Replies: 2 comments 8 replies
-
|
There is an extension that adds position information to the tokens: https://github.com/UziTech/marked-token-position Would that be helpful? |
Beta Was this translation helpful? Give feedback.
-
|
Yes — I think so. Would it be easy to use walkTokens to add a link (as per the above) to the end of every paragraph token? I ask because I've never used walkTokens before and it has always seemed a little mysterious to me. |
Beta Was this translation helpful? Give feedback.
-
|
I think it would be easier to append a link with the renderer. Something like: const renderer: {
paragraph(token) {
const html = Renderer.prototype.paragraph.call(this, token);
return `${html}<a href='customProtocol:///absolute/path/to/file/foo.md:${token.position.start.line}'>Edit</a>\n`
}
};
marked.use({renderer}, markedTokenPosition()); |
Beta Was this translation helpful? Give feedback.
-
|
I added marked-token-position to my program to try it out, and it threw an error message. I've attached the full error message as a file along with a minimal source. What seems odd about the error message is that it says it cannot find a substring even though the substring is present in the original source... The part that throws the error is markdown processed by marked-more-lists. |
Beta Was this translation helpful? Give feedback.
-
|
I was using marked 15.0.6. I've updated to the current version and can compile a document with the marked-token-position extension added. After a little bit of fiddling around, I managed to get it working. Thank you! Is there a way I can over-ride the render for list items, too? That said, I noticed that not every paragraph token gets a start position. It seems that const renderer = {
paragraph(token) {
let html = Renderer.prototype.paragraph.call(this, token);
let edit_link = '';
if (token?.position?.start?.line) {
const value = 'value=' + encodeURIComponent(`${markdownFile}:${token.position.start.line + header_length}`);
const href = 'kmtrigger://macro=Open%20file%20in%20sublime&' + value;
edit_link = `<a class='edit-link' href='${href}'>Edit</a>`;
}
html = html.replace('<p>', `<p>${edit_link}`);
return `${html}\n`
}
}; |
Beta Was this translation helpful? Give feedback.
-
Ya there is a list of renderers you can override in the renderer docs
Can you provide markdown that has a paragraph that doesn't get a position? |
Beta Was this translation helpful? Give feedback.
-
Sure — the attached file is a minimal test case. The HTML generated (ignoring the <body class="">
<p><a class="edit-link" href="kmtrigger://macro=Open%20file%20in%20sublime&value=%2FUsers%2Fjalex%2FDocuments%2FTeaching%2FMarking%2F2024-2025%2FMSc%20Dissertations%2Fminimal.md%3A0"><i class="fa-solid fa-pen-to-square"></i></a>This is a minimal file that shows the problem.</p>
<ol>
<li>
<p>This is a list item that spans
several lines</p>
<p>This causes nested paragraphs inside the list.</p>
</li>
<li>
<p>Another list item</p>
</li>
<li>
<p>And another.</p>
</li>
</ol>
</body>Just for completeness, here's the code I'm using (I changed it a little because the if-condition failed for the valid value of line==0): const renderer = {
paragraph(token) {
let html = Renderer.prototype.paragraph.call(this, token);
let edit_link = '';
if (token?.position?.start?.line !== undefined) {
const value = 'value=' + encodeURIComponent(`${path.resolve(markdownFile)}:${token.position.start.line + header_length}`);
const href = 'kmtrigger://macro=Open%20file%20in%20sublime&' + value;
edit_link = `<a class='edit-link' href='${href}'><i class='fa-solid fa-pen-to-square'></i></a>`;
}
html = html.replace('<p>', `<p>${edit_link}`);
return `${html}`
}
};
import markedTokenPosition from "marked-token-position";
marked.use({ renderer }, markedTokenPosition()); |
Beta Was this translation helpful? Give feedback.
-
|
ah yes the loose list paragraphs are rendered a little differently (from the parser). Lines 109 to 114 in 7b2d430 I'll look into fixing that. The tokens in a list are actually text type not paragraphs since they are sub tokens but I don't know if they need to be. |
Beta Was this translation helpful? Give feedback.
-
|
Here's a short video demo showing both forward and inverse search working between Sublime Text and markdown displayed in Chrome. I did some vibe coding with Claude today and managed to generate a basic Sublime Text plugin that grabs the current line, purifies the text a little bit, and then does a JavaScript injection into Chrome to search through the DOM and scroll the element (if found) into view, with a little animation to highlight the relevant element. Forward.and.inverse.search.in.jmarkdown.mp4 |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Hello,
I was wondering if it was possible for an extension to determine the line number, in the markdown source, where the current 'src' string (given as an argument to the tokenizer) can be found. This would be useful for implementing an inverse-search from the HTML viewer back to the markdown source in an editor.
I've implemented a proof of concept by defining an extension which matches syntax of the form "[Edit N]" (where N is an integer) in the source and turns it into a hyperlink of the form:
And I've configured my Mac so that when the link is clicked on in a web browser, Sublime Text opens the specified file at line N. So that all works great.
Right now, I have to manually type "[Edit N]", explicitly specifying the line number, wherever I want the inverse search functionality to be provided. It would be nice to automate this, but that requires being able to determine the line number in the markdown source from the extension.
Any help would be much appreciated!
Jason
Beta Was this translation helpful? Give feedback.
All reactions