文章编辑器问题

soul-api/cmd/migrate-base64-images/main.go

@@ -0,0 +1,146 @@
+// migrate-base64-images 将 chapters 表中 content 内的 base64 图片提取为文件并替换为 URL
+// 用法：cd soul-api && go run ./cmd/migrate-base64-images [--dry-run]
+// 测试环境：APP_ENV=development 时加载 .env.development，请先在测试库验证
+package main
+
+import (
+	"encoding/base64"
+	"flag"
+	"fmt"
+	"log"
+	"math/rand"
+	"os"
+	"path/filepath"
+	"regexp"
+	"strings"
+	"time"
+
+	"soul-api/internal/config"
+	"soul-api/internal/database"
+	"soul-api/internal/model"
+)
+
+// data:image/png;base64,iVBORw0KG... 或 data:image/jpeg;base64,/9j/4AAQ...
+var base64ImgRe = regexp.MustCompile(`(?i)src=["'](data:image/([^;"']+);base64,([A-Za-z0-9+/=]+))["']`)
+
+func main() {
+	dryRun := flag.Bool("dry-run", false, "仅统计和预览，不写入文件与数据库")
+	flag.Parse()
+
+	cfg, err := config.Load()
+	if err != nil {
+		log.Fatal("load config: ", err)
+	}
+	config.SetCurrent(cfg)
+	if err := database.Init(cfg.DBDSN); err != nil {
+		log.Fatal("database: ", err)
+	}
+
+	uploadDir := cfg.UploadDir
+	if uploadDir == "" {
+		uploadDir = "uploads"
+	}
+	bookImagesDir := filepath.Join(uploadDir, "book-images")
+	if !*dryRun {
+		if err := os.MkdirAll(bookImagesDir, 0755); err != nil {
+			log.Fatal("mkdir book-images: ", err)
+		}
+	}
+
+	db := database.DB()
+	var chapters []model.Chapter
+	if err := db.Select("id", "mid", "section_title", "content").Where("content LIKE ?", "%data:image%").Find(&chapters).Error; err != nil {
+		log.Fatal("query chapters: ", err)
+	}
+
+	log.Printf("找到 %d 篇含 base64 图片的章节", len(chapters))
+	if len(chapters) == 0 {
+		return
+	}
+
+	rand.Seed(time.Now().UnixNano())
+	const letters = "abcdefghijklmnopqrstuvwxyz0123456789"
+	randomStr := func(n int) string {
+		b := make([]byte, n)
+		for i := range b {
+			b[i] = letters[rand.Intn(len(letters))]
+		}
+		return string(b)
+	}
+
+	mimeToExt := map[string]string{
+		"png":  ".png",
+		"jpeg": ".jpg",
+		"jpg":  ".jpg",
+		"gif":  ".gif",
+		"webp": ".webp",
+	}
+
+	totalReplaced := 0
+	totalFiles := 0
+	for _, ch := range chapters {
+		matches := base64ImgRe.FindAllStringSubmatch(ch.Content, -1)
+		if len(matches) == 0 {
+			continue
+		}
+		newContent := ch.Content
+		for _, m := range matches {
+			fullDataURL := m[1]
+			mime := strings.ToLower(strings.TrimSpace(m[2]))
+			b64 := m[3]
+			ext := mimeToExt[mime]
+			if ext == "" {
+				ext = ".png"
+			}
+
+			decoded, err := base64.StdEncoding.DecodeString(b64)
+			if err != nil {
+				log.Printf("  [%s] base64 解码失败: %v", ch.ID, err)
+				continue
+			}
+
+			name := fmt.Sprintf("%d_%s%s", time.Now().UnixNano(), randomStr(6), ext)
+			dst := filepath.Join(bookImagesDir, name)
+			url := "/uploads/" + filepath.ToSlash(filepath.Join("book-images", name))
+
+			if !*dryRun {
+				if err := os.WriteFile(dst, decoded, 0644); err != nil {
+					log.Printf("  [%s] 写入文件失败 %s: %v", ch.ID, name, err)
+					continue
+				}
+			}
+
+			oldSrc := `src="` + fullDataURL + `"`
+			newSrc := `src="` + url + `"`
+			if strings.Contains(newContent, oldSrc) {
+				newContent = strings.Replace(newContent, oldSrc, newSrc, 1)
+			} else {
+				oldSrc2 := `src='` + fullDataURL + `'`
+				newSrc2 := `src="` + url + `"`
+				newContent = strings.Replace(newContent, oldSrc2, newSrc2, 1)
+			}
+			totalFiles++
+			log.Printf("  [%s] %s -> %s (%d bytes)", ch.ID, mime, name, len(decoded))
+		}
+
+		if newContent != ch.Content {
+			totalReplaced++
+			oldLen := len(ch.Content)
+			newLen := len(newContent)
+			if !*dryRun {
+				if err := db.Model(&model.Chapter{}).Where("id = ?", ch.ID).Update("content", newContent).Error; err != nil {
+					log.Printf("  [%s] 更新数据库失败: %v", ch.ID, err)
+					continue
+				}
+			}
+			log.Printf("  [%s] 已更新，content 长度 %d -> %d (减少 %d)", ch.ID, oldLen, newLen, oldLen-newLen)
+		}
+	}
+
+	if *dryRun {
+		log.Printf("[dry-run] 将处理 %d 篇章节，共 %d 张 base64 图片", totalReplaced, totalFiles)
+		log.Printf("[dry-run] 去掉 --dry-run 后执行以实际写入")
+	} else {
+		log.Printf("完成：更新 %d 篇章节，提取 %d 张图片到 uploads/book-images/", totalReplaced, totalFiles)
+	}
+}