Decoded string with html entities does not equal string literal

Question

I tried to implement the solution proposed in this thread to decode a string with html entities, e.g. "foo bar" to "foo bar".

Visually, it seems to work. But my quick Jest test fails:

Expected: "foobar"
Received: "foobar"

  3 | describe('encryption/decodeHtml', () => {
  4 |   it.each([['foo&nbsp;bar', 'foo bar'], ['foo&shy;bar', 'foobar'], ['foo&amp;bar', 'foo&bar']])('should decode html entities', (val, expected) => {
> 5 |     expect(decodeHtml(val)).toEqual(expected);
    |                             ^
  6 |   })
  7 | });
  8 |

A quick Object.is(decodeHtml(' '), ' ') also produces false.

Is there something to JS-Strings I'm not familiar with?

`decodeHtml('foo bar').split("").forEach(c => console.log(c, c.charCodeAt()));` — Andreas, Sep 08 '21 at 08:45
You're right, I forgot about the underlying char codes. I took value equality too literally. — rsmidt, Sep 08 '21 at 08:55

score 0 · Accepted Answer · answered Sep 08 '21 at 08:54

As pointed out by Andreas in the comments, I forgot about the byte representation of the string.

Look at this example:

toBytes('foo bar') -> Uint8Array(7) [102, 111, 111, 32, 98, 97, 114]
toBytes(decodeHtml('foo&nbsp;bar')) -> Uint8Array(8) [102, 111, 111, 194, 160, 98, 97, 114]

In hindsight it pretty obvious because the breaking space and the non breaking space are (of course) different characters.

Decoded string with html entities does not equal string literal

1 Answers1