novel-downloader 1.2.0__py3-none-any.whl → 1.2.2__py3-none-any.whl

novel_downloader/core/requesters/base_async_session.py

@@ -13,7 +13,7 @@ cookie handling, and defines abstract methods for subclasses.
 import abc
 import asyncio
 import time
-from typing import Any, Dict, Optional, Union
+from typing import Any, Dict, Literal, Optional, Union
 
 import aiohttp
 from aiohttp import ClientResponse, ClientSession, ClientTimeout, TCPConnector
@@ -51,13 +51,16 @@ class BaseAsyncSession(AsyncRequesterProtocol, abc.ABC):
 
     Attributes:
         _session (ClientSession): The persistent aiohttp client session.
-        _timeout (int): Timeout for each request in seconds.
+        _timeout (float): Timeout for each request in seconds.
         _retry_times (int): Number of retry attempts on failure.
         _retry_interval (float): Delay (in seconds) between retries.
         _headers (Dict[str, str]): Default HTTP headers to send.
         _cookies (Dict[str, str]): Optional cookie jar for the session.
     """
 
+    def is_async(self) -> Literal[True]:
+        return True
+
     def _init_session(
         self,
         config: RequesterConfig,
@@ -111,7 +114,9 @@ class BaseAsyncSession(AsyncRequesterProtocol, abc.ABC):
         )
 
     @abc.abstractmethod
-    async def get_book_info(self, book_id: str, wait_time: Optional[int] = None) -> str:
+    async def get_book_info(
+        self, book_id: str, wait_time: Optional[float] = None
+    ) -> str:
         """
         Fetch the raw HTML (or JSON) of the book info page asynchronously.
 
@@ -123,7 +128,7 @@ class BaseAsyncSession(AsyncRequesterProtocol, abc.ABC):
 
     @abc.abstractmethod
     async def get_book_chapter(
-        self, book_id: str, chapter_id: str, wait_time: Optional[int] = None
+        self, book_id: str, chapter_id: str, wait_time: Optional[float] = None
     ) -> str:
         """
         Fetch the raw HTML (or JSON) of a single chapter asynchronously.
@@ -135,7 +140,7 @@ class BaseAsyncSession(AsyncRequesterProtocol, abc.ABC):
         """
         ...
 
-    async def get_bookcase(self, wait_time: Optional[int] = None) -> str:
+    async def get_bookcase(self, wait_time: Optional[float] = None) -> str:
         """
         Optional: Retrieve the HTML content of the authenticated user's bookcase page.
         Subclasses that support user login/bookcase should override this.
@@ -238,7 +243,7 @@ class BaseAsyncSession(AsyncRequesterProtocol, abc.ABC):
         return self._session
 
     @property
-    def timeout(self) -> int:
+    def timeout(self) -> float:
         """Return the default timeout setting."""
         return self._timeout
 

novel_downloader/core/requesters/base_browser.py

@@ -11,9 +11,10 @@ specialized purposes.
 
 import abc
 import logging
-from typing import Any, Dict, Optional
+from typing import Any, Dict, Literal, Optional, cast
 
-from DrissionPage import Chromium, ChromiumOptions, ChromiumPage
+from DrissionPage import Chromium, ChromiumOptions
+from DrissionPage._pages.mix_tab import MixTab
 
 from novel_downloader.config.models import RequesterConfig
 from novel_downloader.core.interfaces import RequesterProtocol
@@ -42,6 +43,9 @@ class BaseBrowser(RequesterProtocol, abc.ABC):
         _page (ChromiumPage): The active browser tab.
     """
 
+    def is_async(self) -> Literal[False]:
+        return False
+
     def _init_browser(self, config: RequesterConfig) -> None:
         """
         Initialize the browser with specified options from RequesterConfig.
@@ -99,7 +103,7 @@ class BaseBrowser(RequesterProtocol, abc.ABC):
         Set up the browser instance and open the default tab.
         """
         self._browser = Chromium(self._options)
-        self._page = self._browser.get_tab()
+        self._page = cast(MixTab, self._browser.get_tab())
 
     def login(self, max_retries: int = 3, manual_login: bool = False) -> bool:
         """
@@ -111,7 +115,7 @@ class BaseBrowser(RequesterProtocol, abc.ABC):
         )
 
     @abc.abstractmethod
-    def get_book_info(self, book_id: str, wait_time: Optional[int] = None) -> str:
+    def get_book_info(self, book_id: str, wait_time: Optional[float] = None) -> str:
         """
         Fetch the raw HTML (or JSON) of the book info page.
 
@@ -123,7 +127,7 @@ class BaseBrowser(RequesterProtocol, abc.ABC):
 
     @abc.abstractmethod
     def get_book_chapter(
-        self, book_id: str, chapter_id: str, wait_time: Optional[int] = None
+        self, book_id: str, chapter_id: str, wait_time: Optional[float] = None
     ) -> str:
         """
         Fetch the raw HTML (or JSON) of a single chapter.
@@ -135,7 +139,7 @@ class BaseBrowser(RequesterProtocol, abc.ABC):
         """
         ...
 
-    def get_bookcase(self, wait_time: Optional[int] = None) -> str:
+    def get_bookcase(self, wait_time: Optional[float] = None) -> str:
         """
         Optional: Retrieve the HTML content of the authenticated user's bookcase page.
 
@@ -151,7 +155,7 @@ class BaseBrowser(RequesterProtocol, abc.ABC):
         )
 
     @property
-    def page(self) -> ChromiumPage:
+    def page(self) -> Optional[MixTab]:
         """
         Return the current Chromium page object.
 
@@ -160,7 +164,7 @@ class BaseBrowser(RequesterProtocol, abc.ABC):
         return self._page
 
     @property
-    def browser(self) -> Chromium:
+    def browser(self) -> Optional[Chromium]:
         """
         Return the Chromium browser instance.
 

novel_downloader/core/requesters/base_session.py

@@ -10,7 +10,7 @@ persistent session and supports retries, headers, and timeout configurations.
 """
 
 import abc
-from typing import Any, Dict, Optional, Union
+from typing import Any, Dict, Literal, Optional, Union
 
 import requests
 from requests import Response, Session
@@ -28,9 +28,12 @@ class BaseSession(RequesterProtocol, abc.ABC):
 
     Attributes:
         _session (requests.Session): The persistent HTTP session.
-        _timeout (int): Timeout for each request in seconds.
+        _timeout (float): Timeout for each request in seconds.
     """
 
+    def is_async(self) -> Literal[False]:
+        return False
+
     def _init_session(
         self, config: RequesterConfig, cookies: Optional[Dict[str, str]] = None
     ) -> None:
@@ -81,7 +84,7 @@ class BaseSession(RequesterProtocol, abc.ABC):
         )
 
     @abc.abstractmethod
-    def get_book_info(self, book_id: str, wait_time: Optional[int] = None) -> str:
+    def get_book_info(self, book_id: str, wait_time: Optional[float] = None) -> str:
         """
         Fetch the raw HTML (or JSON) of the book info page.
 
@@ -93,7 +96,7 @@ class BaseSession(RequesterProtocol, abc.ABC):
 
     @abc.abstractmethod
     def get_book_chapter(
-        self, book_id: str, chapter_id: str, wait_time: Optional[int] = None
+        self, book_id: str, chapter_id: str, wait_time: Optional[float] = None
     ) -> str:
         """
         Fetch the raw HTML (or JSON) of a single chapter.
@@ -105,7 +108,7 @@ class BaseSession(RequesterProtocol, abc.ABC):
         """
         ...
 
-    def get_bookcase(self, wait_time: Optional[int] = None) -> str:
+    def get_bookcase(self, wait_time: Optional[float] = None) -> str:
         """
         Optional: Retrieve the HTML content of the authenticated user's bookcase page.
 
@@ -171,7 +174,7 @@ class BaseSession(RequesterProtocol, abc.ABC):
         return self._session
 
     @property
-    def timeout(self) -> int:
+    def timeout(self) -> float:
         """Return the default timeout setting."""
         return self._timeout
 

novel_downloader/core/requesters/common_requester/common_async_session.py

@@ -45,7 +45,9 @@ class CommonAsyncSession(BaseAsyncSession):
         self._site = site
         self._profile = profile
 
-    async def get_book_info(self, book_id: str, wait_time: Optional[int] = None) -> str:
+    async def get_book_info(
+        self, book_id: str, wait_time: Optional[float] = None
+    ) -> str:
         """
         Fetch the raw HTML of the book info page asynchronously.
 
@@ -62,7 +64,7 @@ class CommonAsyncSession(BaseAsyncSession):
         return html
 
     async def get_book_chapter(
-        self, book_id: str, chapter_id: str, wait_time: Optional[int] = None
+        self, book_id: str, chapter_id: str, wait_time: Optional[float] = None
     ) -> str:
         """
         Fetch the raw HTML of a single chapter asynchronously.

novel_downloader/core/requesters/common_requester/common_session.py

@@ -47,7 +47,7 @@ class CommonSession(BaseSession):
         self._site = site
         self._profile = profile
 
-    def get_book_info(self, book_id: str, wait_time: Optional[int] = None) -> str:
+    def get_book_info(self, book_id: str, wait_time: Optional[float] = None) -> str:
         """
         Fetch the raw HTML (or JSON) of the book info page.
 
@@ -64,7 +64,7 @@ class CommonSession(BaseSession):
                 with self.session.get(url, timeout=self.timeout) as response:
                     response.raise_for_status()
                     content = response.text
-                sleep_with_random_delay(base)
+                sleep_with_random_delay(base, add_spread=1.0)
                 return content
             except Exception as e:
                 if attempt == self.retry_times:
@@ -75,7 +75,7 @@ class CommonSession(BaseSession):
         raise RuntimeError("Unexpected error: get_book_info failed without returning")
 
     def get_book_chapter(
-        self, book_id: str, chapter_id: str, wait_time: Optional[int] = None
+        self, book_id: str, chapter_id: str, wait_time: Optional[float] = None
     ) -> str:
         """
         Fetch the raw HTML (or JSON) of a single chapter.
@@ -94,7 +94,7 @@ class CommonSession(BaseSession):
                 with self.session.get(url, timeout=self.timeout) as response:
                     response.raise_for_status()
                     content = response.text
-                sleep_with_random_delay(base)
+                sleep_with_random_delay(base, add_spread=1.0)
                 return content
             except Exception as e:
                 if attempt == self.retry_times:

novel_downloader/core/requesters/qidian_requester/qidian_broswer.py

@@ -15,10 +15,12 @@ import random
 import time
 from typing import Optional
 
+from DrissionPage._elements.chromium_element import ChromiumElement
 from DrissionPage.common import Keys
 
 from novel_downloader.config.models import RequesterConfig
 from novel_downloader.core.requesters.base_browser import BaseBrowser
+from novel_downloader.utils.i18n import t
 from novel_downloader.utils.time_utils import sleep_with_random_delay
 
 logger = logging.getLogger(__name__)
@@ -56,6 +58,8 @@ class QidianBrowser(BaseBrowser):
 
         :return: True if the user appears to be logged in, False otherwise.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         try:
             self._handle_overlay_mask()
             sign_in_elem = self._page.ele("@class=sign-in")
@@ -83,6 +87,8 @@ class QidianBrowser(BaseBrowser):
         :param max_retries: Maximum number of times to try clicking the login button.
         :return: True if login succeeds or is already in place; False otherwise.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         original_url = self._page.url
         try:
             self._page.get("https://www.qidian.com/")
@@ -107,7 +113,8 @@ class QidianBrowser(BaseBrowser):
 
         # return to original page
         try:
-            self._page.get(original_url)
+            if original_url:
+                self._page.get(original_url)
         except Exception as e:
             logger.debug("[auth] Failed to restore page URL: %s", e)
 
@@ -117,6 +124,8 @@ class QidianBrowser(BaseBrowser):
         """
         Detect and close any full-page overlay mask that might block the login UI.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         try:
             mask = self._page.ele("@@tag()=div@@class=mask", timeout=2)
             if not mask:
@@ -143,10 +152,12 @@ class QidianBrowser(BaseBrowser):
 
         :param attempt: The current attempt number (for logging).
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         try:
             logger.debug("[auth] Attempting login click (#%s).", attempt)
             login_btn = self._page.ele("@id=login-btn", timeout=5)
-            if login_btn:
+            if isinstance(login_btn, ChromiumElement):
                 login_btn.click()
                 logger.debug("[auth] Login button clicked.")
             else:
@@ -170,6 +181,8 @@ class QidianBrowser(BaseBrowser):
         :param max_retries: Number of times to check for login success.
         :return: True if login was detected, False otherwise.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         original_headless = self._headless
 
         # 1. Switch to headful mode if needed
@@ -193,13 +206,11 @@ class QidianBrowser(BaseBrowser):
                 logger.info("[auth] Detected successful login.")
                 self._logged_in = True
                 break
-
-            logger.info(
-                "[auth] Attempt %d/%d: Press Enter after completing login...",
-                attempt,
-                max_retries,
+            if attempt == 1:
+                print(t("login_prompt_intro"))
+            input(
+                t("login_prompt_press_enter", attempt=attempt, max_retries=max_retries)
             )
-            input()
         else:
             logger.warning("[auth] Manual login failed after %d attempts.", max_retries)
             self._logged_in = False
@@ -266,7 +277,7 @@ class QidianBrowser(BaseBrowser):
         """
         return self.QIDIAN_BOOKCASE_URL
 
-    def get_book_info(self, book_id: str, wait_time: Optional[int] = None) -> str:
+    def get_book_info(self, book_id: str, wait_time: Optional[float] = None) -> str:
         """
         Retrieve the HTML of a Qidian book info page.
 
@@ -279,14 +290,16 @@ class QidianBrowser(BaseBrowser):
                           If None, uses `self._config.wait_time`.
         :return: The HTML content of the book info page, or an empty string on error.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         url = self._build_book_info_url(book_id)
         try:
             # Navigate and fetch
             self._page.get(url)
 
-            # Randomized human‑like delay
+            # Randomized human-like delay
             base = wait_time if wait_time is not None else self._config.wait_time
-            sleep_with_random_delay(base, base * 0.2)
+            sleep_with_random_delay(base, mul_spread=1.2)
 
             html = str(self._page.html)
             logger.debug("[fetch] Fetched book info for ID %s from %s", book_id, url)
@@ -303,6 +316,8 @@ class QidianBrowser(BaseBrowser):
         :param presses: Number of DOWN key presses.
         :param pause: Seconds to wait between each press.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         for _ in range(presses):
             try:
                 self._page.actions.key_down(Keys.DOWN)
@@ -311,14 +326,14 @@ class QidianBrowser(BaseBrowser):
             time.sleep(pause)
 
     def get_book_chapter(
-        self, book_id: str, chapter_id: str, wait_time: Optional[int] = None
+        self, book_id: str, chapter_id: str, wait_time: Optional[float] = None
     ) -> str:
         """
         Retrieve the HTML content of a specific chapter.
 
         Ensures the user is logged in, navigates to the chapter page,
         waits a randomized delay to mimic human reading, then scrolls
-        to trigger any lazy‑loaded content.
+        to trigger any lazy-loaded content.
 
         :param book_id: The identifier of the book.
         :param chapter_id: The identifier of the chapter.
@@ -326,14 +341,16 @@ class QidianBrowser(BaseBrowser):
                           falls back to `self._config.wait_time`.
         :return: The HTML content of the chapter page, or empty string on error.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         url = self._build_chapter_url(book_id, chapter_id)
         try:
             # 1. Navigate to chapter URL
             self._page.get(url)
 
-            # 2. Randomized human‑like delay
+            # 2. Randomized human-like delay
             base = wait_time if wait_time is not None else self._config.wait_time
-            # sleep_with_random_delay(base, base*0.2)
+            # sleep_with_random_delay(base, mul_spread=1.2)
 
             # 3. Scroll down to load dynamic content
             presses = int(random.uniform(base, base + 5) * 2)
@@ -347,15 +364,17 @@ class QidianBrowser(BaseBrowser):
             logger.warning("[fetch] Error fetching chapter from '%s': %s", url, e)
             return ""
 
-    def get_bookcase(self, wait_time: Optional[int] = None) -> str:
+    def get_bookcase(self, wait_time: Optional[float] = None) -> str:
         """
-        Retrieve the HTML content of the logged‑in user's Qidian bookcase page.
+        Retrieve the HTML content of the logged-in user's Qidian bookcase page.
 
         :param wait_time: Base number of seconds to wait before returning content.
                           If None, falls back to `self._config.wait_time`.
         :return: The HTML markup of the bookcase page, or empty string on error.
         :raises RuntimeError: If the user is not logged in.
         """
+        if self._page is None:
+            raise RuntimeError("Browser page not initialized.")
         if not self._logged_in:
             raise RuntimeError("User not logged in. Please call login() first.")
 
@@ -364,9 +383,9 @@ class QidianBrowser(BaseBrowser):
             # Navigate to the bookcase page
             self._page.get(url)
 
-            # Randomized human‑like delay
+            # Randomized human-like delay
             base = wait_time if wait_time is not None else self._config.wait_time
-            sleep_with_random_delay(base, base * 0.2)
+            sleep_with_random_delay(base, mul_spread=1.2)
 
             html = str(self._page.html)
             logger.debug("[fetch] Fetched bookcase HTML from %s", url)

novel_downloader/core/requesters/qidian_requester/qidian_session.py

@@ -108,7 +108,7 @@ class QidianSession(BaseSession):
         self.get("https://www.qidian.com")
         return True
 
-    def get_book_info(self, book_id: str, wait_time: Optional[int] = None) -> str:
+    def get_book_info(self, book_id: str, wait_time: Optional[float] = None) -> str:
         """
         Fetch the raw HTML of the book info page.
 
@@ -123,7 +123,7 @@ class QidianSession(BaseSession):
             try:
                 resp = self.get(url)
                 resp.raise_for_status()
-                sleep_with_random_delay(base_delay, base_delay * 0.2)
+                sleep_with_random_delay(base_delay, mul_spread=1.2)
                 return resp.text
             except Exception as exc:
                 logger.warning(
@@ -140,7 +140,7 @@ class QidianSession(BaseSession):
         raise RuntimeError("Unexpected fall-through in get_book_info")
 
     def get_book_chapter(
-        self, book_id: str, chapter_id: str, wait_time: Optional[int] = None
+        self, book_id: str, chapter_id: str, wait_time: Optional[float] = None
     ) -> str:
         """
         Fetch the HTML of a single chapter.
@@ -157,7 +157,7 @@ class QidianSession(BaseSession):
             try:
                 resp = self.get(url)
                 resp.raise_for_status()
-                sleep_with_random_delay(base_delay, base_delay * 0.2)
+                sleep_with_random_delay(base_delay, mul_spread=1.2)
                 return resp.text
             except Exception as exc:
                 logger.warning(
@@ -174,7 +174,7 @@ class QidianSession(BaseSession):
 
         raise RuntimeError("Unexpected fall-through in get_book_chapter")
 
-    def get_bookcase(self, wait_time: Optional[int] = None) -> str:
+    def get_bookcase(self, wait_time: Optional[float] = None) -> str:
         """
         Retrieve the user's *bookcase* page.
 
@@ -186,7 +186,7 @@ class QidianSession(BaseSession):
             try:
                 resp = self.get(self.QIDIAN_BOOKCASE_URL, allow_redirects=True)
                 resp.raise_for_status()
-                sleep_with_random_delay(base_delay, base_delay * 0.2)
+                sleep_with_random_delay(base_delay, mul_spread=1.2)
                 return resp.text
             except Exception as exc:
                 logger.warning(

novel_downloader/core/savers/common_saver/common_epub.py

@@ -212,7 +212,7 @@ def common_save_as_epub(
 
     # --- 5. Finalize EPUB ---
     logger.info("%s Building TOC and spine...", TAG)
-    book.toc = tuple(toc_list)
+    book.toc = toc_list
     book.spine = spine
     book.add_item(epub.EpubNcx())
     book.add_item(epub.EpubNav())

novel_downloader/locales/en.json

@@ -58,6 +58,7 @@
   "download_option_site": "Website source, default is '{default}'.",
   "download_using_config": "Using config: {path}",
   "download_site_info": "Site: {site}",
+  "download_site_mode": "Mode: {mode}",
   "download_no_ids": "No book IDs provided. Exiting.",
   "download_fail_get_ids": "Failed to get book IDs from config: {err}",
   "download_only_example": "Only example book IDs found (e.g. '{example}').",
@@ -65,6 +66,9 @@
   "download_downloading": "Downloading book {book_id} from {site}...",
   "download_prompt_parse": "Parse...",
 
+  "login_prompt_intro": "Manual login is required. Please switch to the browser and log in.",
+  "login_prompt_press_enter": "Attempt {attempt}/{max_retries}: Press Enter after completing login in the browser...",
+
   "clean_logs": "Clean log directory",
   "clean_cache": "Clean scripts and browser cache",
   "clean_state": "Clean state files (state.json)",

novel_downloader/locales/zh.json

@@ -58,6 +58,7 @@
   "download_option_site": "网站来源, 默认为 '{default}'",
   "download_using_config": "使用配置: {path}",
   "download_site_info": "站点: {site}",
+  "download_site_mode": "使用模式: {mode}",
   "download_no_ids": "未提供书籍 ID, 正在退出",
   "download_fail_get_ids": "从配置获取书籍 ID 失败: {err}",
   "download_only_example": "只发现示例书籍 ID (例如 '{example}')",
@@ -65,6 +66,9 @@
   "download_downloading": "正在从 {site} 下载书籍 {book_id}...",
   "download_prompt_parse": "结束...",
 
+  "login_prompt_intro": "需要手动登录, 请切换到浏览器窗口完成登录",
+  "login_prompt_press_enter": "第 {attempt}/{max_retries} 次尝试: 请在浏览器中完成登录后按回车键...",
+
   "clean_logs": "清理日志目录",
   "clean_cache": "清理脚本和浏览器缓存",
   "clean_state": "清理状态文件 (state.json)",

novel_downloader/resources/config/settings.yaml

@@ -1,9 +1,9 @@
 # 网络请求层设置
 requests:
-  wait_time: 5                      # 每次请求等待时间 (秒)
+  wait_time: 5.0                    # 每次请求等待时间 (秒)
   retry_times: 3                    # 请求失败重试次数
-  retry_interval: 5
-  timeout: 30                       # 页面加载超时时间 (秒)
+  retry_interval: 5.0
+  timeout: 30.0                     # 页面加载超时时间 (秒)
   max_rps: null                     # 最大请求速率 (requests per second), 为 null 则不限制
   # DrissionPage 专用设置
   headless: false                   # 是否以无头模式启动浏览器
@@ -15,7 +15,7 @@ requests:
 
 # 全局通用设置
 general:
-  request_interval: 5               # 同一本书各章节请求间隔 (秒)
+  request_interval: 5.0             # 同一本书各章节请求间隔 (秒)
   raw_data_dir: "./raw_data"        # 原始章节 HTML/JSON 存放目录
   output_dir: "./downloads"         # 最终输出文件存放目录
   cache_dir: "./novel_cache"        # 本地缓存目录 (字体 / 图片等)
@@ -26,6 +26,18 @@ general:
   debug:
     save_html: false                # 是否将抓取到的原始 HTML 保留到磁盘
     log_level: "INFO"               # 日志级别: DEBUG, INFO, WARNING, ERROR
+  font_ocr:
+    decode_font: false              # 是否尝试本地解码混淆字体
+    use_freq: false                 # 是否使用频率分析
+    ocr_version: "v2.0"             # "v1.0" / "v2.0"
+    use_ocr: true                   # 是否使用 OCR 辅助识别文本
+    use_vec: false                  # 是否使用 Vector 辅助识别文本
+    save_font_debug: false          # 是否保存字体解码调试数据
+    batch_size: 32
+    gpu_mem: 500                    # GPU 显存限制 (MB)
+    gpu_id: null                    # 使用哪个 GPU
+    ocr_weight: 0.6
+    vec_weight: 0.4
 
 # 各站点的特定配置
 sites:
@@ -38,15 +50,6 @@ sites:
       - "0000000000"
     mode: "browser"                 # browser / session
     login_required: true            # 是否需要登录才能访问
-    decode_font: false              # 是否尝试本地解码混淆字体
-    use_freq: false                 # 是否使用频率分析
-    ocr_version: "v2.0"             # "v1.0" / "v2.0"
-    use_ocr: true                   # 是否使用 OCR 辅助识别文本
-    use_vec: false                  # 是否使用 Vector 辅助识别文本
-    save_font_debug: false          # 是否保存字体解码调试数据
-    batch_size: 32
-    ocr_weight: 0.6
-    vec_weight: 0.4
     #
   sample_site:
     book_ids:

novel_downloader/utils/constants.py

@@ -26,11 +26,12 @@ LOGGER_NAME = PACKAGE_NAME  # Root logger name
 # -----------------------------------------------------------------------------
 # Base config directory (e.g. ~/AppData/Local/novel_downloader/)
 BASE_CONFIG_DIR = Path(user_config_dir(APP_DIR_NAME, appauthor=False))
+WORK_DIR = Path.cwd()
 PACKAGE_ROOT: Path = Path(__file__).parent.parent
 LOCALES_DIR: Path = PACKAGE_ROOT / "locales"
 
 # Subdirectories under BASE_CONFIG_DIR
-LOGGER_DIR = BASE_CONFIG_DIR / "logs"
+LOGGER_DIR = WORK_DIR / "logs"
 JS_SCRIPT_DIR = BASE_CONFIG_DIR / "scripts"
 STATE_DIR = BASE_CONFIG_DIR / "state"
 DATA_DIR = BASE_CONFIG_DIR / "data"

novel_downloader/utils/fontocr/ocr_v2.py

@@ -221,6 +221,8 @@ class FontOCRV2:
         use_ocr: bool = True,
         use_vec: bool = False,
         batch_size: int = 32,
+        gpu_mem: int = 500,
+        gpu_id: Optional[int] = None,
         ocr_weight: float = 0.6,
         vec_weight: float = 0.4,
         ocr_version: str = "v1.0",
@@ -232,6 +234,8 @@ class FontOCRV2:
         self.use_ocr = use_ocr
         self.use_vec = use_vec
         self.batch_size = batch_size
+        self.gpu_mem = gpu_mem
+        self.gpu_id = gpu_id
         self.ocr_weight = ocr_weight
         self.vec_weight = vec_weight
         self.ocr_version = ocr_version
@@ -279,6 +283,8 @@ class FontOCRV2:
             rec_batch_num=self.batch_size,
             use_space_char=False,
             use_gpu=gpu_available,
+            gpu_mem=self.gpu_mem,
+            gpu_id=self.gpu_id,
         )
 
     def _load_char_freq_db(self) -> bool:

novel_downloader/utils/time_utils/datetime_utils.py

@@ -106,7 +106,7 @@ def calculate_time_difference(
     """
     Calculate the difference between two datetime values.
 
-    :param from_time_str: Date‐time string "YYYY‑MM‑DD HH:MM:SS" for the start.
+    :param from_time_str: Date‐time string "YYYY-MM-DD HH:MM:SS" for the start.
     :param tz_str:        Timezone of from_time_str, e.g. 'UTC+8'. Defaults to 'UTC'.
     :param to_time_str:   Optional date‐time string for the end; if None, uses now().
     :param to_tz_str:     Timezone of to_time_str. Defaults to 'UTC'.