# pardnchiu/go-rss-reader

> go-rss-reader by Pardn Chiu: a Go terminal RSS reader with reader-mode full-text extraction, SQLite offline cache, and a rolling LLM news digest.

Maintained by 邱敬幃 Pardn Chiu (pardnchiu), Pardn Co., Ltd, Taiwan · Infrastructure Engineering. Source: https://github.com/pardnchiu/go-rss-reader

Full text in one file: [English](https://go-rss-reader.pardn.io/llms-full.txt) · [中文](https://go-rss-reader.pardn.io/zh/llms-full.txt)

## Overview

- [Home](https://go-rss-reader.pardn.io/index.md): go-rss-reader by Pardn Chiu: a Go terminal RSS reader with reader-mode full-text extraction, SQLite offline cache, and a rolling LLM news digest.
- [Getting Started](https://go-rss-reader.pardn.io/getting-started.md): Build go-rss-reader from source with CGO, add your first RSS feed, read full articles in the terminal, and enable the LLM digest.

## Concepts

- [Architecture](https://go-rss-reader.pardn.io/architecture.md): Layer overview of go-rss-reader: TUI app, RSS collector, content extractor, OpenAI client, and SQLite storage, plus cross-cutting rules.
- [Refresh Cycle](https://go-rss-reader.pardn.io/refresh-cycle.md): How one update check runs: launch, Ctrl+R, and the 5-minute ticker, new-article detection, background extraction, then the digest.
- [Feed Collection](https://go-rss-reader.pardn.io/feed-collection.md): How Collector manages feeds and parses RSS 2.0: deduplication by link, the 72-hour window, HTML cleanup, and pubDate formats.
- [Content Extraction](https://go-rss-reader.pardn.io/content-extraction.md): How Extractor pulls title, author, and body from news pages with goquery: noise removal, link-density body selection, and word count.
- [LLM Digest](https://go-rss-reader.pardn.io/llm-digest.md): When the Summary pane digest runs, what gets sent to gpt-4o-mini, the Traditional Chinese output sections, and the streaming API client.
- [Storage](https://go-rss-reader.pardn.io/storage.md): SQLite schema of go-rss-reader: the news, feeds, and data tables, their indexes, and every database.SQLite access method.

## Reference

- [Configuration](https://go-rss-reader.pardn.io/configuration.md): Configure go-rss-reader: RSS_DB_PATH and GO_ENV, database file resolution order, the stored OpenAI API key, and fixed parameters.
- [CLI Reference](https://go-rss-reader.pardn.io/cli-reference.md): go-rss-reader hotkeys, in-app commands (add, rm, apikey, config), pane layout, and status bar messages.

## 中文文件

- [首頁](https://go-rss-reader.pardn.io/zh/index.md): 邱敬幃開發的 Go 終端 RSS 閱讀器，具備閱讀模式全文萃取、SQLite 離線快取與 LLM 滾動新聞概要。
- [快速開始](https://go-rss-reader.pardn.io/zh/getting-started.md): 以 CGO 從原始碼建置 go-rss-reader、加入第一個 RSS 訂閱源、在終端讀全文並啟用 LLM 概要。
- [架構](https://go-rss-reader.pardn.io/zh/architecture.md): go-rss-reader 分層概覽：TUI 應用層、RSS 收集器、全文萃取器、OpenAI 用戶端與 SQLite 儲存及跨切原則。
- [刷新流程](https://go-rss-reader.pardn.io/zh/refresh-cycle.md): 一次檢查更新的完整順序：啟動、Ctrl+R 與 5 分鐘自動檢查、新文章判定、背景全文萃取到產生概要。
- [訂閱源收集](https://go-rss-reader.pardn.io/zh/feed-collection.md): Collector 如何管理訂閱源並解析 RSS 2.0：依連結去重、72 小時時間窗、HTML 清理與 pubDate 格式。
- [全文萃取](https://go-rss-reader.pardn.io/zh/content-extraction.md): Extractor 如何以 goquery 從新聞網頁取出標題、作者與正文：雜訊移除、連結密度判定正文與字數計算。
- [LLM 新聞概要](https://go-rss-reader.pardn.io/zh/llm-digest.md): Summary 欄概要的觸發條件、送交 gpt-4o-mini 的內容、繁體中文輸出段落與串流 API 用戶端。
- [儲存](https://go-rss-reader.pardn.io/zh/storage.md): go-rss-reader 的 SQLite 結構：news、feeds、data 三張資料表、索引與 database.SQLite 所有存取方法。
- [設定](https://go-rss-reader.pardn.io/zh/configuration.md): go-rss-reader 設定：RSS_DB_PATH 與 GO_ENV、資料庫檔案位置判斷順序、儲存的 OpenAI 金鑰與固定參數。
- [命令列參考](https://go-rss-reader.pardn.io/zh/cli-reference.md): go-rss-reader 快捷鍵、介面內指令（add、rm、apikey、config）、介面區塊配置與狀態列訊息。

## Optional

- [Release Notes](https://go-rss-reader.pardn.io/released/index.md): pardnchiu/go-rss-reader changelog by version