Quick question...
Looking at the end of the pipeline, the last three processes are 'AfterDelivery', 'Acknowledge', and 'ToOutgoing'. From reading the comments in the 'AfterDelivery' and 'Acknowledge' source files I would have thought they should come after 'ToOutgoing'. Am i missing something in my logic?? Also should there be a "," after the 'ToOutgoing' entry below?? (just noticed that now).
On a related note - I'm using a modified version of Personalize.py but I'm not too sure about where in the pipeline I should add it. Suggestions? All it does it split up the delivery so each envelope contains an email for exactly one subscriber.
Thanks in advance
Donal DCU
GLOBAL_PIPELINE = [ # These are the modules that do tasks common to all delivery paths. 'SpamDetect', 'Approve', 'Replybot', 'Moderate', 'MimeDel', 'Hold', 'Emergency', 'Tagger', 'CalcRecips', 'AvoidDuplicates', 'Cleanse', 'CookHeaders', # And now we send the message to the digest mbox file, and to the arch and # news queues. Runners will provide further processing of the message, # specific to those delivery paths. 'ToDigest', 'ToArchive', 'ToUsenet', # Now we'll do a few extra things specific to the member delivery # (outgoing) path, finally leaving the message in the outgoing queue. 'AfterDelivery', 'Acknowledge', 'ToOutgoing', ]
"DH" == Donal Hunt <donal.hunt2@mail.dcu.ie> writes:
DH> Looking at the end of the pipeline, the last three processes
DH> are 'AfterDelivery', 'Acknowledge', and 'ToOutgoing'. From
DH> reading the comments in the 'AfterDelivery' and 'Acknowledge'
DH> source files I would have thought they should come after
DH> 'ToOutgoing'. Am i missing something in my logic??You're probably correct. Technically, I think the To* modules should be the last things in GLOBAL_PIPELINE and the out qrunner's own internal pipeline should call AfterDelivery and Acknowledge. But in practice I doubt it matters.
DH> Also should there be a "," after the 'ToOutgoing' entry
DH> below?? (just noticed that now).It's optional! That's actually one of the cool things about Python; the trailing element in a list (or dict or tuple) can have a comma after it. Makes rearranging the list elements so much nicer for cut-n-paste. The one exception is for tuples with only one element: they /must/ have the trailing comma:
tupleone = (1,)
because (1) is syntactically ambiguous.
DH> On a related note - I'm using a modified version of
DH> Personalize.py but I'm not too sure about where in the
DH> pipeline I should add it. Suggestions? All it does it split
DH> up the delivery so each envelope contains an email for exactly
DH> one subscriber.Are you looking at the current cvs? The Personalize.py module is gone now, SMTPDirect.py having subsumed all its functionality for performance reasons.
-Barry
participants (2)
-
barry＠zope.com -
Donal Hunt