Parsing HTML to Markdown AST (#847)

* Force mentions to have a space after the #

* Use types for rendering

* Parse HTML

* Add code block support

* Add table support

* Allow starting heading without a space

* Escape relevant plaintext areas

* Resolve many crashes

* Use better matrix id regex

* Don't match . after id

* Don't parse mentions as links

* Add emote support

* Only emit HTML link if necessary

* Implement review changes
This commit is contained in:
ginnyTheCat 2022-09-16 17:51:53 +02:00 committed by GitHub
parent d43e41e1ba
commit a8f374dd43
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
4 changed files with 265 additions and 46 deletions

View file

@ -204,3 +204,27 @@ export function scaleDownImage(imageFile, width, height) {
img.src = imgURL;
});
}
/**
* @param {sigil} string sigil to search for (for example '@', '#' or '$')
* @param {flags} string regex flags
* @param {prefix} string prefix appended at the beginning of the regex
* @returns {RegExp}
*/
export function idRegex(sigil, flags, prefix) {
const servername = '(?:[a-zA-Z0-9-.]*[a-zA-Z0-9]+|\\[\\S+?\\])(?::\\d+)?';
return new RegExp(`${prefix}(${sigil}\\S+:${servername})`, flags);
}
const matrixToRegex = /^https?:\/\/matrix.to\/#\/(\S+:\S+)/;
/**
* Parses a matrix.to URL into an matrix id.
* This function can later be extended to support matrix: URIs
* @param {string} uri The URI to parse
* @returns {string|null} The id or null if the URI does not match
*/
export function parseIdUri(uri) {
const res = decodeURIComponent(uri).match(matrixToRegex);
if (!res) return null;
return res[1];
}