Open singpolyma opened 12 years ago
It seems like the primary underlying parser should be on ByteString
for this reason. MIME seems to be about bytes and not simply about text. I wonder if this might also improve performance, as emailparse
currently runs an order of magnitude faster.
If the content is uses only 7bit encodings and is embedded as ASCII, then getting it as a
String
is easy, but if it's usingContent-Transfer-Encoding: 8bit
then it is necessary to operate onByteString
until theContent-Type
is known.