xhs-note-extractor 0.1.dev2__py3-none-any.whl → 0.1.dev6__py3-none-any.whl

xhs_note_extractor/extractor.py

@@ -18,6 +18,7 @@ import requests
 import logging
 from typing import Dict, List, Optional, Union
 from urllib.parse import urlparse, parse_qs
+import xml.etree.ElementTree as ET
 
 # 配置日志
 logging.basicConfig(
@@ -35,21 +36,35 @@ class XHSNoteExtractor:
     包括URL解析、设备连接、页面跳转和笔记内容提取。
     """
     
-    def __init__(self, device_serial: Optional[str] = None):
+    def __init__(self, device_serial: Optional[str] = None, enable_time_logging: bool = True):
         """
         初始化小红书笔记提取器
         
         Args:
             device_serial (str, optional): 设备序列号，如果为None则自动连接可用设备
+            enable_time_logging (bool, optional): 是否启用耗时打印，默认为True
             
         Raises:
             RuntimeError: 当没有可用设备时抛出异常
         """
         self.device = None
         self.device_serial = device_serial
+        self.enable_time_logging = enable_time_logging
         if not self.connect_device():
             raise RuntimeError("未找到可用的Android设备，请连接设备后再试")
     
+    def _time_method(self, method_name, start_time):
+        """
+        记录方法执行时间
+        
+        Args:
+            method_name (str): 方法名称
+            start_time (float): 开始时间
+        """
+        if self.enable_time_logging:
+            elapsed_time = time.time() - start_time
+            logger.info(f"[{method_name}] 耗时: {elapsed_time:.3f}秒")
+    
     def connect_device(self) -> bool:
         """
         连接设备
@@ -57,17 +72,35 @@ class XHSNoteExtractor:
         Returns:
             bool: 是否成功连接设备
         """
+        start_time = time.time()
         try:
             if self.device_serial:
                 self.device = u2.connect(self.device_serial)
             else:
                 self.device = u2.connect()
             logger.info(f"✓ 已连接设备: {self.device.serial}")
+            self._time_method("connect_device", start_time)
             return True
         except Exception as e:
             logger.error(f"✗ 设备连接失败: {e}")
+            self._time_method("connect_device", start_time)
             return False
-    
+    def is_device_connected(self) -> bool:
+        """
+        检查设备是否仍然连接
+        
+        Returns:
+            bool: 设备是否连接
+        """
+        if not self.device:
+            return False
+        try:
+            # 通过获取设备信息来验证连接
+            self.device.info
+            return True
+        except:
+            return False
+
     @staticmethod
     def parse_xhs_url(url: str) -> Dict[str, str]:
         """
@@ -82,6 +115,7 @@ class XHSNoteExtractor:
         Raises:
             ValueError: 当URL格式不正确时抛出异常
         """
+        start_time = time.time()
         # 处理xhsdiscover协议格式
         if url.startswith("xhsdiscover://"):
             # 提取note_id
@@ -122,13 +156,22 @@ class XHSNoteExtractor:
                     note_id = path_parts[explore_index + 1]
                 else:
                     raise ValueError("URL中缺少note_id")
+            # 兼容 /discovery/item/ 格式
+            elif 'discovery' in path_parts and 'item' in path_parts:
+                item_index = path_parts.index('item')
+                if item_index + 1 < len(path_parts):
+                    note_id = path_parts[item_index + 1]
+                else:
+                    raise ValueError("URL中缺少note_id")
             else:
-                raise ValueError("URL格式不正确，缺少/explore/路径")
+                raise ValueError("URL格式不正确，缺少/explore/或/discovery/item/路径")
             
             # 提取查询参数中的xsec_token
             query_params = parse_qs(parsed_url.query)
             xsec_token = query_params.get('xsec_token', [''])[0]
             
+            elapsed_time = time.time() - start_time
+            logger.info(f"[parse_xhs_url] 耗时: {elapsed_time:.3f}秒")
             return {
                 "note_id": note_id,
                 "xsec_token": xsec_token,
@@ -136,6 +179,8 @@ class XHSNoteExtractor:
             }
         
         else:
+            elapsed_time = time.time() - start_time
+            logger.info(f"[parse_xhs_url] 耗时: {elapsed_time:.3f}秒")
             raise ValueError("不支持的URL格式")
     
     @staticmethod
@@ -167,12 +212,18 @@ class XHSNoteExtractor:
         Returns:
             str: xhsdiscover协议格式的URL
         """
+        start_time = time.time()
+        result = ""
         if xsec_token:
             original_url = f"http://www.xiaohongshu.com/explore/{note_id}?xsec_token={xsec_token}&xsec_source=pc_feed"
             encoded_url = requests.utils.quote(original_url)
-            return f"xhsdiscover://item/{note_id}?open_url={encoded_url}"
+            result = f"xhsdiscover://item/{note_id}?open_url={encoded_url}"
         else:
-            return f"xhsdiscover://item/{note_id}"
+            result = f"xhsdiscover://item/{note_id}"
+        
+        elapsed_time = time.time() - start_time
+        logger.info(f"[convert_to_xhsdiscover_format] 耗时: {elapsed_time:.3f}秒")
+        return result
     
     def extract_note_data(self, url: Optional[str] = None, note_id: Optional[str] = None, 
                          xsec_token: Optional[str] = None) -> Dict[str, Union[str, List[str]]]:
@@ -191,6 +242,7 @@ class XHSNoteExtractor:
             RuntimeError: 当设备未连接时抛出异常
             Exception: 当提取过程中出现错误时抛出异常
         """
+        start_time = time.time()
         # 如果提供了URL，则先解析它（验证URL有效性）
         if url:
             parsed_data = self.parse_xhs_url(url)
@@ -199,6 +251,7 @@ class XHSNoteExtractor:
             
         # 检查设备是否连接
         if self.device is None:
+            self._time_method("extract_note_data", start_time)
             raise RuntimeError("设备未连接，请先连接设备")
         
         # 构建跳转URL
@@ -216,102 +269,181 @@ class XHSNoteExtractor:
             
             logger.info(f"✓ 成功提取笔记数据，点赞数: {data['likes']}, 图片数: {len(data['image_urls'])}")
             
+            self._time_method("extract_note_data", start_time)
             return data
             
         except Exception as e:
             logger.error(f"✗ 提取笔记数据失败: {e}")
+            self._time_method("extract_note_data", start_time)
             raise
     
     def _get_detail_data(self) -> Dict[str, Union[str, List[str]]]:
         """
         从当前已经打开的小红书详情页提取完整正文、图片和点赞数。
-        这是xhs_utils.get_detail_data的封装版本，保持相同功能。
+        优化版本: 使用 dump_hierarchy 替代遍历，大幅提升速度。
         
         Returns:
             Dict[str, Union[str, List[str]]]: 包含笔记数据的字典
         """
-        logger.info("🔍 进入深度提取模式...")
+        start_time = time.time()
+        logger.info("🔍 进入深度提取模式 (XML优化版)...")
         
-        # 1. 验证是否进入详情页 (增加重试和多关键词检测)
+        # 1. 验证是否进入详情页 & 展开全文
         detail_loaded = False
         detail_keywords = ["说点什么", "写评论", "写点什么", "收藏", "点赞", "评论", "分享", "发弹幕"]
-        for i in range(8):
+        
+        # 尝试点击展开 (预先动作)
+        try:
+            # 快速检查是否有展开按钮
+            for btn_text in ["展开", "查看全部", "全文"]:
+                if self.device(text=btn_text).exists:
+                    self.device(text=btn_text).click()
+                    break
+        except: pass
+
+        # 等待加载完整
+        for i in range(5):
             if any(self.device(textContains=kw).exists or self.device(descriptionContains=kw).exists for kw in detail_keywords):
                 detail_loaded = True
                 break
-            if i == 4:
+            if i == 2:
                 # 可能是视频，点击屏幕中心尝试激活 UI
                 self.device.click(540, 900)
-            time.sleep(1)
+            time.sleep(0.5)
         
         if not detail_loaded:
             logger.warning("⚠ 警告：详情页特征未发现，提取可能不完整")
 
-        # 2. 处理"展开"按钮以获取完整长文
-        for btn_text in ["展开", "查看全部", "全文"]:
-            btn = self.device(text=btn_text)
-            if btn.exists:
-                logger.info(f"[Action] 点击'{btn_text}'")
-                btn.click()
-                time.sleep(1)
-
-        # 3. 提取正文 (多策略拼接)
+        # 2. 获取 UI层级 (核心优化)
+        xml_dump_start = time.time()
+        xml_content = self.device.dump_hierarchy()
+        self._time_method("dump_hierarchy", xml_dump_start)
+        
+        # 3. 解析 XML
+        root = ET.fromstring(xml_content)
+        
         content = ""
-        # 策略 A: 尝试常见 ID
-        desc_el = self.device(resourceIdMatches=".*desc.*|.*content.*")
-        if desc_el.exists:
-            content = desc_el.get_text()
-        
-        # 策略 B: 文本容器遍历 (更稳健)
-        if not content or len(content) < 20:
-            texts = []
-            for el in self.device(className="android.widget.TextView"):
+        likes = "0"
+        collects = "0"
+        comments = "0"
+        author_name = "Unknown"
+        image_urls = []
+        
+        # 收集所有 TextView 节点信息
+        text_nodes = []
+        
+        def parse_nodes(node):
+            if node.attrib.get('class') == 'android.widget.TextView':
+                text = node.attrib.get('text', '')
+                bounds_str = node.attrib.get('bounds', '[0,0][0,0]')
+                # 解析 bounds: [x1,y1][x2,y2]
                 try:
-                    t = el.get_text()
-                    if not t or len(t) < 3: continue
-                    # 过滤坐标：只取屏幕中间内容区
-                    b = el.info.get('bounds', {})
-                    if 200 < b.get('top', 0) < 2100:
-                        if not any(k in t for k in ['收藏', '点赞', '评论', '分享', '发布于', '说点什么', '条评论']):
-                            texts.append(t)
-                except: continue
-            content = "\n".join(texts)
+                    coords = bounds_str.replace('][', ',').replace('[', '').replace(']', '').split(',')
+                    x1, y1, x2, y2 = map(int, coords)
+                    if text:
+                        text_nodes.append({
+                            'text': text,
+                            'l': x1, 't': y1, 'r': x2, 'b': y2,
+                            'cx': (x1 + x2) / 2, 'cy': (y1 + y2) / 2
+                        })
+                except: pass
+            for child in node:
+                parse_nodes(child)
+                
+        parse_nodes(root)
+        
+        # 4. 分析节点数据
+        
+        # A. 作者提取 (寻找 "关注" 附近的文本)
+        # 策略: 找到包含 "关注" 的节点，取其左侧最近的节点
+        follow_node = None
+        for n in text_nodes:
+            if n['text'] in ["关注", "已关注"]:
+                follow_node = n
+                break
+        
+        if follow_node:
+            best_dist = 9999
+            for n in text_nodes:
+                if n == follow_node: continue
+                if n['text'] in ["关注", "已关注"] or len(n['text']) > 30: continue
+                
+                # 垂直接近
+                if abs(n['cy'] - follow_node['cy']) < 100:
+                    # 在左侧
+                    if n['r'] <= follow_node['l'] + 50:
+                        dist = follow_node['l'] - n['r']
+                        if dist < best_dist:
+                            best_dist = dist
+                            author_name = n['text']
+            logger.info(f"✓ 识别到作者: {author_name}")
 
-        # 4. 提取图片 (通过分享链接解析高清图)
-        image_urls = []
+        # B. 互动数据提取 (底部区域)
+        bottom_nodes = [n for n in text_nodes if n['t'] > 2000] # 假设屏幕高度足够
+        bottom_nodes.sort(key=lambda x: x['l']) # 从左到右
+        
+        for n in bottom_nodes:
+            txt = n['text']
+            num_txt = ''.join(c for c in txt if c.isdigit() or c in ['.', 'w', 'W'])
+            if not num_txt: continue
+            
+            cx = n['cx']
+            if 500 < cx < 750:
+                likes = num_txt
+            elif 750 < cx < 900:
+                collects = num_txt
+            elif cx >= 900:
+                comments = num_txt
+
+        # C. 正文提取
+        # 过滤掉非正文内容
+        content_lines = []
+        exclude_keywords = ['收藏', '点赞', '评论', '分享', '发布于', '说点什么', '条评论', '关注', author_name]
+        
+        # 按照垂直位置排序
+        content_nodes = [n for n in text_nodes if 200 < n['t'] < 2000]
+        content_nodes.sort(key=lambda x: x['t'])
+        
+        for n in content_nodes:
+            t = n['text']
+            if len(t) < 2: continue
+            if any(k in t for k in exclude_keywords): continue
+            
+            # 简单的去重策略
+            if content_lines and t in content_lines[-1]: continue
+            content_lines.append(t)
+            
+        content = "\n".join(content_lines)
+
+        # 5. 图片提取 (保持原有逻辑但优化等待)
         try:
+             # 这里还是需要交互，无法纯靠XML
             share_btn = self.device(description="分享")
             if share_btn.exists:
                 share_btn.click()
-                time.sleep(1.5)
-                copy_link_btn = self.device(text="复制链接")
-                if copy_link_btn.exists:
-                    copy_link_btn.click()
+                # 显式等待 "复制链接"
+                copy_link = self.device(text="复制链接")
+                if copy_link.wait(timeout=2.0):
+                    copy_link.click()
+                    # 等待剪贴板更新? 稍微缓一下
                     time.sleep(0.5)
                     share_link = self.device.clipboard
                     if "http" in str(share_link):
                         image_urls = self._fetch_web_images(share_link)
                 else:
+                    logger.warning("未找到复制链接按钮")
                     self.device.press("back")
         except Exception as e:
             logger.warning(f"⚠ 图片提取异常: {e}")
 
-        # 5. 提取点赞数
-        likes = "0"
-        try:
-            for el in self.device(className="android.widget.TextView"):
-                txt = el.get_text() or ""
-                if any(c.isdigit() for c in txt):
-                    b = el.info.get('bounds', {})
-                    if b.get('top', 0) > 2000 and b.get('left', 0) > 500:
-                        likes = ''.join(c for c in txt if c.isdigit() or c in ['.', 'w', 'W'])
-                        if likes: break
-        except: pass
-
+        self._time_method("_get_detail_data", start_time)
         return {
             "content": content,
             "image_urls": image_urls,
-            "likes": likes
+            "likes": likes,
+            "collects": collects,
+            "comments": comments,
+            "author_name": author_name
         }
     
     def _fetch_web_images(self, url: str) -> List[str]:
@@ -324,6 +456,7 @@ class XHSNoteExtractor:
         Returns:
             List[str]: 图片URL列表
         """
+        start_time = time.time()
         try:
             headers = {"User-Agent": "Mozilla/5.0 (iPhone; CPU iPhone OS 14_8 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/14.1.2 Mobile/15E148 Safari/604.1"}
             res = requests.get(url, headers=headers, timeout=10)
@@ -339,8 +472,10 @@ class XHSNoteExtractor:
                 for m in matches:
                     clean_url = m.replace('\\u002F', '/')
                     if clean_url not in found: found.append(clean_url)
+            self._time_method("_fetch_web_images", start_time)
             return found
         except:
+            self._time_method("_fetch_web_images", start_time)
             return []
     
     def save_note_data(self, data: Dict[str, Union[str, List[str]]], 
@@ -354,6 +489,7 @@ class XHSNoteExtractor:
             filename (str): 保存文件名
             note_url (str): 笔记URL
         """
+        start_time = time.time()
         try:
             with open(filename, "w", encoding="utf-8") as f:
                 f.write("=" * 50 + "\n")
@@ -362,8 +498,11 @@ class XHSNoteExtractor:
                 if note_url:
                     f.write(f"笔记URL: {note_url}\n")
                     f.write("=" * 50 + "\n")
-                f.write(f"点赞数: {data['likes']}\n")
-                f.write(f"图片数: {len(data['image_urls'])}\n")
+                f.write(f"作者: {data.get('author_name', 'Unknown')}\n")
+                f.write(f"点赞数: {data.get('likes', '0')}\n")
+                f.write(f"收藏数: {data.get('collects', '0')}\n")
+                f.write(f"评论数: {data.get('comments', '0')}\n")
+                f.write(f"图片数: {len(data.get('image_urls', []))}\n")
                 f.write("=" * 50 + "\n")
                 f.write("【正文内容】\n")
                 f.write(data['content'])
@@ -375,24 +514,32 @@ class XHSNoteExtractor:
                     f.write("=" * 50 + "\n")
             
             logger.info(f"✓ 笔记数据已保存到: {filename}")
+            self._time_method("save_note_data", start_time)
         except Exception as e:
             logger.error(f"✗ 保存笔记数据失败: {e}")
+            self._time_method("save_note_data", start_time)
             raise
 
 
-def extract_note_from_url(url: str, device_serial: Optional[str] = None) -> Dict[str, Union[str, List[str]]]:
+def extract_note_from_url(url: str, device_serial: Optional[str] = None, enable_time_logging: bool = True) -> Dict[str, Union[str, List[str]]]:
     """
     便捷函数：直接从URL提取笔记数据
     
     Args:
         url (str): 小红书笔记URL
         device_serial (str, optional): 设备序列号
+        enable_time_logging (bool, optional): 是否启用耗时打印，默认为True
         
     Returns:
         Dict[str, Union[str, List[str]]]: 笔记数据
     """
-    extractor = XHSNoteExtractor(device_serial=device_serial)
-    return extractor.extract_note_data(url=url)
+    start_time = time.time()
+    logger.info(f"[extract_note_from_url] 开始处理URL: {url}")
+    extractor = XHSNoteExtractor(device_serial=device_serial, enable_time_logging=enable_time_logging)
+    result = extractor.extract_note_data(url=url)
+    elapsed_time = time.time() - start_time
+    logger.info(f"[extract_note_from_url] 总耗时: {elapsed_time:.3f}秒")
+    return result
 
 
 def convert_url_format(url: str) -> str:
@@ -405,8 +552,13 @@ def convert_url_format(url: str) -> str:
     Returns:
         str: 转换后的xhsdiscover协议格式URL
     """
+    start_time = time.time()
+    logger.info(f"[convert_url_format] 开始转换URL: {url}")
     parsed_data = XHSNoteExtractor.parse_xhs_url(url)
-    return XHSNoteExtractor.convert_to_xhsdiscover_format(
+    result = XHSNoteExtractor.convert_to_xhsdiscover_format(
         parsed_data["note_id"], 
         parsed_data["xsec_token"]
-    )
+    )
+    elapsed_time = time.time() - start_time
+    logger.info(f"[convert_url_format] 耗时: {elapsed_time:.3f}秒，结果: {result}")
+    return result

{xhs_note_extractor-0.1.dev2.dist-info → xhs_note_extractor-0.1.dev6.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: xhs-note-extractor
-Version: 0.1.dev2
+Version: 0.1.dev6
 Summary: A Python package for extracting Xiaohongshu (Little Red Book) note data from URLs
 Author-email: JoyCode Agent <agent@joycode.com>
 License: MIT

{xhs_note_extractor-0.1.dev2.dist-info → xhs_note_extractor-0.1.dev6.dist-info}/RECORD

@@ -1,11 +1,11 @@
 xhs_note_extractor/__init__.py,sha256=CjHdqO4W5sj6zbeE7xYkR0_WRfc99G56nR6k2Kmji44,1207
 xhs_note_extractor/_version.py,sha256=59jjKBtTUi_9u6FVZcIpQEDYjyAaqdxzqXyuRuFYKPE,720
 xhs_note_extractor/cli.py,sha256=F5phl4HqnzEe_vTS8vpio_KcZNx4cxmXJnYcQ1FgMbA,2693
-xhs_note_extractor/extractor.py,sha256=Afl-VzMmuRSk82BbAROpIAB6g1BSpC3yRbDJTrO0NCM,14964
+xhs_note_extractor/extractor.py,sha256=NwYWtgU8lX53OVmfHzpLeSWEpZv0mH3UcLcw4K-1Oj4,21136
 xhs_note_extractor/utils.py,sha256=mOVoLknlflzv7aCjXdmeNniQ7P6WNUcjSKjCm8uwFNk,14364
-xhs_note_extractor-0.1.dev2.dist-info/licenses/LICENSE,sha256=VFtWajKKKkgOoX3cMb2upEjsQmDpU85ymhom2bYY_oI,1069
-xhs_note_extractor-0.1.dev2.dist-info/METADATA,sha256=ohcbdSTJ-ms66W5xBEqXIhczNN_7lpk4mqsvpL1fF0g,5525
-xhs_note_extractor-0.1.dev2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-xhs_note_extractor-0.1.dev2.dist-info/entry_points.txt,sha256=1IG34snKfX2pzpLSeDXHqlSVSH8p7bf3eaKQfcwGDk4,60
-xhs_note_extractor-0.1.dev2.dist-info/top_level.txt,sha256=at3SqTdQr3DWMFCL5KM0Ofo_LE88WqADjh8MeFLwwO0,19
-xhs_note_extractor-0.1.dev2.dist-info/RECORD,,
+xhs_note_extractor-0.1.dev6.dist-info/licenses/LICENSE,sha256=VFtWajKKKkgOoX3cMb2upEjsQmDpU85ymhom2bYY_oI,1069
+xhs_note_extractor-0.1.dev6.dist-info/METADATA,sha256=6XtFTabTgXrg7np6dVLtWAn0u5C_rmC10eJ9WYfukyA,5525
+xhs_note_extractor-0.1.dev6.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+xhs_note_extractor-0.1.dev6.dist-info/entry_points.txt,sha256=1IG34snKfX2pzpLSeDXHqlSVSH8p7bf3eaKQfcwGDk4,60
+xhs_note_extractor-0.1.dev6.dist-info/top_level.txt,sha256=at3SqTdQr3DWMFCL5KM0Ofo_LE88WqADjh8MeFLwwO0,19
+xhs_note_extractor-0.1.dev6.dist-info/RECORD,,