diff --git a/scripts/init-db.sql b/scripts/init-db.sql index 32347a8..c5c4ce1 100644 --- a/scripts/init-db.sql +++ b/scripts/init-db.sql @@ -1,272 +1,1450 @@ -- =========================================== -- iConsulting 数据库初始化脚本 +-- 香港移民在线咨询系统 -- =========================================== --- 创建扩展 -CREATE EXTENSION IF NOT EXISTS "uuid-ossp"; -CREATE EXTENSION IF NOT EXISTS "pgcrypto"; - --- 创建数据库 (如果不存在) --- 注意: 此脚本在 docker-entrypoint-initdb.d 中运行时，数据库已由 POSTGRES_DB 环境变量创建 +-- 启用必要的扩展 +CREATE EXTENSION IF NOT EXISTS vector; -- pgvector: 向量存储和相似度搜索 +CREATE EXTENSION IF NOT EXISTS "uuid-ossp"; -- UUID生成函数 -- =========================================== --- 创建表 +-- 用户表 (users) +-- 存储系统用户信息，支持匿名用户和注册用户 -- =========================================== - --- 用户表 -CREATE TABLE IF NOT EXISTS users ( +CREATE TABLE users ( id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - type VARCHAR(20) NOT NULL DEFAULT 'ANONYMOUS', - phone VARCHAR(20) UNIQUE, - email VARCHAR(255) UNIQUE, - password_hash VARCHAR(255), - name VARCHAR(100), - avatar VARCHAR(500), - status VARCHAR(20) NOT NULL DEFAULT 'ACTIVE', - metadata JSONB DEFAULT '{}', - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - updated_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - last_active_at TIMESTAMP WITH TIME ZONE -); - -CREATE INDEX idx_users_phone ON users(phone); -CREATE INDEX idx_users_email ON users(email); -CREATE INDEX idx_users_type ON users(type); -CREATE INDEX idx_users_status ON users(status); - --- 管理员表 -CREATE TABLE IF NOT EXISTS admins ( - id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - username VARCHAR(50) NOT NULL UNIQUE, - password_hash VARCHAR(255) NOT NULL, - name VARCHAR(100) NOT NULL, - email VARCHAR(255), + -- 用户类型: ANONYMOUS(匿名访客), REGISTERED(已注册用户) + type VARCHAR(20) NOT NULL DEFAULT 'ANONYMOUS' + CHECK (type IN ('ANONYMOUS', 'REGISTERED')), + -- 浏览器指纹，用于识别匿名用户 + fingerprint VARCHAR(255), + -- 手机号码（注册用户必填） phone VARCHAR(20), - role VARCHAR(20) NOT NULL DEFAULT 'OPERATOR', - permissions JSONB DEFAULT '[]', + -- 用户昵称 + nickname VARCHAR(100), + -- 头像URL avatar VARCHAR(500), - last_login_at TIMESTAMP WITH TIME ZONE, - last_login_ip VARCHAR(50), - is_active BOOLEAN DEFAULT true, - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - updated_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP + -- 用户来源渠道: WEB, H5, WECHAT_MP, ALIPAY_MP, LINE, WHATSAPP + source_channel VARCHAR(50) DEFAULT 'WEB', + -- 用户来源标识（如推广码、活动ID等） + source_ref VARCHAR(100), + -- 用户标签（JSON数组，如["高净值", "技术人才"]） + tags JSONB DEFAULT '[]', + -- 用户偏好设置 + preferences JSONB DEFAULT '{}', + -- 备注信息（管理员可编辑） + notes TEXT, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 最后活跃时间 + last_active_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() ); -CREATE INDEX idx_admins_username ON admins(username); -CREATE INDEX idx_admins_role ON admins(role); +COMMENT ON TABLE users IS '用户表 - 存储系统所有用户信息'; +COMMENT ON COLUMN users.type IS '用户类型: ANONYMOUS=匿名访客, REGISTERED=已注册用户'; +COMMENT ON COLUMN users.fingerprint IS '浏览器指纹，用于识别和关联匿名用户'; +COMMENT ON COLUMN users.source_channel IS '用户来源渠道，用于统计分析'; +COMMENT ON COLUMN users.tags IS '用户标签，JSON数组格式，用于用户分群'; --- 对话表 -CREATE TABLE IF NOT EXISTS conversations ( +CREATE INDEX idx_users_fingerprint ON users(fingerprint); +CREATE INDEX idx_users_phone ON users(phone); +CREATE INDEX idx_users_type ON users(type); +CREATE INDEX idx_users_source_channel ON users(source_channel); +CREATE INDEX idx_users_created_at ON users(created_at); + +-- =========================================== +-- 对话表 (conversations) +-- 存储用户与AI助手的对话会话 +-- =========================================== +CREATE TABLE conversations ( id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - user_id UUID NOT NULL REFERENCES users(id), - title VARCHAR(500), - status VARCHAR(20) NOT NULL DEFAULT 'ACTIVE', - message_count INTEGER DEFAULT 0, - category VARCHAR(50), - is_paid BOOLEAN DEFAULT false, - satisfaction_score INTEGER, + -- 所属用户ID + user_id UUID REFERENCES users(id) ON DELETE CASCADE, + -- 对话状态: ACTIVE(进行中), ENDED(已结束), ARCHIVED(已归档) + status VARCHAR(20) NOT NULL DEFAULT 'ACTIVE' + CHECK (status IN ('ACTIVE', 'ENDED', 'ARCHIVED')), + -- 对话标题（自动生成或用户设置） + title VARCHAR(255), + -- 对话摘要（AI生成） summary TEXT, - metadata JSONB DEFAULT '{}', - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - updated_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - closed_at TIMESTAMP WITH TIME ZONE + -- 主要咨询的移民类别: QMAS, GEP, IANG, TTPS, CIES, TECHTAS + category VARCHAR(50), + -- 消息数量（通过触发器自动更新） + message_count INT DEFAULT 0, + -- 用户消息数量 + user_message_count INT DEFAULT 0, + -- AI消息数量 + assistant_message_count INT DEFAULT 0, + -- Token消耗统计 + total_input_tokens INT DEFAULT 0, + total_output_tokens INT DEFAULT 0, + -- 对话评分（用户反馈，1-5分） + rating SMALLINT CHECK (rating >= 1 AND rating <= 5), + -- 用户反馈内容 + feedback TEXT, + -- 是否转化为付费（用于统计转化率） + has_converted BOOLEAN DEFAULT FALSE, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 结束时间 + ended_at TIMESTAMP WITH TIME ZONE ); +COMMENT ON TABLE conversations IS '对话表 - 存储用户与AI助手的对话会话'; +COMMENT ON COLUMN conversations.category IS '主要咨询的移民类别，用于分析用户兴趣分布'; +COMMENT ON COLUMN conversations.has_converted IS '是否产生付费转化，用于计算转化率'; +COMMENT ON COLUMN conversations.rating IS '用户对对话的评分，1-5分'; + CREATE INDEX idx_conversations_user_id ON conversations(user_id); CREATE INDEX idx_conversations_status ON conversations(status); CREATE INDEX idx_conversations_category ON conversations(category); -CREATE INDEX idx_conversations_created_at ON conversations(created_at); +CREATE INDEX idx_conversations_created_at ON conversations(created_at DESC); +CREATE INDEX idx_conversations_has_converted ON conversations(has_converted) WHERE has_converted = TRUE; --- 消息表 -CREATE TABLE IF NOT EXISTS messages ( +-- =========================================== +-- 消息表 (messages) +-- 存储对话中的每条消息 +-- =========================================== +CREATE TABLE messages ( id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - conversation_id UUID NOT NULL REFERENCES conversations(id) ON DELETE CASCADE, - role VARCHAR(20) NOT NULL, + -- 所属对话ID + conversation_id UUID REFERENCES conversations(id) ON DELETE CASCADE, + -- 消息角色: user(用户), assistant(AI助手), system(系统) + role VARCHAR(20) NOT NULL CHECK (role IN ('user', 'assistant', 'system')), + -- 消息类型 + type VARCHAR(30) NOT NULL DEFAULT 'TEXT' + CHECK (type IN ('TEXT', 'TOOL_CALL', 'TOOL_RESULT', 'PAYMENT_REQUEST', 'ASSESSMENT_START', 'ASSESSMENT_RESULT')), + -- 消息内容 content TEXT NOT NULL, - token_count INTEGER, - tool_calls JSONB, - citations JSONB, - metadata JSONB DEFAULT '{}', - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP + -- 元数据（工具调用信息、Token使用等） + metadata JSONB, + -- Token使用量 + input_tokens INT, + output_tokens INT, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() ); +COMMENT ON TABLE messages IS '消息表 - 存储对话中的每条消息'; +COMMENT ON COLUMN messages.role IS '消息角色: user=用户发送, assistant=AI回复, system=系统消息'; +COMMENT ON COLUMN messages.type IS '消息类型，用于区分普通文本、工具调用、支付请求等'; +COMMENT ON COLUMN messages.metadata IS '元数据，存储工具调用参数、结果等扩展信息'; + CREATE INDEX idx_messages_conversation_id ON messages(conversation_id); CREATE INDEX idx_messages_role ON messages(role); CREATE INDEX idx_messages_created_at ON messages(created_at); --- 知识文章表 -CREATE TABLE IF NOT EXISTS knowledge_articles ( +-- =========================================== +-- 订单表 (orders) +-- 存储用户购买的服务订单 +-- =========================================== +CREATE TABLE orders ( id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - title VARCHAR(500) NOT NULL, - content TEXT NOT NULL, - summary TEXT, - category VARCHAR(50) NOT NULL, - tags JSONB DEFAULT '[]', - source VARCHAR(20) NOT NULL DEFAULT 'MANUAL', - source_url VARCHAR(1000), - is_published BOOLEAN DEFAULT false, - citation_count INTEGER DEFAULT 0, - quality_score INTEGER DEFAULT 50, - metadata JSONB DEFAULT '{}', - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - updated_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - published_at TIMESTAMP WITH TIME ZONE -); - -CREATE INDEX idx_knowledge_articles_category ON knowledge_articles(category); -CREATE INDEX idx_knowledge_articles_is_published ON knowledge_articles(is_published); -CREATE INDEX idx_knowledge_articles_source ON knowledge_articles(source); - --- 知识块表 (用于向量检索) -CREATE TABLE IF NOT EXISTS knowledge_chunks ( - id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - article_id UUID NOT NULL REFERENCES knowledge_articles(id) ON DELETE CASCADE, - content TEXT NOT NULL, - chunk_index INTEGER NOT NULL, - token_count INTEGER, - embedding_id VARCHAR(100), - metadata JSONB DEFAULT '{}', - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP -); - -CREATE INDEX idx_knowledge_chunks_article_id ON knowledge_chunks(article_id); -CREATE INDEX idx_knowledge_chunks_embedding_id ON knowledge_chunks(embedding_id); - --- 经验表 -CREATE TABLE IF NOT EXISTS experiences ( - id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - experience_type VARCHAR(50) NOT NULL, - content TEXT NOT NULL, - scenario TEXT, - confidence INTEGER NOT NULL DEFAULT 50, - related_category VARCHAR(50), - source_conversation_ids JSONB DEFAULT '[]', - verification_status VARCHAR(20) NOT NULL DEFAULT 'PENDING', - verified_by UUID REFERENCES admins(id), - verified_at TIMESTAMP WITH TIME ZONE, - usage_count INTEGER DEFAULT 0, - positive_count INTEGER DEFAULT 0, - negative_count INTEGER DEFAULT 0, - is_active BOOLEAN DEFAULT true, - metadata JSONB DEFAULT '{}', - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - updated_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP -); - -CREATE INDEX idx_experiences_type ON experiences(experience_type); -CREATE INDEX idx_experiences_status ON experiences(verification_status); -CREATE INDEX idx_experiences_category ON experiences(related_category); -CREATE INDEX idx_experiences_is_active ON experiences(is_active); - --- 支付订单表 -CREATE TABLE IF NOT EXISTS payment_orders ( - id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - user_id UUID NOT NULL REFERENCES users(id), - conversation_id UUID REFERENCES conversations(id), - order_no VARCHAR(50) NOT NULL UNIQUE, + -- 订单号（用于展示，格式：ORD + 年月日 + 序号） + order_no VARCHAR(50) UNIQUE, + -- 所属用户ID + user_id UUID REFERENCES users(id) ON DELETE SET NULL, + -- 关联的对话ID（订单从哪个对话产生） + conversation_id UUID REFERENCES conversations(id) ON DELETE SET NULL, + -- 服务类型: ASSESSMENT(评估), CONSULTATION(咨询), DOCUMENT_REVIEW(文档审核) + service_type VARCHAR(50) NOT NULL + CHECK (service_type IN ('ASSESSMENT', 'CONSULTATION', 'DOCUMENT_REVIEW')), + -- 服务对应的移民类别 + service_category VARCHAR(50), + -- 订单金额 amount DECIMAL(10, 2) NOT NULL, + -- 货币类型 currency VARCHAR(10) NOT NULL DEFAULT 'CNY', - payment_method VARCHAR(20) NOT NULL, - status VARCHAR(20) NOT NULL DEFAULT 'PENDING', - product_type VARCHAR(50) NOT NULL, - product_id VARCHAR(100), - product_name VARCHAR(200), - description TEXT, - external_order_no VARCHAR(100), + -- 订单状态 + status VARCHAR(20) NOT NULL DEFAULT 'CREATED' + CHECK (status IN ('CREATED', 'PENDING_PAYMENT', 'PAID', 'PROCESSING', 'COMPLETED', 'CANCELLED', 'REFUNDED')), + -- 支付方式: ALIPAY, WECHAT, CREDIT_CARD + payment_method VARCHAR(20), + -- 关联的支付ID + payment_id UUID, + -- 优惠券ID + coupon_id UUID, + -- 优惠金额 + discount_amount DECIMAL(10, 2) DEFAULT 0, + -- 实付金额 + paid_amount DECIMAL(10, 2), + -- 支付时间 paid_at TIMESTAMP WITH TIME ZONE, - expired_at TIMESTAMP WITH TIME ZONE, - metadata JSONB DEFAULT '{}', - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - updated_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP + -- 完成时间 + completed_at TIMESTAMP WITH TIME ZONE, + -- 取消时间 + cancelled_at TIMESTAMP WITH TIME ZONE, + -- 取消原因 + cancel_reason TEXT, + -- 退款时间 + refunded_at TIMESTAMP WITH TIME ZONE, + -- 退款原因 + refund_reason TEXT, + -- 订单备注 + notes TEXT, + -- 扩展元数据（评估结果等） + metadata JSONB, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() ); -CREATE INDEX idx_payment_orders_user_id ON payment_orders(user_id); -CREATE INDEX idx_payment_orders_order_no ON payment_orders(order_no); -CREATE INDEX idx_payment_orders_status ON payment_orders(status); -CREATE INDEX idx_payment_orders_external_order_no ON payment_orders(external_order_no); +COMMENT ON TABLE orders IS '订单表 - 存储用户购买的服务订单'; +COMMENT ON COLUMN orders.order_no IS '订单号，用于对外展示，格式：ORD + 年月日 + 序号'; +COMMENT ON COLUMN orders.service_type IS '服务类型: ASSESSMENT=移民评估, CONSULTATION=付费咨询, DOCUMENT_REVIEW=文档审核'; +COMMENT ON COLUMN orders.metadata IS '扩展数据，如评估结果详情等'; --- 账户余额表 -CREATE TABLE IF NOT EXISTS user_balances ( +CREATE INDEX idx_orders_order_no ON orders(order_no); +CREATE INDEX idx_orders_user_id ON orders(user_id); +CREATE INDEX idx_orders_status ON orders(status); +CREATE INDEX idx_orders_service_type ON orders(service_type); +CREATE INDEX idx_orders_created_at ON orders(created_at DESC); +CREATE INDEX idx_orders_paid_at ON orders(paid_at) WHERE paid_at IS NOT NULL; + +-- =========================================== +-- 支付表 (payments) +-- 存储支付交易记录 +-- =========================================== +CREATE TABLE payments ( id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - user_id UUID NOT NULL UNIQUE REFERENCES users(id), - balance DECIMAL(10, 2) NOT NULL DEFAULT 0, - total_recharged DECIMAL(10, 2) NOT NULL DEFAULT 0, - total_consumed DECIMAL(10, 2) NOT NULL DEFAULT 0, - free_quota INTEGER NOT NULL DEFAULT 5, - version INTEGER NOT NULL DEFAULT 1, - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP, - updated_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP -); - -CREATE INDEX idx_user_balances_user_id ON user_balances(user_id); - --- 账户流水表 -CREATE TABLE IF NOT EXISTS balance_transactions ( - id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), - user_id UUID NOT NULL REFERENCES users(id), - type VARCHAR(20) NOT NULL, + -- 关联的订单ID + order_id UUID REFERENCES orders(id) ON DELETE CASCADE, + -- 支付方式 + method VARCHAR(20) NOT NULL CHECK (method IN ('ALIPAY', 'WECHAT', 'CREDIT_CARD')), + -- 支付金额 amount DECIMAL(10, 2) NOT NULL, - balance_before DECIMAL(10, 2) NOT NULL, - balance_after DECIMAL(10, 2) NOT NULL, - reference_type VARCHAR(50), - reference_id UUID, - description TEXT, - created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP + -- 货币类型 + currency VARCHAR(10) NOT NULL DEFAULT 'CNY', + -- 支付状态 + status VARCHAR(20) NOT NULL DEFAULT 'PENDING' + CHECK (status IN ('PENDING', 'PROCESSING', 'COMPLETED', 'FAILED', 'REFUNDED', 'CANCELLED')), + -- 第三方交易号（支付宝/微信/Stripe的交易ID） + transaction_id VARCHAR(255), + -- 支付二维码URL（支付宝/微信） + qr_code_url TEXT, + -- 支付页面URL（Stripe） + payment_url TEXT, + -- 支付过期时间 + expires_at TIMESTAMP WITH TIME ZONE NOT NULL, + -- 实际支付时间 + paid_at TIMESTAMP WITH TIME ZONE, + -- 失败原因 + failed_reason TEXT, + -- 第三方回调原始数据（用于对账和排查问题） + callback_payload JSONB, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() ); -CREATE INDEX idx_balance_transactions_user_id ON balance_transactions(user_id); -CREATE INDEX idx_balance_transactions_type ON balance_transactions(type); -CREATE INDEX idx_balance_transactions_created_at ON balance_transactions(created_at); +COMMENT ON TABLE payments IS '支付表 - 存储支付交易记录'; +COMMENT ON COLUMN payments.transaction_id IS '第三方支付平台的交易号，用于对账'; +COMMENT ON COLUMN payments.callback_payload IS '支付回调的原始数据，用于问题排查和对账'; + +CREATE INDEX idx_payments_order_id ON payments(order_id); +CREATE INDEX idx_payments_status ON payments(status); +CREATE INDEX idx_payments_method ON payments(method); +CREATE INDEX idx_payments_transaction_id ON payments(transaction_id); +CREATE INDEX idx_payments_created_at ON payments(created_at); -- =========================================== --- 初始数据 +-- 分类账/财务流水表 (ledger_entries) +-- 记录所有资金流动，支持财务对账和报表 -- =========================================== +CREATE TABLE ledger_entries ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 流水号（唯一标识） + entry_no VARCHAR(50) UNIQUE NOT NULL, + -- 流水类型 + entry_type VARCHAR(30) NOT NULL + CHECK (entry_type IN ('INCOME', 'REFUND', 'COMMISSION', 'WITHDRAWAL', 'ADJUSTMENT')), + -- 关联订单ID + order_id UUID REFERENCES orders(id) ON DELETE SET NULL, + -- 关联支付ID + payment_id UUID REFERENCES payments(id) ON DELETE SET NULL, + -- 关联用户ID + user_id UUID REFERENCES users(id) ON DELETE SET NULL, + -- 金额（正数为收入，负数为支出） + amount DECIMAL(12, 2) NOT NULL, + -- 货币类型 + currency VARCHAR(10) NOT NULL DEFAULT 'CNY', + -- 账户余额（流水后的账户余额快照） + balance_after DECIMAL(12, 2), + -- 交易渠道: ALIPAY, WECHAT, STRIPE, BANK, MANUAL + channel VARCHAR(30), + -- 第三方交易号 + transaction_id VARCHAR(255), + -- 业务类型: ASSESSMENT, CONSULTATION, DOCUMENT_REVIEW + business_type VARCHAR(50), + -- 业务类别: QMAS, GEP, IANG, TTPS, CIES, TECHTAS + business_category VARCHAR(50), + -- 摘要/描述 + description TEXT, + -- 备注 + notes TEXT, + -- 状态: PENDING(待确认), CONFIRMED(已确认), CANCELLED(已取消) + status VARCHAR(20) NOT NULL DEFAULT 'CONFIRMED' + CHECK (status IN ('PENDING', 'CONFIRMED', 'CANCELLED')), + -- 确认时间 + confirmed_at TIMESTAMP WITH TIME ZONE, + -- 确认人（管理员ID） + confirmed_by UUID, + -- 会计期间（格式：YYYY-MM） + accounting_period VARCHAR(7), + -- 创建时间（流水发生时间） + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); --- 创建默认管理员 (密码: admin123) --- 密码哈希使用 bcrypt 生成: echo -n "admin123" | htpasswd -bnBC 10 "" - | tr -d ':\n' -INSERT INTO admins (id, username, password_hash, name, role, permissions) -VALUES ( - uuid_generate_v4(), - 'admin', - '$2b$10$N9qo8uLOickgx2ZMRZoMyeIjZRGdjGj/n3.QHZWL1xOsV7.Hq5mLe', - '系统管理员', - 'SUPER_ADMIN', - '["*"]' -) ON CONFLICT (username) DO NOTHING; +COMMENT ON TABLE ledger_entries IS '分类账/财务流水表 - 记录所有资金流动，支持财务对账和报表生成'; +COMMENT ON COLUMN ledger_entries.entry_no IS '流水号，格式：LE + 年月日 + 序号'; +COMMENT ON COLUMN ledger_entries.entry_type IS '流水类型: INCOME=收入, REFUND=退款, COMMISSION=佣金, WITHDRAWAL=提现, ADJUSTMENT=调整'; +COMMENT ON COLUMN ledger_entries.balance_after IS '该笔流水后的账户余额快照，用于对账'; +COMMENT ON COLUMN ledger_entries.accounting_period IS '会计期间，格式YYYY-MM，用于月度报表'; + +CREATE INDEX idx_ledger_entries_entry_no ON ledger_entries(entry_no); +CREATE INDEX idx_ledger_entries_entry_type ON ledger_entries(entry_type); +CREATE INDEX idx_ledger_entries_order_id ON ledger_entries(order_id); +CREATE INDEX idx_ledger_entries_user_id ON ledger_entries(user_id); +CREATE INDEX idx_ledger_entries_status ON ledger_entries(status); +CREATE INDEX idx_ledger_entries_accounting_period ON ledger_entries(accounting_period); +CREATE INDEX idx_ledger_entries_created_at ON ledger_entries(created_at DESC); +CREATE INDEX idx_ledger_entries_business_type ON ledger_entries(business_type); -- =========================================== --- 更新触发器 +-- 日统计表 (daily_statistics) +-- 预聚合的每日统计数据，用于快速报表查询 +-- =========================================== +CREATE TABLE daily_statistics ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 统计日期 + stat_date DATE NOT NULL, + -- 统计维度: OVERALL(总体), CHANNEL(渠道), CATEGORY(移民类别) + dimension VARCHAR(30) NOT NULL DEFAULT 'OVERALL', + -- 维度值（如渠道名称、类别代码） + dimension_value VARCHAR(50), + + -- ===== 用户统计 ===== + -- 新增用户数 + new_users INT DEFAULT 0, + -- 新增注册用户数 + new_registered_users INT DEFAULT 0, + -- 活跃用户数 + active_users INT DEFAULT 0, + + -- ===== 对话统计 ===== + -- 新增对话数 + new_conversations INT DEFAULT 0, + -- 总消息数 + total_messages INT DEFAULT 0, + -- 用户消息数 + user_messages INT DEFAULT 0, + -- AI消息数 + assistant_messages INT DEFAULT 0, + -- 平均对话轮次 + avg_conversation_turns DECIMAL(10, 2) DEFAULT 0, + + -- ===== 订单统计 ===== + -- 新增订单数 + new_orders INT DEFAULT 0, + -- 支付成功订单数 + paid_orders INT DEFAULT 0, + -- 订单总金额 + total_order_amount DECIMAL(12, 2) DEFAULT 0, + -- 实收金额 + total_paid_amount DECIMAL(12, 2) DEFAULT 0, + -- 退款订单数 + refunded_orders INT DEFAULT 0, + -- 退款金额 + refund_amount DECIMAL(12, 2) DEFAULT 0, + + -- ===== 转化统计 ===== + -- 咨询转化数（对话转订单） + conversion_count INT DEFAULT 0, + -- 转化率 + conversion_rate DECIMAL(5, 4) DEFAULT 0, + + -- ===== Token消耗统计 ===== + -- 输入Token总数 + total_input_tokens BIGINT DEFAULT 0, + -- 输出Token总数 + total_output_tokens BIGINT DEFAULT 0, + -- 预估API成本（美元） + estimated_api_cost DECIMAL(10, 4) DEFAULT 0, + + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + + -- 唯一约束：同一天同一维度同一维度值只有一条记录 + UNIQUE(stat_date, dimension, dimension_value) +); + +COMMENT ON TABLE daily_statistics IS '日统计表 - 预聚合的每日统计数据，支持多维度分析'; +COMMENT ON COLUMN daily_statistics.dimension IS '统计维度: OVERALL=总体统计, CHANNEL=按渠道, CATEGORY=按移民类别'; +COMMENT ON COLUMN daily_statistics.conversion_rate IS '转化率 = 支付订单数 / 新对话数'; +COMMENT ON COLUMN daily_statistics.estimated_api_cost IS '预估Claude API成本，基于Token消耗计算'; + +CREATE INDEX idx_daily_statistics_stat_date ON daily_statistics(stat_date DESC); +CREATE INDEX idx_daily_statistics_dimension ON daily_statistics(dimension, dimension_value); + +-- =========================================== +-- 月度财务报表 (monthly_financial_reports) +-- 月度汇总的财务数据 +-- =========================================== +CREATE TABLE monthly_financial_reports ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 报表月份（格式：YYYY-MM） + report_month VARCHAR(7) NOT NULL UNIQUE, + + -- ===== 收入统计 ===== + -- 总收入 + total_revenue DECIMAL(12, 2) DEFAULT 0, + -- 评估服务收入 + assessment_revenue DECIMAL(12, 2) DEFAULT 0, + -- 咨询服务收入 + consultation_revenue DECIMAL(12, 2) DEFAULT 0, + -- 其他收入 + other_revenue DECIMAL(12, 2) DEFAULT 0, + + -- ===== 退款统计 ===== + -- 总退款 + total_refunds DECIMAL(12, 2) DEFAULT 0, + -- 净收入（总收入 - 退款） + net_revenue DECIMAL(12, 2) DEFAULT 0, + + -- ===== 成本统计 ===== + -- API成本（Claude） + api_cost DECIMAL(10, 2) DEFAULT 0, + -- 支付手续费 + payment_fees DECIMAL(10, 2) DEFAULT 0, + -- 其他成本 + other_costs DECIMAL(10, 2) DEFAULT 0, + -- 总成本 + total_costs DECIMAL(10, 2) DEFAULT 0, + + -- ===== 利润统计 ===== + -- 毛利润 + gross_profit DECIMAL(12, 2) DEFAULT 0, + -- 毛利率 + gross_margin DECIMAL(5, 4) DEFAULT 0, + + -- ===== 订单统计 ===== + -- 总订单数 + total_orders INT DEFAULT 0, + -- 成功订单数 + successful_orders INT DEFAULT 0, + -- 平均订单金额 + avg_order_amount DECIMAL(10, 2) DEFAULT 0, + + -- ===== 按类别收入明细（JSONB） ===== + revenue_by_category JSONB DEFAULT '{}', + + -- ===== 按渠道收入明细（JSONB） ===== + revenue_by_channel JSONB DEFAULT '{}', + + -- 报表状态: DRAFT(草稿), CONFIRMED(已确认), LOCKED(已锁定) + status VARCHAR(20) NOT NULL DEFAULT 'DRAFT' + CHECK (status IN ('DRAFT', 'CONFIRMED', 'LOCKED')), + -- 确认人 + confirmed_by UUID, + -- 确认时间 + confirmed_at TIMESTAMP WITH TIME ZONE, + -- 备注 + notes TEXT, + + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE monthly_financial_reports IS '月度财务报表 - 月度汇总的财务数据，支持确认和锁定'; +COMMENT ON COLUMN monthly_financial_reports.report_month IS '报表月份，格式YYYY-MM'; +COMMENT ON COLUMN monthly_financial_reports.revenue_by_category IS '按移民类别的收入明细，JSON格式'; +COMMENT ON COLUMN monthly_financial_reports.status IS 'DRAFT=草稿可修改, CONFIRMED=已确认, LOCKED=已锁定不可修改'; + +CREATE INDEX idx_monthly_financial_reports_month ON monthly_financial_reports(report_month DESC); +CREATE INDEX idx_monthly_financial_reports_status ON monthly_financial_reports(status); + +-- =========================================== +-- 审计日志表 (audit_logs) +-- 记录所有重要操作，用于安全审计和问题追踪 +-- =========================================== +CREATE TABLE audit_logs ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 操作者ID（用户或管理员） + actor_id UUID, + -- 操作者类型: USER, ADMIN, SYSTEM + actor_type VARCHAR(20) NOT NULL CHECK (actor_type IN ('USER', 'ADMIN', 'SYSTEM')), + -- 操作者名称/标识 + actor_name VARCHAR(100), + -- 操作类型 + action VARCHAR(50) NOT NULL, + -- 操作对象类型（表名） + entity_type VARCHAR(50) NOT NULL, + -- 操作对象ID + entity_id UUID, + -- 操作前的数据快照 + old_values JSONB, + -- 操作后的数据快照 + new_values JSONB, + -- 变更的字段列表 + changed_fields TEXT[], + -- 操作描述 + description TEXT, + -- 客户端IP地址 + ip_address INET, + -- 用户代理（浏览器信息） + user_agent TEXT, + -- 请求ID（用于追踪） + request_id VARCHAR(100), + -- 操作结果: SUCCESS, FAILED + result VARCHAR(20) DEFAULT 'SUCCESS' CHECK (result IN ('SUCCESS', 'FAILED')), + -- 错误信息（如果失败） + error_message TEXT, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE audit_logs IS '审计日志表 - 记录所有重要操作，用于安全审计'; +COMMENT ON COLUMN audit_logs.action IS '操作类型，如: CREATE, UPDATE, DELETE, LOGIN, LOGOUT, PAYMENT等'; +COMMENT ON COLUMN audit_logs.old_values IS '操作前的数据快照，用于审计和回滚'; +COMMENT ON COLUMN audit_logs.new_values IS '操作后的数据快照'; + +CREATE INDEX idx_audit_logs_actor_id ON audit_logs(actor_id); +CREATE INDEX idx_audit_logs_actor_type ON audit_logs(actor_type); +CREATE INDEX idx_audit_logs_action ON audit_logs(action); +CREATE INDEX idx_audit_logs_entity_type ON audit_logs(entity_type); +CREATE INDEX idx_audit_logs_entity_id ON audit_logs(entity_id); +CREATE INDEX idx_audit_logs_created_at ON audit_logs(created_at DESC); +-- 按时间范围查询优化 +CREATE INDEX idx_audit_logs_created_at_brin ON audit_logs USING BRIN(created_at); + +-- =========================================== +-- 发件箱表 (outbox) +-- 事务发件箱模式，确保消息可靠发送到Kafka +-- =========================================== +CREATE TABLE outbox ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 聚合根类型（如：Order, Payment, User） + aggregate_type VARCHAR(100) NOT NULL, + -- 聚合根ID + aggregate_id UUID NOT NULL, + -- 事件类型（如：OrderCreated, PaymentCompleted） + event_type VARCHAR(100) NOT NULL, + -- 事件负载（JSON格式的事件数据） + payload JSONB NOT NULL, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 处理时间（消息发送到Kafka的时间） + processed_at TIMESTAMP WITH TIME ZONE, + -- 处理次数（用于重试机制） + retry_count INT DEFAULT 0, + -- 最后错误信息 + last_error TEXT +); + +COMMENT ON TABLE outbox IS '发件箱表 - 事务发件箱模式，确保事件消息可靠发送'; +COMMENT ON COLUMN outbox.aggregate_type IS '聚合根类型，对应领域模型'; +COMMENT ON COLUMN outbox.event_type IS '领域事件类型'; + +CREATE INDEX idx_outbox_unprocessed ON outbox(created_at) WHERE processed_at IS NULL; +CREATE INDEX idx_outbox_aggregate ON outbox(aggregate_type, aggregate_id); + +-- =========================================== +-- 知识文档表 (documents) +-- 存储移民知识库文档，用于RAG检索 +-- =========================================== +CREATE TABLE documents ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 文档标题 + title VARCHAR(255) NOT NULL, + -- 文档内容（原始文本） + content TEXT NOT NULL, + -- 移民类别: QMAS, GEP, IANG, TTPS, CIES, TECHTAS, GENERAL + category VARCHAR(50), + -- 标签数组 + tags TEXT[], + -- 来源名称 + source VARCHAR(255), + -- 来源URL + source_url VARCHAR(500), + -- 文档元数据 + metadata JSONB, + -- 文档版本 + version INT DEFAULT 1, + -- 是否启用 + is_active BOOLEAN DEFAULT TRUE, + -- 最后验证时间（确认信息仍然有效） + last_verified_at TIMESTAMP WITH TIME ZONE, + -- 验证人 + verified_by UUID, + -- 创建人 + created_by UUID, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE documents IS '知识文档表 - 存储移民知识库文档，用于RAG检索增强'; +COMMENT ON COLUMN documents.category IS '移民类别，GENERAL表示通用知识'; +COMMENT ON COLUMN documents.last_verified_at IS '最后验证时间，确保信息的时效性'; + +CREATE INDEX idx_documents_category ON documents(category); +CREATE INDEX idx_documents_is_active ON documents(is_active); +CREATE INDEX idx_documents_tags ON documents USING GIN(tags); +CREATE INDEX idx_documents_created_at ON documents(created_at); + +-- =========================================== +-- 文档向量嵌入表 (document_embeddings) +-- 存储文档分块的向量嵌入，用于相似度搜索 +-- =========================================== +CREATE TABLE document_embeddings ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 所属文档ID + document_id UUID REFERENCES documents(id) ON DELETE CASCADE, + -- 分块序号 + chunk_index INT NOT NULL, + -- 分块内容 + content TEXT NOT NULL, + -- 向量嵌入（1536维，对应text-embedding-3-small） + embedding vector(1536), + -- 分块元数据 + metadata JSONB, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE document_embeddings IS '文档向量嵌入表 - 存储文档分块的向量，用于语义搜索'; +COMMENT ON COLUMN document_embeddings.embedding IS '1536维向量，使用text-embedding-3-small模型生成'; + +CREATE INDEX idx_document_embeddings_document_id ON document_embeddings(document_id); +-- IVFFlat索引，用于快速向量相似度搜索 +CREATE INDEX idx_document_embeddings_embedding ON document_embeddings + USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100); + +-- =========================================== +-- 系统配置表 (system_configs) +-- 存储系统配置项，支持动态调整 +-- =========================================== +CREATE TABLE system_configs ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 配置键（唯一） + key VARCHAR(100) UNIQUE NOT NULL, + -- 配置值（JSON格式） + value JSONB NOT NULL, + -- 配置分组: SYSTEM, PROMPT, PAYMENT, NOTIFICATION, FEATURE + config_group VARCHAR(50) DEFAULT 'SYSTEM', + -- 配置描述 + description TEXT, + -- 是否敏感信息（敏感信息不在日志中显示完整值） + is_sensitive BOOLEAN DEFAULT FALSE, + -- 更新人 + updated_by UUID, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE system_configs IS '系统配置表 - 存储可动态调整的系统配置'; +COMMENT ON COLUMN system_configs.config_group IS '配置分组，便于管理和查询'; +COMMENT ON COLUMN system_configs.is_sensitive IS '敏感配置在日志中脱敏显示'; + +CREATE INDEX idx_system_configs_group ON system_configs(config_group); + +-- =========================================== +-- 管理员用户表 (admin_users) +-- 存储后台管理员信息 +-- =========================================== +CREATE TABLE admin_users ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 用户名（唯一） + username VARCHAR(100) UNIQUE NOT NULL, + -- 邮箱 + email VARCHAR(255), + -- 密码哈希（bcrypt） + password_hash VARCHAR(255) NOT NULL, + -- 角色: SUPER_ADMIN, ADMIN, OPERATOR, VIEWER + role VARCHAR(50) NOT NULL CHECK (role IN ('SUPER_ADMIN', 'ADMIN', 'OPERATOR', 'VIEWER')), + -- 姓名 + display_name VARCHAR(100), + -- 手机号 + phone VARCHAR(20), + -- 是否启用 + is_active BOOLEAN DEFAULT TRUE, + -- 最后登录时间 + last_login_at TIMESTAMP WITH TIME ZONE, + -- 最后登录IP + last_login_ip INET, + -- 登录失败次数（用于账户锁定） + failed_login_attempts INT DEFAULT 0, + -- 账户锁定时间 + locked_until TIMESTAMP WITH TIME ZONE, + -- 备注 + notes TEXT, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE admin_users IS '管理员用户表 - 存储后台管理员信息'; +COMMENT ON COLUMN admin_users.role IS 'SUPER_ADMIN=超级管理员, ADMIN=管理员, OPERATOR=操作员, VIEWER=只读用户'; +COMMENT ON COLUMN admin_users.failed_login_attempts IS '连续登录失败次数，超过阈值锁定账户'; + +CREATE INDEX idx_admin_users_username ON admin_users(username); +CREATE INDEX idx_admin_users_email ON admin_users(email); +CREATE INDEX idx_admin_users_role ON admin_users(role); +CREATE INDEX idx_admin_users_is_active ON admin_users(is_active); + +-- =========================================== +-- 经验库表 (experiences) +-- 存储从对话中提取的经验，用于系统自我进化 +-- =========================================== +CREATE TABLE experiences ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 经验类型 + type VARCHAR(50) NOT NULL + CHECK (type IN ('FAQ_PATTERN', 'USER_CONCERN', 'EFFECTIVE_RESPONSE', 'CONVERSION_PATTERN', 'CONFUSION_POINT')), + -- 模式描述 + pattern TEXT NOT NULL, + -- 洞察/经验总结 + insight TEXT NOT NULL, + -- 出现频率 + frequency INT DEFAULT 1, + -- 置信度（0-1） + confidence DECIMAL(3, 2) DEFAULT 0.5 CHECK (confidence >= 0 AND confidence <= 1), + -- 相关移民类别 + category VARCHAR(50), + -- 来源对话ID列表 + source_conversation_ids UUID[], + -- 示例问题 + sample_questions TEXT[], + -- 示例回答 + sample_responses TEXT[], + -- 元数据 + metadata JSONB, + -- 是否已应用 + is_applied BOOLEAN DEFAULT FALSE, + -- 应用时间 + applied_at TIMESTAMP WITH TIME ZONE, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE experiences IS '经验库表 - 存储从对话中提取的经验，支持系统自我进化'; +COMMENT ON COLUMN experiences.type IS '经验类型: FAQ_PATTERN=常见问题模式, USER_CONCERN=用户关注点, EFFECTIVE_RESPONSE=有效回答, CONVERSION_PATTERN=转化模式, CONFUSION_POINT=困惑点'; +COMMENT ON COLUMN experiences.confidence IS '置信度，值越高表示该经验越可靠'; + +CREATE INDEX idx_experiences_type ON experiences(type); +CREATE INDEX idx_experiences_category ON experiences(category); +CREATE INDEX idx_experiences_confidence ON experiences(confidence DESC); +CREATE INDEX idx_experiences_is_applied ON experiences(is_applied); + +-- =========================================== +-- 进化日志表 (evolution_logs) +-- 记录系统配置和行为的变更历史 +-- =========================================== +CREATE TABLE evolution_logs ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 触发人（管理员ID） + triggered_by UUID REFERENCES admin_users(id), + -- 变更类型 + type VARCHAR(50) NOT NULL + CHECK (type IN ('PROMPT_UPDATE', 'KNOWLEDGE_UPDATE', 'RULE_UPDATE', 'BEHAVIOR_UPDATE')), + -- 变更状态 + status VARCHAR(20) NOT NULL DEFAULT 'PROPOSED' + CHECK (status IN ('PROPOSED', 'APPROVED', 'APPLIED', 'ROLLED_BACK', 'REJECTED')), + -- 变更内容 + changes JSONB NOT NULL, + -- 变更原因/说明 + reason TEXT, + -- 关联的经验ID列表 + related_experience_ids UUID[], + -- 审批人 + approved_by UUID REFERENCES admin_users(id), + -- 审批时间 + approved_at TIMESTAMP WITH TIME ZONE, + -- 应用时间 + applied_at TIMESTAMP WITH TIME ZONE, + -- 回滚时间 + rollback_at TIMESTAMP WITH TIME ZONE, + -- 回滚原因 + rollback_reason TEXT, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE evolution_logs IS '进化日志表 - 记录系统配置和行为的变更历史'; +COMMENT ON COLUMN evolution_logs.type IS '变更类型: PROMPT_UPDATE=提示词更新, KNOWLEDGE_UPDATE=知识库更新, RULE_UPDATE=规则更新, BEHAVIOR_UPDATE=行为更新'; +COMMENT ON COLUMN evolution_logs.status IS '变更状态流转: PROPOSED -> APPROVED -> APPLIED, 可回滚为ROLLED_BACK'; + +CREATE INDEX idx_evolution_logs_status ON evolution_logs(status); +CREATE INDEX idx_evolution_logs_type ON evolution_logs(type); +CREATE INDEX idx_evolution_logs_triggered_by ON evolution_logs(triggered_by); +CREATE INDEX idx_evolution_logs_created_at ON evolution_logs(created_at DESC); + +-- =========================================== +-- 验证码表 (verification_codes) +-- 存储手机验证码 +-- =========================================== +CREATE TABLE verification_codes ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 手机号 + phone VARCHAR(20) NOT NULL, + -- 验证码 + code VARCHAR(10) NOT NULL, + -- 过期时间 + expires_at TIMESTAMP WITH TIME ZONE NOT NULL, + -- 是否已使用 + is_used BOOLEAN DEFAULT FALSE, + -- 使用时间 + used_at TIMESTAMP WITH TIME ZONE, + -- IP地址（用于防刷） + ip_address INET, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE verification_codes IS '验证码表 - 存储手机验证码，支持防刷和过期机制'; + +CREATE INDEX idx_verification_codes_phone ON verification_codes(phone); +CREATE INDEX idx_verification_codes_expires_at ON verification_codes(expires_at); +CREATE INDEX idx_verification_codes_ip ON verification_codes(ip_address); + +-- =========================================== +-- 服务定价表 (service_pricing) +-- 存储各类服务的定价信息 +-- =========================================== +CREATE TABLE service_pricing ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 服务类型 + service_type VARCHAR(50) NOT NULL + CHECK (service_type IN ('ASSESSMENT', 'CONSULTATION', 'DOCUMENT_REVIEW')), + -- 移民类别 + category VARCHAR(50), + -- 价格 + price DECIMAL(10, 2) NOT NULL, + -- 原价（用于显示折扣） + original_price DECIMAL(10, 2), + -- 货币 + currency VARCHAR(10) NOT NULL DEFAULT 'CNY', + -- 服务描述 + description TEXT, + -- 服务详情（JSON格式） + details JSONB, + -- 是否启用 + is_active BOOLEAN DEFAULT TRUE, + -- 生效开始时间 + effective_from TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 生效结束时间 + effective_until TIMESTAMP WITH TIME ZONE, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 唯一约束 + UNIQUE(service_type, category) +); + +COMMENT ON TABLE service_pricing IS '服务定价表 - 存储各类服务的定价信息'; +COMMENT ON COLUMN service_pricing.original_price IS '原价，用于显示折扣效果'; +COMMENT ON COLUMN service_pricing.effective_from IS '价格生效开始时间，支持定时调价'; + +CREATE INDEX idx_service_pricing_service_type ON service_pricing(service_type); +CREATE INDEX idx_service_pricing_is_active ON service_pricing(is_active); + +-- =========================================== +-- 优惠券表 (coupons) +-- 存储优惠券信息 +-- =========================================== +CREATE TABLE coupons ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 优惠券码 + code VARCHAR(50) UNIQUE NOT NULL, + -- 优惠券名称 + name VARCHAR(100) NOT NULL, + -- 优惠类型: FIXED(固定金额), PERCENTAGE(百分比) + discount_type VARCHAR(20) NOT NULL CHECK (discount_type IN ('FIXED', 'PERCENTAGE')), + -- 优惠值（固定金额或百分比） + discount_value DECIMAL(10, 2) NOT NULL, + -- 最低消费金额 + min_amount DECIMAL(10, 2) DEFAULT 0, + -- 最高优惠金额（百分比优惠时的上限） + max_discount DECIMAL(10, 2), + -- 适用服务类型（为空表示全部适用） + applicable_services TEXT[], + -- 适用移民类别（为空表示全部适用） + applicable_categories TEXT[], + -- 总发行量 + total_quantity INT, + -- 已使用数量 + used_quantity INT DEFAULT 0, + -- 每人限用次数 + per_user_limit INT DEFAULT 1, + -- 生效时间 + valid_from TIMESTAMP WITH TIME ZONE NOT NULL, + -- 失效时间 + valid_until TIMESTAMP WITH TIME ZONE NOT NULL, + -- 是否启用 + is_active BOOLEAN DEFAULT TRUE, + -- 备注 + notes TEXT, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE coupons IS '优惠券表 - 存储优惠券信息，支持固定金额和百分比折扣'; + +CREATE INDEX idx_coupons_code ON coupons(code); +CREATE INDEX idx_coupons_is_active ON coupons(is_active); +CREATE INDEX idx_coupons_valid_until ON coupons(valid_until); + +-- =========================================== +-- 用户优惠券表 (user_coupons) +-- 存储用户领取/使用的优惠券 +-- =========================================== +CREATE TABLE user_coupons ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 用户ID + user_id UUID REFERENCES users(id) ON DELETE CASCADE, + -- 优惠券ID + coupon_id UUID REFERENCES coupons(id) ON DELETE CASCADE, + -- 状态: AVAILABLE(可用), USED(已使用), EXPIRED(已过期) + status VARCHAR(20) NOT NULL DEFAULT 'AVAILABLE' + CHECK (status IN ('AVAILABLE', 'USED', 'EXPIRED')), + -- 使用的订单ID + used_order_id UUID REFERENCES orders(id) ON DELETE SET NULL, + -- 使用时间 + used_at TIMESTAMP WITH TIME ZONE, + -- 领取时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE user_coupons IS '用户优惠券表 - 记录用户领取和使用优惠券的情况'; + +CREATE INDEX idx_user_coupons_user_id ON user_coupons(user_id); +CREATE INDEX idx_user_coupons_coupon_id ON user_coupons(coupon_id); +CREATE INDEX idx_user_coupons_status ON user_coupons(status); + +-- =========================================== +-- 插入默认数据 -- =========================================== +-- 默认管理员用户 (密码: admin123) +INSERT INTO admin_users (username, email, password_hash, role, display_name) +VALUES ('admin', 'admin@iconsulting.com', '$2b$10$rqKu.wB1E8z9vQzKK0FZMuXz.B7xPc5S7BUTr6G9TkVdEX5C0Q.Wy', 'SUPER_ADMIN', '系统管理员'); + +-- 默认系统配置 +INSERT INTO system_configs (key, value, config_group, description) VALUES +('system_prompt_identity', '"专业、友善、耐心的香港移民顾问"', 'PROMPT', '系统身份定位'), +('system_prompt_style', '"专业但不生硬，用简洁明了的语言解答"', 'PROMPT', '对话风格'), +('system_prompt_rules', '["只回答移民相关问题", "复杂评估建议付费服务", "不做法律承诺"]', 'PROMPT', '系统行为规则'), +('max_free_messages_per_day', '50', 'SYSTEM', '每日免费消息数限制'), +('max_conversation_context_messages', '20', 'SYSTEM', '对话上下文最大消息数'), +('assessment_price_default', '99', 'PAYMENT', '默认评估价格(元)'), +('payment_timeout_minutes', '30', 'PAYMENT', '支付超时时间(分钟)'), +('sms_rate_limit_per_hour', '5', 'SYSTEM', '每小时短信发送限制'), +('enable_anonymous_chat', 'true', 'FEATURE', '是否允许匿名用户聊天'), +('require_phone_for_payment', 'true', 'FEATURE', '支付时是否要求手机验证'); + +-- 默认服务定价 +INSERT INTO service_pricing (service_type, category, price, original_price, currency, description) VALUES +('ASSESSMENT', 'QMAS', 99.00, 199.00, 'CNY', '优才计划移民资格评估'), +('ASSESSMENT', 'GEP', 99.00, 199.00, 'CNY', '专才计划移民资格评估'), +('ASSESSMENT', 'IANG', 79.00, 149.00, 'CNY', '留学IANG移民资格评估'), +('ASSESSMENT', 'TTPS', 99.00, 199.00, 'CNY', '高才通移民资格评估'), +('ASSESSMENT', 'CIES', 199.00, 399.00, 'CNY', '投资移民资格评估'), +('ASSESSMENT', 'TECHTAS', 99.00, 199.00, 'CNY', '科技人才移民资格评估'); + +-- =========================================== +-- 创建函数和触发器 +-- =========================================== + +-- 更新 updated_at 时间戳的通用函数 CREATE OR REPLACE FUNCTION update_updated_at_column() RETURNS TRIGGER AS $$ BEGIN - NEW.updated_at = CURRENT_TIMESTAMP; + NEW.updated_at = NOW(); RETURN NEW; END; $$ language 'plpgsql'; --- 为所有表添加 updated_at 触发器 -DO $$ -DECLARE - t text; -BEGIN - FOR t IN - SELECT table_name - FROM information_schema.columns - WHERE column_name = 'updated_at' - AND table_schema = 'public' - LOOP - EXECUTE format(' - DROP TRIGGER IF EXISTS update_%I_updated_at ON %I; - CREATE TRIGGER update_%I_updated_at - BEFORE UPDATE ON %I - FOR EACH ROW - EXECUTE FUNCTION update_updated_at_column(); - ', t, t, t, t); - END LOOP; -END; -$$ LANGUAGE plpgsql; +-- 为所有需要的表添加更新时间触发器 +CREATE TRIGGER update_users_updated_at BEFORE UPDATE ON users FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_conversations_updated_at BEFORE UPDATE ON conversations FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_orders_updated_at BEFORE UPDATE ON orders FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_payments_updated_at BEFORE UPDATE ON payments FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_documents_updated_at BEFORE UPDATE ON documents FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_system_configs_updated_at BEFORE UPDATE ON system_configs FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_admin_users_updated_at BEFORE UPDATE ON admin_users FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_experiences_updated_at BEFORE UPDATE ON experiences FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_service_pricing_updated_at BEFORE UPDATE ON service_pricing FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_coupons_updated_at BEFORE UPDATE ON coupons FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_ledger_entries_updated_at BEFORE UPDATE ON ledger_entries FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_daily_statistics_updated_at BEFORE UPDATE ON daily_statistics FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); +CREATE TRIGGER update_monthly_financial_reports_updated_at BEFORE UPDATE ON monthly_financial_reports FOR EACH ROW EXECUTE FUNCTION update_updated_at_column(); --- 完成 -SELECT 'Database initialization completed!' as status; +-- 更新对话消息计数的函数 +CREATE OR REPLACE FUNCTION update_conversation_message_count() +RETURNS TRIGGER AS $$ +BEGIN + UPDATE conversations + SET + message_count = message_count + 1, + user_message_count = user_message_count + CASE WHEN NEW.role = 'user' THEN 1 ELSE 0 END, + assistant_message_count = assistant_message_count + CASE WHEN NEW.role = 'assistant' THEN 1 ELSE 0 END, + total_input_tokens = total_input_tokens + COALESCE(NEW.input_tokens, 0), + total_output_tokens = total_output_tokens + COALESCE(NEW.output_tokens, 0) + WHERE id = NEW.conversation_id; + RETURN NEW; +END; +$$ language 'plpgsql'; + +CREATE TRIGGER increment_conversation_message_count + AFTER INSERT ON messages + FOR EACH ROW + EXECUTE FUNCTION update_conversation_message_count(); + +-- 生成订单号的函数 +CREATE OR REPLACE FUNCTION generate_order_no() +RETURNS TRIGGER AS $$ +DECLARE + seq_val INT; +BEGIN + -- 获取当天的序号 + SELECT COALESCE(MAX(CAST(SUBSTRING(order_no FROM 12) AS INT)), 0) + 1 INTO seq_val + FROM orders + WHERE order_no LIKE 'ORD' || TO_CHAR(NOW(), 'YYYYMMDD') || '%'; + + NEW.order_no := 'ORD' || TO_CHAR(NOW(), 'YYYYMMDD') || LPAD(seq_val::TEXT, 6, '0'); + RETURN NEW; +END; +$$ language 'plpgsql'; + +CREATE TRIGGER generate_order_no_trigger + BEFORE INSERT ON orders + FOR EACH ROW + WHEN (NEW.order_no IS NULL) + EXECUTE FUNCTION generate_order_no(); + +-- 生成分类账流水号的函数 +CREATE OR REPLACE FUNCTION generate_ledger_entry_no() +RETURNS TRIGGER AS $$ +DECLARE + seq_val INT; +BEGIN + SELECT COALESCE(MAX(CAST(SUBSTRING(entry_no FROM 11) AS INT)), 0) + 1 INTO seq_val + FROM ledger_entries + WHERE entry_no LIKE 'LE' || TO_CHAR(NOW(), 'YYYYMMDD') || '%'; + + NEW.entry_no := 'LE' || TO_CHAR(NOW(), 'YYYYMMDD') || LPAD(seq_val::TEXT, 6, '0'); + NEW.accounting_period := TO_CHAR(NOW(), 'YYYY-MM'); + RETURN NEW; +END; +$$ language 'plpgsql'; + +CREATE TRIGGER generate_ledger_entry_no_trigger + BEFORE INSERT ON ledger_entries + FOR EACH ROW + WHEN (NEW.entry_no IS NULL) + EXECUTE FUNCTION generate_ledger_entry_no(); + +-- 支付成功后自动创建分类账条目的函数 +CREATE OR REPLACE FUNCTION create_ledger_entry_on_payment() +RETURNS TRIGGER AS $$ +BEGIN + -- 仅当支付状态变为COMPLETED时创建收入流水 + IF NEW.status = 'COMPLETED' AND (OLD.status IS NULL OR OLD.status != 'COMPLETED') THEN + INSERT INTO ledger_entries ( + entry_type, + order_id, + payment_id, + user_id, + amount, + currency, + channel, + transaction_id, + description, + status + ) + SELECT + 'INCOME', + NEW.order_id, + NEW.id, + o.user_id, + NEW.amount, + NEW.currency, + NEW.method, + NEW.transaction_id, + '订单支付: ' || o.order_no || ' - ' || o.service_type || COALESCE(' (' || o.service_category || ')', ''), + 'CONFIRMED' + FROM orders o + WHERE o.id = NEW.order_id; + END IF; + + -- 退款时创建退款流水 + IF NEW.status = 'REFUNDED' AND OLD.status != 'REFUNDED' THEN + INSERT INTO ledger_entries ( + entry_type, + order_id, + payment_id, + user_id, + amount, + currency, + channel, + transaction_id, + description, + status + ) + SELECT + 'REFUND', + NEW.order_id, + NEW.id, + o.user_id, + -NEW.amount, -- 退款为负数 + NEW.currency, + NEW.method, + NEW.transaction_id, + '订单退款: ' || o.order_no, + 'CONFIRMED' + FROM orders o + WHERE o.id = NEW.order_id; + END IF; + + RETURN NEW; +END; +$$ language 'plpgsql'; + +CREATE TRIGGER create_ledger_entry_on_payment_trigger + AFTER UPDATE ON payments + FOR EACH ROW + EXECUTE FUNCTION create_ledger_entry_on_payment(); + +-- =========================================== +-- 创建统计视图 +-- =========================================== + +-- 今日实时统计视图 +CREATE OR REPLACE VIEW v_today_statistics AS +SELECT + COUNT(DISTINCT CASE WHEN u.created_at >= CURRENT_DATE THEN u.id END) AS new_users_today, + COUNT(DISTINCT CASE WHEN u.created_at >= CURRENT_DATE AND u.type = 'REGISTERED' THEN u.id END) AS new_registered_today, + COUNT(DISTINCT CASE WHEN c.created_at >= CURRENT_DATE THEN c.id END) AS new_conversations_today, + COUNT(CASE WHEN m.created_at >= CURRENT_DATE THEN 1 END) AS messages_today, + COUNT(DISTINCT CASE WHEN o.created_at >= CURRENT_DATE THEN o.id END) AS new_orders_today, + COUNT(DISTINCT CASE WHEN o.paid_at >= CURRENT_DATE THEN o.id END) AS paid_orders_today, + COALESCE(SUM(CASE WHEN o.paid_at >= CURRENT_DATE THEN o.paid_amount END), 0) AS revenue_today +FROM users u +FULL OUTER JOIN conversations c ON TRUE +FULL OUTER JOIN messages m ON TRUE +FULL OUTER JOIN orders o ON TRUE; + +COMMENT ON VIEW v_today_statistics IS '今日实时统计视图 - 展示当天的关键指标'; + +-- 移民类别统计视图 +CREATE OR REPLACE VIEW v_category_statistics AS +SELECT + c.category, + COUNT(DISTINCT c.id) AS total_conversations, + COUNT(DISTINCT c.user_id) AS unique_users, + COUNT(DISTINCT CASE WHEN c.has_converted THEN c.id END) AS converted_conversations, + ROUND(COUNT(DISTINCT CASE WHEN c.has_converted THEN c.id END)::NUMERIC / + NULLIF(COUNT(DISTINCT c.id), 0) * 100, 2) AS conversion_rate, + COALESCE(SUM(o.paid_amount), 0) AS total_revenue +FROM conversations c +LEFT JOIN orders o ON c.id = o.conversation_id AND o.status = 'PAID' +WHERE c.category IS NOT NULL +GROUP BY c.category +ORDER BY total_conversations DESC; + +COMMENT ON VIEW v_category_statistics IS '移民类别统计视图 - 按类别汇总对话和转化数据'; + +-- 渠道统计视图 +CREATE OR REPLACE VIEW v_channel_statistics AS +SELECT + u.source_channel, + COUNT(DISTINCT u.id) AS total_users, + COUNT(DISTINCT CASE WHEN u.type = 'REGISTERED' THEN u.id END) AS registered_users, + COUNT(DISTINCT c.id) AS total_conversations, + COALESCE(SUM(o.paid_amount), 0) AS total_revenue, + COUNT(DISTINCT CASE WHEN o.status = 'PAID' THEN o.id END) AS paid_orders +FROM users u +LEFT JOIN conversations c ON u.id = c.user_id +LEFT JOIN orders o ON u.id = o.user_id AND o.status = 'PAID' +GROUP BY u.source_channel +ORDER BY total_users DESC; + +COMMENT ON VIEW v_channel_statistics IS '渠道统计视图 - 按来源渠道汇总用户和收入数据'; + +-- =========================================== +-- 知识文章表 (knowledge_articles) +-- 存储移民相关的知识内容，支持RAG检索 +-- =========================================== +CREATE TABLE knowledge_articles ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 文章标题 + title VARCHAR(500) NOT NULL, + -- 文章内容（纯文本或Markdown） + content TEXT NOT NULL, + -- 内容摘要（用于预览） + summary TEXT, + -- 移民类别: QMAS, GEP, IANG, TTPS, CIES, TechTAS, GENERAL + category VARCHAR(50) NOT NULL, + -- 内容标签（JSON数组） + tags TEXT[] DEFAULT '{}', + -- 来源: MANUAL(手动), CRAWL(爬取), EXTRACT(对话提取), IMPORT(批量导入) + source VARCHAR(20) NOT NULL DEFAULT 'MANUAL' + CHECK (source IN ('MANUAL', 'CRAWL', 'EXTRACT', 'IMPORT')), + -- 来源URL + source_url VARCHAR(1000), + -- 内容向量（用于语义搜索） + embedding VECTOR(1536), + -- 是否已发布 + is_published BOOLEAN DEFAULT FALSE, + -- 引用次数（被对话引用） + citation_count INT DEFAULT 0, + -- 点赞数 + helpful_count INT DEFAULT 0, + -- 点踩数 + unhelpful_count INT DEFAULT 0, + -- 质量评分 0-100 + quality_score INT DEFAULT 50 CHECK (quality_score >= 0 AND quality_score <= 100), + -- 创建者ID + created_by UUID, + -- 更新者ID + updated_by UUID, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE knowledge_articles IS '知识文章表 - 存储移民相关知识，支持RAG语义检索'; +COMMENT ON COLUMN knowledge_articles.category IS '移民类别，用于分类筛选'; +COMMENT ON COLUMN knowledge_articles.source IS '内容来源：MANUAL手动添加,CRAWL网页爬取,EXTRACT对话提取,IMPORT批量导入'; +COMMENT ON COLUMN knowledge_articles.embedding IS '文章向量，1536维，用于pgvector语义搜索'; +COMMENT ON COLUMN knowledge_articles.quality_score IS '质量评分，根据引用和反馈自动计算'; + +CREATE INDEX idx_knowledge_articles_category ON knowledge_articles(category); +CREATE INDEX idx_knowledge_articles_published ON knowledge_articles(is_published); +CREATE INDEX idx_knowledge_articles_quality ON knowledge_articles(quality_score DESC); +CREATE INDEX idx_knowledge_articles_embedding ON knowledge_articles USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100); + +-- =========================================== +-- 知识块表 (knowledge_chunks) +-- 将文章拆分为更小的检索单元，提高RAG精确度 +-- =========================================== +CREATE TABLE knowledge_chunks ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 所属文章ID + article_id UUID NOT NULL REFERENCES knowledge_articles(id) ON DELETE CASCADE, + -- 块内容 + content TEXT NOT NULL, + -- 块序号（在文章中的位置） + chunk_index INT NOT NULL, + -- 块类型: TITLE(标题), PARAGRAPH(段落), LIST(列表), TABLE(表格), CODE(代码), FAQ(问答) + chunk_type VARCHAR(20) NOT NULL DEFAULT 'PARAGRAPH' + CHECK (chunk_type IN ('TITLE', 'PARAGRAPH', 'LIST', 'TABLE', 'CODE', 'FAQ')), + -- 内容向量 + embedding VECTOR(1536), + -- 元数据（如章节标题、前后块链接等） + metadata JSONB DEFAULT '{}', + -- Token数量（估算） + token_count INT DEFAULT 0, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE knowledge_chunks IS '知识块表 - 文章分块，用于精确RAG检索'; +COMMENT ON COLUMN knowledge_chunks.chunk_type IS '块类型，用于理解内容结构'; +COMMENT ON COLUMN knowledge_chunks.metadata IS '元数据，包含章节标题、前后块链接等'; + +CREATE INDEX idx_knowledge_chunks_article ON knowledge_chunks(article_id); +CREATE INDEX idx_knowledge_chunks_embedding ON knowledge_chunks USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100); + +-- =========================================== +-- 用户记忆表 (user_memories) +-- 存储用户的长期记忆，用于个性化对话 +-- =========================================== +CREATE TABLE user_memories ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 用户ID + user_id UUID NOT NULL REFERENCES users(id) ON DELETE CASCADE, + -- 记忆类型 + memory_type VARCHAR(30) NOT NULL + CHECK (memory_type IN ( + 'PERSONAL_INFO', -- 个人信息 + 'WORK_EXPERIENCE', -- 工作经历 + 'EDUCATION', -- 教育背景 + 'LANGUAGE', -- 语言能力 + 'IMMIGRATION_INTENT', -- 移民意向 + 'PREFERRED_CATEGORY', -- 倾向类别 + 'ASSESSMENT_RESULT', -- 评估结果 + 'QUESTION_ASKED', -- 问过的问题 + 'CONCERN', -- 关注点 + 'PREFERENCE', -- 偏好设置 + 'CUSTOM' -- 自定义 + )), + -- 记忆内容 + content TEXT NOT NULL, + -- 重要性 0-100 + importance INT DEFAULT 50 CHECK (importance >= 0 AND importance <= 100), + -- 来源对话ID + source_conversation_id UUID, + -- 相关移民类别 + related_category VARCHAR(50), + -- 内容向量 + embedding VECTOR(1536), + -- 访问次数 + access_count INT DEFAULT 0, + -- 最后访问时间 + last_accessed_at TIMESTAMP WITH TIME ZONE, + -- 是否已过期 + is_expired BOOLEAN DEFAULT FALSE, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE user_memories IS '用户记忆表 - 存储用户长期记忆，支持个性化对话'; +COMMENT ON COLUMN user_memories.memory_type IS '记忆类型，用于分类管理用户信息'; +COMMENT ON COLUMN user_memories.importance IS '重要性评分，影响检索优先级'; +COMMENT ON COLUMN user_memories.is_expired IS '是否过期，用户情况变化时标记'; + +CREATE INDEX idx_user_memories_user ON user_memories(user_id); +CREATE INDEX idx_user_memories_type ON user_memories(memory_type); +CREATE INDEX idx_user_memories_importance ON user_memories(user_id, importance DESC); +CREATE INDEX idx_user_memories_embedding ON user_memories USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100); + +-- =========================================== +-- 系统经验表 (system_experiences) +-- 存储系统从对话中学习到的经验，用于自我进化 +-- =========================================== +CREATE TABLE system_experiences ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 经验类型 + experience_type VARCHAR(30) NOT NULL + CHECK (experience_type IN ( + 'COMMON_QUESTION', -- 常见问题 + 'ANSWER_TEMPLATE', -- 回答模板 + 'CLARIFICATION', -- 澄清方式 + 'USER_PATTERN', -- 用户行为模式 + 'CONVERSION_TRIGGER', -- 转化触发点 + 'KNOWLEDGE_GAP', -- 知识缺口 + 'KNOWLEDGE_UPDATE', -- 知识更新 + 'CONVERSATION_SKILL', -- 对话技巧 + 'OBJECTION_HANDLING', -- 异议处理 + 'CUSTOM' -- 自定义 + )), + -- 经验内容 + content TEXT NOT NULL, + -- 置信度 0-100 + confidence INT DEFAULT 50 CHECK (confidence >= 0 AND confidence <= 100), + -- 应用场景描述 + scenario TEXT NOT NULL, + -- 相关移民类别 + related_category VARCHAR(50), + -- 来源对话ID列表 + source_conversation_ids UUID[] DEFAULT '{}', + -- 验证状态: PENDING(待验证), APPROVED(已通过), REJECTED(已拒绝), DEPRECATED(已弃用) + verification_status VARCHAR(20) NOT NULL DEFAULT 'PENDING' + CHECK (verification_status IN ('PENDING', 'APPROVED', 'REJECTED', 'DEPRECATED')), + -- 验证者ID + verified_by UUID, + -- 验证时间 + verified_at TIMESTAMP WITH TIME ZONE, + -- 使用次数 + usage_count INT DEFAULT 0, + -- 正面反馈次数 + positive_count INT DEFAULT 0, + -- 负面反馈次数 + negative_count INT DEFAULT 0, + -- 内容向量 + embedding VECTOR(1536), + -- 是否激活使用 + is_active BOOLEAN DEFAULT FALSE, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE system_experiences IS '系统经验表 - 从对话中学习的经验，支持系统自我进化'; +COMMENT ON COLUMN system_experiences.experience_type IS '经验类型，用于分类应用'; +COMMENT ON COLUMN system_experiences.confidence IS '置信度，根据来源数量和反馈计算'; +COMMENT ON COLUMN system_experiences.verification_status IS '验证状态，需管理员审核后才能激活'; +COMMENT ON COLUMN system_experiences.is_active IS '是否激活，只有通过审核才能在对话中使用'; + +CREATE INDEX idx_system_experiences_type ON system_experiences(experience_type); +CREATE INDEX idx_system_experiences_status ON system_experiences(verification_status); +CREATE INDEX idx_system_experiences_active ON system_experiences(is_active); +CREATE INDEX idx_system_experiences_confidence ON system_experiences(confidence DESC); +CREATE INDEX idx_system_experiences_embedding ON system_experiences USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100); + +-- =========================================== +-- 管理员表 (admins) +-- 管理后台用户，支持多角色权限 +-- =========================================== +CREATE TABLE admins ( + id UUID PRIMARY KEY DEFAULT uuid_generate_v4(), + -- 用户名（登录用） + username VARCHAR(50) NOT NULL UNIQUE, + -- 密码哈希 + password_hash VARCHAR(255) NOT NULL, + -- 姓名 + name VARCHAR(100) NOT NULL, + -- 邮箱 + email VARCHAR(255), + -- 手机号 + phone VARCHAR(20), + -- 角色: SUPER_ADMIN(超级管理员), ADMIN(管理员), OPERATOR(运营), VIEWER(只读) + role VARCHAR(20) NOT NULL DEFAULT 'OPERATOR' + CHECK (role IN ('SUPER_ADMIN', 'ADMIN', 'OPERATOR', 'VIEWER')), + -- 权限列表（细粒度权限控制） + permissions JSONB DEFAULT '[]', + -- 头像URL + avatar VARCHAR(500), + -- 最后登录时间 + last_login_at TIMESTAMP WITH TIME ZONE, + -- 最后登录IP + last_login_ip VARCHAR(50), + -- 是否启用 + is_active BOOLEAN DEFAULT TRUE, + -- 创建时间 + created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(), + -- 更新时间 + updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW() +); + +COMMENT ON TABLE admins IS '管理员表 - 管理后台用户，支持多角色权限'; +COMMENT ON COLUMN admins.role IS '角色：SUPER_ADMIN超管,ADMIN管理员,OPERATOR运营,VIEWER只读'; +COMMENT ON COLUMN admins.permissions IS '细粒度权限列表，JSON数组格式'; + +CREATE INDEX idx_admins_username ON admins(username); +CREATE INDEX idx_admins_role ON admins(role); +CREATE INDEX idx_admins_active ON admins(is_active); + +-- 插入默认超级管理员（密码: admin123，实际生产环境需要修改） +INSERT INTO admins (username, password_hash, name, role, permissions) VALUES +('admin', '$2b$10$rQNDjKwYXOw8FNrFcD3e0.T8KCqVJLqDQT9gQR2KPnDqPvqK8VpKi', '系统管理员', 'SUPER_ADMIN', '["*"]'); + +-- =========================================== +-- 结束 +-- ===========================================