如何取回创建帖子的原始“原始”数据？

开发
21

If you’re using chrome desktop then you can use Ctrl+Shift+V to paste as plain text instead of notepad.

Also in the admin site settings page, you can disable this fancy pasting behavior by unchecking the “enable rich text paste” option.

2 个赞
22

Thanks, but my goal is to keep the original formatting from microsoft word (in the raw data), not strip it down to plain text.

Currently, Discourse removes a variety of the formatting–such as font size and font type, while leaving others (bold and italics seem to stay there). As a comparison, Gmail composer seems to leave all formatting.

To try to keep more formatting, I suppose an alternative would be to upload a word document (not paste in text to the composer–but upload the actual doc). The problem there, I think, is that discourse does not show the contents of the upload inline (it just displays an upload link).

23

Then paste the raw HTML.

There are different goals involved. Gmail’s goal is to preserve all the formatting, whereas a forum’s goal is to keep a meaningful subset of it while preventing abuse (huge text, blinking text, OVERLY LARGE FONTS, annoying colours, etc.)

As a simple example, here is some simple text’s HTML as generated by Office, present in the clipboard as text/html:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<html>
<head>
	<meta http-equiv="content-type" content="text/html; charset=utf-8"/>
	<title></title>
	<meta name="generator" content="LibreOffice 7.1.2.2 (Linux)"/>
	<style type="text/css">
		@page { size: 21.59cm 27.94cm; margin: 2cm }
		p { margin-bottom: 0.25cm; line-height: 115%; background: transparent }
	</style>
</head>
<body lang="en-CA" link="#000080" vlink="#800000" dir="ltr"><p style="margin-bottom: 0cm; line-height: 100%">
<b>Hello there</b><span style="font-weight: normal"> sunshine </span><i><span style="font-weight: normal">eh</span></i></p>
</body>
</html>

It looks like:

@page { size: 21.59cm 27.94cm; margin: 2cm } p { margin-bottom: 0.25cm; line-height: 115%; background: transparent }

Hello there sunshine eh

But when interpreted by to-markdown.js you get:

**Hello there** sunshine *eh*

Hello there sunshine eh

You cannot, unless you put it there yourself as I did in this post. If you really need it there, hide it in a comment. If you want to later convert it yourself to markdown, use something like pandoc.

2 个赞
24

I support transferring to markdown on the post that is displayed to others for some consistency. It’s the raw entry I’m curious about.

How do you get the raw html of a word doc that you could paste in?

25

Save it as HTML from Word, or copy it to the clipboard and pull the text/html explicitly from the clipboard.

4 个赞
26

If all you care about is being able to reference the raw input at a later point, this component might work for you.

It adds a button to the post menu that shows the raw content on a per-post basis.

5 个赞
27

@jomaxro，评论的原始标记语法是什么？

28

您在问什么？您是在问 markdown 是什么吗？比如 https://commonmark.org/ 您是在问如何获取帖子的原始 markdown 吗？比如 /raw/123

29

对于单个帖子，您可以将 ID 添加到 URL 中。例如，https://meta.discourse.org/raw/189183/27 是您的回复。

2 个赞
30

@pfaffman，是后者。谢谢，@Moin。回想起来，这很明显。

31

@jomaxro，/raw/ 在 discuss.kde.org/c/community/blogs/24 的任何帖子中都无效，因为在 view-source:discuss.kde.org/raw/43656 处，我看到：

<pre>Konqi | 2026-01-23 22:43:37 UTC | #1

&lt;p&gt;Leaking memory is impolite. It’s messy, it can suggest logic bugs, and thanks to AI grifters RAM is expensive.&lt;/p&gt;
&lt;hr&gt;
&lt;small&gt;This is a companion discussion topic for the original entry at &lt;a href="https://nicolasfella.de/posts/detecting-leaks-in-kde-ci/?utm_source=atom_feed"&gt;https://nicolasfella.de/posts/detecting-leaks-in-kde-ci/?utm_source=atom_feed&lt;/a&gt;&lt;/small&gt;

---------
</pre>

然而，当在 discuss.kde.org/t/43656 上交互式展开 details 时，我可以将其展开为：

1276×1076 184 KB

这是否可以在上游 Discourse 端得到可行性的支持？我观察到其他略有定制的 Discourse 实例也有类似情况，因此我提出这个问题。

32

您正试图解决什么问题？

这些主题很特别，因为它们链接到大型帖子。

这是 op 的原始文本。

1 个赞
33

该主题中 OP 的纯文本如下：

○ → curl https://discuss.kde.org/posts/132565/raw
<p>Leaking memory is impolite. It’s messy, it can suggest logic bugs, and thanks to AI grifters RAM is expensive.
</p>
<hr>
<small>This is a companion discussion topic for the original entry at <a href="https://nicolasfella.de/posts/detecting-leaks-in-kde-ci/?utm_source=atom_feed">https://nicolasfella.de/posts/detecting-leaks-in-kde-ci/?utm_source=atom_feed</a></small>

这与“已渲染”版本中显示的内容完全相同。

这不是展开 details 块。点击 Show Full Post 会加载其他内容，这是嵌入内容的完整展开，它不是来自帖子原始内容，而是来自额外的帖子元数据。

您可以在浏览器中点击它时看到网络请求，等效请求如下：

○ → curl -s -H 'accept: application/json' 'https://discuss.kde.org/posts/132565/expand-embed' | jq -r .cooked
<div><div>
            <p>Leaking memory is impolite. It’s messy, it can suggest logic bugs, and thanks to AI grifters RAM is expensive.</p>
…
…
…
<p>LSAN is now enabled for some Frameworks CI builds, but ideally it would be enabled for all KDE projects. And of course any leaks found along the way should be fixed.</p>
<p>Happy leak-fixing!</p>
        </div></div>
<hr>
<small>This is a companion discussion topic for the original entry at <a href='https://nicolasfella.de/posts/detecting-leaks-in-kde-ci/?utm_source=atom_feed'>https://nicolasfella.de/posts/detecting-leaks-in-kde-ci/?utm_source=atom_feed</a></small>

您是想查看嵌入内容吗？如果是，请使用上述内容。如果不是，您期望看到什么？

2 个赞
34

@supermathie，确实如此。非常感谢！显然，我所知道的 Discourse API 远不止这些。

1 个赞