sirchmunk 0.0.1__py3-none-any.whl → 0.0.2__py3-none-any.whl

sirchmunk/api/knowledge.py

@@ -0,0 +1,411 @@
+# Copyright (c) ModelScope Contributors. All rights reserved.
+"""
+Knowledge Base API endpoints
+Provides CRUD and analytics for KnowledgeCluster objects
+"""
+
+from fastapi import APIRouter, HTTPException
+from typing import Optional
+from pydantic import BaseModel
+
+from sirchmunk.storage.knowledge_storage import KnowledgeStorage
+from sirchmunk.schema.knowledge import AbstractionLevel
+
+router = APIRouter(prefix="/api/v1/knowledge", tags=["knowledge"])
+
+# Initialize Knowledge Manager
+km = KnowledgeStorage()
+
+# === Request/Response Models ===
+
+class SearchRequest(BaseModel):
+    query: str
+    limit: int = 10
+
+# === API Endpoints ===
+
+@router.get("/list")
+async def list_knowledge_bases_alias():
+    """Alias for /clusters endpoint (backward compatibility)"""
+    return await get_all_clusters(limit=100)
+
+
+@router.get("/clusters")
+async def get_all_clusters(
+    limit: int = 100,
+    lifecycle: Optional[str] = None,
+    abstraction_level: Optional[str] = None
+):
+    """
+    Get all knowledge clusters with optional filtering
+    
+    Query params:
+        limit: Maximum number of clusters to return
+        lifecycle: Filter by lifecycle (STABLE, EMERGING, CONTESTED, DEPRECATED)
+        abstraction_level: Filter by abstraction level
+    """
+    try:
+        stats = km.get_stats()
+        
+        # Get all clusters by searching with empty query (returns all)
+        all_clusters = []
+        
+        # Fetch clusters using DuckDB directly
+        sql = "SELECT * FROM knowledge_clusters"
+        where_clauses = []
+        params = []
+        
+        if lifecycle:
+            where_clauses.append("lifecycle = ?")
+            params.append(lifecycle.upper())
+        
+        if abstraction_level:
+            where_clauses.append("abstraction_level = ?")
+            params.append(abstraction_level.upper())
+        
+        if where_clauses:
+            sql += " WHERE " + " AND ".join(where_clauses)
+        
+        sql += f" ORDER BY last_modified DESC LIMIT {limit}"
+        
+        try:
+            rows = km.db.fetch_all(sql, params if params else None)
+        except Exception as fetch_error:
+            # If table is missing or schema is out of date, recreate and return empty list.
+            km._create_table()
+            rows = []
+        clusters = []
+        for row in rows:
+            try:
+                clusters.append(km._row_to_cluster(row))
+            except Exception:
+                # Skip malformed rows to avoid failing the whole request
+                continue
+        
+        return {
+            "success": True,
+                "count": len(clusters),
+                "total": stats.get('custom_stats', {}).get('total_clusters', 0),
+                "data": [c.to_dict() for c in clusters]
+            }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@router.get("/clusters/{cluster_id}")
+async def get_cluster(cluster_id: str):
+    """Get a specific knowledge cluster by ID"""
+    try:
+        cluster = await km.get(cluster_id)
+        if not cluster:
+            raise HTTPException(status_code=404, detail="Cluster not found")
+    
+        return {
+            "success": True,
+                "data": cluster.to_dict()
+            }
+    except HTTPException:
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@router.post("/search")
+async def search_clusters(request: SearchRequest):
+    """
+    Search knowledge clusters by query
+    
+    Searches across: id, name, description, content, patterns
+    """
+    try:
+        results = await km.find(request.query, limit=request.limit)
+    
+        return {
+            "success": True,
+                "query": request.query,
+                "count": len(results),
+                "data": [c.to_dict() for c in results]
+            }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@router.get("/stats")
+async def get_knowledge_stats():
+    """
+    Get comprehensive knowledge base statistics
+    
+    Returns:
+        - Total clusters
+        - Lifecycle distribution
+        - Abstraction level distribution
+        - Average confidence
+        - Hotness distribution
+        - Top patterns
+        - Recent activity
+    """
+    try:
+        stats = km.get_stats()
+        custom_stats = stats.get('custom_stats', {})
+        
+        # Get lifecycle distribution
+        lifecycle_dist = custom_stats.get('lifecycle_distribution', {})
+        
+        # Get abstraction level distribution
+        abstraction_dist = {}
+        for level in AbstractionLevel:
+            count_row = km.db.fetch_one(
+                "SELECT COUNT(*) FROM knowledge_clusters WHERE abstraction_level = ?",
+                [level.name]
+            )
+            abstraction_dist[level.name] = count_row[0] if count_row else 0
+        
+        # Get confidence statistics
+        confidence_stats_row = km.db.fetch_one(
+            """
+            SELECT 
+                MIN(confidence) as min_confidence,
+                MAX(confidence) as max_confidence,
+                AVG(confidence) as avg_confidence
+            FROM knowledge_clusters 
+            WHERE confidence IS NOT NULL
+            """
+        )
+        
+        confidence_stats = {
+            "min": confidence_stats_row[0] if confidence_stats_row and confidence_stats_row[0] else 0,
+            "max": confidence_stats_row[1] if confidence_stats_row and confidence_stats_row[1] else 0,
+            "avg": round(confidence_stats_row[2], 4) if confidence_stats_row and confidence_stats_row[2] else 0,
+        }
+        
+        # Get hotness statistics
+        hotness_stats_row = km.db.fetch_one(
+            """
+            SELECT 
+                MIN(hotness) as min_hotness,
+                MAX(hotness) as max_hotness,
+                AVG(hotness) as avg_hotness
+            FROM knowledge_clusters 
+            WHERE hotness IS NOT NULL
+            """
+        )
+        
+        hotness_stats = {
+            "min": hotness_stats_row[0] if hotness_stats_row and hotness_stats_row[0] else 0,
+            "max": hotness_stats_row[1] if hotness_stats_row and hotness_stats_row[1] else 0,
+            "avg": round(hotness_stats_row[2], 4) if hotness_stats_row and hotness_stats_row[2] else 0,
+        }
+        
+        # Get top 10 most recent clusters
+        recent_rows = km.db.fetch_all(
+            """
+            SELECT id, name, last_modified 
+            FROM knowledge_clusters 
+            ORDER BY last_modified DESC 
+            LIMIT 10
+            """
+        )
+        
+        recent_clusters = [
+            {
+                "id": row[0],
+                "name": row[1],
+                "last_modified": row[2]
+            }
+            for row in recent_rows
+        ]
+        
+        # Get top 10 highest confidence clusters
+        top_confidence_rows = km.db.fetch_all(
+            """
+            SELECT id, name, confidence 
+            FROM knowledge_clusters 
+            WHERE confidence IS NOT NULL
+            ORDER BY confidence DESC 
+            LIMIT 10
+            """
+        )
+        
+        top_confidence = [
+            {
+                "id": row[0],
+                "name": row[1],
+                "confidence": row[2]
+            }
+            for row in top_confidence_rows
+        ]
+        
+        # Get top 10 hottest clusters
+        top_hotness_rows = km.db.fetch_all(
+            """
+            SELECT id, name, hotness 
+            FROM knowledge_clusters 
+            WHERE hotness IS NOT NULL
+            ORDER BY hotness DESC 
+            LIMIT 10
+            """
+        )
+        
+        top_hotness = [
+            {
+                "id": row[0],
+                "name": row[1],
+                "hotness": row[2]
+            }
+            for row in top_hotness_rows
+        ]
+        
+        # Timeline data (clusters created per day for last 30 days)
+        timeline_rows = km.db.fetch_all(
+            """
+            SELECT 
+                CAST(create_time AS DATE) as date,
+                COUNT(*) as count
+            FROM knowledge_clusters
+            WHERE create_time >= current_date - INTERVAL '30 days'
+            GROUP BY CAST(create_time AS DATE)
+            ORDER BY date ASC
+            """
+        )
+        
+        timeline = [
+            {
+                "date": str(row[0]),
+                "count": row[1]
+            }
+            for row in timeline_rows
+        ]
+    
+        return {
+            "success": True,
+            "data": {
+                    "overview": {
+                        "total_clusters": custom_stats.get('total_clusters', 0),
+                        "avg_confidence": custom_stats.get('average_confidence', 0),
+                    },
+                    "lifecycle_distribution": lifecycle_dist,
+                    "abstraction_level_distribution": abstraction_dist,
+                    "confidence_stats": confidence_stats,
+                    "hotness_stats": hotness_stats,
+                    "recent_clusters": recent_clusters,
+                    "top_confidence_clusters": top_confidence,
+                    "top_hotness_clusters": top_hotness,
+                    "timeline": timeline,
+                }
+            }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@router.get("/patterns")
+async def get_top_patterns(limit: int = 20):
+    """
+    Get most common patterns across all clusters
+    
+    Query params:
+        limit: Number of top patterns to return
+    """
+    try:
+        # Fetch all patterns and count occurrences
+        rows = km.db.fetch_all("SELECT patterns FROM knowledge_clusters WHERE patterns IS NOT NULL")
+        
+        import json
+        from collections import Counter
+        
+        pattern_counter = Counter()
+        for row in rows:
+            patterns_json = row[0]
+            if patterns_json:
+                patterns = json.loads(patterns_json)
+                pattern_counter.update(patterns)
+        
+        top_patterns = [
+            {"pattern": pattern, "count": count}
+            for pattern, count in pattern_counter.most_common(limit)
+        ]
+    
+        return {
+            "success": True,
+                "count": len(top_patterns),
+                "data": top_patterns
+            }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@router.get("/graph")
+async def get_knowledge_graph():
+    """
+    Get knowledge graph data (nodes and edges)
+    
+    Returns clusters as nodes and related_clusters as edges
+    """
+    try:
+        # Get all clusters
+        rows = km.db.fetch_all(
+            "SELECT id, name, confidence, hotness, lifecycle, abstraction_level, related_clusters FROM knowledge_clusters"
+        )
+        
+        import json
+        
+        nodes = []
+        edges = []
+        
+        for row in rows:
+            cluster_id, name, confidence, hotness, lifecycle, abstraction_level, related_clusters_json = row
+            
+            # Add node
+            nodes.append({
+                "id": cluster_id,
+                "name": name,
+                "confidence": confidence,
+                "hotness": hotness,
+                "lifecycle": lifecycle,
+                "abstraction_level": abstraction_level,
+            })
+            
+            # Add edges
+            if related_clusters_json:
+                related_clusters = json.loads(related_clusters_json)
+                for rc in related_clusters:
+                    edges.append({
+                        "source": cluster_id,
+                        "target": rc["target_cluster_id"],
+                        "weight": rc["weight"],
+                        "type": rc["source"]
+                    })
+    
+        return {
+            "success": True,
+                "data": {
+                    "nodes": nodes,
+                    "edges": edges
+                }
+            }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@router.delete("/clusters/{cluster_id}")
+async def delete_cluster(cluster_id: str):
+    """Delete a knowledge cluster"""
+    try:
+        success = await km.remove(cluster_id)
+        if not success:
+            raise HTTPException(status_code=404, detail="Cluster not found")
+    
+        return {
+            "success": True,
+                "message": f"Cluster {cluster_id} deleted successfully"
+            }
+    except HTTPException:
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@router.delete("/clusters")
+async def clear_all_clusters():
+    """Clear all knowledge clusters (use with caution!)"""
+    try:
+        success = await km.clear()
+        
+        return {
+            "success": success,
+            "message": "All clusters cleared successfully"
+        }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))

sirchmunk/api/main.py

@@ -0,0 +1,120 @@
+# Copyright (c) ModelScope Contributors. All rights reserved.
+"""
+Main FastAPI application for Sirchmunk API
+Combines all API modules and provides centralized configuration
+"""
+
+from fastapi import FastAPI, HTTPException
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import JSONResponse
+import uvicorn
+
+# Import all API routers
+from .knowledge import router as knowledge_router
+from .settings import router as settings_router
+from .history import router as history_router, dashboard_router
+from .chat import router as chat_router
+from .monitor import router as monitor_router
+from .search import router as search_router
+
+# Create FastAPI application
+app = FastAPI(
+    title="Sirchmunk API",
+    description="APIs for Sirchmunk",
+    version="1.0.0",
+    docs_url="/docs",
+    redoc_url="/redoc"
+)
+
+# Configure CORS
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],  # In production, specify actual origins
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+# Include all API routers
+app.include_router(knowledge_router)
+app.include_router(settings_router)
+app.include_router(history_router)
+app.include_router(dashboard_router)
+app.include_router(chat_router)
+app.include_router(monitor_router)
+app.include_router(search_router)
+
+@app.get("/")
+async def root():
+    """Root endpoint with API information"""
+    return {
+        "name": "Sirchmunk API",
+        "version": "1.0.0",
+        "description": "APIs for Sirchmunk",
+        "status": "running",
+        "endpoints": {
+            "search": "/api/v1/search",
+            "knowledge": "/api/v1/knowledge",
+            "settings": "/api/v1/settings",
+            "history": "/api/v1/history",
+            "chat": "/api/v1/chat",
+            "monitor": "/api/v1/monitor"
+        },
+        "documentation": {
+            "swagger": "/docs",
+            "redoc": "/redoc"
+        }
+    }
+
+@app.get("/health")
+async def health_check():
+    """Health check endpoint"""
+    return {
+        "status": "healthy",
+        "timestamp": "2024-01-13T17:30:00Z",
+        "services": {
+            "api": "running",
+            "database": "connected",
+            "llm": "available",
+            "embedding": "available"
+        }
+    }
+
+@app.exception_handler(404)
+async def not_found_handler(request, exc):
+    """Custom 404 handler"""
+    return JSONResponse(
+        status_code=404,
+        content={
+            "success": False,
+            "error": {
+                "code": "NOT_FOUND",
+                "message": "The requested resource was not found",
+                "path": str(request.url.path)
+            }
+        }
+    )
+
+@app.exception_handler(500)
+async def internal_error_handler(request, exc):
+    """Custom 500 handler"""
+    return JSONResponse(
+        status_code=500,
+        content={
+            "success": False,
+            "error": {
+                "code": "INTERNAL_ERROR",
+                "message": "An internal server error occurred",
+                "details": "Please try again later or contact support"
+            }
+        }
+    )
+
+if __name__ == "__main__":
+    uvicorn.run(
+        "main:app",
+        host="0.0.0.0",
+        port=8584,
+        reload=True,
+        log_level="info"
+    )