Add promise <-> monad bridge functions

[char0n]

resolveP, allP, tapP, thenP, catchP, thenCatchP, rejectP, raceP, reduceP, isThenable, isPromise, tryP

[char0n]

resolveP = bind(Promise.resolve, Promise);

[rockymadden]

Something like https://github.com/kirillrogovoy/promised-pipe might also be useful.

[char0n]

@rockymadden isn't pipeP what you're looking for ?

[rockymadden]

If all functions are return promises, yes. The promised-pipe brings the ability to intermix, at any point, sync and async functions.

[char0n]

pipeP does the same. The only requirement is that first function in the pipe must be promise generator (returns promise).

const { add, subtract }  = require('ramda');

const resolveP = Promise.resolve.bind(Promise);
const asyncAdd = val => resolveP(val).then(add(1));

pipeP(
  resolveP,
  add(1),
  subtract(1),
  asyncAdd,
  add(2),
  subtract(1)
)(1);

[rockymadden]

Nice, hmm yeah ignore me then. Thanks! BTW, I see some of what you are laying out here mentioned through out various Ramda issues on the topic of Promises (e.g. ramda/ramda#1869 (comment)). Lovely idea.

[char0n]

NP

[rockymadden]

Thoughts on your resolveP function (or R.bind(Promise.resolve, Promise)) being made into a standard function in adjunct (e.g. resolveP or ofP)? Seems to be generally usable to kick things off as a Promise.

EDIT: Ignore, I see it now as the first function in the list above.

[char0n]

Yep, the plan is to implement it. Checkout out the first comment of this issue. It is the first function to be implemented for this issue ;]

[char0n]

You can provide a pull request for it if you wish and become a contributor. ramda-adjunct ecosystem is ready to receive PR's and I always welcome any help. What do you think ?

[rockymadden]

Yep, love it. I've had a subset of this library internally for a while. Was reading through the Ramda wiki and saw ramda-adjunct, which was exactly what I myself was doing.

[char0n]

Yep, I guess everybody is doing it. This library has a potential saving a lot of development time to a lot of people. Check this article to find out more.

Great, check out the CONTRIBUTING.md and you can start right away.

[char0n]

Further inspirations: https://github.com/krainboltgreene/unction.js?files=1

[char0n]

https://stackoverflow.com/questions/27746304/how-do-i-tell-if-an-object-is-a-promise

[Undistraction]

@char0n You OK if I bump this to 2.7.0?

[char0n]

Bump it up or remove the milestone entirely. Have no usecases for the rest of them right now.

[Undistraction]

@char0n When you get a chance maybe move remaining to separate issues.

[char0n]

Yeah me or anybody else can do that.

Note for me and others: when doing it reference this issue in the child issues just for reference and then close this one.

[guillaumearm]

since #676, additional candidates are :

• mapP
• filterP
• rejectP

IMO, we need to talk about parallel and series behavior, isn't ?

[char0n]

Yes I think we should and I'm very interested in this discussion. But I'm still thinking about this in a context of already implemented reduceP. Does it make sense to have reduceP work in a parallel mode ?

Can you provide example of mapP being parallel and serial ?

[char0n]

Created separate issue for raceP

[char0n]

created separate issue for catchP

[char0n]

created separate issue for thenCatchP

[char0n]

Not going to implement tapP for now.