Pixel Budget

Implement an ImageTransformer, a tool that can hold a set of images and apply some operations to transform them in various different ways.

You are given a sequence of grayscale images and a list of edit scripts. Each edit script is a sequence of named operations applied to one image to produce a new one. Your task is to apply each script and output the resulting image.

---

Image Format

All images are grayscale. Each pixel is an integer in [0, 255]. Images are represented as H x W grids of integers. Pixel values are clamped to [0, 255] after every operation.

---

Operations

Each operation takes the current image as input and produces a new image of the same dimensions. Operations are applied in the order listed in the script. The output image is the result after all operations have been applied.

Name	Arguments
BRIGHTNESS	d (integer)
FLIP_H	(none)
FLIP_V	(none)
CROP	r c h w
ROTATE_90	(none)
THRESHOLD	t (integer)

• BRIGHTNESS d — Add d to every pixel.
• FLIP_H — Flip the image horizontally (mirror left-right).
• FLIP_V — Flip the image vertically (mirror top-bottom).
• CROP r c h w — Extract the h x w sub-image starting at row r, column c. All subsequent operations use this smaller image.
• ROTATE_90 — Rotate the image 90° clockwise.
• THRESHOLD t — Set each pixel to 255 if it is ≥ t, else 0.

---

Input Format

N
<image_0>
<image_1>
...
<image_(N-1)>
S
<script_0>
<script_1>
...
<script_(S-1)>

Each image is given as:

H W
<row 0>
<row 1>
...
<row H-1>

Each script is given as:

img_index K
op_1 [args]
op_2 [args]
...
op_K [args]

where img_index is the 0-indexed image to apply the script to, and K is the number of operations.

---

Output Format

For each script (in order), output the resulting image:

H W
<row 0>
...
<row H-1>

with a blank line between images.

---

Notes

• Clamping is applied after each operation individually, not once at the end.
• ROTATE_90 clockwise means row r, column c maps to row c, column (H-1-r) in the new image (new dimensions are W x H).
• Scripts are independent — each starts from the original image at img_index, not from the result of a previous script.
• Pixel values in output rows are space-separated.