Concatenation/interpolation results in unrecognised character

[nickdunn]

When concatenating two characters together, some combinations result in unrecognised characters being output.

The following test case is a real-world case where I want to append a non-breaking space (\feff) to another character that is used in an icon font.

$char-f: '\66';
$char-g: '\67';

.test-1 {
  content: '#{$char-f}\feff';
}

.test-2 {
  content: '#{$char-g}\feff';
}

// this is broken
.test-3 {
  content: '\feff#{$char-f}';
}

.test-4 {
  content: '\feff#{$char-g}';
}

Output:

@charset "UTF-8";
.test-1 {
  content: "f";
}

.test-2 {
  content: "g";
}

.test-3 {
  content: "󾿿";
}

.test-4 {
  content: "g";
}

(Note that \feff is present in 3 of the 4 examples, but it's invisible).

When \feff is concatenated with \66 a single unrecognised character is returned rather than two.

I've been able to replicate this with the latest https://github.com/sass/node-sass wrapper, as well as pasting the above into http://sass.js.org.

[nickdunn]

Input:

$char: '\66';
content: '\feff#{$char}';

Output (fail):

content: "󾿿";

However if I put a space between them, I get the correct output.

Input:

content: '\feff #{$char-f}';

Output (pass):

content: "f";

Weird that the space gets collapsed?

[nickdunn]

May be related to #1115.

[xzyfer]

Spec added sass/sass-spec#1094

[mgreter]

The space "gets eaten" as described in the css specs:
https://www.w3.org/TR/css-syntax-3/#consume-an-escaped-code-point

Consume as many hex digits as possible, but no more than 5. Note that this means 1-6 hex digits have been consumed in total. If the next input code point is whitespace, consume it as well. Interpret the hex digits as a hexadecimal number. If this number is zero, or is for a surrogate code point, or is greater than the maximum allowed code point, return U+FFFD REPLACEMENT CHARACTER (�). Otherwise, return the code point with that value.

[nickdunn]

I am still able to recreate this issue using libsass 3.5.0.beta.2 (via node-sass 4.5.3)

[nickdunn]

Could this issue be re-opened?

[mgreter]

This is fixed on latest master (or in 3.4.5) after 3.5.0.beta.2 was tagged!

[nickdunn]

Ah, please accept my apologies — I was following the wrong tag. Do you know when this might make it into a release?