Pasting two subsequent empty lines removes the second one about scribe HOT 4 OPEN

You can see here how we convert new line characters into HTML.

I think the fix would be a case of making the regular expression non-greedy, i.e.:

function oldConvert(html) {
  return '<p>' + html.replace(/\n([ \t]*\n)+/g, '</p><p>').replace(/\n/g, '<br>') + '</p>';
}
function newConvert(html) {
  return '<p>' + html.replace(/\n([ \t]*\n)+?/g, '</p><p>').replace(/\n/g, '<br>') + '</p>';
}

oldConvert('1\n\n2\n\n\n3');
// => "<p>1</p><p>2</p><p>3</p>"
oldConvert('1\n\n2\n\n\n\n3');
// => "<p>1</p><p>2</p><p>3</p>"
newConvert('1\n\n2\n\n\n3');
// => "<p>1</p><p>2</p><p><br>3</p>"
newConvert('1\n\n2\n\n\n\n3');
// => "<p>1</p><p>2</p><p></p><p>3</p>"

My regular expression knowledge isn’t the best, so cc'ing @theefer in on this one.

from scribe.

Bump /cc @theefer

from scribe.

The first step is to determine what the expected result should be.

1
2

3

would give

<p>1<br>2</p>
<p>3</p>

as now.

1
2


3

would give

<p>1<br>2</p>
<p><br></p>
<p>3</p>

?

And

1
2



3

would give

<p>1<br>2</p>
<p><br></p>
<p><br></p>
<p>3</p>

?
(note the dummy <br> to make the paragraph selectable; the paste handler may not need to care about this if a subsequent formatter normalises <p></p> to contain a <br>)

In that case, it sounds like we want to replace:

\n by <br>
\n\n by </p><p>
\n\n\n by </p><p></p><p>
\n\n\n\n by </p><p></p><p></p><p>

etc?

Can't think of an easy way to do this except using a function as replace argument:

"1\n2\n\n3\n\n\n4".replace(/\n\n+/g, function(newlines){
  return Array(newlines.length).join('</p><p>'); // fancy repeating pattern, thanks StackOverflow
}).replace(/\n/g, '<br>');
// "1<br>2</p><p>3</p><p></p><p>4"

from scribe.

Note that this does not correctly work if there is whitespace on the empty lines, something the original pattern accounted for. My solution could be extended to work with that somehow I'm sure.

from scribe.