- Add detailed implementation plan (main-plan.md) - Add professional README with project overview - Include original specifications and screenshots - Document technology stack and architecture - Provide setup and deployment instructions
4.2 KiB
Znakovni.hr – Full Application Replication Specification
Purpose
This document specifies the exact functionality, structure, and behavior of the Znakovni.hr web application. The goal is to enable AI programming agents to design and implement a functionally identical copy of the platform, including UX flows, features, and system architecture.
Explicit permission has been granted by the original author to replicate the application for demonstration, research, and AI capability showcasing purposes.
1. Application Overview
Znakovni.hr is a web-based Croatian Sign Language (HZJ) platform that provides:
- A video-based sign language dictionary
- Sentence construction tools using signs
- Video sentence playback
- User accounts and cloud storage
The platform functions as a learning tool, reference dictionary, and sentence composition environment.
2. Main Navigation & Screens
2.1 Sidebar Navigation
Persistent left sidebar with the following items:
- Početna (Home)
- Riječi (Dictionary)
- Znakopis (Sentence Builder)
- Video rečenica (Video Sentence Player)
- Oblak (Cloud Documents)
- Korištenje aplikacije (Help)
- Zajednica (Community)
- Komentari (Comments)
- Prijavi grešku (Bug Report)
- Sign in / Sign out
Sidebar is visible on all authenticated pages.
3. Screens & Functionality
3.1 Home (Početna)
- Intro headline: “Tvoj put u svijet znakovnog jezika”
- Feature overview cards:
- Dictionary
- Sentence composition (Znakopis)
- Video sentences
- Cloud documents
Static informational screen.
3.2 Dictionary (Riječi)
Features
- Free-text search
- Filters:
- Word type
- CEFR level (A1–C2)
- Paginated grid of word cards
Word Card
Each card includes:
- Icon representation
- Word label
- Difficulty color indicator
- Actions:
- Add to sentence
- Info
Word Detail
- Embedded sign video
- Linguistic metadata
- Usage context
3.3 Znakopis (Sentence Builder)
Purpose
Visual assembly of sign-language sentences.
Layout
- Sentence structure panel (left)
- Sentence description editor (right)
Features
- Add words from dictionary
- Reorder via drag or controls
- Remove words
- Multi-page sentence support
- Save as document
Documents may be saved locally or in cloud.
3.4 Video Rečenica (Video Sentence)
Purpose
Playback of complete sentences using video signs.
Layout
- Video player (left)
- Sentence structure view (right)
Controls
- Play / Pause
- Step navigation
- Repeat
- Speed adjustment (optional)
3.5 Oblak (Cloud Documents)
Purpose
Centralized storage of sentence documents.
Features
- Upload documents
- Save Znakopis documents
- Load and edit documents
- Document listing
3.6 Authentication
Login Methods
- Email/password
- Google OAuth
- Microsoft OAuth
Authenticated Capabilities
- Document persistence
- User-specific content
- Session continuity
4. Core Data Models
User
- id
- display_name
- auth_provider
- created_at
Word
- id
- text
- word_type
- cefr_level
- icon_url
- video_url
Sentence
- id
- user_id
- title
- created_at
SentenceItem
- id
- sentence_id
- word_id
- position
Document
- id
- user_id
- sentence_id
- storage_location
- updated_at
5. API Surface (Illustrative)
- GET /api/words
- GET /api/words/{id}
- POST /api/sentences
- PUT /api/sentences/{id}
- GET /api/documents
- POST /api/auth/login
6. Video Handling
- Short video clips per word
- Consistent framing
- Loop-ready
- Neutral background
Sentence playback = sequential video concatenation.
7. Non-Functional Requirements
- Desktop-first responsive UI
- High accessibility
- Fast filtering/search
- Stateless backend
- Secure auth
8. Architecture (Technology-Agnostic)
- SPA frontend
- REST backend
- Relational database
- Object storage for videos
- OAuth + JWT authentication
9. Functional Parity Criteria
- Dictionary search and filters
- Sentence creation
- Video sentence playback
- Cloud document storage
- Authentication
10. Scope
This document defines a 1:1 functional replication of the Znakovni.hr application, intended for controlled demonstration and AI-driven implementation planning.