https://github.com/tk3369/functionalpipes.jl

Functional pipe utility
https://github.com/tk3369/functionalpipes.jl

functional-programming julia

Last synced: 1 day ago
JSON representation

Functional pipe utility

• Host: GitHub
• URL: https://github.com/tk3369/functionalpipes.jl
• Owner: tk3369
• License: mit
• Created: 2019-10-22T06:50:24.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2020-12-13T04:49:50.000Z (almost 4 years ago)
• Last Synced: 2024-07-30T18:34:41.268Z (4 months ago)
• Topics: functional-programming, julia
• Language: Julia
• Homepage:
• Size: 6.84 KB
• Stars: 2
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# FunctionalPipes

_Note: this package is experimental.  Similar packages include Lazy.jl, 

Hose.jl, and MagicUnderscores.jl._

This package provides a simple facility to build functional pipes

that support functions taking more than a single argument.  It also

allows the results from upstream be placed in any argument position.

## Motivation 

The pipe operator `|>` works well as long as the downstream

function takes a single argument.  Otherwise, it becomes quite

awkward to make anonymous functions to work around the issue.

For example:

```julia

[1,2,3] |> v -> filter(x -> x > 1, v) |> sum      # 5

```

## Features

Using the `pipe` function, we can build a computation pipe 

as usual:

```julia

pipe([1,2,3], sum, iseven)  # true

```

However, we can pass prior results to a function in any specific

argument position.  The special token is 🔥 (or `:token`).

```julia

pipe(

    [1,2,3],

    dispatch(filter, x -> x > 1, 🔥),

    sum

)

```

In fact, you can even have multiple tokens:

```julia

pipe(3, dispatch(+, 🔥, 🔥))     # 6

```

If the dispatched function returns `nothing`, then it carries over

the prior result.  This would be useful for functions that only

have side effects.

```julia

julia> pipe(

            1:10,

            collect,

            dispatch(filter, isodd, 🔥),

            dispatch(println, "values: ", 🔥),

            sum,

            dispatch(println, "total: ", 🔥)

        )

values: [1, 3, 5, 7, 9]

total: 25

25

```