Smart Study Task Manager

A JavaFX desktop study assistant designed for students to plan, schedule, and analyze study tasks. The system combines a drag-and-drop weekly calendar, risk-aware task analytics, and an LLM-ready assistant API, while remaining fully functional offline with local persistence.

---

✨ Key Features

📅 Visual Scheduling

• Weekly calendar with compressed morning view (00:00–08:00) and a 26-hour vertical span
• Drag-and-drop task scheduling with 30-minute snapping
• Real-time current-time indicator and clear 24:00 boundary

🧠 Task Modeling

• Task attributes: priority, deadline, estimated duration, optional start time
• Automatic overdue detection and status update
• Postponement tracking (e.g. “已推迟 × N”)

⚠️ Risk & Analytics

• High-risk task detection
• Upcoming-deadline highlighting
• Warning when scheduled end time exceeds deadline

💬 Natural-Language Assistant

• SET mode: create or update tasks via natural language

• CODE mode: generate runnable Java tools (full source code required)

• Context-aware chat interface

  • Enter to send
  • Shift + Enter for newline

🧪 Local Tool Sandbox

• Run any executable JAR placed under Tools/
• Supports both CLI and GUI applications
• Isolated execution with logging

💾 Persistence

• File-based storage using JSON
• User data and settings stored locally under ~/.smart-study
• Fully usable without network access

---

🏗 Architecture Overview

Layered and extensible design, suitable for further research or feature expansion:

ui → controller → service → repository → model
                     ↓
                    api (LLM-ready)

---

📂 Project Structure

src/main/java/edu/study/
├─ ui/
│  └─ SmartTaskWidget.java        # JavaFX UI
├─ controller/
│  └─ TaskController.java         # Orchestration layer
├─ service/
│  └─ TaskService.java            # Core task logic
├─ api/
│  ├─ AssistantAPI.java           # LLM-facing abstraction
│  └─ impl/
│     ├─ OpenAIChatAssistantAPI.java
│     └─ OpenAIChatClient.java    # Chat + SET / CODE handling
├─ repository/                    # File / JSON persistence
├─ model/                         # Task, Priority, Status, etc.
└─ util/
   └─ ToolSandboxRunner.java      # Local JAR execution

Tools/    # Place runnable JAR tools here
logs/     # Chat, CODE generation, and sandbox logs

---

🔧 Prerequisites

• JDK 17+
• Maven 3.8+
• (Optional) OpenAI-compatible API endpoint for LLM features

---

🚀 Quick Start

Run in development mode

mvn javafx:run

Build executable package

mvn -DskipTests package

---

🤖 LLM Configuration (Optional)

Create a .env file or set environment variables:

OPENAI_API_KEY=your_api_key
OPENAI_BASE_URL=https://api.openai.com/v1   # or any compatible endpoint
OPENAI_MODEL=gpt-4o

Assistant Behavior

• Task operations → return SET: followed by natural-language task description
• Code generation → return CODE: + FileName.java + complete Java source (compiled and packaged automatically into Tools/)

If no LLM is available, the system falls back to a rule-based assistant.

---

🧭 Usage Tips

• Calendar

  • Drag tasks to schedule
  • Right-click-drag to move existing blocks
  • Morning hours (0–8h) are visually compressed

• Sandbox

  • Click refresh to reload JARs from Tools/
  • GUI tools open in separate windows
  • Logs are captured automatically

• Chat

  • Context panel displays user profile and next task summary
  • Logs stored under logs/

---

📁 Data & Settings

• Tasks:

  ~/.smart-study/tasks.json
  

• User settings & profile:

  ~/.smart-study/settings.json
  

---

📝 Notes

• CODE generation debug logs:

  logs/code-debug.log
  

• Sandbox execution logs are visible both in UI and log files

---

📄 License

MIT License (Adjust as needed)

Author

Jinnn2 金厚泽 北京航空航天大学