Domovská stránka Prehľadávať Pomoc
Prihlásiť sa
aitori-group
/
aitori-stylist
3
0
Fork 0
Súbory Issues 0 Pull requesty 0 Wiki
Strom: e3e12bd8b6
Branche Tagy
aitori-stylist
/
docs
/
superpowers
/
plans
/
2026-06-28-prisma-v1-baseline.md

2026-06-28-prisma-v1-baseline.md 9.2 KB
História Raw

Prisma V1 Baseline Migration Implementation Plan

For agentic workers: REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (- [ ]) syntax for tracking.

Goal: 将当前稳定版数据库结构收敛为单一的 v1 baseline migration，删除旧增量 migration，并让 Prisma 后续从该基线继续演进。

Architecture: 保持 backend/prisma/schema.prisma 不变，把 backend/prisma/migrations 重置为仅包含一条 20260628_init_v1/migration.sql 的完整建表基线；现有数据库不重建、不重新执行 baseline SQL，而是通过 Prisma 的已应用状态对齐这条基线。实现过程中同步校验 baseline SQL 与 schema.prisma、数据库设计.md 一致，并移除旧 migration 目录。

Tech Stack: Prisma, MySQL, SQL, Node.js

Global Constraints

  • schema.prisma 继续作为当前稳定数据库结构的唯一事实来源。
  • backend/prisma/migrations 目录只保留一条 v1 baseline migration。
  • baseline SQL 必须包含当前稳定结构的完整建表、索引与注释。
  • 旧 migration 全部移除，不保留历史增量。
  • 不修改当前稳定数据库结构语义，只做基线重置。
  • 现有数据库不重新执行 baseline SQL，采用 Prisma 已应用状态对齐。
  • 后续数据库改动必须在 v1 baseline 之上继续新增 migration。

Task 1: 生成并校验 V1 Baseline SQL

Files:

  • Create: backend/prisma/migrations/20260628_init_v1/migration.sql
  • Read: backend/prisma/schema.prisma
  • Read: backend/docs/数据库设计.md

Interfaces:

  • Consumes: backend/prisma/schema.prisma
  • Consumes: backend/docs/数据库设计.md

  • Produces: backend/prisma/migrations/20260628_init_v1/migration.sql

  • [ ] Step 1: 基于当前 schema 与数据库设计文档列出 baseline 必须覆盖的表

    hairstylist_user
customer
customer_treatment_record
ai_hair_simulation
ai_hair_simulation_task
hair_template
hair_template_make_task
system_config

  • [ ] Step 2: 先写 baseline SQL 文件，包含完整建表与索引定义

    CREATE TABLE `hairstylist_user` (
`id` BIGINT UNSIGNED NOT NULL AUTO_INCREMENT COMMENT '发型师主键ID',
`username` VARCHAR(50) NOT NULL COMMENT '发型师用户名',
`email` VARCHAR(100) NOT NULL COMMENT '登录邮箱',
`password` VARCHAR(100) NOT NULL COMMENT '加密后的登录密码',
`create_time` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
`update_time` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
`is_deleted` TINYINT UNSIGNED NOT NULL DEFAULT 0 COMMENT '逻辑删除：0未删除，1已删除',
PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='发型师用户信息表';

CREATE TABLE `hair_template` (
`id` BIGINT UNSIGNED NOT NULL AUTO_INCREMENT COMMENT '模板主键ID',
`template_name` VARCHAR(100) NOT NULL COMMENT '模板名称（必填）',
`cover_url` VARCHAR(500) DEFAULT '' COMMENT '模板封面图URL',
`status` ENUM('DRAFT','PUBLIC') NOT NULL DEFAULT 'DRAFT' COMMENT '模板状态：DRAFT草稿 / PUBLIC公开',
`category` ENUM('LONG','MEDIUM','SHORT','DYE') NOT NULL DEFAULT 'LONG' COMMENT '分类：长发/中长发/短发/染发',
`tags` VARCHAR(500) DEFAULT '' COMMENT '模板标签，逗号分隔存储',
`description` TEXT NULL COMMENT '模板详细说明',
`model_url` VARCHAR(500) DEFAULT '' COMMENT 'AI模特图片URL',
`clothes_url` VARCHAR(500) DEFAULT '' COMMENT '服装素材URL',
`background_url` VARCHAR(500) DEFAULT '' COMMENT '背景图片URL',
`ref_hair_url` VARCHAR(500) DEFAULT '' COMMENT '参考发型图片URL',
`hairstylist_id` BIGINT UNSIGNED NULL COMMENT '创建发型师ID，系统模板为空',
`template_url` VARCHAR(500) DEFAULT '' COMMENT '完整模板预览页面URL',
`used_count` INT UNSIGNED NOT NULL DEFAULT 0 COMMENT '模板累计使用次数',
`is_system_template` ENUM('YES','NO') NOT NULL DEFAULT 'NO' COMMENT '是否系统模板：YES系统模板，NO门店自定义模板',
`create_time` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
`update_time` DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
`is_deleted` TINYINT UNSIGNED NOT NULL DEFAULT 0 COMMENT '逻辑删除：0正常 1删除',
PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='AI发型模板表';

  • [ ] Step 3: 补齐剩余 6 张表、索引和全部注释

    CREATE INDEX `idx_category` ON `hair_template` (`category`, `is_deleted`);
CREATE INDEX `idx_hairstylist_id` ON `hair_template` (`hairstylist_id`, `is_deleted`);
CREATE INDEX `idx_status` ON `hair_template` (`status`, `is_deleted`);

  • [ ] Step 4: 自查 baseline SQL 是否包含 used_count、索引和注释

Run: grep -n "used_count\\|COMMENT\\|CREATE INDEX" backend/prisma/migrations/20260628_init_v1/migration.sql | head -n 20

Expected: 输出包含 used_count、COMMENT、CREATE INDEX

  • [ ] Step 5: 提交 baseline SQL 初稿

    git add backend/prisma/migrations/20260628_init_v1/migration.sql
git commit -m "feat: add prisma v1 baseline migration"

Task 2: 清理旧 migration 并校验目录基线

Files:

  • Delete: backend/prisma/migrations/20260628050000_add_customer_follow_content/migration.sql
  • Delete: backend/prisma/migrations/20260628120000_add_customer_table/migration.sql
  • Delete: backend/prisma/migrations/20260628173000_add_template_used_count/migration.sql
  • Modify: backend/prisma/migrations/

Interfaces:

  • Consumes: backend/prisma/migrations/20260628_init_v1/migration.sql

  • Produces: backend/prisma/migrations/ 目录中仅保留 20260628_init_v1

  • [ ] Step 1: 删除旧 migration 目录

    backend/prisma/migrations/20260628050000_add_customer_follow_content
backend/prisma/migrations/20260628120000_add_customer_table
backend/prisma/migrations/20260628173000_add_template_used_count

  • [ ] Step 2: 列出 migration 目录，确认只剩 baseline

Run: find backend/prisma/migrations -maxdepth 2 -type f | sort

Expected:

backend/prisma/migrations/20260628_init_v1/migration.sql
  • Step 3: 再次校验 baseline 文件存在且旧目录已移除

Run: git status --short backend/prisma/migrations

Expected: 只出现新 baseline 与旧目录删除记录

  • [ ] Step 4: 提交 migration 目录重置

    git add backend/prisma/migrations
git commit -m "chore: reset prisma migrations to v1 baseline"

Task 3: 对齐现有数据库与 Prisma migration 状态

Files:

  • Modify: backend/prisma/migration_lock.toml (if generated by Prisma tooling)
  • Read: current database state via Prisma CLI

Interfaces:

  • Consumes: backend/prisma/migrations/20260628_init_v1/migration.sql

  • Produces: Prisma recognizes 20260628_init_v1 as applied for the current database

  • [ ] Step 1: 先确认当前数据库不需要重新执行 baseline SQL

Run: npx prisma db pull --print

Expected: 输出的结构与 backend/prisma/schema.prisma 主要一致，不出现缺失 used_count 等明显差异

  • Step 2: 将 baseline migration 标记为已应用

Run: npx prisma migrate resolve --applied 20260628_init_v1

Expected: 输出包含 marked as applied 或等价成功提示

  • Step 3: 查看 Prisma migration 状态

Run: npx prisma migrate status

Expected: 输出包含 20260628_init_v1，且当前数据库没有 pending migration

  • [ ] Step 4: 如果 Prisma 生成了额外锁文件或状态文件，确认其已纳入版本控制策略

    backend/prisma/migration_lock.toml

  • [ ] Step 5: 提交 migration 状态对齐改动

    git add backend/prisma
git commit -m "chore: align prisma baseline migration state"

Task 4: 最终一致性校验与文档收口

Files:

  • Read: backend/prisma/schema.prisma
  • Read: backend/prisma/migrations/20260628_init_v1/migration.sql
  • Read: backend/docs/数据库设计.md

Interfaces:

  • Consumes: schema.prisma, 数据库设计.md, 20260628_init_v1/migration.sql

  • Produces: verified v1 baseline layout ready for future migrations

  • [ ] Step 1: 对照 schema.prisma 与 baseline SQL 逐项检查核心表和字段

    重点校验：
- customer.follow_content
- ai_hair_simulation.template_ids
- hair_template.used_count
- system_config.value_type

  • [ ] Step 2: 运行 Prisma 客户端生成，确认 schema 与 migration 目录在当前仓库状态下可工作

Run: npx prisma generate

Expected: PASS，输出包含 Generated Prisma Client

  • Step 3: 运行 migration 状态检查，确认没有残留旧目录引用

Run: npx prisma migrate status

Expected: 不再出现旧 migration 名称，只识别 20260628_init_v1

  • Step 4: 查看最终工作区状态，只确认本次相关文件

Run: git status --short backend/prisma backend/docs/数据库设计.md

Expected: 仅包含本次 baseline 重置相关改动，或已 clean

  • [ ] Step 5: 提交最终收口

    git add backend/prisma backend/docs/数据库设计.md
git commit -m "chore: finalize prisma v1 baseline"