ComfyUI/custom_nodes/research/review_classify.py
2026-04-12 18:10:53 +08:00

99 lines
3.7 KiB
Python

# custom_nodes/research/review_classify.py
"""ReviewClassify node - classify review items by type and severity."""
import json
from typing_extensions import override
from comfy_api.latest import ComfyNode, io
class ReviewClassify(io.ComfyNode):
"""Classify review items by type and severity."""
@classmethod
def define_schema(cls) -> io.Schema:
return io.Schema(
node_id="ReviewClassify",
display_name="Classify Review Items",
category="Research",
inputs=[
io.String.Input(
"review_items",
display_name="Review Items (JSON)",
default="{}",
multiline=True,
),
],
outputs=[
io.String.Output(display_name="Classified Items (JSON)"),
],
)
@classmethod
def execute(cls, review_items: str) -> io.NodeOutput:
try:
data = json.loads(review_items) if review_items else {}
items = data.get("items", [])
except json.JSONDecodeError:
items = []
# Define categories
categories = {
"methodology": {"icon": "🔬", "color": "#388bfd"},
"clarity": {"icon": "📝", "color": "#d29922"},
"error": {"icon": "", "color": "#f85149"},
"missing_info": {"icon": "📋", "color": "#a371f7"},
"major_concern": {"icon": "⚠️", "color": "#f85149"},
"suggestion": {"icon": "💡", "color": "#2ea043"},
"reference": {"icon": "📚", "color": "#58a6ff"},
"general": {"icon": "💬", "color": "#8b949e"},
}
# Classify each item
classified = []
for item in items:
item_type = item.get("type", "general")
severity = item.get("severity", 1)
# Determine action category
if severity >= 3:
action = "needs_revision"
elif severity == 2:
action = "needs_response"
else:
action = "consider"
# Map to manuscript section
section_map = {
"methodology": "Methods",
"error": "Methods",
"clarity": "Writing",
"missing_info": "Introduction",
"reference": "Related Work",
}
classified_item = {
**item,
"category": item_type,
"action": action,
"target_section": section_map.get(item_type, "General"),
"icon": categories.get(item_type, categories["general"])["icon"],
"color": categories.get(item_type, categories["general"])["color"],
"priority": "high" if severity >= 3 else "medium" if severity == 2 else "low",
}
classified.append(classified_item)
result = {
"total_items": len(classified),
"by_severity": {
"high": len([i for i in classified if i["priority"] == "high"]),
"medium": len([i for i in classified if i["priority"] == "medium"]),
"low": len([i for i in classified if i["priority"] == "low"]),
},
"by_action": {
"needs_revision": len([i for i in classified if i["action"] == "needs_revision"]),
"needs_response": len([i for i in classified if i["action"] == "needs_response"]),
"consider": len([i for i in classified if i["action"] == "consider"]),
},
"items": classified,
}
return io.NodeOutput(classified_items=json.dumps(result, indent=2))