add mergeLimit function as mentioned in #374 #375
Conversation
|
This breaks backpressure. You should be using modified version of the current implementation of You all so need to add a |
| * | ||
| * _([txt, md, js]).mergeLimit(2); | ||
| * // => contents of foo.txt, bar.txt, baz.txt and bosh.js in the order | ||
| * they were read, but bosh is not read until the others have comleted |
There was a problem hiding this comment.
Typo in "comleted". Also, this wording suggests bosh won't be read until both bar and baz have completed. I suggest "but bosh is not read until either foo/bar or baz has finished.
|
Finally, let's call this |
dayoadeyemi
force-pushed
the
master
branch
2 times, most recently
from
649928f to
44464fc
Compare
|
Yes that did break back-pressure, Instead of modifying the merge function I've made it so that there is just a transform function before the merge which prevents the stream from pushing any values until something from the current working set has ended. |
| push(err, x); | ||
| // console.log('start', x.id) | ||
| x.once('end', function(){ | ||
| // console.log('end', x.id) |
There was a problem hiding this comment.
Your indentation for x.once is off. Also, I think you want to do index-- in the end handler, right?
Oh, nice. That's a much better idea. |
|
Yea if I don't index-- when the substream ends it will end up only allowing 1 process to run at a time. I've added that in and renamed index to processCount so that it is more descriptive. |
| * | ||
| * @id mergeLimit | ||
| * @section Higher-order Streams | ||
| * @name Stream.mergeWithLimit() |
There was a problem hiding this comment.
Call this @name Stream.mergeWithLimit(n).
The @id field should be @id mergeWithLimit.
You also need to add a @param n field for the parameter.
|
I think this is ready as soon as you fix a few issues with the documentation. |
|
A couple of stray |
|
Cool. Commented Thanks, @dayoadeyemi for the PR! |
Add mergeLimit function. Resolves #374.
No description provided.