Can I use mdx-loader to dynamically load my markdown file within the react component?

[heliannuuthus]

I want to parse all of word entries in my blog(or website).

1. I will create a glossary for record explaintion of word entry.
   
2. for some WEntry(word entry), i will refrer them to slug of word entries in glossary that like a link
   

But now I want to reuse the processor of @docusaurus/mdx-loader, and I want my entries to render in the same way as the rest of my blog, with the same supported syntax

[slorber]

It's difficult for me to understand what you want sorry 😅

[heliannuuthus]

Thank you for your reply ~, I am in my second development project docusaurus-terminology plugin, originally it uses remark (can use unified) to convert hoverText to markdown (detailed code), but I want to convert the content to markdown instead of hoverText, so I want to use the processor built into @docusaurus/mdx-loader to process the content. Because I want content to be consistent with the syntax that normal blogs support

[slorber]

What I understand:

• You have a glossary using MDX docs
• You want that when the user hovers over a term in a docs page, a tooltip displays some content
• The tooltip content should be provided using MDX syntax, similar to the glossary page

---

Unfortunately, I don't think it is possible to do that easily. MDX docs are converted to real React components under the hood. It is actual code, similar to React code you'd write directly, and not "processed strings" that you can serialize. This means that docs pages would have to be able to use the code produced for your terms. Plugins being "sandboxed", our docs plugin pages can't easily use the MDX (React components) of another glossary plugin.

What you want to achieve is likely going to be solved with React Server Components, tracked in #9089

[heliannuuthus]

emmm, I did spend 3 weeks of my free time developing remark plugin + webpack loader combined with custom MDXComponents to complete it. I introduced unist in webpack-loader myself to parse the markdown in the glossary, bypassing @docusaurus/mdx-loader. I now want to use the markdown processor of @docusaurus/mdx-loader in my webpack-loader instead of unist to do the job of parsing the glossary markdown. But I found that @docusaurus/mdx-loader didn't export it 🧐 .

[slorber]

Sorry but it's really hard for me to understand you, please create a minimal repro of your problem so that I can take a look at your problem.

The MDX loader doesn't export everything because things are considered internal implementation details, not public API surface.

However, our package doesn't use "exports" so tech nically you can import implementation details with import ... from "@docusaurus/mdx-loader/lib/..." if you really want to, but keep in mind we might disallow this in the future.