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

novel_downloader/__init__.py

@@ -7,7 +7,7 @@ novel_downloader
 Core package for the Novel Downloader project.
 """
 
-__version__ = "1.2.0"
+__version__ = "1.2.2"
 
 __author__ = "Saudade Z"
 __email__ = "saudadez217@gmail.com"

novel_downloader/cli/download.py

@@ -57,6 +57,8 @@ def download_cli(ctx: Context, book_ids: List[str], site: str) -> None:
     parser_cfg = adapter.get_parser_config()
     saver_cfg = adapter.get_saver_config()
 
+    click.echo(t("download_site_mode", mode=downloader_cfg.mode))
+
     # If no book_ids provided on the command line, try to load them from config
     if not book_ids:
         try:

novel_downloader/config/adapter.py

@@ -23,6 +23,7 @@ from .models import (
     RequesterConfig,
     SaverConfig,
 )
+from .site_rules import load_site_rules
 
 
 class ConfigAdapter:
@@ -38,19 +39,43 @@ class ConfigAdapter:
         self._config = config
         self._site = site
 
+        site_rules = load_site_rules()  # -> Dict[str, SiteRules]
+        self._supported_sites = set(site_rules.keys())
+
     def set_site(self, site: str) -> None:
         """
         切换当前适配的站点
         """
         self._site = site
 
+    def _get_site_cfg(self) -> Dict[str, Any]:
+        """
+        统一获取站点配置:
+
+        1. 先尝试从 self._config["sites"][self._site] 取配置
+        2. 如果没有配置, 且 self._site 在 self._supported_sites 中, 则取 sites["common"]
+        3. 否则返回空 dict
+        """
+        sites_cfg = self._config.get("sites", {}) or {}
+
+        # 1. site-specific config
+        if self._site in sites_cfg:
+            return sites_cfg[self._site] or {}
+
+        # 2. fallback to "common" only if site is supported
+        if self._site in self._supported_sites:
+            return sites_cfg.get("common", {}) or {}
+
+        # 3. completely unsupported site
+        return {}
+
     def get_requester_config(self) -> RequesterConfig:
         """
         从 config["requests"] 中读取通用请求配置 (含 DrissionPage 设置)
         返回 RequesterConfig 实例
         """
         req = self._config.get("requests", {})
-        site_cfg = self._config.get("sites", {}).get(self._site, {})
+        site_cfg = self._get_site_cfg()
         return RequesterConfig(
             wait_time=req.get("wait_time", 5),
             retry_times=req.get("retry_times", 3),
@@ -73,7 +98,7 @@ class ConfigAdapter:
         """
         gen = self._config.get("general", {})
         debug = gen.get("debug", {})
-        site_cfg = self._config.get("sites", {}).get(self._site, {})
+        site_cfg = self._get_site_cfg()
         return DownloaderConfig(
             request_interval=gen.get("request_interval", 5),
             raw_data_dir=gen.get("raw_data_dir", "./raw_data"),
@@ -93,18 +118,21 @@ class ConfigAdapter:
         config["sites"][site] 中读取解析器相关配置, 返回 ParserConfig 实例
         """
         gen = self._config.get("general", {})
-        site_cfg = self._config.get("sites", {}).get(self._site, {})
+        font_ocr = gen.get("font_ocr", {})
+        site_cfg = self._get_site_cfg()
         return ParserConfig(
             cache_dir=gen.get("cache_dir", "./cache"),
-            decode_font=site_cfg.get("decode_font", False),
-            use_freq=site_cfg.get("use_freq", False),
-            use_ocr=site_cfg.get("use_ocr", True),
-            use_vec=site_cfg.get("use_vec", False),
-            ocr_version=site_cfg.get("ocr_version", "v1.0"),
-            save_font_debug=site_cfg.get("save_font_debug", False),
-            batch_size=site_cfg.get("batch_size", 32),
-            ocr_weight=site_cfg.get("ocr_weight", 0.6),
-            vec_weight=site_cfg.get("vec_weight", 0.4),
+            decode_font=font_ocr.get("decode_font", False),
+            use_freq=font_ocr.get("use_freq", False),
+            use_ocr=font_ocr.get("use_ocr", True),
+            use_vec=font_ocr.get("use_vec", False),
+            ocr_version=font_ocr.get("ocr_version", "v1.0"),
+            save_font_debug=font_ocr.get("save_font_debug", False),
+            batch_size=font_ocr.get("batch_size", 32),
+            gpu_mem=font_ocr.get("gpu_mem", 500),
+            gpu_id=font_ocr.get("gpu_id", None),
+            ocr_weight=font_ocr.get("ocr_weight", 0.6),
+            vec_weight=font_ocr.get("vec_weight", 0.4),
             mode=site_cfg.get("mode", "session"),
         )
 
@@ -136,7 +164,7 @@ class ConfigAdapter:
         """
         从 config["sites"][site]["book_ids"] 中提取目标书籍列表
         """
-        site_cfg = self._config.get("sites", {}).get(self._site, {})
+        site_cfg = self._get_site_cfg()
         raw_ids = site_cfg.get("book_ids", [])
 
         if isinstance(raw_ids, str):

novel_downloader/config/models.py

@@ -24,10 +24,10 @@ from typing import Any, Dict, List, Literal, Optional, TypedDict
 # === Requesters ===
 @dataclass
 class RequesterConfig:
-    wait_time: int = 5
+    wait_time: float = 5.0
     retry_times: int = 3
-    retry_interval: int = 5
-    timeout: int = 30
+    retry_interval: float = 5.0
+    timeout: float = 30.0
     headless: bool = True
     user_data_folder: str = ""
     profile_name: str = ""
@@ -41,7 +41,7 @@ class RequesterConfig:
 # === Downloaders ===
 @dataclass
 class DownloaderConfig:
-    request_interval: int = 5
+    request_interval: float = 5.0
     raw_data_dir: str = "./raw_data"
     cache_dir: str = "./novel_cache"
     download_workers: int = 4
@@ -63,6 +63,8 @@ class ParserConfig:
     use_vec: bool = False
     ocr_version: str = "v1.0"
     batch_size: int = 32
+    gpu_mem: int = 500
+    gpu_id: Optional[int] = None
     ocr_weight: float = 0.6
     vec_weight: float = 0.4
     save_font_debug: bool = False
@@ -133,16 +135,19 @@ class ChapterFieldRules(TypedDict):
     steps: List[RuleStep]
 
 
-class VolumesRules(TypedDict, total=False):
-    has_volume: bool  # 是否存在卷，false=未分卷
+class VolumesRulesOptional(TypedDict, total=False):
     volume_selector: str  # 有卷时选择 volume 块的 selector
-    chapter_selector: str  # 选择 chapter 节点的 selector
     volume_name_steps: List[RuleStep]
-    chapter_steps: List[ChapterFieldRules]  # 提取章节信息的步骤列表
     volume_mode: str  # Optional: "normal" (default) or "mixed"
     list_selector: str  # Optional: If "mixed" mode, parent container selector
 
 
+class VolumesRules(VolumesRulesOptional):
+    has_volume: bool  # 是否存在卷，false=未分卷
+    chapter_selector: str  # 选择 chapter 节点的 selector
+    chapter_steps: List[ChapterFieldRules]  # 提取章节信息的步骤列表
+
+
 class BookInfoRules(TypedDict, total=False):
     book_name: FieldRules
     author: FieldRules

novel_downloader/core/downloaders/base_async_downloader.py

@@ -94,7 +94,7 @@ class BaseAsyncDownloader(AsyncDownloaderProtocol, abc.ABC):
         return self._config.login_required
 
     @property
-    def request_interval(self) -> int:
+    def request_interval(self) -> float:
         return self._config.request_interval
 
     async def prepare(self) -> None:

novel_downloader/core/downloaders/common_downloader.py

@@ -67,8 +67,7 @@ class CommonDownloader(BaseDownloader):
         cache_base = self.cache_dir / site / book_id
         info_path = raw_base / "book_info.json"
         chapter_dir = raw_base / "chapters"
-        if save_html:
-            chapters_html_dir = cache_base / "html"
+        chapters_html_dir = cache_base / "html"
 
         raw_base.mkdir(parents=True, exist_ok=True)
         chapter_dir.mkdir(parents=True, exist_ok=True)

novel_downloader/core/downloaders/qidian_downloader.py

@@ -87,8 +87,7 @@ class QidianDownloader(BaseDownloader):
         info_path = raw_base / "book_info.json"
         chapter_dir = raw_base / "chapters"
         encrypted_chapter_dir = raw_base / "encrypted_chapters"
-        if save_html:
-            chapters_html_dir = cache_base / "html"
+        chapters_html_dir = cache_base / "html"
 
         raw_base.mkdir(parents=True, exist_ok=True)
         chapter_dir.mkdir(parents=True, exist_ok=True)

novel_downloader/core/factory/downloader_factory.py

@@ -14,7 +14,7 @@ based on the site name and parser mode specified in the configuration.
 To add support for new sites or modes, extend the `_site_map` accordingly.
 """
 
-from typing import Union
+from typing import Union, cast
 
 from novel_downloader.config import DownloaderConfig, load_site_rules
 from novel_downloader.core.downloaders import (
@@ -137,13 +137,15 @@ def get_downloader(
     :raises TypeError: If the provided requester does not match the required protocol
                     for the chosen mode (sync vs async).
     """
-    mode = config.mode.lower()
-    if mode == "async":
-        if not isinstance(requester, AsyncRequesterProtocol):
-            raise TypeError("Async mode requires an AsyncRequesterProtocol")
-        return get_async_downloader(requester, parser, saver, site, config)
-    if mode in ("browser", "session"):
-        if not isinstance(requester, RequesterProtocol):
-            raise TypeError("Sync mode requires a RequesterProtocol")
-        return get_sync_downloader(requester, parser, saver, site, config)
-    raise ValueError(f"Unknown mode '{config.mode}' for site '{site}'")
+    if requester.is_async():
+        if config.mode.lower() != "async":
+            raise TypeError("Requester is async, but config.mode is not 'async'")
+        async_requester = cast(AsyncRequesterProtocol, requester)
+        return get_async_downloader(async_requester, parser, saver, site, config)
+    else:
+        if config.mode.lower() not in ("browser", "session"):
+            raise TypeError(
+                "Requester is sync, but config.mode is not 'browser' or 'session'"
+            )
+        sync_requester = cast(RequesterProtocol, requester)
+        return get_sync_downloader(sync_requester, parser, saver, site, config)

novel_downloader/core/interfaces/async_requester_protocol.py

@@ -9,7 +9,7 @@ for book info pages, individual chapters, managing request lifecycle,
 and optionally retrieving a user's authenticated bookcase — all in async style.
 """
 
-from typing import Optional, Protocol, runtime_checkable
+from typing import Literal, Optional, Protocol, runtime_checkable
 
 
 @runtime_checkable
@@ -21,6 +21,9 @@ class AsyncRequesterProtocol(Protocol):
     and manage login/shutdown asynchronously.
     """
 
+    def is_async(self) -> Literal[True]:
+        ...
+
     async def login(self, max_retries: int = 3, manual_login: bool = False) -> bool:
         """
         Attempt to log in asynchronously.
@@ -28,7 +31,9 @@ class AsyncRequesterProtocol(Protocol):
         """
         ...
 
-    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.
 
@@ -39,7 +44,7 @@ class AsyncRequesterProtocol(Protocol):
         ...
 
     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.
@@ -51,7 +56,7 @@ class AsyncRequesterProtocol(Protocol):
         """
         ...
 
-    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 asynchronously.

novel_downloader/core/interfaces/requester_protocol.py

@@ -9,7 +9,7 @@ for book info pages, individual chapters, managing request lifecycle,
 and optionally retrieving a user's authenticated bookcase.
 """
 
-from typing import Optional, Protocol, runtime_checkable
+from typing import Literal, Optional, Protocol, runtime_checkable
 
 
 @runtime_checkable
@@ -20,13 +20,16 @@ class RequesterProtocol(Protocol):
       - a specific chapter page.
     """
 
+    def is_async(self) -> Literal[False]:
+        ...
+
     def login(self, max_retries: int = 3, manual_login: bool = False) -> bool:
         """
         Attempt to log in
         """
         ...
 
-    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.
 
@@ -37,7 +40,7 @@ class RequesterProtocol(Protocol):
         ...
 
     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.
@@ -55,7 +58,7 @@ class RequesterProtocol(Protocol):
         """
         ...
 
-    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.
 

novel_downloader/core/parsers/base_parser.py

@@ -45,14 +45,14 @@ class BaseParser(ParserProtocol, abc.ABC):
         self._base_cache_dir = Path(config.cache_dir)
 
     @abc.abstractmethod
-    def parse_book_info(self, html: str) -> Dict[str, Any]:
+    def parse_book_info(self, html_str: str) -> Dict[str, Any]:
         """
         Parse a book info page and extract metadata and chapter structure.
 
         Depending on the site structure, the return dict may include a
         flat `chapters` list or nested `volumes` with chapter groups.
 
-        :param html: Raw HTML of the book info page.
+        :param html_str: Raw HTML of the book info page.
         :return: Parsed metadata and chapter structure as a dictionary.
         """
         ...
@@ -62,7 +62,7 @@ class BaseParser(ParserProtocol, abc.ABC):
         """
         Parse a single chapter page and extract clean text or simplified HTML.
 
-        :param html: Raw HTML of the chapter page.
+        :param html_str: Raw HTML of the chapter page.
         :param chapter_id: Identifier of the chapter being parsed.
         :return: Cleaned chapter content as plain text or minimal HTML.
         """

novel_downloader/core/parsers/common_parser/helper.py

@@ -188,7 +188,7 @@ class HTMLExtractor:
                     current = sep.join(current)
 
             elif t == "attr":
-                name = step.get("attr")
+                name = step.get("attr") or ""
                 if isinstance(current, list):
                     current = [elem.get(name, "") for elem in current]
                 elif isinstance(current, Tag):
@@ -216,9 +216,9 @@ class HTMLExtractor:
         """
         list_selector = volume_rule.get("list_selector")
         volume_selector = volume_rule.get("volume_selector")
-        chapter_selector = volume_rule.get("chapter_selector")
         volume_name_steps = volume_rule.get("volume_name_steps")
-        chapter_steps_list = volume_rule.get("chapter_steps")
+        chapter_selector = volume_rule["chapter_selector"]
+        chapter_steps_list = volume_rule["chapter_steps"]
 
         if not (
             list_selector and volume_selector and chapter_selector and volume_name_steps
@@ -241,6 +241,8 @@ class HTMLExtractor:
         for elem in list_area.find_all(
             [volume_selector, chapter_selector], recursive=True
         ):
+            if not isinstance(elem, Tag):
+                continue
             if elem.name == volume_selector:
                 extractor = HTMLExtractor(str(elem))
                 volume_name = extractor.extract_field(volume_name_steps)
@@ -257,9 +259,9 @@ class HTMLExtractor:
         return volumes
 
     def extract_volume_blocks(self, volume_rule: VolumesRules) -> List[Dict[str, Any]]:
-        volume_selector = volume_rule["volume_selector"]
+        volume_selector = volume_rule.get("volume_selector")
+        volume_name_steps = volume_rule.get("volume_name_steps")
         chapter_selector = volume_rule["chapter_selector"]
-        volume_name_steps = volume_rule["volume_name_steps"]
         chapter_steps_list = volume_rule["chapter_steps"]
         if not (volume_selector and volume_name_steps):
             raise ValueError(

novel_downloader/core/parsers/qidian_parser/browser/chapter_encrypted.py

@@ -229,7 +229,7 @@ def extract_paragraphs_recursively(
     if chapter_id:
         main_id = f"c-{chapter_id}"
         main_tag = soup.find("main", id=main_id)
-        if not main_tag:
+        if not isinstance(main_tag, Tag):
             return []
     else:
         main_tag = soup

novel_downloader/core/parsers/qidian_parser/browser/main_parser.py

@@ -60,6 +60,8 @@ class QidianBrowserParser(BaseParser):
                 use_ocr=config.use_ocr,
                 use_vec=config.use_vec,
                 batch_size=config.batch_size,
+                gpu_mem=config.gpu_mem,
+                gpu_id=config.gpu_id,
                 ocr_weight=config.ocr_weight,
                 vec_weight=config.vec_weight,
                 font_debug=config.save_font_debug,
@@ -67,14 +69,14 @@ class QidianBrowserParser(BaseParser):
             self._font_debug_dir = self._base_cache_dir / "font_debug"
             self._font_debug_dir.mkdir(parents=True, exist_ok=True)
 
-    def parse_book_info(self, html: str) -> Dict[str, Any]:
+    def parse_book_info(self, html_str: str) -> Dict[str, Any]:
         """
         Parse a book info page and extract metadata and chapter structure.
 
-        :param html: Raw HTML of the book info page.
+        :param html_str: Raw HTML of the book info page.
         :return: Parsed metadata and chapter structure as a dictionary.
         """
-        return parse_book_info(html)
+        return parse_book_info(html_str)
 
     def parse_chapter(self, html_str: str, chapter_id: str) -> Dict[str, Any]:
         """

novel_downloader/core/parsers/qidian_parser/session/chapter_encrypted.py

@@ -245,7 +245,7 @@ def extract_paragraphs_recursively(
     if chapter_id > 0:
         main_id = f"c-{chapter_id}"
         main_tag = soup.find("main", id=main_id)
-        if not main_tag:
+        if not isinstance(main_tag, Tag):
             return []
     else:
         main_tag = soup

novel_downloader/core/parsers/qidian_parser/session/main_parser.py

@@ -63,6 +63,8 @@ class QidianSessionParser(BaseParser):
                 use_ocr=config.use_ocr,
                 use_vec=config.use_vec,
                 batch_size=config.batch_size,
+                gpu_mem=config.gpu_mem,
+                gpu_id=config.gpu_id,
                 ocr_weight=config.ocr_weight,
                 vec_weight=config.vec_weight,
                 font_debug=config.save_font_debug,
@@ -70,14 +72,14 @@ class QidianSessionParser(BaseParser):
             self._font_debug_dir = self._base_cache_dir / "font_debug"
             self._font_debug_dir.mkdir(parents=True, exist_ok=True)
 
-    def parse_book_info(self, html: str) -> Dict[str, Any]:
+    def parse_book_info(self, html_str: str) -> Dict[str, Any]:
         """
         Parse a book info page and extract metadata and chapter structure.
 
-        :param html: Raw HTML of the book info page.
+        :param html_str: Raw HTML of the book info page.
         :return: Parsed metadata and chapter structure as a dictionary.
         """
-        return parse_book_info(html)
+        return parse_book_info(html_str)
 
     def parse_chapter(self, html_str: str, chapter_id: str) -> Dict[str, Any]:
         """

novel_downloader/core/parsers/qidian_parser/shared/book_info_parser.py

@@ -41,6 +41,58 @@ def _get_volume_name(vol_div: Tag) -> str:
     return text.split(chr(183))[0].strip()
 
 
+def safe_select_text(
+    soup: Tag,
+    selector: str,
+    *,
+    separator: str = "",
+    strip: bool = False,
+    default: str = "",
+) -> str:
+    """
+    Safely select the first element matching a CSS selector and return its text.
+
+    :param soup: A BeautifulSoup Tag or sub-tree to query.
+    :param selector: A CSS selector string.
+    :param separator: Separator to use between strings when joining.
+    :param strip: Whether to strip whitespace from the result.
+    :param default: Value to return if no element is found.
+    :return: The element's text, or `default` if not found.
+    """
+    tag = soup.select_one(selector)
+    return (
+        tag.get_text(separator=separator, strip=strip)
+        if isinstance(tag, Tag)
+        else default
+    )
+
+
+def safe_select_attr(
+    soup: Tag,
+    selector: str,
+    attr: str,
+    *,
+    default: str = "",
+) -> str:
+    """
+    Safely select the first element matching a CSS selector and return one attributes.
+
+    :param soup: A BeautifulSoup Tag or sub-tree to query.
+    :param selector: A CSS selector string.
+    :param attr: The attribute name to retrieve from the selected element.
+    :param default: Value to return if no element or attribute is found.
+    :return: The attribute's value stripped of whitespace, or `default` if not found.
+    """
+    tag = soup.select_one(selector)
+    if isinstance(tag, Tag) and attr in tag.attrs:
+        value = tag.attrs[attr]
+        if isinstance(value, list):
+            return " ".join(value).strip()
+        elif isinstance(value, str):
+            return value.strip()
+    return default
+
+
 def parse_book_info(html_str: str) -> Dict[str, Any]:
     """
     Extract metadata: title, author, cover_url, update_time, status,
@@ -52,27 +104,24 @@ def parse_book_info(html_str: str) -> Dict[str, Any]:
     info: Dict[str, Any] = {}
     try:
         soup = html_to_soup(html_str)
-        info["book_name"] = soup.select_one("em#bookName").get_text(strip=True)
-        info["author"] = soup.select_one("a.writer").get_text(strip=True)
-        info["cover_url"] = soup.select_one("div.book-img img")["src"].strip()
+        info["book_name"] = safe_select_text(soup, "em#bookName", strip=True)
+        info["author"] = safe_select_text(soup, "a.writer", strip=True)
+        info["cover_url"] = safe_select_attr(soup, "div.book-img img", "src")
         info["update_time"] = (
-            soup.select_one("span.book-update-time")
-            .get_text(strip=True)
+            safe_select_text(soup, "span.book-update-time", strip=True)
             .replace("更新时间", "")
             .strip()
         )
-        info["serial_status"] = soup.select_one("span.blue").get_text(strip=True)
-        # word count via regex
-        match = re.search(
-            r"<em>([\d.]+)</em>\s*<cite>(.*?)字</cite>",
-            html_str,
+        info["serial_status"] = safe_select_text(soup, "span.blue", strip=True)
+
+        # Word count via regex fallback
+        match = re.search(r"<em>([\d.]+)</em>\s*<cite>(.*?)字</cite>", html_str)
+        info["word_count"] = (
+            f"{match.group(1)}{match.group(2)}字" if match else "Unknown"
         )
-        if match:
-            info["word_count"] = match.group(1) + match.group(2) + "字"
-        else:
-            info["word_count"] = "Unknown"
-        info["summary"] = soup.select_one("div.book-intro p").get_text(
-            separator="\n", strip=True
+
+        info["summary"] = safe_select_text(
+            soup, "div.book-intro p", separator="\n", strip=True
         )
         # volumes
         vols = []
@@ -81,11 +130,18 @@ def parse_book_info(html_str: str) -> Dict[str, Any]:
             chaps = []
             for li in vol_div.select("li"):
                 a = li.select_one("a")
+                if not isinstance(a, Tag) or "href" not in a.attrs:
+                    continue
+                href_val = a["href"]
+                if isinstance(href_val, list):
+                    href = href_val[0].strip()
+                else:
+                    href = str(href_val).strip()
                 chaps.append(
                     {
                         "title": a.get_text(strip=True),
-                        "url": a["href"].strip(),
-                        "chapterId": _chapter_url_to_id(a["href"]),
+                        "url": href,
+                        "chapterId": _chapter_url_to_id(href),
                     }
                 )
             vols.append({"volume_name": name, "chapters": chaps})

novel_downloader/core/parsers/qidian_parser/shared/helpers.py

@@ -16,7 +16,7 @@ import json
 import logging
 from typing import Any, Dict, Union
 
-from bs4 import BeautifulSoup
+from bs4 import BeautifulSoup, Tag
 
 logger = logging.getLogger(__name__)
 
@@ -103,7 +103,7 @@ def find_ssr_page_context(soup: BeautifulSoup) -> Dict[str, Any]:
     """
     try:
         tag = soup.find("script", id="vite-plugin-ssr_pageContext")
-        if tag and tag.string:
+        if isinstance(tag, Tag) and tag.string:
             data: Dict[str, Any] = json.loads(tag.string.strip())
             return data
     except Exception as e: