func BenchmarkOpenGraph_ProcessHTML(b *testing.B) {
og := opengraph.NewOpenGraph()
b.ReportAllocs()
b.SetBytes(int64(len(html)))
for i := 0; i < b.N; i++ {
if err := og.ProcessHTML(strings.NewReader(html)); err != nil {
b.Fatal(err)
}
}
}
func TestOpenGraphProcessHTML(t *testing.T) {
og := opengraph.NewOpenGraph()
err := og.ProcessHTML(strings.NewReader(html))
if err != nil {
t.Fatal(err)
}
if og.Type != "article" {
t.Error("type parsed incorrectly")
}
if len(og.Title) == 0 {
t.Error("title parsed incorrectly")
}
if len(og.URL) == 0 {
t.Error("url parsed incorrectly")
}
if len(og.Description) == 0 {
t.Error("description parsed incorrectly")
}
if len(og.Images) == 0 {
t.Error("images parsed incorrectly")
} else {
if len(og.Images[0].URL) == 0 {
t.Error("image url parsed incorrectly")
}
}
if len(og.Locale) == 0 {
t.Error("locale parsed incorrectly")
}
if len(og.SiteName) == 0 {
t.Error("site name parsed incorrectly")
}
if og.Article == nil {
t.Error("articles parsed incorrectly")
} else {
ev, _ := time.Parse(time.RFC3339, "2015-08-18T19:12:38+00:00")
if !og.Article.PublishedTime.Equal(ev) {
t.Error("article published time parsed incorrectly")
}
}
}
func main() {
h := `<html><head><meta property="og:type" content="article" />
<meta property="og:title" content="WordPress 4.3 "Billie"" />
<meta property="og:url" content="https://wordpress.org/news/2015/08/billie/" /></head><body></body></html>`
og := opengraph.NewOpenGraph()
doc, err := html.Parse(strings.NewReader(h))
if err != nil {
fmt.Println(err)
return
}
var parseHead func(*html.Node)
parseHead = func(n *html.Node) {
for c := n.FirstChild; c != nil; c = c.NextSibling {
if c.Type == html.ElementNode && c.Data == "meta" {
m := make(map[string]string)
for _, a := range c.Attr {
m[a.Key] = a.Val
}
og.ProcessMeta(m)
}
}
}
var f func(*html.Node)
f = func(n *html.Node) {
for c := n.FirstChild; c != nil; c = c.NextSibling {
if c.Type == html.ElementNode {
if c.Data == "head" {
parseHead(c)
continue
} else if c.Data == "body" { // OpenGraph is only in head, so we don't need body
break
}
}
f(c)
}
}
f(doc)
fmt.Printf("Type: %s\n", og.Type)
fmt.Printf("Title: %s\n", og.Title)
fmt.Printf("URL: %s\n", og.URL)
fmt.Printf("String/JSON Representation: %s\n", og)
}
func main() {
html := `<html><head><meta property="og:type" content="article" />
<meta property="og:title" content="WordPress 4.3 "Billie"" />
<meta property="og:url" content="https://wordpress.org/news/2015/08/billie/" /></head><body></body></html>`
og := opengraph.NewOpenGraph()
err := og.ProcessHTML(strings.NewReader(html))
if err != nil {
fmt.Println(err)
return
}
fmt.Printf("Type: %s\n", og.Type)
fmt.Printf("Title: %s\n", og.Title)
fmt.Printf("URL: %s\n", og.URL)
fmt.Printf("String/JSON Representation: %s\n", og)
}
func TestOpenGraphProcessMeta(t *testing.T) {
og := opengraph.NewOpenGraph()
og.ProcessMeta(map[string]string{"property": "og:type", "content": "book"})
if og.Type != "book" {
t.Error("wrong og:type processing")
}
og.ProcessMeta(map[string]string{"property": "book:isbn", "content": "123456"})
if og.Book == nil {
t.Error("wrong book type processing")
} else {
if og.Book.ISBN != "123456" {
t.Error("wrong book isbn processing")
}
}
og.ProcessMeta(map[string]string{"property": "article:section", "content": "testsection"})
if og.Article != nil {
t.Error("article processed when it should not be")
}
og.ProcessMeta(map[string]string{"property": "book:author:first_name", "content": "John"})
if og.Book != nil {
if len(og.Book.Authors) == 0 {
t.Error("book author was not processed")
} else {
if og.Book.Authors[0].FirstName != "John" {
t.Error("author first name was processed incorrectly")
}
}
}
}
// NewHTMLInfo return new instance of HTMLInfo
func NewHTMLInfo() *HTMLInfo {
info := &HTMLInfo{AllowOembedFetching: true, AllowMainContentExtraction: true, OGInfo: opengraph.NewOpenGraph(), AcceptLanguage: "en-us"}
return info
}