Browse Source

Move extractURLS function

renovate/configure
Yota Toyama 8 years ago
parent
commit
fedffd1784
  1. 37
      file_checker.go
  2. 0
      file_checker_test.go
  3. 38
      main.go

37
file_checker.go

@ -75,3 +75,40 @@ func (r fileResult) Ok() bool {
return true
}
func extractURLs(n *html.Node) []string {
ss := make(map[string]bool)
ns := make([]*html.Node, 0, 1024)
ns = append(ns, n)
for len(ns) > 0 {
i := len(ns) - 1
n := ns[i]
ns = ns[:i]
if n.Type == html.ElementNode {
switch n.Data {
case "a":
for _, a := range n.Attr {
if a.Key == "href" && isURL(a.Val) {
ss[a.Val] = true
break
}
}
case "img":
for _, a := range n.Attr {
if a.Key == "src" && isURL(a.Val) {
ss[a.Val] = true
break
}
}
}
}
for n := n.FirstChild; n != nil; n = n.NextSibling {
ns = append(ns, n)
}
}
return stringSetToSlice(ss)
}

0
main_test.go → file_checker_test.go

38
main.go

@ -6,7 +6,6 @@ import (
"time"
"github.com/docopt/docopt-go"
"golang.org/x/net/html"
)
const usage = `Link checker for Markdown and HTML
@ -57,43 +56,6 @@ func main() {
}
}
func extractURLs(n *html.Node) []string {
ss := make(map[string]bool)
ns := make([]*html.Node, 0, 1024)
ns = append(ns, n)
for len(ns) > 0 {
i := len(ns) - 1
n := ns[i]
ns = ns[:i]
if n.Type == html.ElementNode {
switch n.Data {
case "a":
for _, a := range n.Attr {
if a.Key == "href" && isURL(a.Val) {
ss[a.Val] = true
break
}
}
case "img":
for _, a := range n.Attr {
if a.Key == "src" && isURL(a.Val) {
ss[a.Val] = true
break
}
}
}
}
for n := n.FirstChild; n != nil; n = n.NextSibling {
ns = append(ns, n)
}
}
return stringSetToSlice(ss)
}
func isURL(s string) bool {
u, err := url.Parse(s)
return err == nil && (u.Scheme == "http" || u.Scheme == "https")

Loading…
Cancel
Save