I think this is something env dependent

Fix problematic in-place mutations for dicts, lists, and sets

.github/workflows/publish.yml

@@ -7,15 +7,19 @@ on:
     paths:
       - "pyproject.toml"
 
+permissions:
+  issues: write
+
 jobs:
   publish-node:
     name: Publish Custom Node to registry
     runs-on: ubuntu-latest
+    if: ${{ github.repository_owner == 'aria1th' }}
     steps:
       - name: Check out code
         uses: actions/checkout@v4
       - name: Publish Custom Node
-        uses: Comfy-Org/publish-node-action@main
+        uses: Comfy-Org/publish-node-action@v1
         with:
           ## Add your own personal access token to your Github Repository secrets and reference it here.
-          personal_access_token: ${{ secrets.REGISTRY_ACCESS_TOKEN }}
+          personal_access_token: ${{ secrets.REGISTRY_ACCESS_TOKEN }}

README.md

@@ -0,0 +1,16 @@
+Logic Utilities for someone who wants
+
+~~prime list calculation in comfyui~~
+Proper documentation is being prepared, however there are too many nodes
+![image](https://github.com/user-attachments/assets/8e388417-6912-41d7-98fa-798b50eacfda)
+
+## Tests
+
+Run from the repo root:
+
+`python -m unittest discover -s tests`
+
+## Notes
+
+- Auto-install runs when loaded in ComfyUI. To disable it, set `COMFYUI_LOGICUTILS_SKIP_INSTALL=1`.
+- To force CPU/GPU tagger dependency selection, set `COMFYUI_LOGICUTILS_IMGUTILS_VARIANT=cpu` or `COMFYUI_LOGICUTILS_IMGUTILS_VARIANT=gpu`.

autonode.py

@@ -80,18 +80,18 @@ def validate(container):
         function_kwargs = list(function_kwargs) + ["self"]
         # if args/kwargs are in function kwargs, warn and skip
         if "args" in function_kwargs:
-            print(f"Warning: args in function arguments in {cls.__name__}, skipping argument validation")
+            #print(f"Warning: args in function arguments in {cls.__name__}, skipping argument validation")
             continue
         if "kwargs" in function_kwargs:
-            print(f"Warning: kwargs in function arguments in {cls.__name__}, skipping argument validation")
+            #print(f"Warning: kwargs in function arguments in {cls.__name__}, skipping argument validation")
             continue
         # input kwargs are subset of function kwargs
         if not set(input_keys).issubset(function_kwargs):
             raise Exception(f"INPUT_TYPES and function arguments must match in {cls.__name__}, input_types: {input_keys}, function arguments: {function_kwargs}")
         # if not exact match, print warning
         if len(set(input_keys)) != len(set(function_kwargs)):
-            print(f"Warning: INPUT_TYPES and function arguments don't match in {cls.__name__}, input_types: {input_keys}, function arguments: {function_kwargs}")
-            
+            #print(f"Warning: INPUT_TYPES and function arguments don't match in {cls.__name__}, input_types: {input_keys}, function arguments: {function_kwargs}")
+            pass
 # AllTrue class hijacks the isinstance, issubclass, bool, str, jsonserializable, eq, ne methods to always return True
 class AllTrue(str):
     def __init__(self, representation=None) -> None:

auxilary.py

@@ -1,6 +1,7 @@
 from .imgio.converter import PILHandlingHodes
 from .autonode import node_wrapper, get_node_names_mappings, validate, anytype, PILImage
 from .utils.tagger import get_tags, tagger_keys
+from PIL import Image, ImageFilter
 
 auxilary_classes = []
 auxilary_node = node_wrapper(auxilary_classes)
@@ -49,6 +50,129 @@ class GetRatingFromTextNode:
             }
         }
 
+def pixelate(image, pixelation_factor=0.1):
+    # Downscale the image
+    small = image.resize(
+        (int(image.width * pixelation_factor), int(image.height * pixelation_factor)),
+        resample=Image.NEAREST
+    )
+    # Upscale back to original size
+    return small.resize(image.size, Image.NEAREST)
+
+
+def pixelate_target_tiles(image, max_tiles=100):
+    aspect_ratio = image.width / image.height
+    # solve h * a * h = max_tiles
+    h = int((max_tiles / aspect_ratio) ** 0.5)
+    w = int(aspect_ratio * h)
+    h, w = max(4, h), max(4, w)
+    # Downscale the image
+    small = image.resize(
+        (w, h),
+        resample=Image.NEAREST,
+    )
+    # Upscale back to original size
+    return small.resize(image.size, Image.NEAREST)
+
+
+@auxilary_node
+class CensorImageByRating:
+    FUNCTION = "censor_image"
+    RETURN_TYPES = ("IMAGE",)
+    CATEGORY = "image"
+    custom_name = "Censor Image by Rating"
+
+    @staticmethod
+    @PILHandlingHodes.output_wrapper
+    def censor_image(image, rating_threshold, censor_method, model_name=None):
+        # Convert input to a PIL image
+        image = PILHandlingHodes.handle_input(image)
+        result_dict = get_tags(image, model_name=model_name)
+        rating = result_dict['rating']
+        # If rating is general, no censorship required
+        if rating.lower() == "general":
+            return (image,)
+        censor_image = False
+        if rating_threshold == "general":
+            # censor if not general
+            if rating.lower() != "general":
+                censor_image = True
+        elif rating_threshold == "sensitive":
+            # censor if not general or sensitive
+            if rating.lower() not in ["general", "sensitive"]:
+                censor_image = True
+        elif rating_threshold == "questionable":
+            # censor if not general, sensitive or questionable
+            if rating.lower() not in ["general", "sensitive", "questionable"]:
+                censor_image = True
+        elif rating_threshold == "explicit":
+            return (image,) # why are you using this?
+        if censor_image:
+            if censor_method.lower() == "white":
+                # Return a white image of the same size
+                censored_image = Image.new("RGB", image.size, (255, 255, 255))
+                return (censored_image,)
+
+            elif censor_method.lower() == "blur":
+                # Apply a strong blur (you can adjust radius as needed)
+                censored_image = image.filter(ImageFilter.GaussianBlur(radius=40))
+                return (censored_image,)
+            elif censor_method.lower() == "pixelate":
+                # first, blur
+                censored_image = image.filter(ImageFilter.GaussianBlur(radius=20))
+                censored_image = pixelate_target_tiles(censored_image, max_tiles=100)
+                return (censored_image,)
+
+        # If unknown method is provided, just return the original image
+        return (image,)
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "image": ("IMAGE",),
+                "rating_threshold": (["general", "sensitive", "questionable", "explicit"],),
+                "censor_method": (["blur", "white","pixelate"],),
+            },
+            "optional": {
+                "model_name": (tagger_keys, {"default": tagger_keys[0]}),
+            }
+        }
+
+@auxilary_node
+class FilterTagsNode:
+    """
+    Filters tags, given a list of tags splitted by ",".
+    We assume the input text is splittable by "," (or separator). Then, if any tags contain the filter, we remove matching tags.
+    """
+    FUNCTION = "filter_tags"
+    RETURN_TYPES = ("STRING",)
+    CATEGORY = "safety"
+    custom_name = "Filter Tags"
+    @staticmethod
+    def filter_tags(tags, filter_tags, separator):
+        filter_tags = filter_tags.split(",")
+        filter_tags = [tag.strip() for tag in filter_tags]
+        tags = tags.split(separator)
+        tags = [tag.strip() for tag in tags]
+        filtered = []
+        for tag in tags:
+            if all(filter_tag not in tag for filter_tag in filter_tags):
+                filtered.append(tag)
+        return (separator.join(filtered), )
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "tags": ("STRING",),
+                "filter_tags": ("STRING",),
+            },
+            # optional separator
+            "optional": {
+                "separator": ("STRING", {"default": ","}),
+            }
+        }
+
 @auxilary_node
 class GetTagsAboveThresholdNode:
     FUNCTION = "get_tags_above_threshold"
@@ -96,7 +220,7 @@ class GetTagsAboveThresholdFromTextNode:
                 "model_name": (tagger_keys, {"default": tagger_keys[0]}),
             }
         }
-        
+
 @auxilary_node
 class GetCharactersAboveThresholdNode:
     FUNCTION = "get_tags_above_threshold"
@@ -172,6 +296,31 @@ class GetAllTagsAboveThresholdNode:
                 "model_name": (tagger_keys, {"default": tagger_keys[0]}),
             }
         }
-
+@auxilary_node
+class GetAllTagsExceptCharacterAboveThresholdNode:
+    FUNCTION = "get_tags"
+    RETURN_TYPES = ("STRING",)
+    CATEGORY = "tagger"
+    custom_name = "Get All Tags Above Threshold Except Characters"
+    @staticmethod
+    def get_tags(image, threshold, replace, model_name):
+        image = PILHandlingHodes.handle_input(image)
+        result = get_tags(image, threshold=threshold, replace=replace, model_name=model_name)
+        result_list = []
+        result_list.append(result['rating'])
+        result_list.extend(result['tags'])
+        return (", ".join(result_list), )
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "image": ("IMAGE",),
+            },
+            "optional": {
+                "threshold": ("FLOAT", {"default": 0.4}),
+                "replace": ("BOOLEAN", {"default": False}),
+                "model_name": (tagger_keys, {"default": tagger_keys[0]}),
+            }
+        }
 CLASS_MAPPINGS, CLASS_NAMES = get_node_names_mappings(auxilary_classes)
-validate(auxilary_classes)
+validate(auxilary_classes)

conversion.py

@@ -9,8 +9,12 @@ node = node_wrapper(classes)
 conversion_operators = {
     "Int" : int,
     "Float" : float,
-    "Bool" : bool,
-    "String" : str
+    "Boolean" : bool,
+    "String" : str,
+    "Dict": dict,
+    "List": list,
+    "Tuple": tuple,
+    "Set": set,
 }
 def create_class(type_to):
     class_name = "ConvertAny2{}".format(type_to)
@@ -83,9 +87,9 @@ class ConvertComboToString:
     CATEGORY = "Logic Gates"
     custom_name = "Convert Combo to String"
     def convertComboToString(self, combo, separator):
-        if isinstance(combo, (str, float, int, bool)):
-            return (combo,)
-        return (separator.join(combo),)
+        if isinstance(combo, (list, tuple)):
+            return (separator.join(str(item) for item in combo),)
+        return (str(combo),)
 
 for type_to in conversion_operators:
     create_class(type_to)

crypto.py

@@ -0,0 +1,182 @@
+import os, io
+import numpy as np
+from .imgio.converter import PILHandlingHodes
+from .autonode import node_wrapper, get_node_names_mappings, validate
+
+from PIL import Image
+try:
+    from Crypto.PublicKey import RSA
+    from Crypto.Cipher import AES, PKCS1_OAEP
+    from Crypto.Random import get_random_bytes
+except ImportError:
+    print("Crypto library not found. Please install pycryptodome.")
+    raise
+
+import torch
+from base64 import b64encode, b64decode
+
+# List of classes to register
+secure_classes = []
+secure_node = node_wrapper(secure_classes)
+
+@secure_node
+class SecureBase64Encrypt:
+    """
+    Encrypt an image as a base64 string using RSA public key + AES.
+    - images: Only the first image is used.
+    - public_key_pem: RSA public key (PEM string).
+    Outputs: 'encrypted_base64' string that SecureWebPDecrypt can decrypt.
+    """
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "images": ("IMAGE",),
+                "public_key_pem": ("STRING", {"multiline": True, "default": ""}),
+            }
+        }
+
+    RETURN_TYPES = ("STRING",)
+    RETURN_NAMES = ("encrypted_base64",)
+    FUNCTION = "encrypted_base64"
+    CATEGORY = "image"
+    custom_name = "Secure Base64 Encrypt"
+    OUTPUT_NODE = True
+    RESULT_NODE = True
+
+    def encrypted_base64(self, images, public_key_pem):
+        # Check input
+        if images is None or not len(images) or images[0] is None:
+            raise ValueError("No image provided.")
+        # Load RSA public key
+        rsa_key = RSA.import_key(public_key_pem)
+        cipher_rsa = PKCS1_OAEP.new(rsa_key)
+
+        # Take first image (if there's a batch dimension, pick index [0])
+        img_tensor = images[0].clone().detach()
+        if img_tensor.ndim == 4:
+            img_tensor = img_tensor[0]
+
+        # Convert [0..1] float => [0..255] uint8
+        img_array = (255.0 * img_tensor.clamp(0, 1).cpu().numpy()).astype("uint8")
+
+        # If shape is (C,H,W), transpose to (H,W,C). If shape is (H,W,C), leave as is.
+        if img_array.ndim == 3:
+            # If the first dimension is small (1,3,4), interpret as channels-first
+            if img_array.shape[0] in [1,3,4] and img_array.shape[-1] not in [1,3,4]:
+                img_array = np.transpose(img_array, (1,2,0))
+
+        # The critical fix: ensure array is contiguous
+        img_array = np.ascontiguousarray(img_array)
+
+        # Create a PIL Image from array
+        pil_img = Image.fromarray(img_array, mode="RGB")
+
+        # Save in-memory as lossless WebP
+        buffer = io.BytesIO()
+        pil_img.save(buffer, format="WEBP", lossless=True)
+        image_bytes = buffer.getvalue()
+
+        # Generate a random AES session key, encrypt it with RSA
+        session_key = get_random_bytes(16)
+        enc_session_key = cipher_rsa.encrypt(session_key)
+
+        # Encrypt the image bytes with AES-EAX
+        cipher_aes = AES.new(session_key, AES.MODE_EAX)
+        ciphertext, tag = cipher_aes.encrypt_and_digest(image_bytes)
+
+        # Build custom envelope
+        encrypted_blob = (
+            b"ENCWEBP" +
+            len(enc_session_key).to_bytes(2, "big") +
+            enc_session_key +
+            bytes([len(cipher_aes.nonce)]) + cipher_aes.nonce +
+            bytes([len(tag)]) + tag +
+            ciphertext
+        )
+
+        # Base64-encode the blob
+        encrypted_base64 = b64encode(encrypted_blob).decode("utf-8")
+
+        return (encrypted_base64,)
+
+
+@secure_node
+class SecureWebPDecrypt:
+    """
+    Decrypt an encrypted WebP image (or list of them) produced by SecureBase64Encrypt.
+    Returns a single IMAGE (first one).
+    """
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "encrypted_base64": ("STRING", {"multiline": True, "default": ""}),
+                "private_key_pem": ("STRING", {"multiline": True, "default": ""}),
+            }
+        }
+
+    RETURN_TYPES = ("IMAGE",)
+    RETURN_NAMES = ("Decrypted_Image",)
+    FUNCTION = "decrypt_image"
+    CATEGORY = "image"
+    custom_name = "Secure WebP Decrypt"
+    @PILHandlingHodes.output_wrapper
+    def decrypt_image(self, encrypted_base64, private_key_pem):
+        # Convert to list if single string
+        if encrypted_base64 is None:
+            encrypted_base64 = []
+        if isinstance(encrypted_base64, str):
+            encrypted_base64 = [encrypted_base64]
+        elif not isinstance(encrypted_base64, (list, tuple)):
+            raise ValueError("encrypted_base64 must be string or list/tuple.")
+
+        # Import RSA private key
+        if isinstance(private_key_pem, str):
+            private_key_pem = private_key_pem.encode("utf-8")
+        rsa_key = RSA.import_key(private_key_pem)
+        cipher_rsa = PKCS1_OAEP.new(rsa_key)
+
+        for b64_item in encrypted_base64:
+            data = b64decode(b64_item)
+            if data[:7] != b"ENCWEBP":
+                raise ValueError("Invalid encrypted WebP data (missing header).")
+
+            idx = 7
+            enc_key_len = int.from_bytes(data[idx : idx + 2], "big")
+            idx += 2
+
+            enc_session_key = data[idx : idx + enc_key_len]
+            idx += enc_key_len
+
+            nonce_len = data[idx]
+            idx += 1
+            nonce = data[idx : idx + nonce_len]
+            idx += nonce_len
+
+            tag_len = data[idx]
+            idx += 1
+            tag = data[idx : idx + tag_len]
+            idx += tag_len
+
+            ciphertext = data[idx:]
+
+            # RSA-decrypt the AES session key
+            session_key = cipher_rsa.decrypt(enc_session_key)
+
+            # AES-EAX decrypt
+            cipher_aes = AES.new(session_key, AES.MODE_EAX, nonce=nonce)
+            plaintext = cipher_aes.decrypt(ciphertext)
+            try:
+                cipher_aes.verify(tag)
+            except ValueError:
+                raise ValueError("Decryption failed: data tampered or wrong key.")
+
+            # plaintext -> PIL
+            pil_img = Image.open(io.BytesIO(plaintext)).convert("RGB")
+            return (pil_img, )
+
+
+# Register node classes with ComfyUI
+CLASS_MAPPINGS, CLASS_NAMES = get_node_names_mappings(secure_classes)
+validate(secure_classes)

external.py

@@ -30,7 +30,7 @@ class SDWebuiAPINode:
                 "height": ("INT", {"default": 1024}),
                 "hr_scale": ("FLOAT", {"default": 1.5}),
                 "hr_upscale": ("STRING", {"default": "Latent"}),
-                "enable_hr": ("BOOL", {"default": False}),
+                "enable_hr": ("BOOLEAN", {"default": False}),
                 "cfg_scale": ("INT", {"default": 7}),
             }
         }
@@ -59,7 +59,7 @@ class SDWebuiAPIFallbackNode:
                 "height": ("INT", {"default": 1024}),
                 "hr_scale": ("FLOAT", {"default": 1.5}),
                 "hr_upscale": ("STRING", {"default": "Latent"}),
-                "enable_hr": ("BOOL", {"default": False}),
+                "enable_hr": ("BOOLEAN", {"default": False}),
                 "cfg_scale": ("INT", {"default": 7}),
             }
         }

imgio/converter.py

@@ -1,3 +1,4 @@
+from typing import Union, List
 from PIL import Image
 import numpy as np
 import base64
@@ -5,11 +6,144 @@ import torch
 import requests
 import os
 from io import BytesIO
+import gzip
+import re
+from urllib.parse import urlparse
+
+
+def handle_rgba_composite(
+    image: Image.Image, background_color=(255, 255, 255), as_rgba=False
+) -> Image.Image:
+    """
+    Convert RGBA image to RGB image using alpha_composite.
+    """
+    mode = image.mode
+    if as_rgba:
+        return image.convert("RGBA") # universal format
+    if mode == "RGB":
+        return image
+    if mode == "RGBA":
+        # Create a white RGBA background
+        background = Image.new("RGBA", image.size, (255, 255, 255, 255))
+        # Composite the original image over the white background
+        composed = Image.alpha_composite(background, image)
+        # Convert back to RGB (now that background is flattened)
+        return composed.convert("RGB")
+    elif mode == "LA":
+        # "LA" is 8-bit grayscale + alpha.
+        rgba_image = image.convert("RGBA")
+        background = Image.new("RGBA", rgba_image.size, (*background_color, 255))
+        composed = Image.alpha_composite(background, rgba_image)
+        return composed.convert("RGB")
+
+    # 3. "L" or "1" = Grayscale or Black/White, "P" = Palette
+    elif mode in ["L", "1", "P"]:
+        # Simply converting to "RGB" is usually enough.
+        return image.convert("RGB")
+
+    # 4. "CMYK", "YCbCr", "HSV", etc.
+    elif mode in ["CMYK", "YCbCr", "HSV"]:
+        # Typically, a .convert("RGB") is enough if you just need an RGB version.
+        return image.convert("RGB")
+    print(f"Warning: Unhandled image mode: {mode}. Converting to RGB.")
+    return image.convert("RGB")
+
+def fetch_image_securely(image_url: str,
+                        allowed_schemes=('http', 'https'),
+                        max_file_size=5_000_000,
+                        request_timeout=30):
+    """
+    Fetches an image from the given URL securely.
+
+    This function:
+    1. Validates the URL scheme (only http/https).
+    2. Blocks private IP/loopback addresses to prevent SSRF attacks.
+    3. Streams data to avoid excessive memory usage.
+    4. Checks MIME type, size limits, and optionally handles form-encoded image data.
+
+    :param image_url: URL of the image to retrieve (e.g., an S3-signed URL).
+    :param allowed_schemes: A tuple of allowed URL schemes (default: ('http', 'https')).
+    :param max_file_size: Max size (in bytes) of the file to download.
+    :param request_timeout: Timeout (in seconds) for the request.
+    :return: PIL Image object if successful, else raises an exception.
+    """
+
+    # -- 1. Validate scheme to avoid unexpected protocols  --
+    parsed = urlparse(image_url)
+    if parsed.scheme not in allowed_schemes:
+        raise ValueError(f"Invalid or disallowed URL scheme: {parsed.scheme}")
+
+    # -- 2. Prevent local network (SSRF) attacks by blocking private or loopback addresses  --
+    #    This is a simplified check. Consider using a library for robust IP parsing if needed.
+    ip_like_pattern = r'^(\d{1,3}\.){3}\d{1,3}$'
+    hostname = parsed.hostname
+    if (
+        hostname is None
+        or hostname.lower() in ("localhost", "127.0.0.1", "::1")
+        or (re.match(ip_like_pattern, hostname) and hostname.startswith("10."))
+        or hostname.startswith("192.168.")
+        or hostname.startswith("172.16.")
+        or hostname.startswith("172.17.")
+        or hostname.startswith("172.18.")
+        or hostname.startswith("172.19.")
+        or hostname.startswith("172.2")  # covers 172.20 - 172.31
+        or hostname.startswith("172.3")
+    ):
+        raise ValueError("URL resolves to a private or loopback address, which is disallowed.")
+
+    # -- 3. Retrieve the response with a timeout and stream  --
+    #    This handles the S3 URL just like any other public HTTPS link.
+    with requests.get(image_url, timeout=request_timeout, stream=True) as response:
+        response.raise_for_status()
+
+        # -- 4. Check Content-Type in headers  --
+        content_type = response.headers.get('Content-Type', '').lower()
+
+        # If it's a direct image...
+        if content_type.startswith("image/"):
+            # -- 5. Check Content-Length against max_file_size  --
+            content_length = response.headers.get('Content-Length')
+            if content_length and int(content_length) > max_file_size:
+                raise ValueError(
+                    f"File is too large: {int(content_length)} bytes. "
+                    f"Max allowed is {max_file_size} bytes."
+                )
+
+            data = BytesIO()
+            downloaded = 0
+            chunk_size = 8192
+            for chunk in response.iter_content(chunk_size=chunk_size):
+                downloaded += len(chunk)
+                if downloaded > max_file_size:
+                    raise ValueError(
+                        f"File exceeded the maximum allowed size of {max_file_size} bytes."
+                    )
+                data.write(chunk)
+
+            # Reset the buffer and open with PIL
+            data.seek(0)
+            return Image.open(data)
+
+        # If the server reports x-www-form-urlencoded, parse for embedded image data
+        elif content_type == "application/x-www-form-urlencoded":
+            # let PIL handle the parsing
+            try:
+                return Image.open(BytesIO(response.content))
+            except Exception as e:
+                raise ValueError(
+                    f"Failed to parse x-www-form-urlencoded data as image: {e}"
+                )
+
+        else:
+            # Some other content type we don't handle
+            raise ValueError(
+                f"Unsupported Content-Type or not an image: {content_type}"
+            )
 
 class IOConverter:
     """
     Classify the input data type.
-    
+
     Assumes the inputs to be following:
 
     - PIL Image
@@ -17,9 +151,10 @@ class IOConverter:
     - torch tensor
     - string (path to image)
     - base64 string (which can be decoded to bytes and then to image)
+    - gzip-compressed base64 string
     - URL (which can be downloaded to image)
-    
-    do NOT pass unsafe URLs / base64 strings, as it may cause security issues.
+
+    Do NOT pass unsafe URLs / base64 strings, as it may cause security issues.
     """
 
     class InputType:
@@ -28,8 +163,9 @@ class IOConverter:
         TORCH = "TORCH"
         STRING = "STRING"
         BASE64 = "BASE64"
+        GZIP_BASE64 = "GZIP_BASE64"
         URL = "URL"
-    
+
     def __init__(self):
         raise Exception("This class should not be instantiated.")
 
@@ -49,102 +185,209 @@ class IOConverter:
             elif input_data.startswith("http://") or input_data.startswith("https://"):
                 return IOConverter.InputType.URL
             else:
-                raise Exception(f"Invalid string input, {input_data:10}")
+                # Attempt to detect base64-encoded data
+                try:
+                    decoded_data = base64.b64decode(input_data, validate=True)
+                    # Check for gzip magic number
+                    if decoded_data[:2] == b'\x1f\x8b':
+                        return IOConverter.InputType.GZIP_BASE64
+                    else:
+                        return IOConverter.InputType.BASE64
+                except Exception:
+                    raise Exception(f"Invalid string input, cannot be decoded as base64.")
         else:
             raise Exception(f"Invalid input type, {type(input_data)}")
-    
+    @staticmethod
+    def match_dtype(array_or_tensor, is_tensor=False):
+        # if all value is between 0 and 1, multiply by 255 and convert to uint8
+        # however already uint8, skip
+        # check dtype first
+        if array_or_tensor.dtype == np.uint8 or array_or_tensor.dtype == torch.uint8:
+            return array_or_tensor
+
+        if array_or_tensor.min() >= 0 and array_or_tensor.max() <= 1:
+            multiplied = array_or_tensor * 255
+            if not is_tensor:
+                return multiplied.astype(np.uint8)
+            else:
+                return multiplied.to(torch.uint8)
+        return array_or_tensor
+
     @staticmethod
     def convert_to_pil(input_data):
         input_type = IOConverter.classify(input_data)
         if input_type == IOConverter.InputType.PIL:
-            return input_data
+            return handle_rgba_composite(input_data)
         elif input_type == IOConverter.InputType.NUMPY:
             # [1, 1216, 832, 3], '<f4'] -> [1216, 832, 3], 'uint8'
             # if not first element is 1, then it is a batch of images so warning
             if input_data.shape[0] != 1:
-                print("Warning: Batch of images detected, taking first image")
-            input_data = input_data[0] * 255.0 if input_data.dtype == np.float32 else input_data[0]
-            input_data = input_data.astype('uint8')
-            return Image.fromarray(input_data)
+                result = []
+                for i in range(input_data.shape[0]):
+                    np_array = IOConverter.match_dtype(input_data[i])
+                    result.append(handle_rgba_composite(Image.fromarray(np_array)))
+                return result # return list of PIL images
+            input_data = IOConverter.match_dtype(input_data[0])
+            return handle_rgba_composite(Image.fromarray(input_data))
         elif input_type == IOConverter.InputType.TORCH:
             # same as above
             if input_data.shape[0] != 1:
-                print("Warning: Batch of images detected, taking first image")
-            input_data = input_data[0].cpu().numpy() * 255.0
-            input_data = input_data.astype('uint8')
-            return Image.fromarray(input_data)
+                result = []
+                for i in range(input_data.shape[0]):
+                    np_array = (
+                        IOConverter.match_dtype(input_data[i], is_tensor=True)
+                        .cpu()
+                        .numpy()
+                    )
+                    result.append(handle_rgba_composite(Image.fromarray(np_array)))
+                return result
+            input_data = IOConverter.match_dtype(input_data[0], is_tensor=True)
+            np_array = input_data.cpu().numpy()
+            return handle_rgba_composite(Image.fromarray(np_array))
         elif input_type == IOConverter.InputType.STRING:
             return Image.open(input_data)
+        elif input_type == IOConverter.InputType.GZIP_BASE64:
+            decoded_data = IOConverter.read_base64(input_data)
+            decompressed_data = gzip.decompress(decoded_data)
+            partial_result = Image.open(BytesIO(decompressed_data))
+            result = handle_rgba_composite(partial_result)
+            return result
         elif input_type == IOConverter.InputType.BASE64:
-            return Image.open(BytesIO(base64.b64decode(input_data)))
+            decoded_data = IOConverter.read_base64(input_data)
+            partial_result = Image.open(BytesIO(decoded_data))
+            result = handle_rgba_composite(partial_result)
+            return result
         elif input_type == IOConverter.InputType.URL:
-            response = requests.get(input_data)
-            return Image.open(BytesIO(response.content))
+            partial_result = fetch_image_securely(input_data)
+            result = handle_rgba_composite(partial_result)
+            return result
         else:
             raise Exception(f"Invalid input type, {input_type}")
+
     @staticmethod
-    def to_tensor(pil_image):
+    def to_rgb_tensor(pil_image):
         if pil_image.mode == "I":
-            pil_image = pil_image.point(lambda i: i * (1/255)) # convert to float
+            pil_image = pil_image.point(lambda i: i * (1/255))  # convert to float
+        pil_image = handle_rgba_composite(pil_image)
         np_array = np.array(pil_image).astype(np.float32) / 255.0
-        image = torch.from_numpy(np_array)[None,][0]
-        image = image[None,] # to batch
-        return image
+        tensor = torch.from_numpy(np_array)
+        tensor = tensor.unsqueeze(0)  # Add batch dimension
+        # assert 4-dimensional tensor, B,C,H,W
+        if len(tensor.shape) != 4:
+            raise Exception(f"Invalid tensor shape, expected 4-dimensional tensor, got {tensor.shape}")
+        return tensor
+
     @staticmethod
-    def convert_to_tensor(input_data):
+    def to_rgba_tensor(pil_image):
+        if pil_image.mode == "I":
+            pil_image = pil_image.point(lambda i: i * (1/255))  # convert to float
+        pil_image = handle_rgba_composite(pil_image, as_rgba=True)
+        np_array = np.array(pil_image).astype(np.float32) / 255.0
+        tensor = torch.from_numpy(np_array)
+        tensor = tensor.unsqueeze(0)  # Add batch dimension
+        # assert 4-dimensional tensor, B,C,H,W
+        if len(tensor.shape) != 4:
+            raise Exception(f"Invalid tensor shape, expected 4-dimensional tensor, got {tensor.shape}")
+        return tensor
+
+    @staticmethod
+    def read_base64(base64_string: str) -> bytes:
+        return base64.b64decode(base64_string)
+
+    @staticmethod
+    def read_maybe_gzip_base64(base64_string: str) -> bytes:
+        decoded_data = base64.b64decode(base64_string)
+        if decoded_data[:2] == b'\x1f\x8b':
+            result = gzip.decompress(decoded_data)
+        else:
+            result = decoded_data
+        # to string
+        return result.decode('utf-8')
+
+    @staticmethod
+    def convert_to_rgb_tensor(input_data, rgba=False):
+        if not rgba:
+            output_func = IOConverter.to_rgb_tensor
+        else:
+            output_func = IOConverter.to_rgba_tensor
         input_type = IOConverter.classify(input_data)
         if input_type == IOConverter.InputType.PIL:
-            return IOConverter.to_tensor(input_data)
+            return output_func(input_data)
         elif input_type == IOConverter.InputType.NUMPY:
-            return torch.from_numpy(input_data)
+            # if all values are 0~1, skip
+            if input_data.min() >= 0 and input_data.max() <= 1:
+                np_array = input_data.astype(np.float32)
+            else:
+                np_array = input_data.astype(np.float32) / 255.0
+            tensor = torch.from_numpy(np_array)
+            tensor = tensor.unsqueeze(0)  # Add batch dimension
+            return tensor
         elif input_type == IOConverter.InputType.TORCH:
             return input_data
         elif input_type == IOConverter.InputType.STRING:
             image = Image.open(input_data)
-            return IOConverter.to_tensor(image)
+            return output_func(image)
+        elif input_type == IOConverter.InputType.GZIP_BASE64:
+            image = IOConverter.convert_to_pil(input_data)
+            return output_func(image)
         elif input_type == IOConverter.InputType.BASE64:
-            image = Image.open(BytesIO(base64.b64decode(input_data)))
-            return IOConverter.to_tensor(image)
+            image = IOConverter.convert_to_pil(input_data)
+            return output_func(image)
         elif input_type == IOConverter.InputType.URL:
-            response = requests.get(input_data)
-            response.raise_for_status()
-            image = Image.open(BytesIO(response.content))
-            return IOConverter.to_tensor(image)
-        else:
-            raise Exception(f"Invalid input type, {input_type}")
-    
-    @staticmethod
-    def convert_to_base64(input_data):
-        input_type = IOConverter.classify(input_data)
-        if input_type == IOConverter.InputType.PIL:
-            buffered = BytesIO()
-            input_data.save(buffered, format="PNG")
-            return base64.b64encode(buffered.getvalue()).decode("utf-8")
-        elif input_type == IOConverter.InputType.NUMPY:
-            return IOConverter.convert_to_base64(Image.fromarray(input_data))
-        elif input_type == IOConverter.InputType.TORCH:
-            return IOConverter.convert_to_base64(Image.fromarray(input_data.cpu().numpy()))
-        elif input_type == IOConverter.InputType.STRING:
-            return IOConverter.convert_to_base64(Image.open(input_data))
-        elif input_type == IOConverter.InputType.BASE64:
-            return input_data
-        elif input_type == IOConverter.InputType.URL:
-            response = requests.get(input_data)
-            return base64.b64encode(response.content).decode("utf-8")
+            image = fetch_image_securely(input_data)
+            return output_func(image)
         else:
             raise Exception(f"Invalid input type, {input_type}")
 
+    @staticmethod
+    def convert_to_base64(input_data, format="PNG", quality=100, gzip_compress=False):
+        pil_image = IOConverter.convert_to_pil(input_data)
+        buffered = BytesIO()
+        save_params = {'format': format}
+        if format.upper() in ['JPEG', 'JPG']:
+            save_params['quality'] = quality
+        pil_image.save(buffered, **save_params)
+        buffered.seek(0)
+        if gzip_compress:
+            compressed_buffer = BytesIO()
+            with gzip.GzipFile(fileobj=compressed_buffer, mode='wb') as f:
+                f.write(buffered.getvalue())
+            compressed_buffer.seek(0)
+            base64_data = base64.b64encode(compressed_buffer.getvalue()).decode('utf-8')
+        else:
+            base64_data = base64.b64encode(buffered.getvalue()).decode('utf-8')
+        return base64_data
+
+    @staticmethod
+    def string_to_base64(input_string, gzip_compress=False):
+        if gzip_compress:
+            compressed_buffer = BytesIO()
+            with gzip.GzipFile(fileobj=compressed_buffer, mode='wb') as f:
+                f.write(input_string.encode())
+            compressed_buffer.seek(0)
+            base64_data = base64.b64encode(compressed_buffer.getvalue()).decode('utf-8')
+        else:
+            base64_data = base64.b64encode(input_string.encode()).decode('utf-8')
+        return base64_data
+
 class PILHandlingHodes:
     @staticmethod
-    def handle_input(tensor_or_image):
+    def handle_input(tensor_or_image) -> Union[Image.Image, List[Image.Image]]:
         pil_image = IOConverter.convert_to_pil(tensor_or_image)
         return pil_image
+
     @staticmethod
-    def handle_output_as_pil(pil_image):
+    def handle_output_as_pil(pil_image: Image.Image) -> Image.Image:
         return pil_image
+
     @staticmethod
-    def handle_output_as_tensor(pil_image):
-        return IOConverter.convert_to_tensor(pil_image)
+    def handle_output_as_tensor(pil_image: Image.Image, rgba=False) -> torch.Tensor:
+        return IOConverter.convert_to_rgb_tensor(pil_image, rgba=rgba)
+
+    @staticmethod
+    def handle_output_as_rgba_tensor(pil_image: Image.Image) -> torch.Tensor:
+        return IOConverter.convert_to_rgb_tensor(pil_image, rgba=True)
+
     @staticmethod
     def output_wrapper(func):
         def wrapped(*args, **kwargs):
@@ -158,3 +401,29 @@ class PILHandlingHodes:
             return tuple(tuples_collect)
         return wrapped
 
+    @staticmethod
+    def rgba_output_wrapper(func):
+        def wrapped(*args, **kwargs):
+            outputs = func(*args, **kwargs)
+            tuples_collect = []
+            for output in outputs:
+                if isinstance(output, (Image.Image, torch.Tensor)):
+                    tuples_collect.append(PILHandlingHodes.handle_output_as_rgba_tensor(output))
+                else:
+                    tuples_collect.append(output)
+            return tuple(tuples_collect)
+        return wrapped
+
+    @staticmethod
+    def to_base64(anything, quality=100, format="PNG", gzip_compress=False):
+        base64_data = IOConverter.convert_to_base64(anything, format=format, quality=quality, gzip_compress=gzip_compress)
+        return base64_data
+
+    @staticmethod
+    def string_to_base64(input_string, gzip_compress=False):
+        base64_data = IOConverter.string_to_base64(input_string, gzip_compress=gzip_compress)
+        return base64_data
+
+    @staticmethod
+    def maybe_gzip_base64_to_string(base64_string):
+        return IOConverter.read_maybe_gzip_base64(base64_string)

install.py

@@ -1,4 +1,5 @@
 #https://github.com/ltdrdata/ComfyUI-Impact-Pack/blob/Main/install.py
+import os
 import sys
 import subprocess
 import threading
@@ -31,30 +32,71 @@ def process_wrap(cmd_str, cwd=None, handler=None):
     return process.wait()
 
 if "python_embeded" in sys.executable or "python_embedded" in sys.executable: #standalone python version
-    pip_install = [sys.executable, '-s', '-m', 'pip', 'install', "-U"]
+    pip_install = [sys.executable, '-s', '-m', 'pip', 'install', "-U", "--default-timeout=1000"]
 else:
-    pip_install = [sys.executable, '-m', 'pip', 'install', "-U"]
+    pip_install = [sys.executable, '-m', 'pip', 'install', "-U", "--default-timeout=1000"]
+
+def _imgutils_install_candidates() -> list[str]:
+    """
+    Decide which dghs-imgutils package variant to install first.
+
+    - `dghs-imgutils[gpu]` includes optional GPU-related dependencies.
+    - If GPU detection fails, default to CPU variant first.
+    """
+    variant = os.environ.get("COMFYUI_LOGICUTILS_IMGUTILS_VARIANT", "").strip().lower()
+    if variant in {"gpu", "cuda"}:
+        return ["dghs-imgutils[gpu]", "dghs-imgutils"]
+    if variant in {"cpu"}:
+        return ["dghs-imgutils", "dghs-imgutils[gpu]"]
+
+    # auto
+    try:
+        import torch
+
+        if getattr(torch, "cuda", None) is not None and torch.cuda.is_available():
+            return ["dghs-imgutils[gpu]", "dghs-imgutils"]
+    except Exception:
+        pass
+    return ["dghs-imgutils", "dghs-imgutils[gpu]"]
 
 def initialization():
     try:
         import piexif
-    except ImportError:
+    except Exception:
+        print("piexif not found, installing...")
         run_installation("piexif")
     try:
         import chardet
-    except ImportError:
+    except Exception:
+        print("chardet not found, installing...")
         run_installation("chardet")
     try:
         from imgutils.tagging import get_wd14_tags
-    except ImportError:
-        run_installation("dghs-imgutils[gpu]")
+    except Exception:
+        print("imgutils not found, installing...")
+        for candidate in _imgutils_install_candidates():
+            print(f"Trying to install {candidate}...")
+            run_installation(candidate)
+            try:
+                from imgutils.tagging import get_wd14_tags  # noqa: F401
+                break
+            except Exception:
+                continue
+    try:
+        from Crypto.PublicKey import RSA
+    except Exception:
+        print("pycryptodome not found, installing...")
+        run_installation("pycryptodome")
 
 def run_installation(pkg_name: str):
     print(f"Installing {pkg_name}...")
-    if process_wrap(pip_install + [pkg_name]) == 0:
-        print(f"Successfully installed {pkg_name}")
-    else:
-        print(f"Failed to install {pkg_name}")
+    try:
+        if process_wrap(pip_install + [pkg_name]) == 0:
+            print(f"Successfully installed {pkg_name}")
+        else:
+            print(f"Failed to install {pkg_name}")
+    except Exception as e:
+        print(f"Failed to install {pkg_name}: {e}")
 
 if __name__ == "__main__":
     initialization()

logic_gates.py

@@ -8,25 +8,6 @@ from .autonode import node_wrapper, get_node_names_mappings, validate, anytype
 classes = []
 node = node_wrapper(classes)
 
-@node
-class LogicGateCompare:
-    """
-    Returns 1 if input1 > input2, 0 otherwise
-    """
-    RETURN_TYPES = ("BOOL",)
-    @classmethod
-    def INPUT_TYPES(s):
-        return {
-        "required": {
-            "input1": (anytype, {"default": 0.0}),
-            "input2": (anytype, {"default": 0.0}),
-        }
-    }
-    FUNCTION = "compareFloat"
-    CATEGORY = "Logic Gates"
-    custom_name = "ABiggerThanB"
-    def compareFloat(self, input1, input2):
-        return (True if input1 > input2 else False,)
 @node
 class LogicGateInvertBasic:
     """
@@ -86,7 +67,9 @@ class LogicGateBitwiseShift:
         # validate input2
         if abs(input2) > 32:
             raise ValueError("input2 must be between -32 and 32")
-        return (input1 << input2,)
+        if input2 >= 0:
+            return (input1 << input2,)
+        return (input1 >> abs(input2),)
 @node
 class LogicGateBitwiseAnd:
     """
@@ -171,7 +154,7 @@ class LogicGateCompare:
     """
     Returns 1 if input1 > input2, 0 otherwise
     """
-    RETURN_TYPES = ("BOOL",)
+    RETURN_TYPES = ("BOOLEAN",)
     @classmethod
     def INPUT_TYPES(s):
         return {
@@ -190,7 +173,7 @@ class LogicGateCompareString:
     """
     Returns if given regex (1) is found in given string (2)
     """
-    RETURN_TYPES = ("BOOL",)
+    RETURN_TYPES = ("BOOLEAN",)
     @classmethod
     def INPUT_TYPES(s):
         return {
@@ -204,6 +187,26 @@ class LogicGateCompareString:
     custom_name = "AContainsB(String)"
     def compareString(self, regex, input2):
         return (True if re.search(regex, input2) else False,)
+
+@node
+class GetLengthString:
+    """
+    Returns the length of the input string
+    """
+    RETURN_TYPES = ("INT",)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+        "required": {
+            "string": ("STRING", {"default": ""}),
+        }
+    }
+    FUNCTION = "lengthString"
+    CATEGORY = "Logic Gates"
+    custom_name = "Length of String"
+    def lengthString(self, string):
+        return (len(string),)
+
 @node
 class StaticNumberInt:
     """
@@ -263,7 +266,7 @@ class LogicGateAnd:
     """
     Returns 1 if all inputs are True, 0 otherwise
     """
-    RETURN_TYPES = ("BOOL",)
+    RETURN_TYPES = ("BOOLEAN",)
     @classmethod
     def INPUT_TYPES(s):
         return {
@@ -282,7 +285,7 @@ class LogicGateOr:
     """
     Returns 1 if any input is True, 0 otherwise
     """
-    RETURN_TYPES = ("BOOL",)
+    RETURN_TYPES = ("BOOLEAN",)
     @classmethod
     def INPUT_TYPES(s):
         return {
@@ -295,7 +298,7 @@ class LogicGateOr:
     CATEGORY = "Logic Gates"
     custom_name = "AOrBGate"
     def or_(self, input1, input2):
-        return (True if input1 and input2 else False,)
+        return (True if input1 or input2 else False,)
 @node
 class LogicGateEither:
     """
@@ -365,9 +368,9 @@ class ReplaceString:
     def INPUT_TYPES(s):
         return {
         "required": {
-            "String": ("STRING", {"default": ""}),
-            "Regex": ("STRING", {"default": ""}),
-            "ReplaceWith": ("STRING", {"default": ""}),
+            "String": ("STRING", {"default": ""}), # input string
+            "Regex": ("STRING", {"default": ""}), # regex to search for
+            "ReplaceWith": ("STRING", {"default": ""}), # string to replace with
         }
     }
     FUNCTION = "replace"
@@ -402,7 +405,5 @@ class MemoryNode:
             self.memory_value = input1
         return (self.memory_value,)
 
-
-
 CLASS_MAPPINGS, CLASS_NAMES = get_node_names_mappings(classes)
 validate(classes)

math_nodes.py

@@ -1,3 +1,4 @@
+import random
 from .autonode import node_wrapper, get_node_names_mappings, validate, anytype
 classes = []
 node = node_wrapper(classes)
@@ -137,12 +138,162 @@ class PowerNode:
     CATEGORY = "Math"
     custom_name = "Power"
     def power(self, input1, power):
-        # validate power with log scale, prevent overflow
-        log_val = math.log(abs(input1), 10)
-        if log_val * power > 100 or log_val == 0:
+        abs_input = abs(input1)
+        # fast paths for values that won't overflow digit-wise
+        if abs_input == 0:
+            if power < 0:
+                raise ZeroDivisionError("0 cannot be raised to a negative power")
+            return (math.pow(input1, power),)
+        if abs_input == 1:
+            return (math.pow(input1, power),)
+
+        # validate power with log10 scale, prevent huge magnitudes
+        log10_abs = math.log10(abs_input)
+        if (log10_abs * power) > 100:
             raise OverflowError("Power is too large, exceeds 100 digits")
         return (math.pow(input1, power),)
 
+@node
+class SigmoidNode:
+    """
+    Returns the sigmoid of a number
+    """
+    RETURN_TYPES = ("FLOAT",)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+        "required": {
+            "input1": ("FLOAT",),
+        }
+    }
+    FUNCTION = "sigmoid"
+    CATEGORY = "Math"
+    custom_name = "Sigmoid"
+    def sigmoid(self, input1):
+        return (1 / (1 + math.exp(-input1)),)
+def is_prime_small(n: int) -> bool:
+    """
+    Deterministic check for primality for smaller n.
+    Skips multiples of 2 and 3, then checks i, i+2, i+4 up to sqrt(n).
+    """
+    if n < 2:
+        return False
+    if n in (2, 3):
+        return True
+    if n % 2 == 0 or n % 3 == 0:
+        return n == 2 or n == 3
+    
+    # 6k ± 1 optimization
+    limit = int(math.isqrt(n))  # integer sqrt
+    i = 5
+    while i <= limit:
+        if n % i == 0 or n % (i + 2) == 0:
+            return False
+        i += 6
+    return True
+
+def miller_rabin_test(d: int, n: int) -> bool:
+    """ One round of the Miller-Rabin test with a random base 'a'. """
+    a = random.randrange(2, n - 1)
+    x = pow(a, d, n)  # a^d % n
+    if x == 1 or x == n - 1:
+        return True
+    
+    # Keep squaring x while d does not reach n-1
+    while d != n - 1:
+        x = (x * x) % n
+        d <<= 1  # d *= 2
+        if x == 1:
+            return False
+        if x == n - 1:
+            return True
+    return False
+
+def is_prime_miller_rabin(n: int, k: int = 5) -> bool:
+    """
+    Miller-Rabin primality test with k rounds (probabilistic).
+    Good enough for big integers in practice.
+    """
+    # Handle small or trivial cases
+    if n < 2:
+        return False
+    # check small primes quickly
+    for small_prime in [2, 3, 5, 7, 11, 13, 17, 19, 23, 29]:
+        if n == small_prime:
+            return True
+        if n % small_prime == 0 and n != small_prime:
+            return False
+
+    # Write n - 1 as d * 2^r
+    d = n - 1
+    while d % 2 == 0:
+        d //= 2
+
+    # Witness loop
+    for _ in range(k):
+        if not miller_rabin_test(d, n):
+            return False
+    return True
+
+@node
+class IsPrimeNode:
+    """
+    Checks if an integer is prime.
+
+    - If the integer |value| < threshold, uses a deterministic small-check (trial division).
+    - Otherwise, uses a Miller-Rabin pseudoprime test for a faster check (probabilistic).
+
+    Returns a BOOLEAN (True if prime, False if composite).
+    """
+    FUNCTION = "is_prime"
+    RETURN_TYPES = ("BOOLEAN",)
+    CATEGORY = "Math"
+    custom_name = "Is Prime?"
+
+    @staticmethod
+    def is_prime(value: int, threshold: int = 10_000_000, miller_rabin_rounds: int = 5):
+        # handle negative or zero
+        if value < 2:
+            return (False,)
+
+        if value < threshold:
+            # use small prime check
+            return (is_prime_small(value),)
+        else:
+            # use Miller-Rabin
+            return (is_prime_miller_rabin(value, k=miller_rabin_rounds),)
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "value": ("INT", {"default": 1, "min": -9999999999, "max": 9999999999, "step": 1}),
+            },
+            "optional": {
+                "threshold": ("INT", {"default": 10_000_000, "min": 1, "max": 9999999999, "step": 1}),
+                "miller_rabin_rounds": ("INT", {"default": 5, "min": 1, "max": 50, "step": 1}),
+            }
+        }
+@node
+class RAMPNode:
+    """
+    Returns the ramp of a number
+    """
+    RETURN_TYPES = ("FLOAT",)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+        "required": {
+            "input1": ("FLOAT",),
+        }
+    }
+    FUNCTION = "ramp"
+    CATEGORY = "Math"
+    custom_name = "RAMP"
+    def ramp(self, input1):
+        return (max(0, input1),)
+
+@node
 class ModuloNode:
     """
     Returns the modulo of a number

nodes.py

@@ -1,33 +1,75 @@
-
-from .install import initialization
-
-initialization()
-
-from .logic_gates import CLASS_MAPPINGS as LogicMapping, CLASS_NAMES as LogicNames
-from .randomness import CLASS_MAPPINGS as RandomMapping, CLASS_NAMES as RandomNames
-from .conversion import CLASS_MAPPINGS as ConversionMapping, CLASS_NAMES as ConversionNames
-from .math_nodes import CLASS_MAPPINGS as MathMapping, CLASS_NAMES as MathNames
-from .io_node import CLASS_MAPPINGS as IOMapping, CLASS_NAMES as IONames
-from .auxilary import CLASS_MAPPINGS as AuxilaryMapping, CLASS_NAMES as AuxilaryNames
-from .external import CLASS_MAPPINGS as ExternalMapping, CLASS_NAMES as ExternalNames
-
-NODE_CLASS_MAPPINGS = {
-}
-NODE_CLASS_MAPPINGS.update(IOMapping)
-NODE_CLASS_MAPPINGS.update(LogicMapping)
-NODE_CLASS_MAPPINGS.update(RandomMapping)
-NODE_CLASS_MAPPINGS.update(ConversionMapping)
-NODE_CLASS_MAPPINGS.update(MathMapping)
-NODE_CLASS_MAPPINGS.update(ExternalMapping)
-NODE_CLASS_MAPPINGS.update(AuxilaryMapping)
-
-NODE_DISPLAY_NAME_MAPPINGS = {
-
-}
-NODE_DISPLAY_NAME_MAPPINGS.update(IONames)
-NODE_DISPLAY_NAME_MAPPINGS.update(LogicNames)
-NODE_DISPLAY_NAME_MAPPINGS.update(RandomNames)
-NODE_DISPLAY_NAME_MAPPINGS.update(ConversionNames)
-NODE_DISPLAY_NAME_MAPPINGS.update(MathNames)
-NODE_DISPLAY_NAME_MAPPINGS.update(ExternalNames)
-NODE_DISPLAY_NAME_MAPPINGS.update(AuxilaryNames)
+import os
+
+from .install import initialization
+
+
+def _running_in_comfyui() -> bool:
+    try:
+        import folder_paths  # noqa: F401
+    except Exception:
+        return False
+    return True
+
+
+_IN_COMFYUI = _running_in_comfyui()
+_SKIP_INSTALL = os.environ.get("COMFYUI_LOGICUTILS_SKIP_INSTALL", "").strip().lower() in {
+    "1",
+    "true",
+    "yes",
+}
+
+if _IN_COMFYUI and not _SKIP_INSTALL:
+    initialization()
+else:
+    print("Skipping ComfyUI-LogicUtils installation.")
+
+from .logic_gates import CLASS_MAPPINGS as LogicMapping, CLASS_NAMES as LogicNames
+from .randomness import CLASS_MAPPINGS as RandomMapping, CLASS_NAMES as RandomNames
+from .conversion import CLASS_MAPPINGS as ConversionMapping, CLASS_NAMES as ConversionNames
+from .math_nodes import CLASS_MAPPINGS as MathMapping, CLASS_NAMES as MathNames
+if _IN_COMFYUI:
+    from .io_node import CLASS_MAPPINGS as IOMapping, CLASS_NAMES as IONames
+else:
+    IOMapping = {}
+    IONames = {}
+from .auxilary import CLASS_MAPPINGS as AuxilaryMapping, CLASS_NAMES as AuxilaryNames
+from .external import CLASS_MAPPINGS as ExternalMapping, CLASS_NAMES as ExternalNames
+
+
+
+NODE_CLASS_MAPPINGS = {
+}
+NODE_CLASS_MAPPINGS.update(IOMapping)
+NODE_CLASS_MAPPINGS.update(LogicMapping)
+NODE_CLASS_MAPPINGS.update(RandomMapping)
+NODE_CLASS_MAPPINGS.update(ConversionMapping)
+NODE_CLASS_MAPPINGS.update(MathMapping)
+NODE_CLASS_MAPPINGS.update(ExternalMapping)
+NODE_CLASS_MAPPINGS.update(AuxilaryMapping)
+
+
+
+NODE_DISPLAY_NAME_MAPPINGS = {
+
+}
+NODE_DISPLAY_NAME_MAPPINGS.update(IONames)
+NODE_DISPLAY_NAME_MAPPINGS.update(LogicNames)
+NODE_DISPLAY_NAME_MAPPINGS.update(RandomNames)
+NODE_DISPLAY_NAME_MAPPINGS.update(ConversionNames)
+NODE_DISPLAY_NAME_MAPPINGS.update(MathNames)
+NODE_DISPLAY_NAME_MAPPINGS.update(ExternalNames)
+NODE_DISPLAY_NAME_MAPPINGS.update(AuxilaryNames)
+
+
+try:
+    from .pystructure import CLASS_MAPPINGS as PyStructureMapping, CLASS_NAMES as PyStructureNames
+    NODE_CLASS_MAPPINGS.update(PyStructureMapping)
+    NODE_DISPLAY_NAME_MAPPINGS.update(PyStructureNames)
+except Exception:
+    pass
+try:
+    from .crypto import CLASS_MAPPINGS as SecureMapping, CLASS_NAMES as SecureNames
+    NODE_CLASS_MAPPINGS.update(SecureMapping)
+    NODE_DISPLAY_NAME_MAPPINGS.update(SecureNames)
+except Exception:
+    pass

pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "comfyui-logicutils"
 description = "Logical Utils (compare, string, boolean operations) for ComfyUI"
-version = "1.2.0"
+version = "1.7.3"
 license = "MIT"
 
 [project.urls]
@@ -12,3 +12,4 @@ Repository = "https://github.com/aria1th/ComfyUI-LogicUtils"
 PublisherId = "angelbottomless"
 DisplayName = "ComfyUI-LogicUtils"
 Icon = ""
+

randomness.py

@@ -1,11 +1,172 @@
+import math
 import random
+import uuid
+import time
 from .autonode import node_wrapper, get_node_names_mappings, validate
 
+
 classes = []
 node = node_wrapper(classes)
 
+class RandomGuaranteedClass:
+    OUTPUT_NODE = True
+    RESULT_NODE = True
+    @classmethod
+    def IS_CHANGED(s, *args, **kwargs):
+       return float("NaN")
+    
 @node
-class UniformRandomFloat:
+class SystemRandomFloat(RandomGuaranteedClass):
+    """
+    Random number generator using system randomness
+    """
+    def __init__(self):
+        pass
+    @staticmethod
+    def generate(min_val=0.0, max_val=1.0, precision=0):
+        instance = random.SystemRandom(time.time())
+        value = instance.uniform(min_val, max_val)
+        if precision > 0:
+            value = round(value, precision)
+        return (value,)
+    RETURN_TYPES = ("FLOAT",)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "min_val": ("FLOAT", { "default": 0.0, "min": -999999999, "max": 999999999.0, "step": 0.01, "display": "number" }),
+                "max_val": ("FLOAT", { "default": 1.0, "min": -999999999, "max": 999999999.0, "step": 0.01, "display": "number" }),
+                "precision": ("INT", { "default": 0, "min": 0, "max": 10, "step": 1, "display": "number" }),
+            },
+        }
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "System Random Float"
+
+@node
+class DimensionSelectorWithSeedNode:
+    """
+    Finds (width, height) such that width*height is near (resolution^2),
+    ratio = width/height is in [min_ratio, max_ratio],
+    both are multiples of 'multiples', and uses 'seed' for random tie-break.
+    """
+    RETURN_TYPES = ("INT", "INT")
+    
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "resolution": ("INT", {"default": 1024}),
+                "min_ratio": ("FLOAT", {"default": 0.6}),
+                "max_ratio": ("FLOAT", {"default": 1.6}),
+                "multiples": ("INT", {"default": 32}),
+                "seed": ("INT", {"default": 0}),
+            }
+        }
+
+    FUNCTION = "select_dimensions"
+    CATEGORY = "Logic Gates"
+    custom_name = "Random Width/Height with Resolution"
+
+    def select_dimensions(self, resolution, min_ratio, max_ratio, multiples, seed):
+        # For reproducible randomness
+        random.seed(seed)
+
+        desired_area = resolution * resolution
+        ratio = random.uniform(min_ratio, max_ratio)
+        # width * height = resolution^2, width/height = ratio
+        # thus h**2 = resolution^2 / ratio
+        height = int(math.sqrt(desired_area / ratio))
+        width = int(desired_area / height)
+        # round to nearest multiple
+        div_h = height / multiples
+        div_w = width / multiples
+        height = round(div_h) * multiples
+        width = round(div_w) * multiples
+        # if width * height > resolution^2, reduce width or height
+        if width * height > desired_area:
+            if random.choice([True, False]):
+                width = width - multiples
+            else:
+                height = height - multiples
+        return (width, height)
+        
+
+@node
+class SystemRandomInt(RandomGuaranteedClass):
+    """
+    Random number generator using system randomness
+    Generates an integer value between 0 and 2^32-1
+    """
+
+    def __init__(self):
+        pass
+    @staticmethod
+    def generate(min_val=0, max_val=2**63-1):
+        instance = random.SystemRandom(time.time())
+        value = instance.randint(min_val, max_val)
+        return (value,)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "min_val": (
+                    "INT",
+                    {
+                        "default": 0,
+                        "min": -(2**63-1),
+                        "max": 2**63-1,
+                        "step": 1,
+                        "display": "number",
+                    },
+                ),
+                "max_val": (
+                    "INT",
+                    {
+                        "default": 2**63-1,
+                        "min": -(2**63-1),
+                        "max": 2**63-1,
+                        "step": 1,
+                        "display": "number",
+                    },
+                ),
+            },
+        }
+    RETURN_TYPES = ("INT",)
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "System Random Int"
+
+
+@node
+class SystemUUIDGenerator(RandomGuaranteedClass):
+    """
+    Generates a random UUID
+    """
+    def __init__(self):
+        pass
+    @staticmethod
+    def generate(length=36):
+        value = uuid.uuid4()
+        value = str(value)
+        if length < 36:
+            value = value[:length]
+        return (value,)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "length": ("INT", { "default": 36, "min": 1, "max": 36, "step": 1, "display": "number" }),
+            },
+        }
+    RETURN_TYPES = ("STRING",)
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "UUID Generator"
+
+
+@node
+class UniformRandomFloat(RandomGuaranteedClass):
     """
     Selects a random float from min to max
     Fallbacks to default if min is greater than max
@@ -14,7 +175,7 @@ class UniformRandomFloat:
         pass
     def generate(self, min_val, max_val, decimal_places, seed=0):
         if min_val > max_val:
-            return min_val
+            return (min_val,)
         instance = random.Random(seed)
         value = instance.uniform(min_val, max_val)
         # prune to decimal places - 0 = int, 1 = 1 decimal place,...
@@ -28,15 +189,186 @@ class UniformRandomFloat:
                 "min_val": ("FLOAT", { "default": 0.0, "min": -999999999, "max": 999999999.0, "step": 0.02, "display": "number" }),
                 "max_val": ("FLOAT", { "default": 1.0, "min": -999999999, "max": 999999999.0, "step": 0.02, "display": "number" }),
                 "decimal_places": ("INT", { "default": 1, "min": 0, "max": 10, "step": 1, "display": "number" }),
-                "seed" : ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "seed" : ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("FLOAT",)
     FUNCTION = "generate"
     CATEGORY = "Logic Gates"
     custom_name = "Uniform Random Float"
+
 @node
-class UniformRandomInt:
+class TriangularRandomFloat(RandomGuaranteedClass):
+    """
+    Selects a random float from min to max
+    Fallbacks to default if min is greater than max
+    """
+    def __init__(self):
+        pass
+    def generate(self, low, high, mode, seed=0):
+        if low > high:
+            return (low,)
+        instance = random.Random(seed)
+        value = instance.triangular(low, high, mode)
+        return (value,)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "low": ("FLOAT", { "default": 0.0, "min": -999999999, "max": 999999999.0, "step": 0.02, "display": "number" }),
+                "high": ("FLOAT", { "default": 1.0, "min": -999999999, "max": 999999999.0, "step": 0.02, "display": "number" }),
+                "mode": ("FLOAT", { "default": 0.5, "min": -999999999, "max": 999999999.0, "step": 0.02, "display": "number" }),
+                "seed" : ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
+            },
+        }
+    RETURN_TYPES = ("FLOAT",)
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "Triangular Random Float"
+
+@node
+class WeightedRandomChoice(RandomGuaranteedClass):
+    """
+    Randomly choose one item from a list with weights.
+    The input string is parsed as "value|weight$value2|weight2..."
+    Example: "apple|10$banana|1$orange|3"
+    """
+    def __init__(self):
+        pass
+
+    def generate(self, input_string, separator, seed=0):
+        # Example input: "apple|10$banana|1$orange|3"
+        # Split by '$' -> ["apple|10", "banana|1", "orange|3"]
+        items = input_string.split(separator)
+        choices = []
+        weights = []
+        for item in items:
+            if '|' in item:
+                val, wt = item.split('|', 1)
+                choices.append(val)
+                try:
+                    weights.append(float(wt))
+                except ValueError:
+                    weights.append(1.0)
+            else:
+                # fallback if no weight specified
+                choices.append(item)
+                weights.append(1.0)
+
+        instance = random.Random(seed)
+        chosen = instance.choices(population=choices, weights=weights, k=1)[0]
+        return (chosen,)
+
+    RETURN_TYPES = ("STRING",)
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "input_string": ("STRING", {"default": "apple|10$banana|1$orange|3", "display": "text"}),
+                "separator": ("STRING", {"default": "$", "display": "text"}),
+                "seed": ("INT", {"default": 0, "min": 0, "max": 2**63-1, "step": 1, "display": "number"}),
+            }
+        }
+
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "Weighted Random Choice"
+
+@node
+class RandomGaussianFloat(RandomGuaranteedClass):
+    """
+    Generates a random float from a normal (Gaussian) distribution
+    with specified mean and std_dev.
+    """
+    def __init__(self):
+        pass
+
+    def generate(self, mean, std_dev, decimal_places, seed=0):
+        instance = random.Random(seed)
+        value = instance.gauss(mean, std_dev)
+        value = round(value, decimal_places)
+        return (value,)
+
+    RETURN_TYPES = ("FLOAT",)
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "mean": ("FLOAT", {"default": 0.0, "min": -999999999, "max": 999999999.0, "step": 0.01}),
+                "std_dev": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 999999999.0, "step": 0.01}),
+                "decimal_places": ("INT", {"default": 2, "min": 0, "max": 10, "step": 1}),
+                "seed": ("INT", {"default": 0, "min": 0, "max": 2**63-1, "step": 1}),
+            },
+        }
+
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "Random Gaussian Float"
+
+@node
+class SystemRandomGaussianFloat(RandomGuaranteedClass):
+    """
+    Generates a random float from a normal (Gaussian) distribution
+    with specified mean and std_dev.
+    """
+    def __init__(self):
+        pass
+
+    def generate(self, mean, std_dev, decimal_places):
+        instance = random.SystemRandom(time.time())
+        value = instance.gauss(mean, std_dev)
+        value = round(value, decimal_places)
+        return (value,)
+
+    RETURN_TYPES = ("FLOAT",)
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "mean": ("FLOAT", {"default": 0.0, "min": -999999999, "max": 999999999.0, "step": 0.01}),
+                "std_dev": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 999999999.0, "step": 0.01}),
+                "decimal_places": ("INT", {"default": 2, "min": 0, "max": 10, "step": 1}),
+            },
+        }
+
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "System Random Gaussian Float"
+
+@node
+class ProbabilityGate(RandomGuaranteedClass):
+    """
+    Returns TRUE with probability p, FALSE otherwise.
+    """
+    def __init__(self):
+        pass
+
+    def generate(self, probability, seed=0):
+        instance = random.Random(seed)
+        value = instance.random()  # uniform in [0,1)
+        return (value < probability,)
+
+    RETURN_TYPES = ("BOOLEAN",)
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "probability": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}),
+                "seed": ("INT", {"default": 0, "min": 0, "max": 2**63-1, "step": 1}),
+            },
+        }
+
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "Probability Gate"
+
+
+@node
+class UniformRandomInt(RandomGuaranteedClass):
     """
     Selects a random int from min to max
     Fallbacks to default if min is greater than max
@@ -45,7 +377,7 @@ class UniformRandomInt:
         pass
     def generate(self, min_val, max_val, seed=0):
         if min_val > max_val:
-            return min_val
+            return (min_val,)
         instance = random.Random(seed)
         value = instance.randint(min_val, max_val)
         #print(f"Selected {value} from {min_val} to {max_val}")
@@ -56,7 +388,7 @@ class UniformRandomInt:
             "required": {
                 "min_val": ("INT", { "default": 0, "min": -999999999, "max": 999999999, "step": 1, "display": "number" }),
                 "max_val": ("INT", { "default": 1, "min": -999999999, "max": 999999999, "step": 1, "display": "number" }),
-                "seed" : ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "seed" : ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("INT",)
@@ -64,7 +396,7 @@ class UniformRandomInt:
     CATEGORY = "Logic Gates"
     custom_name = "Uniform Random Int"
 @node
-class UniformRandomChoice:
+class UniformRandomChoice(RandomGuaranteedClass):
     """
     Parses input string with separator '$' and returns a random choice
     separator can be changed in the input
@@ -82,7 +414,7 @@ class UniformRandomChoice:
             "required": {
                 "input_string": ("STRING", { "default": "a$b$c", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "seed" : ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "seed" : ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("STRING",)
@@ -108,7 +440,7 @@ class ManualChoiceString:
             "required": {
                 "input_string": ("STRING", { "default": "a$b$c", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "index": ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "index": ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("STRING",)
@@ -135,7 +467,7 @@ class ManualChoiceInt:
             "required": {
                 "input_string": ("STRING", { "default": "1$2$3", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "index": ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "index": ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("INT",)
@@ -162,7 +494,7 @@ class ManualChoiceFloat:
             "required": {
                 "input_string": ("STRING", { "default": "1.0$2.0$3.0", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "index": ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "index": ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("FLOAT",)
@@ -171,7 +503,7 @@ class ManualChoiceFloat:
     custom_name = "Manual Choice Float"
 
 @node
-class RandomShuffleInt:
+class RandomShuffleInt(RandomGuaranteedClass):
     """
     Get the shuffled list of integers from start to end
     Input types and output types are lists of ints
@@ -190,7 +522,7 @@ class RandomShuffleInt:
             "required": {
                 "input_string": ("STRING", { "default": "1$2$3", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "seed" : ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "seed" : ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("STRING",)
@@ -198,7 +530,7 @@ class RandomShuffleInt:
     CATEGORY = "Logic Gates"
     custom_name = "Random Shuffle Int"
 @node
-class RandomShuffleFloat:
+class RandomShuffleFloat(RandomGuaranteedClass):
     """
     Get the shuffled list of floats from start to end
     Input types and output types are lists of floats
@@ -217,7 +549,7 @@ class RandomShuffleFloat:
             "required": {
                 "input_string": ("STRING", { "default": "1.0$2.0$3.0", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "seed" : ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "seed" : ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("STRING",)
@@ -225,7 +557,7 @@ class RandomShuffleFloat:
     CATEGORY = "Logic Gates"
     custom_name = "Random Shuffle Float"
 @node
-class RandomShuffleString:
+class RandomShuffleString(RandomGuaranteedClass):
     """
     Get the shuffled list of strings from start to end
     Input types and output types are lists of strings
@@ -244,7 +576,7 @@ class RandomShuffleString:
             "required": {
                 "input_string": ("STRING", { "default": "a$b$c", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "seed" : ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
+                "seed" : ("INT", { "default": 0, "min": 0, "max": (2**63-1), "step": 1, "display": "number" }),
             },
         }
     RETURN_TYPES = ("STRING",)
@@ -253,13 +585,73 @@ class RandomShuffleString:
     custom_name = "Random Shuffle String"
 
 @node
-class YieldableIteratorString:
+class CounterInteger(RandomGuaranteedClass):
+    """
+    Generates a counter that increments by 1
+    """
+    def __init__(self):
+        self.counter = None
+    def generate(self, reset, start):
+        if self.counter is None:
+            self.counter = start
+        if reset:
+            self.counter = 0
+        self.counter += 1
+        return (int(self.counter),)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "start": ("FLOAT", { "default": 0.0, "min": -(2**63-1), "max": (2**63-1), "step": 1.0, "display": "number" }),
+            },
+            "optional": {
+                "reset": ("BOOLEAN", { "default": False }),
+            },
+        }
+    
+    RETURN_TYPES = ("INT",)
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "Counter Integer"
+    
+@node
+class CounterFloat(RandomGuaranteedClass):
+    """
+    Generates a counter that increments by 1
+    """
+    def __init__(self):
+        self.counter = None
+    def generate(self, reset, start, step):
+        if self.counter is None:
+            self.counter = start
+        if reset:
+            self.counter = start
+        self.counter += step
+        return (self.counter,)
+    @classmethod
+    def INPUT_TYPES(s):
+        return {
+            "required": {
+                "start": ("FLOAT", { "default": 0.0, "min": -(2**63-1), "max": (2**63-1), "step": 1.0, "display": "number" }),
+            },
+            "optional": {
+                "reset": ("BOOLEAN", {"default": False}),
+                "step": ("FLOAT", { "default": 1.0, "min": -(2**63-1), "max": (2**63-1), "step": 1.0, "display": "number" }),
+            },
+        }
+    RETURN_TYPES = ("FLOAT",)
+    FUNCTION = "generate"
+    CATEGORY = "Logic Gates"
+    custom_name = "Counter Float"
+
+@node
+class YieldableIteratorString(RandomGuaranteedClass):
     """
     Yields sequentially from the input list (with separator)
     If reset is True, then it starts from the beginning
     """
     def __init__(self):
-        self.index = 0
+        self.index = -1
     def generate(self, input_string, separator, reset):
         choices = input_string.split(separator)
         if reset:
@@ -276,7 +668,7 @@ class YieldableIteratorString:
             "required": {
                 "input_string": ("STRING", { "default": "a$b$c", "display": "text" }),
                 "separator": ("STRING", { "default": "$", "display": "text" }),
-                "reset": ("INT", { "default": 0, "min": 0, "max": 1, "step": 1, "display": "number" }),
+                "reset": ("BOOLEAN", {"default": False}),
             },
         }
     RETURN_TYPES = ("STRING",)
@@ -285,7 +677,7 @@ class YieldableIteratorString:
     custom_name = "Yieldable Iterator String"
 
 @node
-class YieldableIteratorInt:
+class YieldableIteratorInt(RandomGuaranteedClass):
     """
     Yields sequentially with start, end, step
     Resets if reset is True
@@ -300,21 +692,21 @@ class YieldableIteratorInt:
         if reset:
             self.iterator = None
         if self.iterator is None:
-            self.iterator = range(start, end, step)
+            self.iterator = iter(range(start, end, step))
         try:
             value = next(self.iterator)
         except StopIteration:
-            self.iterator = range(start, end, step)
+            self.iterator = iter(range(start, end, step))
             value = next(self.iterator)
         return (value,)
     @classmethod
     def INPUT_TYPES(s):
         return {
             "required": {
-                "start": ("INT", { "default": 0, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
-                "end": ("INT", { "default": 10, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
-                "step": ("INT", { "default": 1, "min": 0, "max": 9999999999, "step": 1, "display": "number" }),
-                "reset": ("INT", { "default": 0, "min": 0, "max": 1, "step": 1, "display": "number" }),
+                "start": ("INT", { "default": 0, "min": -(2**63-1), "max": (2**63-1), "step": 1, "display": "number" }),
+                "end": ("INT", { "default": 10, "min": -(2**63-1), "max": (2**63-1), "step": 1, "display": "number" }),
+                "step": ("INT", { "default": 1, "min": -(2**63-1), "max": (2**63-1), "step": 1, "display": "number" }),
+                "reset": ("BOOLEAN", {"default": False}),
             },
         }
 

tests/__init__.py

@@ -0,0 +1,5 @@
+"""Unit tests for ComfyUI-LogicUtils.
+
+These tests are intentionally runnable outside a full ComfyUI install.
+"""
+

tests/import_utils.py

@@ -0,0 +1,32 @@
+from __future__ import annotations
+
+import importlib
+import sys
+import types
+from pathlib import Path
+
+
+_PKG_NAME = "comfyui_logicutils"
+
+
+def ensure_local_package() -> str:
+    """Expose the repo's sources as an importable package for unit tests.
+
+    The upstream folder name contains a hyphen, which isn't a valid Python import name.
+    We create an in-memory package module (with a __path__) so intra-package relative
+    imports like `from .autonode import ...` work normally.
+    """
+    if _PKG_NAME in sys.modules:
+        return _PKG_NAME
+
+    repo_root = Path(__file__).resolve().parents[1]
+    package = types.ModuleType(_PKG_NAME)
+    package.__path__ = [str(repo_root)]
+    sys.modules[_PKG_NAME] = package
+    return _PKG_NAME
+
+
+def import_local(module: str):
+    pkg = ensure_local_package()
+    return importlib.import_module(f"{pkg}.{module}")
+

tests/test_conversion.py

@@ -0,0 +1,35 @@
+import unittest
+
+from import_utils import import_local
+
+
+class TestConversion(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.conversion = import_local("conversion")
+
+    def test_generated_conversion_nodes_exist(self):
+        expected = {
+            "ConvertAny2Int",
+            "ConvertAny2Float",
+            "ConvertAny2Boolean",
+            "ConvertAny2String",
+            "ConvertAny2Dict",
+            "ConvertAny2List",
+            "ConvertAny2Tuple",
+            "ConvertAny2Set",
+        }
+        self.assertTrue(expected.issubset(set(self.conversion.CLASS_MAPPINGS.keys())))
+
+    def test_convert_combo_to_string_always_returns_string(self):
+        Node = self.conversion.CLASS_MAPPINGS["ConvertComboToString"]
+        node = Node()
+        self.assertEqual(node.convertComboToString(["a", "b"], "|"), ("a|b",))
+        self.assertEqual(node.convertComboToString([1, 2, 3], ","), ("1,2,3",))
+        self.assertEqual(node.convertComboToString(123, "|"), ("123",))
+
+    def test_string_list_to_combo(self):
+        Node = self.conversion.CLASS_MAPPINGS["StringListToCombo"]
+        node = Node()
+        self.assertEqual(node.stringListToCombo("a$b$c", "$", 1), ("b",))
+        self.assertEqual(node.stringListToCombo("abc", "$", 0), ("abc",))

tests/test_crypto.py

@@ -0,0 +1,38 @@
+import unittest
+
+import numpy as np
+import torch
+from Crypto.PublicKey import RSA
+
+from import_utils import import_local
+
+
+class TestCryptoNodes(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.crypto = import_local("crypto")
+
+    def test_encrypt_decrypt_roundtrip(self):
+        Encrypt = self.crypto.CLASS_MAPPINGS["SecureBase64Encrypt"]
+        Decrypt = self.crypto.CLASS_MAPPINGS["SecureWebPDecrypt"]
+
+        key = RSA.generate(1024)
+        private_pem = key.export_key().decode("utf-8")
+        public_pem = key.publickey().export_key().decode("utf-8")
+
+        # Create a small deterministic RGB image tensor with values aligned to 1/255.
+        arr = np.zeros((8, 8, 3), dtype=np.uint8)
+        arr[0, 0] = [255, 0, 0]
+        arr[0, 1] = [0, 255, 0]
+        arr[0, 2] = [0, 0, 255]
+        img = torch.from_numpy(arr.astype(np.float32) / 255.0).unsqueeze(0)
+
+        enc = Encrypt()
+        encrypted_b64 = enc.encrypted_base64(img, public_pem)[0]
+        self.assertIsInstance(encrypted_b64, str)
+
+        dec = Decrypt()
+        decrypted = dec.decrypt_image(encrypted_b64, private_pem)[0]
+        self.assertTrue(torch.is_tensor(decrypted))
+        self.assertEqual(tuple(decrypted.shape), tuple(img.shape))
+        self.assertTrue(torch.allclose(decrypted, img, atol=1 / 255, rtol=0))

tests/test_imgio_converter.py

@@ -0,0 +1,42 @@
+import unittest
+
+import numpy as np
+import torch
+from PIL import Image
+
+from import_utils import import_local
+
+
+class TestImgIOConverter(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.converter = import_local("imgio.converter")
+
+    def test_handle_rgba_composite_outputs_rgb(self):
+        img = Image.new("RGBA", (4, 4), (255, 0, 0, 128))
+        out = self.converter.handle_rgba_composite(img)
+        self.assertEqual(out.mode, "RGB")
+
+    def test_classify_pil_numpy_torch(self):
+        IOConverter = self.converter.IOConverter
+        img = Image.new("RGB", (2, 2), (0, 0, 0))
+        arr = np.zeros((1, 2, 2, 3), dtype=np.uint8)
+        ten = torch.zeros((1, 2, 2, 3), dtype=torch.float32)
+
+        self.assertEqual(IOConverter.classify(img), IOConverter.InputType.PIL)
+        self.assertEqual(IOConverter.classify(arr), IOConverter.InputType.NUMPY)
+        self.assertEqual(IOConverter.classify(ten), IOConverter.InputType.TORCH)
+
+    def test_base64_roundtrip_pil(self):
+        IOConverter = self.converter.IOConverter
+        img = Image.new("RGB", (3, 5), (10, 20, 30))
+        b64 = IOConverter.convert_to_base64(img, format="PNG")
+        out = IOConverter.convert_to_pil(b64)
+        self.assertEqual(out.size, img.size)
+        self.assertEqual(out.mode, "RGB")
+
+    def test_gzip_base64_string_roundtrip(self):
+        IOConverter = self.converter.IOConverter
+        text = "hello world"
+        b64 = IOConverter.string_to_base64(text, gzip_compress=True)
+        self.assertEqual(IOConverter.read_maybe_gzip_base64(b64), text)

tests/test_io_nodes_extras.py

@@ -0,0 +1,57 @@
+import tempfile
+import unittest
+from pathlib import Path
+
+import torch
+from PIL import Image
+
+from import_utils import import_local
+
+
+class TestIoNodesExtras(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.io_node = import_local("io_node")
+
+    def test_comma_rejoin_defaults(self):
+        Node = self.io_node.CLASS_MAPPINGS["CommaRejoinNode"]
+        self.assertEqual(Node.comma_rejoin("a,b , c"), ("a, b, c",))
+        self.assertEqual(Node.comma_rejoin("  a ,  b  ,c "), ("a, b, c",))
+
+    def test_comma_rejoin_custom_separators(self):
+        Node = self.io_node.CLASS_MAPPINGS["CommaRejoinNode"]
+        self.assertEqual(
+            Node.comma_rejoin("a| b |c", split_separator="|", join_separator="| "),
+            ("a| b| c",),
+        )
+
+    def test_random_image_from_folder(self):
+        Node = self.io_node.CLASS_MAPPINGS["RandomImageFromFolderNode"]
+
+        with tempfile.TemporaryDirectory() as tmp:
+            root = Path(tmp)
+            sub = root / "sub"
+            sub.mkdir()
+
+            img1 = root / "a.png"
+            img2 = root / "b.jpg"
+            img3 = sub / "c.png"
+
+            Image.new("RGB", (8, 6), (10, 20, 30)).save(img1)
+            Image.new("RGB", (8, 6), (40, 50, 60)).save(img2)
+            Image.new("RGB", (8, 6), (70, 80, 90)).save(img3)
+
+            # Non-recursive should only pick from root
+            tensor, path = Node.random_image_from_folder(
+                str(root), "all", False, seed=0
+            )
+            self.assertTrue(torch.is_tensor(tensor))
+            self.assertEqual(tuple(tensor.shape), (1, 6, 8, 3))
+            self.assertIn(Path(path), {img1, img2})
+
+            # Recursive should include subfolder images
+            tensor, path = Node.random_image_from_folder(str(root), "png", True, seed=0)
+            self.assertTrue(torch.is_tensor(tensor))
+            self.assertEqual(tuple(tensor.shape), (1, 6, 8, 3))
+            self.assertIn(Path(path), {img1, img3})
+

tests/test_logic_gates.py

@@ -0,0 +1,44 @@
+import unittest
+
+from import_utils import import_local
+
+
+class TestLogicGates(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.logic_gates = import_local("logic_gates")
+
+    def test_mappings_have_unique_keys(self):
+        module = self.logic_gates
+        self.assertEqual(len(module.CLASS_MAPPINGS), len(set(module.CLASS_MAPPINGS.keys())))
+        self.assertEqual(len(module.CLASS_MAPPINGS), len(module.classes))
+
+    def test_bitwise_shift_supports_negative_shift(self):
+        Shift = self.logic_gates.CLASS_MAPPINGS["LogicGateBitwiseShift"]
+        node = Shift()
+        self.assertEqual(node.bitwiseShift(8, 1), (16,))
+        self.assertEqual(node.bitwiseShift(8, -1), (4,))
+
+    def test_bitwise_shift_validates_range(self):
+        Shift = self.logic_gates.CLASS_MAPPINGS["LogicGateBitwiseShift"]
+        node = Shift()
+        with self.assertRaises(ValueError):
+            node.bitwiseShift(1, 33)
+
+    def test_compare_gate(self):
+        Compare = self.logic_gates.CLASS_MAPPINGS["LogicGateCompare"]
+        node = Compare()
+        self.assertEqual(node.compareInt(2, 1), (True,))
+        self.assertEqual(node.compareInt(1, 2), (False,))
+
+    def test_memory_node_flip_flop(self):
+        Memory = self.logic_gates.CLASS_MAPPINGS["MemoryNode"]
+        node = Memory()
+        self.assertEqual(node.memory("a", 0), ("a",))
+        self.assertEqual(node.memory("b", 0), ("a",))
+        self.assertEqual(node.memory("b", 1), ("b",))
+
+    def test_replace_string_regex(self):
+        Replace = self.logic_gates.CLASS_MAPPINGS["ReplaceString"]
+        node = Replace()
+        self.assertEqual(node.replace("hello", "l+", "x"), ("hexo",))

tests/test_math_nodes.py

@@ -0,0 +1,46 @@
+import unittest
+from unittest.mock import patch
+
+from import_utils import import_local
+
+
+class TestMathNodes(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.math_nodes = import_local("math_nodes")
+
+    def test_power_allows_one_and_zero(self):
+        Power = self.math_nodes.CLASS_MAPPINGS["PowerNode"]
+        node = Power()
+        self.assertEqual(node.power(1, 5), (1.0,))
+        self.assertEqual(node.power(0, 2), (0.0,))
+        self.assertEqual(node.power(0, 0), (1.0,))
+        with self.assertRaises(ZeroDivisionError):
+            node.power(0, -1)
+
+    def test_power_overflow_guard(self):
+        Power = self.math_nodes.CLASS_MAPPINGS["PowerNode"]
+        node = Power()
+        with self.assertRaises(OverflowError):
+            node.power(10, 101)
+
+    def test_modulo_node_is_registered(self):
+        self.assertIn("ModuloNode", self.math_nodes.CLASS_MAPPINGS)
+        Modulo = self.math_nodes.CLASS_MAPPINGS["ModuloNode"]
+        self.assertEqual(Modulo().modulo(10, 3), (1,))
+
+    def test_is_prime_small(self):
+        is_prime_small = self.math_nodes.is_prime_small
+        self.assertFalse(is_prime_small(0))
+        self.assertFalse(is_prime_small(1))
+        self.assertTrue(is_prime_small(2))
+        self.assertTrue(is_prime_small(3))
+        self.assertFalse(is_prime_small(4))
+        self.assertTrue(is_prime_small(7919))
+
+    def test_is_prime_miller_rabin_deterministic_seed(self):
+        is_prime_miller_rabin = self.math_nodes.is_prime_miller_rabin
+        # Force a deterministic base for stable tests.
+        with patch("random.randrange", return_value=2):
+            self.assertTrue(is_prime_miller_rabin(1_000_000_007, k=3))
+            self.assertFalse(is_prime_miller_rabin(1517, k=3))

tests/test_nodes_import.py

@@ -0,0 +1,21 @@
+import unittest
+
+from import_utils import import_local
+
+
+class TestNodesImport(unittest.TestCase):
+    def test_nodes_imports_outside_comfyui(self):
+        nodes = import_local("nodes")
+        self.assertTrue(nodes.NODE_CLASS_MAPPINGS)
+        self.assertIn("LogicGateCompare", nodes.NODE_CLASS_MAPPINGS)
+
+        try:
+            import folder_paths  # noqa: F401
+            in_comfyui = True
+        except ModuleNotFoundError:
+            in_comfyui = False
+
+        if not in_comfyui:
+            # io_node is intentionally skipped outside ComfyUI.
+            self.assertNotIn("SleepNodeAny", nodes.NODE_CLASS_MAPPINGS)
+

tests/test_randomness.py

@@ -0,0 +1,38 @@
+import unittest
+
+from import_utils import import_local
+
+
+class TestRandomness(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.randomness = import_local("randomness")
+
+    def test_uniform_random_float_fallback_is_tuple(self):
+        Node = self.randomness.CLASS_MAPPINGS["UniformRandomFloat"]
+        self.assertEqual(Node().generate(2.0, 1.0, 1, seed=0), (2.0,))
+
+    def test_uniform_random_int_fallback_is_tuple(self):
+        Node = self.randomness.CLASS_MAPPINGS["UniformRandomInt"]
+        self.assertEqual(Node().generate(2, 1, seed=0), (2,))
+
+    def test_triangular_random_float_fallback_is_tuple(self):
+        Node = self.randomness.CLASS_MAPPINGS["TriangularRandomFloat"]
+        self.assertEqual(Node().generate(2.0, 1.0, 1.5, seed=0), (2.0,))
+
+    def test_yieldable_iterator_int_is_iterable_and_wraps(self):
+        Node = self.randomness.CLASS_MAPPINGS["YieldableIteratorInt"]
+        it = Node()
+
+        self.assertEqual(it.generate(0, 3, 1, True), (0,))
+        self.assertEqual(it.generate(0, 3, 1, False), (1,))
+        self.assertEqual(it.generate(0, 3, 1, False), (2,))
+        self.assertEqual(it.generate(0, 3, 1, False), (0,))  # wraps
+
+    def test_yieldable_iterator_string_starts_at_first(self):
+        Node = self.randomness.CLASS_MAPPINGS["YieldableIteratorString"]
+        it = Node()
+
+        self.assertEqual(it.generate("a$b$c", "$", False), ("a",))
+        self.assertEqual(it.generate("a$b$c", "$", False), ("b",))
+        self.assertEqual(it.generate("a$b$c", "$", True), ("a",))

utils/tagger.py

@@ -1,9 +1,25 @@
 try:
     from imgutils.tagging import get_wd14_tags
     from imgutils.tagging.wd14 import MODEL_NAMES as tagger_model_names
-except ImportError:
-    def get_wd14_tags(image_path):
-        raise Exception("Tagger feature not available, please install dghs-imgutils")
+except Exception as e:
+    _tagger_import_error = e
+
+    def get_wd14_tags(image_path, model_name=None):
+        raise RuntimeError(
+            "Tagger feature not available. Install 'dghs-imgutils' to enable it."
+        ) from _tagger_import_error
+    tagger_model_names = {
+        "EVA02_Large": None,
+        "ViT_Large": None,
+        "SwinV2": None,
+        "ConvNext": None,
+        "ConvNextV2": None,
+        "ViT": None,
+        "MOAT": None,
+        "SwinV2_v3": None,
+        "ConvNext_v3": None,
+        "ViT_v3": None,
+    }
 from typing import Union
 from PIL import Image
 
@@ -27,5 +43,4 @@ def get_tags(image_path:Union[str, Image.Image], threshold:float = 0.4, replace:
         result['tags'] = [replace_underscore(tag) for tag in result['tags']]
         result['chars'] = [replace_underscore(tag) for tag in result['chars']]
     return result
-
 tagger_keys = list(tagger_model_names.keys())