#!/bin/bash

# 种子数据验证脚本
# 用于验证种子数据的完整性和正确性

set -e

# 配置
DB_PATH="./data/photography.db"

# 颜色定义
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
NC='\033[0m' # No Color

# 日志函数
log_info() {
    echo -e "${BLUE}[INFO]${NC} $1"
}

log_success() {
    echo -e "${GREEN}[SUCCESS]${NC} $1"
}

log_warning() {
    echo -e "${YELLOW}[WARNING]${NC} $1"
}

log_error() {
    echo -e "${RED}[ERROR]${NC} $1"
}

# 检查数据库是否存在
check_database() {
    if [ ! -f "$DB_PATH" ]; then
        log_error "数据库文件不存在: $DB_PATH"
        exit 1
    fi
    log_info "数据库文件存在"
}

# 测试基础数据统计
test_basic_stats() {
    log_info "测试基础数据统计..."
    
    local user_count=$(sqlite3 "$DB_PATH" "SELECT COUNT(*) FROM user;")
    local category_count=$(sqlite3 "$DB_PATH" "SELECT COUNT(*) FROM category;")
    local photo_count=$(sqlite3 "$DB_PATH" "SELECT COUNT(*) FROM photo;")
    
    echo "用户数量: $user_count (期望: 6)"
    echo "分类数量: $category_count (期望: 9)"
    echo "照片数量: $photo_count (期望: 30)"
    
    # 验证数据量
    if [ "$user_count" -eq 6 ] && [ "$category_count" -eq 9 ] && [ "$photo_count" -ge 30 ]; then
        log_success "基础数据统计正确"
    else
        log_error "基础数据统计不正确"
        return 1
    fi
}

# 测试分类ID连续性
test_category_ids() {
    log_info "测试分类ID连续性..."
    
    local missing_ids=$(sqlite3 "$DB_PATH" "
    WITH RECURSIVE series(x) AS (
        SELECT 1
        UNION ALL
        SELECT x+1 FROM series WHERE x < 9
    )
    SELECT COUNT(*) FROM series 
    WHERE x NOT IN (SELECT id FROM category);
    ")
    
    if [ "$missing_ids" -eq 0 ]; then
        log_success "分类ID连续性正确"
    else
        log_error "分类ID不连续，缺失 $missing_ids 个ID"
        return 1
    fi
}

# 测试用户数据完整性
test_user_integrity() {
    log_info "测试用户数据完整性..."
    
    # 检查管理员用户
    local admin_count=$(sqlite3 "$DB_PATH" "SELECT COUNT(*) FROM user WHERE username='admin';")
    if [ "$admin_count" -eq 1 ]; then
        log_success "管理员用户存在"
    else
        log_error "管理员用户不存在或重复"
        return 1
    fi
    
    # 检查用户名唯一性
    local duplicate_users=$(sqlite3 "$DB_PATH" "
    SELECT COUNT(*) FROM (
        SELECT username FROM user GROUP BY username HAVING COUNT(*) > 1
    );
    ")
    
    if [ "$duplicate_users" -eq 0 ]; then
        log_success "用户名唯一性正确"
    else
        log_error "存在重复用户名"
        return 1
    fi
}

# 测试照片数据完整性
test_photo_integrity() {
    log_info "测试照片数据完整性..."
    
    # 检查所有照片都有有效的用户ID
    local invalid_user_photos=$(sqlite3 "$DB_PATH" "
    SELECT COUNT(*) FROM photo 
    WHERE user_id NOT IN (SELECT id FROM user);
    ")
    
    if [ "$invalid_user_photos" -eq 0 ]; then
        log_success "照片用户关联正确"
    else
        log_error "存在 $invalid_user_photos 张照片的用户ID无效"
        return 1
    fi
    
    # 检查所有照片都有有效的分类ID
    local invalid_category_photos=$(sqlite3 "$DB_PATH" "
    SELECT COUNT(*) FROM photo 
    WHERE category_id NOT IN (SELECT id FROM category);
    ")
    
    if [ "$invalid_category_photos" -eq 0 ]; then
        log_success "照片分类关联正确"
    else
        log_error "存在 $invalid_category_photos 张照片的分类ID无效"
        return 1
    fi
}

# 测试分类照片分布
test_category_distribution() {
    log_info "测试分类照片分布..."
    
    # 检查是否有分类没有照片
    local empty_categories=$(sqlite3 "$DB_PATH" "
    SELECT COUNT(*) FROM category 
    WHERE id NOT IN (SELECT DISTINCT category_id FROM photo);
    ")
    
    if [ "$empty_categories" -eq 0 ]; then
        log_success "所有分类都有照片"
    else
        log_warning "有 $empty_categories 个分类没有照片"
    fi
    
    # 显示各分类照片数量
    echo "各分类照片数量分布:"
    sqlite3 "$DB_PATH" "
    SELECT 
        c.name || ': ' || COUNT(p.id) || '张' as distribution
    FROM category c
    LEFT JOIN photo p ON c.id = p.category_id
    GROUP BY c.id, c.name
    ORDER BY COUNT(p.id) DESC;
    "
}

# 测试用户照片分布
test_user_distribution() {
    log_info "测试用户照片分布..."
    
    # 检查是否有用户没有照片
    local users_without_photos=$(sqlite3 "$DB_PATH" "
    SELECT COUNT(*) FROM user 
    WHERE id NOT IN (SELECT DISTINCT user_id FROM photo);
    ")
    
    if [ "$users_without_photos" -eq 0 ]; then
        log_success "所有用户都有照片"
    else
        log_warning "有 $users_without_photos 个用户没有照片"
    fi
    
    # 显示各用户照片数量
    echo "各用户照片数量分布:"
    sqlite3 "$DB_PATH" "
    SELECT 
        u.username || ': ' || COUNT(p.id) || '张' as distribution
    FROM user u
    LEFT JOIN photo p ON u.id = p.user_id
    GROUP BY u.id, u.username
    ORDER BY COUNT(p.id) DESC;
    "
}

# 测试数据时间合理性
test_data_timestamps() {
    log_info "测试数据时间合理性..."
    
    # 检查是否有未来时间的数据
    local future_photos=$(sqlite3 "$DB_PATH" "
    SELECT COUNT(*) FROM photo 
    WHERE created_at > datetime('now');
    ")
    
    if [ "$future_photos" -eq 0 ]; then
        log_success "照片时间戳合理"
    else
        log_warning "有 $future_photos 张照片的时间戳在未来"
    fi
}

# 性能测试
test_performance() {
    log_info "测试基础查询性能..."
    
    # 测试分类查询
    local start_time=$(date +%s)
    sqlite3 "$DB_PATH" "SELECT * FROM category;" > /dev/null
    local end_time=$(date +%s)
    local category_time=$((end_time - start_time))
    
    # 测试照片查询
    start_time=$(date +%s)
    sqlite3 "$DB_PATH" "SELECT * FROM photo LIMIT 10;" > /dev/null
    end_time=$(date +%s)
    local photo_time=$((end_time - start_time))
    
    echo "分类查询时间: ${category_time}s"
    echo "照片查询时间: ${photo_time}s"
    
    if [ "$category_time" -lt 2 ] && [ "$photo_time" -lt 2 ]; then
        log_success "查询性能良好"
    else
        log_warning "查询性能可能需要优化"
    fi
}

# 主函数
main() {
    log_info "开始种子数据验证测试"
    echo "=================================================="
    
    local test_count=0
    local passed_count=0
    
    # 运行所有测试
    tests=(
        "check_database"
        "test_basic_stats"
        "test_category_ids"
        "test_user_integrity"
        "test_photo_integrity"
        "test_category_distribution"
        "test_user_distribution"
        "test_data_timestamps"
        "test_performance"
    )
    
    for test in "${tests[@]}"; do
        test_count=$((test_count + 1))
        echo ""
        if $test; then
            passed_count=$((passed_count + 1))
        fi
    done
    
    echo ""
    echo "=================================================="
    echo "测试完成: $passed_count/$test_count 通过"
    
    if [ "$passed_count" -eq "$test_count" ]; then
        log_success "所有测试通过！种子数据质量良好"
        exit 0
    else
        log_warning "部分测试失败，请检查数据质量"
        exit 1
    fi
}

# 脚本入口
main "$@"