A Tour of Go - Exercise: Web Crawler

crawler.go

package main

import (
	"fmt"
	"sync"
)

type Fetcher interface {
	// Fetch returns the body of URL and
	// a slice of URLs found on that page.
	Fetch(url string) (body string, urls []string, err error)
}

type Cache struct {
	visited map[string]bool
	mux sync.Mutex
}


// Crawl uses fetcher to recursively crawl
// pages starting with url, to a maximum of depth.
func Crawl(url string, depth int, fetcher Fetcher,
		   ch chan response, cache Cache) {
	defer close(ch)
	if depth <= 0 {
		return
	}
	cache.mux.Lock()
	if cache.visited[url] {
		cache.mux.Unlock()
		return
	}
	cache.visited[url] = true
	cache.mux.Unlock()
	
	body, urls, err := fetcher.Fetch(url)
	if err != nil {
		fmt.Println(err)
		return
	}
	
	ch <- response{url, body}
	result := make([]chan response, len(urls))
	for i, u := range urls {
        	result[i] = make(chan response)
		go Crawl(u, depth-1, fetcher, result[i], cache)
	}
	
	for i := range result {
		for resp := range result[i] {
			ch <- resp
		}
	}
	
	return
}

func main() {
	var ch = make(chan response)
	go Crawl("http://golang.org/", 4, fetcher, ch,
	         cache, Cache{visited: make(map[string] bool)})
	for resp := range ch {
		fmt.Printf("found: %s %q\n", resp.url, resp.body)
	}
}

type response struct {
	url string
	body string
}

// fakeFetcher is Fetcher that returns canned results.
type fakeFetcher map[string]*fakeResult

type fakeResult struct {
	body string
	urls []string
}

func (f fakeFetcher) Fetch(url string) (string, []string, error) {
	if res, ok := f[url]; ok {
		return res.body, res.urls, nil
	}
	return "", nil, fmt.Errorf("not found: %s", url)
}

// fetcher is a populated fakeFetcher.
var fetcher = fakeFetcher{
	"http://golang.org/": &fakeResult{
		"The Go Programming Language",
		[]string{
			"http://golang.org/pkg/",
			"http://golang.org/cmd/",
		},
	},
	"http://golang.org/pkg/": &fakeResult{
		"Packages",
		[]string{
			"http://golang.org/",
			"http://golang.org/cmd/",
			"http://golang.org/pkg/fmt/",
			"http://golang.org/pkg/os/",
		},
	},
	"http://golang.org/pkg/fmt/": &fakeResult{
		"Package fmt",
		[]string{
			"http://golang.org/",
			"http://golang.org/pkg/",
		},
	},
	"http://golang.org/pkg/os/": &fakeResult{
		"Package os",
		[]string{
			"http://golang.org/",
			"http://golang.org/pkg/",
		},
	},
}

Why does one Crawl have 5 arguments, and the other 6?

5 Here: go Crawl(u, depth-1, fetcher, result[i], cache)
5 Here: go Crawl("http://golang.org/", 4, fetcher, ch, cache, Cache{visited: make(map[string] bool)})

I think this code has 2 mistake:

1. function Crawl should defined like this: func Crawl(url string, depth int, fetcher Fetcher, ch chan response, cache *Cache), Crawl shouldn't passes lock by value: main.Cache contains sync.Mutex.
2. line 60 maybe is a slip of a pen, Crawl function should call like this: var cache = &Cache{visited: make(map[string] bool)} go Crawl("http://golang.org/", 4, fetcher, ch, cache)

The recursive call that defers closing the channel in the Crawl function is very interesting to me. As channels seem to behave like they are passed by reference, I thought that the calls to close() would result it a panic as the recursive calls returned, closing an already closed channel, but it didn't. The source code for go channels held the answer.

In the implementation, a channel is a struct called hchan which has many fields. The important fields are buf and closed. buf is a pointer to an array of the elements held in the channel. closed is an integer which I think indicates whether the channel is closed or not. Thus when the struct is passed to a function by value, both buf and closed are copied. Since buf is a pointer, a reference to the values to stored in the channel is retained, making the channel behave as if it were passed by reference. Since closed is an integer, this means you can close both the copied channel inside of the function and the original channel outside of the function without a panic. This is why the code above works like it does, allowing access to the same channel data, but allowing each recursive copy of the channel to be closed without a panic.

@cedarmora But in fact the channels passed to each recursive calls are distinct, made from make([]chan response). There will not be any channel closed twice.

I ended up writing something similar but I suppose it has a clearer logic. The above code looks correct to me through.

package main

import (
	"fmt"
	"sync"
)

type Fetcher interface {
	// Fetch returns the body of URL and
	// a slice of URLs found on that page.
	Fetch(url string) (body string, urls []string, err error)
}

// Crawl uses fetcher to recursively crawl
// pages starting with url, to a maximum of depth.
func Crawl(url string, depth int, fetcher Fetcher, c chan bool) {
	defer close(c)
	
	if depth <= 0 {
		return
	}
	
	mtx.Lock()
	found := cache[url]
	mtx.Unlock()
	
	if found {
		return
	}
	
	mtx.Lock()
	cache[url] = true
	mtx.Unlock()
	
	body, urls, err := fetcher.Fetch(url)
	if err != nil {
		fmt.Println(err)
		return
	}
	
	fmt.Printf("found: %s %q\n", url, body)
	
	var chans []chan bool

	for i, u := range urls {
		chans = append(chans, make(chan bool))
		go Crawl(u, depth-1, fetcher, chans[i])
	}
	
	for _, cc := range chans {
		<- cc
	}
	
	return
}

func main() {
	c := make(chan bool)
	go Crawl("https://golang.org/", 4, fetcher, c)
	<- c
}

// fakeFetcher is Fetcher that returns canned results.
type fakeFetcher map[string]*fakeResult

type fakeResult struct {
	body string
	urls []string
}

func (f fakeFetcher) Fetch(url string) (string, []string, error) {
	if res, ok := f[url]; ok {
		return res.body, res.urls, nil
	}
	return "", nil, fmt.Errorf("not found: %s", url)
}

var mtx sync.Mutex
var cache = make(map[string] bool)

// fetcher is a populated fakeFetcher.
var fetcher = fakeFetcher{
	"https://golang.org/": &fakeResult{
		"The Go Programming Language",
		[]string{
			"https://golang.org/pkg/",
			"https://golang.org/cmd/",
		},
	},
	"https://golang.org/pkg/": &fakeResult{
		"Packages",
		[]string{
			"https://golang.org/",
			"https://golang.org/cmd/",
			"https://golang.org/pkg/fmt/",
			"https://golang.org/pkg/os/",
		},
	},
	"https://golang.org/pkg/fmt/": &fakeResult{
		"Package fmt",
		[]string{
			"https://golang.org/",
			"https://golang.org/pkg/",
		},
	},
	"https://golang.org/pkg/os/": &fakeResult{
		"Package os",
		[]string{
			"https://golang.org/",
			"https://golang.org/pkg/",
		},
	},
}

with highlight

package main

import (
	"fmt"
	"sync"
)

type Fetcher interface {
	// Fetch returns the body of URL and
	// a slice of URLs found on that page.
	Fetch(url string) (body string, urls []string, err error)
}

// Crawl uses fetcher to recursively crawl
// pages starting with url, to a maximum of depth.
func Crawl(url string, depth int, fetcher Fetcher, c chan bool) {
	defer close(c)
	
	if depth <= 0 {
		return
	}
	
	mtx.Lock()
	found := cache[url]
	mtx.Unlock()
	
	if found {
		return
	}
	
	mtx.Lock()
	cache[url] = true
	mtx.Unlock()
	
	body, urls, err := fetcher.Fetch(url)
	if err != nil {
		fmt.Println(err)
		return
	}
	
	fmt.Printf("found: %s %q\n", url, body)
	
	var chans []chan bool

	for i, u := range urls {
		chans = append(chans, make(chan bool))
		go Crawl(u, depth-1, fetcher, chans[i])
	}
	
	for _, cc := range chans {
		<- cc
	}
	
	return
}

func main() {
	c := make(chan bool)
	go Crawl("https://golang.org/", 4, fetcher, c)
	<- c
}

// fakeFetcher is Fetcher that returns canned results.
type fakeFetcher map[string]*fakeResult

type fakeResult struct {
	body string
	urls []string
}

func (f fakeFetcher) Fetch(url string) (string, []string, error) {
	if res, ok := f[url]; ok {
		return res.body, res.urls, nil
	}
	return "", nil, fmt.Errorf("not found: %s", url)
}

var mtx sync.Mutex
var cache = make(map[string] bool)

// fetcher is a populated fakeFetcher.
var fetcher = fakeFetcher{
	"https://golang.org/": &fakeResult{
		"The Go Programming Language",
		[]string{
			"https://golang.org/pkg/",
			"https://golang.org/cmd/",
		},
	},
	"https://golang.org/pkg/": &fakeResult{
		"Packages",
		[]string{
			"https://golang.org/",
			"https://golang.org/cmd/",
			"https://golang.org/pkg/fmt/",
			"https://golang.org/pkg/os/",
		},
	},
	"https://golang.org/pkg/fmt/": &fakeResult{
		"Package fmt",
		[]string{
			"https://golang.org/",
			"https://golang.org/pkg/",
		},
	},
	"https://golang.org/pkg/os/": &fakeResult{
		"Package os",
		[]string{
			"https://golang.org/",
			"https://golang.org/pkg/",
		},
	},
}