gsh-digital-services/seaweedFS
5
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

s3tables: optimize regex usage and improve version token uniqueness

Browse Source 
- Pre-compile regex patterns as package-level variables to avoid re-compilation overhead on every call
- Add a random component to version token generation to reduce collision probability under high concurrency
This commit is contained in:
Chris Lu
2026-01-28 11:44:56 -08:00
parent 1c0d37e15a
commit a3af5eb77a
1 changed files with 11 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
weed/s3api/s3tables/utils.go
View File

@@ -2,6 +2,7 @@ package s3tables
import ( import (
"fmt" "fmt"
"math/rand"
"net/url" "net/url"
"path" "path"
"regexp" "regexp"
@@ -9,13 +10,18 @@ import (
"time" "time"
) )
var (
bucketARNPattern = regexp.MustCompile(`^arn:aws:s3tables:[^:]*:[^:]*:bucket/([a-z0-9_-]+)$`)
tableARNPattern = regexp.MustCompile(`^arn:aws:s3tables:[^:]*:[^:]*:bucket/([a-z0-9_-]+)/table/([^/]+)/([a-z0-9_]+)$`)
bucketNamePattern = regexp.MustCompile(`^[a-z0-9_-]+$`)
)
// ARN parsing functions // ARN parsing functions
// parseBucketNameFromARN extracts bucket name from table bucket ARN // parseBucketNameFromARN extracts bucket name from table bucket ARN
// ARN format: arn:aws:s3tables:{region}:{account}:bucket/{bucket-name} // ARN format: arn:aws:s3tables:{region}:{account}:bucket/{bucket-name}
func parseBucketNameFromARN(arn string) (string, error) { func parseBucketNameFromARN(arn string) (string, error) {
pattern := regexp.MustCompile(`^arn:aws:s3tables:[^:]*:[^:]*:bucket/([a-z0-9_-]+)$`) matches := bucketARNPattern.FindStringSubmatch(arn)
matches := pattern.FindStringSubmatch(arn)
if len(matches) != 2 { if len(matches) != 2 {
return "", fmt.Errorf("invalid bucket ARN: %s", arn) return "", fmt.Errorf("invalid bucket ARN: %s", arn)
} }
@@ -26,8 +32,7 @@ func parseBucketNameFromARN(arn string) (string, error) {
// ARN format: arn:aws:s3tables:{region}:{account}:bucket/{bucket-name}/table/{namespace}/{table-name} // ARN format: arn:aws:s3tables:{region}:{account}:bucket/{bucket-name}/table/{namespace}/{table-name}
func parseTableFromARN(arn string) (bucketName, namespace, tableName string, err error) { func parseTableFromARN(arn string) (bucketName, namespace, tableName string, err error) {
// Updated regex to align with namespace validation (single-segment) // Updated regex to align with namespace validation (single-segment)
pattern := regexp.MustCompile(`^arn:aws:s3tables:[^:]*:[^:]*:bucket/([a-z0-9_-]+)/table/([^/]+)/([a-z0-9_]+)$`) matches := tableARNPattern.FindStringSubmatch(arn)
matches := pattern.FindStringSubmatch(arn)
if len(matches) != 4 { if len(matches) != 4 {
return "", "", "", fmt.Errorf("invalid table ARN: %s", arn) return "", "", "", fmt.Errorf("invalid table ARN: %s", arn)
} }
@@ -92,13 +97,12 @@ type tableMetadataInternal struct {
// isValidBucketName validates bucket name characters // isValidBucketName validates bucket name characters
// Bucket names must contain only lowercase letters, numbers, hyphens, and underscores // Bucket names must contain only lowercase letters, numbers, hyphens, and underscores
func isValidBucketName(name string) bool { func isValidBucketName(name string) bool {
pattern := regexp.MustCompile(`^[a-z0-9_-]+$`) return bucketNamePattern.MatchString(name)
return pattern.MatchString(name)
} }
// generateVersionToken generates a unique version token // generateVersionToken generates a unique version token
func generateVersionToken() string { func generateVersionToken() string {
return fmt.Sprintf("%d", time.Now().UnixNano()) return fmt.Sprintf("%d_%d", time.Now().UnixNano(), rand.Int63())
} }
// splitPath splits a path into directory and name components using stdlib // splitPath splits a path into directory and name components using stdlib