ExperionCrawler/futurePlan/End-to-End P&ID Graph Pipeline/test_pipeline_phase3.py

import json
import sys
import os
import asyncio
import networkx as nx

# 경로 설정을 위해 현재 파일의 디렉토리를 sys.path에 추가
current_dir = os.path.dirname(os.path.abspath(__file__))
sys.path.append(current_dir)

from pid_geometric_extractor import PidGeometricExtractor
from pid_topology_builder import PidTopologyBuilder
from pid_intelligent_mapper import IntelligentMapper, validate_mapping

async def run_full_pipeline():
    # 1. 경로 설정
    input_dxf = os.path.join(current_dir, "No-10_Plant_PID.dxf")
    geo_json_path = os.path.join(current_dir, "shared_geo_data.json")
    graph_json_path = os.path.join(current_dir, "pid_graph_topology.json")
    mapping_result_path = os.path.join(current_dir, "pid_final_mapping.json")

    # --- Phase 1: Geometric Extraction ---
    print("\n--- Phase 1: Geometric Extraction ---")
    try:
        extractor = PidGeometricExtractor(input_dxf)
        extractor.extract_and_save(geo_json_path)
        print(f"Geometric data saved to {geo_json_path}")
    except Exception as e:
        print(f"Phase 1 failed: {e}")
        return

    # --- Phase 2: Topology Modeling ---
    print("\n--- Phase 2: Topology Modeling ---")
    try:
        with open(geo_json_path, 'r', encoding='utf-8') as f:
            geometric_data = json.load(f)
        
        builder = PidTopologyBuilder(
            geometric_data=geometric_data, 
            all_extracted_tags=[], 
            config={'dist_threshold': 50.0, 'tag_threshold': 100.0}
        )
        builder.build_graph()
        builder.save_graph(graph_json_path)
        print(f"Graph topology saved to {graph_json_path}")
    except Exception as e:
        print(f"Phase 2 failed: {e}")
        return

    # --- Phase 3: Intelligent Mapping ---
    print("\n--- Phase 3: Intelligent Mapping ---")
    try:
        # 1. 그래프 로드
        with open(graph_json_path, 'r', encoding='utf-8') as f:
            graph_data = json.load(f)
        
        # NetworkX 그래프 복원 (node_link_data 형식 대응)
        from networkx.readwrite import json_graph
        G = json_graph.node_link_graph(graph_data)

        # 2. 시스템 태그 리스트 (실제로는 API나 DB에서 가져와야 함)
        # 테스트를 위한 가상 태그 리스트
        system_tags = [
            "PT-101.PV", "PT-102.PV", "FT-201.PV", "LT-301.PV", 
            "P-101.STATUS", "P-101.SPEED", "V-101.OPEN", "V-101.CLOSE",
            "T-101.TEMP", "TK-101.LEVEL"
        ]

        # 3. 매퍼 초기화 (API Key는 환경변수나 설정파일에서 가져오는 것을 권장)
        api_key = os.getenv("OPENAI_API_KEY", "your-api-key-here")
        mapper = IntelligentMapper(G, system_tags, api_key=api_key)

        # 4. 노드 분류 및 매핑 실행
        nodes = list(G.nodes())
        transmitter_nodes = [n for n in nodes if "Transmitter" in G.nodes[n].get('type', '')]
        valve_nodes = [n for n in nodes if "Valve" in G.nodes[n].get('type', '')]
        equipment_nodes = [n for n in nodes if "Equipment" in G.nodes[n].get('type', '') or "Pump" in G.nodes[n].get('type', '')]

        print(f"Mapping {len(transmitter_nodes)} transmitters, {len(valve_nodes)} valves, {len(equipment_nodes)} equipment...")

        # 비동기 실행
        results = await asyncio.gather(
            mapper.extract_transmitters(transmitter_nodes),
            mapper.extract_valves(valve_nodes),
            mapper.extract_equipment(equipment_nodes)
        )

        # 결과 통합
        final_mapping_raw = {}
        for res in results:
            final_mapping_raw.update(res)

        # 5. 검증 및 최종 결과 정리
        # 가상 메타데이터 (실제로는 시스템에서 조회)
        mock_metadata = {
            "PT-101.PV": {"unit": "bar", "description": "Pressure Transmitter 101"},
            "P-101.STATUS": {"unit": "", "description": "Pump 101 Status"},
        }

        final_results = []
        for node_id, mapping in final_mapping_raw.items():
            symbol_type = G.nodes[node_id].get('type', 'Unknown')
            tag = mapping.resolved_tag
            meta = mock_metadata.get(tag, {"unit": "", "description": ""})
            
            is_valid, val_msg = validate_mapping(tag, symbol_type, meta)
            
            final_results.append({
                "node_id": node_id,
                "symbol_type": symbol_type,
                "original_text": G.nodes[node_id].get('value', ''),
                "resolved_tag": tag,
                "confidence": mapping.confidence,
                "reason": mapping.reason,
                "validation": {
                    "is_valid": is_valid,
                    "message": val_msg
                }
            })

        # 6. 결과 저장
        with open(mapping_result_path, 'w', encoding='utf-8') as f:
            json.dump(final_results, f, indent=4, ensure_ascii=False)
        
        print(f"Final mapping results saved to {mapping_result_path}")
        print(f"Successfully mapped {len(final_results)} nodes.")

    except Exception as e:
        print(f"Phase 3 failed: {e}")
        import traceback
        traceback.print_exc()

if __name__ == "__main__":
    asyncio.run(run_full_pipeline())