|
| 1 | +# CLAUDE.md |
| 2 | + |
| 3 | +This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository. |
| 4 | + |
| 5 | +## Project Overview |
| 6 | + |
| 7 | +This is the **@moleculer/database** package - an advanced database access service for the Moleculer microservices framework. It provides a mixin that adds CRUD operations, field validation, data transformation, populating, scoping, multi-tenancy, and other database features to Moleculer services. |
| 8 | + |
| 9 | +## Key Architecture Components |
| 10 | + |
| 11 | +### Core Structure |
| 12 | +- **Mixin Pattern**: The service is implemented as a Moleculer mixin that other services can use |
| 13 | +- **Multi-Adapter Support**: Supports NeDB, MongoDB, and Knex (SQL databases) adapters |
| 14 | +- **Field-Based Schema**: Uses a comprehensive field definition system similar to Fastest Validator |
| 15 | +- **Transformation Pipeline**: Data flows through validation → transformation → storage → retrieval → transformation |
| 16 | + |
| 17 | +### Main Files |
| 18 | +- `src/index.js` - Main mixin factory function with lifecycle hooks |
| 19 | +- `src/actions.js` - Auto-generated CRUD actions (find, list, get, create, update, etc.) |
| 20 | +- `src/methods.js` - Core database methods used by actions and custom implementations |
| 21 | +- `src/validation.js` - Field validation and sanitization logic |
| 22 | +- `src/transform.js` - Data transformation (encoding/decoding, populating, field filtering) |
| 23 | +- `src/adapters/` - Database adapter implementations (base, mongodb, knex, nedb) |
| 24 | +- `src/schema.js` - Field schema processing and validation schema generation |
| 25 | + |
| 26 | +### Adapter Pattern |
| 27 | +- `src/adapters/base.js` - Abstract base adapter |
| 28 | +- `src/adapters/mongodb.js` - MongoDB adapter |
| 29 | +- `src/adapters/knex.js` - SQL databases via Knex.js |
| 30 | +- `src/adapters/nedb.js` - NeDB (in-memory/file-based) adapter |
| 31 | + |
| 32 | +## Development Commands |
| 33 | + |
| 34 | +### Testing |
| 35 | +- `npm test` - Run all tests (unit, integration, leak detection) |
| 36 | +- `npm run test:unit` - Run unit tests only |
| 37 | +- `npm run test:integration` - Run integration tests with coverage |
| 38 | +- `npm run test:leak` - Run memory leak detection tests |
| 39 | +- `npm run ci:unit` - Watch mode for unit tests |
| 40 | +- `npm run ci:integration` - Watch mode for integration tests |
| 41 | +- `npm run ci:leak` - Watch mode for leak detection tests |
| 42 | + |
| 43 | +### Development |
| 44 | +- `npm run dev` - Start example service in development mode |
| 45 | +- `npm run lint` - Run ESLint on source code, examples, and tests |
| 46 | +- `npm run bench` - Run benchmarks |
| 47 | +- `npm run bench:watch` - Run benchmarks in watch mode |
| 48 | + |
| 49 | +### Maintenance |
| 50 | +- `npm run deps` - Check and update dependencies interactively |
| 51 | +- `npm run ci-update-deps` - Auto-update minor version dependencies |
| 52 | + |
| 53 | +## Database Adapters |
| 54 | + |
| 55 | +When working with database functionality, understand that: |
| 56 | +- **NeDB** is used for development/testing (default if no adapter specified) |
| 57 | +- **MongoDB** supports full document operations and nested field querying |
| 58 | +- **Knex** supports SQL databases (MySQL, PostgreSQL, SQLite, etc.) but converts objects/arrays to JSON strings |
| 59 | + |
| 60 | +## Field Definitions |
| 61 | + |
| 62 | +The service uses a comprehensive field definition system. Each field can have: |
| 63 | +- Basic validation properties (type, required, min, max, etc.) |
| 64 | +- Database properties (columnName, columnType, primaryKey) |
| 65 | +- Permission properties (readPermission, permission) |
| 66 | +- Lifecycle hooks (onCreate, onUpdate, onRemove) |
| 67 | +- Transformation functions (get, set, validate) |
| 68 | +- Populate configurations for relationships |
| 69 | + |
| 70 | +## Multi-Tenancy Support |
| 71 | + |
| 72 | +The service supports three multi-tenancy modes: |
| 73 | +1. **Record-based**: Same table with tenant ID field + scopes |
| 74 | +2. **Table-based**: Different tables per tenant |
| 75 | +3. **Database-based**: Different databases per tenant |
| 76 | + |
| 77 | +Implement via the `getAdapterByContext` method to customize adapter creation per context. |
| 78 | + |
| 79 | +## Testing Integration Services |
| 80 | + |
| 81 | +When testing services that use this mixin: |
| 82 | +- Use NeDB adapter for simple unit tests (no external dependencies) |
| 83 | +- Use real databases for integration tests (see `test/docker-compose.yml`) |
| 84 | +- Test both single operations and batch operations |
| 85 | +- Test scoping, populating, and multi-tenancy if used |
| 86 | +- Check soft delete behavior if implemented |
| 87 | + |
| 88 | +## Common Patterns |
| 89 | + |
| 90 | +### Service Implementation |
| 91 | +```javascript |
| 92 | +const DbService = require("@moleculer/database").Service; |
| 93 | + |
| 94 | +module.exports = { |
| 95 | + name: "posts", |
| 96 | + mixins: [DbService({ adapter: "MongoDB" })], |
| 97 | + settings: { |
| 98 | + fields: { |
| 99 | + id: { type: "string", primaryKey: true, columnName: "_id" }, |
| 100 | + title: { type: "string", required: true }, |
| 101 | + // ... more fields |
| 102 | + } |
| 103 | + } |
| 104 | +} |
| 105 | +``` |
| 106 | + |
| 107 | +### Custom Methods |
| 108 | +Services often implement custom methods that use the built-in methods like `findEntities`, `createEntity`, `updateEntity`, etc. |
| 109 | + |
| 110 | +### Adapter Connection |
| 111 | +Adapters connect lazily on first use, supporting multi-tenancy scenarios where connection details depend on context. |
0 commit comments