On 21/Jul/11 14:28, Matus UHLAR - fantomas wrote: > On 19.07.11 19:53, Alessandro Vesely wrote: >>Yes, _was_ until version 57.1, announced on 14 October 2007 (The line >>saying "Invoke Courier filters after closing all control files.") > > Hmmm, does that apply for data file too? ;-) > > However this will be what I missed to find - I remember once I've read > in docs that a filter must not alter the files, but I wasn't able to > find that later. > > Is there any recommended way to alter the data file? > (e.g. create new D* in tmp, move it to msgs/ ?)
I didn't find any. The file is already in some tmp, so I concatenate the filter name and its (forked) process id to the data filename, e.g. /var/courier/localstatedir/tmp/131126/D6144051zdkimfilter2766 After rewriting, I rename (mv) such name to the original data filename. Should errors occur, I'd try and delete the rewritten file (never happened.) > It currently seems to cause courier to drop the mail (hmm, how?) The practice used to consist in marking all recipients as delivered, so that courier won't act on the message. > and pipe modified mail to sendmail. I'd be satisfied if it said courier > to just accept the mail "200 Ok" but current drweb-courier seems to lack > this functionality now. Without sources, you had better have it done by the upstream developer. As a workaround, you could insert a wrap-filter that calls drweb-courier rather than having it called by courier directly. The wrapper could pass a dummy control file to drweb, and thus control the destiny of a message. You'd need a dummy sendmail as well... Is that worth? Why doesn't drweb at least publish an API? -- ------------------------------------------------------------------------------ 5 Ways to Improve & Secure Unified Communications Unified Communications promises greater efficiencies for business. UC can improve internal communications as well as offer faster, more efficient ways to interact with customers and streamline customer service. Learn more! http://www.accelacomm.com/jaw/sfnl/114/51426253/ _______________________________________________ courier-users mailing list [email protected] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users
