entari-plugin-hyw 4.0.0rc5__py3-none-any.whl

entari_plugin_hyw/browser/engines/searxng.py

@@ -0,0 +1,137 @@
+
+import urllib.parse
+import re
+from typing import List, Dict, Any
+from loguru import logger
+from .base import SearchEngine
+
+class SearXNGEngine(SearchEngine):
+    """
+    Parser for DuckDuckGo and SearXNG results.
+    Handles both Markdown (from Crawl4AI) and HTML (fallback).
+    """
+    
+    def build_url(self, query: str, limit: int = 10) -> str:
+        encoded_query = urllib.parse.quote(query)
+        # Default fallback if not configurable per instance, but usually this is what we support as "searxng"
+        base = "https://lite.duckduckgo.com/lite/"
+        return f"{base}?q={encoded_query}"
+
+    def parse(self, content: str) -> List[Dict[str, Any]]:
+        # Prioritize HTML parsing if content looks like HTML
+        if "<html" in content.lower() or "<!doctype" in content.lower() or "<div" in content.lower():
+            results = self._parse_html(content)
+            if results:
+                return results
+
+        # Fallback to Markdown
+        return self._parse_markdown(content)
+
+    def _parse_html(self, content: str) -> List[Dict[str, Any]]:
+        results = []
+        seen_urls = set()
+        
+        # Simple regex for DDG Lite / SearXNG HTML structure
+        link_regex = re.compile(r'<a[^>]+href=["\'](http[^"\']+)["\'][^>]*>(.*?)</a>', re.IGNORECASE | re.DOTALL)
+        
+        pos = 0
+        while True:
+            match = link_regex.search(content, pos)
+            if not match:
+                break
+            
+            href = match.group(1)
+            title_html = match.group(2)
+            
+            # Clean title
+            title = re.sub(r'<[^>]+>', '', title_html).strip()
+            
+            pos = match.end()
+            
+            # Filter junk
+            if "search" in href and "q=" in href: continue 
+            if "google.com" in href or "bing.com" in href: continue
+            if href in seen_urls: continue
+            
+            # Look ahead for snippet
+            snippet_chunk = content[pos:pos+1000]
+            snippet_match = re.search(r'(.*?)<a', snippet_chunk, re.DOTALL | re.IGNORECASE)
+            raw_snippet = snippet_match.group(1) if snippet_match else snippet_chunk
+            
+            # Clean HTML tags from snippet
+            snippet = re.sub(r'<[^>]+>', ' ', raw_snippet)
+            snippet = re.sub(r'\s+', ' ', snippet).strip()
+            
+            # No truncation as per user request (or very generous limit)
+            snippet = snippet[:5000]
+            
+            # Valid result check
+            if title and len(title) > 2 and snippet:
+                # Extract images from the result block (rough heuristic)
+                images = []
+                img_matches = re.findall(r'<img[^>]+src=["\'](http[^"\']+)["\']', snippet_match.group(0) if snippet_match else snippet_chunk)
+                for img_url in img_matches:
+                    if not any(x in img_url for x in ['favicon', 'icon', 'tracking', 'pixel']):
+                         images.append(img_url)
+                
+                results.append({
+                    "title": title,
+                    "url": href,
+                    "domain": urllib.parse.urlparse(href).hostname or "",
+                    "content": snippet,
+                    "images": images[:3] # Limit per result
+                })
+                seen_urls.add(href)
+                
+        logger.info(f"SearXNG Parser(HTML) found {len(results)} results.")
+        return results
+
+    def _parse_markdown(self, content: str) -> List[Dict[str, Any]]:
+        results = []
+        seen_urls = set()
+        
+        # Link regex: [Title](URL)
+        link_regex = re.compile(r'\[(.*?)\]\((https?://.*?)\)')
+        
+        lines = content.split('\n')
+        current_result = None
+        
+        for line in lines:
+            line = line.strip()
+            if not line: continue
+            
+            # Check for link
+            match = link_regex.search(line)
+            if match:
+                # Save previous result
+                if current_result:
+                    results.append(current_result)
+                
+                title, href = match.groups()
+                
+                # Filter junk
+                if "search" in href and "q=" in href: continue 
+                if "google.com" in href or "bing.com" in href: continue 
+                if href in seen_urls: 
+                    current_result = None
+                    continue
+                    
+                seen_urls.add(href)
+                
+                current_result = {
+                    "title": title,
+                    "url": href,
+                    "domain": urllib.parse.urlparse(href).hostname or "",
+                    "content": "" 
+                }
+            elif current_result:
+                # Append snippet
+                if len(current_result["content"]) < 5000:
+                    current_result["content"] += " " + line
+        
+        # Append last
+        if current_result:
+             results.append(current_result)
+        
+        logger.info(f"SearXNG Parser(Markdown) found {len(results)} results.")
+        return results

entari_plugin_hyw/browser/landing.html

@@ -0,0 +1,172 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+
+<head>
+    <meta charset="UTF-8">
+    <title>entari-plugin-hyw Browser</title>
+    <style>
+        :root {
+            --theme-color: #ef4444;
+            --text-primary: #2c2c2e;
+            --text-body: #3a3a3c;
+            --text-muted: #86868b;
+            --bg-border: #f2f2f2;
+        }
+
+        body {
+            background-color: var(--bg-border);
+            color: var(--text-primary);
+            font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, sans-serif;
+            margin: 0;
+            padding: 0;
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            height: 100vh;
+            overflow: hidden;
+        }
+
+        #main-container {
+            width: 560px;
+            background: white;
+            padding: 40px;
+            box-shadow: 0 4px 6px -1px rgba(0, 0, 0, 0.1), 0 2px 4px -1px rgba(0, 0, 0, 0.06);
+            position: relative;
+            transform: scale(1.1);
+            /* Slightly larger for visibility */
+        }
+
+        .corner-badge {
+            position: absolute;
+            top: -10px;
+            left: -10px;
+            height: 28px;
+            padding: 0 12px;
+            background-color: var(--theme-color);
+            color: white;
+            display: flex;
+            align-items: center;
+            font-size: 12px;
+            font-weight: 800;
+            text-transform: uppercase;
+            letter-spacing: 0.5px;
+            box-shadow: 0 2px 4px 0 rgba(0, 0, 0, 0.15);
+            z-index: 10;
+        }
+
+        h1 {
+            font-size: 32px;
+            font-weight: 900;
+            text-transform: uppercase;
+            letter-spacing: -1px;
+            margin: 0 0 24px 0;
+            line-height: 1.1;
+        }
+
+        .brand-text {
+            color: var(--theme-color);
+        }
+
+        .section-title {
+            font-size: 17px;
+            font-bold: 700;
+            text-transform: uppercase;
+            letter-spacing: -0.5px;
+            margin-bottom: 12px;
+            color: var(--text-primary);
+        }
+
+        .content-box {
+            padding-top: 24px;
+            border-top: 1px solid #eee;
+            margin-top: 32px;
+        }
+
+        .summary {
+            font-size: 15px;
+            line-height: 1.6;
+            color: var(--text-body);
+            text-align: justify;
+        }
+
+        .summary-en {
+            margin-top: 16px;
+            color: var(--text-muted);
+            font-size: 13px;
+            font-style: italic;
+            line-height: 1.5;
+        }
+
+        b {
+            color: var(--theme-color);
+            font-weight: 700;
+        }
+
+        /* Subtle loading animation to show it's "live" */
+        .progress-bar {
+            height: 3px;
+            background: #f3f3f3;
+            width: 100%;
+            margin-top: 32px;
+            position: relative;
+            overflow: hidden;
+        }
+
+        .progress-fill {
+            position: absolute;
+            height: 100%;
+            background: var(--theme-color);
+            width: 30%;
+            left: -30%;
+            animation: moveProgress 2s infinite ease-in-out;
+        }
+
+        @keyframes moveProgress {
+            0% {
+                left: -30%;
+                width: 20%;
+            }
+
+            50% {
+                width: 50%;
+            }
+
+            100% {
+                left: 100%;
+                width: 20%;
+            }
+        }
+    </style>
+</head>
+
+<body>
+    <div id="main-container">
+        <!-- Corner Badge Style -->
+        <div class="corner-badge">
+            Status: Ready
+        </div>
+
+        <h1>entari-plugin-<span class="brand-text">hyw</span></h1>
+
+        <div class="content-box">
+            <div class="section-title">Browser Service</div>
+            <div class="summary">
+                这是一个受 <b>entari-plugin-hyw</b> 插件控制的自动化浏览器实例。<br>
+                它负责网络搜索、内容爬取以及卡片 UI 的实时渲染。<br>
+                请勿关闭此窗口，以确保插件功能正常运行。
+            </div>
+
+            <div class="summary-en">
+                This is an automated browser instance controlled by the <b>entari-plugin-hyw</b> plugin.
+                It handles web searches, content crawling, and real-time Card UI rendering.
+                Please do not close this window to ensure the plugin functions correctly.
+            </div>
+        </div>
+
+        <div class="progress-bar">
+            <div class="progress-fill"></div>
+        </div>
+    </div>
+</body>
+
+</html>

entari_plugin_hyw/browser/manager.py

@@ -0,0 +1,153 @@
+"""
+Shared Browser Manager (DrissionPage)
+
+Manages a single ChromiumPage browser instance.
+Supports multi-tab concurrency via DrissionPage's built-in tab management.
+"""
+
+import threading
+from typing import Optional, Any
+from loguru import logger
+from DrissionPage import ChromiumPage, ChromiumOptions
+from DrissionPage.errors import PageDisconnectedError
+
+class SharedBrowserManager:
+    """
+    Manages a shared DrissionPage Chromium browser.
+    """
+    
+    _instance: Optional["SharedBrowserManager"] = None
+    _lock = threading.Lock()
+    
+    def __init__(self, headless: bool = True):
+        self.headless = headless
+        self._page: Optional[ChromiumPage] = None
+        self._starting = False
+        self._tab_lock = threading.Lock()
+    
+    @classmethod
+    def get_instance(cls, headless: bool = True) -> "SharedBrowserManager":
+        """Get or create the singleton SharedBrowserManager."""
+        with cls._lock:
+            if cls._instance is None:
+                cls._instance = cls(headless=headless)
+            return cls._instance
+    
+    def start(self) -> bool:
+        """Start the Chromium browser."""
+        if self._page is not None:
+            # Check if alive
+            try:
+                if self._page.run_cdp('Browser.getVersion'):
+                    return True
+            except (PageDisconnectedError, Exception):
+                self._page = None
+        
+        with self._lock:
+            if self._starting:
+                return False
+            self._starting = True
+        
+        try:
+            logger.info("SharedBrowserManager: Starting DrissionPage browser...")
+            
+            # Configure options
+            co = ChromiumOptions()
+            co.headless(self.headless)
+            co.auto_port()  # Auto find available port
+            
+            # Anti-detection settings 
+            co.set_argument('--no-sandbox')
+            co.set_argument('--disable-gpu')
+            # Essential for loading local files and avoiding CORS issues
+            co.set_argument('--allow-file-access-from-files')
+            co.set_argument('--disable-web-security')
+            # Hide scrollbars globally
+            co.set_argument('--hide-scrollbars')
+            # 十万的原因是滚动条屏蔽(大概吧)
+            co.set_argument('--window-size=1280,20000')
+            self._page = ChromiumPage(addr_or_opts=co)
+            
+            # Show Landing Page
+            try:
+                import os
+                landing_path = os.path.join(os.path.dirname(__file__), 'landing.html')
+                if os.path.exists(landing_path):
+                    self._page.get(f"file://{landing_path}")
+            except Exception as e:
+                logger.warning(f"SharedBrowserManager: Failed to show landing page: {e}")
+
+            logger.success(f"SharedBrowserManager: Browser ready (port={self._page.address})")
+            return True
+            
+        except Exception as e:
+            logger.error(f"SharedBrowserManager: Failed to start browser: {e}")
+            self._page = None
+            raise
+        finally:
+            self._starting = False
+    
+    @property
+    def page(self) -> Optional[ChromiumPage]:
+        """Get the main ChromiumPage instance."""
+        if self._page is None:
+            self.start()
+        return self._page
+
+    def new_tab(self, url: str) -> Any:
+        """Thread-safe tab creation."""
+        with self._tab_lock:
+            page = self.page
+            if not page:
+                raise RuntimeError("Browser not available")
+            return page.new_tab(url)
+    
+    def close(self):
+        """Shutdown the browser."""
+        with self._lock:
+            if self._page:
+                try:
+                    self._page.quit()
+                    logger.info("SharedBrowserManager: Browser closed.")
+                except Exception as e:
+                    logger.warning(f"SharedBrowserManager: Error closing browser: {e}")
+                finally:
+                    self._page = None
+
+    @staticmethod
+    def hide_scrollbars(page: ChromiumPage):
+        """
+        Robustly hide scrollbars using CDP commands.
+        This eliminates the reserved space/gutter that standard CSS might miss.
+        """
+        try:
+            # Emulation.setScrollbarsHidden is a CDP command
+            page.run_cdp('Emulation.setScrollbarsHidden', hidden=True)
+            logger.debug("SharedBrowserManager: CDP scrollbars hidden.")
+        except Exception as e:
+            logger.warning(f"SharedBrowserManager: Failed to hide scrollbars via CDP: {e}")
+
+
+# Module-level singleton accessor
+_shared_manager: Optional[SharedBrowserManager] = None
+
+
+def get_shared_browser_manager(headless: bool = True) -> SharedBrowserManager:
+    """
+    Get or create the shared browser manager.
+    """
+    global _shared_manager
+    
+    if _shared_manager is None:
+        _shared_manager = SharedBrowserManager.get_instance(headless=headless)
+        _shared_manager.start()
+    
+    return _shared_manager
+
+
+def close_shared_browser():
+    """Close the shared browser manager."""
+    global _shared_manager
+    if _shared_manager:
+        _shared_manager.close()
+        _shared_manager = None