Markdown variables consume too many characters if parentheses are nested with whitespace before final )

[5e-Cleric]

Not sure what is causing this issue, but it is a bug of some sort.

Some combination of links and parentheses inside tables merges cells together, sometimes splitting words.

the following code will work as a demonstration:

|  title 1  | title 2 | title 3 | title 4|
|-----------|---------|---------|--------|
|[foo](bar) |  Ipsum  |    )    |   )    |

results in this:

[5e-Cleric]

@Gazook89 has reportedly tried this combination in other markdown editors without being able to replicate this behaviour, therefore, we are drawn to the conclusion the bug must come from our own code, most likely our table handling code in \shared\naturalcrit\markdown.js.

Pointed out by @calculuschild, the error might be new and related to variables.

[Gazook89]

I think the issue is a bad regex.

in markdown.js:

	const inlineDefRegex  = /([!$]?\[((?!\s*\])(?:\\.|[^\[\]\\])+)\])\(([^\n]+)\)/;                         //Matches 8, 9[10](11)

should be changed to

	const inlineDefRegex  = /([!$]?\[((?!\s*\])(?:\\.|[^\[\]\\])+)\])\(([^\n)]+)\)/;                         //Matches 8, 9[10](11)

which adds a ) in the disallowed characters in the third matching group (([^\n)]+))

[calculuschild]

Discussed on Gitter, the real issue is this:

As easy as let content = match[11] ? match[11] : null;, just removing the trim and regex substitution, since we need to do the parenthesis matching to occur before the trim.

homebrewery/shared/naturalcrit/markdown.js

Line 599 in dd82a1b

let content = match[11] ? match[11].trim().replace(/\s+/g, ' ') : null; // Trim edge spaces and shorten blocks of whitespace to 1 space

The trim will then happen afterward here. Note that the if(i > -1) will always be true since there is no opportunity for i to be equal or less than -1. That was leftover from an earlier piece of code, and so we can just pull the inside of the if out.

homebrewery/shared/naturalcrit/markdown.js

Lines 615 to 618 in dd82a1b

	if(i > -1) {
	combinedRegex.lastIndex = combinedRegex.lastIndex - (content.length - i);
	content = content.slice(0, i).trim().replace(/\s+/g, ' ');
	}