feat: store packaged user data under data directory

2026-07-07 14:12:32 +08:00
parent caa04a9aa7
commit 148c4a73eb
33 changed files with 602 additions and 196 deletions
@@ -1,24 +1,64 @@
 """Application-level configuration for cmshopee.

-This module owns `config.json` plus `config/ai_models.json`. `config.json`
-stores local app settings and AI role/generation parameters. AI provider
-definitions and local plaintext API keys live in ignored `config/ai_models.json`.
+This module owns `data/config.json` plus `data/config/ai_models.json`.
+`config.json` stores local app settings and AI role/generation parameters.
+AI provider definitions and local plaintext API keys live in ignored
+`data/config/ai_models.json`.
 """

 import copy
 import json
 import os
+import shutil
+import sys
+import tempfile
 import urllib.error
 import urllib.parse
 import urllib.request


-CONFIG_PATH = "config.json"
-AI_MODELS_PATH = os.path.join("config", "ai_models.json")
-CMHUB_CONFIG_PATH = os.path.join("config", "cmhub.json")
+DATA_DIR_NAME = "data"
+
+
+def app_base_dir() -> str:
+    """Return the application directory used as the parent of local data."""
+
+    if getattr(sys, "frozen", False):
+        return os.path.dirname(os.path.abspath(sys.executable))
+    return os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+
+
+def default_data_dir(base_dir=None) -> str:
+    """Return the default local user data directory."""
+
+    return os.path.abspath(os.path.join(base_dir or app_base_dir(), DATA_DIR_NAME))
+
+
+CONFIG_PATH = os.path.join(default_data_dir(), "config.json")
+AI_MODELS_PATH = os.path.join(default_data_dir(), "config", "ai_models.json")
+CMHUB_CONFIG_PATH = os.path.join(default_data_dir(), "config", "cmhub.json")
 CATEGORIES = {"text", "image"}
 API_TYPES = {"chat", "images_edits", "auto"}
 AI_BACKENDS = {"direct", "cmhub"}
+RUNTIME_CONFIG_KEYS = {
+    "config_path",
+    "ai_models_path",
+    "cmhub_config_path",
+    "data_dir",
+}
+LEGACY_USER_DATA_PATHS = (
+    "config.json",
+    "config",
+    "cmshopee.db",
+    "cmshopee.db-wal",
+    "cmshopee.db-shm",
+    "db.sqlite",
+    "chrome_user_data_dir",
+    "images",
+    "logs",
+    "prompts",
+    "title_prompt.txt",
+)

 DEFAULT_CONFIG = {
    "chrome_path": r"C:\Program Files\Google\Chrome\Application\chrome.exe",
@@ -104,6 +144,180 @@ class ConfigError(RuntimeError):
    """Raised when app configuration is missing or malformed."""


+class DataMigrationConflictError(ConfigError):
+    """Raised when old-layout data cannot be moved into data/ safely."""
+
+
+class DataDirectoryWriteError(ConfigError):
+    """Raised when the local data directory is not writable."""
+
+
+def _data_dir_for_config_path(path) -> str:
+    config_path = os.path.abspath(path or CONFIG_PATH)
+    if config_path == os.path.abspath(CONFIG_PATH):
+        return default_data_dir()
+    return os.path.dirname(config_path)
+
+
+def data_dir(config=None) -> str:
+    """Return the absolute local user data directory."""
+
+    if isinstance(config, dict):
+        if config.get("data_dir"):
+            return os.path.abspath(str(config["data_dir"]))
+        if config.get("config_path"):
+            return _data_dir_for_config_path(config["config_path"])
+    return default_data_dir()
+
+
+def _strip_runtime_config_keys(config):
+    if not isinstance(config, dict):
+        return config
+    return {
+        key: copy.deepcopy(value)
+        for key, value in config.items()
+        if key not in RUNTIME_CONFIG_KEYS
+    }
+
+
+def _runtime_paths(config_path, data_dir_path=None):
+    root = os.path.abspath(data_dir_path or _data_dir_for_config_path(config_path))
+    return {
+        "config_path": os.path.abspath(config_path or os.path.join(root, "config.json")),
+        "data_dir": root,
+        "ai_models_path": os.path.join(root, "config", "ai_models.json"),
+        "cmhub_config_path": os.path.join(root, "config", "cmhub.json"),
+    }
+
+
+def _with_runtime_paths(config, path):
+    result = copy.deepcopy(config)
+    result.update(_runtime_paths(path))
+    return result
+
+
+def _data_parent_dir(config=None) -> str:
+    return os.path.dirname(data_dir(config))
+
+
+def resolve_data_path(path, config=None) -> str:
+    """Resolve a user-data path under data_dir unless it is already absolute."""
+
+    text = str(path or "").strip()
+    if not text:
+        return ""
+    if os.path.isabs(text):
+        return os.path.abspath(text)
+    normalized = os.path.normpath(text)
+    first_part = normalized.split(os.sep, 1)[0]
+    if first_part == DATA_DIR_NAME:
+        return os.path.abspath(os.path.join(_data_parent_dir(config), normalized))
+    return os.path.abspath(os.path.join(data_dir(config), normalized))
+
+
+def data_path(*parts, config=None) -> str:
+    return os.path.abspath(os.path.join(data_dir(config), *[str(part) for part in parts]))
+
+
+def ai_models_config_path(config=None) -> str:
+    if isinstance(config, dict) and config.get("ai_models_path"):
+        return os.path.abspath(str(config["ai_models_path"]))
+    return data_path("config", "ai_models.json", config=config)
+
+
+def cmhub_config_file_path(config=None) -> str:
+    if isinstance(config, dict) and config.get("cmhub_config_path"):
+        return os.path.abspath(str(config["cmhub_config_path"]))
+    return data_path("config", "cmhub.json", config=config)
+
+
+def title_prompt_path(config=None) -> str:
+    return data_path("title_prompt.txt", config=config)
+
+
+def cover_prompts_dir(config=None) -> str:
+    return data_path("prompts", "cover", config=config)
+
+
+def diagnostic_log_dir(config=None) -> str:
+    return data_path("logs", config=config)
+
+
+def _is_within(path, parent) -> bool:
+    try:
+        return os.path.commonpath([os.path.abspath(path), os.path.abspath(parent)]) == os.path.abspath(parent)
+    except ValueError:
+        return False
+
+
+def migrate_legacy_user_data(base_dir=None, data_dir_path=None):
+    """Move old exe-top-level local data into data_dir without overwriting."""
+
+    base = os.path.abspath(base_dir or app_base_dir())
+    target_root = os.path.abspath(data_dir_path or default_data_dir(base))
+    if base == target_root:
+        return []
+    os.makedirs(target_root, exist_ok=True)
+
+    moves = []
+    seen_sources = set()
+    for relative in LEGACY_USER_DATA_PATHS:
+        source = os.path.abspath(os.path.join(base, relative))
+        if source in seen_sources or not os.path.exists(source):
+            continue
+        seen_sources.add(source)
+        if _is_within(source, target_root):
+            continue
+        if not _is_within(source, base):
+            raise ConfigError(f"旧数据路径不在程序目录内: {source}")
+        target = os.path.abspath(os.path.join(target_root, relative))
+        if not _is_within(target, target_root):
+            raise ConfigError(f"迁移目标路径不在 data 目录内: {target}")
+        moves.append((relative, source, target))
+
+    conflicts = [relative for relative, _source, target in moves if os.path.exists(target)]
+    if conflicts:
+        raise DataMigrationConflictError(
+            "检测到旧布局数据和 data 目录内数据同时存在，无法自动迁移。"
+            "请先手动合并或备份后再启动。冲突项: " + "、".join(conflicts)
+        )
+
+    moved = []
+    for relative, source, target in moves:
+        os.makedirs(os.path.dirname(target), exist_ok=True)
+        shutil.move(source, target)
+        moved.append(relative)
+    return moved
+
+
+def ensure_writable_data_dir(path=None) -> str:
+    root = os.path.abspath(path or default_data_dir())
+    try:
+        os.makedirs(root, exist_ok=True)
+        with tempfile.NamedTemporaryFile(
+            mode="w",
+            encoding="utf-8",
+            prefix=".cmshopee_write_test_",
+            dir=root,
+            delete=False,
+        ) as fh:
+            marker = fh.name
+            fh.write("ok")
+        os.remove(marker)
+    except Exception as exc:
+        raise DataDirectoryWriteError(
+            f"数据目录不可写：{root}。请把程序放到可写目录，勿放 Program Files。"
+        ) from exc
+    return root
+
+
+def prepare_data_dir(base_dir=None, data_dir_path=None, migrate=True) -> str:
+    root = os.path.abspath(data_dir_path or default_data_dir(base_dir))
+    if migrate:
+        migrate_legacy_user_data(base_dir=base_dir, data_dir_path=root)
+    return ensure_writable_data_dir(root)
+
+
 def default_config() -> dict:
    """Return a new copy of the default config."""

@@ -241,7 +455,7 @@ def _normalize_cmhub_config(config):
    if config is None:
        config = {}
    if not isinstance(config, dict):
-        raise ConfigError("config/cmhub.json 必须是对象")
+        raise ConfigError("cmhub 配置必须是对象")
    return {"api_key": str(config.get("api_key", "") or "")}


@@ -279,7 +493,7 @@ def get_cmhub_api_key(path=CMHUB_CONFIG_PATH, masked=False) -> str:
 def save_config(config, path=CONFIG_PATH) -> dict:
    """Persist config to JSON and return the normalized config."""

-    normalized = _deep_merge(DEFAULT_CONFIG, config)
+    normalized = _deep_merge(DEFAULT_CONFIG, _strip_runtime_config_keys(config))
    _normalize_config_values(normalized)
    _assert_no_secrets(normalized)
    directory = os.path.dirname(os.path.abspath(path))
@@ -288,7 +502,7 @@ def save_config(config, path=CONFIG_PATH) -> dict:
    with open(path, "w", encoding="utf-8") as fh:
        json.dump(normalized, fh, ensure_ascii=False, indent=2)
        fh.write("\n")
-    return normalized
+    return _with_runtime_paths(normalized, path)


 def load_config(path=CONFIG_PATH) -> dict:
@@ -304,7 +518,7 @@ def load_config(path=CONFIG_PATH) -> dict:
    normalized = _deep_merge(DEFAULT_CONFIG, loaded)
    _normalize_config_values(normalized)
    _assert_no_secrets(normalized)
-    return normalized
+    return _with_runtime_paths(normalized, path)


 def update_config(updates, path=CONFIG_PATH) -> dict:
@@ -323,15 +537,18 @@ def chrome_path(config=None) -> str:


 def user_data_root(config=None) -> str:
-    return _config_or_load(config).get("user_data_root", "chrome_user_data_dir")
+    cfg = _config_or_load(config)
+    return resolve_data_path(cfg.get("user_data_root", "chrome_user_data_dir"), cfg)


 def image_dir(config=None) -> str:
-    return _config_or_load(config).get("image_dir", "images")
+    cfg = _config_or_load(config)
+    return resolve_data_path(cfg.get("image_dir", "images"), cfg)


 def db_path(config=None) -> str:
-    return _config_or_load(config).get("db_path", "cmshopee.db")
+    cfg = _config_or_load(config)
+    return resolve_data_path(cfg.get("db_path", "cmshopee.db"), cfg)


 def default_debug_port(config=None) -> int:
@@ -446,7 +663,7 @@ def _normalize_ai_model(model):
 def _normalize_ai_models_config(config):
    models = config.get("models") if isinstance(config, dict) else None
    if not isinstance(models, list):
-        raise ConfigError("config/ai_models.json 必须包含 models 列表")
+        raise ConfigError("AI 模型清单必须包含 models 列表")
    normalized = {"models": [_normalize_ai_model(model) for model in models]}
    _assert_unique_model_names(normalized["models"])
    _assert_required_categories(normalized["models"])