我正在尝试构建一个用 Go 编写的云函数,它将使用 Google Cloud Functions 基础架构中可用的 ImageMagick 库将多个图像合成和操作成一个结果输出图像。
问题的根源在于我想使用的 ImageMagick 函数是可用的,但它需要多个不同的输入才能工作。我的输入是存储桶中的对象。
os/exec Cmd 结构允许您通过使用“ExtraFiles”数组来执行此操作,并且我知道如何将这些额外文件提供给我的 ImageMagick 命令。
但是,“ExtraFiles”数组只想存储 os.File 的实例,而 GCP 存储客户端会在您打开文件时为您提供“Reader”实例。
backgroundBlob := storageClient.Bucket(inputBucket).Object(background)
backgroundImg, err := backgroundBlob.NewReader(ctx)
if err != nil {
return "", fmt.Errorf("backgroundImg: %v", err)
}
foregroundBlob := storageClient.Bucket(inputBucket).Object(foreground)
foregroundImg, err := foregroundBlob.NewReader(ctx)
if err != nil {
return "", fmt.Errorf("foregroundImg: %v", err)
}
outputBlob := storageClient.Bucket(outputBucket).Object(output)
outputImg := outputBlob.NewWriter(ctx)
defer outputImg.Close()
// Set up some additional file handles since we're dealing with more inputs than STDIN Can cope with
cmd := exec.Command("convert", "fd:3", "fd:4", "-composite", "fd:5")
cmd.ExtraFiles = append(cmd.ExtraFiles,backgroundImg)
cmd.ExtraFiles = append(cmd.ExtraFiles,foregroundImg)
cmd.ExtraFiles = append(cmd.ExtraFiles,outputImg)
if err := cmd.Run(); err != nil {
return "", fmt.Errorf("cmd.Run: %v", err)
}
log.Printf("Blurred image has been uploaded to %s", outputBlob.ObjectName())
return outputBlob.ObjectName(), nil
}```
So, from where I am at the moment, I need to do one of the following two things:
1. Figure out how to get Google's storage API to treat/use objects in their storage buckets as "os.File"s
OR
2. Figure out how to execute my "convert" command using multiple Readers as input, rather than leveraging the ExtraFiles array.
If anybody out there has any ideas on how to achieve either of the above, or has alternate ways of solving this problem, I would be very grateful to hear them!