(pa, pb):
-        matrix.append([p1[0], p1[1], 1, 0, 0, 0, -p2[0] * p1[0], -p2[0] * p1[1]])
-        matrix.append([0, 0, 0, p1[0], p1[1], 1, -p2[1] * p1[0], -p2[1] * p1[1]])
-
-    A = numpy.matrix(matrix, dtype=float)
-    B = numpy.array(pb).reshape(8)
-
-    res = numpy.dot(numpy.linalg.inv(A.T * A) * A.T, B)
-    return numpy.array(res).reshape(8)