Open dbnski opened 5 years ago
It was just something quick I put together while trying to use your library in a PoC, but I see it needs some polishing to be able to pass through tests. If you are interested in merging this patch, I can update the PR when I have some spare time.
The original code uses signed
char
type for variables that hold the character being processed. It breaks 8-bit ASCII characters used in many European languages, but also prevents the library from correctly handling unescaped UTF8 characters. The attached patch addresses both these problems and further improves support for UTF8-encoded strings. I am aware that updating all methods wasn't necessary to make this work, but I thought perhaps it was a good idea to still do so for consistency.