305 lines
9.7 KiB
Go
305 lines
9.7 KiB
Go
// 生成结果（data URL / 远程地址）的持久化与远程下载。
|
||
//
|
||
// 远程下载经过出站地址校验（validateRemoteResourceURL），防止 SSRF。
|
||
|
||
package app
|
||
|
||
import (
|
||
"bytes"
|
||
"context"
|
||
"encoding/base64"
|
||
"encoding/json"
|
||
"errors"
|
||
"fmt"
|
||
"image"
|
||
"io"
|
||
"mime"
|
||
"net/http"
|
||
"net/url"
|
||
"path"
|
||
"path/filepath"
|
||
"strings"
|
||
"time"
|
||
)
|
||
|
||
func localObjectKey(userID string, kind string, fileName string, mimeType string, now time.Time) string {
|
||
ext := resourceFileExtension(fileName, mimeType, kind)
|
||
return path.Join("users", safeObjectSegment(userID), kind, now.Format("2006/01/02"), newID()+ext)
|
||
}
|
||
|
||
func (s *Service) persistGeneratedMediaResult(userID string, result map[string]interface{}) (map[string]interface{}, error) {
|
||
return s.persistGeneratedMediaResultMode(userID, result, false, true)
|
||
}
|
||
|
||
func (s *Service) persistLegacyGeneratedMediaResult(userID string, result map[string]interface{}) (map[string]interface{}, error) {
|
||
return s.persistGeneratedMediaResultMode(userID, result, true, false)
|
||
}
|
||
|
||
func (s *Service) persistGeneratedMediaResultMode(userID string, result map[string]interface{}, skipInvalidDataURL bool, enforceQuota bool) (map[string]interface{}, error) {
|
||
if result == nil {
|
||
return map[string]interface{}{}, nil
|
||
}
|
||
encoded, err := json.Marshal(result)
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
var normalized map[string]interface{}
|
||
if err := json.Unmarshal(encoded, &normalized); err != nil {
|
||
return nil, err
|
||
}
|
||
value, err := s.persistGeneratedMediaValueMode(userID, normalized, skipInvalidDataURL, enforceQuota)
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
return value.(map[string]interface{}), nil
|
||
}
|
||
|
||
func (s *Service) persistGeneratedMediaValue(userID string, value interface{}) (interface{}, error) {
|
||
return s.persistGeneratedMediaValueMode(userID, value, false, true)
|
||
}
|
||
|
||
func (s *Service) persistGeneratedMediaValueMode(userID string, value interface{}, skipInvalidDataURL bool, enforceQuota bool) (interface{}, error) {
|
||
switch item := value.(type) {
|
||
case []interface{}:
|
||
for index, child := range item {
|
||
stored, err := s.persistGeneratedMediaValueMode(userID, child, skipInvalidDataURL, enforceQuota)
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
item[index] = stored
|
||
}
|
||
return item, nil
|
||
case map[string]interface{}:
|
||
if raw := inlineMediaValue(item); raw != "" {
|
||
mimeType, data, err := s.decodeDataURL(raw)
|
||
if err != nil && !skipInvalidDataURL {
|
||
return nil, err
|
||
}
|
||
if err == nil {
|
||
kind := normalizeResourceKind("", mimeType)
|
||
width, height := intValue(item["width"]), intValue(item["height"])
|
||
if kind == "image" && (width <= 0 || height <= 0) {
|
||
width, height = imageDimensions(data)
|
||
}
|
||
quotaDay := ""
|
||
if enforceQuota {
|
||
quotaDay, err = s.reserveGeneratedResourceQuota(userID, int64(len(data)))
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
}
|
||
resource, _, err := s.storeResource(userID, kind, "generated."+extensionFromMimeType(mimeType), mimeType, int64(len(data)), width, height, int64(intValue(item["durationMs"])), bytes.NewReader(data), nil, false)
|
||
if err != nil {
|
||
if enforceQuota {
|
||
s.releaseUserUploadQuota(userID, quotaDay, int64(len(data)))
|
||
}
|
||
return nil, fmt.Errorf("生成内容写入资源存储失败：%w", err)
|
||
}
|
||
if enforceQuota {
|
||
s.commitUserUploadQuota(userID, int64(len(data)))
|
||
}
|
||
resourceURL := resourceFileURL(resource.ID)
|
||
for _, key := range []string{"dataUrl", "content", "url", "coverUrl"} {
|
||
if text, ok := item[key].(string); ok && (text == raw || strings.HasPrefix(text, "blob:")) {
|
||
item[key] = resourceURL
|
||
}
|
||
}
|
||
if _, ok := item["dataUrl"]; ok {
|
||
item["dataUrl"] = resourceURL
|
||
}
|
||
item["url"] = resourceURL
|
||
item["storageKey"] = "resource:" + resource.ID
|
||
item["resourceId"] = resource.ID
|
||
item["bytes"] = resource.Size
|
||
item["mimeType"] = resource.MimeType
|
||
item["width"] = resource.Width
|
||
item["height"] = resource.Height
|
||
}
|
||
}
|
||
for key, child := range item {
|
||
stored, err := s.persistGeneratedMediaValueMode(userID, child, skipInvalidDataURL, enforceQuota)
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
item[key] = stored
|
||
}
|
||
return item, nil
|
||
default:
|
||
return value, nil
|
||
}
|
||
}
|
||
|
||
func inlineMediaValue(item map[string]interface{}) string {
|
||
for _, key := range []string{"dataUrl", "content", "url", "coverUrl"} {
|
||
if text, ok := item[key].(string); ok && (strings.HasPrefix(text, "data:image/") || strings.HasPrefix(text, "data:video/") || strings.HasPrefix(text, "data:audio/")) {
|
||
return text
|
||
}
|
||
}
|
||
return ""
|
||
}
|
||
|
||
func (s *Service) decodeDataURL(value string) (string, []byte, error) {
|
||
header, encoded, ok := strings.Cut(value, ",")
|
||
if !ok || !strings.HasPrefix(header, "data:") || !strings.HasSuffix(strings.ToLower(header), ";base64") {
|
||
return "", nil, fmt.Errorf("%w：格式无效", errInvalidGeneratedDataURL)
|
||
}
|
||
mimeType := strings.TrimSuffix(strings.TrimPrefix(header, "data:"), ";base64")
|
||
data, err := base64.StdEncoding.DecodeString(encoded)
|
||
if err != nil {
|
||
return "", nil, fmt.Errorf("%w：base64 解码失败：%v", errInvalidGeneratedDataURL, err)
|
||
}
|
||
policy, err := s.RuntimePolicy()
|
||
if err != nil {
|
||
return "", nil, err
|
||
}
|
||
if int64(len(data)) > megabytes(policy.Resource.GeneratedFileMB) {
|
||
return "", nil, fmt.Errorf("单个生成资源超过 %dMB", policy.Resource.GeneratedFileMB)
|
||
}
|
||
return mimeType, data, nil
|
||
}
|
||
|
||
func intValue(value interface{}) int {
|
||
switch number := value.(type) {
|
||
case float64:
|
||
return int(number)
|
||
case int:
|
||
return number
|
||
case int64:
|
||
return int(number)
|
||
default:
|
||
return 0
|
||
}
|
||
}
|
||
|
||
type remoteResourcePayload struct {
|
||
url string
|
||
endpoint string
|
||
fileName string
|
||
mimeType string
|
||
data []byte
|
||
}
|
||
|
||
func downloadRemoteResource(rawURL string, maxBytes int64) (remoteResourcePayload, error) {
|
||
parsed, err := validateRemoteResourceURL(rawURL)
|
||
if err != nil {
|
||
return remoteResourcePayload{}, err
|
||
}
|
||
client := OutboundHTTPClient(3 * time.Minute)
|
||
req, err := http.NewRequest(http.MethodGet, parsed.String(), nil)
|
||
if err != nil {
|
||
return remoteResourcePayload{}, err
|
||
}
|
||
resp, err := client.Do(req)
|
||
if err != nil {
|
||
// 把下载层网络错误统一映射为可公开的 502/504 文案，避免在恢复接口上被 failService
|
||
// 折叠成通用的“系统处理失败”。同时保留真实原因在 Cause 里供诊断日志追踪。
|
||
if errors.Is(err, context.DeadlineExceeded) {
|
||
return remoteResourcePayload{}, WrapAppError(http.StatusGatewayTimeout, "远程资源下载超时，请检查地址是否可访问或稍后重试", err)
|
||
}
|
||
return remoteResourcePayload{}, WrapAppError(http.StatusBadGateway, "远程资源下载失败，请检查地址是否可访问", err)
|
||
}
|
||
defer resp.Body.Close()
|
||
if resp.StatusCode < 200 || resp.StatusCode >= 300 {
|
||
return remoteResourcePayload{}, WrapAppError(http.StatusBadGateway, fmt.Sprintf("远程资源下载失败：%s", resp.Status), nil)
|
||
}
|
||
data, err := io.ReadAll(io.LimitReader(resp.Body, maxBytes))
|
||
if err != nil {
|
||
if errors.Is(err, context.DeadlineExceeded) {
|
||
return remoteResourcePayload{}, WrapAppError(http.StatusGatewayTimeout, "远程资源下载超时，请检查地址是否可访问或稍后重试", err)
|
||
}
|
||
return remoteResourcePayload{}, WrapAppError(http.StatusBadGateway, "远程资源读取失败，请检查地址是否可访问", err)
|
||
}
|
||
if int64(len(data)) >= maxBytes {
|
||
return remoteResourcePayload{}, BadAuthRequest(fmt.Sprintf("远程资源必须小于 %s", formatStorageLimit(maxBytes)))
|
||
}
|
||
mimeType := strings.TrimSpace(resp.Header.Get("Content-Type"))
|
||
if idx := strings.Index(mimeType, ";"); idx >= 0 {
|
||
mimeType = strings.TrimSpace(mimeType[:idx])
|
||
}
|
||
if mimeType == "" || mimeType == "application/octet-stream" {
|
||
mimeType = http.DetectContentType(data)
|
||
}
|
||
fileName := path.Base(parsed.Path)
|
||
if fileName == "" || fileName == "." || !strings.Contains(fileName, ".") {
|
||
fileName = "resource." + extensionFromMimeType(mimeType)
|
||
}
|
||
return remoteResourcePayload{url: parsed.String(), endpoint: parsed.Host, fileName: fileName, mimeType: mimeType, data: data}, nil
|
||
}
|
||
|
||
func openRemoteResource(rawURL string) (io.ReadCloser, error) {
|
||
parsed, err := validateRemoteResourceURL(rawURL)
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
client := OutboundHTTPClient(90 * time.Second)
|
||
resp, err := client.Get(parsed.String())
|
||
if err != nil {
|
||
return nil, err
|
||
}
|
||
if resp.StatusCode < 200 || resp.StatusCode >= 300 {
|
||
defer resp.Body.Close()
|
||
return nil, fmt.Errorf("远程资源读取失败：%s", resp.Status)
|
||
}
|
||
return resp.Body, nil
|
||
}
|
||
|
||
func validateRemoteResourceURL(rawURL string) (*url.URL, error) {
|
||
return ValidateOutboundURL(rawURL)
|
||
}
|
||
|
||
func extensionFromMimeType(mimeType string) string {
|
||
if strings.Contains(mimeType, "png") {
|
||
return "png"
|
||
}
|
||
if strings.Contains(mimeType, "jpeg") {
|
||
return "jpg"
|
||
}
|
||
if strings.Contains(mimeType, "webp") {
|
||
return "webp"
|
||
}
|
||
if strings.Contains(mimeType, "gif") {
|
||
return "gif"
|
||
}
|
||
if strings.Contains(mimeType, "mp4") {
|
||
return "mp4"
|
||
}
|
||
if strings.Contains(mimeType, "webm") {
|
||
return "webm"
|
||
}
|
||
if strings.Contains(mimeType, "mpeg") {
|
||
return "mp3"
|
||
}
|
||
if strings.Contains(mimeType, "wav") {
|
||
return "wav"
|
||
}
|
||
return "bin"
|
||
}
|
||
|
||
func resourceFileExtension(fileName string, mimeType string, kind string) string {
|
||
if ext := strings.ToLower(filepath.Ext(strings.TrimSpace(fileName))); ext != "" && ext != "." {
|
||
return ext
|
||
}
|
||
cleanMimeType := strings.TrimSpace(strings.Split(mimeType, ";")[0])
|
||
if extensions, err := mime.ExtensionsByType(cleanMimeType); err == nil && len(extensions) > 0 {
|
||
return strings.ToLower(extensions[0])
|
||
}
|
||
switch kind {
|
||
case "image":
|
||
return ".png"
|
||
case "video":
|
||
return ".mp4"
|
||
case "audio":
|
||
return ".mp3"
|
||
default:
|
||
return ".bin"
|
||
}
|
||
}
|
||
|
||
func imageDimensions(data []byte) (int, int) {
|
||
config, _, err := image.DecodeConfig(bytes.NewReader(data))
|
||
if err != nil {
|
||
return 0, 0
|
||
}
|
||
return config.Width, config.Height
|
||
}
|