E-mail can transmit text in different text formats, including Hypertext Markup Language (HTML), RTF, and plain text. Various software components can impose different text format requirements for content to be stored or displayed to the user, and text format conversion might be necessary to comply with these requirements. For example, an e-mail client might be configured to compose e-mail in HTML, RTF, or plain text, and support dynamically changing formats during composition.
General format conversion can introduce noticeable (and unwanted) changes in content formatting. Therefore, it is imperative not only to aim for high-fidelity conversions to RTF, but also to find a mechanism to recover the content in its original format. This algorithm is used to encode meta information from (or about) the original format (HTML or plain text) within RTF, so that if conversion back to the original form is necessary, it can be very close to the original content.