Question

I’m trying to use go instead of python for my website on google app engine. But I keep getting this error with my script when I test locally.

panic: runtime error: invalid memory address or nil pointer dereference

I’m pretty confused, however it will run without error if I comment out

channel <- buffer[0:dat]

So I must be using channels incorrectly, Any help?

Edit:

This is the working code, many thanks to Kevin Ballard for helping me get this one.

package defp

import (
    "fmt"
    "http"
    "os"
)

func getContent(filename string, channel chan []byte) {
    file, err := os.OpenFile(filename, os.O_RDONLY, 0666)
    defer file.Close()
    if err == nil {
        fmt.Printf("FILE FOUND : " + filename + " \n")
        buffer := make([]byte, 16)
        dat, err := file.Read(buffer)
        for err == nil {
            fmt.Printf("herp")
            channel <- buffer[0:dat]
            buffer = make([]byte, 16)
            dat, err = file.Read(buffer)
        }
        close(channel)
        fmt.Printf("DONE READING\n")
    } else {
        fmt.Printf("FILE NOT FOUND : " + filename + " \n")
    }
}
func writeContent(w http.ResponseWriter, channel chan []byte) {
    fmt.Printf("ATTEMPTING TO WRITE CONTENT\n")
    go func() {
        for bytes := range channel {
            w.Write(bytes)
            fmt.Printf("BYTES RECEIVED\n")
        }
    }()
    fmt.Printf("FINISHED WRITING\n")
}
func load(w http.ResponseWriter, path string) {
    fmt.Printf("ATTEMPTING LOAD " + path + "\n")
    channel := make(chan []byte, 50)
    writeContent(w, channel)
    getContent(path, channel)
}
func handle(w http.ResponseWriter, r *http.Request) {
    fmt.Printf("HANDLING REQUEST FOR " + r.URL.Path[1:] + "\n")
    load(w, r.URL.Path[1:])
}
func init() {
    http.HandleFunc("/", handle)
}

Answer 1

The reason why your program sometimes panics is that it is sometimes writing to w http.ResponseWriter after the program exits the load function. The http package automatically closes the http.ResponseWriter when the program exits the handler function. In function writeContent, the program will sometimes attempt to write to a closed http.ResponseWriter.

BTW: You can make the program source code much smaller if you use the io.Copy function.

To always get predictable behavior, make sure that all work that you want the program to perform in response to a HTTP request is done before you exit the handler function. For example:

func writeContent(w http.ResponseWriter, channel chan []byte) {
    fmt.Printf("ATTEMPTING TO WRITE CONTENT\n")
    for bytes := range channel {
            w.Write(bytes)
            fmt.Printf("BYTES RECEIVED\n")
    }
    fmt.Printf("FINISHED WRITING\n")
}

func load(w http.ResponseWriter, path string) {
    fmt.Printf("ATTEMPTING LOAD " + path + "\n")
    channel := make(chan []byte)
    workDone := make(chan byte)
    go func() {
            writeContent(w, channel)
            workDone <- 1 //Send an arbitrary value
    }()
    go func() {
            getContent(path, channel)
            workDone <- 2 //Send an arbitrary value
    }()
    <-workDone
    <-workDone
}

func handle(w http.ResponseWriter, r *http.Request) {
    fmt.Printf("HANDLING REQUEST FOR " + r.URL.Path[1:] + "\n")
    load(w, r.URL.Path[1:])
}