home *** CD-ROM | disk | FTP | other *** search
- Path: news.ox.ac.uk!news
- From: Nick Christie <nick.christie@oucs.ox.ac.uk>
- Newsgroups: comp.sys.amiga.networking
- Subject: Re: EMAIL Formats
- Date: 12 Feb 1996 20:26:36 GMT
- Organization: Oxford University Computing Services
- Message-ID: <4fo7ps$2g@news.ox.ac.uk>
- References: <19960210.7C1DE70.A64A@ccubb.com>
- NNTP-Posting-Host: nick.oucs.ox.ac.uk
- Mime-Version: 1.0
- Content-Type: text/plain; charset=us-ascii
- Content-Transfer-Encoding: 7bit
- X-Mailer: Mozilla 1.2N (Windows; I; 16bit)
-
- devans@ccubb.com (David Evans) wrote:
- >Basically, the standard email box for a user contains each letter appended to
- >the other. The format of a mail item appears to consist of basically the
- >"header" followed by the text of the item. The second "letter" in a persons
- >mailbox follows the last letters text and begins with another "header".
-
- Many systems adopt the Unix semi-standard mailbox format of appended
- messages, delineated by a line that starts with "From " (note space, not
- colon) and followed, usually, by an email address and date-time stamp.
- The next line is the first line of the message header, which continues
- until the next empty line. Lines up to the next "From " delineator are
- the message body.
-
- For example,
- |From user@pophost Mon, 12 Feb 1996 20:20:00 +0000
- |Received: by foo@bar with SMTP...
- |[more header fields...]
- |
- |The message body starts here and continues until the next "From ".
-
- The "From " bit is a bastard to parse, as that phrase could quite
- conceivably be at the beginning of a line in the message body.
- And you can't check for the following email address or date-time
- stamp either, as those can take many forms, too.
-
- AFAIK, you're stuck with searching for a line that begins with "From ",
- followed a line that starts with one or more alphanumerics followed by
- a colon (the first field name in the header, in other words). Yuck!
-
- If anybody knows a better rule, I'd very interested in hearing about it.
-
- >I have seen some headers have a "byte" count, but its not consistant. I
- >suspect that possibly the message ID has the byte count encrypted but I can't
- >seem to find docs on how to decrypt it if I am correct.
-
- No, the message ID does not have anything encrypted in it.
- It's simply a relatively unique identifier chosen by the emailer.
-
- >Anyone have any ideas?
-
- Occasionally, but most of the time I just play dumb like everybody else.
- :-)
-
- Nick
- --
- Nick Christie, IT Support, mailto://nick.christie@oucs.ox.ac.uk
- Oxford University Computing Services, http://sable.ox.ac.uk/~nick
- 13 Banbury Road, Oxford OX2 6NN. Opinions expressed are my own.
- "There are no more bugs in my program." -- Anon.
-
-
-