Paste_Image.png
本文会介绍下Eloquent\Collection，这个是什么呢？这是我们平常使用Eloquent中get,all返回的结果集。
collection缘来
我们首先来看一段简单的代码：
$books = Book::all();
$titles = [];
foreach ($books as $book){
if ($book->pages_count > 8){
$titles[] = $book->title;
}
}
这段代码意图其实非常明确，就是获取超过8的书名，再看下面一段代码：
$titles = [];
foreach ($books as $book){
if ($book->publisher_id == 2){
$titles[] = $book->title;
}
}
此处是获取作者是2的书名，所有这些代码都有同样的loop逻辑，我们完全可以抽取出来，于是就有了下面的函数：
function map($input, $func)
{
$result = [];
foreach ($input as $each)
{
$result[] = $func($each);
}
return $result;
}
map($books, function($book){
if ($book->publisher_id == 2){
return $book->title;
}
});
这只是展示了一个简单的map模式，还有其他更多方便的集合操作方法，我们将其抽取出来，于是就出现了Collection，网上有个讲Collection的课程，不过太贵了，买不起。
其实Collection的总体思想感觉就是函数式编程，Tell, Don’t Ask，客户端在使用上不再是想着怎么做how，而是想着what to do，一直有个神一样存在的系列文章没去读，今天看到collection的文章，有了冲动去看的，文章地址：Category Theory for Programmers: The Preface，等最近看完orm系列就去看这个的。我们还是接着讲collection。
collection使用
在使用collection的原则上，我们遵守当代码出现loop的时候，我们就停下来想下，是否可以通过collection来解决。
first
三种使用方式
public function testFirstReturnsFirstItemInCollection()
{
$c = new Collection(['foo', 'bar']);
$this->assertEquals('foo', $c->first());
}
public function testFirstWithCallback()
{
$data = new Collection(['foo', 'bar', 'baz']);
$result = $data->first(function ($value) {
return $value === 'bar';
});
$this->assertEquals('bar', $result);
}
public function testFirstWithCallbackAndDefault()
{
$data = new Collection(['foo', 'bar']);
$result = $data->first(function ($value) {
return $value === 'baz';
}, 'default');
$this->assertEquals('default', $result);
}
last
和first使用方式相同
public function testLastReturnsLastItemInCollection()
{
$c = new Collection(['foo', 'bar']);
$this->assertEquals('bar', $c->last());
}
public function testLastWithCallback()
{
$data = new Collection([100, 200, 300]);
$result = $data->last(function ($value) {
return $value < 250;
});
$this->assertEquals(200, $result);
$result = $data->last(function ($value, $key) {
return $key < 2;
});
$this->assertEquals(200, $result);
}
public function testLastWithCallbackAndDefault()
{
$data = new Collection(['foo', 'bar']);
$result = $data->last(function ($value) {
return $value === 'baz';
}, 'default');
$this->assertEquals('default', $result);
}
map
map是对loop的抽离，对于集合中每个元素做完操作后，再返回新的元素。
public function testMap()
{
$data = new Collection(['first' => 'taylor', 'last' => 'otwell']);
$data = $data->map(function ($item, $key) {
return $key.'-'.strrev($item);
});
$this->assertEquals(['first' => 'first-rolyat', 'last' => 'last-llewto'], $data->all());
}
each
遍历元组进行操作，不返回元素操作后的结果，当遇到返回false的时候，结束遍历。
public function testEach()
{
$c = new Collection($original = [1, 2, 'foo' => 'bar', 'bam' => 'baz']);
$result = [];
$c->each(function ($item, $key) use (&$result) {
$result[$key] = $item;
});
$this->assertEquals($original, $result);
$result = [];
$c->each(function ($item, $key) use (&$result) {
$result[$key] = $item;
if (is_string($key)) {
return false;
}
});
$this->assertEquals([1, 2, 'foo' => 'bar'], $result);
}
filter
遍历集合，只将符合条件的留下，集合中元素的性质不会变，如果集合中是product，返回的也是product，不会像map那样，返回price
public function testFilter()
{
$c = new Collection([['id' => 1, 'name' => 'Hello'], ['id' => 2, 'name' => 'World']]);
$this->assertEquals([1 => ['id' => 2, 'name' => 'World']], $c->filter(function ($item) {
return $item['id'] == 2;
})->all());
$c = new Collection(['', 'Hello', '', 'World']);
$this->assertEquals(['Hello', 'World'], $c->filter()->values()->toArray());
$c = new Collection(['id' => 1, 'first' => 'Hello', 'second' => 'World']);
$this->assertEquals(['first' => 'Hello', 'second' => 'World'], $c->filter(function ($item, $key) {
return $key != 'id';
})->all());
}
reduce
reduce将一个集合中的元素做遍历，返回为一个单子的元素
public function testReduce()
{
$data = new Collection([1, 2, 3]);
$this->assertEquals(6, $data->reduce(function ($carry, $element) {
return $carry += $element;
}));
}
flatten
flatten意为平坦，可以将任意嵌套的array变为同层级的，通过参数depth，可以指定平坦的层级
public function testFlatten()
{
// Flat arrays are unaffected
$c = new Collection(['#foo', '#bar', '#baz']);
$this->assertEquals(['#foo', '#bar', '#baz'], $c->flatten()->all());
// Nested arrays are flattened with existing flat items
$c = new Collection([['#foo', '#bar'], '#baz']);
$this->assertEquals(['#foo', '#bar', '#baz'], $c->flatten()->all());
// Sets of nested arrays are flattened
$c = new Collection([['#foo', '#bar'], ['#baz']]);
$this->assertEquals(['#foo', '#bar', '#baz'], $c->flatten()->all());
// Deeply nested arrays are flattened
$c = new Collection([['#foo', ['#bar']], ['#baz']]);
$this->assertEquals(['#foo', '#bar', '#baz'], $c->flatten()->all());
}
public function testFlattenWithDepth()
{
// No depth flattens recursively
$c = new Collection([['#foo', ['#bar', ['#baz']]], '#zap']);
$this->assertEquals(['#foo', '#bar', '#baz', '#zap'], $c->flatten()->all());
// Specifying a depth only flattens to that depth
$c = new Collection([['#foo', ['#bar', ['#baz']]], '#zap']);
$this->assertEquals(['#foo', ['#bar', ['#baz']], '#zap'], $c->flatten(1)->all());
$c = new Collection([['#foo', ['#bar', ['#baz']]], '#zap']);
$this->assertEquals(['#foo', '#bar', ['#baz'], '#zap'], $c->flatten(2)->all());
}
FlatMap
flatMap类似于做了先map后flat的操作
public function testFlatMap()
{
$data = new Collection([
['name' => 'taylor', 'hobbies' => ['programming', 'basketball']],
['name' => 'adam', 'hobbies' => ['music', 'powerlifting']],
]);
$data = $data->flatMap(function ($person) {
return $person['hobbies'];
});
$this->assertEquals(['programming', 'basketball', 'music', 'powerlifting'], $data->all());
}
zip
吧两个结构一样的array像拉拉链一样做合并
public function testZip()
{
$c = new Collection([1, 2, 3]);
$c = $c->zip(new Collection([4, 5, 6]));
$this->assertInstanceOf(Collection::class, $c);
$this->assertInstanceOf(Collection::class, $c[0]);
$this->assertInstanceOf(Collection::class, $c[1]);
$this->assertInstanceOf(Collection::class, $c[2]);
$this->assertCount(3, $c);
$this->assertEquals([1, 4], $c[0]->all());
$this->assertEquals([2, 5], $c[1]->all());
$this->assertEquals([3, 6], $c[2]->all());
$c = new Collection([1, 2, 3]);
$c = $c->zip([4, 5, 6], [7, 8, 9]);
$this->assertCount(3, $c);
$this->assertEquals([1, 4, 7], $c[0]->all());
$this->assertEquals([2, 5, 8], $c[1]->all());
$this->assertEquals([3, 6, 9], $c[2]->all());
$c = new Collection([1, 2, 3]);
$c = $c->zip([4, 5, 6], [7]);
$this->assertCount(3, $c);
$this->assertEquals([1, 4, 7], $c[0]->all());
$this->assertEquals([2, 5, null], $c[1]->all());
$this->assertEquals([3, 6, null], $c[2]->all());
}
pluck
pluck接受两个参数，如果传递了第二个参数，则以第二个参数为key
public function testPluckWithArrayAndObjectValues()
{
$data = new Collection([(object) ['name' => 'taylor', 'email' => 'foo'], ['name' => 'dayle', 'email' => 'bar']]);
$this->assertEquals(['taylor' => 'foo', 'dayle' => 'bar'], $data->pluck('email', 'name')->all());
$this->assertEquals(['foo', 'bar'], $data->pluck('email')->all());
}
更详细的可以参考https://laravel.tw/docs/5.2/collections
一些建议
我们使用collection的取值的时候，如果没有对应的值，我们可以提供default值，此时可以在default中直接抛出异常
return $this->checkers->first(function ($i, $checker) use ( $file) {
return $checker->canCheck($file);
}, function () {
throw new Exception("No matching style checker found!");
});
我们有时候为了连贯操作，即使前一个出错了，我们也不希望返回一个null object，我们希望能返回一个空对象，但是这个对象实现了一个空操作，意图如下：
$this->getObject($input)->check();
此处getObject($input)可能返回是一个实现了check操作的空对象，这时候就可以使用Macroable trait 的东西。
public function testMacroable()
{
// Foo() macro : unique values starting with A
Collection::macro('foo', function () {
return $this->filter(function ($item) {
return strpos($item, 'a') === 0;
})
->unique()
->values();
});
$c = new Collection(['a', 'a', 'aa', 'aaa', 'bar']);
$this->assertSame(['a', 'aa', 'aaa'], $c->foo()->all());
}
更多内容大家可以去看文章Refactoring to Collection — Notes
