nxxmdata/backend/analyze-foreign-keys.js

/**
 * 分析数据库中表之间的外键关系
 * @file analyze-foreign-keys.js
 * @description 识别所有外键约束和引用关系，为ID重排做准备
 */

const { sequelize } = require('./config/database-simple');
const { QueryTypes } = require('sequelize');

async function analyzeForeignKeys() {
  try {
    console.log('=== 分析数据库外键关系 ===\n');
    
    // 获取所有外键约束
    const foreignKeys = await sequelize.query(`
      SELECT 
        kcu.TABLE_NAME as table_name,
        kcu.COLUMN_NAME as column_name,
        kcu.REFERENCED_TABLE_NAME as referenced_table,
        kcu.REFERENCED_COLUMN_NAME as referenced_column,
        rc.CONSTRAINT_NAME as constraint_name,
        rc.UPDATE_RULE as update_rule,
        rc.DELETE_RULE as delete_rule
      FROM information_schema.KEY_COLUMN_USAGE kcu
      JOIN information_schema.REFERENTIAL_CONSTRAINTS rc 
        ON kcu.CONSTRAINT_NAME = rc.CONSTRAINT_NAME
        AND kcu.TABLE_SCHEMA = rc.CONSTRAINT_SCHEMA
      WHERE kcu.TABLE_SCHEMA = DATABASE()
        AND kcu.REFERENCED_TABLE_NAME IS NOT NULL
      ORDER BY kcu.TABLE_NAME, kcu.COLUMN_NAME
    `, { type: QueryTypes.SELECT });
    
    console.log(`发现 ${foreignKeys.length} 个外键关系:\n`);
    
    const relationshipMap = new Map();
    const tablesWithForeignKeys = new Set();
    const referencedTables = new Set();
    
    foreignKeys.forEach(fk => {
      const key = `${fk.table_name}.${fk.column_name}`;
      const reference = `${fk.referenced_table}.${fk.referenced_column}`;
      
      relationshipMap.set(key, {
        table: fk.table_name,
        column: fk.column_name,
        referencedTable: fk.referenced_table,
        referencedColumn: fk.referenced_column,
        constraintName: fk.constraint_name,
        updateRule: fk.update_rule,
        deleteRule: fk.delete_rule
      });
      
      tablesWithForeignKeys.add(fk.table_name);
      referencedTables.add(fk.referenced_table);
      
      console.log(`🔗 ${fk.table_name}.${fk.column_name} -> ${fk.referenced_table}.${fk.referenced_column}`);
      console.log(`   约束名: ${fk.constraint_name}`);
      console.log(`   更新规则: ${fk.update_rule}`);
      console.log(`   删除规则: ${fk.delete_rule}`);
      console.log('');
    });
    
    // 分析每个外键字段的数据分布
    console.log('\n=== 外键字段数据分布 ===\n');
    
    const foreignKeyStats = [];
    
    for (const [key, relationship] of relationshipMap) {
      const { table, column, referencedTable, referencedColumn } = relationship;
      
      try {
        // 获取外键字段的统计信息
        const stats = await sequelize.query(`
          SELECT 
            COUNT(*) as total_count,
            COUNT(DISTINCT ${column}) as unique_count,
            MIN(${column}) as min_value,
            MAX(${column}) as max_value,
            COUNT(CASE WHEN ${column} IS NULL THEN 1 END) as null_count
          FROM ${table}
        `, { type: QueryTypes.SELECT });
        
        const stat = stats[0];
        
        // 检查引用完整性
        const integrityCheck = await sequelize.query(`
          SELECT COUNT(*) as invalid_references
          FROM ${table} t
          WHERE t.${column} IS NOT NULL 
            AND t.${column} NOT IN (
              SELECT ${referencedColumn} 
              FROM ${referencedTable} 
              WHERE ${referencedColumn} IS NOT NULL
            )
        `, { type: QueryTypes.SELECT });
        
        const invalidRefs = parseInt(integrityCheck[0].invalid_references);
        
        const fkStat = {
          table,
          column,
          referencedTable,
          referencedColumn,
          totalCount: parseInt(stat.total_count),
          uniqueCount: parseInt(stat.unique_count),
          minValue: stat.min_value,
          maxValue: stat.max_value,
          nullCount: parseInt(stat.null_count),
          invalidReferences: invalidRefs,
          hasIntegrityIssues: invalidRefs > 0
        };
        
        foreignKeyStats.push(fkStat);
        
        console.log(`📊 ${table}.${column} -> ${referencedTable}.${referencedColumn}:`);
        console.log(`   - 总记录数: ${fkStat.totalCount}`);
        console.log(`   - 唯一值数: ${fkStat.uniqueCount}`);
        console.log(`   - 值范围: ${fkStat.minValue} - ${fkStat.maxValue}`);
        console.log(`   - NULL值数: ${fkStat.nullCount}`);
        console.log(`   - 无效引用: ${fkStat.invalidReferences}`);
        console.log(`   - 完整性问题: ${fkStat.hasIntegrityIssues ? '是' : '否'}`);
        console.log('');
        
      } catch (error) {
        console.log(`❌ ${table}.${column}: 分析失败 - ${error.message}`);
      }
    }
    
    // 生成依赖关系图
    console.log('\n=== 表依赖关系 ===\n');
    
    const dependencyGraph = new Map();
    
    foreignKeys.forEach(fk => {
      if (!dependencyGraph.has(fk.table_name)) {
        dependencyGraph.set(fk.table_name, new Set());
      }
      dependencyGraph.get(fk.table_name).add(fk.referenced_table);
    });
    
    // 计算更新顺序（拓扑排序）
    const updateOrder = [];
    const visited = new Set();
    const visiting = new Set();
    
    function topologicalSort(table) {
      if (visiting.has(table)) {
        console.log(`⚠️  检测到循环依赖: ${table}`);
        return;
      }
      if (visited.has(table)) {
        return;
      }
      
      visiting.add(table);
      
      const dependencies = dependencyGraph.get(table) || new Set();
      for (const dep of dependencies) {
        topologicalSort(dep);
      }
      
      visiting.delete(table);
      visited.add(table);
      updateOrder.push(table);
    }
    
    // 对所有表进行拓扑排序
    const allTables = new Set([...tablesWithForeignKeys, ...referencedTables]);
    for (const table of allTables) {
      topologicalSort(table);
    }
    
    console.log('建议的ID重排顺序（被引用的表优先）:');
    updateOrder.reverse().forEach((table, index) => {
      const deps = dependencyGraph.get(table);
      const depList = deps ? Array.from(deps).join(', ') : '无';
      console.log(`${index + 1}. ${table} (依赖: ${depList})`);
    });
    
    // 汇总报告
    console.log('\n=== 汇总报告 ===');
    console.log(`外键关系总数: ${foreignKeys.length}`);
    console.log(`涉及外键的表: ${tablesWithForeignKeys.size}`);
    console.log(`被引用的表: ${referencedTables.size}`);
    
    const tablesWithIssues = foreignKeyStats.filter(stat => stat.hasIntegrityIssues);
    if (tablesWithIssues.length > 0) {
      console.log(`\n⚠️  发现完整性问题的表:`);
      tablesWithIssues.forEach(stat => {
        console.log(`- ${stat.table}.${stat.column}: ${stat.invalidReferences} 个无效引用`);
      });
    } else {
      console.log('\n✅ 所有外键关系完整性正常');
    }
    
    return {
      foreignKeys,
      foreignKeyStats,
      updateOrder: updateOrder.reverse(),
      relationshipMap,
      tablesWithIssues
    };
    
  } catch (error) {
    console.error('分析外键关系失败:', error);
    throw error;
  } finally {
    await sequelize.close();
  }
}

// 如果直接运行此脚本
if (require.main === module) {
  analyzeForeignKeys()
    .then(() => {
      console.log('\n分析完成!');
      process.exit(0);
    })
    .catch(error => {
      console.error('分析失败:', error);
      process.exit(1);
    });
}

module.exports = { analyzeForeignKeys };