Enhancements

[nitishkansal]

1. Few enhancements in Regex used to parse the header values.
2. Get the filename using python email module instead of getting it from content id or content disposition as we have seen some email services don't follow the standard and content id is not present at all in the attachment part.
3. Stopped some noise in error logging as that error was already handled but it was still adding some unnecessary noise in the logs, Ideally there should be logger setting which can be used to suppress some specific log levels.
4. Change the way how attachment is identified as we have seen some email services send the email with html and plain text with content id and they are also treated as attachment but they should be treated as email body.
5. Change the way how payload is handled, because when get_payload() decodes 7bit or 8bit or without any Content-Transfer-Encoding then it encodes them with raw-unicode-escape which leave the message unreadable and makes it some gibberish characters. so when we call get_payload() with decode, we are again checking if Content-Transfer-Encoding was one of the culprits then we decode them again with raw-unicode-escape so that we get the message as it was sent before sending it to ported_string().

We checked the test suite as well added in the library but that test suite just check if email parsing is working fine or not but it does not check if encoding was maintained or not. So test suite is working but emails were left with gibberish characters.

PS: We have been using this library for almost 8-9 months now and it has been great to use. But we were still facing encoding issues so we had to investigate and made some changes to library and we have been using these updates for almost 6-7 months now and now our encoding complaints are reduced by 95%-96% Main problems we were facing with unicode characters and different Content-Type, which is mostly resolved now.

[The-Hidden-Hand]

@fedelemantuano
Hello, are you accepting pull requests?

[fedelemantuano]

The PR look good. I'm looking inside it.