youtube handling + special site handling framework

cmd/imagorvideo/main.go

@@ -13,6 +13,7 @@ import (
 func main() {
 	var server = config.CreateServer(
 		os.Args[1:],
+		//imagorvideoextended.LoaderConfig,
 		imagorvideoextended.Config,
 		imagorvideo.Config,
 		vipsconfig.WithVips,

config.go

@@ -6,6 +6,10 @@ import (
 	"go.uber.org/zap"
 )
 
+func LoaderConfig(fs *flag.FlagSet, cb func() (*zap.Logger, bool)) imagor.Option {
+	return imagor.WithLoaders(NewSpecialLoader())
+}
+
 // Config imagorvideo config.Option
 func Config(fs *flag.FlagSet, cb func() (*zap.Logger, bool)) imagor.Option {
 	var (

loader.go

@@ -0,0 +1,43 @@
+package imagorvideoextended
+
+import (
+	"github.com/cshum/imagor"
+	"net/http"
+	"net/url"
+	"strings"
+)
+
+/*
+	This was started to handle YouTube, however it turns out they leak just enough info to be useful in the processing
+	phase. This can still be used for sites that have more info in the URL than the connection, though.
+	Currently unloaded to eke out some more perf
+*/
+
+type LoaderOption func(h *SpecialLoader)
+
+// SpecialLoader HTTP Loader implements imagor.Loader interface
+type SpecialLoader struct {
+}
+
+func NewSpecialLoader(options ...LoaderOption) *SpecialLoader {
+	loader := &SpecialLoader{}
+	for _, option := range options {
+		option(loader)
+	}
+	return loader
+}
+
+func (loader *SpecialLoader) Get(_ *http.Request, key string) (*imagor.Blob, error) {
+	components, err := url.Parse(key)
+	if err != nil {
+		return nil, imagor.ErrInvalid
+	}
+	hostname := components.Hostname()
+	if trimmed, ok := strings.CutPrefix(hostname, "www."); ok {
+		hostname = trimmed
+	}
+	switch components.Hostname() {
+	default:
+		return nil, imagor.ErrInvalid
+	}
+}

processor.go

@@ -9,7 +9,9 @@ import (
 	"github.com/gabriel-vasile/mimetype"
 	"go.uber.org/zap"
 	"io"
+	"net/url"
 	"os"
+	"regexp"
 	"strings"
 )
 
@@ -41,6 +43,37 @@ func (p *Processor) Shutdown(_ context.Context) error {
 	return nil
 }
 
+var youtubeIdRegex = regexp.MustCompile("[^\"&?\\/\\s\\.=]{11}")
+
+func specialUrl(path string) (hostname string, specialHandler string, specialData map[string]string, err error) {
+	specialData = make(map[string]string)
+	components, err := url.Parse(path)
+
+	if err != nil {
+		return
+	}
+	hostname = components.Host
+	if trimmed, ok := strings.CutPrefix(hostname, "www."); ok {
+		hostname = trimmed
+	}
+
+	switch hostname {
+	case "youtu.be":
+		fallthrough
+	case "youtube.com":
+		fallthrough
+	case "m.youtube.com":
+		id := youtubeIdRegex.FindString(path)
+		println(id)
+		if id != "" {
+			specialHandler = "youtube"
+			specialData["id"] = id
+		}
+	}
+
+	return hostname, specialHandler, specialData, nil
+}
+
 func subThumbnail(url string) string {
 	key := os.Getenv("IMAGOR_SECRET")
 	params := imagorpath.Params{
@@ -102,12 +135,17 @@ func (p *Processor) Process(ctx context.Context, in *imagor.Blob, params imagorp
 		return nil, err
 	}
 
+	hostname, specialHandler, specialData, _ := specialUrl(params.Image)
+
 	doc, err := htmlquery.Parse(strings.NewReader(string(all[:])))
 	meta := Metadata{
 		Format:         strings.TrimPrefix(mime.Extension(), "."),
 		Title:          "",
 		Description:    "",
 		Image:          "",
+		Hostname:       hostname,
+		SpecialHandler: specialHandler,
+		SpecialData:    specialData,
 	}
 	metaTags := htmlquery.Find(doc, "//meta[@property]")
 	for _, metaTag := range metaTags {
@@ -152,6 +190,9 @@ type Metadata struct {
 	Title          string            `json:"title"`
 	Description    string            `json:"description"`
 	Image          string            `json:"image"`
+	Hostname       string            `json:"hostname"`
+	SpecialHandler string            `json:"special_handler"`
+	SpecialData    map[string]string `json:"special_data"`
 }
 
 var transPixel = []byte("\x47\x49\x46\x38\x39\x61\x01\x00\x01\x00\x80\x00\x00\x00\x00\x00\x00\x00\x00\x21\xF9\x04\x01\x00\x00\x00\x00\x2C\x00\x00\x00\x00\x01\x00\x01\x00\x00\x02\x02\x44\x01\x00\x3B")