OK, I need to test, looking at the feedparser code, it appears to be capable out outputting each article individually at least. There are other libraries that do output JSON directly as well.
Personally, I would also like to find a better way to handle both incoming feeds and feed merging & republishing. Not been using feeds so much over the last couple of years, partly due to the limited good options for offline aggregation/reading with a complete article not the truncated ones so many feeds produce. And partly because I'm no longer regularly travelling by public transport.
I greatly miss Yahoo! Pipes which made feed manipulation really easy. Really, it should be possible to reproduce some of those features in Node-RED.
Thanks for the concise explanation. I will try to reproduce this flow here in my instance.
In that part, why is Key Variable defined in that way (msg.alf99226....)?
Excuse my sincerity, but I'll follow his example at first simply because he was more clear and illustrative about how trying to make that work with loop nodes. I didn't really understand your explanation using split node.
If there was an example similar to his (with images, a concrete example, etc), but with split node, it would facilitate my comprehension (remember that i'm newbie in Node-RED)...
I imported your demo flow, just changing the fake webhook url to a real one and the rss feed to Lorem RSS (to try to test with a highly changing feed) and got some JSON parse errors like in the image:
Give me a chance! You can't make the current node do it as it isn't written that way. I started to prepare a test node but totally sidetracked myself by setting up a new GitHub repo that will let me more easily do experiments like this. It also has some build tools (which is what took me so long to get working, the whole evening).
Now I have it ready and installed, I can copy the core feedparser node and experiment with it.
Hopefully it will also be useful to other people who want to get started on creating new nodes.
EDIT: I noticed that your feedparser node had a entry point (point at left), that the feedparser node that i installed (node-red-node-feedparser) didn't had.
Got intrigued and searched for feedparser in the Manage Pallete "store". Ends up that there is some "forks" of this node available as well, and some have the left point like the yours.
Ah, I meant to mention that but it was late and I forgot. That was because I wanted to be able to trigger it manually. So I hacked that into my test version. To be honest, it doesn't make a lot of difference. The library that is used actually takes into account a cache so you have to restart node-red anyway if you want the full feed gain.