Bug: conversion of literals can replace the wrong content

[lehni]

After having worked a bit with this part of the code, it just occurred to me that there could be this theoretical scenario where the use of replace() will replace the wrong part of the content, because it's run on the full result, not just the current element:

Input

p
  | Just some text mentioning
  code div#id
  | or perhaps it talks about
  code div.class
div(id="id")
div(class="class")

Expected Output

p
  | Just some text mentioning
  code div#id
  | or perhaps it talks about
  code div.class
#id
.class

Actual Output

p
  | Just some text mentioning
  code #id
  | or perhaps it talks about
  code .class
div#id
div.class

I have ideas for how to fix this, will work on that next.

[Shinigami92]

Just in case, if you create a new PR, could you now use directly the repo instead of the fork? Then I can also help work and it's easier for me to checkout and review and also resolve merge conflicts 😄

[lehni]

Yes I was thinking already that I should have done that yesterday. Will do so with this one!

[Shinigami92]

Are these two things?

code div#id instead of code #id
and
div#id instead of #id <-- This one was known to me since initial project setup XD but I was too lazy cause a second format resolved it :D

[lehni]

The point is that when you have div#id anywhere in your text content, it'll get replaced if it appears before the actual tag that was supposed to get replaced. This probably makes that clearer:

p
  | Just some text mentioning "div#id"
  | or perhaps it talks about "div.class"
div(id="id")
div(class="class")

The use of code tags in my example above doesn't have anything to do with the problem