validate tags on copy object and add regex for validating tags

4 files changed, 60 insertions, 26 deletions

diff --git a/weed/s3api/s3api_object_copy_handlers.go b/weed/s3api/s3api_object_copy_handlers.go
index 9157748f6..8e62deb94 100644
--- a/weed/s3api/s3api_object_copy_handlers.go
+++ b/weed/s3api/s3api_object_copy_handlers.go
@@ -45,7 +45,11 @@ func (s3a *S3ApiServer) CopyObjectHandler(w http.ResponseWriter, r *http.Request
 			s3err.WriteErrorResponse(w, r, s3err.ErrInvalidCopySource)
 			return
 		}
-		entry.Extended = processMetadataBytes(r.Header, entry.Extended, replaceMeta, replaceTagging)
+		entry.Extended, err = processMetadataBytes(r.Header, entry.Extended, replaceMeta, replaceTagging)
+		if err != nil {
+			glog.Errorf("CopyObjectHandler ValidateTags error %s: %v", r.URL, err)
+			s3err.WriteErrorResponse(w, r, s3err.ErrInvalidTag)
+		}
 		err = s3a.touch(dir, name, entry)
 		if err != nil {
 			s3err.WriteErrorResponse(w, r, s3err.ErrInvalidCopySource)
@@ -252,7 +256,7 @@ func processMetadata(reqHeader, existing http.Header, replaceMeta, replaceTaggin
 	return
 }
 
-func processMetadataBytes(reqHeader http.Header, existing map[string][]byte, replaceMeta, replaceTagging bool) (metadata map[string][]byte) {
+func processMetadataBytes(reqHeader http.Header, existing map[string][]byte, replaceMeta, replaceTagging bool) (metadata map[string][]byte, err error) {
 	metadata = make(map[string][]byte)
 
 	if sc := existing[s3_constants.AmzStorageClass]; len(sc) > 0 {
@@ -277,16 +281,18 @@ func processMetadataBytes(reqHeader http.Header, existing map[string][]byte, rep
 			}
 		}
 	}
-
 	if replaceTagging {
 		if tags := reqHeader.Get(s3_constants.AmzObjectTagging); tags != "" {
-			for _, v := range strings.Split(tags, "&") {
-				tag := strings.Split(v, "=")
-				if len(tag) == 2 {
-					metadata[s3_constants.AmzObjectTagging+"-"+tag[0]] = []byte(tag[1])
-				} else if len(tag) == 1 {
-					metadata[s3_constants.AmzObjectTagging+"-"+tag[0]] = nil
-				}
+			parsedTags, err := parseTagsHeader(tags)
+			if err != nil {
+				return nil, err
+			}
+			err = validateTags(parsedTags)
+			if err != nil {
+				return nil, err
+			}
+			for k, v := range parsedTags {
+				metadata[s3_constants.AmzObjectTagging+"-"+k] = []byte(v)
 			}
 		}
 	} else {
diff --git a/weed/s3api/s3api_object_copy_handlers_test.go b/weed/s3api/s3api_object_copy_handlers_test.go
index 610b29a6b..3cb2b0562 100644
--- a/weed/s3api/s3api_object_copy_handlers_test.go
+++ b/weed/s3api/s3api_object_copy_handlers_test.go
@@ -367,7 +367,7 @@ func TestProcessMetadataBytes(t *testing.T) {
 		reqHeader := transferHToHeader(tc.request)
 		existing := transferHToBytesArr(tc.existing)
 		replaceMeta, replaceTagging := replaceDirective(reqHeader)
-		extends := processMetadataBytes(reqHeader, existing, replaceMeta, replaceTagging)
+		extends, _ := processMetadataBytes(reqHeader, existing, replaceMeta, replaceTagging)
 
 		result := transferBytesArrToH(extends)
 		fmtTagging(result, tc.want)
diff --git a/weed/s3api/s3api_object_tagging_handlers.go b/weed/s3api/s3api_object_tagging_handlers.go
index 9fde0309c..35da1cc20 100644
--- a/weed/s3api/s3api_object_tagging_handlers.go
+++ b/weed/s3api/s3api_object_tagging_handlers.go
@@ -62,22 +62,10 @@ func (s3a *S3ApiServer) PutObjectTaggingHandler(w http.ResponseWriter, r *http.R
 		return
 	}
 	tags := tagging.ToTags()
-	if len(tags) > 10 {
-		glog.Errorf("PutObjectTaggingHandler tags %s: %d tags more than 10", r.URL, len(tags))
+	err = validateTags(tags)
+	if err != nil {
+		glog.Errorf("PutObjectTaggingHandler ValidateTags error %s: %v", r.URL, err)
 		s3err.WriteErrorResponse(w, r, s3err.ErrInvalidTag)
-		return
-	}
-	for k, v := range tags {
-		if len(k) > 128 {
-			glog.Errorf("PutObjectTaggingHandler tags %s: tag key %s longer than 128", r.URL, k)
-			s3err.WriteErrorResponse(w, r, s3err.ErrInvalidTag)
-			return
-		}
-		if len(v) > 256 {
-			glog.Errorf("PutObjectTaggingHandler tags %s: tag value %s longer than 256", r.URL, v)
-			s3err.WriteErrorResponse(w, r, s3err.ErrInvalidTag)
-			return
-		}
 	}
 
 	if err = s3a.setTags(dir, name, tagging.ToTags()); err != nil {
diff --git a/weed/s3api/tags.go b/weed/s3api/tags.go
index 979e5a80c..06c9e4b80 100644
--- a/weed/s3api/tags.go
+++ b/weed/s3api/tags.go
@@ -2,6 +2,9 @@ package s3api
 
 import (
 	"encoding/xml"
+	"fmt"
+	"regexp"
+	"strings"
 )
 
 type Tag struct {
@@ -37,3 +40,40 @@ func FromTags(tags map[string]string) (t *Tagging) {
 	}
 	return
 }
+
+func parseTagsHeader(tags string) (map[string]string, error) {
+	var parsedTags map[string]string
+	for _, v := range strings.Split(tags, "&") {
+		tag := strings.Split(v, "=")
+		if len(tag) == 2 {
+			parsedTags[tag[0]] = tag[1]
+		} else if len(tag) == 1 {
+			parsedTags[tag[0]] = ""
+		}
+	}
+	return nil, nil
+}
+
+func validateTags(tags map[string]string) error {
+	if len(tags) > 10 {
+		return fmt.Errorf("validate tags: %d tags more than 10", len(tags))
+	}
+	for k, v := range tags {
+		if len(k) > 128 {
+			return fmt.Errorf("validate tags: tag key %s longer than 128", k)
+		}
+		validateKey, err := regexp.MatchString(`^([\p{L}\p{Z}\p{N}_.:/=+\-@]*)$`, k)
+		if !validateKey && err != nil {
+			return fmt.Errorf("validate tags key %s error %w ", k, err)
+		}
+		if len(v) > 256 {
+			return fmt.Errorf("validate tags: tag value %s longer than 256", v)
+		}
+		validateValue, err := regexp.MatchString(`^([\p{L}\p{Z}\p{N}_.:/=+\-@]*)$`, v)
+		if !validateValue && err != nil {
+			return fmt.Errorf("validate tags value %s error %w ", v, err)
+		}
+	}
+
+	return nil
+}