在http中输出一个.mp4文件，将问题从数据库拖到浏览器

Question

answered

I am having a hard time with Mongodb and Gridfs, using it with Go's http package. I am trying to store a .mp4 file into Gridfs, then pull it out into the browser for playback.

Heres what I am doing now. It successfully pulls the file from database, I could even write it correctly to a download location.

// Connect to database
// Session to database

func movie(w http.ResponseWriter r *http.Request) {
    file, err := db.GridFS("fs").Open("movie.mp4")
    if err != nil {
        log.Println(err)
    }
    defer file.Close()

    w.Header.Set("Content-type", "video/mp4")
    if _, err := io.Copy(w, file); err != nil {
    log.Println(err)
    } 
    // I am trying to send it to the browser.
    // I want to achieve same thing as, http://localhost/view/movie.mp4, 
    as if you did that.
}

If the file was on the server, I would just do something like this. But instead I am trying to store it in Mongodb, for all around easier use involving metadata.

func movie(w http.ResponseWriter r *http.Request) {
    http.ServeFile("./uploads/movie.mp4") // Easy
}

The browser is receiving something, but its just malformed or corrupted. Just shows the video player with an error message. Any help would be appreciated, I have only been programming a week.

Here is a picture of the error, no console error messages.

Unless someone has an alternative to storing video files for playback in browser somewhere other than MongoDB or Amazon S3. Please let me know, thanks.

Answer 1

You may want to check http.ServeContent. It will handle all the mess (content-type, content-length, partial data, cache) automaticlly and save you a lot of time. It takes a ReadSeeker to serve, which GridFile already implented. So your code may simply change to the below.

func movie(w http.ResponseWriter r *http.Request) {
    file, err := db.GridFS("fs").Open("movie.mp4")
    if err != nil {
        log.Println(err)
    }
    defer file.Close()

    http.ServeContent(w,r,"movie.mp4",file.UploadDate(),file)

}

If that doesn't work, please use tools like curl or wget to download the served content and compare it to the orignal one (in db).