feat(agent): implement resolveMediaRefs with streaming base64 and filetype detection

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-06-12 18:08:54 +00:00 · 2026-03-03 14:01:52 +08:00
parent 559cef3d5b
commit 6fd65825e7
2 changed files with 261 additions and 0 deletions
@@ -0,0 +1,121 @@
+// PicoClaw - Ultra-lightweight personal AI agent
+// Inspired by and based on nanobot: https://github.com/HKUDS/nanobot
+// License: MIT
+//
+// Copyright (c) 2026 PicoClaw contributors
+
+package agent
+
+import (
+	"bytes"
+	"encoding/base64"
+	"io"
+	"os"
+	"strings"
+
+	"github.com/h2non/filetype"
+	"github.com/sipeed/picoclaw/pkg/logger"
+	"github.com/sipeed/picoclaw/pkg/media"
+	"github.com/sipeed/picoclaw/pkg/providers"
+)
+
+// resolveMediaRefs replaces media:// refs in message Media fields with base64 data URLs.
+// Uses streaming base64 encoding (file handle → encoder → buffer) to avoid holding
+// both raw bytes and encoded string in memory simultaneously.
+// Returns a new slice; original messages are not mutated.
+func resolveMediaRefs(messages []providers.Message, store media.MediaStore, maxSize int) []providers.Message {
+	if store == nil {
+		return messages
+	}
+
+	result := make([]providers.Message, len(messages))
+	copy(result, messages)
+
+	for i, m := range result {
+		if len(m.Media) == 0 {
+			continue
+		}
+
+		resolved := make([]string, 0, len(m.Media))
+		for _, ref := range m.Media {
+			if !strings.HasPrefix(ref, "media://") {
+				resolved = append(resolved, ref)
+				continue
+			}
+
+			localPath, meta, err := store.ResolveWithMeta(ref)
+			if err != nil {
+				logger.WarnCF("agent", "Failed to resolve media ref", map[string]any{
+					"ref":   ref,
+					"error": err.Error(),
+				})
+				continue
+			}
+
+			info, err := os.Stat(localPath)
+			if err != nil {
+				logger.WarnCF("agent", "Failed to stat media file", map[string]any{
+					"path":  localPath,
+					"error": err.Error(),
+				})
+				continue
+			}
+			if info.Size() > int64(maxSize) {
+				logger.WarnCF("agent", "Media file too large, skipping", map[string]any{
+					"path":     localPath,
+					"size":     info.Size(),
+					"max_size": maxSize,
+				})
+				continue
+			}
+
+			// Determine MIME type: prefer metadata, fallback to magic-bytes detection
+			mime := meta.ContentType
+			if mime == "" {
+				kind, err := filetype.MatchFile(localPath)
+				if err != nil || kind == filetype.Unknown {
+					logger.WarnCF("agent", "Unknown media type, skipping", map[string]any{
+						"path": localPath,
+					})
+					continue
+				}
+				mime = kind.MIME.Value
+			}
+
+			// Streaming base64: open file → base64 encoder → buffer
+			// Peak memory: ~1.33x file size (buffer only, no raw bytes copy)
+			f, err := os.Open(localPath)
+			if err != nil {
+				logger.WarnCF("agent", "Failed to open media file", map[string]any{
+					"path":  localPath,
+					"error": err.Error(),
+				})
+				continue
+			}
+
+			prefix := "data:" + mime + ";base64,"
+			encodedLen := base64.StdEncoding.EncodedLen(int(info.Size()))
+			var buf bytes.Buffer
+			buf.Grow(len(prefix) + encodedLen)
+			buf.WriteString(prefix)
+
+			encoder := base64.NewEncoder(base64.StdEncoding, &buf)
+			if _, err := io.Copy(encoder, f); err != nil {
+				f.Close()
+				logger.WarnCF("agent", "Failed to encode media file", map[string]any{
+					"path":  localPath,
+					"error": err.Error(),
+				})
+				continue
+			}
+			encoder.Close()
+			f.Close()
+
+			resolved = append(resolved, buf.String())
+		}
+
+		result[i].Media = resolved
+	}
+
+	return result
+}