* Baurjan Ismagulov <[EMAIL PROTECTED]> [2002-08-28 10:18]: > I've looked at procmail a while ago for body > translation, but couldn't find a working solution. > > 1. All three inboxes I use are IMAP; how can I use procmail with it? I > can define a macro that would do what I want, but how would it take > the input, and where it should store the output?
a local process (procmail) cannot work on remote data (mail on IMAP server). so i expect it just won't work. i'd use fetchmail to download only the necessary messages and then filter them locally - with procmail. > 2. Matching the messages by From: or To: isn't reliable since > the same person can send messages in different encodings. > E.g., message from a web-based client can be encoded in koi8-r > and flagged as iso-8859-1, and message from a windows-based > MUA can be encoded in windows-1251 and flagged as iso-8859-1. > Let alone rare cases where the same person can send mail > either in koi8-r or in iso-8859-9, both tagged as us-ascii. procmail is not a repair tool for programs which incorrectly declare data. and yet you can use it to change data - if you know how. > 3. What to do with raw 8-bit data in headers? raw 8bit data in headers is not allows, i think. rfc822 and all that. do you receive any such mails? (i suppose you do because otherwise it would not make much sense.) > That is why I think it would be much better to have manual charset > override for headers, just like the one we have for bodies. > IMHO, this functionality should be implemented in MUA. this would only work around the problems created by bad mail clients, wouldn't it? > However, mutt is the best MUA > with regard to bad input tolerance. I think it > lacks this feature, which is already implemented > in the source, but is inapplicable to headers. I somehow the prospect of mutt becoming a repair tool for the brokenness of other mailers does not make me happy. i'd rather tell other users to not use the broken mailers at all and thereby let them die. life's too short for bad software. Sven
