Hmm I'm actually not sure what are you trying to achieve here? The new test ("Encoding using internal modules: utf8 with surrogates in separate chunks") is OK, but to make it work, we need to change the library itself (i.e. adjust the utf8 codec), not adjust back the inputs into a single chunk in the formatSurrogate function in tests.
Hmm I'm actually not sure what are you trying to achieve here? The new test ("Encoding using internal modules: utf8 with surrogates in separate chunks") is OK, but to make it work, we need to change the library itself (i.e. adjust the utf8 codec), not adjust back the inputs into a single chunk in the
formatSurrogate
function in tests.