Start adding ahash fingerprinting support

go.mod

@@ -28,6 +28,7 @@ require (
 
 require (
 	github.com/davecgh/go-spew v1.1.1 // indirect
+	github.com/devedge/imagehash v0.0.0-20180324030135-7061aa3b4066 // indirect
 	github.com/kr/pretty v0.3.1 // indirect
 	github.com/mattn/go-colorable v0.1.13 // indirect
 	github.com/mattn/go-isatty v0.0.20 // indirect

go.sum

@@ -29,6 +29,8 @@ github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ3
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
+github.com/devedge/imagehash v0.0.0-20180324030135-7061aa3b4066 h1:+QbuEqZjC9bIWKkf73EQ5oaIA7g/lB6fE7uFuEV0SeY=
+github.com/devedge/imagehash v0.0.0-20180324030135-7061aa3b4066/go.mod h1:FdoOQDHSR0xYTQCl+G8ZhqsB5dKYbyxVWUoUFW7F0Lw=
 github.com/disintegration/imaging v1.6.2 h1:w1LecBlG2Lnp8B3jk5zSuNqd7b4DXhcjwek1ei82L+c=
 github.com/disintegration/imaging v1.6.2/go.mod h1:44/5580QXChDfwIclfc/PCwrr44amcmDAg8hxG0Ewe4=
 github.com/dustin/go-humanize v1.0.0/go.mod h1:HtrtbFcZ19U5GC7JDqmcUSB87Iq5E25KnS6fMYU6eOk=

pkg/config/config.go

@@ -186,6 +186,8 @@ type SiteConfig struct {
 	GeoIPType    string
 	GeoIPOptions map[string]any
 	Captcha      CaptchaConfig
+
+	FingerprinterOptions map[string]map[string]any
 }
 
 type CaptchaConfig struct {

pkg/gcsql/tables.go

@@ -54,12 +54,15 @@ type Board struct {
 // FileBan contains the information associated with a specific file ban.
 // table: DBPREFIXfile_ban
 type FileBan struct {
-	ID        int       // sql: `id`
-	BoardID   *int      // sql: `board_id`
-	StaffID   int       // sql: `staff_id`
-	StaffNote string    // sql: `staff_note`
-	IssuedAt  time.Time // sql: `issued_at`
-	Checksum  string    // sql: `checksum`
+	ID            int       // sql: `id`
+	BoardID       *int      // sql: `board_id`
+	StaffID       int       // sql: `staff_id`
+	StaffNote     string    // sql: `staff_note`
+	IssuedAt      time.Time // sql: `issued_at`
+	Checksum      string    // sql: `checksum`
+	Fingerprinter *string   // sql: `fingerprinter`
+	BanIP         bool      // sql: `ban_ip`
+	BanIPMessage  *string   // sql: `ban_ip_message`
 }
 
 type filenameOrUsernameBanBase struct {

pkg/posting/uploads/fingerprinting/ahash.go

@@ -0,0 +1,105 @@
+package fingerprinting
+
+import (
+	"fmt"
+	"image"
+	"path"
+	"strings"
+
+	"github.com/devedge/imagehash"
+	"github.com/gochan-org/gochan/pkg/gcsql"
+)
+
+type ahashHandler struct {
+	hashVideoThumb bool // if true and the upload is a video, hash the thumb
+	hashLength     int
+}
+
+func (ah *ahashHandler) Init(options map[string]any) error {
+	var ok bool
+	for key, val := range options {
+		switch strings.ToLower(key) {
+		case "hashvideothumb":
+			fallthrough
+		case "hashvideothumbnail":
+			ah.hashVideoThumb, ok = val.(bool)
+			if !ok {
+				return fmt.Errorf("invalid value type for %q, expected boolean, got %T", key, val)
+			}
+		case "hashlength":
+			ah.hashLength, ok = val.(int)
+			if !ok {
+				return fmt.Errorf("invalid value type for %q, expected voolean, got %T", key, val)
+			}
+		}
+	}
+	if ah.hashLength < 1 {
+		ah.hashLength = defaultHashLength
+	}
+	return nil
+}
+
+func (ah *ahashHandler) getImage(source *FingerprintSource) (image.Image, error) {
+	if source.Img == nil {
+		if source.FilePath == "" {
+			file, _, err := source.Request.FormFile("imagefile")
+			if err != nil {
+				return nil, err
+			}
+			source.Img, _, err = image.Decode(file)
+			return source.Img, err
+		}
+	}
+	return nil, nil
+}
+
+func (ah *ahashHandler) CheckFile(source *FingerprintSource, board string) (*gcsql.FileBan, error) {
+	img, err := ah.getImage(source)
+	if err != nil {
+		return nil, err
+	}
+	ba, err := imagehash.Ahash(img, ah.hashLength)
+	if err != nil {
+		return nil, err
+	}
+	const query = `SELECT id,board_id,staff_id,staff_note,issued_at,checksum,fingerprinter,ban_ip,ban_message
+	FROM DBPREFIXfile_ban WHERE fingerprinter = 'ahash' AND checksum = ? LIMIT 1`
+
+	var fileBan gcsql.FileBan
+	err = gcsql.QueryRowSQL(query, []any{fmt.Sprintf("%x", ba)}, []any{
+		&fileBan.ID, &fileBan.BoardID, &fileBan.StaffID, &fileBan.StaffNote,
+		&fileBan.IssuedAt, &fileBan.Checksum, &fileBan.Fingerprinter,
+		&fileBan.BanIP, &fileBan.BanIPMessage,
+	})
+	return &fileBan, err
+}
+
+func (ah *ahashHandler) IsCompatible(upload *gcsql.Upload) bool {
+	switch strings.ToLower(path.Ext(upload.OriginalFilename)) {
+	case ".jpg":
+		fallthrough
+	case ".jpeg":
+		fallthrough
+	case ".png":
+		fallthrough
+	case ".gif":
+		fallthrough
+	case ".tif":
+		fallthrough
+	case ".tiff":
+		fallthrough
+	case ".bmp":
+		fallthrough
+	case ".webp":
+		return true
+	case ".mp4":
+		fallthrough
+	case ".webm":
+		return ah.hashVideoThumb
+	}
+	return false
+}
+
+func (ahashHandler) Close() error {
+	return nil
+}

pkg/posting/uploads/fingerprinting/fingerprinting.go

@@ -0,0 +1,50 @@
+package fingerprinting
+
+import (
+	"fmt"
+	"image"
+	"net/http"
+
+	"github.com/gochan-org/gochan/pkg/gcsql"
+)
+
+const (
+	defaultHashLength = 16
+)
+
+var (
+	fingerprinterHandlers map[string]UploadFingerprinter
+)
+
+type UploadFingerprinter interface {
+	// Init initializes the fingerprinter, and accepts options assumed to be in gochan.json
+	Init(options map[string]any) error
+	// IsCompatible returns true if the fingerprinter is able to handle the incoming upload
+	IsCompatible(upload *gcsql.Upload) bool
+	// CheckFile scans the incoming file and scans it against files in DBPREFIXfile_ban
+	// with the fingerprinter's id. It returns true if the file is banned
+	CheckFile(source *FingerprintSource, board string) (*gcsql.FileBan, error)
+	// Close closes the fingerprinter. This may or may not be necessary
+	Close() error
+}
+
+type FingerprintSource struct {
+	FilePath string
+	Img      image.Image
+	Request  *http.Request
+}
+
+// RegisterFingerprinter registers the given id. It must be in the configuration,
+// in the FingerprinterOptions map
+func RegisterFingerprinter(id string, handler UploadFingerprinter) error {
+	_, ok := fingerprinterHandlers[id]
+	if ok {
+		return fmt.Errorf("a fingerprinter has already been registered to the ID %q", id)
+	}
+	fingerprinterHandlers[id] = handler
+	return nil
+}
+
+func init() {
+	RegisterFingerprinter("ahash", &ahashHandler{})
+}

pkg/posting/uploads/processimage.go

@@ -85,7 +85,7 @@ func processImage(upload *gcsql.Upload, post *gcsql.Post, board string, filePath
 			return err
 		}
 		if err = syscall.Symlink(path.Join(documentRoot, "spoiler.png"), thumbPath); err != nil {
-			errEv.Err(err).
+			errEv.Err(err).Caller().
 				Str("thumbPath", thumbPath).
 				Msg("Error creating symbolic link to thumbnail path")
 			return err