feat: sync-on-change optimization & update config to GLM-4.7-FlashX

scripts/sync_hf.py

@@ -40,44 +40,10 @@ class OpenClawHFSync:
         
         self.enabled = True
         self.api = HfApi(token=HF_TOKEN)
+        self.checksums = {}
         self._ensure_repo()
 
-    def _upload_folder(self, local_dir, remote_path, msg, ignore=None):
-        if not local_dir.exists(): return
-        patterns = ignore or EXCLUDE_PATTERNS
-        
-        # Debug: list files being considered for credentials to ensure they are found
-        if "credentials" in str(local_dir):
-            try:
-                files = list(local_dir.glob("**/*"))
-                print(f"[SYNC] Scanning credentials dir: {local_dir}, found {len(files)} items")
-                for f in files:
-                    if f.is_file():
-                        print(f"[SYNC]   - {f.relative_to(local_dir)}")
-            except Exception:
-                pass
 
-        try:
-            self.api.upload_folder(
-                folder_path=str(local_dir),
-                path_in_repo=remote_path,
-                repo_id=HF_REPO_ID,
-                repo_type="dataset",
-                token=HF_TOKEN,
-                commit_message=msg,
-                commit_description=f"Automated sync at {datetime.now().isoformat()}",
-                ignore_patterns=patterns,
-            )
-        except Exception as e:
-            print(f"[SYNC] Upload folder error ({local_dir}): {e}")
-        if not HF_TOKEN:
-            print("[SYNC] WARNING: HF_TOKEN not set in environment. Persistence disabled.")
-            self.enabled = False
-            return
-        
-        self.enabled = True
-        self.api = HfApi(token=HF_TOKEN)
-        self._ensure_repo()
 
     def _ensure_repo(self):
         try:
@@ -112,6 +78,13 @@ class OpenClawHFSync:
 
         print("[SYNC] Load complete.")
         self._debug_list_files()
+        self._init_checksums()
+
+    def _init_checksums(self):
+        for name in ["credentials", "workspace", "agents"]:
+            p = OPENCLAW_HOME / name
+            if p.exists():
+                self.checksums[str(p)] = self._get_dir_checksum(p)
 
     def _patch_config(self):
         """Ensure critical config settings are present and valid after restore."""
@@ -151,22 +124,30 @@ class OpenClawHFSync:
                         print("[SYNC] Removed '/dev/null' from plugins.locations")
                         modified = True
             
-            # 2. Enforce Model to GLM-4-Plus (User's desired state from context)
-            # Check if we need to add the provider as well
+            # 2. Enforce Model to GLM-4-Flash (User's desired state from context)
             if 'agents' not in data: data['agents'] = {}
             if 'defaults' not in data['agents']: data['agents']['defaults'] = {}
             if 'model' not in data['agents']['defaults']: data['agents']['defaults']['model'] = {}
             
-            current_model = data['agents']['defaults']['model'].get('primary')
-            if current_model != 'zhipu/glm-4-plus':
-                print(f"[SYNC] Updating model.primary: {current_model} -> zhipu/glm-4-plus")
-                data['agents']['defaults']['model']['primary'] = 'zhipu/glm-4-plus'
-                modified = True
-
-            # Ensure Zhipu provider exists if missing (basic check)
-            # This is complex to patch fully, but we ensure 'zhipu' key exists in providers if needed
-            # For now, relying on User to have set it up or defaults. 
+            # Update Provider Zhipu
+            if 'models' not in data: data['models'] = {}
+            if 'providers' not in data['models']: data['models']['providers'] = {}
+            
+            # Force Zhipu config
+            data['models']['providers']['zhipu'] = {
+                "baseUrl": "https://api.z.ai/api/paas/v4", 
+                "apiKey": "${ZHIPU_API_KEY}",
+                "api": "openai-chat", 
+                "models": [
+                    { "id": "GLM-4.7-FlashX", "name": "GLM-4.7 FlashX" }
+                ]
+            }
             
+            # Set default model
+            print(f"[SYNC] Updating model.primary -> zhipu/GLM-4.7-FlashX")
+            data['agents']['defaults']['model']['primary'] = 'zhipu/GLM-4.7-FlashX'
+            modified = True
+
             if modified:
                 with open(config_path, 'w') as f:
                     json.dump(data, f, indent=2)
@@ -177,6 +158,26 @@ class OpenClawHFSync:
         except Exception as e:
             print(f"[SYNC] Failed to patch config: {e}")
 
+    def _get_dir_checksum(self, directory, patterns=None):
+        """Calculate a simple checksum of a directory based on mtime and size of files."""
+        if not directory.exists(): return None
+        import hashlib
+        hash_md5 = hashlib.md5()
+        # TODO: Implement pattern filtering for checksum if needed, 
+        # but for now simple walk is enough as upload_folder handles ignores.
+        # We just want to avoid calling upload_folder if nothing changed.
+        for root, dirs, files in os.walk(directory):
+            for name in sorted(files):
+                if name.endswith(".pyc") or name == "models.json": continue # meaningful ignores
+                fp = Path(root) / name
+                try:
+                    stats = fp.stat()
+                    # Hash path, mtime, size
+                    s = f"{fp.relative_to(directory)}:{stats.st_mtime}:{stats.st_size}".encode('utf-8')
+                    hash_md5.update(s)
+                except: pass
+        return hash_md5.hexdigest()
+
     def _download_config(self):
         config_local = OPENCLAW_HOME / "openclaw.json"
         try:
@@ -331,7 +332,17 @@ class OpenClawHFSync:
             print(f"[SYNC] Upload file error ({local_path}): {e}")
 
     def _upload_folder(self, local_dir, remote_path, msg, ignore=None):
-        if not local_dir.exists(): return
+        # Checksum check
+        current_sum = self._get_dir_checksum(local_dir)
+        last_sum = self.checksums.get(str(local_dir))
+        
+        if current_sum == last_sum:
+            # print(f"[SYNC] Skipping {local_dir.name} (no changes)")
+            return
+            
+        print(f"[SYNC] Changes detected in {local_dir.name}, syncing...")
+        self.checksums[str(local_dir)] = current_sum
+        
         patterns = ignore or EXCLUDE_PATTERNS
         
         # Debug: list files being considered for credentials to ensure they are found