Add SplitImageToTileList and ImageMergeTileList nodes.

[comfyanonymous]

With these you can split an image into tiles, do operations and then combine it back to a single image.

[coderabbitai]

📝 Walkthrough

Walkthrough

This pull request adds two new image processing nodes to the ComfyUI image utilities. The first node, SplitImageToTileList, divides an input image into a list of tiles using configurable width, height, and overlap parameters. The second node, ImageMergeTileList, reconstructs a single image from a tile list with specified dimensions and overlap handling. Both nodes share grid coordinate generation logic and implement per-tile weighting using sine-based masks for blending. The math module is imported to support the weighting calculations. The updated node registry exposes both new nodes publicly.

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title directly and concisely describes the main change: adding two new image tile manipulation nodes to the codebase.
Description check	✅ Passed	The description is directly related to the changeset, explaining the purpose of the two new nodes and how they work together for tile-based image operations.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@comfy_extras/nodes_images.py`:
- Around line 735-737: In execute (and before calling get_grid_coords) validate
that overlap is strictly less than both tile_width and tile_height (or clamp it
to a safe value like min(tile_dim-1, overlap)); if invalid, raise a clear
ValueError mentioning tile_width/tile_height and overlap to fail fast, or
silently clamp overlap to a computed safe stride and document that
behavior—update the execute signature logic around image, tile_width,
tile_height, overlap to perform this check and then call get_grid_coords with
the validated/clamped overlap.
- Around line 796-808: The code currently reads image_list[0] unsafely and can
silently produce black regions when tiles are missing or mismatched; update the
execute method to validate inputs: first assert image_list is non-empty and
raise a clear ValueError if empty, then read first_tile = image_list[0] and
capture its shape/device/dtype, call cls.get_grid_coords(w,h,t_w,t_h,ovlp) as
now, and then verify len(image_list) == len(coords) raising a descriptive error
if not; also iterate the tiles in image_list and validate each has the same
batch/channel dims, identical t_h/t_w (or compatible), device and dtype as
first_tile and raise a ValueError listing the offending index(s) if any mismatch
is found so the method fails fast instead of producing silent corruption.

---

ℹ️ Review info

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 7a7debc and 022d3d7.

📒 Files selected for processing (1)

• comfy_extras/nodes_images.py

[coderabbitai]

⚠️ Potential issue | 🟠 Major

Validate overlap against tile size to prevent runaway tiling.
When overlap >= tile_width/height, stride collapses to 1, exploding tile counts (e.g., tile_width=64 with default overlap=128). Add a guard to fail fast or clamp.

🛡️ Proposed guard

     `@classmethod`
     def execute(cls, image, tile_width, tile_height, overlap):
+        if overlap >= tile_width or overlap >= tile_height:
+            raise ValueError("overlap must be smaller than tile_width and tile_height")
         b, h, w, c = image.shape
         coords = cls.get_grid_coords(w, h, tile_width, tile_height, overlap)

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@comfy_extras/nodes_images.py` around lines 735 - 737, In execute (and before
calling get_grid_coords) validate that overlap is strictly less than both
tile_width and tile_height (or clamp it to a safe value like min(tile_dim-1,
overlap)); if invalid, raise a clear ValueError mentioning
tile_width/tile_height and overlap to fail fast, or silently clamp overlap to a
computed safe stride and document that behavior—update the execute signature
logic around image, tile_width, tile_height, overlap to perform this check and
then call get_grid_coords with the validated/clamped overlap.

[coderabbitai]

⚠️ Potential issue | 🟡 Minor

Fail fast on empty or mismatched tile lists.
image_list[0] will throw on empty input, and mismatched counts silently create black regions. Add validation to avoid silent corruption.

🧪 Suggested validation

     `@classmethod`
     def execute(cls, image_list, final_width, final_height, overlap):
+        if not image_list:
+            raise ValueError("image_list is empty")
         w = final_width[0]
         h = final_height[0]
         ovlp = overlap[0]
         feather_str = 1.0
@@
         coords = cls.get_grid_coords(w, h, t_w, t_h, ovlp)
+        if len(image_list) != len(coords):
+            raise ValueError(f"Expected {len(coords)} tiles, got {len(image_list)}")

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@comfy_extras/nodes_images.py` around lines 796 - 808, The code currently
reads image_list[0] unsafely and can silently produce black regions when tiles
are missing or mismatched; update the execute method to validate inputs: first
assert image_list is non-empty and raise a clear ValueError if empty, then read
first_tile = image_list[0] and capture its shape/device/dtype, call
cls.get_grid_coords(w,h,t_w,t_h,ovlp) as now, and then verify len(image_list) ==
len(coords) raising a descriptive error if not; also iterate the tiles in
image_list and validate each has the same batch/channel dims, identical t_h/t_w
(or compatible), device and dtype as first_tile and raise a ValueError listing
the offending index(s) if any mismatch is found so the method fails fast instead
of producing silent corruption.