If the message suggests a file name to use for the imbedded part, that name is cleaned of potential problem characters and used for the output file. If the suggested filename includes subdirectories, they will be created as necessary. If the message does not suggest a file name, the names "part1", "part2", etc are used in sequence.
If the imbedded part was preceded with textual information, that information is also written to a file. The file is named the same as the imbedded part, with any filename extension replaced with ".desc".
To decode a MIME message, first save it to a text file. If possible, save it with all headers included. Munpack can decode some MIME files when the headers are missing or incomplete, other files it cannot decode without having the information in the headers. In general, messages which have a statement at the beginning that they are in MIME format can be decoded without the headers. Messages which have been split into multiple parts generally require all headers in order to be reassembled and decoded.
Some LAN-based mail systems and some mail providers (including America Online, as of the writing of this document) place the mail headers at the bottom of the message, instead of at the top of the message. If you are having problems decoding a MIME message on such a system, you need to convert the mail back into the standard format by removing the system's nonstandard headers and moving the standard Internet headers at the top of the message (separated from the message body with a blank line).
There must be exactly one message per file. Munpack cannot deal with multiple messages in a single file, to decode things correctly it must know when one message ends and the next one begins.
To decode a message, run the command:
where "file" is the name of the file containing the message. More than one filename may be specified, munpack will try to decode the message in each file. For more information on ways to run munpack, see the section "OPTIONS" above.