155 lines
3.6 KiB
Go
155 lines
3.6 KiB
Go
// Copyright 2017 Frédéric Guillot. All rights reserved.
|
|
// Use of this source code is governed by the Apache 2.0
|
|
// license that can be found in the LICENSE file.
|
|
|
|
package icon
|
|
|
|
import (
|
|
"encoding/base64"
|
|
"fmt"
|
|
"io"
|
|
"io/ioutil"
|
|
"strings"
|
|
|
|
"github.com/miniflux/miniflux/crypto"
|
|
"github.com/miniflux/miniflux/http"
|
|
"github.com/miniflux/miniflux/logger"
|
|
"github.com/miniflux/miniflux/model"
|
|
"github.com/miniflux/miniflux/url"
|
|
|
|
"github.com/PuerkitoBio/goquery"
|
|
)
|
|
|
|
// FindIcon try to find the website's icon.
|
|
func FindIcon(websiteURL string) (*model.Icon, error) {
|
|
rootURL := url.RootURL(websiteURL)
|
|
client := http.NewClient(rootURL)
|
|
response, err := client.Get()
|
|
if err != nil {
|
|
return nil, fmt.Errorf("unable to download website index page: %v", err)
|
|
}
|
|
|
|
if response.HasServerFailure() {
|
|
return nil, fmt.Errorf("unable to download website index page: status=%d", response.StatusCode)
|
|
}
|
|
|
|
iconURL, err := parseDocument(rootURL, response.Body)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
if strings.HasPrefix(iconURL, "data:") {
|
|
return parseImageDataURL(iconURL)
|
|
}
|
|
|
|
logger.Debug("[FindIcon] Fetching icon => %s", iconURL)
|
|
icon, err := downloadIcon(iconURL)
|
|
if err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
return icon, nil
|
|
}
|
|
|
|
func parseDocument(websiteURL string, data io.Reader) (string, error) {
|
|
queries := []string{
|
|
"link[rel='shortcut icon']",
|
|
"link[rel='Shortcut Icon']",
|
|
"link[rel='icon shortcut']",
|
|
"link[rel='icon']",
|
|
}
|
|
|
|
doc, err := goquery.NewDocumentFromReader(data)
|
|
if err != nil {
|
|
return "", fmt.Errorf("unable to read document: %v", err)
|
|
}
|
|
|
|
var iconURL string
|
|
for _, query := range queries {
|
|
doc.Find(query).Each(func(i int, s *goquery.Selection) {
|
|
if href, exists := s.Attr("href"); exists {
|
|
iconURL = href
|
|
}
|
|
})
|
|
|
|
if iconURL != "" {
|
|
break
|
|
}
|
|
}
|
|
|
|
if iconURL == "" {
|
|
iconURL = url.RootURL(websiteURL) + "favicon.ico"
|
|
} else {
|
|
iconURL, _ = url.AbsoluteURL(websiteURL, iconURL)
|
|
}
|
|
|
|
return iconURL, nil
|
|
}
|
|
|
|
func downloadIcon(iconURL string) (*model.Icon, error) {
|
|
client := http.NewClient(iconURL)
|
|
response, err := client.Get()
|
|
if err != nil {
|
|
return nil, fmt.Errorf("unable to download iconURL: %v", err)
|
|
}
|
|
|
|
if response.HasServerFailure() {
|
|
return nil, fmt.Errorf("unable to download icon: status=%d", response.StatusCode)
|
|
}
|
|
|
|
body, err := ioutil.ReadAll(response.Body)
|
|
if err != nil {
|
|
return nil, fmt.Errorf("unable to read downloaded icon: %v", err)
|
|
}
|
|
|
|
if len(body) == 0 {
|
|
return nil, fmt.Errorf("downloaded icon is empty, iconURL=%s", iconURL)
|
|
}
|
|
|
|
icon := &model.Icon{
|
|
Hash: crypto.HashFromBytes(body),
|
|
MimeType: response.ContentType,
|
|
Content: body,
|
|
}
|
|
|
|
return icon, nil
|
|
}
|
|
|
|
func parseImageDataURL(value string) (*model.Icon, error) {
|
|
colon := strings.Index(value, ":")
|
|
semicolon := strings.Index(value, ";")
|
|
comma := strings.Index(value, ",")
|
|
|
|
if colon <= 0 || semicolon <= 0 || comma <= 0 {
|
|
return nil, fmt.Errorf(`icon: invalid data url "%s"`, value)
|
|
}
|
|
|
|
mimeType := value[colon+1 : semicolon]
|
|
encoding := value[semicolon+1 : comma]
|
|
data := value[comma+1:]
|
|
|
|
if encoding != "base64" {
|
|
return nil, fmt.Errorf(`icon: unsupported data url encoding "%s"`, value)
|
|
}
|
|
|
|
if !strings.HasPrefix(mimeType, "image/") {
|
|
return nil, fmt.Errorf(`icon: invalid mime type "%s"`, mimeType)
|
|
}
|
|
|
|
blob, err := base64.StdEncoding.DecodeString(data)
|
|
if err != nil {
|
|
return nil, fmt.Errorf(`icon: invalid data "%s" (%v)`, value, err)
|
|
}
|
|
|
|
if len(blob) == 0 {
|
|
return nil, fmt.Errorf(`icon: empty data "%s"`, value)
|
|
}
|
|
|
|
icon := &model.Icon{
|
|
Hash: crypto.HashFromBytes(blob),
|
|
Content: blob,
|
|
MimeType: mimeType,
|
|
}
|
|
|
|
return icon, nil
|
|
}
|