The discussion on the Feature Request for filtering feed items as they are imported (http://drupal.org/node/160692) advises to build your own processor. However, I have a similar concern as noted in this Feature Request for parser hooks: http://drupal.org/node/162922 - that copy/pasting the code means endless "keep up" maintenance.

Would it make sense to have processor hooks and/or "post processor" hooks in a processor?

Processor hooks means I get a chance to execute code on each node as it is being processed. By "Post Processor" hooks I mean feedapi notifies me when it is done processing and maybe passes me a list of items just added, which I can reprocess.

I could see where both hooks could have good uses that are not mutually exclusive.

Comments

alex_b’s picture

alex_b commented

You can hook into feedapi_item e. g. through the node api.

However, there are two issues here:

1) nodeapi provides no standard way of NOT saving a node (you mentioned this issue: http://drupal.org/node/160692)
2) currently feedapi doesn't provide a standard for passing on the result of one processing stage to the other: http://drupal.org/node/184333

Where would you copy/paste code? Could you explain this a little closer?

Alex

lyricnz’s picture

lyricnz commented

Isn't this a dupe of http://drupal.org/node/162922 ?

aron novak’s picture

aron novak
Primary language Hungarian
Location Hungary, Budapest
commented
Status: Active » Closed (works as designed)

FeedAPI was designed differently.
See this:
http://drupal.org/project/new_aggregator
Here the processors access to each other results, so no need to hook into at the most of the cases.