3
93%
master: 93%

Ran 25 Jul 2020 04:49AM UTC

Files 26

Run time 2s

Badge

Embed ▾

Committed 25 Jul 2020 04:32AM UTC coverage: 64.749% (-21.0%) from 85.79%

Job # JULIA_NUM_THREADS=2 COVERALLS_PARALLEL=true

Build Type

push

travis-ci-com

Committed by

web-flow

Commit Message

Add foldxl, foldxt and foldxd (#369)

This PR adds `foldxl`, `foldxt` and `foldxd` as the new (experimental)
entry points for calling the extended folds defined in Transducers.jl.
The plan is to deprecate `reduce` and `dreduce` in favor of `foldxt`
and `foldxd`.

* Motivation for dedicated entry points

Often times I need to add `Map(identity)` just to dispatch to the
fold defined in Transducers.jl (e.g,
`foldl(TeeRF(min, max), Map(identity), xs)`).  Adding
Transducers.jl-specific entry points like `foldxd` makes
`Map(identity)` unnecessary.  Furthermore, decoupling entry point
from `Base.foldl` let us have a curried version
`itr |> xf |> foldxl(+)` which is quite useful with the new `|>`
pattern.

* Motivation for deprecating `reduce`

* As of #72, `sum(foldable)` calls `foldl(add_sum, foldable)` and
not `reduce(add_sum, foldable)`.  This could be a bit surprising
since `sum` usually calls `reduce`.

* The `Base` implementation of `reduce` is not threaded.  So, using
it for calling multi-threaded implementation can be confusing.

Run Details

889 of 1373 relevant lines covered (64.75%)

16548.79 hits per line

JuliaFolds / Transducers.jl / 1852 / 3
93%
master: 93%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 1852.3 (JULIA_NUM_THREADS=2 COVERALLS_PARALLEL=true)

JuliaFolds / Transducers.jl / 1852 / 3 93% master: 93%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 1852.3 (JULIA_NUM_THREADS=2 COVERALLS_PARALLEL=true)

JuliaFolds / Transducers.jl / 1852 / 3
93%
master: 93%

README BADGES
x