imagorvideoextended/processor.go

package imagorvideoextended

import (
	"context"
	"crypto/sha256"
	"github.com/antchfx/htmlquery"
	"github.com/cshum/imagor"
	"github.com/cshum/imagor/imagorpath"
	"github.com/gabriel-vasile/mimetype"
	"go.uber.org/zap"
	"io"
	"net/url"
	"os"
	"regexp"
	"strings"
)

// Processor for imagorvideo that implements imagor.Processor interface
type Processor struct {
	Logger        *zap.Logger
	Debug         bool
	FallbackImage string
}

// NewProcessor creates Processor
func NewProcessor(options ...Option) *Processor {
	p := &Processor{
		Logger: zap.NewNop(),
	}
	for _, option := range options {
		option(p)
	}
	return p
}

// Startup implements imagor.Processor interface
func (p *Processor) Startup(_ context.Context) error {
	return nil
}

// Shutdown implements imagor.Processor interface
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{
		Image:  url,
		FitIn:  true,
		Width:  400,
		Height: 400,
	}
	path := imagorpath.Generate(params, imagorpath.NewHMACSigner(sha256.New, 0, key))
	return path
}

// Process implements imagor.Processor interface
func (p *Processor) Process(ctx context.Context, in *imagor.Blob, params imagorpath.Params, load imagor.LoadFunc) (out *imagor.Blob, err error) {
	defer func() {
		if err == nil || out != nil {
			return
		}
		if _, ok := err.(imagor.ErrForward); ok {
			return
		}
		err = imagor.NewError(err.Error(), 406)
		// fallback image on error
		out = imagor.NewBlobFromBytes(transPixel)
		if p.FallbackImage != "" {
			if o, e := load(p.FallbackImage); e == nil {
				out = o
			}
		}
	}()

	var mime = mimetype.Detect(in.Sniff())

	if typ := mime.String(); !strings.HasPrefix(typ, "text/html") {
		// forward identical for non video nor audio
		err = imagor.ErrForward{Params: params}
		out = in
		return
	}

	rs, size, err := in.NewReadSeeker()
	if err != nil {
		return
	}
	defer func() {
		_ = rs.Close()
	}()
	if size <= 0 && rs != nil {
		// size must be known
		if size, err = rs.Seek(0, io.SeekEnd); err != nil {
			return
		}
		if _, err = rs.Seek(0, io.SeekStart); err != nil {
			return
		}
	}
	all, err := in.ReadAll()
	if err != nil {
		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 {
		var property = htmlquery.SelectAttr(metaTag, "property")
		var val = htmlquery.SelectAttr(metaTag, "content")

		switch property {
		case "og:image":
			fallthrough
		case "twitter:image:src":
			meta.Image = subThumbnail(val)
			break

		case "og:title":
			fallthrough
		case "twitter:title":
			meta.Title = val
			break

		case "twitter:description":
			fallthrough
		case "og:description":
			meta.Description = val
		}
	}
	if meta.Title == "" {
		title := htmlquery.FindOne(doc, "//title")
		if title != nil {
			meta.Title = htmlquery.InnerText(title)
		} else {
			meta.Title = in.FilePath()
		}
	}

	out = imagor.NewBlobFromJsonMarshal(meta)
	return
}

// Metadata imagorvideo metadata
type Metadata struct {
	Format         string            `json:"format"`
	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")