feat: Add Gemini API translation support

[kapitalismho]

Description

Summary

Adds Gemini API translation support to mimiuchi.

Motivation

I first discovered mimiuchi at Vket and was impressed by its fast speech recognition. However, I always felt the translation feature was lacking—it felt like a shame to leave such a great program as-is. So I decided to build it myself.

What I Built

• Full Gemini API integration (communication logic, GUI, etc.)

• Fully customizable system prompt for flexibility

Real-World Testing

• After using it for about two days:
  • Latency: ~3-5 seconds from speech to translated output
  • Quality: Good translation accuracy

Technical Approach

I tried to keep the existing code as intact as possible for a non-invasive integration.

For example, since Gemini uses BCP-47 language codes while mimiuchi uses NLLB-200 codes, I created a mapping layer between them instead of modifying the existing code.

What is the purpose of this pull request?

• Bug fix
• New Feature
• Documentation update
• Other