🐛 fix shortcut's dump & load

src/arclet/alconna/_internal/_analyser.py

@@ -288,7 +288,7 @@ def __repr__(self):
 
     def shortcut(
         self, argv: Argv[TDC], data: list[Any], short: Arparma | InnerShortcutArgs, reg: Match | None = None
-    ) -> Arparma[TDC]:
+    ) -> Arparma[TDC] | None:
         """处理被触发的快捷命令
 
         Args:
@@ -298,11 +298,13 @@ def shortcut(
             reg (Match | None): 可能的正则匹配结果
 
         Returns:
-            Arparma[TDC]: Arparma 解析结果
+            Arparma[TDC] | None: Arparma 解析结果
 
         Raises:
             ParamsUnmatched: 若不允许快捷命令后随其他参数，则抛出此异常
         """
+        self.reset()
+
         if isinstance(short, Arparma):
             return short
 
@@ -326,7 +328,7 @@ def shortcut(
             argv.ndata = 0
             argv.current_index = 0
             argv.addon(data)
-        return self.process(argv)
+        return
 
     def process(self, argv: Argv[TDC]) -> Arparma[TDC]:
         """主体解析函数, 应针对各种情况进行解析
@@ -366,8 +368,12 @@ def process(self, argv: Argv[TDC]) -> Arparma[TDC]:
                 argv.context[SHORTCUT_ARGS] = short
                 argv.context[SHORTCUT_REST] = rest
                 argv.context[SHORTCUT_REGEX_MATCH] = mat
-                self.reset()
-                return self.shortcut(argv, rest, short, mat)
+
+                if arp := self.shortcut(argv, rest, short, mat):
+                    return arp
+
+                self.header_result = self.header_handler(self.command_header, argv)
+                self.header_result.origin = _next
 
         except RuntimeError as e:
             exc = InvalidParam(lang.require("header", "error").format(target=argv.release(recover=True)[0]))

src/arclet/alconna/manager.py

@@ -10,6 +10,7 @@
 from datetime import datetime
 from typing import TYPE_CHECKING, Any, Match, MutableSet, Union
 from weakref import WeakValueDictionary
+from pathlib import Path
 
 from nepattern import TPattern
 from tarina import LRU, lang
@@ -47,7 +48,6 @@ def max_count(self) -> int:
     __shortcuts: dict[str, tuple[dict[str, Union[Arparma, InnerShortcutArgs]], dict[str, Union[Arparma, InnerShortcutArgs]]]]
 
     def __init__(self):
-        self.cache_path = f"{__file__.replace('manager.py', '')}manager_cache.db"
         self.sign = "ALCONNA::"
         self.current_count = 0
 
@@ -71,24 +71,49 @@ def _del():
 
         weakref.finalize(self, _del)
 
-    def load_cache(self) -> None:
+    def load_shortcuts(self, file: str | Path | None = None) -> None:
         """加载缓存"""
+        path = Path(file or (Path.cwd() / "shortcut.db"))
         with contextlib.suppress(FileNotFoundError, KeyError):
-            with shelve.open(self.cache_path) as db:
-                self.__shortcuts = dict(db["shortcuts"])  # type: ignore
-
-    def dump_cache(self) -> None:
+            with shelve.open(path.resolve().as_posix()) as db:
+                data: dict[str, tuple[dict, dict]] = dict(db["shortcuts"])  # type: ignore
+            for cmd, shorts in data.items():
+                _data = self.__shortcuts.setdefault(cmd, ({}, {}))
+                for key, short in shorts[0].items():
+                    if isinstance(short, dict):
+                        _data[0][key] = InnerShortcutArgs.load(short)
+                    else:
+                        _data[0][key] = short
+                for key, short in shorts[1].items():
+                    if isinstance(short, dict):
+                        _data[1][key] = InnerShortcutArgs.load(short)
+                    else:
+                        _data[1][key] = short
+
+    load_cache = load_shortcuts
+
+    def dump_shortcuts(self, file: str | Path | None = None) -> None:
         """保存缓存"""
         data = {}
-        for key, short in self.__shortcuts.items():
-            if isinstance(short, dict):
-                data[key] = {k: v for k, v in short.items() if k != "wrapper"}
-            else:
-                data[key] = short
-        with shelve.open(self.cache_path) as db:
+        for cmd, shorts in self.__shortcuts.items():
+            _data = data.setdefault(cmd, ({}, {}))
+            for key, short in shorts[0].items():
+                if isinstance(short, InnerShortcutArgs):
+                    _data[0][key] = short.dump()
+                else:
+                    _data[0][key] = short
+            for key, short in shorts[1].items():
+                if isinstance(short, InnerShortcutArgs):
+                    _data[1][key] = short.dump()
+                else:
+                    _data[1][key] = short
+        path = Path(file or (Path.cwd() / "shortcut.db"))
+        with shelve.open(path.resolve().as_posix()) as db:
             db["shortcuts"] = data
         data.clear()
 
+    dump_cache = dump_shortcuts
+
     @property
     def get_loaded_namespaces(self):
         """获取所有命名空间

src/arclet/alconna/typing.py

@@ -104,6 +104,28 @@ def __init__(
     def __repr__(self):
         return f"ShortcutArgs({self.command!r}, args={self.args!r}, fuzzy={self.fuzzy}, prefix={self.prefix})"
 
+    def dump(self):
+        return {
+            "command": self.command,
+            "args": self.args,
+            "fuzzy": self.fuzzy,
+            "prefix": self.prefix,
+            "prefixes": self.prefixes,
+            "flags": self.flags,
+        }
+
+    @classmethod
+    def load(cls, data: dict[str, Any]) -> InnerShortcutArgs:
+        return cls(
+            data["command"],
+            data.get("args"),
+            data.get("fuzzy", True),
+            data.get("prefix", False),
+            data.get("prefixes"),
+            data.get("wrapper"),
+            data.get("flags", 0),
+        )
+
 
 @runtime_checkable
 class DataCollection(Protocol[DataUnit]):

tests/core_test.py

@@ -466,6 +466,7 @@ def test_shortcut():
         # 构造体缩写传入；{i} 将被可能的正则匹配替换
         alc16.shortcut(r"TEST(\d+)(.+)", {"args": ["{0}", "bar {1}"]})
         res = alc16.parse("TEST123aa")
+        assert res.header_match.origin == "TEST123aa"
         assert res.matched is True
         assert res.foo == 123
         assert res.baz == "aa"
@@ -488,9 +489,11 @@ def test_shortcut():
         alc16_1.shortcut("echo", command="exec print({%0})")
         alc16_1.shortcut("echo1", command="exec print(\\'{*\n}\\')")
         res5 = alc16_1.parse("echo 123")
+        assert res5.header_match.origin == "echo"
         assert res5.content == "print(123)"
         assert not alc16_1.parse("echo 123 456").matched
         res6 = alc16_1.parse(["echo1", "123", "456 789"])
+        assert res6.header_match.origin == "echo1"
         assert res6.content == "print('123\n456\n789')"
         res7 = alc16_1.parse([123])
         assert not res7.matched