166 lines
5.5 KiB
Markdown
166 lines
5.5 KiB
Markdown
# Claro Future Features Roadmap
|
|
|
|
> **For Hermes:** Use subagent-driven-development skill to implement this plan task-by-task.
|
|
|
|
**Goal:** Grow Claro from a beginner scripting language into a stable accessible platform without breaking the learning-friendly core.
|
|
|
|
**Architecture:** Keep the interpreter small and stable. Add major systems behind explicit commands, feature flags, and testable layers. Prefer beginner-readable commands first, then deeper professional tooling around them.
|
|
|
|
**Tech Stack:** C99 interpreter, `.claro` tests/golden outputs, optional helper tools in Python, optional platform libraries such as SDL only after isolated build support exists.
|
|
|
|
---
|
|
|
|
> Current note: Some tasks in this implementation plan have since been partially completed or moved into dedicated status docs. For the current high-level view, start with `ROADMAP.md`, `CURRENT_STATUS.md`, and `COMPLETE_PLATFORM_ROADMAP.md`.
|
|
|
|
## Priority order
|
|
|
|
1. **Advanced static typing foundation** — safest immediate win; improves learner feedback and catches mistakes before running.
|
|
2. **Professional IDE foundations** — language metadata, diagnostics, formatter, LSP/documentation support.
|
|
3. **Concurrency as real runtime feature** — cooperative tasks first; avoid native threads until semantics are clear.
|
|
4. **Production-grade web programming** — HTTP primitives, JSON, routing/server design after safety model.
|
|
5. **GUI/game engine features** — separate stable placeholder graphics from real optional SDL/GUI backends.
|
|
6. **Online package registry** — design package metadata/security before network install.
|
|
7. **Advanced compiler optimization** — only after semantics stabilize; start with bytecode/AST planning, not premature optimization.
|
|
|
|
---
|
|
|
|
## Track A: Advanced Static Typing
|
|
|
|
### Task A1: Add typed LIST and MAP declarations
|
|
|
|
**Objective:** Accept and check beginner-readable container type declarations such as `SET names LIST OF TEXT` and `SET scores MAP OF NUMBER`.
|
|
|
|
**Files:**
|
|
- Modify: `src/claro.c`
|
|
- Test: `tests/40_static_container_good.claro`
|
|
- Test: `tests/40_static_container_good.out`
|
|
- Test: `tests/typecheck_container_bad.claro`
|
|
|
|
**Steps:**
|
|
1. Add a failing good test with `LIST OF TEXT` and `MAP OF NUMBER` declarations.
|
|
2. Add a failing bad typecheck test where a `LIST OF NUMBER` receives text.
|
|
3. Teach `claro_is_type_word`, `claro_type_matches`, and the typecheck parser to preserve simple container annotations.
|
|
4. Verify `claro test` and `claro typecheck tests/typecheck_container_bad.claro`.
|
|
|
|
### Task A2: Add clearer type mismatch messages
|
|
|
|
**Objective:** Make type errors explain the variable, expected type, actual value kind, and line number in plain language.
|
|
|
|
**Files:**
|
|
- Modify: `src/claro.c`
|
|
- Test: `tests/typecheck_bad.claro`
|
|
|
|
**Steps:**
|
|
1. Capture current bad output.
|
|
2. Improve wording without changing successful behavior.
|
|
3. Update tests/docs only after verifying the new wording is clearer.
|
|
|
|
---
|
|
|
|
## Track B: Professional IDE Foundations
|
|
|
|
### Task B1: Expand `claro ide` JSON
|
|
|
|
**Objective:** Output commands, snippets, type names, and file extension metadata for editor plugins.
|
|
|
|
**Files:**
|
|
- Modify: `src/claro.c`
|
|
- Test: `tests/41_ide_metadata.claro` or a CLI validation script.
|
|
|
|
**Steps:**
|
|
1. Add expected JSON fields: `commands`, `snippets`, `diagnostics`, `version`.
|
|
2. Validate with Python `json.loads`.
|
|
3. Document in `docs/IDE.md`.
|
|
|
|
### Task B2: Build minimal LSP helper
|
|
|
|
**Objective:** Provide a Python helper that reads Claro files and returns diagnostics using `claro check`/`typecheck`.
|
|
|
|
**Files:**
|
|
- Create/modify: `support/lsp/claro_lsp_stub.py`
|
|
- Create: `docs/IDE.md`
|
|
|
|
---
|
|
|
|
## Track C: Real Concurrency
|
|
|
|
### Task C1: Define cooperative task semantics
|
|
|
|
**Objective:** Document `START TASK`, `YIELD`, `WAIT TASK`, and `TASK STATUS` without native threads.
|
|
|
|
**Files:**
|
|
- Create: `docs/CONCURRENCY.md`
|
|
- Test: future `tests/42_tasks_cooperative.claro`
|
|
|
|
**Rule:** Start cooperative scheduling before OS threads. This keeps beginner programs deterministic.
|
|
|
|
---
|
|
|
|
## Track D: Web Programming
|
|
|
|
### Task D1: Stabilize HTTP client commands
|
|
|
|
**Objective:** Keep `HTTP CHECK`, `HTTP GET`, and `HTTP SAVE` predictable and testable with `claro://` URLs.
|
|
|
|
**Files:**
|
|
- Modify: `docs/WEB_SERVER_PLAN.md`
|
|
- Tests: existing networking tests
|
|
|
|
### Task D2: Design local web server syntax
|
|
|
|
**Objective:** Propose beginner-friendly server syntax, but do not implement sockets until parser/runtime tests are ready.
|
|
|
|
Example future shape:
|
|
|
|
```claro
|
|
WEB SERVER 8080
|
|
WEB GET "/" DO home
|
|
TEACH home request
|
|
RETURN "Hello"
|
|
END
|
|
WEB START
|
|
```
|
|
|
|
---
|
|
|
|
## Track E: GUI/Game Engine
|
|
|
|
### Task E1: Separate placeholder graphics from real SDL
|
|
|
|
**Objective:** Keep stable no-op graphics commands documented separately from optional SDL backends.
|
|
|
|
**Files:**
|
|
- Modify: `docs/SDL12.md`
|
|
- Create: `docs/GRAPHICS.md`
|
|
|
|
### Task E2: Optional SDL prototype branch
|
|
|
|
**Objective:** Only after build scripts can find SDL 1.2 reliably, add `CLARO_WITH_SDL12` support behind a compile flag.
|
|
|
|
---
|
|
|
|
## Track F: Online Package Registry
|
|
|
|
### Task F1: Define package manifest and lock security
|
|
|
|
**Objective:** Design package metadata, versions, hashes, and allowed sources before downloads.
|
|
|
|
**Files:**
|
|
- Create: `docs/PACKAGE_REGISTRY.md`
|
|
|
|
**Rule:** No online package install until hash verification and safe paths are implemented.
|
|
|
|
---
|
|
|
|
## Track G: Optimization
|
|
|
|
### Task G1: Add timing benchmark command
|
|
|
|
**Objective:** Measure before optimizing.
|
|
|
|
**Files:**
|
|
- Create: `tests/benchmarks/`
|
|
- Create: `docs/PERFORMANCE.md`
|
|
|
|
**Rule:** No optimizer work without benchmarks.
|