Hi 👋

Previously, I opened #42 to correctly unescape whitespace HTML entity codes.
However, that the approach already present turned out to be not very scalable, as many valid entities are still not decoded correctly.

For example:

ï  ->  ï
°  ->  °

These entities are automatically decoded in GitHub’s markdown preview, but not in marked-react, which leads to inconsistent rendering in our application (some entities decoded, others not).

Proposed solution

Use html-entities, a lightweight package without dependencies, to safely decode all valid HTML entities.

Example

Input:

Café – temperature: 23°C

Expected output:

Café – temperature: 23°C

Actual output (current):

Café – temperature: 23°C

Would you consider adding this improvement?
I’d be happy to open a PR if this approach sounds good.