@rawan-temo/prisma-zod-generator 3.3.1-config-coercion.0

package/CLAUDE.md

@@ -0,0 +1,121 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Overview
+
+This is a Prisma Generator that automatically generates Zod schemas from Prisma schema files. It reads Prisma schema definitions and transforms them into TypeScript Zod validation schemas for API endpoint validation.
+
+## Key Commands
+
+**IMPORTANT**: Always use `pnpm`, never use `npm` or `npx`.
+
+## Windows + WSL Policy
+
+- On Windows hosts, run repository commands through WSL by default.
+- Prefer: `wsl.exe bash -lc "<command>"`.
+- Only use PowerShell/CMD when explicitly requested or when a task is Windows-only.
+
+### Build and Generate
+- `pnpm exec tsc` - Compile TypeScript to JavaScript in the `lib/` directory
+- `pnpm prisma generate` - Generate Zod schemas from Prisma schema
+- `pnpm run gen-example` - Build and generate schemas in one command
+
+### Development
+- `pnpm run check-uncommitted` - Check for uncommitted changes before publishing
+- `pnpm run package:publish` - Full publish process (update, check, package, publish)
+
+### Coverage
+
+`pnpm run test:coverage` locally, `test:coverage:ci` in CI. Both run under c8; there is no
+in-process coverage script left, and that is deliberate.
+
+The reason is worth knowing before trusting any coverage figure here. Most of the suite exercises
+the generator by spawning `prisma generate`, which runs `node ./lib/generator.js` in a **child
+process**, and in-process V8 coverage cannot see that at all. Measured with plain Vitest the
+engine — transformer, generators/, prisma-generator — reads close to zero while being the
+best-covered code in the repo, and the whole of `src/` looks like ~20% against a real ~89%. Nine
+scripts used to report that number, including the `test:coverage` the contributing guide points
+at; they are gone or repointed.
+
+Both scripts pass `--exclude-after-remap`, which is the part that makes it work: without it c8
+matches its filters against the pre-map `lib/` paths, drops every child's data, and `--all`
+backfills the src files at 0% — a result that looks real and is not.
+
+`test:coverage:ci` differs only in its thresholds, calibrated for a checkout without the private
+`src/pro` submodule, where 18 files skip. Both sets of thresholds sit below the figure measured in
+*both* conditions, so either script passes either way.
+
+The report needs a raised heap (the scripts set it): it merges well over a thousand raw V8
+coverage files and OOMs on the default.
+
+## Architecture
+
+### Core Components
+
+**Generator Entry Point**: `src/generator.ts` - Simple shebang script that imports the main logic
+
+**Main Logic**: `src/index.ts` - Entry point that orchestrates the generation process
+
+**Transformer**: `src/transformer.ts` - Core class that transforms Prisma DMMF into Zod schemas
+- Handles model operations (CRUD), enums, object schemas
+- Manages imports and schema relationships
+- Supports MongoDB, PostgreSQL, MySQL providers
+- Handles Select/Include generation when enabled
+
+**Helper Functions**: `src/helpers/` directory contains specialized helpers:
+- `model-helpers.ts` - Model-specific transformations
+- `aggregate-helpers.ts` - Aggregate operation support
+- `include-helpers.ts` - Include schema generation
+- `select-helpers.ts` - Select schema generation
+- `mongodb-helpers.ts` - MongoDB-specific logic
+
+**Utilities**: `src/utils/` directory:
+- `writeFileSafely.ts` - Safe file writing with directory creation
+- `formatFile.ts` - Code formatting using Prettier
+- `writeIndexFile.ts` - Index file generation
+- `naming-resolver.ts` - Naming patterns for models, enums and exports
+
+### Generated Output Structure
+
+The generator creates schemas in the output directory (default: `./generated/schemas/`):
+- `enums/` - Enum validation schemas
+- `objects/` - Object validation schemas for input types
+- Root level - Operation schemas (findMany, create, update, etc.)
+- `index.ts` - Barrel export file
+
+### Key Features
+
+- **Provider Support**: Handles PostgreSQL, MySQL, MongoDB, SQL Server
+- **Aggregate Operations**: Supports count, min, max, avg, sum aggregations
+- **Relation Handling**: Manages complex model relationships
+- **Select/Include**: Optional generation of select and include schemas
+- **Type Safety**: Full TypeScript support with Prisma type imports
+- **Custom Output**: Configurable output directory and Prisma client path
+
+## Development Notes
+
+- The project uses `private: true` in package.json - this is a generator, not a published package
+- TypeScript compilation outputs to `lib/` directory
+- The `package/` directory contains the packaged version for distribution
+- Generator options can be configured in the Prisma schema generator block
+- Test schemas are generated in `prisma/generated/schemas/` for development/testing
+
+## Contribution Workflow
+
+- Package manager: always use `pnpm`.
+- Formatting & linting: run `pnpm format` then `pnpm lint` before pushing; commit any resulting changes.
+- Conventional Commits with scope: `type(scope): subject`.
+  - Choose `type` based on impact: `feat`, `fix`, `docs`, `refactor`, `chore`, `style`, etc.
+  - Example: `fix(variants): include .js in ESM index imports`.
+- Commit in small, logical units and add/update tests for behavior changes.
+- Pull requests: use the preinstalled GitHub CLI with a Markdown body file.
+  - Create a new body file (e.g., `.github/PR_BODY.md`) containing: Summary, Problem, Fix, Validation, and linked issues (e.g., `Fixes #234`).
+  - Open the PR: `gh pr create --base master --head <branch> --title "fix(scope): concise title" --body-file .github/PR_BODY.md`.
+  - Prefer keeping PR body files untracked unless explicitly requested.
+- Ensure `pnpm build`, `pnpm test`, `pnpm format:check`, and `pnpm lint` succeed prior to push/PR.
+
+## Local Overrides
+
+- Use `CLAUDE.local.md` for private, repository-local Claude instructions.
+- Keep `CLAUDE.local.md` untracked.

package/CONTRIBUTING.md

@@ -0,0 +1,279 @@
+# Contributing to Prisma Zod Generator
+
+Thank you for your interest in contributing to PZG! This guide will help you get started.
+
+## Code of Conduct
+
+Be respectful, collaborative, and constructive. We're all here to build great tools together.
+
+## Repository Structure
+
+This project uses a **hybrid architecture** with public core and private pro features:
+
+```
+prisma-zod-generator/ (PUBLIC)
+├── src/
+│   ├── index.ts              # Core generator (public, MIT)
+│   ├── transformer.ts        # Core logic (public, MIT)
+│   ├── license.ts            # License validation (public)
+│   └── pro/                  # Pro features (PRIVATE SUBMODULE)
+├── tests/                    # Tests (public)
+├── docs/                     # Documentation (public)
+└── README.md                 # Main docs (public)
+```
+
+**Pro Features Repository** (PRIVATE):
+- `prisma-zod-generator-pro` - Private submodule at `src/pro/`
+- Access restricted to core team members
+- Commercial license
+
+## For Core Contributors (Public Features)
+
+### Setup
+
+```bash
+# Clone the repository
+git clone https://github.com/omar-dulaimi/prisma-zod-generator.git
+cd prisma-zod-generator
+
+# Install dependencies
+pnpm install
+
+# Build
+pnpm build
+```
+
+**Note:** You won't have access to the `src/pro/` submodule. That's expected! You can contribute to core features without it.
+
+### Development Workflow
+
+1. **Create a branch**
+   ```bash
+   git checkout -b feat/your-feature
+   ```
+
+2. **Make changes** to core files (anything except `src/pro/`)
+
+3. **Test your changes**
+   ```bash
+   pnpm typecheck
+   pnpm lint
+   pnpm test
+   ```
+
+4. **Commit with conventional commits**
+   ```bash
+   git add .
+   git commit -m "feat(core): add new Zod helper"
+   ```
+
+5. **Push and create PR**
+   ```bash
+   git push origin feat/your-feature
+   gh pr create --base master --head feat/your-feature
+   ```
+
+### Areas to Contribute
+
+- 🐛 **Bug fixes** - Fix issues in core generator
+- ✨ **Core features** - New schema types, Zod helpers
+- 📚 **Documentation** - Improve docs, add examples
+- 🧪 **Tests** - Add test coverage
+- 🎨 **Examples** - Create usage examples
+
+## For Pro Developers (Team Members)
+
+### Setup with Pro Submodule
+
+```bash
+# Clone the repository
+git clone https://github.com/omar-dulaimi/prisma-zod-generator.git
+cd prisma-zod-generator
+
+# Initialize submodules (requires private repo access)
+pnpm setup
+# Or manually:
+# git submodule update --init --recursive
+
+# Install dependencies
+pnpm install
+
+# Build (includes pro features)
+pnpm build
+```
+
+### Pro Development Workflow
+
+1. **Update pro submodule** (get latest changes)
+   ```bash
+   pnpm sync:pro
+   ```
+
+2. **Create feature branch in pro repo**
+   ```bash
+   cd src/pro
+   git checkout -b feat/new-pro-feature
+   cd ../..
+   ```
+
+3. **Make changes** to pro features
+
+4. **Test changes**
+   ```bash
+   pnpm typecheck
+   pnpm lint
+   pnpm test
+   ```
+
+5. **Commit to pro repo**
+   ```bash
+   cd src/pro
+   git add .
+   git commit -m "feat(policies): add new policy type"
+   git push origin feat/new-pro-feature
+   cd ../..
+   ```
+
+6. **Create PR in private repo**
+   - Go to `https://github.com/omar-dulaimi/prisma-zod-generator-pro`
+   - Create PR for your branch
+   - Get review and merge
+
+7. **Update main repo with new submodule commit**
+   ```bash
+   git add src/pro
+   git commit -m "chore(pro): update pro submodule to latest"
+   git push
+   ```
+
+### Pro Feature Guidelines
+
+- All pro features must have license checks
+- Add comprehensive tests
+- Update pro documentation
+- Consider backward compatibility
+- Performance matters (pro features run in prod)
+
+## Commit Conventions
+
+We use **Conventional Commits** format:
+
+```
+<type>(<scope>): <subject>
+
+<body>
+
+<footer>
+```
+
+> **Do not write the CI-skip marker in a commit body.** GitHub matches it anywhere in
+> the head commit's message, so it disables every workflow for that commit — and the
+> body carries into the squash-merge commit, so the merge skips CI too. This has
+> already cost us once: a pull request whose body quoted the marker while explaining
+> it landed on `master` with zero checks and zero workflow runs, and nothing reported
+> a problem, because from GitHub's side nothing had gone wrong.
+>
+> Mention it without the brackets when describing it. A `commit-msg` hook rejects the
+> bracketed form in a body; the subject line, where `semantic-release` writes it,
+> still works.
+
+### Types
+- `feat`: New feature
+- `fix`: Bug fix
+- `docs`: Documentation changes
+- `refactor`: Code refactoring
+- `test`: Adding tests
+- `chore`: Maintenance tasks
+- `perf`: Performance improvements
+
+### Scopes
+- `core`: Core generator features
+- `pro`: Pro features (team members only)
+- `policies`: Policy feature
+- `server-actions`: Server Actions feature
+- `drift-guard`: Drift Guard feature
+- etc.
+
+### Examples
+```bash
+feat(core): add support for BigInt fields
+fix(policies): resolve PII redaction edge case
+docs(readme): update installation instructions
+chore(deps): update Prisma to v6.18.0
+```
+
+## Testing Requirements
+
+All PRs must include tests:
+
+```bash
+# Run EVERY test file (the gate that must pass before opening a PR)
+pnpm test
+
+# Run one file, or one test by name — pass vitest arguments straight through
+pnpm test tests/config.test.ts
+pnpm test -t 'should respect field exclusions'
+
+# Same suite on a single worker, for a failure that only shows up in parallel
+pnpm test:sequential
+
+# Run with coverage
+pnpm test:coverage
+
+# Type check
+pnpm typecheck
+
+# Lint
+pnpm lint
+```
+
+`pnpm test` runs every file matched by the vitest `include` glob. There is
+deliberately no curated-subset script: a hand-maintained file list drifts, and a
+new test file silently stops being run. Filter on the command line instead.
+
+Scratch `test-env-*` directories are swept automatically before and after each
+run. Set `KEEP_TEST_ENVS=1` to keep them for debugging.
+
+## Before Submitting PR
+
+✅ **Checklist:**
+- [ ] Every test file passes (`pnpm test`)
+- [ ] Type check passes (`pnpm typecheck`)
+- [ ] Linting passes (`pnpm lint`)
+- [ ] Formatted code (`pnpm format`)
+- [ ] Conventional commit messages
+- [ ] Updated documentation (if needed)
+- [ ] Added tests for new features
+- [ ] No breaking changes (or clearly documented)
+
+## PR Review Process
+
+1. **Automated checks** run on every PR
+2. **Core team review** (usually within 1-2 days)
+3. **Revisions** if requested
+4. **Merge** to master once approved
+
+## Release Process
+
+Releases are automated using **semantic-release**:
+
+1. Commits to `master` trigger release workflow
+2. Version bump based on commit types
+3. Changelog generated automatically
+4. Package published to npm
+
+## Questions?
+
+- **Discussions:** Use GitHub Discussions for questions
+- **Issues:** Report bugs via GitHub Issues
+- **Support:** Pro members receive priority assistance via the onboarding support channel; all other questions should go through GitHub Issues
+
+## License
+
+By contributing, you agree that your contributions will be licensed under the MIT License for core features.
+
+Pro features are proprietary and require separate contributor agreement.
+
+---
+
+**Thank you for contributing to Prisma Zod Generator!** 🎉

package/DOCKER.md

@@ -0,0 +1,84 @@
+# Docker Setup
+
+This repository includes a Docker Compose setup that brings up one server per
+supported database provider.
+
+> **The test suite does not need any of this.** Nothing under `tests/` opens a
+> database connection — the provider suites only run `prisma generate`, which
+> reads the schema and never connects. `pnpm test` passes on a machine with no
+> Docker installed. These services are here for manually checking generated
+> schemas against a real server.
+
+## Quick Start
+
+1. **Start all databases:**
+   ```bash
+   pnpm docker:up
+   ```
+
+2. **Copy environment variables:**
+   ```bash
+   cp .env.example .env
+   ```
+
+3. **Stop databases:**
+   ```bash
+   pnpm docker:down
+   ```
+
+## Available Services
+
+The Docker Compose setup includes the following services that match the existing `.env.example` configuration:
+
+| Service | Port | Database | Username | Password |
+|---------|------|----------|----------|----------|
+| PostgreSQL | 5432 | test | user | password |
+| MySQL | 3306 | test | user | password |
+| MongoDB | 27017 | test | - | - |
+| SQL Server | 1433 | test | sa | password |
+
+## Docker Commands
+
+- `pnpm docker:up` - Start all database services
+- `pnpm docker:down` - Stop all database services
+- `pnpm docker:logs` - View logs from all services
+- `pnpm docker:ps` - Check status of all services
+- `pnpm docker:reset` - Reset all data and restart services
+
+To start a single service, name it: `docker compose up -d postgresql`.
+
+## Running the tests
+
+Just `pnpm test` — no databases required. To run only the provider suites:
+
+```bash
+pnpm test tests/multi-provider/
+pnpm test tests/multi-provider/ -t 'PostgreSQL Provider Tests'
+```
+
+## Environment Variables
+
+The Docker setup uses the same environment variables as defined in `.env.example`:
+
+- `POSTGRESQL_URL="postgresql://user:password@localhost:5432/test"`
+- `MYSQL_URL="mysql://user:password@localhost:3306/test"`
+- `MONGODB_URL="mongodb://localhost:27017/test"`
+- `SQLSERVER_URL="sqlserver://localhost:1433;database=test;user=sa;password=password;encrypt=true;trustServerCertificate=true"`
+
+## Health Checks
+
+All database services include health checks to ensure they're ready before running tests. You can check the status with:
+
+```bash
+pnpm docker:ps
+```
+
+## Data Persistence
+
+Database data is persisted in Docker volumes. To reset all data:
+
+```bash
+pnpm docker:reset
+```
+
+This setup ensures consistent testing environments across different machines and makes it easy to test against multiple database providers without manual setup.

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2022-2025 Omar Dulaimi
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.