"""
PostgreSQL全文搜索高級應(yīng)用示例
"""
import psycopg2
from psycopg2.extras import Json, DictCursor
import json
from typing import List, Dict, Any, Optional
import logging
from datetime import datetime

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


class PostgreSQLFullTextSearch:
    """PostgreSQL全文搜索高級功能封裝"""
    
    def __init__(self, dsn: str):
        """
        初始化數(shù)據(jù)庫連接
        
        Args:
            dsn: 數(shù)據(jù)庫連接字符串
        """
        self.conn = psycopg2.connect(dsn)
        self.conn.autocommit = False
        self.cursor = self.conn.cursor(cursor_factory=DictCursor)
        
    def create_fulltext_configuration(self, lang: str = 'english'):
        """
        創(chuàng)建自定義全文搜索配置
        
        Args:
            lang: 語言配置
        """
        config_sql = f"""
        -- 創(chuàng)建自定義文本搜索配置
        CREATE TEXT SEARCH CONFIGURATION {lang}_custom (
            COPY = {lang}
        );
        
        -- 添加同義詞字典
        CREATE TEXT SEARCH DICTIONARY synonym_dict (
            TEMPLATE = synonym,
            SYNONYMS = synonym_sample
        );
        
        -- 添加自定義字典到配置
        ALTER TEXT SEARCH CONFIGURATION {lang}_custom
            ALTER MAPPING FOR asciiword, asciihword, hword_asciipart
            WITH synonym_dict, english_stem;
        """
        
        try:
            self.cursor.execute(config_sql)
            self.conn.commit()
            logger.info(f"全文搜索配置創(chuàng)建成功: {lang}_custom")
        except Exception as e:
            self.conn.rollback()
            logger.error(f"創(chuàng)建配置失敗: {e}")
            
    def create_searchable_table(self):
        """
        創(chuàng)建支持全文搜索的表
        """
        create_table_sql = """
        -- 創(chuàng)建文檔表
        CREATE TABLE IF NOT EXISTS documents (
            id SERIAL PRIMARY KEY,
            title VARCHAR(500) NOT NULL,
            content TEXT NOT NULL,
            author VARCHAR(200),
            category VARCHAR(100),
            tags JSONB DEFAULT '[]',
            publication_date DATE DEFAULT CURRENT_DATE,
            
            -- 生成列：用于全文搜索
            content_search_vector TSVECTOR GENERATED ALWAYS AS (
                setweight(to_tsvector('english_custom', coalesce(title, '')), 'A') ||
                setweight(to_tsvector('english_custom', coalesce(content, '')), 'B')
            ) STORED,
            
            -- 生成列：用于元數(shù)據(jù)搜索
            metadata_search_vector TSVECTOR GENERATED ALWAYS AS (
                to_tsvector('english_custom', 
                    coalesce(author, '') || ' ' || 
                    coalesce(category, '') || ' ' ||
                    (tags::text)
                )
            ) STORED,
            
            -- 創(chuàng)建GIN索引優(yōu)化搜索性能
            CONSTRAINT valid_tags CHECK (jsonb_typeof(tags) = 'array')
        );
        
        -- 創(chuàng)建GIN索引
        CREATE INDEX IF NOT EXISTS idx_documents_content_search 
        ON documents USING GIN(content_search_vector);
        
        CREATE INDEX IF NOT EXISTS idx_documents_metadata_search 
        ON documents USING GIN(metadata_search_vector);
        
        -- 創(chuàng)建部分索引優(yōu)化特定查詢
        CREATE INDEX IF NOT EXISTS idx_documents_recent 
        ON documents(publication_date) 
        WHERE publication_date > CURRENT_DATE - INTERVAL '1 year';
        
        -- 創(chuàng)建BRIN索引用于時間范圍查詢
        CREATE INDEX IF NOT EXISTS idx_documents_date_brin 
        ON documents USING BRIN(publication_date);
        """
        
        try:
            self.cursor.execute(create_table_sql)
            self.conn.commit()
            logger.info("全文搜索表創(chuàng)建成功")
        except Exception as e:
            self.conn.rollback()
            logger.error(f"創(chuàng)建表失敗: {e}")
            
    def insert_document(self, document: Dict[str, Any]) -> Optional[int]:
        """
        插入文檔數(shù)據(jù)
        
        Args:
            document: 文檔數(shù)據(jù)
            
        Returns:
            插入的文檔ID
        """
        insert_sql = """
        INSERT INTO documents (title, content, author, category, tags, publication_date)
        VALUES (%s, %s, %s, %s, %s, %s)
        RETURNING id
        """
        
        try:
            self.cursor.execute(
                insert_sql,
                (
                    document.get('title'),
                    document.get('content'),
                    document.get('author'),
                    document.get('category'),
                    Json(document.get('tags', [])),
                    document.get('publication_date')
                )
            )
            doc_id = self.cursor.fetchone()['id']
            self.conn.commit()
            logger.info(f"文檔插入成功，ID: {doc_id}")
            return doc_id
        except Exception as e:
            self.conn.rollback()
            logger.error(f"插入文檔失敗: {e}")
            return None
            
    def search_documents(
        self, 
        query: str,
        categories: List[str] = None,
        start_date: str = None,
        end_date: str = None,
        min_relevance: float = 0.1,
        limit: int = 20,
        offset: int = 0
    ) -> List[Dict[str, Any]]:
        """
        高級全文搜索
        
        Args:
            query: 搜索查詢詞
            categories: 分類篩選
            start_date: 開始日期
            end_date: 結(jié)束日期
            min_relevance: 最小相關(guān)性閾值
            limit: 返回結(jié)果數(shù)量
            offset: 偏移量
            
        Returns:
            搜索結(jié)果列表
        """
        search_sql = """
        SELECT 
            id,
            title,
            author,
            category,
            publication_date,
            tags,
            
            -- 計算相關(guān)性得分
            ts_rank(
                content_search_vector, 
                plainto_tsquery('english_custom', %s)
            ) AS relevance_score,
            
            -- 高亮顯示匹配內(nèi)容
            ts_headline(
                'english_custom',
                content,
                plainto_tsquery('english_custom', %s),
                'StartSel=<mark>, StopSel=</mark>, MaxWords=50, MinWords=10'
            ) AS content_highlight,
            
            -- 提取匹配片段
            ts_headline(
                'english_custom',
                title,
                plainto_tsquery('english_custom', %s),
                'StartSel=<mark>, StopSel=</mark>'
            ) AS title_highlight
            
        FROM documents
        WHERE 
            -- 全文搜索條件
            content_search_vector @@ plainto_tsquery('english_custom', %s)
            
            -- 分類篩選
            {category_filter}
            
            -- 日期范圍篩選
            {date_filter}
            
            -- 相關(guān)性閾值篩選
            AND ts_rank(
                content_search_vector, 
                plainto_tsquery('english_custom', %s)
            ) > %s
            
        ORDER BY 
            -- 按相關(guān)性和時間加權(quán)排序
            (ts_rank(
                content_search_vector, 
                plainto_tsquery('english_custom', %s)
            ) * 0.7 + 
            (CASE WHEN publication_date > CURRENT_DATE - INTERVAL '30 days' 
                  THEN 0.3 ELSE 0 END)) DESC,
            publication_date DESC
            
        LIMIT %s OFFSET %s
        """
        
        # 構(gòu)建動態(tài)WHERE條件
        category_filter = ""
        date_filter = ""
        params = [query, query, query, query]
        
        if categories:
            placeholders = ', '.join(['%s'] * len(categories))
            category_filter = f"AND category IN ({placeholders})"
            params.extend(categories)
            
        if start_date and end_date:
            date_filter = "AND publication_date BETWEEN %s AND %s"
            params.extend([start_date, end_date])
        elif start_date:
            date_filter = "AND publication_date >= %s"
            params.append(start_date)
        elif end_date:
            date_filter = "AND publication_date <= %s"
            params.append(end_date)
            
        # 添加剩余參數(shù)
        params.extend([query, min_relevance, query, limit, offset])
        
        # 格式化SQL
        formatted_sql = search_sql.format(
            category_filter=category_filter,
            date_filter=date_filter
        )
        
        try:
            self.cursor.execute(formatted_sql, params)
            results = self.cursor.fetchall()
            
            # 轉(zhuǎn)換為字典列表
            return [
                {
                    'id': row['id'],
                    'title': row['title'],
                    'author': row['author'],
                    'category': row['category'],
                    'publication_date': row['publication_date'],
                    'tags': row['tags'],
                    'relevance_score': float(row['relevance_score']),
                    'content_highlight': row['content_highlight'],
                    'title_highlight': row['title_highlight']
                }
                for row in results
            ]
            
        except Exception as e:
            logger.error(f"搜索失敗: {e}")
            return []
            
    def search_similar_documents(self, doc_id: int, limit: int = 10) -> List[Dict[str, Any]]:
        """
        查找相似文檔（基于內(nèi)容相似度）
        
        Args:
            doc_id: 參考文檔ID
            limit: 返回結(jié)果數(shù)量
            
        Returns:
            相似文檔列表
        """
        similarity_sql = """
        WITH target_doc AS (
            SELECT content_search_vector 
            FROM documents 
            WHERE id = %s
        )
        SELECT 
            d.id,
            d.title,
            d.author,
            d.category,
            
            -- 計算余弦相似度
            (d.content_search_vector <=> td.content_search_vector) AS similarity,
            
            -- 提取共同標(biāo)簽
            (
                SELECT jsonb_agg(tag)
                FROM jsonb_array_elements_text(d.tags) AS tag
                WHERE tag IN (
                    SELECT jsonb_array_elements_text(td.tags)
                    FROM documents td 
                    WHERE td.id = %s
                )
            ) AS common_tags
            
        FROM documents d, target_doc td
        WHERE d.id != %s
        ORDER BY similarity DESC
        LIMIT %s
        """
        
        try:
            self.cursor.execute(similarity_sql, (doc_id, doc_id, doc_id, limit))
            results = self.cursor.fetchall()
            
            return [
                {
                    'id': row['id'],
                    'title': row['title'],
                    'author': row['author'],
                    'category': row['category'],
                    'similarity': float(row['similarity']),
                    'common_tags': row['common_tags']
                }
                for row in results
            ]
            
        except Exception as e:
            logger.error(f"查找相似文檔失敗: {e}")
            return []
            
    def get_search_statistics(self, time_period: str = '1 month') -> Dict[str, Any]:
        """
        獲取搜索統(tǒng)計信息
        
        Args:
            time_period: 統(tǒng)計時間周期
            
        Returns:
            統(tǒng)計信息字典
        """
        stats_sql = """
        -- 總文檔數(shù)
        SELECT COUNT(*) as total_documents FROM documents;
        
        -- 按分類統(tǒng)計
        SELECT 
            category,
            COUNT(*) as count,
            ROUND(COUNT(*) * 100.0 / (SELECT COUNT(*) FROM documents), 2) as percentage
        FROM documents 
        WHERE category IS NOT NULL
        GROUP BY category 
        ORDER BY count DESC;
        
        -- 時間分布統(tǒng)計
        SELECT 
            DATE_TRUNC('month', publication_date) as month,
            COUNT(*) as documents_count
        FROM documents
        WHERE publication_date > CURRENT_DATE - INTERVAL %s
        GROUP BY DATE_TRUNC('month', publication_date)
        ORDER BY month DESC;
        
        -- 標(biāo)簽使用統(tǒng)計
        SELECT 
            tag,
            COUNT(*) as usage_count
        FROM documents, jsonb_array_elements_text(tags) as tag
        GROUP BY tag
        ORDER BY usage_count DESC
        LIMIT 20;
        """
        
        try:
            stats = {}
            
            # 執(zhí)行多個統(tǒng)計查詢
            self.cursor.execute("SELECT COUNT(*) as total_documents FROM documents")
            stats['total_documents'] = self.cursor.fetchone()['total_documents']
            
            self.cursor.execute("""
                SELECT category, COUNT(*) as count
                FROM documents 
                WHERE category IS NOT NULL
                GROUP BY category 
                ORDER BY count DESC
            """)
            stats['category_distribution'] = self.cursor.fetchall()
            
            self.cursor.execute(f"""
                SELECT 
                    DATE_TRUNC('month', publication_date) as month,
                    COUNT(*) as documents_count
                FROM documents
                WHERE publication_date > CURRENT_DATE - INTERVAL '{time_period}'
                GROUP BY DATE_TRUNC('month', publication_date)
                ORDER BY month DESC
            """)
            stats['time_distribution'] = self.cursor.fetchall()
            
            self.cursor.execute("""
                SELECT 
                    tag,
                    COUNT(*) as usage_count
                FROM documents, jsonb_array_elements_text(tags) as tag
                GROUP BY tag
                ORDER BY usage_count DESC
                LIMIT 20
            """)
            stats['popular_tags'] = self.cursor.fetchall()
            
            return stats
            
        except Exception as e:
            logger.error(f"獲取統(tǒng)計信息失敗: {e}")
            return {}
            
    def optimize_search_indexes(self):
        """
        優(yōu)化全文搜索索引
        """
        optimize_sql = """
        -- 重新構(gòu)建GIN索引以提高搜索性能
        REINDEX INDEX CONCURRENTLY idx_documents_content_search;
        REINDEX INDEX CONCURRENTLY idx_documents_metadata_search;
        
        -- 更新表統(tǒng)計信息
        ANALYZE documents;
        
        -- 清理索引膨脹
        VACUUM ANALYZE documents;
        
        -- 更新全文搜索配置字典
        ALTER TEXT SEARCH CONFIGURATION english_custom 
        REFRESH VERSION;
        """
        
        try:
            # 分步執(zhí)行優(yōu)化操作
            self.cursor.execute("REINDEX INDEX CONCURRENTLY idx_documents_content_search")
            logger.info("內(nèi)容搜索索引重建完成")
            
            self.cursor.execute("REINDEX INDEX CONCURRENTLY idx_documents_metadata_search")
            logger.info("元數(shù)據(jù)搜索索引重建完成")
            
            self.cursor.execute("ANALYZE documents")
            logger.info("表統(tǒng)計信息更新完成")
            
            self.cursor.execute("VACUUM ANALYZE documents")
            logger.info("表清理完成")
            
            self.conn.commit()
            logger.info("全文搜索索引優(yōu)化完成")
            
        except Exception as e:
            self.conn.rollback()
            logger.error(f"索引優(yōu)化失敗: {e}")
            
    def close(self):
        """關(guān)閉數(shù)據(jù)庫連接"""
        if self.cursor:
            self.cursor.close()
        if self.conn:
            self.conn.close()
        logger.info("數(shù)據(jù)庫連接已關(guān)閉")


def example_usage():
    """使用示例"""
    # 數(shù)據(jù)庫連接字符串
    dsn = "dbname=testdb user=postgres password=password host=localhost port=5432"
    
    # 創(chuàng)建全文搜索實例
    search = PostgreSQLFullTextSearch(dsn)
    
    try:
        # 1. 創(chuàng)建全文搜索配置
        search.create_fulltext_configuration('english')
        
        # 2. 創(chuàng)建搜索表
        search.create_searchable_table()
        
        # 3. 插入示例文檔
        sample_documents = [
            {
                'title': 'PostgreSQL Performance Optimization',
                'content': 'PostgreSQL provides advanced optimization techniques including query planning, indexing strategies, and configuration tuning.',
                'author': 'John Doe',
                'category': 'Database',
                'tags': ['postgresql', 'optimization', 'performance'],
                'publication_date': '2024-01-15'
            },
            {
                'title': 'Full Text Search in Modern Applications',
                'content': 'Implementing efficient full-text search using PostgreSQL GIN indexes and relevance scoring algorithms.',
                'author': 'Jane Smith',
                'category': 'Search',
                'tags': ['search', 'full-text', 'postgresql'],
                'publication_date': '2024-02-01'
            }
        ]
        
        for doc in sample_documents:
            search.insert_document(doc)
        
        # 4. 執(zhí)行高級搜索
        print("執(zhí)行全文搜索...")
        results = search.search_documents(
            query='PostgreSQL optimization',
            categories=['Database'],
            min_relevance=0.05,
            limit=10
        )
        
        print(f"找到 {len(results)} 個結(jié)果:")
        for result in results:
            print(f"- {result['title']} (相關(guān)性: {result['relevance_score']:.3f})")
        
        # 5. 查找相似文檔
        if results:
            similar = search.search_similar_documents(results[0]['id'])
            print(f"\n相似文檔: {len(similar)} 個")
        
        # 6. 獲取統(tǒng)計信息
        stats = search.get_search_statistics()
        print(f"\n總文檔數(shù): {stats.get('total_documents', 0)}")
        
    finally:
        search.close()


if __name__ == "__main__":
    example_usage()

"""
PostgreSQL查詢性能分析與優(yōu)化工具
"""
import psycopg2
from psycopg2.extras import DictCursor
import time
from typing import Dict, List, Any, Optional, Tuple
import statistics
import json
from datetime import datetime, timedelta
import logging

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


class QueryPerformanceAnalyzer:
    """查詢性能分析器"""
    
    def __init__(self, dsn: str):
        self.conn = psycopg2.connect(dsn)
        self.cursor = self.conn.cursor(cursor_factory=DictCursor)
        self.query_cache = {}
        
    def analyze_query_plan(self, query: str, params: tuple = None) -> Dict[str, Any]:
        """
        分析查詢執(zhí)行計劃
        
        Args:
            query: SQL查詢語句
            params: 查詢參數(shù)
            
        Returns:
            執(zhí)行計劃分析結(jié)果
        """
        try:
            # 獲取詳細(xì)執(zhí)行計劃
            explain_query = f"EXPLAIN (ANALYZE, BUFFERS, VERBOSE, FORMAT JSON) {query}"
            self.cursor.execute(explain_query, params)
            plan_result = self.cursor.fetchone()[0]
            
            plan = plan_result[0]['Plan']
            return self._parse_execution_plan(plan)
            
        except Exception as e:
            logger.error(f"分析執(zhí)行計劃失敗: {e}")
            return {}
    
    def _parse_execution_plan(self, plan: Dict[str, Any]) -> Dict[str, Any]:
        """
        解析執(zhí)行計劃
        
        Args:
            plan: 執(zhí)行計劃字典
            
        Returns:
            解析后的分析結(jié)果
        """
        analysis = {
            'operation_type': plan.get('Node Type'),
            'relation_name': plan.get('Relation Name'),
            'alias': plan.get('Alias'),
            'startup_cost': plan.get('Startup Cost'),
            'total_cost': plan.get('Total Cost'),
            'plan_rows': plan.get('Plan Rows'),
            'plan_width': plan.get('Plan Width'),
            'actual_rows': plan.get('Actual Rows'),
            'actual_time': plan.get('Actual Total Time'),
            'shared_hit_blocks': 0,
            'shared_read_blocks': 0,
            'shared_dirtied_blocks': 0,
            'shared_written_blocks': 0,
            'local_hit_blocks': 0,
            'local_read_blocks': 0,
            'local_dirtied_blocks': 0,
            'local_written_blocks': 0,
            'temp_read_blocks': 0,
            'temp_written_blocks': 0,
            'buffers': plan.get('Shared Hit Blocks', 0) + plan.get('Shared Read Blocks', 0),
            'children': [],
            'issues': []
        }
        
        # 解析緩沖區(qū)使用情況
        if 'Shared Hit Blocks' in plan:
            analysis['shared_hit_blocks'] = plan['Shared Hit Blocks']
        if 'Shared Read Blocks' in plan:
            analysis['shared_read_blocks'] = plan['Shared Read Blocks']
        
        # 遞歸解析子節(jié)點
        if 'Plans' in plan:
            for child_plan in plan['Plans']:
                child_analysis = self._parse_execution_plan(child_plan)
                analysis['children'].append(child_analysis)
        
        # 識別潛在問題
        self._identify_issues(analysis, plan)
        
        return analysis
    
    def _identify_issues(self, analysis: Dict[str, Any], plan: Dict[str, Any]):
        """識別查詢計劃中的潛在問題"""
        
        # 檢查全表掃描
        if analysis['operation_type'] == 'Seq Scan' and analysis['actual_rows'] > 10000:
            analysis['issues'].append({
                'type': 'FULL_TABLE_SCAN',
                'severity': 'HIGH',
                'message': '檢測到大量行的全表掃描',
                'suggestion': '考慮添加合適的索引'
            })
        
        # 檢查嵌套循環(huán)連接
        if analysis['operation_type'] == 'Nested Loop' and analysis['actual_rows'] > 1000:
            analysis['issues'].append({
                'type': 'INEFFICIENT_JOIN',
                'severity': 'MEDIUM',
                'message': '嵌套循環(huán)連接可能效率較低',
                'suggestion': '考慮使用Hash Join或Merge Join'
            })
        
        # 檢查排序操作
        if analysis['operation_type'] == 'Sort' and analysis['actual_rows'] > 10000:
            analysis['issues'].append({
                'type': 'LARGE_SORT',
                'severity': 'MEDIUM',
                'message': '大規(guī)模排序操作',
                'suggestion': '考慮添加索引以避免排序'
            })
        
        # 檢查緩沖區(qū)命中率
        total_blocks = analysis['shared_hit_blocks'] + analysis['shared_read_blocks']
        if total_blocks > 0:
            hit_ratio = analysis['shared_hit_blocks'] / total_blocks
            if hit_ratio < 0.9:
                analysis['issues'].append({
                    'type': 'LOW_BUFFER_HIT',
                    'severity': 'MEDIUM',
                    'message': f'緩沖區(qū)命中率較低: {hit_ratio:.2%}',
                    'suggestion': '考慮增加shared_buffers或優(yōu)化查詢'
                })
    
    def benchmark_query(self, query: str, params: tuple = None, 
                       iterations: int = 10) -> Dict[str, Any]:
        """
        基準(zhǔn)測試查詢性能
        
        Args:
            query: SQL查詢語句
            params: 查詢參數(shù)
            iterations: 迭代次數(shù)
            
        Returns:
            性能基準(zhǔn)測試結(jié)果
        """
        execution_times = []
        row_counts = []
        
        try:
            # 預(yù)熱緩存
            self.cursor.execute(query, params)
            _ = self.cursor.fetchall()
            
            # 執(zhí)行基準(zhǔn)測試
            for i in range(iterations):
                start_time = time.perf_counter()
                self.cursor.execute(query, params)
                rows = self.cursor.fetchall()
                end_time = time.perf_counter()
                
                execution_times.append(end_time - start_time)
                row_counts.append(len(rows))
            
            # 分析執(zhí)行計劃
            plan_analysis = self.analyze_query_plan(query, params)
            
            # 計算統(tǒng)計信息
            stats = {
                'iterations': iterations,
                'total_time': sum(execution_times),
                'avg_time': statistics.mean(execution_times),
                'min_time': min(execution_times),
                'max_time': max(execution_times),
                'std_dev': statistics.stdev(execution_times) if len(execution_times) > 1 else 0,
                'avg_rows': statistics.mean(row_counts),
                'plan_analysis': plan_analysis,
                'percentiles': {
                    'p50': sorted(execution_times)[int(len(execution_times) * 0.5)],
                    'p90': sorted(execution_times)[int(len(execution_times) * 0.9)],
                    'p95': sorted(execution_times)[int(len(execution_times) * 0.95)],
                    'p99': sorted(execution_times)[int(len(execution_times) * 0.99)],
                }
            }
            
            return stats
            
        except Exception as e:
            logger.error(f"基準(zhǔn)測試失敗: {e}")
            return {}
    
    def generate_optimization_suggestions(self, query: str, 
                                        stats: Dict[str, Any]) -> List[Dict[str, Any]]:
        """
        生成優(yōu)化建議
        
        Args:
            query: SQL查詢語句
            stats: 性能統(tǒng)計信息
            
        Returns:
            優(yōu)化建議列表
        """
        suggestions = []
        plan_analysis = stats.get('plan_analysis', {})
        
        # 基于執(zhí)行時間建議
        avg_time = stats.get('avg_time', 0)
        if avg_time > 1.0:  # 超過1秒
            suggestions.append({
                'priority': 'HIGH',
                'area': 'PERFORMANCE',
                'suggestion': '查詢執(zhí)行時間較長，考慮優(yōu)化查詢或添加索引',
                'estimated_impact': 'HIGH'
            })
        
        # 基于執(zhí)行計劃建議
        for issue in plan_analysis.get('issues', []):
            suggestions.append({
                'priority': issue['severity'],
                'area': 'QUERY_PLAN',
                'suggestion': issue['suggestion'],
                'estimated_impact': 'MEDIUM'
            })
        
        # 基于統(tǒng)計信息建議
        if stats.get('std_dev', 0) / stats.get('avg_time', 1) > 0.5:
            suggestions.append({
                'priority': 'MEDIUM',
                'area': 'CONSISTENCY',
                'suggestion': '查詢執(zhí)行時間波動較大，可能存在并發(fā)或資源競爭問題',
                'estimated_impact': 'MEDIUM'
            })
        
        return suggestions


class IndexOptimizer:
    """索引優(yōu)化器"""
    
    def __init__(self, dsn: str):
        self.conn = psycopg2.connect(dsn)
        self.cursor = self.conn.cursor(cursor_factory=DictCursor)
    
    def analyze_table_indexes(self, table_name: str) -> List[Dict[str, Any]]:
        """
        分析表索引
        
        Args:
            table_name: 表名
            
        Returns:
            索引分析結(jié)果
        """
        query = """
        SELECT 
            i.relname as index_name,
            am.amname as index_type,
            idx.indisunique as is_unique,
            idx.indisprimary as is_primary,
            idx.indisexclusion as is_exclusion,
            idx.indisclustered as is_clustered,
            idx.indisvalid as is_valid,
            idx.indpred as partial_index_predicate,
            pg_relation_size(i.oid) as index_size_bytes,
            pg_size_pretty(pg_relation_size(i.oid)) as index_size,
            pg_stat_get_numscans(i.oid) as scan_count,
            pg_stat_get_tuples_returned(i.oid) as tuples_returned,
            pg_stat_get_tuples_fetched(i.oid) as tuples_fetched,
            
            -- 索引定義
            pg_get_indexdef(idx.indexrelid) as index_definition,
            
            -- 索引列
            array_to_string(array_agg(a.attname), ', ') as index_columns
            
        FROM pg_index idx
        JOIN pg_class i ON i.oid = idx.indexrelid
        JOIN pg_class t ON t.oid = idx.indrelid
        JOIN pg_am am ON i.relam = am.oid
        JOIN pg_attribute a ON a.attrelid = t.oid AND a.attnum = ANY(idx.indkey)
        
        WHERE t.relname = %s
        GROUP BY i.relname, am.amname, idx.indisunique, idx.indisprimary,
                 idx.indisexclusion, idx.indisclustered, idx.indisvalid,
                 idx.indpred, i.oid, idx.indexrelid
        ORDER BY pg_relation_size(i.oid) DESC
        """
        
        try:
            self.cursor.execute(query, (table_name,))
            indexes = self.cursor.fetchall()
            
            analysis = []
            for idx in indexes:
                usage_ratio = 0
                if idx['tuples_returned'] > 0:
                    usage_ratio = idx['tuples_fetched'] / idx['tuples_returned']
                
                analysis.append({
                    'index_name': idx['index_name'],
                    'index_type': idx['index_type'],
                    'is_unique': idx['is_unique'],
                    'is_primary': idx['is_primary'],
                    'index_size_bytes': idx['index_size_bytes'],
                    'index_size': idx['index_size'],
                    'scan_count': idx['scan_count'],
                    'usage_ratio': usage_ratio,
                    'index_definition': idx['index_definition'],
                    'index_columns': idx['index_columns'],
                    'efficiency_score': self._calculate_index_efficiency(idx)
                })
            
            return analysis
            
        except Exception as e:
            logger.error(f"分析索引失敗: {e}")
            return []
    
    def _calculate_index_efficiency(self, index_info: Dict[str, Any]) -> float:
        """
        計算索引效率評分
        
        Args:
            index_info: 索引信息
            
        Returns:
            效率評分 (0-100)
        """
        score = 100.0
        
        # 基于使用率扣分
        if index_info['scan_count'] == 0:
            score -= 50  # 從未使用
        
        # 基于大小扣分
        size_mb = index_info['index_size_bytes'] / (1024 * 1024)
        if size_mb > 1000:  # 超過1GB
            score -= 30
        elif size_mb > 100:  # 超過100MB
            score -= 15
        
        # 基于唯一性加分
        if index_info['is_unique']:
            score += 10
        
        return max(0, min(100, score))
    
    def suggest_index_improvements(self, table_name: str, 
                                 query_patterns: List[str]) -> List[Dict[str, Any]]:
        """
        基于查詢模式建議索引改進(jìn)
        
        Args:
            table_name: 表名
            query_patterns: 查詢模式列表
            
        Returns:
            索引改進(jìn)建議
        """
        suggestions = []
        existing_indexes = self.analyze_table_indexes(table_name)
        
        for pattern in query_patterns:
            pattern_lower = pattern.lower()
            
            # 提取WHERE子句中的列
            where_start = pattern_lower.find('where ')
            if where_start != -1:
                where_clause = pattern_lower[where_start + 6:]
                
                # 簡單提取列名（實際應(yīng)用中應(yīng)使用SQL解析器）
                import re
                column_matches = re.findall(r'(\w+)\s*[=<>!]', where_clause)
                
                for column in column_matches:
                    # 檢查是否已有索引
                    has_index = False
                    for idx in existing_indexes:
                        if column in idx['index_columns'].lower():
                            has_index = True
                            break
                    
                    if not has_index:
                        suggestions.append({
                            'table': table_name,
                            'column': column,
                            'suggestion': f'在 {column} 列上創(chuàng)建索引',
                            'estimated_impact': 'HIGH',
                            'sql': f'CREATE INDEX idx_{table_name}_{column} ON {table_name}({column})'
                        })
        
        return suggestions


class PostgreSQLConfigOptimizer:
    """PostgreSQL配置優(yōu)化器"""
    
    def __init__(self, dsn: str):
        self.conn = psycopg2.connect(dsn)
        self.cursor = self.conn.cursor(cursor_factory=DictCursor)
    
    def analyze_current_config(self) -> Dict[str, Any]:
        """
        分析當(dāng)前配置
        
        Returns:
            配置分析結(jié)果
        """
        config_queries = {
            'basic_settings': """
                SELECT name, setting, unit, context, vartype
                FROM pg_settings
                WHERE name IN (
                    'shared_buffers', 'work_mem', 'maintenance_work_mem',
                    'effective_cache_size', 'max_connections'
                )
            """,
            'performance_settings': """
                SELECT name, setting, unit, context, vartype
                FROM pg_settings
                WHERE name LIKE '%cost%' OR name LIKE '%join%' OR name LIKE '%parallel%'
                ORDER BY name
            """,
            'wal_settings': """
                SELECT name, setting, unit, context, vartype
                FROM pg_settings
                WHERE name LIKE 'wal_%'
                ORDER BY name
            """,
            'statistics': """
                SELECT 
                    datname as database_name,
                    numbackends as active_connections,
                    xact_commit as transactions_committed,
                    xact_rollback as transactions_rolled_back,
                    blks_read as blocks_read,
                    blks_hit as blocks_hit,
                    tup_returned as tuples_returned,
                    tup_fetched as tuples_fetched,
                    tup_inserted as tuples_inserted,
                    tup_updated as tuples_updated,
                    tup_deleted as tuples_deleted
                FROM pg_stat_database
                WHERE datname = current_database()
            """
        }
        
        analysis = {}
        
        try:
            for category, query in config_queries.items():
                self.cursor.execute(query)
                analysis[category] = self.cursor.fetchall()
            
            # 計算緩沖區(qū)命中率
            if 'statistics' in analysis and analysis['statistics']:
                stats = analysis['statistics'][0]
                blocks_hit = stats['blocks_hit']
                blocks_read = stats['blocks_read']
                total_blocks = blocks_hit + blocks_read
                
                if total_blocks > 0:
                    analysis['buffer_hit_ratio'] = blocks_hit / total_blocks
                else:
                    analysis['buffer_hit_ratio'] = 0
            
            return analysis
            
        except Exception as e:
            logger.error(f"分析配置失敗: {e}")
            return {}
    
    def generate_config_recommendations(self, 
                                      system_memory_gb: float = 16,
                                      expected_connections: int = 100) -> List[Dict[str, Any]]:
        """
        生成配置優(yōu)化建議
        
        Args:
            system_memory_gb: 系統(tǒng)總內(nèi)存（GB）
            expected_connections: 預(yù)期最大連接數(shù)
            
        Returns:
            配置優(yōu)化建議列表
        """
        recommendations = []
        current_config = self.analyze_current_config()
        
        # 共享緩沖區(qū)建議（通常為系統(tǒng)內(nèi)存的25%）
        recommended_shared_buffers = f"{int(system_memory_gb * 0.25 * 1024)}MB"
        recommendations.append({
            'parameter': 'shared_buffers',
            'current_value': self._get_config_value(current_config, 'shared_buffers'),
            'recommended_value': recommended_shared_buffers,
            'reason': f'設(shè)置為系統(tǒng)內(nèi)存({system_memory_gb}GB)的25%以優(yōu)化緩存性能',
            'impact': 'HIGH'
        })
        
        # 工作內(nèi)存建議
        recommended_work_mem = f"{int(system_memory_gb * 1024 / expected_connections / 4)}MB"
        recommendations.append({
            'parameter': 'work_mem',
            'current_value': self._get_config_value(current_config, 'work_mem'),
            'recommended_value': recommended_work_mem,
            'reason': f'基于{expected_connections}個并發(fā)連接和系統(tǒng)內(nèi)存計算',
            'impact': 'MEDIUM'
        })
        
        # 維護(hù)工作內(nèi)存建議
        recommended_maintenance_work_mem = f"{int(system_memory_gb * 0.1 * 1024)}MB"
        recommendations.append({
            'parameter': 'maintenance_work_mem',
            'current_value': self._get_config_value(current_config, 'maintenance_work_mem'),
            'recommended_value': recommended_maintenance_work_mem,
            'reason': '設(shè)置為系統(tǒng)內(nèi)存的10%以優(yōu)化維護(hù)操作性能',
            'impact': 'MEDIUM'
        })
        
        # 有效緩存大小建議
        recommended_effective_cache_size = f"{int(system_memory_gb * 0.5 * 1024)}MB"
        recommendations.append({
            'parameter': 'effective_cache_size',
            'current_value': self._get_config_value(current_config, 'effective_cache_size'),
            'recommended_value': recommended_effective_cache_size,
            'reason': '設(shè)置為系統(tǒng)內(nèi)存的50%以幫助查詢規(guī)劃器做出更好的決策',
            'impact': 'MEDIUM'
        })
        
        # 基于緩沖區(qū)命中率的建議
        hit_ratio = current_config.get('buffer_hit_ratio', 0)
        if hit_ratio < 0.9:
            recommendations.append({
                'parameter': 'BUFFER_HIT_RATIO',
                'current_value': f'{hit_ratio:.2%}',
                'recommended_value': '>90%',
                'reason': '緩沖區(qū)命中率較低，可能影響查詢性能',
                'impact': 'HIGH',
                'additional_suggestions': [
                    '增加shared_buffers',
                    '優(yōu)化熱點查詢',
                    '考慮使用pg_prewarm擴(kuò)展'
                ]
            })
        
        return recommendations
    
    def _get_config_value(self, config_analysis: Dict[str, Any], 
                         param_name: str) -> str:
        """獲取配置參數(shù)值"""
        if 'basic_settings' in config_analysis:
            for setting in config_analysis['basic_settings']:
                if setting['name'] == param_name:
                    return f"{setting['setting']} {setting['unit'] or ''}".strip()
        
        return '未找到'
    
    def close(self):
        """關(guān)閉連接"""
        if self.cursor:
            self.cursor.close()
        if self.conn:
            self.conn.close()


def comprehensive_performance_analysis(dsn: str):
    """綜合性能分析示例"""
    print("=" * 60)
    print("PostgreSQL性能綜合分析")
    print("=" * 60)
    
    # 1. 查詢性能分析
    print("\n1. 查詢性能分析")
    print("-" * 40)
    
    analyzer = QueryPerformanceAnalyzer(dsn)
    
    test_query = """
    SELECT 
        u.username,
        COUNT(o.id) as order_count,
        SUM(o.amount) as total_amount,
        AVG(o.amount) as avg_amount
    FROM users u
    JOIN orders o ON u.id = o.user_id
    WHERE u.created_at > CURRENT_DATE - INTERVAL '1 year'
    GROUP BY u.id, u.username
    HAVING COUNT(o.id) > 5
    ORDER BY total_amount DESC
    LIMIT 100
    """
    
    benchmark_results = analyzer.benchmark_query(test_query, iterations=5)
    
    if benchmark_results:
        print(f"平均執(zhí)行時間: {benchmark_results['avg_time']:.3f}秒")
        print(f"最小執(zhí)行時間: {benchmark_results['min_time']:.3f}秒")
        print(f"最大執(zhí)行時間: {benchmark_results['max_time']:.3f}秒")
        print(f"標(biāo)準(zhǔn)差: {benchmark_results['std_dev']:.3f}秒")
        
        # 生成優(yōu)化建議
        suggestions = analyzer.generate_optimization_suggestions(
            test_query, benchmark_results
        )
        
        print(f"\n優(yōu)化建議 ({len(suggestions)}條):")
        for i, suggestion in enumerate(suggestions, 1):
            print(f"{i}. [{suggestion['priority']}] {suggestion['suggestion']}")
    
    # 2. 索引分析
    print("\n2. 索引分析")
    print("-" * 40)
    
    index_optimizer = IndexOptimizer(dsn)
    table_name = "users"  # 假設(shè)的表名
    
    index_analysis = index_optimizer.analyze_table_indexes(table_name)
    
    if index_analysis:
        print(f"表 '{table_name}' 的索引分析:")
        for idx in index_analysis[:5]:  # 顯示前5個索引
            print(f"  - {idx['index_name']}: {idx['index_size']}, "
                  f"使用率: {idx['usage_ratio']:.2f}, "
                  f"效率評分: {idx['efficiency_score']:.1f}")
    
    # 3. 配置分析
    print("\n3. 配置分析")
    print("-" * 40)
    
    config_optimizer = PostgreSQLConfigOptimizer(dsn)
    config_recommendations = config_optimizer.generate_config_recommendations(
        system_memory_gb=16,
        expected_connections=200
    )
    
    print("配置優(yōu)化建議:")
    for rec in config_recommendations:
        print(f"  - {rec['parameter']}: {rec['current_value']} → {rec['recommended_value']}")
    
    # 4. 綜合報告
    print("\n4. 綜合性能報告")
    print("-" * 40)
    
    overall_score = 85.0  # 示例評分
    bottlenecks = [
        "查詢響應(yīng)時間波動較大",
        "部分索引使用率較低",
        "緩沖區(qū)命中率需要優(yōu)化"
    ]
    
    print(f"總體性能評分: {overall_score}/100")
    print("\n主要瓶頸:")
    for bottleneck in bottlenecks:
        print(f"  ? {bottleneck}")
    
    print("\n優(yōu)化優(yōu)先級:")
    print("  1. 優(yōu)化高響應(yīng)時間查詢")
    print("  2. 調(diào)整數(shù)據(jù)庫配置參數(shù)")
    print("  3. 重建低效率索引")
    
    # 清理資源
    analyzer.cursor.close()
    analyzer.conn.close()
    config_optimizer.close()


if __name__ == "__main__":
    # 數(shù)據(jù)庫連接字符串
    dsn = "dbname=testdb user=postgres password=password host=localhost port=5432"
    
    comprehensive_performance_analysis(dsn)

"""
PostgreSQL并發(fā)控制與鎖優(yōu)化示例
"""
import psycopg2
from psycopg2.extras import DictCursor
import threading
import time
from typing import List, Dict, Any
import logging
from datetime import datetime

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


class ConcurrentTransactionTest:
    """并發(fā)事務(wù)測試"""
    
    def __init__(self, dsn: str):
        self.dsn = dsn
        self.results = []
        self.lock = threading.Lock()
    
    def run_concurrent_updates(self, num_threads: int = 10):
        """
        運行并發(fā)更新測試
        
        Args:
            num_threads: 并發(fā)線程數(shù)
        """
        print(f"開始并發(fā)更新測試，線程數(shù): {num_threads}")
        print("=" * 60)
        
        threads = []
        
        # 初始化測試數(shù)據(jù)
        self._setup_test_data()
        
        # 創(chuàng)建并啟動線程
        for i in range(num_threads):
            thread = threading.Thread(
                target=self._update_account_balance,
                args=(i, 100 + i, 50.0),  # 賬戶ID從100開始
                name=f"Transaction-{i}"
            )
            threads.append(thread)
            thread.start()
        
        # 等待所有線程完成
        for thread in threads:
            thread.join()
        
        # 驗證結(jié)果
        self._verify_results()
        
        print(f"\n測試完成，總事務(wù)數(shù): {num_threads}")
        print(f"成功事務(wù): {len([r for r in self.results if r['success']])}")
        print(f"失敗事務(wù): {len([r for r in self.results if not r['success']])}")
    
    def _setup_test_data(self):
        """設(shè)置測試數(shù)據(jù)"""
        conn = psycopg2.connect(self.dsn)
        cursor = conn.cursor()
        
        try:
            # 創(chuàng)建測試表
            cursor.execute("""
                CREATE TABLE IF NOT EXISTS accounts (
                    id SERIAL PRIMARY KEY,
                    account_number VARCHAR(50) UNIQUE,
                    balance DECIMAL(15, 2) DEFAULT 0.0,
                    version INTEGER DEFAULT 0,
                    last_updated TIMESTAMP DEFAULT CURRENT_TIMESTAMP
                );
            """)
            
            # 插入測試賬戶
            for i in range(10):
                account_id = 100 + i
                cursor.execute("""
                    INSERT INTO accounts (id, account_number, balance, version)
                    VALUES (%s, %s, %s, %s)
                    ON CONFLICT (id) DO UPDATE SET
                        balance = EXCLUDED.balance,
                        version = EXCLUDED.version
                """, (account_id, f"ACC{account_id}", 1000.0, 0))
            
            conn.commit()
            logger.info("測試數(shù)據(jù)設(shè)置完成")
            
        except Exception as e:
            conn.rollback()
            logger.error(f"設(shè)置測試數(shù)據(jù)失敗: {e}")
        finally:
            cursor.close()
            conn.close()
    
    def _update_account_balance(self, thread_id: int, account_id: int, amount: float):
        """
        更新賬戶余額
        
        Args:
            thread_id: 線程ID
            account_id: 賬戶ID
            amount: 更新金額
        """
        conn = None
        cursor = None
        
        try:
            conn = psycopg2.connect(self.dsn)
            cursor = conn.cursor()
            
            # 設(shè)置事務(wù)隔離級別（可測試不同級別）
            cursor.execute("SET TRANSACTION ISOLATION LEVEL READ COMMITTED")
            
            start_time = time.time()
            
            # 方法1: 使用行級鎖
            # cursor.execute("""
            #     SELECT balance FROM accounts 
            #     WHERE id = %s FOR UPDATE
            # """, (account_id,))
            
            # 方法2: 樂觀鎖（版本控制）
            cursor.execute("""
                SELECT balance, version FROM accounts 
                WHERE id = %s
            """, (account_id,))
            
            result = cursor.fetchone()
            if not result:
                raise Exception(f"賬戶 {account_id} 不存在")
            
            current_balance, current_version = result
            new_balance = current_balance + amount
            
            # 模擬一些處理時間
            time.sleep(0.01)
            
            # 使用樂觀鎖更新
            cursor.execute("""
                UPDATE accounts 
                SET balance = %s, 
                    version = version + 1,
                    last_updated = CURRENT_TIMESTAMP
                WHERE id = %s AND version = %s
            """, (new_balance, account_id, current_version))
            
            # 檢查是否更新成功
            if cursor.rowcount == 0:
                # 樂觀鎖沖突，重試或失敗
                conn.rollback()
                success = False
                error_msg = "樂觀鎖沖突，版本不匹配"
            else:
                conn.commit()
                success = True
                error_msg = None
            
            end_time = time.time()
            duration = end_time - start_time
            
            # 記錄結(jié)果
            with self.lock:
                self.results.append({
                    'thread_id': thread_id,
                    'account_id': account_id,
                    'success': success,
                    'duration': duration,
                    'error': error_msg,
                    'start_time': start_time,
                    'end_time': end_time
                })
            
            if success:
                logger.debug(f"線程 {thread_id}: 賬戶 {account_id} 更新成功，耗時 {duration:.3f}秒")
            else:
                logger.warning(f"線程 {thread_id}: 賬戶 {account_id} 更新失敗 - {error_msg}")
            
        except Exception as e:
            if conn:
                conn.rollback()
            
            with self.lock:
                self.results.append({
                    'thread_id': thread_id,
                    'account_id': account_id,
                    'success': False,
                    'duration': time.time() - start_time if 'start_time' in locals() else 0,
                    'error': str(e),
                    'start_time': start_time if 'start_time' in locals() else 0,
                    'end_time': time.time()
                })
            
            logger.error(f"線程 {thread_id} 異常: {e}")
            
        finally:
            if cursor:
                cursor.close()
            if conn:
                conn.close()
    
    def _verify_results(self):
        """驗證測試結(jié)果"""
        conn = psycopg2.connect(self.dsn)
        cursor = conn.cursor()
        
        try:
            # 檢查賬戶最終狀態(tài)
            cursor.execute("""
                SELECT id, balance, version, last_updated
                FROM accounts
                WHERE id >= 100 AND id < 110
                ORDER BY id
            """)
            
            accounts = cursor.fetchall()
            
            print("\n賬戶最終狀態(tài):")
            print("-" * 60)
            print(f"{'賬戶ID':<10} {'余額':<15} {'版本':<10} {'最后更新'}")
            print("-" * 60)
            
            for acc in accounts:
                print(f"{acc[0]:<10} {float(acc[1]):<15.2f} {acc[2]:<10} {acc[3]}")
            
            # 分析并發(fā)問題
            success_count = len([r for r in self.results if r['success']])
            total_count = len(self.results)
            
            if success_count < total_count:
                print(f"\n發(fā)現(xiàn)并發(fā)問題: {total_count - success_count} 個事務(wù)失敗")
                print("\n失敗事務(wù)詳情:")
                for result in self.results:
                    if not result['success']:
                        print(f"  線程 {result['thread_id']}: {result['error']}")
            
            # 計算性能指標(biāo)
            if self.results:
                durations = [r['duration'] for r in self.results if r['success']]
                if durations:
                    avg_duration = sum(durations) / len(durations)
                    max_duration = max(durations)
                    min_duration = min(durations)
                    
                    print(f"\n性能指標(biāo):")
                    print(f"  平均事務(wù)時間: {avg_duration:.3f}秒")
                    print(f"  最長事務(wù)時間: {max_duration:.3f}秒")
                    print(f"  最短事務(wù)時間: {min_duration:.3f}秒")
                    print(f"  吞吐量: {success_count / sum(durations):.2f} 事務(wù)/秒")
            
        except Exception as e:
            logger.error(f"驗證結(jié)果失敗: {e}")
        finally:
            cursor.close()
            conn.close()


class LockMonitor:
    """鎖監(jiān)控器"""
    
    def __init__(self, dsn: str):
        self.dsn = dsn
    
    def get_current_locks(self) -> List[Dict[str, Any]]:
        """
        獲取當(dāng)前鎖信息
        
        Returns:
            鎖信息列表
        """
        conn = psycopg2.connect(self.dsn)
        cursor = conn.cursor(cursor_factory=DictCursor)
        
        try:
            lock_query = """
            SELECT 
                -- 鎖信息
                pl.pid as process_id,
                pl.mode as lock_mode,
                pl.granted as is_granted,
                pl.fastpath as is_fastpath,
                
                -- 事務(wù)信息
                pa.query as current_query,
                pa.state as query_state,
                pa.wait_event_type as wait_event_type,
                pa.wait_event as wait_event,
                pa.backend_start as backend_start_time,
                pa.xact_start as transaction_start_time,
                pa.query_start as query_start_time,
                
                -- 被鎖對象信息
                pl.relation::regclass as locked_relation,
                pl.page as locked_page,
                pl.tuple as locked_tuple,
                pl.virtualxid as virtual_transaction_id,
                pl.transactionid as transaction_id,
                pl.classid::regclass as locked_class,
                pl.objid as locked_object_id,
                pl.objsubid as locked_object_subid,
                
                -- 等待圖信息
                pg_blocking_pids(pl.pid) as blocking_pids,
                
                -- 附加信息
                now() - pa.query_start as query_duration,
                now() - pa.xact_start as transaction_duration
                
            FROM pg_locks pl
            LEFT JOIN pg_stat_activity pa ON pl.pid = pa.pid
            WHERE pl.pid <> pg_backend_pid()  -- 排除當(dāng)前連接
            ORDER BY 
                pl.granted DESC,  -- 先顯示未授予的鎖
                transaction_duration DESC,
                query_duration DESC
            """
            
            cursor.execute(lock_query)
            locks = cursor.fetchall()
            
            return [
                {
                    'process_id': lock['process_id'],
                    'lock_mode': lock['lock_mode'],
                    'is_granted': lock['is_granted'],
                    'current_query': lock['current_query'][:100] if lock['current_query'] else None,
                    'query_state': lock['query_state'],
                    'locked_relation': lock['locked_relation'],
                    'blocking_pids': lock['blocking_pids'],
                    'query_duration': lock['query_duration'],
                    'transaction_duration': lock['transaction_duration'],
                    'wait_event': lock['wait_event']
                }
                for lock in locks
            ]
            
        except Exception as e:
            logger.error(f"獲取鎖信息失敗: {e}")
            return []
        finally:
            cursor.close()
            conn.close()
    
    def analyze_lock_contention(self) -> Dict[str, Any]:
        """
        分析鎖爭用情況
        
        Returns:
            鎖爭用分析報告
        """
        locks = self.get_current_locks()
        
        analysis = {
            'total_locks': len(locks),
            'granted_locks': len([l for l in locks if l['is_granted']]),
            'waiting_locks': len([l for l in locks if not l['is_granted']]),
            'lock_modes': {},
            'wait_chains': [],
            'long_running_transactions': [],
            'potential_deadlocks': []
        }
        
        # 統(tǒng)計鎖模式
        for lock in locks:
            mode = lock['lock_mode']
            analysis['lock_modes'][mode] = analysis['lock_modes'].get(mode, 0) + 1
        
        # 識別等待鏈
        waiting_processes = [l for l in locks if not l['is_granted']]
        for waiter in waiting_processes:
            if waiter['blocking_pids']:
                chain = {
                    'waiting_pid': waiter['process_id'],
                    'blocking_pids': waiter['blocking_pids'],
                    'lock_mode': waiter['lock_mode'],
                    'wait_time': waiter['query_duration']
                }
                analysis['wait_chains'].append(chain)
        
        # 識別長時間運行的事務(wù)
        for lock in locks:
            if lock['transaction_duration'] and lock['transaction_duration'].total_seconds() > 60:
                analysis['long_running_transactions'].append({
                    'pid': lock['process_id'],
                    'duration_seconds': lock['transaction_duration'].total_seconds(),
                    'query': lock['current_query']
                })
        
        return analysis
    
    def kill_blocking_processes(self, threshold_seconds: int = 300):
        """
        終止阻塞時間過長的進(jìn)程
        
        Args:
            threshold_seconds: 阻塞時間閾值（秒）
        """
        conn = psycopg2.connect(self.dsn)
        cursor = conn.cursor()
        
        try:
            # 查找阻塞時間過長的進(jìn)程
            kill_query = """
            SELECT pid, query, now() - xact_start as duration
            FROM pg_stat_activity
            WHERE pid IN (
                SELECT DISTINCT unnest(pg_blocking_pids(pid))
                FROM pg_stat_activity
                WHERE wait_event_type = 'Lock'
                AND state = 'active'
                AND now() - query_start > INTERVAL '%s seconds'
            )
            AND state = 'active'
            """
            
            cursor.execute(kill_query % threshold_seconds)
            blocking_processes = cursor.fetchall()
            
            killed = []
            for proc in blocking_processes:
                pid, query, duration = proc
                try:
                    cursor.execute("SELECT pg_terminate_backend(%s)", (pid,))
                    killed.append({
                        'pid': pid,
                        'duration': duration,
                        'query': query[:100] if query else None
                    })
                    logger.warning(f"終止阻塞進(jìn)程 {pid}，已運行 {duration}")
                except Exception as e:
                    logger.error(f"終止進(jìn)程 {pid} 失敗: {e}")
            
            conn.commit()
            return killed
            
        except Exception as e:
            conn.rollback()
            logger.error(f"終止阻塞進(jìn)程失敗: {e}")
            return []
        finally:
            cursor.close()
            conn.close()


def test_concurrency_scenarios():
    """測試不同并發(fā)場景"""
    dsn = "dbname=testdb user=postgres password=password host=localhost port=5432"
    
    print("并發(fā)控制測試")
    print("=" * 60)
    
    # 場景1：高并發(fā)更新
    print("\n場景1: 高并發(fā)更新測試")
    test1 = ConcurrentTransactionTest(dsn)
    test1.run_concurrent_updates(num_threads=20)
    
    # 場景2：鎖監(jiān)控
    print("\n\n場景2: 鎖監(jiān)控分析")
    print("-" * 40)
    
    monitor = LockMonitor(dsn)
    lock_analysis = monitor.analyze_lock_contention()
    
    print(f"總鎖數(shù): {lock_analysis['total_locks']}")
    print(f"已授予鎖: {lock_analysis['granted_locks']}")
    print(f"等待鎖: {lock_analysis['waiting_locks']}")
    
    if lock_analysis['wait_chains']:
        print("\n等待鏈:")
        for chain in lock_analysis['wait_chains'][:5]:  # 顯示前5個
            print(f"  進(jìn)程 {chain['waiting_pid']} 等待 {chain['blocking_pids']}")
    
    if lock_analysis['long_running_transactions']:
        print("\n長時間運行事務(wù):")
        for txn in lock_analysis['long_running_transactions'][:3]:
            print(f"  進(jìn)程 {txn['pid']}: 已運行 {txn['duration_seconds']:.0f}秒")
    
    # 場景3：死鎖處理建議
    print("\n\n場景3: 死鎖預(yù)防建議")
    print("-" * 40)
    
    recommendations = [
        "1. 使用合適的索引減少鎖競爭范圍",
        "2. 保持事務(wù)簡短，盡快提交",
        "3. 使用顯式鎖（SELECT FOR UPDATE）時按固定順序訪問資源",
        "4. 設(shè)置合理的鎖超時（lock_timeout）",
        "5. 考慮使用樂觀鎖（版本控制）替代悲觀鎖",
        "6. 使用較低的隔離級別（READ COMMITTED）",
        "7. 監(jiān)控和調(diào)整max_connections參數(shù)",
        "8. 定期分析并優(yōu)化長時間運行的事務(wù)"
    ]
    
    for rec in recommendations:
        print(f"  ? {rec}")


if __name__ == "__main__":
    test_concurrency_scenarios()

"""
電商系統(tǒng)PostgreSQL高級特性綜合應(yīng)用示例
"""
import psycopg2
from psycopg2.extras import Json, DictCursor
import json
from typing import List, Dict, Any, Optional
from datetime import datetime, timedelta
import logging
from decimal import Decimal

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


class ECommerceDatabase:
    """電商系統(tǒng)數(shù)據(jù)庫設(shè)計"""
    
    def __init__(self, dsn: str):
        self.conn = psycopg2.connect(dsn)
        self.cursor = self.conn.cursor(cursor_factory=DictCursor)
    
    def create_schema(self):
        """創(chuàng)建電商系統(tǒng)數(shù)據(jù)庫架構(gòu)"""
        schema_sql = """
        -- 啟用必要擴(kuò)展
        CREATE EXTENSION IF NOT EXISTS "uuid-ossp";
        CREATE EXTENSION IF NOT EXISTS "pg_trgm";
        CREATE EXTENSION IF NOT EXISTS "btree_gin";
        
        -- 1. 產(chǎn)品表（使用JSONB存儲變體屬性）
        CREATE TABLE IF NOT EXISTS products (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            sku VARCHAR(100) UNIQUE NOT NULL,
            name VARCHAR(500) NOT NULL,
            description TEXT,
            category_id UUID,
            brand VARCHAR(200),
            
            -- JSONB存儲動態(tài)屬性
            attributes JSONB DEFAULT '{}',
            
            -- 價格信息
            base_price DECIMAL(12, 2) NOT NULL,
            discount_price DECIMAL(12, 2),
            
            -- 庫存信息
            stock_quantity INTEGER DEFAULT 0,
            reserved_quantity INTEGER DEFAULT 0,
            
            -- 搜索優(yōu)化字段
            search_vector TSVECTOR GENERATED ALWAYS AS (
                setweight(to_tsvector('english', coalesce(name, '')), 'A') ||
                setweight(to_tsvector('english', coalesce(description, '')), 'B') ||
                setweight(to_tsvector('english', coalesce(brand, '')), 'C')
            ) STORED,
            
            -- 時間戳
            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            
            -- 約束
            CONSTRAINT positive_price CHECK (base_price >= 0),
            CONSTRAINT positive_stock CHECK (stock_quantity >= 0),
            CONSTRAINT valid_discount CHECK (
                discount_price IS NULL OR 
                (discount_price >= 0 AND discount_price <= base_price)
            )
        );
        
        -- 產(chǎn)品表索引
        CREATE INDEX IF NOT EXISTS idx_products_sku ON products(sku);
        CREATE INDEX IF NOT EXISTS idx_products_category ON products(category_id);
        CREATE INDEX IF NOT EXISTS idx_products_price ON products(base_price);
        CREATE INDEX IF NOT EXISTS idx_products_search ON products USING GIN(search_vector);
        CREATE INDEX IF NOT EXISTS idx_products_attributes ON products USING GIN(attributes);
        CREATE INDEX IF NOT EXISTS idx_products_brand ON products(brand);
        
        -- 2. 產(chǎn)品變體表（范圍類型用于尺寸）
        CREATE TABLE IF NOT EXISTS product_variants (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            product_id UUID REFERENCES products(id) ON DELETE CASCADE,
            variant_name VARCHAR(200) NOT NULL,
            
            -- 使用范圍類型表示尺寸范圍
            size_range numrange,
            weight_range numrange,
            
            -- JSONB存儲變體特定屬性
            variant_attributes JSONB DEFAULT '{}',
            
            -- 價格調(diào)整
            price_adjustment DECIMAL(10, 2) DEFAULT 0,
            additional_cost DECIMAL(10, 2) DEFAULT 0,
            
            -- 庫存跟蹤
            variant_stock INTEGER DEFAULT 0,
            min_order_quantity INTEGER DEFAULT 1,
            max_order_quantity INTEGER,
            
            -- 約束
            CONSTRAINT valid_size_range CHECK (
                size_range IS NULL OR 
                (lower(size_range) >= 0 AND upper(size_range) > lower(size_range))
            ),
            CONSTRAINT valid_quantity CHECK (
                min_order_quantity > 0 AND 
                (max_order_quantity IS NULL OR max_order_quantity >= min_order_quantity)
            )
        );
        
        -- 變體表索引
        CREATE INDEX IF NOT EXISTS idx_variants_product ON product_variants(product_id);
        CREATE INDEX IF NOT EXISTS idx_variants_size ON product_variants USING GIST(size_range);
        
        -- 3. 分類表（使用遞歸CTE支持層級結(jié)構(gòu)）
        CREATE TABLE IF NOT EXISTS categories (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            name VARCHAR(200) NOT NULL,
            slug VARCHAR(200) UNIQUE NOT NULL,
            description TEXT,
            parent_id UUID REFERENCES categories(id),
            sort_order INTEGER DEFAULT 0,
            
            -- JSONB存儲分類屬性
            category_attributes JSONB DEFAULT '{}',
            
            -- 層級路徑（物化路徑模式）
            path VARCHAR(1000),
            level INTEGER DEFAULT 0,
            
            -- 時間戳
            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            
            -- 約束
            CONSTRAINT no_self_parent CHECK (id != parent_id)
        );
        
        -- 分類表索引
        CREATE INDEX IF NOT EXISTS idx_categories_parent ON categories(parent_id);
        CREATE INDEX IF NOT EXISTS idx_categories_path ON categories(path);
        CREATE INDEX IF NOT EXISTS idx_categories_slug ON categories(slug);
        
        -- 4. 訂單表（使用分區(qū)表）
        CREATE TABLE IF NOT EXISTS orders_partitioned (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            order_number VARCHAR(50) UNIQUE NOT NULL,
            customer_id UUID NOT NULL,
            status VARCHAR(50) NOT NULL,
            
            -- JSONB存儲訂單元數(shù)據(jù)
            order_metadata JSONB DEFAULT '{}',
            
            -- 金額信息
            subtotal DECIMAL(12, 2) NOT NULL,
            tax_amount DECIMAL(12, 2) DEFAULT 0,
            shipping_amount DECIMAL(12, 2) DEFAULT 0,
            discount_amount DECIMAL(12, 2) DEFAULT 0,
            total_amount DECIMAL(12, 2) NOT NULL,
            
            -- 時間信息
            ordered_at TIMESTAMP NOT NULL,
            shipped_at TIMESTAMP,
            delivered_at TIMESTAMP,
            
            -- 約束
            CONSTRAINT positive_amounts CHECK (
                subtotal >= 0 AND
                tax_amount >= 0 AND
                shipping_amount >= 0 AND
                discount_amount >= 0 AND
                total_amount >= 0
            )
        ) PARTITION BY RANGE (ordered_at);
        
        -- 創(chuàng)建訂單分區(qū)（每月一個分區(qū)）
        CREATE TABLE IF NOT EXISTS orders_2024_01 
        PARTITION OF orders_partitioned
        FOR VALUES FROM ('2024-01-01') TO ('2024-02-01');
        
        CREATE TABLE IF NOT EXISTS orders_2024_02 
        PARTITION OF orders_partitioned
        FOR VALUES FROM ('2024-02-01') TO ('2024-03-01');
        
        -- 訂單表索引
        CREATE INDEX IF NOT EXISTS idx_orders_customer ON orders_partitioned(customer_id);
        CREATE INDEX IF NOT EXISTS idx_orders_status ON orders_partitioned(status);
        CREATE INDEX IF NOT EXISTS idx_orders_date ON orders_partitioned(ordered_at);
        CREATE INDEX IF NOT EXISTS idx_orders_metadata ON orders_partitioned USING GIN(order_metadata);
        
        -- 5. 訂單項表
        CREATE TABLE IF NOT EXISTS order_items (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            order_id UUID REFERENCES orders_partitioned(id) ON DELETE CASCADE,
            product_id UUID REFERENCES products(id),
            variant_id UUID REFERENCES product_variants(id),
            
            -- 產(chǎn)品快照（避免產(chǎn)品信息變更影響歷史訂單）
            product_snapshot JSONB NOT NULL,
            
            -- 購買信息
            quantity INTEGER NOT NULL,
            unit_price DECIMAL(12, 2) NOT NULL,
            discount_percentage DECIMAL(5, 2) DEFAULT 0,
            item_total DECIMAL(12, 2) NOT NULL,
            
            -- 約束
            CONSTRAINT positive_quantity CHECK (quantity > 0),
            CONSTRAINT positive_unit_price CHECK (unit_price >= 0)
        );
        
        -- 訂單項索引
        CREATE INDEX IF NOT EXISTS idx_order_items_order ON order_items(order_id);
        CREATE INDEX IF NOT EXISTS idx_order_items_product ON order_items(product_id);
        
        -- 6. 客戶表
        CREATE TABLE IF NOT EXISTS customers (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            email VARCHAR(255) UNIQUE NOT NULL,
            phone VARCHAR(50),
            first_name VARCHAR(100),
            last_name VARCHAR(100),
            
            -- JSONB存儲客戶屬性
            customer_profile JSONB DEFAULT '{}',
            
            -- 地址信息（使用JSONB存儲多個地址）
            addresses JSONB DEFAULT '[]',
            
            -- 賬戶信息
            is_active BOOLEAN DEFAULT TRUE,
            loyalty_points INTEGER DEFAULT 0,
            customer_tier VARCHAR(50) DEFAULT 'STANDARD',
            
            -- 時間戳
            registered_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            last_login_at TIMESTAMP,
            
            -- 約束
            CONSTRAINT valid_email CHECK (email ~* '^[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Z|a-z]{2,}$')
        );
        
        -- 客戶表索引
        CREATE INDEX IF NOT EXISTS idx_customers_email ON customers(email);
        CREATE INDEX IF NOT EXISTS idx_customers_name ON customers(last_name, first_name);
        CREATE INDEX IF NOT EXISTS idx_customers_profile ON customers USING GIN(customer_profile);
        CREATE INDEX IF NOT EXISTS idx_customers_tier ON customers(customer_tier);
        
        -- 7. 庫存變更日志（使用BRIN索引）
        CREATE TABLE IF NOT EXISTS inventory_logs (
            id BIGSERIAL PRIMARY KEY,
            product_id UUID REFERENCES products(id),
            variant_id UUID REFERENCES product_variants(id),
            
            -- 變更信息
            change_type VARCHAR(50) NOT NULL,
            quantity_change INTEGER NOT NULL,
            previous_quantity INTEGER,
            new_quantity INTEGER,
            
            -- 關(guān)聯(lián)信息
            reference_id UUID,
            reference_type VARCHAR(100),
            
            -- 時間戳
            changed_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            changed_by UUID,
            
            -- 備注
            notes TEXT
        ) WITH (fillfactor = 90);
        
        -- 庫存日志索引（使用BRIN適合時間序列）
        CREATE INDEX IF NOT EXISTS idx_inventory_logs_time 
        ON inventory_logs USING BRIN(changed_at);
        
        CREATE INDEX IF NOT EXISTS idx_inventory_logs_product 
        ON inventory_logs(product_id, changed_at);
        
        -- 8. 產(chǎn)品評論表（使用數(shù)組和全文搜索）
        CREATE TABLE IF NOT EXISTS product_reviews (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            product_id UUID REFERENCES products(id) ON DELETE CASCADE,
            customer_id UUID REFERENCES customers(id),
            order_item_id UUID REFERENCES order_items(id),
            
            -- 評分
            rating INTEGER NOT NULL CHECK (rating >= 1 AND rating <= 5),
            
            -- 評論內(nèi)容
            title VARCHAR(500),
            review_text TEXT NOT NULL,
            
            -- 數(shù)組存儲優(yōu)點/缺點
            pros TEXT[],
            cons TEXT[],
            
            -- 元數(shù)據(jù)
            is_verified_purchase BOOLEAN DEFAULT FALSE,
            helpful_votes INTEGER DEFAULT 0,
            not_helpful_votes INTEGER DEFAULT 0,
            
            -- 全文搜索向量
            review_vector TSVECTOR GENERATED ALWAYS AS (
                to_tsvector('english', 
                    coalesce(title, '') || ' ' || 
                    coalesce(review_text, '')
                )
            ) STORED,
            
            -- 時間戳
            reviewed_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            
            -- 約束
            CONSTRAINT one_review_per_order_item UNIQUE (order_item_id)
        );
        
        -- 評論表索引
        CREATE INDEX IF NOT EXISTS idx_reviews_product ON product_reviews(product_id);
        CREATE INDEX IF NOT EXISTS idx_reviews_rating ON product_reviews(rating);
        CREATE INDEX IF NOT EXISTS idx_reviews_search ON product_reviews USING GIN(review_vector);
        CREATE INDEX IF NOT EXISTS idx_reviews_pros ON product_reviews USING GIN(pros);
        CREATE INDEX IF NOT EXISTS idx_reviews_date ON product_reviews(reviewed_at);
        
        -- 9. 促銷規(guī)則表（使用復(fù)雜約束和JSONB）
        CREATE TABLE IF NOT EXISTS promotion_rules (
            id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
            name VARCHAR(200) NOT NULL,
            description TEXT,
            
            -- 規(guī)則條件（JSONB存儲復(fù)雜規(guī)則）
            conditions JSONB NOT NULL,
            
            -- 折扣信息
            discount_type VARCHAR(50) NOT NULL,
            discount_value DECIMAL(10, 2),
            discount_percentage DECIMAL(5, 2),
            max_discount_amount DECIMAL(12, 2),
            
            -- 時間范圍
            valid_from TIMESTAMP NOT NULL,
            valid_until TIMESTAMP,
            
            -- 使用限制
            usage_limit INTEGER,
            per_customer_limit INTEGER,
            minimum_order_amount DECIMAL(12, 2),
            
            -- 狀態(tài)
            is_active BOOLEAN DEFAULT TRUE,
            
            -- 元數(shù)據(jù)
            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
            
            -- 約束
            CONSTRAINT valid_discount_values CHECK (
                (discount_type = 'AMOUNT' AND discount_value IS NOT NULL) OR
                (discount_type = 'PERCENTAGE' AND discount_percentage IS NOT NULL)
            ),
            CONSTRAINT valid_discount_percentage CHECK (
                discount_percentage IS NULL OR 
                (discount_percentage >= 0 AND discount_percentage <= 100)
            )
        );
        
        -- 促銷規(guī)則索引
        CREATE INDEX IF NOT EXISTS idx_promotions_active 
        ON promotion_rules(is_active, valid_from, valid_until);
        
        CREATE INDEX IF NOT EXISTS idx_promotions_conditions 
        ON promotion_rules USING GIN(conditions);
        
        -- 10. 物化視圖：產(chǎn)品統(tǒng)計
        CREATE MATERIALIZED VIEW IF NOT EXISTS product_statistics AS
        SELECT 
            p.id as product_id,
            p.name as product_name,
            p.category_id,
            p.base_price,
            
            -- 銷售統(tǒng)計
            COUNT(DISTINCT oi.order_id) as total_orders,
            SUM(oi.quantity) as total_units_sold,
            SUM(oi.item_total) as total_revenue,
            
            -- 庫存統(tǒng)計
            p.stock_quantity,
            p.reserved_quantity,
            p.stock_quantity - p.reserved_quantity as available_quantity,
            
            -- 評分統(tǒng)計
            COALESCE(AVG(pr.rating), 0) as average_rating,
            COUNT(pr.id) as review_count,
            
            -- 時間統(tǒng)計
            MAX(o.ordered_at) as last_sale_date
            
        FROM products p
        LEFT JOIN order_items oi ON p.id = oi.product_id
        LEFT JOIN orders_partitioned o ON oi.order_id = o.id
        LEFT JOIN product_reviews pr ON p.id = pr.product_id
        
        GROUP BY p.id, p.name, p.category_id, p.base_price,
                 p.stock_quantity, p.reserved_quantity
        
        WITH DATA;
        
        -- 物化視圖索引
        CREATE UNIQUE INDEX IF NOT EXISTS idx_product_stats_product 
        ON product_statistics(product_id);
        
        CREATE INDEX IF NOT EXISTS idx_product_stats_category 
        ON product_statistics(category_id);
        
        CREATE INDEX IF NOT EXISTS idx_product_stats_revenue 
        ON product_statistics(total_revenue DESC);
        
        -- 創(chuàng)建刷新物化視圖的函數(shù)
        CREATE OR REPLACE FUNCTION refresh_product_statistics()
        RETURNS TRIGGER AS $$
        BEGIN
            REFRESH MATERIALIZED VIEW CONCURRENTLY product_statistics;
            RETURN NULL;
        END;
        $$ LANGUAGE plpgsql;
        """
        
        try:
            # 分步執(zhí)行架構(gòu)創(chuàng)建
            statements = schema_sql.split(';')
            for statement in statements:
                statement = statement.strip()
                if statement:
                    self.cursor.execute(statement)
            
            self.conn.commit()
            logger.info("電商系統(tǒng)數(shù)據(jù)庫架構(gòu)創(chuàng)建成功")
            
        except Exception as e:
            self.conn.rollback()
            logger.error(f"創(chuàng)建架構(gòu)失敗: {e}")
            raise
    
    def search_products(
        self,
        query: Optional[str] = None,
        category_id: Optional[str] = None,
        min_price: Optional[float] = None,
        max_price: Optional[float] = None,
        brands: Optional[List[str]] = None,
        min_rating: Optional[float] = None,
        in_stock_only: bool = False,
        sort_by: str = 'relevance',
        limit: int = 20,
        offset: int = 0
    ) -> List[Dict[str, Any]]:
        """
        高級產(chǎn)品搜索
        
        Args:
            query: 搜索關(guān)鍵詞
            category_id: 分類ID
            min_price: 最低價格
            max_price: 最高價格
            brands: 品牌列表
            min_rating: 最低評分
            in_stock_only: 僅顯示有貨商品
            sort_by: 排序方式
            limit: 返回數(shù)量
            offset: 偏移量
            
        Returns:
            產(chǎn)品列表
        """
        search_sql = """
        SELECT 
            p.id,
            p.sku,
            p.name,
            p.description,
            p.brand,
            p.base_price,
            p.discount_price,
            p.stock_quantity,
            p.reserved_quantity,
            p.attributes,
            
            -- 計算可用庫存
            p.stock_quantity - p.reserved_quantity as available_quantity,
            
            -- 計算折扣率
            CASE 
                WHEN p.discount_price IS NOT NULL 
                THEN ROUND((1 - p.discount_price / p.base_price) * 100, 1)
                ELSE 0
            END as discount_percentage,
            
            -- 獲取評分信息
            COALESCE(ps.average_rating, 0) as average_rating,
            COALESCE(ps.review_count, 0) as review_count,
            
            -- 計算相關(guān)性得分（如果有關(guān)鍵詞）
            {relevance_score}
            
        FROM products p
        LEFT JOIN product_statistics ps ON p.id = ps.product_id
        
        WHERE 1=1
            {search_condition}
            {category_condition}
            {price_condition}
            {brand_condition}
            {rating_condition}
            {stock_condition}
            
        {order_clause}
        
        LIMIT %s OFFSET %s
        """
        
        # 構(gòu)建查詢條件
        conditions = []
        params = []
        
        # 全文搜索條件
        if query:
            conditions.append("p.search_vector @@ plainto_tsquery('english', %s)")
            params.append(query)
        
        # 分類條件
        if category_id:
            # 獲取分類及其所有子分類
            subcategories = self._get_all_subcategories(category_id)
            if subcategories:
                placeholders = ', '.join(['%s'] * len(subcategories))
                conditions.append(f"p.category_id IN ({placeholders})")
                params.extend(subcategories)
        
        # 價格條件
        if min_price is not None:
            conditions.append("p.base_price >= %s")
            params.append(min_price)
        if max_price is not None:
            conditions.append("p.base_price <= %s")
            params.append(max_price)
        
        # 品牌條件
        if brands:
            placeholders = ', '.join(['%s'] * len(brands))
            conditions.append(f"p.brand IN ({placeholders})")
            params.extend(brands)
        
        # 評分條件
        if min_rating is not None:
            conditions.append("COALESCE(ps.average_rating, 0) >= %s")
            params.append(min_rating)
        
        # 庫存條件
        if in_stock_only:
            conditions.append("(p.stock_quantity - p.reserved_quantity) > 0")
        
        # 構(gòu)建相關(guān)性得分計算
        relevance_score = ""
        if query:
            relevance_score = """
            , ts_rank(
                p.search_vector, 
                plainto_tsquery('english', %s)
            ) as relevance_score
            """
        
        # 構(gòu)建排序子句
        order_clause_map = {
            'relevance': "ORDER BY relevance_score DESC NULLS LAST",
            'price_asc': "ORDER BY p.base_price ASC",
            'price_desc': "ORDER BY p.base_price DESC",
            'rating': "ORDER BY ps.average_rating DESC NULLS LAST",
            'popularity': "ORDER BY ps.total_units_sold DESC NULLS LAST",
            'newest': "ORDER BY p.created_at DESC"
        }
        
        order_clause = order_clause_map.get(sort_by, "ORDER BY p.created_at DESC")
        
        # 格式化SQL
        formatted_sql = search_sql.format(
            relevance_score=relevance_score,
            search_condition=f"AND {' AND '.join(conditions)}" if conditions else "",
            category_condition="",
            price_condition="",
            brand_condition="",
            rating_condition="",
            stock_condition="",
            order_clause=order_clause
        )
        
        # 添加分頁參數(shù)
        params.extend([limit, offset])
        
        try:
            self.cursor.execute(formatted_sql, params)
            products = self.cursor.fetchall()
            
            return [
                {
                    'id': str(product['id']),
                    'sku': product['sku'],
                    'name': product['name'],
                    'brand': product['brand'],
                    'base_price': float(product['base_price']),
                    'discount_price': float(product['discount_price']) if product['discount_price'] else None,
                    'available_quantity': product['available_quantity'],
                    'discount_percentage': product['discount_percentage'],
                    'average_rating': float(product['average_rating']),
                    'review_count': product['review_count'],
                    'attributes': product['attributes']
                }
                for product in products
            ]
            
        except Exception as e:
            logger.error(f"產(chǎn)品搜索失敗: {e}")
            return []
    
    def _get_all_subcategories(self, category_id: str) -> List[str]:
        """
        獲取分類及其所有子分類
        
        Args:
            category_id: 分類ID
            
        Returns:
            子分類ID列表
        """
        recursive_sql = """
        WITH RECURSIVE category_tree AS (
            -- 基礎(chǔ)分類
            SELECT id, parent_id
            FROM categories
            WHERE id = %s
            
            UNION ALL
            
            -- 遞歸獲取子分類
            SELECT c.id, c.parent_id
            FROM categories c
            INNER JOIN category_tree ct ON c.parent_id = ct.id
        )
        SELECT id FROM category_tree
        """
        
        try:
            self.cursor.execute(recursive_sql, (category_id,))
            results = self.cursor.fetchall()
            return [str(row['id']) for row in results]
        except Exception as e:
            logger.error(f"獲取子分類失敗: {e}")
            return []
    
    def get_product_recommendations(
        self, 
        product_id: str, 
        customer_id: Optional[str] = None,
        limit: int = 10
    ) -> List[Dict[str, Any]]:
        """
        獲取產(chǎn)品推薦
        
        Args:
            product_id: 產(chǎn)品ID
            customer_id: 客戶ID（可選）
            limit: 返回數(shù)量
            
        Returns:
            推薦產(chǎn)品列表
        """
        recommendations_sql = """
        -- 基于多種策略的混合推薦
        
        (
            -- 策略1: 同品牌產(chǎn)品
            SELECT 
                p.id,
                p.name,
                p.brand,
                p.base_price,
                'same_brand' as recommendation_reason,
                0.7 as recommendation_score
            FROM products p
            WHERE p.brand = (
                SELECT brand FROM products WHERE id = %s
            )
            AND p.id != %s
            AND (p.stock_quantity - p.reserved_quantity) > 0
            LIMIT 3
        )
        
        UNION ALL
        
        (
            -- 策略2: 同分類熱門產(chǎn)品
            SELECT 
                p.id,
                p.name,
                p.brand,
                p.base_price,
                'popular_in_category' as recommendation_reason,
                ps.total_units_sold::float / 
                    (SELECT MAX(total_units_sold) FROM product_statistics) as recommendation_score
            FROM products p
            JOIN product_statistics ps ON p.id = ps.product_id
            WHERE p.category_id = (
                SELECT category_id FROM products WHERE id = %s
            )
            AND p.id != %s
            AND (p.stock_quantity - p.reserved_quantity) > 0
            ORDER BY ps.total_units_sold DESC
            LIMIT 3
        )
        
        UNION ALL
        
        (
            -- 策略3: 經(jīng)常一起購買的產(chǎn)品
            SELECT 
                p.id,
                p.name,
                p.brand,
                p.base_price,
                'frequently_bought_together' as recommendation_reason,
                COUNT(DISTINCT oi2.order_id)::float / 
                    (SELECT COUNT(DISTINCT oi3.order_id) 
                     FROM order_items oi3 
                     WHERE oi3.product_id = %s) as recommendation_score
            FROM order_items oi1
            JOIN order_items oi2 ON oi1.order_id = oi2.order_id
            JOIN products p ON oi2.product_id = p.id
            WHERE oi1.product_id = %s
            AND oi2.product_id != %s
            AND (p.stock_quantity - p.reserved_quantity) > 0
            GROUP BY p.id, p.name, p.brand, p.base_price
            HAVING COUNT(DISTINCT oi2.order_id) >= 2
            ORDER BY recommendation_score DESC
            LIMIT 3
        )
        
        {customer_based_recommendations}
        
        ORDER BY recommendation_score DESC
        LIMIT %s
        """
        
        # 如果有客戶ID，添加個性化推薦
        customer_recommendations = ""
        if customer_id:
            customer_recommendations = """
            UNION ALL
            
            (
                -- 策略4: 基于客戶購買歷史的推薦
                SELECT 
                    p.id,
                    p.name,
                    p.brand,
                    p.base_price,
                    'based_on_your_purchases' as recommendation_reason,
                    COUNT(DISTINCT oi.order_id)::float / 10 as recommendation_score
                FROM order_items oi
                JOIN products p ON oi.product_id = p.id
                WHERE oi.order_id IN (
                    SELECT order_id 
                    FROM order_items 
                    WHERE product_id = %s
                )
                AND p.id != %s
                AND (p.stock_quantity - p.reserved_quantity) > 0
                GROUP BY p.id, p.name, p.brand, p.base_price
                HAVING COUNT(DISTINCT oi.order_id) >= 1
                ORDER BY recommendation_score DESC
                LIMIT 2
            )
            """
        
        # 格式化SQL
        formatted_sql = recommendations_sql.format(
            customer_based_recommendations=customer_recommendations
        )
        
        # 準(zhǔn)備參數(shù)
        params = [product_id, product_id, product_id, product_id, 
                 product_id, product_id, product_id]
        
        if customer_id:
            params.extend([customer_id, customer_id, product_id, product_id])
        
        params.append(limit)
        
        try:
            self.cursor.execute(formatted_sql, params)
            recommendations = self.cursor.fetchall()
            
            return [
                {
                    'id': str(rec['id']),
                    'name': rec['name'],
                    'brand': rec['brand'],
                    'price': float(rec['base_price']),
                    'recommendation_reason': rec['recommendation_reason'],
                    'recommendation_score': float(rec['recommendation_score'])
                }
                for rec in recommendations
            ]
            
        except Exception as e:
            logger.error(f"獲取推薦失敗: {e}")
            return []
    
    def create_order(
        self,
        customer_id: str,
        items: List[Dict[str, Any]],
        shipping_address: Dict[str, Any],
        promotion_code: Optional[str] = None
    ) -> Optional[Dict[str, Any]]:
        """
        創(chuàng)建訂單
        
        Args:
            customer_id: 客戶ID
            items: 訂單項列表
            shipping_address: 配送地址
            promotion_code: 促銷代碼
            
        Returns:
            創(chuàng)建的訂單信息
        """
        try:
            # 開始事務(wù)
            self.conn.autocommit = False
            
            # 生成訂單號
            order_number = f"ORD-{datetime.now().strftime('%Y%m%d')}-{self._generate_order_suffix()}"
            
            # 計算訂單金額
            order_calculation = self._calculate_order_amounts(items, promotion_code)
            
            # 插入訂單
            order_sql = """
            INSERT INTO orders_partitioned (
                order_number, customer_id, status, order_metadata,
                subtotal, tax_amount, shipping_amount, 
                discount_amount, total_amount, ordered_at
            )
            VALUES (%s, %s, 'PENDING', %s, %s, %s, %s, %s, %s, CURRENT_TIMESTAMP)
            RETURNING id, order_number, total_amount
            """
            
            self.cursor.execute(
                order_sql,
                (
                    order_number,
                    customer_id,
                    Json({
                        'shipping_address': shipping_address,
                        'promotion_code': promotion_code
                    }),
                    order_calculation['subtotal'],
                    order_calculation['tax_amount'],
                    order_calculation['shipping_amount'],
                    order_calculation['discount_amount'],
                    order_calculation['total_amount']
                )
            )
            
            order_result = self.cursor.fetchone()
            order_id = order_result['id']
            
            # 插入訂單項
            for item in items:
                # 獲取產(chǎn)品快照
                product_snapshot = self._get_product_snapshot(item['product_id'])
                
                # 插入訂單項
                item_sql = """
                INSERT INTO order_items (
                    order_id, product_id, variant_id,
                    product_snapshot, quantity, unit_price,
                    discount_percentage, item_total
                )
                VALUES (%s, %s, %s, %s, %s, %s, %s, %s)
                """
                
                self.cursor.execute(
                    item_sql,
                    (
                        order_id,
                        item['product_id'],
                        item.get('variant_id'),
                        Json(product_snapshot),
                        item['quantity'],
                        item['unit_price'],
                        item.get('discount_percentage', 0),
                        item['item_total']
                    )
                )
                
                # 更新庫存
                self._update_inventory(
                    product_id=item['product_id'],
                    variant_id=item.get('variant_id'),
                    quantity_change=-item['quantity'],
                    reference_id=order_id,
                    reference_type='ORDER',
                    change_type='SALE'
                )
            
            # 提交事務(wù)
            self.conn.commit()
            
            return {
                'order_id': str(order_id),
                'order_number': order_result['order_number'],
                'total_amount': float(order_result['total_amount']),
                'items_count': len(items)
            }
            
        except Exception as e:
            self.conn.rollback()
            logger.error(f"創(chuàng)建訂單失敗: {e}")
            return None
    
    def _calculate_order_amounts(
        self,
        items: List[Dict[str, Any]],
        promotion_code: Optional[str] = None
    ) -> Dict[str, float]:
        """計算訂單金額"""
        subtotal = sum(item['item_total'] for item in items)
        
        # 應(yīng)用促銷折扣
        discount_amount = 0
        if promotion_code:
            # 這里可以添加促銷邏輯
            pass
        
        # 計算稅費（簡化示例）
        tax_amount = subtotal * 0.1  # 10%稅率
        
        # 計算運費（簡化示例）
        shipping_amount = 5.99 if subtotal < 50 else 0
        
        # 計算總額
        total_amount = subtotal + tax_amount + shipping_amount - discount_amount
        
        return {
            'subtotal': subtotal,
            'tax_amount': tax_amount,
            'shipping_amount': shipping_amount,
            'discount_amount': discount_amount,
            'total_amount': total_amount
        }
    
    def _get_product_snapshot(self, product_id: str) -> Dict[str, Any]:
        """獲取產(chǎn)品快照"""
        snapshot_sql = """
        SELECT 
            id, sku, name, description, brand,
            base_price, attributes
        FROM products
        WHERE id = %s
        """
        
        self.cursor.execute(snapshot_sql, (product_id,))
        product = self.cursor.fetchone()
        
        return {
            'product_id': str(product['id']),
            'sku': product['sku'],
            'name': product['name'],
            'brand': product['brand'],
            'price_at_time_of_purchase': float(product['base_price']),
            'attributes': product['attributes']
        }
    
    def _update_inventory(
        self,
        product_id: str,
        variant_id: Optional[str],
        quantity_change: int,
        reference_id: str,
        reference_type: str,
        change_type: str
    ):
        """更新庫存"""
        # 獲取當(dāng)前庫存
        if variant_id:
            stock_sql = """
            SELECT variant_stock as current_quantity
            FROM product_variants
            WHERE id = %s
            """
            self.cursor.execute(stock_sql, (variant_id,))
        else:
            stock_sql = """
            SELECT stock_quantity as current_quantity
            FROM products
            WHERE id = %s
            """
            self.cursor.execute(stock_sql, (product_id,))
        
        result = self.cursor.fetchone()
        if not result:
            raise Exception("產(chǎn)品不存在")
        
        current_quantity = result['current_quantity']
        new_quantity = current_quantity + quantity_change
        
        if new_quantity < 0:
            raise Exception("庫存不足")
        
        # 更新庫存
        if variant_id:
            update_sql = """
            UPDATE product_variants
            SET variant_stock = %s
            WHERE id = %s
            """
            self.cursor.execute(update_sql, (new_quantity, variant_id))
        else:
            update_sql = """
            UPDATE products
            SET stock_quantity = %s
            WHERE id = %s
            """
            self.cursor.execute(update_sql, (new_quantity, product_id))
        
        # 記錄庫存變更
        log_sql = """
        INSERT INTO inventory_logs (
            product_id, variant_id, change_type,
            quantity_change, previous_quantity, new_quantity,
            reference_id, reference_type
        )
        VALUES (%s, %s, %s, %s, %s, %s, %s, %s)
        """
        
        self.cursor.execute(
            log_sql,
            (
                product_id,
                variant_id,
                change_type,
                quantity_change,
                current_quantity,
                new_quantity,
                reference_id,
                reference_type
            )
        )
    
    def _generate_order_suffix(self) -> str:
        """生成訂單號后綴"""
        import random
        import string
        
        # 生成6位隨機(jī)字母數(shù)字
        return ''.join(random.choices(
            string.ascii_uppercase + string.digits, k=6
        ))
    
    def refresh_materialized_views(self):
        """刷新物化視圖"""
        refresh_sql = """
        REFRESH MATERIALIZED VIEW CONCURRENTLY product_statistics;
        """
        
        try:
            self.cursor.execute(refresh_sql)
            self.conn.commit()
            logger.info("物化視圖刷新成功")
        except Exception as e:
            self.conn.rollback()
            logger.error(f"刷新物化視圖失敗: {e}")
    
    def close(self):
        """關(guān)閉連接"""
        if self.cursor:
            self.cursor.close()
        if self.conn:
            self.conn.close()
        logger.info("數(shù)據(jù)庫連接已關(guān)閉")


def demonstrate_ecommerce_features():
    """演示電商系統(tǒng)高級特性"""
    dsn = "dbname=ecommerce user=postgres password=password host=localhost port=5432"
    
    print("電商系統(tǒng)PostgreSQL高級特性演示")
    print("=" * 60)
    
    # 創(chuàng)建數(shù)據(jù)庫實例
    db = ECommerceDatabase(dsn)
    
    try:
        # 1. 創(chuàng)建架構(gòu)
        print("\n1. 創(chuàng)建數(shù)據(jù)庫架構(gòu)...")
        db.create_schema()
        print("   架構(gòu)創(chuàng)建完成")
        
        # 2. 演示搜索功能
        print("\n2. 演示高級產(chǎn)品搜索...")
        products = db.search_products(
            query="wireless headphone",
            min_price=50,
            max_price=200,
            min_rating=4.0,
            in_stock_only=True,
            sort_by='rating',
            limit=5
        )
        
        print(f"   找到 {len(products)} 個產(chǎn)品:")
        for product in products:
            print(f"   - {product['name']} (評分: {product['average_rating']:.1f}, "
                  f"價格: ${product['base_price']:.2f})")
        
        # 3. 演示推薦系統(tǒng)
        if products:
            print("\n3. 演示產(chǎn)品推薦系統(tǒng)...")
            recommendations = db.get_product_recommendations(
                product_id=products[0]['id'],
                limit=5
            )
            
            print(f"   為 '{products[0]['name']}' 的推薦:")
            for rec in recommendations:
                print(f"   - {rec['name']} (原因: {rec['recommendation_reason']}, "
                      f"得分: {rec['recommendation_score']:.2f})")
        
        # 4. 演示訂單創(chuàng)建
        print("\n4. 演示訂單創(chuàng)建...")
        # 這里需要實際的產(chǎn)品數(shù)據(jù)，所以只是演示代碼結(jié)構(gòu)
        print("   訂單創(chuàng)建功能就緒")
        
        # 5. 刷新物化視圖
        print("\n5. 刷新物化視圖...")
        db.refresh_materialized_views()
        print("   物化視圖刷新完成")
        
        print("\n演示完成!")
        
    except Exception as e:
        print(f"演示過程中出錯: {e}")
    finally:
        db.close()


if __name__ == "__main__":
    demonstrate_ecommerce_features()

"""
PostgreSQL高級特性與性能優(yōu)化完整示例
"""
import psycopg2
from psycopg2.extras import DictCursor, Json
import time
from datetime import datetime, timedelta
from typing import Dict, List, Any, Optional
import json
import logging

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


class PostgreSQLAdvancedOptimizer:
    """PostgreSQL高級優(yōu)化器"""
    
    def __init__(self, dsn: str):
        """
        初始化優(yōu)化器
        
        Args:
            dsn: 數(shù)據(jù)庫連接字符串
        """
        self.dsn = dsn
        self.conn = psycopg2.connect(dsn)
        self.cursor = self.conn.cursor(cursor_factory=DictCursor)
    
    def analyze_database_health(self) -> Dict[str, Any]:
        """
        全面分析數(shù)據(jù)庫健康狀態(tài)
        
        Returns:
            數(shù)據(jù)庫健康報告
        """
        health_report = {
            'timestamp': datetime.now().isoformat(),
            'overall_score': 0,
            'categories': {},
            'issues': [],
            'recommendations': []
        }
        
        # 分析各個維度
        categories = [
            ('連接與并發(fā)', self._analyze_connections),
            ('查詢性能', self._analyze_query_performance),
            ('索引效率', self._analyze_index_efficiency),
            ('緩存性能', self._analyze_cache_performance),
            ('存儲效率', self._analyze_storage_efficiency),
            ('配置優(yōu)化', self._analyze_configuration)
        ]
        
        total_score = 0
        category_count = 0
        
        for category_name, analyzer_func in categories:
            try:
                category_result = analyzer_func()
                health_report['categories'][category_name] = category_result
                
                if 'score' in category_result:
                    total_score += category_result['score']
                    category_count += 1
                
                if 'issues' in category_result:
                    health_report['issues'].extend(category_result['issues'])
                
                if 'recommendations' in category_result:
                    health_report['recommendations'].extend(category_result['recommendations'])
                    
            except Exception as e:
                logger.error(f"分析{category_name}失敗: {e}")
                health_report['issues'].append({
                    'category': category_name,
                    'severity': 'ERROR',
                    'message': f'分析失敗: {str(e)}'
                })
        
        # 計算總體評分
        if category_count > 0:
            health_report['overall_score'] = total_score / category_count
        
        # 排序問題和建議
        health_report['issues'] = sorted(
            health_report['issues'], 
            key=lambda x: {'CRITICAL': 3, 'HIGH': 2, 'MEDIUM': 1, 'LOW': 0}.get(x.get('severity', 'LOW'), 0),
            reverse=True
        )
        
        health_report['recommendations'] = sorted(
            health_report['recommendations'],
            key=lambda x: x.get('priority', 3),
            reverse=False
        )
        
        return health_report
    
    def _analyze_connections(self) -> Dict[str, Any]:
        """分析連接與并發(fā)"""
        analysis = {
            'score': 100,
            'metrics': {},
            'issues': [],
            'recommendations': []
        }
        
        try:
            # 獲取連接統(tǒng)計
            self.cursor.execute("""
                SELECT 
                    COUNT(*) as total_connections,
                    COUNT(*) FILTER (WHERE state = 'active') as active_connections,
                    COUNT(*) FILTER (WHERE state = 'idle') as idle_connections,
                    COUNT(*) FILTER (WHERE wait_event_type IS NOT NULL) as waiting_connections,
                    MAX(now() - backend_start) as oldest_connection_age
                FROM pg_stat_activity
                WHERE pid <> pg_backend_pid()
            """)
            conn_stats = self.cursor.fetchone()
            
            # 獲取配置參數(shù)
            self.cursor.execute("""
                SELECT setting::integer as max_connections
                FROM pg_settings
                WHERE name = 'max_connections'
            """)
            max_conns = self.cursor.fetchone()['max_connections']
            
            # 計算連接使用率
            conn_usage = conn_stats['total_connections'] / max_conns
            analysis['metrics'] = {
                'total_connections': conn_stats['total_connections'],
                'active_connections': conn_stats['active_connections'],
                'idle_connections': conn_stats['idle_connections'],
                'waiting_connections': conn_stats['waiting_connections'],
                'connection_usage_percentage': round(conn_usage * 100, 1),
                'max_connections': max_conns
            }
            
            # 分析問題
            if conn_usage > 0.8:
                analysis['score'] -= 30
                analysis['issues'].append({
                    'severity': 'HIGH',
                    'message': f'連接使用率過高: {conn_usage:.1%}',
                    'details': '接近最大連接數(shù)限制'
                })
                analysis['recommendations'].append({
                    'priority': 1,
                    'action': '考慮增加max_connections參數(shù)或優(yōu)化連接池配置'
                })
            
            if conn_stats['waiting_connections'] > 5:
                analysis['score'] -= 20
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'等待連接數(shù)較多: {conn_stats["waiting_connections"]}',
                    'details': '可能存在鎖爭用或資源競爭'
                })
            
            if conn_stats['oldest_connection_age'] and \
               conn_stats['oldest_connection_age'].total_seconds() > 3600:
                analysis['score'] -= 10
                analysis['issues'].append({
                    'severity': 'LOW',
                    'message': f'存在長時間連接: {conn_stats["oldest_connection_age"]}',
                    'details': '考慮優(yōu)化連接生命周期'
                })
            
        except Exception as e:
            logger.error(f"連接分析失敗: {e}")
            analysis['score'] = 0
        
        return analysis
    
    def _analyze_query_performance(self) -> Dict[str, Any]:
        """分析查詢性能"""
        analysis = {
            'score': 100,
            'metrics': {},
            'issues': [],
            'recommendations': []
        }
        
        try:
            # 獲取慢查詢統(tǒng)計
            self.cursor.execute("""
                WITH query_stats AS (
                    SELECT 
                        query,
                        calls,
                        total_time,
                        mean_time,
                        rows,
                        100.0 * shared_blks_hit / nullif(shared_blks_hit + shared_blks_read, 0) as buffer_hit_rate,
                        ROW_NUMBER() OVER (ORDER BY total_time DESC) as time_rank
                    FROM pg_stat_statements
                    WHERE query NOT LIKE '%pg_stat_statements%'
                    AND calls > 0
                )
                SELECT 
                    COUNT(*) as total_queries,
                    SUM(calls) as total_calls,
                    SUM(total_time) as total_time_ms,
                    AVG(mean_time) as avg_query_time_ms,
                    PERCENTILE_CONT(0.95) WITHIN GROUP (ORDER BY mean_time) as p95_query_time_ms,
                    SUM(CASE WHEN mean_time > 100 THEN 1 ELSE 0 END) as slow_queries_count,
                    AVG(buffer_hit_rate) as avg_buffer_hit_rate
                FROM query_stats
            """)
            query_stats = self.cursor.fetchone()
            
            analysis['metrics'] = {
                'total_queries': query_stats['total_queries'],
                'total_calls': query_stats['total_calls'],
                'total_time_seconds': round(query_stats['total_time_ms'] / 1000, 1),
                'avg_query_time_ms': round(query_stats['avg_query_time_ms'], 2),
                'p95_query_time_ms': round(query_stats['p95_query_time_ms'], 2),
                'slow_queries_count': query_stats['slow_queries_count'],
                'avg_buffer_hit_rate': round(query_stats['avg_buffer_hit_rate'], 1)
            }
            
            # 分析問題
            if query_stats['avg_query_time_ms'] > 50:
                analysis['score'] -= 20
                analysis['issues'].append({
                    'severity': 'HIGH',
                    'message': f'平均查詢時間較高: {query_stats["avg_query_time_ms"]:.2f}ms',
                    'details': '可能存在查詢優(yōu)化空間'
                })
                analysis['recommendations'].append({
                    'priority': 1,
                    'action': '分析并優(yōu)化最耗時的查詢'
                })
            
            if query_stats['slow_queries_count'] > 10:
                analysis['score'] -= 15
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'發(fā)現(xiàn) {query_stats["slow_queries_count"]} 個慢查詢',
                    'details': '定義: 平均執(zhí)行時間 > 100ms'
                })
            
            if query_stats['avg_buffer_hit_rate'] < 90:
                analysis['score'] -= 10
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'平均緩沖區(qū)命中率較低: {query_stats["avg_buffer_hit_rate"]:.1f}%',
                    'details': '建議增加shared_buffers或優(yōu)化查詢'
                })
            
        except Exception as e:
            logger.error(f"查詢性能分析失敗: {e}")
            analysis['score'] = 0
        
        return analysis
    
    def _analyze_index_efficiency(self) -> Dict[str, Any]:
        """分析索引效率"""
        analysis = {
            'score': 100,
            'metrics': {},
            'issues': [],
            'recommendations': []
        }
        
        try:
            # 獲取索引使用統(tǒng)計
            self.cursor.execute("""
                WITH index_stats AS (
                    SELECT 
                        schemaname,
                        tablename,
                        indexname,
                        idx_scan as index_scans,
                        idx_tup_read as tuples_read,
                        idx_tup_fetch as tuples_fetched,
                        pg_relation_size(indexname::regclass) as index_size_bytes,
                        CASE 
                            WHEN idx_scan = 0 THEN 0
                            ELSE idx_tup_fetch::float / idx_tup_read * 100
                        END as index_efficiency
                    FROM pg_stat_user_indexes
                    WHERE schemaname NOT LIKE 'pg_%'
                )
                SELECT 
                    COUNT(*) as total_indexes,
                    SUM(index_size_bytes) as total_index_size_bytes,
                    SUM(CASE WHEN index_scans = 0 THEN 1 ELSE 0 END) as unused_indexes_count,
                    AVG(index_efficiency) as avg_index_efficiency,
                    PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY index_efficiency) as median_index_efficiency
                FROM index_stats
            """)
            index_stats = self.cursor.fetchone()
            
            analysis['metrics'] = {
                'total_indexes': index_stats['total_indexes'],
                'total_index_size_gb': round(index_stats['total_index_size_bytes'] / (1024**3), 2),
                'unused_indexes_count': index_stats['unused_indexes_count'],
                'avg_index_efficiency': round(index_stats['avg_index_efficiency'], 1),
                'median_index_efficiency': round(index_stats['median_index_efficiency'], 1)
            }
            
            # 分析問題
            unused_ratio = index_stats['unused_indexes_count'] / index_stats['total_indexes'] \
                if index_stats['total_indexes'] > 0 else 0
            
            if unused_ratio > 0.1:
                analysis['score'] -= 25
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'未使用索引比例較高: {unused_ratio:.1%}',
                    'details': f'{index_stats["unused_indexes_count"]} 個索引從未使用'
                })
                analysis['recommendations'].append({
                    'priority': 2,
                    'action': '考慮刪除未使用的索引以節(jié)省空間和提高寫入性能'
                })
            
            if index_stats['avg_index_efficiency'] < 50:
                analysis['score'] -= 15
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'平均索引效率較低: {index_stats["avg_index_efficiency"]:.1f}%',
                    'details': '索引可能不是最優(yōu)選擇'
                })
            
        except Exception as e:
            logger.error(f"索引效率分析失敗: {e}")
            analysis['score'] = 0
        
        return analysis
    
    def _analyze_cache_performance(self) -> Dict[str, Any]:
        """分析緩存性能"""
        analysis = {
            'score': 100,
            'metrics': {},
            'issues': [],
            'recommendations': []
        }
        
        try:
            # 獲取緩存統(tǒng)計
            self.cursor.execute("""
                SELECT 
                    -- 共享緩沖區(qū)命中率
                    CASE 
                        WHEN blks_hit + blks_read > 0 
                        THEN blks_hit::float / (blks_hit + blks_read) * 100
                        ELSE 0 
                    END as shared_buffer_hit_rate,
                    
                    -- TOAST緩沖區(qū)命中率
                    CASE 
                        WHEN toast_blks_hit + toast_blks_read > 0 
                        THEN toast_blks_hit::float / (toast_blks_hit + toast_blks_read) * 100
                        ELSE 0 
                    END as toast_buffer_hit_rate,
                    
                    -- 臨時文件使用
                    temp_files,
                    temp_bytes,
                    
                    -- 檢查點統(tǒng)計
                    checkpoints_timed,
                    checkpoints_req,
                    checkpoint_write_time,
                    checkpoint_sync_time
                    
                FROM pg_stat_bgwriter
            """)
            cache_stats = self.cursor.fetchone()
            
            # 獲取緩沖區(qū)配置
            self.cursor.execute("""
                SELECT 
                    name,
                    setting,
                    unit
                FROM pg_settings
                WHERE name IN ('shared_buffers', 'effective_cache_size')
            """)
            cache_config = {row['name']: row for row in self.cursor.fetchall()}
            
            analysis['metrics'] = {
                'shared_buffer_hit_rate': round(cache_stats['shared_buffer_hit_rate'], 1),
                'toast_buffer_hit_rate': round(cache_stats['toast_buffer_hit_rate'], 1),
                'temp_files_count': cache_stats['temp_files'],
                'temp_files_size_gb': round(cache_stats['temp_bytes'] / (1024**3), 2),
                'shared_buffers': cache_config.get('shared_buffers', {}).get('setting', 'N/A'),
                'effective_cache_size': cache_config.get('effective_cache_size', {}).get('setting', 'N/A')
            }
            
            # 分析問題
            if cache_stats['shared_buffer_hit_rate'] < 90:
                analysis['score'] -= 20
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'共享緩沖區(qū)命中率較低: {cache_stats["shared_buffer_hit_rate"]:.1f}%',
                    'details': '建議增加shared_buffers或優(yōu)化工作集'
                })
                analysis['recommendations'].append({
                    'priority': 2,
                    'action': '考慮增加shared_buffers參數(shù)'
                })
            
            if cache_stats['temp_files'] > 100:
                analysis['score'] -= 15
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'臨時文件使用較多: {cache_stats["temp_files"]} 個文件',
                    'details': '可能存在排序或哈希操作溢出到磁盤'
                })
                analysis['recommendations'].append({
                    'priority': 2,
                    'action': '增加work_mem參數(shù)以減少臨時文件使用'
                })
            
        except Exception as e:
            logger.error(f"緩存性能分析失敗: {e}")
            analysis['score'] = 0
        
        return analysis
    
    def _analyze_storage_efficiency(self) -> Dict[str, Any]:
        """分析存儲效率"""
        analysis = {
            'score': 100,
            'metrics': {},
            'issues': [],
            'recommendations': []
        }
        
        try:
            # 獲取表膨脹信息
            self.cursor.execute("""
                SELECT 
                    schemaname,
                    tablename,
                    n_dead_tup as dead_tuples,
                    n_live_tup as live_tuples,
                    CASE 
                        WHEN n_live_tup > 0 
                        THEN n_dead_tup::float / n_live_tup * 100
                        ELSE 0 
                    END as dead_tuple_ratio,
                    last_vacuum,
                    last_autovacuum,
                    last_analyze,
                    last_autoanalyze
                FROM pg_stat_user_tables
                WHERE schemaname NOT LIKE 'pg_%'
                ORDER BY dead_tuple_ratio DESC
                LIMIT 10
            """)
            table_stats = self.cursor.fetchall()
            
            # 獲取數(shù)據(jù)庫大小
            self.cursor.execute("""
                SELECT 
                    pg_database_size(current_database()) as database_size_bytes,
                    pg_size_pretty(pg_database_size(current_database())) as database_size_pretty
            """)
            db_size = self.cursor.fetchone()
            
            analysis['metrics'] = {
                'database_size': db_size['database_size_pretty'],
                'tables_analyzed': len(table_stats),
                'top_tables_by_dead_tuples': [
                    {
                        'table': f"{row['schemaname']}.{row['tablename']}",
                        'dead_tuple_ratio': round(row['dead_tuple_ratio'], 1),
                        'dead_tuples': row['dead_tuples'],
                        'last_vacuum': row['last_vacuum']
                    }
                    for row in table_stats[:5]
                ]
            }
            
            # 分析問題
            high_dead_tables = [
                row for row in table_stats 
                if row['dead_tuple_ratio'] > 20
            ]
            
            if high_dead_tables:
                analysis['score'] -= 25
                analysis['issues'].append({
                    'severity': 'MEDIUM',
                    'message': f'發(fā)現(xiàn) {len(high_dead_tables)} 個表死元組比例超過20%',
                    'details': '可能導(dǎo)致查詢性能下降和存儲空間浪費'
                })
                analysis['recommendations'].append({
                    'priority': 2,
                    'action': '對高死元組比例的表執(zhí)行VACUUM操作'
                })
            
        except Exception as e:
            logger.error(f"存儲效率分析失敗: {e}")
            analysis['score'] = 0
        
        return analysis
    
    def _analyze_configuration(self) -> Dict[str, Any]:
        """分析配置優(yōu)化"""
        analysis = {
            'score': 100,
            'metrics': {},
            'issues': [],
            'recommendations': []
        }
        
        try:
            # 獲取關(guān)鍵配置參數(shù)
            self.cursor.execute("""
                SELECT 
                    name,
                    setting,
                    unit,
                    context,
                    vartype,
                    source
                FROM pg_settings
                WHERE name IN (
                    'shared_buffers',
                    'work_mem',
                    'maintenance_work_mem',
                    'effective_cache_size',
                    'max_connections',
                    'checkpoint_timeout',
                    'checkpoint_completion_target',
                    'wal_buffers',
                    'random_page_cost',
                    'seq_page_cost',
                    'effective_io_concurrency'
                )
                ORDER BY name
            """)
            configs = {row['name']: row for row in self.cursor.fetchall()}
            
            analysis['metrics'] = {
                'key_parameters': configs
            }
            
            # 檢查配置合理性
            issues = []
            recommendations = []
            
            # 檢查shared_buffers（應(yīng)為系統(tǒng)內(nèi)存的25%）
            if 'shared_buffers' in configs:
                shared_buffers = configs['shared_buffers']['setting']
                if shared_buffers.endswith('MB'):
                    mb_value = int(shared_buffers[:-2])
                    if mb_value < 128:  # 小于128MB
                        issues.append('shared_buffers設(shè)置可能過小')
                        recommendations.append('考慮增加shared_buffers到系統(tǒng)內(nèi)存的25%')
            
            # 檢查work_mem
            if 'work_mem' in configs:
                work_mem = configs['work_mem']['setting']
                if work_mem.endswith('kB'):
                    kb_value = int(work_mem[:-2])
                    if kb_value < 4096:  # 小于4MB
                        issues.append('work_mem設(shè)置可能過小')
                        recommendations.append('適當(dāng)增加work_mem以減少臨時文件使用')
            
            # 檢查checkpoint配置
            if 'checkpoint_timeout' in configs:
                checkpoint_timeout = int(configs['checkpoint_timeout']['setting'])
                if checkpoint_timeout > 900:  # 超過15分鐘
                    issues.append('checkpoint_timeout設(shè)置過長')
                    recommendations.append('考慮減少checkpoint_timeout以降低恢復(fù)時間')
            
            if issues:
                analysis['score'] -= len(issues) * 10
                for issue in issues:
                    analysis['issues'].append({
                        'severity': 'MEDIUM',
                        'message': issue
                    })
                
                for rec in recommendations:
                    analysis['recommendations'].append({
                        'priority': 3,
                        'action': rec
                    })
            
        except Exception as e:
            logger.error(f"配置分析失敗: {e}")
            analysis['score'] = 0
        
        return analysis
    
    def generate_optimization_report(self) -> Dict[str, Any]:
        """
        生成優(yōu)化報告
        
        Returns:
            優(yōu)化報告
        """
        health_report = self.analyze_database_health()
        
        report = {
            'summary': {
                'overall_score': health_report['overall_score'],
                'assessment': self._get_assessment(health_report['overall_score']),
                'total_issues': len(health_report['issues']),
                'total_recommendations': len(health_report['recommendations'])
            },
            'detailed_analysis': health_report['categories'],
            'critical_issues': [
                issue for issue in health_report['issues']
                if issue.get('severity') in ['CRITICAL', 'HIGH']
            ],
            'optimization_plan': self._create_optimization_plan(health_report),
            'execution_checklist': self._create_execution_checklist()
        }
        
        return report
    
    def _get_assessment(self, score: float) -> str:
        """根據(jù)評分獲取評估結(jié)果"""
        if score >= 90:
            return 'EXCELLENT'
        elif score >= 80:
            return 'GOOD'
        elif score >= 70:
            return 'FAIR'
        elif score >= 60:
            return 'NEEDS_IMPROVEMENT'
        else:
            return 'POOR'
    
    def _create_optimization_plan(self, health_report: Dict[str, Any]) -> List[Dict[str, Any]]:
        """創(chuàng)建優(yōu)化計劃"""
        plan = []
        
        # 按優(yōu)先級排序建議
        sorted_recommendations = sorted(
            health_report['recommendations'],
            key=lambda x: x.get('priority', 3)
        )
        
        for i, rec in enumerate(sorted_recommendations[:10], 1):  # 取前10個
            plan.append({
                'step': i,
                'action': rec['action'],
                'priority': rec.get('priority', 3),
                'estimated_effort': self._estimate_effort(rec['action']),
                'expected_impact': self._estimate_impact(rec['action'])
            })
        
        return plan
    
    def _estimate_effort(self, action: str) -> str:
        """估計實施難度"""
        low_effort_keywords = ['調(diào)整', '設(shè)置', '啟用', '禁用']
        medium_effort_keywords = ['優(yōu)化', '重構(gòu)', '重建', '遷移']
        high_effort_keywords = ['重寫', '重構(gòu)架構(gòu)', '數(shù)據(jù)遷移', '集群擴(kuò)展']
        
        action_lower = action.lower()
        
        if any(keyword in action_lower for keyword in high_effort_keywords):
            return 'HIGH'
        elif any(keyword in action_lower for keyword in medium_effort_keywords):
            return 'MEDIUM'
        else:
            return 'LOW'
    
    def _estimate_impact(self, action: str) -> str:
        """估計影響程度"""
        high_impact_keywords = ['性能提升', '顯著改善', '根本解決', '關(guān)鍵修復(fù)']
        medium_impact_keywords = ['優(yōu)化', '改進(jìn)', '增強(qiáng)', '調(diào)整']
        low_impact_keywords = ['微調(diào)', '小優(yōu)化', '維護(hù)', '清理']
        
        action_lower = action.lower()
        
        if any(keyword in action_lower for keyword in high_impact_keywords):
            return 'HIGH'
        elif any(keyword in action_lower for keyword in medium_impact_keywords):
            return 'MEDIUM'
        else:
            return 'LOW'
    
    def _create_execution_checklist(self) -> List[Dict[str, Any]]:
        """創(chuàng)建執(zhí)行檢查清單"""
        checklist = [
            {
                'phase': '準(zhǔn)備階段',
                'tasks': [
                    {'task': '備份數(shù)據(jù)庫', 'completed': False},
                    {'task': '驗證備份完整性', 'completed': False},
                    {'task': '準(zhǔn)備回滾計劃', 'completed': False},
                    {'task': '安排維護(hù)窗口', 'completed': False}
                ]
            },
            {
                'phase': '實施階段',
                'tasks': [
                    {'task': '應(yīng)用配置變更', 'completed': False},
                    {'task': '執(zhí)行索引優(yōu)化', 'completed': False},
                    {'task': '運行VACUUM操作', 'completed': False},
                    {'task': '更新統(tǒng)計信息', 'completed': False}
                ]
            },
            {
                'phase': '驗證階段',
                'tasks': [
                    {'task': '驗證性能改進(jìn)', 'completed': False},
                    {'task': '運行回歸測試', 'completed': False},
                    {'task': '監(jiān)控系統(tǒng)穩(wěn)定性', 'completed': False},
                    {'task': '更新文檔記錄', 'completed': False}
                ]
            }
        ]
        
        return checklist
    
    def close(self):
        """關(guān)閉連接"""
        if self.cursor:
            self.cursor.close()
        if self.conn:
            self.conn.close()
        logger.info("數(shù)據(jù)庫連接已關(guān)閉")


def demonstrate_optimization():
    """演示優(yōu)化功能"""
    dsn = "dbname=testdb user=postgres password=password host=localhost port=5432"
    
    print("PostgreSQL高級優(yōu)化演示")
    print("=" * 60)
    
    optimizer = PostgreSQLAdvancedOptimizer(dsn)
    
    try:
        # 生成健康報告
        print("\n生成數(shù)據(jù)庫健康報告...")
        health_report = optimizer.analyze_database_health()
        
        print(f"\n總體評分: {health_report['overall_score']:.1f}/100")
        print(f"發(fā)現(xiàn)問題: {len(health_report['issues'])} 個")
        print(f"優(yōu)化建議: {len(health_report['recommendations'])} 條")
        
        # 顯示關(guān)鍵問題
        critical_issues = [
            issue for issue in health_report['issues']
            if issue.get('severity') in ['CRITICAL', 'HIGH']
        ]
        
        if critical_issues:
            print("\n關(guān)鍵問題:")
            for issue in critical_issues[:3]:
                print(f"  ? [{issue['severity']}] {issue['message']}")
        
        # 顯示高優(yōu)先級建議
        high_priority_recs = [
            rec for rec in health_report['recommendations']
            if rec.get('priority', 3) == 1
        ]
        
        if high_priority_recs:
            print("\n高優(yōu)先級建議:")
            for rec in high_priority_recs[:3]:
                print(f"  ? {rec['action']}")
        
        # 生成詳細(xì)報告
        print("\n\n生成詳細(xì)優(yōu)化報告...")
        report = optimizer.generate_optimization_report()
        
        print(f"\n優(yōu)化計劃 ({len(report['optimization_plan'])} 個步驟):")
        for step in report['optimization_plan'][:5]:
            print(f"  步驟{step['step']}: {step['action']}")
            print(f"     優(yōu)先級: {step['priority']}, 難度: {step['estimated_effort']}, "
                  f"影響: {step['expected_impact']}")
        
        print("\n執(zhí)行檢查清單:")
        for phase in report['execution_checklist']:
            print(f"  {phase['phase']}:")
            for task in phase['tasks']:
                status = '?' if task['completed'] else '○'
                print(f"    {status} {task['task']}")
        
        print("\n優(yōu)化演示完成!")
        
    except Exception as e:
        print(f"演示過程中出錯: {e}")
    finally:
        optimizer.close()


if __name__ == "__main__":
    demonstrate_optimization()

"""
PostgreSQL代碼質(zhì)量自查工具
"""
import re
import ast
from typing import List, Dict, Any, Set
import logging

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


class PostgreSQLCodeChecker:
    """PostgreSQL代碼質(zhì)量檢查器"""
    
    def __init__(self):
        self.rules = {
            'sql_injection': self.check_sql_injection,
            'connection_management': self.check_connection_management,
            'transaction_handling': self.check_transaction_handling,
            'index_usage': self.check_index_usage,
            'data_type_validation': self.check_data_type_validation,
            'error_handling': self.check_error_handling,
            'performance_anti_patterns': self.check_performance_anti_patterns
        }
        
    def check_file(self, filepath: str) -> Dict[str, List[Dict[str, Any]]]:
        """
        檢查Python文件中的PostgreSQL相關(guān)代碼
        
        Args:
            filepath: Python文件路徑
            
        Returns:
            檢查結(jié)果
        """
        with open(filepath, 'r', encoding='utf-8') as f:
            content = f.read()
        
        issues = {}
        
        for rule_name, rule_func in self.rules.items():
            rule_issues = rule_func(content, filepath)
            if rule_issues:
                issues[rule_name] = rule_issues
        
        return issues
    
    def check_sql_injection(self, content: str, filepath: str) -> List[Dict[str, Any]]:
        """
        檢查SQL注入漏洞
        
        Args:
            content: 文件內(nèi)容
            filepath: 文件路徑
            
        Returns:
            問題列表
        """
        issues = []
        
        # 查找可能的字符串拼接SQL
        patterns = [
            (r'execute\s*\(.*?\s*%\s*.*?\)', '使用字符串格式化執(zhí)行SQL'),
            (r'execute\s*\(.*?\s*\+\s*.*?\)', '使用字符串拼接執(zhí)行SQL'),
            (r'executemany\s*\(.*?\s*%\s*.*?\)', '使用字符串格式化執(zhí)行批量SQL'),
            (r'f"SELECT.*{.*}.*"', '使用f-string直接嵌入變量'),
        ]
        
        lines = content.split('\n')
        for i, line in enumerate(lines, 1):
            for pattern, description in patterns:
                if re.search(pattern, line, re.IGNORECASE):
                    issues.append({
                        'line': i,
                        'severity': 'CRITICAL',
                        'message': f'潛在的SQL注入漏洞: {description}',
                        'suggestion': '使用參數(shù)化查詢（%s占位符）',
                        'code_snippet': line.strip()[:100]
                    })
        
        return issues
    
    def check_connection_management(self, content: str, filepath: str) -> List[Dict[str, Any]]:
        """
        檢查數(shù)據(jù)庫連接管理
        
        Args:
            content: 文件內(nèi)容
            filepath: 文件路徑
            
        Returns:
            問題列表
        """
        issues = []
        lines = content.split('\n')
        
        # 查找連接創(chuàng)建但不關(guān)閉的情況
        connect_pattern = r'psycopg2\.connect\(|connect\('
        close_pattern = r'\.close\(\)'
        
        in_function = False
        function_start = 0
        connect_lines = []
        
        for i, line in enumerate(lines, 1):
            # 檢測函數(shù)開始
            if line.strip().startswith('def '):
                in_function = True
                function_start = i
                connect_lines = []
            
            # 檢測連接創(chuàng)建
            if re.search(connect_pattern, line):
                connect_lines.append(i)
            
            # 檢測連接關(guān)閉
            if re.search(close_pattern, line) and 'close' in line:
                if connect_lines:
                    connect_lines.pop()
            
            # 檢測函數(shù)結(jié)束
            if line.strip() == '' or i == len(lines):
                if in_function and connect_lines:
                    for connect_line in connect_lines:
                        issues.append({
                            'line': connect_line,
                            'severity': 'HIGH',
                            'message': '數(shù)據(jù)庫連接可能未正確關(guān)閉',
                            'suggestion': '確保在finally塊中關(guān)閉連接',
                            'context': f'函數(shù)開始于第{function_start}行'
                        })
                in_function = False
        
        return issues
    
    def check_transaction_handling(self, content: str, filepath: str) -> List[Dict[str, Any]]:
        """
        檢查事務(wù)處理
        
        Args:
            content: 文件內(nèi)容
            filepath: 文件路徑
            
        Returns:
            問題列表
        """
        issues = []
        lines = content.split('\n')
        
        # 查找沒有明確事務(wù)管理的操作
        write_operations = [
            'INSERT', 'UPDATE', 'DELETE', 'CREATE', 'ALTER', 'DROP'
        ]
        
        transaction_keywords = [
            'BEGIN', 'COMMIT', 'ROLLBACK', 'autocommit',
            'set_session', 'transaction'
        ]
        
        in_write_operation = False
        has_transaction_control = False
        
        for i, line in enumerate(lines, 1):
            line_upper = line.upper()
            
            # 檢查是否有寫操作
            if any(op in line_upper for op in write_operations):
                in_write_operation = True
            
            # 檢查是否有事務(wù)控制
            if any(keyword in line for keyword in transaction_keywords):
                has_transaction_control = True
            
            # 如果是空行或注釋，檢查之前的操作
            if line.strip() == '' or line.strip().startswith('#'):
                if in_write_operation and not has_transaction_control:
                    issues.append({
                        'line': i - 1,
                        'severity': 'MEDIUM',
                        'message': '寫操作沒有顯式的事務(wù)管理',
                        'suggestion': '使用明確的事務(wù)控制（BEGIN/COMMIT/ROLLBACK）',
                        'context': '多個寫操作應(yīng)該在同一事務(wù)中'
                    })
                
                in_write_operation = False
                has_transaction_control = False
        
        return issues
    
    def check_index_usage(self, content: str, filepath: str) -> List[Dict[str, Any]]:
        """
        檢查索引使用
        
        Args:
            content: 文件內(nèi)容
            filepath: 文件路徑
            
        Returns:
            問題列表
        """
        issues = []
        lines = content.split('\n')
        
        # 查找可能受益于索引的查詢模式
        index_patterns = [
            (r'WHERE\s+\w+\s*=\s*', '等值查詢'),
            (r'WHERE\s+\w+\s+IN\s*\(', 'IN列表查詢'),
            (r'WHERE\s+\w+\s+LIKE\s+\'', 'LIKE查詢（可能前綴匹配）'),
            (r'ORDER\s+BY\s+\w+', '排序操作'),
            (r'GROUP\s+BY\s+\w+', '分組操作'),
            (r'JOIN\s+\w+\s+ON\s+\w+\s*=\s*\w+', '連接操作')
        ]
        
        for i, line in enumerate(lines, 1):
            line_upper = line.upper()
            
            # 跳過注釋
            if line.strip().startswith('#'):
                continue
            
            for pattern, description in index_patterns:
                if re.search(pattern, line_upper, re.IGNORECASE):
                    issues.append({
                        'line': i,
                        'severity': 'LOW',
                        'message': f'查詢可能受益于索引: {description}',
                        'suggestion': '考慮在相關(guān)列上創(chuàng)建索引',
                        'code_snippet': line.strip()[:100]
                    })
                    break
        
        return issues
    
    def generate_report(self, issues: Dict[str, List[Dict[str, Any]]]) -> str:
        """生成檢查報告"""
        report_lines = []
        report_lines.append("=" * 60)
        report_lines.append("PostgreSQL代碼質(zhì)量檢查報告")
        report_lines.append("=" * 60)
        
        total_issues = sum(len(rule_issues) for rule_issues in issues.values())
        report_lines.append(f"\n總共發(fā)現(xiàn) {total_issues} 個問題\n")
        
        severity_counts = {'CRITICAL': 0, 'HIGH': 0, 'MEDIUM': 0, 'LOW': 0}
        
        for rule_name, rule_issues in issues.items():
            if rule_issues:
                report_lines.append(f"\n{rule_name.upper()} ({len(rule_issues)}個問題):")
                report_lines.append("-" * 40)
                
                for issue in rule_issues:
                    severity = issue.get('severity', 'UNKNOWN')
                    severity_counts[severity] = severity_counts.get(severity, 0) + 1
                    
                    report_lines.append(
                        f"[{severity}] 第{issue['line']}行: {issue['message']}"
                    )
                    if 'suggestion' in issue:
                        report_lines.append(f"    建議: {issue['suggestion']}")
                    if 'code_snippet' in issue:
                        report_lines.append(f"    代碼: {issue['code_snippet']}")
                    report_lines.append("")
        
        # 添加嚴(yán)重性統(tǒng)計
        report_lines.append("\n嚴(yán)重性統(tǒng)計:")
        report_lines.append("-" * 40)
        for severity, count in severity_counts.items():
            if count > 0:
                report_lines.append(f"  {severity}: {count} 個問題")
        
        return '\n'.join(report_lines)


def check_postgresql_best_practices():
    """PostgreSQL最佳實踐檢查示例"""
    checker = PostgreSQLCodeChecker()
    
    # 示例代碼
    sample_code = """
import psycopg2

# 不好的示例：字符串拼接SQL（SQL注入風(fēng)險）
def bad_example(user_input):
    conn = psycopg2.connect("dbname=test")
    cursor = conn.cursor()
    
    # SQL注入漏洞
    query = f"SELECT * FROM users WHERE username = '{user_input}'"
    cursor.execute(query)
    
    # 連接未關(guān)閉
    return cursor.fetchall()

# 好的示例：參數(shù)化查詢
def good_example(username):
    conn = None
    cursor = None
    try:
        conn = psycopg2.connect("dbname=test")
        cursor = conn.cursor()
        
        # 參數(shù)化查詢
        query = "SELECT * FROM users WHERE username = %s"
        cursor.execute(query, (username,))
        
        return cursor.fetchall()
    finally:
        if cursor:
            cursor.close()
        if conn:
            conn.close()

# 事務(wù)處理示例
def transaction_example(user_data):
    conn = psycopg2.connect("dbname=test")
    cursor = conn.cursor()
    
    try:
        # 開始事務(wù)
        conn.autocommit = False
        
        # 多個寫操作
        cursor.execute(
            "INSERT INTO users (name, email) VALUES (%s, %s)",
            (user_data['name'], user_data['email'])
        )
        
        cursor.execute(
            "INSERT INTO logs (action) VALUES (%s)",
            ('user_created',)
        )
        
        # 提交事務(wù)
        conn.commit()
        
    except Exception as e:
        # 回滾事務(wù)
        conn.rollback()
        raise e
    finally:
        cursor.close()
        conn.close()

# 可能受益于索引的查詢
def query_without_index():
    conn = psycopg2.connect("dbname=test")
    cursor = conn.cursor()
    
    # 這個查詢可能受益于索引
    cursor.execute("""
        SELECT * FROM orders 
        WHERE customer_id = %s 
        AND order_date > %s
        ORDER BY order_date DESC
    """, (123, '2024-01-01'))
    
    return cursor.fetchall()
    """
    
    # 將示例代碼寫入臨時文件
    import tempfile
    with tempfile.NamedTemporaryFile(mode='w', suffix='.py', delete=False) as f:
        f.write(sample_code)
        temp_file = f.name
    
    try:
        # 檢查代碼
        issues = checker.check_file(temp_file)
        report = checker.generate_report(issues)
        print(report)
    finally:
        import os
        os.unlink(temp_file)


if __name__ == "__main__":
    check_postgresql_best_practices()