Golang 实现超大文件读取的两种方法

func ReadFile(filePath string) []byte{
    content, err := ioutil.ReadFile(filePath)
    if err != nil {
        log.Println("Read error")
    }
    return content
} 

func ReadFile(filePath string, handle func(string)) error {
    f, err := os.Open(filePath)
    defer f.Close()
    if err != nil {
        return err
    }
    buf := bufio.NewReader(f)
 
    for {
        line, err := buf.ReadLine("\n")
        line = strings.TrimSpace(line)
        handle(line)
        if err != nil {
            if err == io.EOF{
                return nil
            }
            return err
        }
        return nil
    }
}

func ReadBigFile(fileName string, handle func([]byte)) error {
    f, err := os.Open(fileName)
    if err != nil {
        fmt.Println("can't opened this file")
        return err
    }
    defer f.Close()
    s := make([]byte, 4096)
    for {
        switch nr, err := f.Read(s[:]); true {
        case nr < 0:
            fmt.Fprintf(os.Stderr, "cat: error reading: %s\n

package main
import (
	"bufio"
	"fmt"
	"io"
	//"math"
	"os"
	"strings"
	"sync"
	"time"
)
func main() {
	
	var (
		s time.Time //当前时间
		file *os.File
		fileStat os.FileInfo
		err error
		lastLineSize int64
	)
	s = time.Now()
	if file, err = os.Open("/Users/zhangsan/Downloads/log.txt");err != nil{
		fmt.Println(err)
	}
	defer func() {
		err = file.Close() //close after checking err
	}()
	//queryStartTime, err := time.Parse("2006-01-02T15:04:05.0000Z", startTimeArg)
	//if err != nil {
	//	fmt.Println("Could not able to parse the start time", startTimeArg)
	//	return
	//}
	//
	//queryFinishTime, err := time.Parse("2006-01-02T15:04:05.0000Z", finishTimeArg)
	//if err != nil {
	//	fmt.Println("Could not able to parse the finish time", finishTimeArg)
	//	return
	//}
	
	if fileStat, err = file.Stat();err != nil {
		return
	}
	fileSize := fileStat.Size()//72849354767
	offset := fileSize - 1
	//检测是不是都是空行 只有\n
	for {
		var (
			b []byte
			n int
			char string
		)
		b = make([]byte, 1)
		//从指定位置读取
		if n, err = file.ReadAt(b, offset);err != nil {
			fmt.Println("Error reading file ", err)
			break
		}
		char = string(b[0])
		if char == "\n" {
			break
		}
		offset--
		//获取一行的大小
		lastLineSize += int64(n)
	}
	var (
		lastLine []byte
		logSlice []string
		logSlice1 []string
	)
	//初始化一行大小的空间
	lastLine = make([]byte, lastLineSize)
	_, err = file.ReadAt(lastLine, offset)
	if err != nil {
		fmt.Println("Could not able to read last line with offset", offset, "and lastline size", lastLineSize)
		return
	}
	//根据条件进行区分
	logSlice = strings.Split(strings.Trim(string(lastLine),"\n"),"next_pay_date")
	logSlice1  = strings.Split(logSlice[1],"\"")
	if logSlice1[2] == "2021-06-15"{
		Process(file)
	}
	fmt.Println("\nTime taken - ", time.Since(s))
		fmt.Println(err)
}
func Process(f *os.File) error {
	//读取数据的key，减小gc压力
	linesPool := sync.Pool{New: func() interface{} {
		lines := make([]byte, 250*1024)
		return lines
	}}
	//读取回来的数据池
	stringPool := sync.Pool{New: func() interface{} {
		lines := ""
		return lines
	}}
	//一个文件对象本身是实现了io.Reader的 使用bufio.NewReader去初始化一个Reader对象，存在buffer中的，读取一次就会被清空
	r := bufio.NewReader(f) //
	//设置读取缓冲池大小 默认16
	r = bufio.NewReaderSize(r,250 *1024)
	var wg sync.WaitGroup
	for {
		buf := linesPool.Get().([]byte)
		//读取Reader对象中的内容到[]byte类型的buf中
		n, err := r.Read(buf)
		buf = buf[:n]
		if n == 0 {
			if err != nil {
				fmt.Println(err)
				break
			}
			if err == io.EOF {
				break
			}
			return err
		}
		//补齐剩下没满足的剩余
		nextUntillNewline, err := r.ReadBytes('\n')
		//fmt.Println(string(nextUntillNewline))
		if err != io.EOF {
			buf = append(buf, nextUntillNewline...)
		}
		wg.Add(1)
		go func() {
			ProcessChunk(buf, &linesPool, &stringPool)
			wg.Done()
		}()
	}
	wg.Wait()
	return nil
}
func ProcessChunk(chunk []byte, linesPool *sync.Pool,stringPool *sync.Pool) {
//做相应的处理
}

go run test2.go "2020-01-01T00:00:00.0000Z" "2020-02-02T00:00:00.0000Z" /Users/zhangsan/go/class="lazy" data-src/workspace/test/log.log

EOF
Time taken -  20.023517675s
<nil>