fixed # 适配AstrBot

2026-10-29 21:59:58 +08:00 · 2026-10-23 19:12:16 +08:00
parent d0c1741940
commit 685cff9293
49 changed files with 1557 additions and 165 deletions
@@ -0,0 +1,120 @@
+"""固定摘要的外部素材缓存；下载后校验，原子替换，不解压到磁盘。"""
+
+import hashlib
+from importlib import resources
+import json
+import logging
+from pathlib import Path
+import tempfile
+import threading
+import time
+from urllib.parse import urlsplit
+from urllib.request import Request, urlopen
+from zipfile import ZipFile
+
+logger = logging.getLogger(__name__)
+_LOCK = threading.Lock()
+SOURCES = ('Github', 'Gitea', 'farm_service')
+MAX_BYTES = 64 * 1024 * 1024
+
+
+def expected_manifest():
+    return json.loads(resources.files(__package__).joinpath(
+        'data', 'asset-manifest.json',
+    ).read_bytes())
+
+
+def validate_options(source, service_url):
+    if source not in SOURCES:
+        raise ValueError('Unknown resource source')
+    if service_url:
+        checked_url(service_url)
+    if source == 'farm_service' and not service_url:
+        raise ValueError('farm_service address required')
+
+
+def checked_url(url):
+    parsed = urlsplit(url)
+    if (parsed.scheme != 'https' or not parsed.hostname or parsed.username
+            or parsed.password or parsed.query or parsed.fragment):
+        raise ValueError('Resource URL requires public HTTPS without credentials')
+    return url
+
+
+def _copy_remote(url, destination, limit):
+    request = Request(checked_url(url), headers={'User-Agent': 'Farm-assets/1'})
+    started = time.monotonic()
+    with urlopen(request, timeout=15) as response:
+        checked_url(response.geturl())
+        length = response.headers.get('Content-Length')
+        if length is not None and int(length) > limit:
+            raise ValueError('Resource download exceeds limit')
+        total = 0
+        while True:
+            chunk = response.read(min(65536, limit-total+1))
+            if not chunk:
+                break
+            total += len(chunk)
+            if total > limit or time.monotonic()-started > 60:
+                raise ValueError('Resource download limit or timeout')
+            destination.write(chunk)
+
+
+def verify_bundle(path, manifest):
+    path = Path(path)
+    if (not path.is_file() or path.stat().st_size != manifest['size_bytes']
+            or not 0 < manifest['size_bytes'] <= MAX_BYTES):
+        raise ValueError('Resource size mismatch')
+    digest = hashlib.sha256()
+    with path.open('rb') as stream:
+        for chunk in iter(lambda: stream.read(65536), b''):
+            digest.update(chunk)
+    if digest.hexdigest() != manifest['sha256']:
+        raise ValueError('Resource SHA256 mismatch')
+    metadata = json.loads(resources.files(__package__).joinpath(
+        'data', 'metadata.json',
+    ).read_bytes())
+    with ZipFile(path) as archive:
+        entries = archive.infolist()
+        names = [item.filename for item in entries]
+        if (len(names) != len(set(names))
+                or set(names) != set(metadata['images'])
+                or sum(item.file_size for item in entries) > 256*1024*1024
+                or any(item.is_dir() or item.flag_bits & 1
+                       for item in entries)):
+            raise ValueError('Resource contents mismatch')
+    return path
+
+
+def ensure_assets(data_dir, source='Github', service_url=''):
+    """每次启动校验缓存；仅缺失或不匹配时访问所选源。"""
+    validate_options(source, service_url)
+    expected = expected_manifest()
+    target = (Path(data_dir) / 'assets' / expected['sha256'] / 'assets.zip')
+    with _LOCK:
+        try:
+            return verify_bundle(target, expected)
+        except (OSError, ValueError):
+            pass
+        target.parent.mkdir(parents=True, exist_ok=True)
+        # 清单绑定在插件内；远端不能通过更换清单放行不兼容的素材。
+        if source == 'farm_service':
+            url = (service_url.rstrip('/') + '/v1/assets/'
+                   + expected['sha256'] + '/assets.zip')
+        else:
+            url = expected['sources'][source.lower()]
+        temporary = None
+        try:
+            with tempfile.NamedTemporaryFile(
+                dir=target.parent, suffix='.download', delete=False,
+            ) as stream:
+                temporary = Path(stream.name)
+                _copy_remote(url, stream, expected['size_bytes'])
+            verify_bundle(temporary, expected)
+            temporary.replace(target)
+            logger.info('farm assets ready source=%s sha256=%s',
+                        source, expected['sha256'])
+            return target
+        finally:
+            if temporary is not None:
+                temporary.unlink(missing_ok=True)