Job results to be pass across pipeline about pypette

[csurfer]

@jfunez: You can do that now too, with python queue (https://docs.python.org/3.6/library/queue.html) . Queues in python are thread-safe and so you can push things into queue from one thread and pull from another thread and in this way you can achieve communication between threads.

Does this answer your question or did you have anything else in mind?

[jfunez]

@csurfer not exactly what I have in mind... Maybe this example could help to understand:

In a simple pipeline with three jobs:

J1 -> J2 -> J3

1. In J1, I would collect the data from external source (DB, API, etc)
2. In J2,I would run a data manipulations process with the data returned from J1, then J2 return the data transformed
3. In J3 I would persist the result if the data given by J2 have some specific value

I know it could be possible to achieve it with a queue as you said, but maybe the Job could return an iterator/generator that would be passed to the next Job in the pipeline, the the later could make a decision without having all the data in a single "global" queue. you know the output of one element is the input of the next one

HTH

[ndemou]

@jfunez I love pyfunctional for working exactly that way. seq(generator).map(...).map(...)...

[jfunez]

Hi @ndemou
PyFunctional looks very interesting!!! Thanks for the tip, I'll take a look more deeply for sure!