New nodes and continue restructuring

__init__.py

@@ -2,6 +2,7 @@ from .nodes.nodes import *
 from .nodes.curve_nodes import *
 from .nodes.batchcrop_nodes import *
 from .nodes.audioscheduler_nodes import *
+from .nodes.image_nodes import *
 NODE_CLASS_MAPPINGS = {
     #constants
     "INTConstant": INTConstant,
@@ -80,6 +81,7 @@ NODE_CLASS_MAPPINGS = {
     "ScaleBatchPromptSchedule": ScaleBatchPromptSchedule,
     "CameraPoseVisualizer": CameraPoseVisualizer,
     "JoinStrings": JoinStrings,
+    "JoinStringMulti": JoinStringMulti,
     "Sleep": Sleep,
     "VRAM_Debug" : VRAM_Debug,
     "SomethingToString" : SomethingToString,

nodes/curve_nodes.py

@@ -217,6 +217,7 @@ class MaskOrImageToWeight:
                     'list',
                     'pandas series',
                     'tensor',
+                    'string'
                 ],
                 {
                 "default": 'list'
@@ -228,7 +229,7 @@ class MaskOrImageToWeight:
             },
 
         }
-    RETURN_TYPES = ("FLOAT",)
+    RETURN_TYPES = ("FLOAT", "STRING",)
     FUNCTION = "execute"
     CATEGORY = "KJNodes"
     DESCRIPTION = """
@@ -249,18 +250,17 @@ and returns that as the selected output type.
                   
         # Convert mean_values to the specified output_type
         if output_type == 'list':
-            return mean_values,
+            out = mean_values,
         elif output_type == 'pandas series':
             try:
                 import pandas as pd
             except:
                 raise Exception("MaskOrImageToWeight: pandas is not installed. Please install pandas to use this output_type")
-            return pd.Series(mean_values),
+            out = pd.Series(mean_values),
         elif output_type == 'tensor':
-            return torch.tensor(mean_values, dtype=torch.float32),
-        else:
-            raise ValueError(f"Unsupported output_type: {output_type}")
-
+            out = torch.tensor(mean_values, dtype=torch.float32),
+        return (out, [str(value) for value in mean_values],)
+    
 class WeightScheduleConvert:
 
     @classmethod
@@ -287,7 +287,7 @@ class WeightScheduleConvert:
              },
              
         }
-    RETURN_TYPES = ("FLOAT",)
+    RETURN_TYPES = ("FLOAT", "STRING",)
     FUNCTION = "execute"
     CATEGORY = "KJNodes"
     DESCRIPTION = """
@@ -344,18 +344,18 @@ Converts different value lists/series to another type.
             float_values = float_values * repeat
 
         if output_type == 'list':
-            return float_values,
+            out = float_values,
         elif output_type == 'pandas series':
-            return pd.Series(float_values),
+            out = pd.Series(float_values),
         elif output_type == 'tensor':
             if input_type == 'pandas series':
-                return torch.tensor(float_values.values, dtype=torch.float32),
+                out = torch.tensor(float_values.values, dtype=torch.float32),
             else:   
-                return torch.tensor(float_values, dtype=torch.float32),
+                out = torch.tensor(float_values, dtype=torch.float32),
         elif output_type == 'match_input':
-            return float_values,
-        else:
-            raise ValueError(f"Unsupported output_type: {output_type}")
+            out = float_values,
+        return (out, [str(value) for value in float_values],)
+        
 
 class FloatToMask:
 

web/js/jsnodes.js

@@ -73,6 +73,28 @@ app.registerExtension({
 						});
 					}
 					break;
+			case "JoinStringMulti":
+				nodeType.prototype.onNodeCreated = function () {
+				this._type = "STRING"
+				this.inputs_offset = nodeData.name.includes("selective")?1:0
+				this.addWidget("button", "Update inputs", null, () => {
+					if (!this.inputs) {
+						this.inputs = [];
+					}
+					const target_number_of_inputs = this.widgets.find(w => w.name === "inputcount")["value"];
+						if(target_number_of_inputs===this.inputs.length)return; // already set, do nothing
+
+						if(target_number_of_inputs < this.inputs.length){
+							for(let i = this.inputs.length; i>=this.inputs_offset+target_number_of_inputs; i--)
+									this.removeInput(i)
+						}
+						else{
+							for(let i = this.inputs.length+1-this.inputs_offset; i <= target_number_of_inputs; ++i)
+								this.addInput(`string_${i}`, this._type)
+							}
+						});
+					}
+					break;
 			case "SoundReactive":
 				nodeType.prototype.onNodeCreated = function () {
 					let audioContext;