goroutine didn't take effect in Crawl example of 'A Tour of Go' - go

Question

As the hits mentioned in Crawl example of 'A Tour of Go', I modified the Crawl function and just wonder why the 'go Crawl' failed to spawn another thread as only one url was found printed out.
Is there anything wrong with my modification?
List my modification as below,
// Crawl uses fetcher to recursively crawl
// pages starting with url, to a maximum of depth.
func Crawl(url string, depth int, fetcher Fetcher) {
// TODO: Fetch URLs in parallel.
// TODO: Don't fetch the same URL twice.
// This implementation doesn't do either:
if depth <= 0 {
fmt.Printf("depth <= 0 return")
return
}
body, urls, err := fetcher.Fetch(url)
if err != nil {
fmt.Println(err)
return
}
fmt.Printf("found: %s %q\n", url, body)
crawled.mux.Lock()
crawled.c[url]++
crawled.mux.Unlock()
for _, u := range urls {
//crawled.mux.Lock()
if cnt, ok := crawled.c[u]; ok {
cnt++
} else {
fmt.Println("go ...", u)
go Crawl(u, depth-1, fetcher)
}
//crawled.mux.Unlock()
//Crawl(u, depth-1, fetcher)
}
return
}
type crawledUrl struct {
c map[string]int
mux sync.Mutex
}
var crawled = crawledUrl{c: make(map[string]int)}

Answer 1

In your program, you have no any synchronized tool for your go routines.
So the behavior of this code is undefined. Perhaps main go thread will end soon.
Please remember that the main go routine will never block to wait other go routines for termination, only if you explicitly use some kind of util to synchronize the execution of go routines.
Such as channels or useful sync utils.
Let me help to give a version.
type fetchState struct {
mu sync.Mutex
fetched map[string]bool
}
func (f *fetchState) CheckAndMark(url string) bool {
defer f.mu.Unlock()
f.mu.Lock()
if f.fetched[url] {
return true
}
f.fetched[url] = true
return false
}
func mkFetchState() *fetchState {
f := &fetchState{}
f.fetched = make(map[string]bool)
return f
}
func CrawlConcurrentMutex(url string, fetcher Fetcher, f *fetchState) {
if f.CheckAndMark(url) {
return
}
body, urls, err := fetcher.Fetch(url)
if err != nil {
fmt.Println(err)
return
}
fmt.Printf("found: %s %q\n", url, body)
var done sync.WaitGroup
for _, u := range urls {
done.Add(1)
go func(u string) {
defer done.Done()
CrawlConcurrentMutex(u, fetcher, f)
}(u) // Without the u argument there is a race
}
done.Wait()
return
}
Please pay attention to the usage of sync.WaitGroup, please refer the doc and you can understand the whole story.