Files
spark/.serena/memories/spark_project_overview.md

47 lines
1.9 KiB
Markdown
Raw Normal View History

feat(migration): Migrate from Node.js to Deno runtime Major migration to Deno runtime following the nupst project pattern: Core Changes: - Created deno.json configuration with tasks, imports, and settings - Created mod.ts as main entry point with Deno permissions - Updated all TypeScript imports from .js to .ts extensions - Replaced Node.js APIs (process.exit) with Deno equivalents (Deno.exit) - Updated path imports to use @std/path from JSR Dependencies: - Migrated all npm dependencies to use npm: prefix in import map - Added Deno standard library imports (@std/path, @std/assert) - Configured import aliases for all @push.rocks and @serve.zone packages Build & Distribution: - Created install.sh for downloading pre-compiled binaries - Created uninstall.sh for clean system removal - Created scripts/compile-all.sh for multi-platform compilation - Supports Linux (x64, ARM64), macOS (x64, ARM64), Windows (x64) Testing: - Migrated tests to Deno test framework using @std/assert - Created test.simple.ts for basic verification - Updated test structure to use Deno.test instead of tap CI/CD: - Created .gitea/workflows/ci.yml for type checking, linting, and builds - Created .gitea/workflows/release.yml for automated releases - Created .gitea/release-template.md for release documentation Cleanup: - Removed package.json, pnpm-lock.yaml, tsconfig.json - Removed Node.js CLI files (cli.js, cli.child.ts, cli.ts.js) - Removed dist_ts/ compiled output directory - Removed npmextra.json configuration This migration enables standalone binary distribution without Node.js runtime dependency while maintaining all existing functionality.
2025-10-23 23:22:16 +00:00
# Spark Project Overview
## Project Purpose
Spark is a comprehensive tool for maintaining and configuring servers, integrating with Docker and supporting advanced task scheduling, targeted at the Servezone infrastructure and used by @serve.zone/cloudly as a cluster node server system manager.
## Tech Stack
- **Language**: TypeScript
- **Runtime**: Node.js (currently)
- **Package Manager**: pnpm
- **Build Tool**: @git.zone/tsbuild
- **Test Framework**: @git.zone/tstest with @push.rocks/tapbundle
- **CLI Framework**: @push.rocks/smartcli
- **Version**: 1.2.2
## Directory Structure
```
spark/
├── ts/ # TypeScript source files
├── test/ # Test files (single test.nonci.ts)
├── dist_ts/ # Compiled TypeScript output
├── cli.js # CLI entry point
├── cli.child.ts # Child process CLI
├── cli.ts.js # TypeScript CLI wrapper
└── package.json # Dependencies and scripts
```
## Key Dependencies
- **@serve.zone/api**: API client for Servezone
- **@serve.zone/interfaces**: Interface definitions
- **@apiclient.xyz/docker**: Docker API client
- **@push.rocks/*** packages: Various utilities (smartlog, smartfile, smartcli, smartdaemon, etc.)
## Main Components
1. **CLI** (spark.cli.ts): Command-line interface with commands like installdaemon, updatedaemon, asdaemon
2. **Spark** (spark.classes.spark.ts): Main application class
3. **TaskManager** (spark.classes.taskmanager.ts): Task scheduling
4. **UpdateManager** (spark.classes.updatemanager.ts): Service updates
5. **Config** (spark.classes.config.ts): Configuration management
## Commands
- `pnpm build`: Build the TypeScript code
- `pnpm test`: Run tests
- `spark installdaemon`: Install as system daemon
- `spark updatedaemon`: Update daemon service
- `spark asdaemon`: Run as daemon
- `spark logs`: View daemon logs
- `spark prune`: Clean up resources