mdast utility to selectively hide and reveal mdast nodes
mdast-util-hidden
This is a small mdast utility for hiding nodes from transformers by mapping them to Hidden pseudo-node instances.
This is a low level project used by remark-ignore and remark-renumber-references.
Due to the nature of the unified ecosystem, this package is ESM only and cannot be
require'd.
To install:
npm install mdast-util-hiddenSuppose we have the following Markdown file example.md:
# Hello
Some _emphasis_, **importance**, and `code`.
# GoodbyeRepresented by the following AST (with position data omitted for brevity):
{
"type": "root",
"children": [
{
"type": "heading",
"depth": 1,
"children": [{ "type": "text", "value": "Hello" }]
},
{
"type": "paragraph",
"children": [
{ "type": "text", "value": "Some " },
{
"type": "emphasis",
"children": [{ "type": "text", "value": "emphasis" }]
},
{ "type": "text", "value": ", " },
{
"type": "strong",
"children": [{ "type": "text", "value": "importance" }]
},
{ "type": "text", "value": ", and " },
{ "type": "inlineCode", "value": "code" },
{ "type": "text", "value": "." }
]
},
{
"type": "heading",
"depth": 1,
"children": [{ "type": "text", "value": "Goodbye" }]
}
]
}Then running the following JavaScript:
import fs from 'node:fs';
import { unified } from 'unified';
import remarkParse from 'remark-parse';
import { visit, SKIP } from 'unist-util-visit';
import { hide } from 'mdast-util-hidden';
const doc = fs.readFileSync('example.md');
const tree = unified().use(remarkParse).parse(doc);
visit(tree, 'heading', (node, index, parent) => {
if (index !== undefined && parent !== undefined) {
// β¦ Other stuff
hide({ nodes: [node], index, parent });
// Do not traverse node, continue with the next node at index + 1
return [SKIP, index + 1];
}
});
console.dir(tree, { depth: null });Yields:
{
"type": "root",
"children": [
{
"type": "hidden",
"hiddenChildren": [
{
"type": "heading",
"depth": 1,
"children": [{ "type": "text", "value": "Hello" }]
}
]
},
{
"type": "paragraph",
"children": [
{ "type": "text", "value": "Some " },
{
"type": "emphasis",
"children": [{ "type": "text", "value": "emphasis" }]
},
{ "type": "text", "value": ", " },
{
"type": "strong",
"children": [{ "type": "text", "value": "importance" }]
},
{ "type": "text", "value": ", and " },
{ "type": "inlineCode", "value": "code" },
{ "type": "text", "value": "." }
]
},
{
"type": "hidden",
"hiddenChildren": [
{
"type": "heading",
"depth": 1,
"children": [{ "type": "text", "value": "Goodbye" }]
}
]
}
]
}Note
The elements of hiddenChildren will, if later revealed, have their
position data stripped off using unist-util-remove-position (making
them generated nodes). This is because position data can be invalidated
at any time due to it being impossible to know how later plugins will
manipulate the AST. Missing position data will not crash any remark plugins
that properly follow the unist spec.
And running the following JavaScript:
import fs from 'node:fs';
import { unified } from 'unified';
import remarkParse from 'remark-parse';
import { visit, SKIP } from 'unist-util-visit';
import { hide, visitAndReveal, reveal } from 'mdast-util-hidden';
const doc = fs.readFileSync('example.md');
const tree = unified().use(remarkParse).parse(doc);
visit(tree, 'heading', (node, index, parent) => {
if (index !== undefined && parent !== undefined) {
hide({ nodes: [node], index, parent });
return [SKIP, index + 1];
}
});
// LATER:
visit(tree, 'hidden', (node, index, parent) => {
if (index !== undefined && parent !== undefined) {
reveal({ nodes: [node], index, parent });
// β¦ Other stuff
// Do not traverse node, continue with the node now at index (NOT index + 1)
return [SKIP, index];
}
});
// OR:
visitAndReveal({ tree });
// OR:
visitAndReveal({
tree,
visitor: (node, index, parent) => {
// reveal is called for you
// β¦ Other stuff
// returns [SKIP, index] unless this function returns a defined value
}
});
console.dir(tree, { depth: null });Yields the original AST with the hidden headings restored (but stripped of their
position data):
{
"type": "root",
"children": [
{
"type": "heading",
"depth": 1,
"children": [{ "type": "text", "value": "Hello" }]
},
{
"type": "paragraph",
"children": [
{ "type": "text", "value": "Some " },
{
"type": "emphasis",
"children": [{ "type": "text", "value": "emphasis" }]
},
{ "type": "text", "value": ", " },
{
"type": "strong",
"children": [{ "type": "text", "value": "importance" }]
},
{ "type": "text", "value": ", and " },
{ "type": "inlineCode", "value": "code" },
{ "type": "text", "value": "." }
]
},
{
"type": "heading",
"depth": 1,
"children": [{ "type": "text", "value": "Goodbye" }]
}
]
}When using mdast-util-hidden in your own remark plugin, it's generally a good
idea to mark your Hidden nodes with a unique symbol and ensure that you only
call reveal and visitAndReveal on the Hidden nodes marked with said
symbol. Otherwise, if someone is using your plugin and remark-ignore
concurrently, it could be the case that your plugin accidentally reveals nodes
that were hidden by remark-ignore, causing erroneous transformations against the
user's wishes.
An example of using a unique symbol to mark ownership of Hidden nodes can be
found in the remark-renumber-references source code.
Detailed interface information can be found under docs/.
Hidden
interface Hidden extends Node {
type: 'hidden';
hiddenChildren: MdastContent[];
}Hidden (Node) represents an abstract interface in mdast containing mdast content that is hidden (and thus protected) from other transformers.
Hidden nodes are always generated, cannot be serialized to markdown, and cannot be derived from markdown directly.
- remark-ignore β use comments to exclude one or more nodes from transformation.
- remark-renumber-references β renumber numeric reference-style link ids
contiguously starting from
[1]. - rehype-ignore β ignore content display via HTML comments.
Further documentation can be found under docs/.
This is an ESM-only package built by Babel for use in Node.js
versions that are not end-of-life. For TypeScript users, this package supports
both "Node10" and "Node16" module resolution strategies.
Expand details
That means ESM source will load this package via import { ... } from ... or
await import(...) and CJS source will load this package via dynamic
import(). This has several benefits, the foremost being: less code
shipped/smaller package size, avoiding dual package
hazard entirely, distributables are not
packed/bundled/uglified, and a drastically less complex build process.
The glaring downside, which may or may not be relevant, is that CJS consumers
cannot require() this package and can only use import() in an asynchronous
context. This means, in effect, CJS consumers may not be able to use this
package at all.
Each entry point (i.e. ENTRY) in package.json's
exports[ENTRY] object includes one or more export
conditions. These entries may or may not include: an
exports[ENTRY].types condition pointing to a type
declaration file for TypeScript and IDEs, a
exports[ENTRY].module condition pointing to
(usually ESM) source for Webpack/Rollup, a exports[ENTRY].node and/or
exports[ENTRY].default condition pointing to (usually CJS2) source for Node.js
require/import and for browsers and other environments, and other
conditions not enumerated here. Check the
package.json file to see which export conditions are
supported.
Note that, regardless of the { "type": "..." } specified in
package.json, any JavaScript files written in ESM
syntax (including distributables) will always have the .mjs extension. Note
also that package.json may include the
sideEffects key, which is almost always false for
optimal tree shaking where appropriate.
See LICENSE.
New issues and pull requests are always welcome and greatly appreciated! π€© Just as well, you can star π this project to let me know you found it useful! βπΏ Or buy me a beer, I'd appreciate it. Thank you!
See CONTRIBUTING.md and SUPPORT.md for more information.
See the table of contributors.
