Remove unnecessary blank line for consistent formatting

Co-authored-by: jags111 <5968619+jags111@users.noreply.github.com>

.gitignore

@@ -1,10 +1,22 @@
-# Python cache files
+# Python
 __pycache__/
-*.pyc
-*.pyo
-*.pyd
-
-# Compiled files
+*.py[cod]
+*$py.class
 *.so
-*.dll
-*.dylib
+.Python
+
+# Virtual environments
+venv/
+ENV/
+env/
+
+# IDEs
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+
+# OS
+.DS_Store
+Thumbs.db

efficiency_nodes.py

@@ -53,6 +53,7 @@ from .py import city96_latent_upscaler
 from .py import ttl_nn_latent_upscaler
 from .py import bnk_tiled_samplers
 from .py import bnk_adv_encode
+from .py.bnk_adv_encode import normalize_prompt_text
 sys.path.remove(my_dir)
 
 from comfy import samplers
@@ -71,6 +72,9 @@ SCHEDULERS = samplers.KSampler.SCHEDULERS + ["AYS SD1", "AYS SDXL", "AYS SVD", "
 def encode_prompts(positive_prompt, negative_prompt, token_normalization, weight_interpretation, clip, clip_skip,
                    refiner_clip, refiner_clip_skip, ascore, is_sdxl, empty_latent_width, empty_latent_height,
                    return_type="both"):
+    # Ensure prompts are valid strings to prevent tokenization errors
+    positive_prompt = normalize_prompt_text(positive_prompt)
+    negative_prompt = normalize_prompt_text(negative_prompt)
 
     positive_encoded = negative_encoded = refiner_positive_encoded = refiner_negative_encoded = None
 
@@ -326,10 +330,6 @@ class TSC_LoRA_Stacker:
     FUNCTION = "lora_stacker"
     CATEGORY = "Efficiency Nodes/Stackers"
 
-    @classmethod
-    def IS_CHANGED(cls, **kwargs):
-        return float("nan")
-
     def lora_stacker(self, input_mode, lora_count, lora_stack=None, **kwargs):
 
         # Extract values from kwargs

py/bnk_adv_encode.py

@@ -6,6 +6,34 @@ from math import gcd
 from comfy import model_management
 from comfy.sdxl_clip import SDXLClipModel, SDXLRefinerClipModel, SDXLClipG
 
+def normalize_prompt_text(text):
+    """
+    Normalize prompt text to prevent tokenization errors.
+    
+    Converts None, empty strings, or whitespace-only strings to a single space.
+    Ensures the input is a string type by converting non-string values.
+    This function is designed to handle edge cases gracefully without crashing,
+    which is important for ComfyUI workflows where users might have empty prompts.
+    
+    Parameters:
+        text: The input prompt text to normalize. Can be of any type, though
+              string, None, or convertible types are expected.
+    
+    Returns:
+        str: A normalized string that is safe to pass to the tokenizer.
+             Returns " " (single space) for None, empty, or whitespace-only inputs.
+             Returns the original text unchanged if it's a valid non-empty string.
+             Returns str(text) for non-string types.
+    """
+    if text is None:
+        return " "
+    if not isinstance(text, str):
+        # Convert non-string types to string
+        text = str(text)
+    if not text.strip():
+        return " "
+    return text
+
 def _grouper(n, iterable):
     it = iter(iterable)
     while True:
@@ -237,6 +265,8 @@ def prepareXL(embs_l, embs_g, pooled, clip_balance):
         return embs_g, pooled
 
 def advanced_encode(clip, text, token_normalization, weight_interpretation, w_max=1.0, clip_balance=.5, apply_to_pooled=True):
+    # Ensure text is a valid string to prevent tokenization errors
+    text = normalize_prompt_text(text)
     tokenized = clip.tokenize(text, return_word_ids=True)
     if isinstance(clip.cond_stage_model, (SDXLClipModel, SDXLRefinerClipModel, SDXLClipG)):
         embs_l = None
@@ -265,6 +295,9 @@ def advanced_encode(clip, text, token_normalization, weight_interpretation, w_ma
                                            lambda x: (clip.encode_from_tokens({'l': x}), None),
                                            w_max=w_max)
 def advanced_encode_XL(clip, text1, text2, token_normalization, weight_interpretation, w_max=1.0, clip_balance=.5, apply_to_pooled=True):
+    # Ensure texts are valid strings to prevent tokenization errors
+    text1 = normalize_prompt_text(text1)
+    text2 = normalize_prompt_text(text2)
     tokenized1 = clip.tokenize(text1, return_word_ids=True)
     tokenized2 = clip.tokenize(text2, return_word_ids=True)