Merge pull request #17382 from caesarxuchao/rewrite-linkchecker

Auto commit by PR queue bot

Godeps/Godeps.json

@@ -762,6 +762,11 @@
 			"ImportPath": "github.com/mitchellh/mapstructure",
 			"Rev": "740c764bc6149d3f1806231418adb9f52c11bcbf"
 		},
+		{
+			"ImportPath": "github.com/mvdan/xurls",
+			"Comment": "v0.8.0-14-g1b768d7",
+			"Rev": "1b768d7c393abd8e8dda1458385a57becd4b2d4e"
+		},
 		{
 			"ImportPath": "github.com/mxk/go-flowrate/flowrate",
 			"Rev": "cca7078d478f8520f85629ad7c68962d31ed7682"

Godeps/LICENSES.md

@@ -61,6 +61,7 @@ github.com/matttproud/golang_protobuf_extensions | Apache-2
 github.com/mesos/mesos-go | Apache-2
 github.com/miekg/dns | spdxBSD3
 github.com/mitchellh/mapstructure | MITname
+github.com/mvdan/xurls | spdxBSD3
 github.com/mxk/go-flowrate | spdxBSD3
 github.com/onsi/ginkgo | spdxMIT
 github.com/onsi/gomega | spdxMIT

Godeps/_workspace/src/github.com/mvdan/xurls/.gitignore

@@ -0,0 +1,3 @@
+cmd/xurls/xurls
+generate/tldsgen/tldsgen
+generate/regexgen/regexgen

Godeps/_workspace/src/github.com/mvdan/xurls/.travis.yml

@@ -0,0 +1,5 @@
+language: go
+
+go:
+  - 1.4.3
+  - 1.5.1

Godeps/_workspace/src/github.com/mvdan/xurls/LICENSE

@@ -0,0 +1,27 @@
+Copyright (c) 2015, Daniel Martí. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   * Redistributions of source code must retain the above copyright
+notice, this list of conditions and the following disclaimer.
+   * Redistributions in binary form must reproduce the above
+copyright notice, this list of conditions and the following disclaimer
+in the documentation and/or other materials provided with the
+distribution.
+   * Neither the name of Google Inc. nor the names of its
+contributors may be used to endorse or promote products derived from
+this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Godeps/_workspace/src/github.com/mvdan/xurls/README.md

@@ -0,0 +1,31 @@
+# xurls
+
+[![GoDoc](https://godoc.org/github.com/mvdan/xurls?status.svg)](https://godoc.org/github.com/mvdan/xurls) [![Travis](https://travis-ci.org/mvdan/xurls.svg?branch=master)](https://travis-ci.org/mvdan/xurls)
+
+Extract urls from text using regular expressions.
+
+	go get github.com/mvdan/xurls
+
+```go
+import "github.com/mvdan/xurls"
+
+func main() {
+	xurls.Relaxed.FindString("Do gophers live in golang.org?")
+	// "golang.org"
+	xurls.Relaxed.FindAllString("foo.com is http://foo.com/.", -1)
+	// []string{"foo.com", "http://foo.com/"}
+	xurls.Strict.FindAllString("foo.com is http://foo.com/.", -1)
+	// []string{"http://foo.com/"}
+}
+```
+
+#### cmd/xurls
+
+Reads text and prints one url per line.
+
+	go get github.com/mvdan/xurls/cmd/xurls
+
+```shell
+$ echo "Do gophers live in http://golang.org?" | xurls
+http://golang.org
+```

Godeps/_workspace/src/github.com/mvdan/xurls/cmd/xurls/main.go

@@ -0,0 +1,83 @@
+// Copyright (c) 2015, Daniel Martí <mvdan@mvdan.cc>
+// See LICENSE for licensing information
+
+package main
+
+import (
+	"bufio"
+	"flag"
+	"fmt"
+	"os"
+	"regexp"
+
+	"github.com/mvdan/xurls"
+)
+
+var (
+	matching = flag.String("m", "", "")
+	relaxed  = flag.Bool("r", false, "")
+)
+
+func init() {
+	flag.Usage = func() {
+		p := func(format string, a ...interface{}) {
+			fmt.Fprintf(os.Stderr, format, a...)
+		}
+		p("Usage: xurls [-h] [files]\n\n")
+		p("If no files are given, it reads from standard input.\n\n")
+		p("   -m <regexp>   only match urls whose scheme matches a regexp\n")
+		p("                    example: 'https?://|mailto:'\n")
+		p("   -r            also match urls without a scheme (relaxed)\n")
+	}
+}
+
+func scanPath(re *regexp.Regexp, path string) error {
+	r := os.Stdin
+	if path != "-" {
+		f, err := os.Open(path)
+		if err != nil {
+			return err
+		}
+		defer f.Close()
+		r = f
+	}
+	scanner := bufio.NewScanner(r)
+	scanner.Split(bufio.ScanWords)
+	for scanner.Scan() {
+		word := scanner.Text()
+		for _, match := range re.FindAllString(word, -1) {
+			fmt.Println(match)
+		}
+	}
+	return scanner.Err()
+}
+
+func main() {
+	flag.Parse()
+	if *relaxed && *matching != "" {
+		errExit(fmt.Errorf("-r and -m at the same time don't make much sense"))
+	}
+	re := xurls.Strict
+	if *relaxed {
+		re = xurls.Relaxed
+	} else if *matching != "" {
+		var err error
+		if re, err = xurls.StrictMatchingScheme(*matching); err != nil {
+			errExit(err)
+		}
+	}
+	args := flag.Args()
+	if len(args) == 0 {
+		args = []string{"-"}
+	}
+	for _, path := range args {
+		if err := scanPath(re, path); err != nil {
+			errExit(err)
+		}
+	}
+}
+
+func errExit(err error) {
+	fmt.Fprintf(os.Stderr, "%v\n", err)
+	os.Exit(1)
+}

Godeps/_workspace/src/github.com/mvdan/xurls/generate/regexgen/main.go

@@ -0,0 +1,70 @@
+// Copyright (c) 2015, Daniel Martí <mvdan@mvdan.cc>
+// See LICENSE for licensing information
+
+package main
+
+import (
+	"log"
+	"os"
+	"sort"
+	"strings"
+	"text/template"
+
+	"golang.org/x/net/idna"
+
+	"github.com/mvdan/xurls"
+)
+
+const path = "regex.go"
+
+var regexTmpl = template.Must(template.New("regex").Parse(`// Generated by regexgen
+
+package xurls
+
+const ({{ range $key, $value := . }}
+	{{$key}} = ` + "`" + `{{$value}}` + "`" + `{{end}}
+)
+`))
+
+func writeRegex(tlds []string) error {
+	allTldsSet := make(map[string]struct{})
+	add := func(tld string) {
+		if _, e := allTldsSet[tld]; e {
+			log.Fatalf("Duplicate TLD: %s", tld)
+		}
+		allTldsSet[tld] = struct{}{}
+	}
+	for _, tldlist := range [...][]string{tlds, xurls.PseudoTLDs} {
+		for _, tld := range tldlist {
+			add(tld)
+			asciiTld, err := idna.ToASCII(tld)
+			if err != nil {
+				return err
+			}
+			if asciiTld != tld {
+				add(asciiTld)
+			}
+		}
+	}
+	var allTlds []string
+	for tld := range allTldsSet {
+		allTlds = append(allTlds, tld)
+	}
+	sort.Strings(allTlds)
+	f, err := os.Create(path)
+	if err != nil {
+		return err
+	}
+	defer f.Close()
+	return regexTmpl.Execute(f, map[string]string{
+		"gtld       ": `(?i)(` + strings.Join(allTlds, `|`) + `)(?-i)`,
+		"otherScheme": `(?i)(` + strings.Join(xurls.SchemesNoAuthority, `|`) + `)(?-i):`,
+	})
+}
+
+func main() {
+	log.Printf("Generating %s...", path)
+	if err := writeRegex(xurls.TLDs); err != nil {
+		log.Fatalf("Could not write %s: %v", path, err)
+	}
+}

Godeps/_workspace/src/github.com/mvdan/xurls/generate/tldsgen/main.go

@@ -0,0 +1,140 @@
+// Copyright (c) 2015, Daniel Martí <mvdan@mvdan.cc>
+// See LICENSE for licensing information
+
+package main
+
+import (
+	"bufio"
+	"errors"
+	"log"
+	"net/http"
+	"os"
+	"regexp"
+	"sort"
+	"strings"
+	"sync"
+	"text/template"
+)
+
+const path = "tlds.go"
+
+var tldsTmpl = template.Must(template.New("tlds").Parse(`// Generated by tldsgen
+
+package xurls
+
+// TLDs is a sorted list of all public top-level domains.
+//
+// Sources:{{range $_, $url := .URLs}}
+//  * {{$url}}{{end}}
+var TLDs = []string{
+{{range $_, $tld := .TLDs}}` + "\t`" + `{{$tld}}` + "`" + `,
+{{end}}}
+`))
+
+func cleanTld(tld string) string {
+	tld = strings.ToLower(tld)
+	if strings.HasPrefix(tld, "xn--") {
+		return ""
+	}
+	return tld
+}
+
+func fetchFromURL(url, pat string) {
+	defer wg.Done()
+	log.Printf("Fetching %s", url)
+	resp, err := http.Get(url)
+	if err == nil && resp.StatusCode >= 400 {
+		err = errors.New(resp.Status)
+	}
+	if err != nil {
+		errChan <- err
+		return
+	}
+	defer resp.Body.Close()
+	scanner := bufio.NewScanner(resp.Body)
+	re := regexp.MustCompile(pat)
+	for scanner.Scan() {
+		line := scanner.Text()
+		tld := re.FindString(line)
+		tld = cleanTld(tld)
+		if tld == "" {
+			continue
+		}
+		tldChan <- tld
+	}
+	if err := scanner.Err(); err != nil {
+		errChan <- err
+	}
+}
+
+var (
+	wg      sync.WaitGroup
+	tldChan = make(chan string)
+	errChan = make(chan error)
+)
+
+func tldList() ([]string, []string, error) {
+	var urls []string
+	fromURL := func(url, pat string) {
+		urls = append(urls, url)
+		wg.Add(1)
+		go fetchFromURL(url, pat)
+	}
+	fromURL("https://data.iana.org/TLD/tlds-alpha-by-domain.txt",
+		`^[^#]+$`)
+	fromURL("https://publicsuffix.org/list/effective_tld_names.dat",
+		`^[^/.]+$`)
+
+	tldSet := make(map[string]struct{})
+	anyError := false
+	go func() {
+		for {
+			select {
+			case tld := <-tldChan:
+				tldSet[tld] = struct{}{}
+			case err := <-errChan:
+				log.Printf("%v", err)
+				anyError = true
+			}
+		}
+	}()
+	wg.Wait()
+
+	if anyError {
+		return nil, nil, errors.New("there were some errors while fetching the TLDs")
+	}
+
+	tlds := make([]string, 0, len(tldSet))
+	for tld := range tldSet {
+		tlds = append(tlds, tld)
+	}
+
+	sort.Strings(tlds)
+	return tlds, urls, nil
+}
+
+func writeTlds(tlds, urls []string) error {
+	f, err := os.Create(path)
+	if err != nil {
+		return err
+	}
+	defer f.Close()
+	return tldsTmpl.Execute(f, struct {
+		TLDs []string
+		URLs []string
+	}{
+		TLDs: tlds,
+		URLs: urls,
+	})
+}
+
+func main() {
+	tlds, urls, err := tldList()
+	if err != nil {
+		log.Fatalf("Could not get TLD list: %v", err)
+	}
+	log.Printf("Generating %s...", path)
+	if err := writeTlds(tlds, urls); err != nil {
+		log.Fatalf("Could not write path: %v", err)
+	}
+}

Godeps/_workspace/src/github.com/mvdan/xurls/schemes.go

@@ -0,0 +1,14 @@
+package xurls
+
+// SchemesNoAuthority is a sorted list of some well-known url schemes that are
+// followed by ":" instead of "://". Since these are more prone to false
+// positives, we limit their matching.
+var SchemesNoAuthority = []string{
+	`bitcoin`, // Bitcoin
+	`file`,    // Files
+	`magnet`,  // Torrent magnets
+	`mailto`,  // Mail
+	`sms`,     // SMS
+	`tel`,     // Telephone
+	`xmpp`,    // XMPP
+}

Godeps/_workspace/src/github.com/mvdan/xurls/tlds_pseudo.go

@@ -0,0 +1,22 @@
+package xurls
+
+// PseudoTLDs is a sorted list of some widely used unofficial TLDs.
+//
+// Sources:
+//  * https://en.wikipedia.org/wiki/Pseudo-top-level_domain
+//  * https://en.wikipedia.org/wiki/Category:Pseudo-top-level_domains
+//  * https://tools.ietf.org/html/draft-grothoff-iesg-special-use-p2p-names-00
+//  * https://www.iana.org/assignments/special-use-domain-names/special-use-domain-names.xhtml
+var PseudoTLDs = []string{
+	`bit`,       // Namecoin
+	`example`,   // Example domain
+	`exit`,      // Tor exit node
+	`gnu`,       // GNS by public key
+	`i2p`,       // I2P network
+	`invalid`,   // Invalid domain
+	`local`,     // Local network
+	`localhost`, // Local network
+	`onion`,     // Tor hidden services
+	`test`,      // Test domain
+	`zkey`,      // GNS domain name
+}

Godeps/_workspace/src/github.com/mvdan/xurls/xurls.go

@@ -0,0 +1,66 @@
+// Copyright (c) 2015, Daniel Martí <mvdan@mvdan.cc>
+// See LICENSE for licensing information
+
+// Package xurls extracts urls from plain text using regular expressions.
+package xurls
+
+import "regexp"
+
+//go:generate go run generate/tldsgen/main.go
+//go:generate go run generate/regexgen/main.go
+
+const (
+	letter    = `\p{L}`
+	number    = `\p{N}`
+	iriChar   = letter + number
+	currency  = `\p{Sc}`
+	otherSymb = `\p{So}`
+	endChar   = iriChar + `/\-+_&~*%=#` + currency
+	midChar   = endChar + `@.,:;'?!|` + otherSymb
+	wellParen = `\([` + midChar + `]*(\([` + midChar + `]*\)[` + midChar + `]*)*\)`
+	wellBrack = `\[[` + midChar + `]*(\[[` + midChar + `]*\][` + midChar + `]*)*\]`
+	wellBrace = `\{[` + midChar + `]*(\{[` + midChar + `]*\}[` + midChar + `]*)*\}`
+	wellAll   = wellParen + `|` + wellBrack + `|` + wellBrace
+	pathCont  = `([` + midChar + `]*(` + wellAll + `|[` + endChar + `])+)+`
+	comScheme = `[a-zA-Z][a-zA-Z.\-+]*://`
+	scheme    = `(` + comScheme + `|` + otherScheme + `)`
+
+	iri      = `[` + iriChar + `]([` + iriChar + `\-]*[` + iriChar + `])?`
+	domain   = `(` + iri + `\.)+`
+	octet    = `(25[0-5]|2[0-4][0-9]|1[0-9]{2}|[1-9][0-9]|[0-9])`
+	ipv4Addr = `\b` + octet + `\.` + octet + `\.` + octet + `\.` + octet + `\b`
+	ipv6Addr = `([0-9a-fA-F]{1,4}:([0-9a-fA-F]{1,4}:([0-9a-fA-F]{1,4}:([0-9a-fA-F]{1,4}:([0-9a-fA-F]{1,4}:[0-9a-fA-F]{0,4}|:[0-9a-fA-F]{1,4})?|(:[0-9a-fA-F]{1,4}){0,2})|(:[0-9a-fA-F]{1,4}){0,3})|(:[0-9a-fA-F]{1,4}){0,4})|:(:[0-9a-fA-F]{1,4}){0,5})((:[0-9a-fA-F]{1,4}){2}|:(25[0-5]|(2[0-4]|1[0-9]|[1-9])?[0-9])(\.(25[0-5]|(2[0-4]|1[0-9]|[1-9])?[0-9])){3})|(([0-9a-fA-F]{1,4}:){1,6}|:):[0-9a-fA-F]{1,4}|([0-9a-fA-F]{1,4}:){7}:`
+	ipAddr   = `(` + ipv4Addr + `|` + ipv6Addr + `)`
+	site     = domain + gtld
+	hostName = `(` + site + `|` + ipAddr + `)`
+	port     = `(:[0-9]*)?`
+	path     = `(/|/` + pathCont + `?|\b|$)`
+	webURL   = hostName + port + path
+
+	strict  = `(\b` + scheme + pathCont + `)`
+	relaxed = `(` + strict + `|` + webURL + `)`
+)
+
+var (
+	// Relaxed matches all the urls it can find.
+	Relaxed = regexp.MustCompile(relaxed)
+	// Strict only matches urls with a scheme to avoid false positives.
+	Strict = regexp.MustCompile(strict)
+)
+
+func init() {
+	Relaxed.Longest()
+	Strict.Longest()
+}
+
+// StrictMatchingScheme produces a regexp that matches urls like Strict but
+// whose scheme matches the given regular expression.
+func StrictMatchingScheme(exp string) (*regexp.Regexp, error) {
+	strictMatching := `(\b(?i)(` + exp + `)(?-i)` + pathCont + `)`
+	re, err := regexp.Compile(strictMatching)
+	if err != nil {
+		return nil, err
+	}
+	re.Longest()
+	return re, nil
+}

Makefile

@@ -49,7 +49,6 @@ verify:
 	hack/verify-generated-deep-copies.sh
 	hack/verify-generated-docs.sh
 	hack/verify-swagger-spec.sh
-	hack/verify-linkcheck.sh
 	hack/verify-flags-underscore.py
 	hack/verify-godeps.sh $(BRANCH)
 	hack/verify-godep-licenses.sh $(BRANCH)

cmd/linkcheck/links.go

@@ -14,56 +14,155 @@ See the License for the specific language governing permissions and
 limitations under the License.
 */
 
-// This command checks if the hyperlinks in files are valid. It checks the files
-// with 'fileSuffix' in 'rootDir' for URLs that match 'prefix'. It trims the
-// 'prefix' from the URL, uses what's left as the relative path to repoRoot to
-// verify if the link is valid. For example:
-// $ linkcheck --root-dir=${TYPEROOT} --repo-root=${KUBE_ROOT} \
-// --file-suffix=types.go --prefix=http://releases.k8s.io/HEAD
+// This tool extracts the links from types.go and .md files, visits the link and
+// checks the status code of the response.
+// Usage:
+// $ linkcheck --root-dir=${ROOT}
 
 package main
 
 import (
 	"fmt"
 	"io/ioutil"
+	"net/http"
 	"os"
-	"path"
 	"path/filepath"
 	"regexp"
+	"strconv"
 	"strings"
+	"time"
 
+	"github.com/mvdan/xurls"
 	flag "github.com/spf13/pflag"
 )
 
 var (
-	httpRE *regexp.Regexp
-
 	rootDir    = flag.String("root-dir", "", "Root directory containing documents to be processed.")
-	repoRoot   = flag.String("repo-root", "", `Root directory of k8s repository.`)
-	fileSuffix = flag.String("file-suffix", "", "suffix of files to be checked")
-	prefix     = flag.String("prefix", "", "Longest common prefix of the link URL, e.g., http://release.k8s.io/HEAD/ for links in pkg/api/types.go")
+	fileSuffix = flag.StringSlice("file-suffix", []string{"types.go", ".md"}, "suffix of files to be checked")
+	// URLs matching the patterns in the regWhiteList won't be checked. Patterns
+	// of dummy URLs should be added to the list to avoid false alerts. Also,
+	// patterns of URLs that we don't care about can be added here to improve
+	// efficiency.
+	regWhiteList = []*regexp.Regexp{
+		regexp.MustCompile(`https://kubernetes-site\.appspot\.com`),
+		// skip url that doesn't start with an English alphabet, e.g., URLs with IP addresses.
+		regexp.MustCompile(`https?://[^A-Za-z].*`),
+		regexp.MustCompile(`https?://localhost.*`),
+	}
+	// URLs listed in the fullURLWhiteList won't be checked. This separated from
+	// the RegWhiteList to improve efficiency. This list includes dummy URLs that
+	// are hard to be generalized by a regex, and URLs that will cause false alerts.
+	fullURLWhiteList = map[string]struct{}{
+		"http://github.com/some/repo.git": {},
+		// This URL returns 404 when visited by this tool, but it works fine if visited by a browser.
+		"http://stackoverflow.com/questions/ask?tags=kubernetes":                                            {},
+		"https://github.com/$YOUR_GITHUB_USERNAME/kubernetes.git":                                           {},
+		"https://github.com/$YOUR_GITHUB_USERNAME/kubernetes":                                               {},
+		"http://storage.googleapis.com/kubernetes-release/release/v${K8S_VERSION}/bin/darwin/amd64/kubectl": {},
+		// It seems this server expects certain User-Agent value, it works fine with Chrome, but returns 404 if we issue a plain cURL to it.
+		"http://supervisord.org/":         {},
+		"http://kubernetes.io/vX.Y/docs":  {},
+		"http://kubernetes.io/vX.Y/docs/": {},
+		"http://kubernetes.io/vX.Y/":      {},
+	}
+
+	visitedURLs    = map[string]struct{}{}
+	htmlpreviewReg = regexp.MustCompile(`https://htmlpreview\.github\.io/\?`)
+	httpOrhttpsReg = regexp.MustCompile(`https?.*`)
 )
 
-func newWalkFunc(invalidLink *bool) filepath.WalkFunc {
+func newWalkFunc(invalidLink *bool, client *http.Client) filepath.WalkFunc {
 	return func(filePath string, info os.FileInfo, err error) error {
-		if !strings.HasSuffix(info.Name(), *fileSuffix) {
+		hasSuffix := false
+		for _, suffix := range *fileSuffix {
+			hasSuffix = hasSuffix || strings.HasSuffix(info.Name(), suffix)
+		}
+		if !hasSuffix {
 			return nil
 		}
+
 		fileBytes, err := ioutil.ReadFile(filePath)
 		if err != nil {
 			return err
 		}
 		foundInvalid := false
-		matches := httpRE.FindAllSubmatch(fileBytes, -1)
-		for _, match := range matches {
-			// match[1] should look like docs/devel/api-conventions.md
-			if _, err := os.Stat(path.Join(*repoRoot, string(match[1]))); err != nil {
-				fmt.Fprintf(os.Stderr, "Link is not valid: %s\n", string(match[0]))
+		allURLs := xurls.Strict.FindAll(fileBytes, -1)
+		fmt.Fprintf(os.Stdout, "\nChecking file %s\n", filePath)
+	URL:
+		for _, URL := range allURLs {
+			// Don't check non http/https URL
+			if !httpOrhttpsReg.Match(URL) {
+				continue
+			}
+			for _, whiteURL := range regWhiteList {
+				if whiteURL.Match(URL) {
+					continue URL
+				}
+			}
+			if _, found := fullURLWhiteList[string(URL)]; found {
+				continue
+			}
+			// remove the htmlpreview Prefix
+			processedURL := htmlpreviewReg.ReplaceAll(URL, []byte{})
+
+			// check if we have visited the URL.
+			if _, found := visitedURLs[string(processedURL)]; found {
+				continue
+			}
+			visitedURLs[string(processedURL)] = struct{}{}
+
+			retry := 0
+			const maxRetry int = 3
+			backoff := 100
+			for retry < maxRetry {
+				fmt.Fprintf(os.Stdout, "Visiting %s\n", string(processedURL))
+				// Use verb HEAD to increase efficiency. However, some servers
+				// do not handle HEAD well, so we need to try a GET to avoid
+				// false alert.
+				resp, err := client.Head(string(processedURL))
+				// URLs with mock host or mock port will cause error. If we report
+				// the error here, people need to add the mock URL to the white
+				// list every time they add a mock URL, which will be a maintenance
+				// nightmare. Hence, we decide to only report 404 to catch the
+				// cases where host and port are legit, but path is not, which
+				// is the most common mistake in our docs.
+				if err != nil {
+					break
+				}
+				if resp.StatusCode == 429 {
+					retryAfter := resp.Header.Get("Retry-After")
+					if seconds, err := strconv.Atoi(retryAfter); err != nil {
+						backoff = seconds + 10
+					}
+					fmt.Fprintf(os.Stderr, "Got %d visiting %s, retry after %d seconds.\n", resp.StatusCode, string(URL), backoff)
+					time.Sleep(time.Duration(backoff) * time.Second)
+					backoff *= 2
+					retry++
+				} else if resp.StatusCode == 404 {
+					// We only check for 404 error for now. 401, 403 errors are hard to handle.
+
+					// We need to try a GET to avoid false alert.
+					resp, err = client.Get(string(processedURL))
+					if err != nil {
+						break
+					}
+					if resp.StatusCode != 404 {
+						continue URL
+					}
+
+					foundInvalid = true
+					fmt.Fprintf(os.Stderr, "Failed: in file %s, Got %d visiting %s\n", filePath, resp.StatusCode, string(URL))
+					break
+				} else {
+					break
+				}
+			}
+			if retry == maxRetry {
 				foundInvalid = true
+				fmt.Fprintf(os.Stderr, "Failed: in file %s, still got 429 visiting %s after %d retries\n", filePath, string(URL), maxRetry)
 			}
 		}
 		if foundInvalid {
-			fmt.Fprintf(os.Stderr, "Found invalid links in %s\n", filePath)
 			*invalidLink = true
 		}
 		return nil
@@ -72,14 +171,16 @@ func newWalkFunc(invalidLink *bool) filepath.WalkFunc {
 
 func main() {
 	flag.Parse()
-	httpRE = regexp.MustCompile(*prefix + `(.*\.md)`)
 
-	if *rootDir == "" || *repoRoot == "" || *prefix == "" {
+	if *rootDir == "" {
 		flag.Usage()
 		os.Exit(2)
 	}
+	client := http.Client{
+		Timeout: time.Duration(5 * time.Second),
+	}
 	invalidLink := false
-	if err := filepath.Walk(*rootDir, newWalkFunc(&invalidLink)); err != nil {
+	if err := filepath.Walk(*rootDir, newWalkFunc(&invalidLink, &client)); err != nil {
 		fmt.Fprintf(os.Stderr, "Fail: %v.\n", err)
 		os.Exit(2)
 	}

docs/devel/how-to-doc.md

@@ -18,10 +18,6 @@
 If you are using a released version of Kubernetes, you should
 refer to the docs that go with that version.
 
-<strong>
-The latest 1.1.x release of this document can be found
-[here](http://releases.k8s.io/release-1.1/docs/devel/how-to-doc.md).
-
 Documentation for other releases can be found at
 [releases.k8s.io](http://releases.k8s.io).
 </strong>

docs/proposals/node-allocatable.md

@@ -65,7 +65,7 @@ reservation grows), or running multiple Kubelets on a single node.
 ![image](node-allocatable.png)
 
 1. **Node Capacity** - Already provided as
-   [`NodeStatus.Capacity`](https://htmlpreview.github.io/?https://github.com/kubernetes/kubernetes/HEAD/docs/api-reference/v1/definitions.html#_v1_nodestatus),
+   [`NodeStatus.Capacity`](https://htmlpreview.github.io/?https://github.com/kubernetes/kubernetes/blob/HEAD/docs/api-reference/v1/definitions.html#_v1_nodestatus),
    this is total capacity read from the node instance, and assumed to be constant.
 2. **System-Reserved** (proposed) - Compute resources reserved for processes which are not managed by
    Kubernetes. Currently this covers all the processes lumped together in the `/system` raw
@@ -81,7 +81,7 @@ reservation grows), or running multiple Kubelets on a single node.
 #### Allocatable
 
 Add `Allocatable` (4) to
-[`NodeStatus`](https://htmlpreview.github.io/?https://github.com/kubernetes/kubernetes/HEAD/docs/api-reference/v1/definitions.html#_v1_nodestatus):
+[`NodeStatus`](https://htmlpreview.github.io/?https://github.com/kubernetes/kubernetes/blob/HEAD/docs/api-reference/v1/definitions.html#_v1_nodestatus):
 
 ```
 type NodeStatus struct {

hack/after-build/verify-linkcheck.sh

@@ -24,15 +24,36 @@ source "${KUBE_ROOT}/hack/lib/init.sh"
 kube::golang::setup_env
 linkcheck=$(kube::util::find-binary "linkcheck")
 
-TYPEROOT="${KUBE_ROOT}/pkg/api/"
-"${linkcheck}" "--root-dir=${TYPEROOT}" "--repo-root=${KUBE_ROOT}" "--file-suffix=types.go" "--prefix=http://releases.k8s.io/HEAD" && ret=0 || ret=$?
-if [[ $ret -eq 1 ]]; then
-  echo "links in ${TYPEROOT} is out of date."
-  exit 1
-fi
-if [[ $ret -gt 1 ]]; then
-  echo "Error running linkcheck"
-  exit 1
+kube::util::ensure-temp-dir
+OUTPUT="${KUBE_TEMP}"/linkcheck-output
+cleanup() {
+	rm -rf "${OUTPUT}"
+}
+trap "cleanup" EXIT SIGINT
+mkdir -p "$OUTPUT"
+
+APIROOT="${KUBE_ROOT}/pkg/api/"
+APISROOT="${KUBE_ROOT}/pkg/apis/"
+DOCROOT="${KUBE_ROOT}/docs/"
+ROOTS=($APIROOT $APISROOT $DOCROOT)
+found_invalid=false
+for root in "${ROOTS[@]}"; do
+  "${linkcheck}" "--root-dir=${root}" 2> >(tee -a "${OUTPUT}/error" >&2) && ret=0 || ret=$?
+  if [[ $ret -eq 1 ]]; then
+    echo "Failed: found invalid links in ${root}."
+    found_invalid=true
+  fi
+  if [[ $ret -gt 1 ]]; then
+    echo "Error running linkcheck"
+    exit 1
+  fi
+done
+
+if [ ${found_invalid} = true ]; then
+  echo "Summary of invalid links:"
+  cat ${OUTPUT}/error
 fi
 
+trap "cleanup" EXIT SIGINT
+
 # ex: ts=2 sw=2 et filetype=sh

hack/verify-all.sh

@@ -60,7 +60,7 @@ if $SILENT ; then
 fi
 
 # remove protobuf until it is part of direct generation
-EXCLUDE="verify-godeps.sh verify-godep-licenses.sh verify-generated-protobuf.sh"
+EXCLUDE="verify-godeps.sh verify-godep-licenses.sh verify-generated-protobuf.sh verify-linkcheck.sh"
 
 ret=0
 for t in `ls $KUBE_ROOT/hack/verify-*.sh`

hooks/pre-commit

@@ -122,16 +122,6 @@ else
 fi
 echo "${reset}"
 
-echo -ne "Checking for links in API descriptions... "
-if ! hack/after-build/verify-linkcheck.sh > /dev/null; then
-  echo "${red}ERROR!"
-  echo "Some links in pkg/api/.*types.go are outdated. They require a manual fix."
-  exit_code=1
-else
-  echo "${green}OK"
-fi
-echo "${reset}"
-
 echo -ne "Checking for docs that need updating... "
 if ! hack/after-build/verify-generated-docs.sh > /dev/null; then
   echo "${red}ERROR!"

shippable.yml

@@ -56,7 +56,6 @@ install:
   - ./hack/verify-generated-docs.sh
   - ./hack/verify-generated-swagger-docs.sh
   - ./hack/verify-swagger-spec.sh
-  - ./hack/verify-linkcheck.sh
 
 script:
   # Disable coverage collection on pull requests