document/TranslateProject
2
0
Fork 0
mirror of https://github.com/LCTT/TranslateProject.git synced 2025-01-25 23:11:02 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #10889 from Flowsnow/functional

Browse Source 
[translated by Flowsnow] Getting started with functional programming in Python using the toolz library
This commit is contained in:
Xingyu.Wang 2018-10-26 07:53:03 +08:00 committed by GitHub
commit 97462e4451
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 98 additions and 94 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

94
sources/tech/20181023 Getting started with functional programming in Python using the toolz library.md
View File

@ -1,94 +0,0 @@
translating by Flowsnow
Getting started with functional programming in Python using the toolz library
======
The toolz library allows you to manipulate functions, making it easier to understand and test code.
![](https://opensource.com/sites/default/files/styles/image-full-size/public/lead-images/programming-code-keyboard-laptop-music-headphones.png?itok=EQZ2WKzy)
In the second of a two-part series, we continue to explore how we can import ideas from functional programming methodology into Python to have the best of both worlds.
In the previous post, we covered [immutable data structures][1]. Those allow us to write "pure" functions, or functions that have no side effects, merely accepting some arguments and returning a result while maintaining decent performance.
In this post, we build on that using the `toolz` library. This library has functions that manipulate such functions, and they work especially well with pure functions. In the functional programming world, these are often referred to as "higher-order functions" since they take functions as arguments and return functions as results.
Let's start with this:
```
def add_one_word(words, word):
    return words.set(words.get(word, 0) + 1)
```
This function assumes that its first argument is an immutable dict-like object, and it returns a new dict-like object with the relevant place incremented: It's a simple frequency counter.
However, it is useful only if we apply it to a stream of words and reduce. We have access to a reducer in the built-in module `functools`. `functools.reduce(function, stream, initializer)`.
We want a function that, applied to a stream, will return a frequency count.
We start by using `toolz.curry`:
```
add_all_words = curry(functools.reduce, add_one_word)
```
With this version, we will need to supply the initializer. However, we can't just add `pyrsistent.m` to the `curry`; it is in the wrong order.
```
add_all_words_flipped = flip(add_all_words)
```
The `flip` higher-level function returns a function that calls the original, with arguments flipped.
```
get_all_words = add_all_words_flipped(pyrsistent.m())
```
We take advantage of the fact that `flip` auto-curries its argument to give it a starting value: an empty dictionary.
Now we can do `get_all_words(word_stream)` and get a frequency dictionary. However, how do we get a word stream? Python files are by line streams.
```
def to_words(lines):
    for line in lines:
        yield from line.split()
```
After testing each function by itself, we can combine them:
```
words_from_file = toolz.compose(get_all_words, to_words)
```
In this case, the composition being of just being two functions was straightforward to read: Apply `to_words` first, then apply `get_all_words` to the result. The prose, it seems, is in the inverse of the code.
This matters when we start taking composability seriously. It is sometimes possible to write the code as a sequence of units, test each individually, and finally, compose them all. If there are several elements, the ordering of compose can get tricky to understand.
The `toolz` library borrows from the Unix command line and uses `pipe` as a function that does the same, but in the reverse order.
```
words_from_file = toolz.pipe(to_words, get_all_words)
```
Now it reads more intuitively: Pipe the input into `to_words`, and pipe the results into `get_all_words`. On a command line, the equivalent would look like this:
```
$ cat files | to_words | get_all_words
```
The `toolz` library allows us to manipulate functions, slicing, dicing, and composing them to make our code easier to understand and to test.
--------------------------------------------------------------------------------
via: https://opensource.com/article/18/10/functional-programming-python-toolz
作者：[Moshe Zadka][a]
选题：[lujun9972][b]
译者：[译者ID](https://github.com/译者ID)
校对：[校对者ID](https://github.com/校对者ID)
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
[a]: https://opensource.com/users/moshez
[b]: https://github.com/lujun9972
[1]: https://opensource.com/article/18/10/functional-programming-python-immutable-data-structures

98
translated/tech/20181023 Getting started with functional programming in Python using the toolz library.md Normal file
View File

@ -0,0 +1,98 @@
使用Python的toolz库开始函数式编程
======
toolz库允许你操作函数，使其更容易理解，更容易测试代码。
![](https://opensource.com/sites/default/files/styles/image-full-size/public/lead-images/programming-code-keyboard-laptop-music-headphones.png?itok=EQZ2WKzy)
在这个由两部分组成的系列文章的第二部分中，我们将继续探索如何将函数式编程方法中的好想法引入到 Python中，以实现两全其美。
在上一篇文章中，我们介绍了[不可变数据结构][1]。 这些数据结构使得我们可以编写“纯”函数，或者说是没有副作用的函数，仅仅接受一些参数并返回结果，同时保持良好的性能。
在这篇文章中，我们使用 toolz 库来构建。 这个库具有操作此类函数的函数，并且它们在纯函数中表现得特别好。 在函数式编程世界中，它们通常被称为“高阶函数”，因为它们将函数作为参数，将函数作为结果返回。
让我们从这里开始：
```
def add_one_word(words, word):
    return words.set(words.get(word, 0) + 1)
```
这个函数假设它的第一个参数是一个不可变的类似字典的对象，它返回一个新的类似字典的在相关位置递增的对象：这就是一个简单的频率计数器。
但是，只有将它应用于单词流并做归纳时才有用。 我们可以使用内置模块 `functools` 中的归纳器。 `functools.reduce(function, stream, initializer)`
我们想要一个函数，应用于流，并且能能返回频率计数。
我们首先使用 `toolz.curry` 函数：
```
add_all_words = curry(functools.reduce, add_one_word)
```
使用此版本，我们需要提供初始化程序。 但是，我们不能只将 `pyrsistent.m` 函数添加到 `curry` 函数中中； 因为这个顺序是错误的。
```
add_all_words_flipped = flip(add_all_words)
```
The `flip` higher-level function returns a function that calls the original, with arguments flipped.
`flip` 这个高阶函数返回一个调用原始函数的函数，并且翻转参数顺序。
```
get_all_words = add_all_words_flipped(pyrsistent.m())
```
我们利用 `flip` 自动调整其参数的特性给它一个初始值：一个空字典。
现在我们可以执行 `get_all_words(word_stream)` 这个函数来获取频率字典。 但是，我们如何获得一个单词流呢？ Python文件是行流的。
```
def to_words(lines):
    for line in lines:
        yield from line.split()
```
在单独测试每个函数后，我们可以将它们组合在一起：
```
words_from_file = toolz.compose(get_all_words, to_words)
```
在这种情况下，组合只是使两个函数很容易阅读：首先将文件的行流应用于 `to_words`，然后将 `get_all_words` 应用于 `to_words` 的结果。 散文似乎与代码相反。
当我们开始认真对待可组合性时，这很重要。 有时可以将代码编写为一个单元序列，单独测试每个单元，最后将它们全部组合。 如果有几个组合元素时，组合的顺序可能就很难理解。
`toolz` 库借用了 Unix 命令行的做法，并使用 `pipe` 作为执行相同操作的函数，但顺序相反。
```
words_from_file = toolz.pipe(to_words, get_all_words)
```
Now it reads more intuitively: Pipe the input into `to_words`, and pipe the results into `get_all_words`. On a command line, the equivalent would look like this:
现在读起来更直观了：将输入传递到 `to_words`，并将结果传递给 `get_all_words`。 在命令行上，等效写法如下所示：
```
$ cat files | to_words | get_all_words
```
The `toolz` library allows us to manipulate functions, slicing, dicing, and composing them to make our code easier to understand and to test.
`toolz` 库允许我们操作函数，切片，分割和组合，以使我们的代码更容易理解和测试。
--------------------------------------------------------------------------------
via: https://opensource.com/article/18/10/functional-programming-python-toolz
作者：[Moshe Zadka][a]
选题：[lujun9972][b]
译者：[Flowsnow](https://github.com/Flowsnow)
校对：[校对者ID](https://github.com/校对者ID)
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
[a]: https://opensource.com/users/moshez
[b]: https://github.com/lujun9972
[1]: https://opensource.com/article/18/10/functional-programming-python-immutable-data-structures