用過 Java 的同學(xué)都熟悉 Stream API,那么在 Go 里我們可以用類似的方式處理集合數(shù)據(jù)嗎?本文給大家介紹 go-zero 內(nèi)置的 Stream API,為了幫助理解,函數(shù)主要分為三類:獲取操作、中間處理操作、終結(jié)操作。
什么是流處理
如果有 java 使用經(jīng)驗(yàn)的同學(xué)一定會對 java8 的 Stream 贊不絕口,極大的提高了們對于集合類型數(shù)據(jù)的處理能力。
1
2
3
4
|
int sum = widgets.stream() .filter(w -> w.getColor() == RED) .mapToInt(w -> w.getWeight()) .sum(); |
Stream 能讓我們支持鏈?zhǔn)秸{(diào)用和函數(shù)編程的風(fēng)格來實(shí)現(xiàn)數(shù)據(jù)的處理,看起來數(shù)據(jù)像是在流水線一樣不斷的實(shí)時流轉(zhuǎn)加工,最終被匯總。Stream 的實(shí)現(xiàn)思想就是將數(shù)據(jù)處理流程抽象成了一個數(shù)據(jù)流,每次加工后返回一個新的流供使用。
Stream 功能定義
動手寫代碼之前,先想清楚,把需求理清楚是最重要的一步,我們嘗試代入作者的視角來思考整個組件的實(shí)現(xiàn)流程。首先把底層實(shí)現(xiàn)的邏輯放一下 ,先嘗試從零開始進(jìn)行功能定義 stream 功能。
Stream 的工作流程其實(shí)也屬于生產(chǎn)消費(fèi)者模型,整個流程跟工廠中的生產(chǎn)流程非常相似,嘗試先定義一下 Stream 的生命周期:
- 創(chuàng)建階段/數(shù)據(jù)獲取(原料)
- 加工階段/中間處理(流水線加工)
- 匯總階段/終結(jié)操作(最終產(chǎn)品)
下面圍繞 stream 的三個生命周期開始定義 API:
創(chuàng)建階段
為了創(chuàng)建出數(shù)據(jù)流 stream 這一抽象對象,可以理解為構(gòu)造器。
我們支持三種方式構(gòu)造 stream,分別是:切片轉(zhuǎn)換,channel 轉(zhuǎn)換,函數(shù)式轉(zhuǎn)換。
注意這個階段的方法都是普通的公開方法,并不綁定 Stream 對象。
1
2
3
4
5
6
7
8
9
10
11
|
// 通過可變參數(shù)模式創(chuàng)建 stream func Just(items ... interface {}) Stream // 通過 channel 創(chuàng)建 stream func Range(source <-chan interface {}) Stream // 通過函數(shù)創(chuàng)建 stream func From(generate GenerateFunc) Stream // 拼接 stream func Concat(s Stream, others ...Stream) Stream |
加工階段
加工階段需要進(jìn)行的操作往往對應(yīng)了我們的業(yè)務(wù)邏輯,比如:轉(zhuǎn)換,過濾,去重,排序等等。
這個階段的 API 屬于 method 需要綁定到 Stream 對象上。
結(jié)合常用的業(yè)務(wù)場景進(jìn)行如下定義:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
|
// 去除重復(fù)item Distinct(keyFunc KeyFunc) Stream // 按條件過濾item Filter(filterFunc FilterFunc, opts ...Option) Stream // 分組 Group(fn KeyFunc) Stream // 返回前n個元素 Head(n int64) Stream // 返回后n個元素 Tail(n int64) Stream // 轉(zhuǎn)換對象 Map(fn MapFunc, opts ...Option) Stream // 合并item到slice生成新的stream Merge() Stream // 反轉(zhuǎn) Reverse() Stream // 排序 Sort(fn LessFunc) Stream // 作用在每個item上 Walk(fn WalkFunc, opts ...Option) Stream // 聚合其他Stream Concat(streams ...Stream) Stream |
加工階段的處理邏輯都會返回一個新的 Stream 對象,這里有個基本的實(shí)現(xiàn)范式
匯總階段
匯總階段其實(shí)就是我們想要的處理結(jié)果,比如:是否匹配,統(tǒng)計(jì)數(shù)量,遍歷等等。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
// 檢查是否全部匹配 AllMatch(fn PredicateFunc) bool // 檢查是否存在至少一項(xiàng)匹配 AnyMatch(fn PredicateFunc) bool // 檢查全部不匹配 NoneMatch(fn PredicateFunc) bool // 統(tǒng)計(jì)數(shù)量 Count() int // 清空stream Done() // 對所有元素執(zhí)行操作 ForAll(fn ForAllFunc) // 對每個元素執(zhí)行操作 ForEach(fn ForEachFunc) |
梳理完組件的需求邊界后,我們對于即將要實(shí)現(xiàn)的 Stream 有了更清晰的認(rèn)識。在我的認(rèn)知里面真正的架構(gòu)師對于需求的把握以及后續(xù)演化能達(dá)到及其精準(zhǔn)的地步,做到這一點(diǎn)離不開對需求的深入思考以及洞穿需求背后的本質(zhì)。通過代入作者的視角來模擬復(fù)盤整個項(xiàng)目的構(gòu)建流程,學(xué)習(xí)作者的思維方法論這正是我們學(xué)習(xí)開源項(xiàng)目最大的價值所在。
好了,我們嘗試定義出完整的 Stream 接口全貌以及函數(shù)。
接口的作用不僅僅是模版作用,還在于利用其抽象能力搭建項(xiàng)目整體的框架而不至于一開始就陷入細(xì)節(jié),能快速的將我們的思考過程通過接口簡潔的表達(dá)出來,學(xué)會養(yǎng)成自頂向下的思維方法從宏觀的角度來觀察整個系統(tǒng),一開始就陷入細(xì)節(jié)則很容易拔劍四顧心茫然。。。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
|
rxOptions struct { unlimitedWorkers bool workers int } Option func(opts *rxOptions) // key生成器 //item - stream中的元素 KeyFunc func(item interface {}) interface {} // 過濾函數(shù) FilterFunc func(item interface {}) bool // 對象轉(zhuǎn)換函數(shù) MapFunc func(intem interface {}) interface {} // 對象比較 LessFunc func(a, b interface {}) bool // 遍歷函數(shù) WalkFunc func(item interface {}, pip chan<- interface {}) // 匹配函數(shù) PredicateFunc func(item interface {}) bool // 對所有元素執(zhí)行操作 ForAllFunc func(pip <-chan interface {}) // 對每個item執(zhí)行操作 ForEachFunc func(item interface {}) // 對每個元素并發(fā)執(zhí)行操作 ParallelFunc func(item interface {}) // 對所有元素執(zhí)行聚合操作 ReduceFunc func(pip <-chan interface {}) ( interface {}, error) // item生成函數(shù) GenerateFunc func(source <-chan interface {}) Stream interface { // 去除重復(fù)item Distinct(keyFunc KeyFunc) Stream // 按條件過濾item Filter(filterFunc FilterFunc, opts ...Option) Stream // 分組 Group(fn KeyFunc) Stream // 返回前n個元素 Head(n int64) Stream // 返回后n個元素 Tail(n int64) Stream // 獲取第一個元素 First() interface {} // 獲取最后一個元素 Last() interface {} // 轉(zhuǎn)換對象 Map(fn MapFunc, opts ...Option) Stream // 合并item到slice生成新的stream Merge() Stream // 反轉(zhuǎn) Reverse() Stream // 排序 Sort(fn LessFunc) Stream // 作用在每個item上 Walk(fn WalkFunc, opts ...Option) Stream // 聚合其他Stream Concat(streams ...Stream) Stream // 檢查是否全部匹配 AllMatch(fn PredicateFunc) bool // 檢查是否存在至少一項(xiàng)匹配 AnyMatch(fn PredicateFunc) bool // 檢查全部不匹配 NoneMatch(fn PredicateFunc) bool // 統(tǒng)計(jì)數(shù)量 Count() int // 清空stream Done() // 對所有元素執(zhí)行操作 ForAll(fn ForAllFunc) // 對每個元素執(zhí)行操作 ForEach(fn ForEachFunc) } |
channel() 方法用于獲取 Stream 管道屬性,因?yàn)樵诰唧w實(shí)現(xiàn)時我們面向的是接口對象所以暴露一個私有方法 read 出來。
1
2
|
// 獲取內(nèi)部的數(shù)據(jù)容器channel,內(nèi)部方法 channel() chan interface {} |
實(shí)現(xiàn)思路
功能定義梳理清楚了,接下來考慮幾個工程實(shí)現(xiàn)的問題。
如何實(shí)現(xiàn)鏈?zhǔn)秸{(diào)用
鏈?zhǔn)秸{(diào)用,創(chuàng)建對象用到的 builder 模式可以達(dá)到鏈?zhǔn)秸{(diào)用效果。實(shí)際上 Stream 實(shí)現(xiàn)類似鏈?zhǔn)降男Ч硪彩且粯拥模看握{(diào)用完后都創(chuàng)建一個新的 Stream 返回給用戶。
1
2
3
4
|
// 去除重復(fù)item Distinct(keyFunc KeyFunc) Stream // 按條件過濾item Filter(filterFunc FilterFunc, opts ...Option) Stream |
如何實(shí)現(xiàn)流水線的處理效果
所謂的流水線可以理解為數(shù)據(jù)在 Stream 中的存儲容器,在 go 中我們可以使用 channel 作為數(shù)據(jù)的管道,達(dá)到 Stream 鏈?zhǔn)秸{(diào)用執(zhí)行多個操作時異步非阻塞效果。
如何支持并行處理
數(shù)據(jù)加工本質(zhì)上是在處理 channel 中的數(shù)據(jù),那么要實(shí)現(xiàn)并行處理無非是并行消費(fèi) channel 而已,利用 goroutine 協(xié)程、WaitGroup 機(jī)制可以非常方便的實(shí)現(xiàn)并行處理。
go-zero 實(shí)現(xiàn)
core/fx/stream.go
go-zero 中關(guān)于 Stream 的實(shí)現(xiàn)并沒有定義接口,不過沒關(guān)系底層實(shí)現(xiàn)時邏輯是一樣的。
為了實(shí)現(xiàn) Stream 接口我們定義一個內(nèi)部的實(shí)現(xiàn)類,其中 source 為 channel 類型,模擬流水線功能。
1
2
3
|
Stream struct { source <-chan interface {} } |
創(chuàng)建 API
channel 創(chuàng)建 Range
通過 channel 創(chuàng)建 stream
1
2
3
4
5
|
func Range(source <-chan interface {}) Stream { return Stream{ source: source, } } |
可變參數(shù)模式創(chuàng)建 Just
通過可變參數(shù)模式創(chuàng)建 stream,channel 寫完后及時 close 是個好習(xí)慣。
1
2
3
4
5
6
7
8
|
func Just(items ... interface {}) Stream { source := make(chan interface {}, len(items)) for _, item := range items { source <- item } close(source) return Range(source) } |
函數(shù)創(chuàng)建 From
通過函數(shù)創(chuàng)建 Stream
1
2
3
4
5
6
7
8
|
func From(generate GenerateFunc) Stream { source := make(chan interface {}) threading.GoSafe(func() { defer close(source) generate(source) }) return Range(source) } |
因?yàn)樯婕巴獠總魅氲暮瘮?shù)參數(shù)調(diào)用,執(zhí)行過程并不可用因此需要捕捉運(yùn)行時異常防止 panic 錯誤傳導(dǎo)到上層導(dǎo)致應(yīng)用崩潰。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
func Recover(cleanups ...func()) { for _, cleanup := range cleanups { cleanup() } if r := recover(); r != nil { logx.ErrorStack(r) } } func RunSafe(fn func()) { defer rescue.Recover() fn() } func GoSafe(fn func()) { go RunSafe(fn) } |
拼接 Concat
拼接其他 Stream 創(chuàng)建一個新的 Stream,調(diào)用內(nèi)部 Concat method 方法,后文將會分析 Concat 的源碼實(shí)現(xiàn)。
1
2
3
|
func Concat(s Stream, others ...Stream) Stream { return s.Concat(others...) } |
加工 API
去重 Distinct
因?yàn)閭魅氲氖呛瘮?shù)參數(shù)KeyFunc func(item interface{}) interface{}
意味著也同時支持按照業(yè)務(wù)場景自定義去重,本質(zhì)上是利用 KeyFunc 返回的結(jié)果基于 map 實(shí)現(xiàn)去重。
函數(shù)參數(shù)非常強(qiáng)大,能極大的提升靈活性。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
|
func (s Stream) Distinct(keyFunc KeyFunc) Stream { source := make(chan interface {}) threading.GoSafe(func() { // channel記得關(guān)閉是個好習(xí)慣 defer close(source) keys := make(map[ interface {}]lang.PlaceholderType) for item := range s.source { // 自定義去重邏輯 key := keyFunc(item) // 如果key不存在,則將數(shù)據(jù)寫入新的channel if _, ok := keys[key]; !ok { source <- item keys[key] = lang.Placeholder } } }) return Range(source) } |
使用案例:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
|
// 1 2 3 4 5 Just( 1 , 2 , 3 , 3 , 4 , 5 , 5 ).Distinct(func(item interface {}) interface {} { return item }).ForEach(func(item interface {}) { t.Log(item) }) // 1 2 3 4 Just( 1 , 2 , 3 , 3 , 4 , 5 , 5 ).Distinct(func(item interface {}) interface {} { uid := item.( int ) // 對大于4的item進(jìn)行特殊去重邏輯,最終只保留一個>3的item if uid > 3 { return 4 } return item }).ForEach(func(item interface {}) { t.Log(item) }) |
過濾 Filter
通過將過濾邏輯抽象成 FilterFunc,然后分別作用在 item 上根據(jù) FilterFunc 返回的布爾值決定是否寫回新的 channel 中實(shí)現(xiàn)過濾功能,實(shí)際的過濾邏輯委托給了 Walk method。
Option 參數(shù)包含兩個選項(xiàng):
- unlimitedWorkers 不限制協(xié)程數(shù)量
- workers 限制協(xié)程數(shù)量
1
2
3
4
5
6
7
8
9
|
FilterFunc func(item interface {}) bool func (s Stream) Filter(filterFunc FilterFunc, opts ...Option) Stream { return s.Walk(func(item interface {}, pip chan<- interface {}) { if filterFunc(item) { pip <- item } }, opts...) } |
使用示例:
1
2
3
4
5
6
7
8
9
|
func TestInternalStream_Filter(t *testing.T) { // 保留偶數(shù) 2,4 channel := Just( 1 , 2 , 3 , 4 , 5 ).Filter(func(item interface {}) bool { return item.( int )% 2 == 0 }).channel() for item := range channel { t.Log(item) } } |
遍歷執(zhí)行 Walk
walk 英文意思是步行,這里的意思是對每個 item 都執(zhí)行一次 WalkFunc 操作并將結(jié)果寫入到新的 Stream 中。
這里注意一下因?yàn)閮?nèi)部采用了協(xié)程機(jī)制異步執(zhí)行讀取和寫入數(shù)據(jù)所以新的 Stream 中 channel 里面的數(shù)據(jù)順序是隨機(jī)的。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
|
// item-stream中的item元素 // pipe-item符合條件則寫入pipe WalkFunc func(item interface {}, pipe chan<- interface {}) func (s Stream) Walk(fn WalkFunc, opts ...Option) Stream { option := buildOptions(opts...) if option.unlimitedWorkers { return s.walkUnLimited(fn, option) } return s.walkLimited(fn, option) } func (s Stream) walkUnLimited(fn WalkFunc, option *rxOptions) Stream { // 創(chuàng)建帶緩沖區(qū)的channel // 默認(rèn)為16,channel中元素超過16將會被阻塞 pipe := make(chan interface {}, defaultWorkers) go func() { var wg sync.WaitGroup for item := range s.source { // 需要讀取s.source的所有元素 // 這里也說明了為什么channel最后寫完記得完畢 // 如果不關(guān)閉可能導(dǎo)致協(xié)程一直阻塞導(dǎo)致泄漏 // 重要, 不賦值給val是個典型的并發(fā)陷阱,后面在另一個goroutine里使用了 val := item wg.Add( 1 ) // 安全模式下執(zhí)行函數(shù) threading.GoSafe(func() { defer wg.Done() fn(item, pipe) }) } wg.Wait() close(pipe) }() // 返回新的Stream return Range(pipe) } func (s Stream) walkLimited(fn WalkFunc, option *rxOptions) Stream { pipe := make(chan interface {}, option.workers) go func() { var wg sync.WaitGroup // 控制協(xié)程數(shù)量 pool := make(chan lang.PlaceholderType, option.workers) for item := range s.source { // 重要, 不賦值給val是個典型的并發(fā)陷阱,后面在另一個goroutine里使用了 val := item // 超過協(xié)程限制時將會被阻塞 pool <- lang.Placeholder // 這里也說明了為什么channel最后寫完記得完畢 // 如果不關(guān)閉可能導(dǎo)致協(xié)程一直阻塞導(dǎo)致泄漏 wg.Add( 1 ) // 安全模式下執(zhí)行函數(shù) threading.GoSafe(func() { defer func() { wg.Done() //執(zhí)行完成后讀取一次pool釋放一個協(xié)程位置 <-pool }() fn(item, pipe) }) } wg.Wait() close(pipe) }() return Range(pipe) } |
使用案例:
返回的順序是隨機(jī)的。
1
2
3
4
5
6
7
8
|
func Test_Stream_Walk(t *testing.T) { // 返回 300,100,200 Just( 1 , 2 , 3 ).Walk(func(item interface {}, pip chan<- interface {}) { pip <- item.( int ) * 100 }, WithWorkers( 3 )).ForEach(func(item interface {}) { t.Log(item) }) } |
分組 Group
通過對 item 匹配放入 map 中。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
KeyFunc func(item interface {}) interface {} func (s Stream) Group(fn KeyFunc) Stream { groups := make(map[ interface {}][] interface {}) for item := range s.source { key := fn(item) groups[key] = append(groups[key], item) } source := make(chan interface {}) go func() { for _, group := range groups { source <- group } close(source) }() return Range(source) } |
獲取前 n 個元素 Head
n 大于實(shí)際數(shù)據(jù)集長度的話將會返回全部元素
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
|
func (s Stream) Head(n int64) Stream { if n < 1 { panic( "n must be greather than 1" ) } source := make(chan interface {}) go func() { for item := range s.source { n-- // n值可能大于s.source長度,需要判斷是否>=0 if n >= 0 { source <- item } // let successive method go ASAP even we have more items to skip // why we don't just break the loop, because if break, // this former goroutine will block forever, which will cause goroutine leak. // n==0說明source已經(jīng)寫滿可以進(jìn)行關(guān)閉了 // 既然source已經(jīng)滿足條件了為什么不直接進(jìn)行break跳出循環(huán)呢? // 作者提到了防止協(xié)程泄漏 // 因?yàn)槊看尾僮髯罱K都會產(chǎn)生一個新的Stream,舊的Stream永遠(yuǎn)也不會被調(diào)用了 if n == 0 { close(source) break } } // 上面的循環(huán)跳出來了說明n大于s.source實(shí)際長度 // 依舊需要顯示關(guān)閉新的source if n > 0 { close(source) } }() return Range(source) } |
使用示例:
1
2
3
4
5
6
7
|
// 返回1,2 func TestInternalStream_Head(t *testing.T) { channel := Just( 1 , 2 , 3 , 4 , 5 ).Head( 2 ).channel() for item := range channel { t.Log(item) } } |
獲取后 n 個元素 Tail
這里很有意思,為了確保拿到最后 n 個元素使用環(huán)形切片 Ring 這個數(shù)據(jù)結(jié)構(gòu),先了解一下 Ring 的實(shí)現(xiàn)。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
|
// 環(huán)形切片 type Ring struct { elements [] interface {} index int lock sync.Mutex } func NewRing(n int ) *Ring { if n < 1 { panic( "n should be greather than 0" ) } return &Ring{ elements: make([] interface {}, n), } } // 添加元素 func (r *Ring) Add(v interface {}) { r.lock.Lock() defer r.lock.Unlock() // 將元素寫入切片指定位置 // 這里的取余實(shí)現(xiàn)了循環(huán)寫效果 r.elements[r.index%len(r.elements)] = v // 更新下次寫入位置 r.index++ } // 獲取全部元素 // 讀取順序保持與寫入順序一致 func (r *Ring) Take() [] interface {} { r.lock.Lock() defer r.lock.Unlock() var size int var start int // 當(dāng)出現(xiàn)循環(huán)寫的情況時 // 開始讀取位置需要通過去余實(shí)現(xiàn),因?yàn)槲覀兿Mx取出來的順序與寫入順序一致 if r.index > len(r.elements) { size = len(r.elements) // 因?yàn)槌霈F(xiàn)循環(huán)寫情況,當(dāng)前寫入位置index開始為最舊的數(shù)據(jù) start = r.index % len(r.elements) } else { size = r.index } elements := make([] interface {}, size) for i := 0 ; i < size; i++ { // 取余實(shí)現(xiàn)環(huán)形讀取,讀取順序保持與寫入順序一致 elements[i] = r.elements[(start+i)%len(r.elements)] } return elements } |
總結(jié)一下環(huán)形切片的優(yōu)點(diǎn):
- 支持自動滾動更新
- 節(jié)省內(nèi)存
環(huán)形切片能實(shí)現(xiàn)固定容量滿的情況下舊數(shù)據(jù)不斷被新數(shù)據(jù)覆蓋,由于這個特性可以用于讀取 channel 后 n 個元素。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
|
func (s Stream) Tail(n int64) Stream { if n < 1 { panic( "n must be greather than 1" ) } source := make(chan interface {}) go func() { ring := collection.NewRing( int (n)) // 讀取全部元素,如果數(shù)量>n環(huán)形切片能實(shí)現(xiàn)新數(shù)據(jù)覆蓋舊數(shù)據(jù) // 保證獲取到的一定最后n個元素 for item := range s.source { ring.Add(item) } for _, item := range ring.Take() { source <- item } close(source) }() return Range(source) } |
那么為什么不直接使用 len(source) 長度的切片呢?
答案是節(jié)省內(nèi)存。凡是涉及到環(huán)形類型的數(shù)據(jù)結(jié)構(gòu)時都具備一個優(yōu)點(diǎn)那就省內(nèi)存,能做到按需分配資源。
使用示例:
1
2
3
4
5
6
7
8
9
10
11
12
|
func TestInternalStream_Tail(t *testing.T) { // 4,5 channel := Just( 1 , 2 , 3 , 4 , 5 ).Tail( 2 ).channel() for item := range channel { t.Log(item) } // 1,2,3,4,5 channel2 := Just( 1 , 2 , 3 , 4 , 5 ).Tail( 6 ).channel() for item := range channel2 { t.Log(item) } } |
元素轉(zhuǎn)換Map
元素轉(zhuǎn)換,內(nèi)部由協(xié)程完成轉(zhuǎn)換操作,注意輸出channel并不保證按原序輸出。
1
2
3
4
5
6
|
MapFunc func(intem interface {}) interface {} func (s Stream) Map(fn MapFunc, opts ...Option) Stream { return s.Walk(func(item interface {}, pip chan<- interface {}) { pip <- fn(item) }, opts...) } |
使用示例:
1
2
3
4
5
6
7
8
|
func TestInternalStream_Map(t *testing.T) { channel := Just( 1 , 2 , 3 , 4 , 5 , 2 , 2 , 2 , 2 , 2 , 2 ).Map(func(item interface {}) interface {} { return item.( int ) * 10 }).channel() for item := range channel { t.Log(item) } } |
合并 Merge
實(shí)現(xiàn)比較簡單,我考慮了很久沒想到有什么場景適合這個方法。
1
2
3
4
5
6
7
8
9
|
func (s Stream) Merge() Stream { var items [] interface {} for item := range s.source { items = append(items, item) } source := make(chan interface {}, 1 ) source <- items return Range(source) } |
反轉(zhuǎn) Reverse
反轉(zhuǎn) channel 中的元素。反轉(zhuǎn)算法流程是:
- 找到中間節(jié)點(diǎn)
- 節(jié)點(diǎn)兩邊開始兩兩交換
注意一下為什么獲取 s.source 時用切片來接收呢? 切片會自動擴(kuò)容,用數(shù)組不是更好嗎?
其實(shí)這里是不能用數(shù)組的,因?yàn)椴恢?Stream 寫入 source 的操作往往是在協(xié)程異步寫入的,每個 Stream 中的 channel 都可能在動態(tài)變化,用流水線來比喻 Stream 工作流程的確非常形象。
1
2
3
4
5
6
7
8
9
10
11
|
func (s Stream) Reverse() Stream { var items [] interface {} for item := range s.source { items = append(items, item) } for i := len(items)/ 2 - 1 ; i >= 0 ; i-- { opp := len(items) - 1 - i items[i], items[opp] = items[opp], items[i] } return Just(items...) } |
使用示例:
1
2
3
4
5
6
|
func TestInternalStream_Reverse(t *testing.T) { channel := Just( 1 , 2 , 3 , 4 , 5 ).Reverse().channel() for item := range channel { t.Log(item) } } |
排序 Sort
內(nèi)網(wǎng)調(diào)用 slice 官方包的排序方案,傳入比較函數(shù)實(shí)現(xiàn)比較邏輯即可。
1
2
3
4
5
6
7
8
9
10
11
|
func (s Stream) Sort(fn LessFunc) Stream { var items [] interface {} for item := range s.source { items = append(items, item) } sort.Slice(items, func(i, j int ) bool { return fn(i, j) }) return Just(items...) } |
使用示例:
1
2
3
4
5
6
7
8
9
|
// 5,4,3,2,1 func TestInternalStream_Sort(t *testing.T) { channel := Just( 1 , 2 , 3 , 4 , 5 ).Sort(func(a, b interface {}) bool { return a.( int ) > b.( int ) }).channel() for item := range channel { t.Log(item) } } |
拼接 Concat
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
func (s Stream) Concat(steams ...Stream) Stream { // 創(chuàng)建新的無緩沖channel source := make(chan interface {}) go func() { // 創(chuàng)建一個waiGroup對象 group := threading.NewRoutineGroup() // 異步從原channel讀取數(shù)據(jù) group.Run(func() { for item := range s.source { source <- item } }) // 異步讀取待拼接Stream的channel數(shù)據(jù) for _, stream := range steams { // 每個Stream開啟一個協(xié)程 group.Run(func() { for item := range stream.channel() { source <- item } }) } // 阻塞等待讀取完成 group.Wait() close(source) }() // 返回新的Stream return Range(source) } |
匯總 API
全部匹配 AllMatch
1
2
3
4
5
6
7
8
9
10
11
|
func (s Stream) AllMatch(fn PredicateFunc) bool { for item := range s.source { if !fn(item) { // 需要排空 s.source,否則前面的goroutine可能阻塞 go drain(s.source) return false } } return true } |
任意匹配 AnyMatch
1
2
3
4
5
6
7
8
9
10
11
|
func (s Stream) AnyMatch(fn PredicateFunc) bool { for item := range s.source { if fn(item) { // 需要排空 s.source,否則前面的goroutine可能阻塞 go drain(s.source) return true } } return false } |
一個也不匹配 NoneMatch
1
2
3
4
5
6
7
8
9
10
11
|
func (s Stream) NoneMatch(fn func(item interface {}) bool) bool { for item := range s.source { if fn(item) { // 需要排空 s.source,否則前面的goroutine可能阻塞 go drain(s.source) return false } } return true } |
數(shù)量統(tǒng)計(jì) Count
1
2
3
4
5
6
7
|
func (s Stream) Count() int { var count int for range s.source { count++ } return count } |
清空 Done
1
2
3
4
|
func (s Stream) Done() { // 排空 channel,防止 goroutine 阻塞泄露 drain(s.source) } |
迭代全部元素 ForAll
1
2
3
|
func (s Stream) ForAll(fn ForAllFunc) { fn(s.source) } |
迭代每個元素 ForEach
1
2
3
|
func (s Stream) ForAll(fn ForAllFunc) { fn(s.source) } |
小結(jié)
至此 Stream 組件就全部實(shí)現(xiàn)完了,核心邏輯是利用 channel 當(dāng)做管道,數(shù)據(jù)當(dāng)做水流,不斷的用協(xié)程接收/寫入數(shù)據(jù)到 channel 中達(dá)到異步非阻塞的效果。
回到開篇提到的問題,未動手前想要實(shí)現(xiàn)一個 stream 難度似乎非常大,很難想象在 go 中 300 多行的代碼就能實(shí)現(xiàn)如此強(qiáng)大的組件。
實(shí)現(xiàn)高效的基礎(chǔ)來源三個語言特性:
- channel
- 協(xié)程
- 函數(shù)式編程
參考資料
項(xiàng)目地址
https://github.com/zeromicro/go-zero
到此這篇關(guān)于Go 通過 Map/Filter/ForEach 等流式 API 高效處理數(shù)據(jù)的文章就介紹到這了,更多相關(guān)go 流式 API 處理數(shù)據(jù)內(nèi)容請搜索服務(wù)器之家以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持服務(wù)器之家!
原文鏈接:https://www.cnblogs.com/kevinwan/p/15761172.html