get meta from html page

server.go

package main

import (
	"encoding/json"
	"net/http"
	"net/url"

	"golang.org/x/net/html"
	"io"
)

func main() {

	http.HandleFunc(`/read`, func(rw http.ResponseWriter, req *http.Request) {
		rw.Header().Set(`Content-Type`, `application/json`)

		err := req.ParseForm()
		if err != nil {
			rw.WriteHeader(http.StatusBadRequest)
			json.NewEncoder(rw).Encode(map[string]string{"error": err.Error()})
			return
		}

		link := req.FormValue(`link`)
		if link == "" {
			rw.WriteHeader(http.StatusBadRequest)
			json.NewEncoder(rw).Encode(map[string]string{"error": `empty value of link`})
			return
		}

		if _, err := url.Parse(link); err != nil {
			rw.WriteHeader(http.StatusBadRequest)
			json.NewEncoder(rw).Encode(map[string]string{"error": err.Error()})
			return
		}

		resp, err := http.Get(link)
		if err != nil {
			//proxy status and err
			rw.WriteHeader(resp.StatusCode)
			json.NewEncoder(rw).Encode(map[string]string{"error": err.Error()})
			return
		}
		defer resp.Body.Close()

		meta := extract(resp.Body)
		rw.WriteHeader(http.StatusOK)
		json.NewEncoder(rw).Encode(meta)
		return
	})

	// little help %)
	println("call like: \n$ curl -XPOST 'http://localhost:4567/read' -d link='https://github.com/golang/go'")
	println(`{"title":"golang/go","description":"go - The Go programming language","image":"https://avatars1.githubusercontent.com/u/4314092?v=3\u0026s=400","site_name":"GitHub"}`)

	err := http.ListenAndServe(`:4567`, nil)
	if err != nil {
		panic(err)
	}

}

type HTMLMeta struct {
	Title       string `json:"title"`
	Description string `json:"description"`
	Image       string `json:"image"`
	SiteName    string `json:"site_name"`
}

func extract(resp io.Reader) *HTMLMeta {
	z := html.NewTokenizer(resp)

	titleFound := false

	hm := new(HTMLMeta)

	for {
		tt := z.Next()
		switch tt {
		case html.ErrorToken:
			return hm
		case html.StartTagToken, html.SelfClosingTagToken:
			t := z.Token()
			if t.Data == `body` {
				return hm
			}
			if t.Data == "title" {
				titleFound = true
			}
			if t.Data == "meta" {
				desc, ok := extractMetaProperty(t, "description")
				if ok {
					hm.Description = desc
				}

				ogTitle, ok := extractMetaProperty(t, "og:title")
				if ok {
					hm.Title = ogTitle
				}

				ogDesc, ok := extractMetaProperty(t, "og:description")
				if ok {
					hm.Description = ogDesc
				}

				ogImage, ok := extractMetaProperty(t, "og:image")
				if ok {
					hm.Image = ogImage
				}

				ogSiteName, ok := extractMetaProperty(t, "og:site_name")
				if ok {
					hm.SiteName = ogSiteName
				}
			}
		case html.TextToken:
			if titleFound {
				t := z.Token()
				hm.Title = t.Data
				titleFound = false
			}
		}
	}
	return hm
}

func extractMetaProperty(t html.Token, prop string) (content string, ok bool) {
	for _, attr := range t.Attr {
		if attr.Key == "property" && attr.Val == prop {
			ok = true
		}

		if attr.Key == "content" {
			content = attr.Val
		}
	}

	return
}

Very clean and impressive 👍

👍