Lectern — 阅读时用到的工具

Lectern 是读网页文章的工具集合。 · 阅读追踪：自动认出文章页，记录每次专注了多久、读了多少、读到哪一段；重新打开时跳回上次的位置。 · 划词翻译与讲解：选中英文，给出结合上下文的翻译和用法讲解，句子里的生词逐个讲开。 · 截图翻译：框选图片里的文字，在本机识别后翻译。 · 复习：查过的词自动做成卡片，用 FSRS 间隔重复安排复习。 · 多设备：可以连接自己部署的同步服务器，在电脑和安卓 App 之间合并记录。 使用前请注意： · 需要你自己的 MiniMax API Key，在设置页填写。没有 Key 时不会记录文章，也不能翻译。 · 填了 Key 之后，你打开的网页的网址、标题和正文会自动发给模型，用来判断它是不是文章。可以在设置页的「文章记录黑名单」里排除网站。 · 设备同步需要自己部署后端，不启用时所有数据只留在本机。 · 界面目前只有中文。 源代码与说明：https://github.com/xbghc/lectern Lectern is a reading desk: the tools you need while reading articles on the web, in one place. · Reading tracker: detects article pages, records how long you focused, how much you read and where you stopped, and jumps back there next time. · Translate and explain: select English text to get a translation in context, a usage note, and each unfamiliar word explained. · Screenshot translation: draw a box around text in an image; it is recognized on your device and then translated. · Review: looked-up words become cards scheduled with FSRS spaced repetition. · Multiple devices: optionally connect a sync server you host yourself to merge records between computers and the Android app. Before you install: · You need your own MiniMax API key, entered in settings. Without a key nothing is recorded or translated. · Once a key is set, the URL, title and text of pages you open are sent to the model automatically to decide whether each page is an article. Exclude sites with the "article blocklist" in settings. · Device sync requires a backend you deploy yourself. With sync off, all data stays on your device. · The interface is currently Chinese only. Source code and documentation: https://github.com/xbghc/lectern