# khanh02-vin/docdich architecture

> Architecture diagram of the GitHub repository khanh02-vin/docdich, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-17). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/khanh02-vin/docdich
- Repository: https://github.com/khanh02-vin/docdich

## Overview

Neko Manga Translator is a Firefox extension that detects manga images in a page, sends them through a relay to a vision-backed translation service, then renders translated text as a browser overlay. The backend-mit tree also contains the general image-translation engine and API; backend-translate supplies glossary, memory, and style support. The README documents the extension request flow, while sampled backend source confirms API and pipeline responsibilities; several extension and auxiliary-service links therefore rely on README evidence rather than inspected implementation.

Main entry points: [`backend-mit/manga_translator/manga_translator.py`](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/manga_translator.py) (Image Pipeline), [`extension-neko/mvp/extension/content.js`](https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/content.js) (Image Queue), [`extension-neko/mvp/mit_relay.py`](https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/mit_relay.py) (Translation Relay), [`backend-mit/server/main.py`](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/main.py) (MIT API)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_extension["Firefox Extension"]
  node_content["Image Queue<br/>[content.js]"]
  node_cache[("Page Cache<br/>[content.js]")]
  node_background["Request Broker<br/>[background.js]"]
  node_relay["Translation Relay<br/>[mit_relay.py]"]
  node_overlay["Text Overlay<br/>[content.js]"]
end

subgraph group_api["Translation API"]
  node_mitapi["MIT API<br/>[main.py]"]
  node_vision["Vision Translation<br/>[vision.py]"]
  node_queue["Work Queue<br/>[myqueue.py]"]
  node_worker["Translation Instance<br/>[instance.py]"]
  node_serializer["JSON Serializer<br/>[to_json.py]"]
end

subgraph group_engine["Translation Engine"]
  node_engine["Image Pipeline"]
  node_detection["Text Detection<br/>[__init__.py]"]
  node_ocr["Text Recognition<br/>[__init__.py]"]
  node_translation["Text Translation<br/>[chatgpt.py]"]
  node_inpainting["Bubble Inpainting<br/>[__init__.py]"]
  node_render["Image Rendering<br/>[__init__.py]"]
end

subgraph group_support["Translation Support"]
  node_translateapi["Support API<br/>[main.py]"]
  node_personalization[("Glossary Memory Style")]
  node_localmodel["Local NLLB<br/>[server.py]"]
end

node_page(("Manga Page"))

node_page -->|"scans images"| node_content
node_content -->|"checks cache"| node_cache
node_content -->|"dispatches request"| node_background
node_background -->|"forwards request"| node_relay
node_relay -->|"posts image"| node_mitapi
node_mitapi -->|"routes vision request"| node_vision
node_vision -->|"enqueues work"| node_queue
node_queue -->|"dispatches task"| node_worker
node_worker -->|"runs translation"| node_engine
node_engine -->|"dispatches detection"| node_detection
node_engine -->|"dispatches OCR"| node_ocr
node_engine -->|"dispatches translation"| node_translation
node_engine -->|"dispatches inpainting"| node_inpainting
node_engine -->|"dispatches rendering"| node_render
node_mitapi -->|"serializes result"| node_serializer
node_serializer -->|"returns metadata"| node_relay
node_relay -->|"returns translations"| node_content
node_content -->|"renders overlay"| node_overlay
node_translateapi -.->|"uses support services"| node_personalization
node_translateapi -.->|"can select local provider"| node_localmodel

click node_content "https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/content.js"
click node_cache "https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/content.js"
click node_background "https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/background.js"
click node_relay "https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/mit_relay.py"
click node_overlay "https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/content.js"
click node_mitapi "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/main.py"
click node_vision "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/vision.py"
click node_queue "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/myqueue.py"
click node_worker "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/instance.py"
click node_serializer "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/to_json.py"
click node_engine "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/manga_translator.py"
click node_detection "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/detection/__init__.py"
click node_ocr "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/ocr/__init__.py"
click node_translation "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/translators/chatgpt.py"
click node_inpainting "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/inpainting/__init__.py"
click node_render "https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/rendering/__init__.py"
click node_translateapi "https://github.com/khanh02-vin/docdich/blob/master/backend-translate/apps/backend/main.py"
click node_personalization "https://github.com/khanh02-vin/docdich/tree/master/backend-translate/services"
click node_localmodel "https://github.com/khanh02-vin/docdich/blob/master/backend-translate/services/local_translation/server.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_content,node_cache,node_background,node_relay,node_overlay toneBlue
class node_mitapi,node_vision,node_queue,node_worker,node_serializer toneAmber
class node_engine,node_detection,node_ocr,node_translation,node_inpainting,node_render toneMint
class node_translateapi,node_personalization,node_localmodel toneRose
class node_page toneIndigo
```

## Components

### Firefox Extension

- [Image Queue](https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/content.js): `extension-neko/mvp/extension/content.js`
- [Page Cache](https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/content.js): `extension-neko/mvp/extension/content.js`
- [Request Broker](https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/background.js): `extension-neko/mvp/extension/background.js`
- [Translation Relay](https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/mit_relay.py): `extension-neko/mvp/mit_relay.py`
- [Text Overlay](https://github.com/khanh02-vin/docdich/blob/master/extension-neko/mvp/extension/content.js): `extension-neko/mvp/extension/content.js`

### Translation API

- [MIT API](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/main.py): `backend-mit/server/main.py`
- [Vision Translation](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/vision.py): `backend-mit/server/vision.py`
- [Work Queue](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/myqueue.py): `backend-mit/server/myqueue.py`
- [Translation Instance](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/instance.py): `backend-mit/server/instance.py`
- [JSON Serializer](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/server/to_json.py): `backend-mit/server/to_json.py`

### Translation Engine

- [Image Pipeline](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/manga_translator.py): `backend-mit/manga_translator/manga_translator.py`
- [Text Detection](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/detection/__init__.py): `backend-mit/manga_translator/detection/__init__.py`
- [Text Recognition](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/ocr/__init__.py): `backend-mit/manga_translator/ocr/__init__.py`
- [Text Translation](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/translators/chatgpt.py): `backend-mit/manga_translator/translators/chatgpt.py`
- [Bubble Inpainting](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/inpainting/__init__.py): `backend-mit/manga_translator/inpainting/__init__.py`
- [Image Rendering](https://github.com/khanh02-vin/docdich/blob/master/backend-mit/manga_translator/rendering/__init__.py): `backend-mit/manga_translator/rendering/__init__.py`

### Translation Support

- [Support API](https://github.com/khanh02-vin/docdich/blob/master/backend-translate/apps/backend/main.py): `backend-translate/apps/backend/main.py`
- [Glossary Memory Style](https://github.com/khanh02-vin/docdich/tree/master/backend-translate/services): `backend-translate/services`
- [Local NLLB](https://github.com/khanh02-vin/docdich/blob/master/backend-translate/services/local_translation/server.py): `backend-translate/services/local_translation/server.py`

### Other components

- **Manga Page**

## Connections

- Manga Page → Image Queue: scans images
- Image Queue → Page Cache: checks cache
- Image Queue → Request Broker: dispatches request
- Request Broker → Translation Relay: forwards request
- Translation Relay → MIT API: posts image
- MIT API → Vision Translation: routes vision request (evidence: `backend-mit/server/main.py`)
- Vision Translation → Work Queue: enqueues work (evidence: `backend-mit/server/vision.py`)
- Work Queue → Translation Instance: dispatches task (evidence: `backend-mit/server/myqueue.py`)
- Translation Instance → Image Pipeline: runs translation (evidence: `backend-mit/server/instance.py`)
- Image Pipeline → Text Detection: dispatches detection (evidence: `backend-mit/manga_translator/manga_translator.py`)
- Image Pipeline → Text Recognition: dispatches OCR (evidence: `backend-mit/manga_translator/manga_translator.py`)
- Image Pipeline → Text Translation: dispatches translation (evidence: `backend-mit/manga_translator/manga_translator.py`)
- Image Pipeline → Bubble Inpainting: dispatches inpainting (evidence: `backend-mit/manga_translator/manga_translator.py`)
- Image Pipeline → Image Rendering: dispatches rendering (evidence: `backend-mit/manga_translator/manga_translator.py`)
- MIT API → JSON Serializer: serializes result (evidence: `backend-mit/server/main.py`)
- JSON Serializer → Translation Relay: returns metadata
- Translation Relay → Image Queue: returns translations
- Image Queue → Text Overlay: renders overlay
- Support API → Glossary Memory Style: uses support services (evidence: `backend-translate/apps/backend/main.py`)
- Support API → Local NLLB: can select local provider
