# Human Language Mobile Integration Guide

This guide explains how to integrate the new **Human Language** feature in the **Yimaru learner mobile app** (not admin).

It is designed to keep the existing non-language hierarchy intact while introducing a dedicated CEFR-based flow for language learning.

---

## 1) Scope and Goals

### What is new

- A dedicated backend API namespace for Human Language:
  - `GET /api/v1/course-management/human-language/courses/:courseId/lessons?cefr_level=A1..C3`
  - `POST /api/v1/course-management/human-language/lessons` (admin/instructor side)
  - `PATCH /api/v1/course-management/human-language/lessons/:id` (admin/instructor side)
- CEFR levels are fixed to:
  - `A1, A2, A3, B1, B2, B3, C1, C2, C3`
- No custom sub-levels in Human Language flow.

### What remains unchanged

- Existing non-human-language content hierarchy and APIs.
- Existing course/category/sub-course endpoints for non-language domains (programming, etc.).

---

## 2) Target Hierarchy for Human Language

- Course Category (Human Language)
  - Course (e.g., English)
    - CEFR Lesson Unit (A1..C3 only)
      - Intro/lesson videos
      - Practices
        - Audio questions

Backend implementation stores CEFR lesson units using the existing sub-course model, with CEFR mapped internally and validated strictly.

---

## 3) Authentication and Access

All Human Language endpoints are under `/api/v1` and require bearer token auth.

- Header:
  - `Authorization: Bearer <access_token>`

### Permission notes

- **Learner mobile app** typically needs only the `GET` endpoint.
- `POST` and `PATCH` are content-management endpoints and should generally be used by admin/instructor clients, not learner clients.

If learner roles do not currently have `learning_tree.get`, coordinate RBAC assignment for read-only access to the `GET` endpoint.

---

## 4) Endpoint Contracts

## 4.1 Fetch lessons by CEFR level (learner-facing)

### Request

`GET /api/v1/course-management/human-language/courses/{courseId}/lessons?cefr_level={A1|A2|A3|B1|B2|B3|C1|C2|C3}`

### Example

`GET /api/v1/course-management/human-language/courses/12/lessons?cefr_level=B1`

### Response (success)

```json
{
  "message": "Human-language lessons retrieved successfully",
  "data": {
    "course_id": 12,
    "course_title": "English",
    "cefr_level": "B1",
    "lessons": [
      {
        "id": 201,
        "course_id": 12,
        "title": "B1 Module 1",
        "description": "Intermediate conversational patterns",
        "thumbnail": "https://.../thumb.jpg",
        "display_order": 1,
        "level": "B1",
        "video_count": 3,
        "practice_count": 2,
        "videos": [
          {
            "id": 9001,
            "title": "B1 Intro",
            "description": "Lesson intro",
            "video_url": "https://...",
            "duration": 420,
            "resolution": "1080p",
            "display_order": 1
          }
        ],
        "practices": [
          {
            "id": 4401,
            "title": "B1 Speaking Practice",
            "status": "PUBLISHED",
            "question_count": 10
          }
        ]
      }
    ]
  },
  "success": true,
  "status_code": 200,
  "metadata": null
}
```

### Validation errors

- Invalid CEFR level:
  - `400` with error message: `Use one of: A1,A2,A3,B1,B2,B3,C1,C2,C3`
- Course not in human language category:
  - `400` with error message indicating invalid human-language course.

---

## 4.2 Create Human Language lesson unit (admin/instructor)

### Request

`POST /api/v1/course-management/human-language/lessons`

```json
{
  "course_id": 12,
  "title": "A2 Module 1",
  "description": "A2 speaking fundamentals",
  "thumbnail": "https://...",
  "display_order": 1,
  "cefr_level": "A2"
}
```

### Response

Returns created lesson metadata with CEFR-normalized level behavior.

---

## 4.3 Update Human Language lesson unit (admin/instructor)

### Request

`PATCH /api/v1/course-management/human-language/lessons/{lessonId}`

```json
{
  "title": "A2 Module 1 - Updated",
  "description": "Updated description",
  "cefr_level": "A3",
  "is_active": true
}
```

---

## 5) Recommended Mobile App Integration Flow

## Step 1: Discover courses under Human Language category

Use existing endpoints:

1. `GET /course-management/categories`
2. Pick category where name indicates human language.
3. `GET /course-management/categories/:categoryId/courses`

## Step 2: Select CEFR level in UI

Use static list in app:

- `A1, A2, A3, B1, B2, B3, C1, C2, C3`

Do not allow custom levels in mobile UI.

## Step 3: Fetch lessons by selected CEFR level

Call:

- `GET /course-management/human-language/courses/:courseId/lessons?cefr_level=<selected>`

Render grouped lessons with nested videos/practices from response.

## Step 4: Navigate to learning/practice screens

- Lesson video playback uses returned `videos[]`.
- Practice entry uses returned `practices[]` IDs and existing practice-question endpoints.

---

## 6) Mobile UI/UX Recommendations

- Use CEFR tabs/segmented control (`A1`...`C3`) at top.
- Cache last selected level per course.
- Show empty state per level:
  - "No lessons available for this level yet."
- Sort by `display_order` and maintain backend order.
- Show badges:
  - video count
  - practice count

---

## 7) Error Handling and Resilience

- `400` invalid level: reset selection to previous valid CEFR value and show toast/snackbar.
- `401/403`: trigger token refresh / re-login flow.
- `5xx`: show retry UI with exponential backoff.
- If category/course fetch succeeds but level fetch fails, keep course visible and allow manual retry.

---

## 8) Data Model Mapping (Mobile DTO)

Suggested DTO for learner app:

```ts
type CefrLevel = "A1"|"A2"|"A3"|"B1"|"B2"|"B3"|"C1"|"C2"|"C3";

interface HumanLanguageLessonDTO {
  id: number;
  courseId: number;
  title: string;
  description?: string;
  thumbnail?: string;
  order: number;
  level: CefrLevel;
  videoCount: number;
  practiceCount: number;
  videos: {
    id: number;
    title: string;
    url: string;
    duration: number;
    order: number;
  }[];
  practices: {
    id: number;
    title: string;
    status: string;
    questionCount: number;
  }[];
}
```

---

## 9) Backward Compatibility

- Existing non-language content (programming and other categories) continues to use current APIs and hierarchy unchanged.
- New Human Language endpoint is additive and isolated.
- Mobile app can progressively enable the new flow for language categories only.

---

## 10) QA Checklist for Mobile Team

- [ ] Category discovery correctly identifies Human Language category.
- [ ] CEFR selector only allows A1..C3.
- [ ] Fetch by CEFR level returns only matching lessons.
- [ ] Video/practice counts match rendered lists.
- [ ] Empty-level state works.
- [ ] Unauthorized/session-expired flow works.
- [ ] Non-language courses still load via existing app flow.

---

## 11) Example Call Matrix

- Load language categories/courses:
  - `GET /course-management/categories`
  - `GET /course-management/categories/:id/courses`
- Load A1 lessons:
  - `GET /course-management/human-language/courses/:courseId/lessons?cefr_level=A1`
- Load B2 lessons:
  - `GET /course-management/human-language/courses/:courseId/lessons?cefr_level=B2`

---

For backend ownership questions, refer to:

- `internal/web_server/handlers/course_management.go`
- `internal/web_server/routes.go`