golang判断文本文件是否是BOM格式

在Go语言中，我们可以通过读取文本文件的前几个字节来识别它是否是BOM格式的文件。BOM（Byte Order Mark）是UTF编码标准中的一部分，用于标示文本文件的编码顺序。对于不同类型的UTF编码（UTF-8, UTF-16, UTF-32），BOM的值是不同的。

UTF-8

package main

import (
        "fmt"
        "io/ioutil"
        "os"
)

func checkBOMUTF8(file string) bool {
        f, err := os.Open(file)
        if err != nil {
            fmt.Println("Error:", err)
            os.Exit(1)
        }

        defer f.Close()

        // Read the first three bytes
        b := make([]byte, 3)
        _, err = f.Read(b)
        if err != nil {
            fmt.Println("Error:", err)
            os.Exit(1)
        }

        // Check if the bytes match the UTF-8 BOM
        if b[0] == 0xEF && b[1] == 0xBB && b[2] == 0xBF {
            return true
        }

        return false
}

func main() {
        if checkBOMUTF8("test.txt") {
            fmt.Println("The file is in BOM format.")
        } else {
            fmt.Println("The file is not in BOM format.")
        }
}

UTF-16

package main

import (
        "fmt"
        "io/ioutil"
        "os"
)

func checkBOMUTF16(file string) bool {
        f, err := os.Open(file)
        if err != nil {
            fmt.Println("Error:", err)
            os.Exit(1)
        }

        defer f.Close()

        // Read the first two bytes
        b := make([]byte, 2)
        _, err = f.Read(b)
        if err != nil {
            fmt.Println("Error:", err)
            os.Exit(1)
        }

        // Check if the bytes match the UTF-16 BOM (Little Endian)
        if b[0] == 0xFF && b[1] == 0xFE {
            return true
        }

        // Check if the bytes match the UTF-16 BOM (Big Endian)
        if b[0] == 0xFE && b[1] == 0xFF {
            return true
        }

        return false
}

func main() {
        if checkBOMUTF16("test.txt") {
            fmt.Println("The file is in UTF-16 BOM format.")
        } else {
            fmt.Println("The file is not in UTF-16 BOM format.")
        }
}

UTF-32

package main

import (
        "fmt"
        "io/ioutil"
        "os"
)

func checkBOMUTF32(file string) bool {
        f, err := os.Open(file)
        if err != nil {
            fmt.Println("Error:", err)
            os.Exit(1)
        }

        defer f.Close()

        // Read the first four bytes
        b := make([]byte, 4)
        _, err = f.Read(b)
        if err != nil {
            fmt.Println("Error:", err)
            os.Exit(1)
        }

        // Check if the bytes match the UTF-32 BOM (Little Endian)
        if b[0] == 0xFF && b[1] == 0xFE && b[2] == 0x00 && b[3] == 0x00 {
            return true
        }

        // Check if the bytes match the UTF-32 BOM (Big Endian)
        if b[0] == 0x00 && b[1] == 0x00 && b[2] == 0xFE && b[3] == 0xFF {
            return true
        }

        return false
}

func main() {
        if checkBOMUTF32("test.txt") {
            fmt.Println("The file is in UTF-32 BOM format.")
        } else {
            fmt.Println("The file is not in UTF-32 BOM format.")
        }
}