Struct smtp_message::DataUnescaper

pub struct DataUnescaper { /* private fields */ }

Helper struct to unescape a data stream.

Note that one unescaper should be used for a single data stream. Creating a DataUnescaper is basically free, and not creating a new one would probably lead to initial \r\n being handled incorrectly.

Implementations

impl DataUnescaper

pub fn new(is_preceded_by_crlf: bool) -> DataUnescaper

Creates a DataUnescaper.

The is_preceded_by_crlf argument is used to indicate whether, before the first buffer that is fed into unescape, the unescaper should assume that a \r\n was present.

Usually, one will want to set true as an argument, as starting a DataUnescaper mid-line is a rare use case.

pub fn unescape(&mut self, data: &mut [u8]) -> DataUnescapeRes

Unescapes data coming from an EscapedDataReader.

This takes a data argument. It will modify the data argument, removing the escaping that could happen with it, and then returns a DataUnescapeRes.

It is possible that the end of data does not land on a boundary that allows yet to know whether data should be output or not. This is the reason why this returns a DataUnescapeRes. The returned value will contain:

• .written, which is the number of unescaped bytes that have been written in data — that is, data[..res.written] is the unescaped data, and
• .unhandled_idx, which is the number of bytes at the end of data that could not be handled yet for lack of more information — that is, data[res.unhandled_idx..] is data that should be at the beginning of the next call to data_unescape.

Note that the unhandled data’s length is never going to be longer than 4 bytes long (“\r\n.\r”, the longest sequence that can’t be interpreted yet), so it should not be an issue to just copy it to the next buffer’s start.