crystalwindow 4.5__py3-none-any.whl → 4.7__py3-none-any.whl

crystalwindow/websearch.py

@@ -1,5 +1,6 @@
 """
-CrystalWindow WebSearch module — Google (Serper.dev) Edition
+WebBrowse — Universal Web Search Module
+A modern, clean search helper for any Python application.
 """
 
 from __future__ import annotations
@@ -8,94 +9,101 @@ import webbrowser
 from dataclasses import dataclass
 from typing import List, Tuple, Optional, Any
 
-# Optional Qt integration
-try:
-    from PySide6.QtWidgets import (
-        QWidget, QVBoxLayout, QLineEdit, QPushButton, QTextEdit,
-        QListWidget, QListWidgetItem, QLabel, QHBoxLayout
-    )
-    from PySide6.QtCore import Qt, QUrl
-    from PySide6.QtGui import QDesktopServices
-    PYSIDE6_AVAILABLE = True
-except Exception:
-    PYSIDE6_AVAILABLE = False
 
-
-# --------------------------
-# Data result structure
-# --------------------------
+# ================================
+# Dataclass: Unified Search Result
+# ================================
 
 @dataclass
-class WebSearchResult:
-    text: str
-    links: List[Tuple[str, str]]  # (title, url)
-    raw: Any = None
-
-
-# --------------------------
-# Main WebSearch class
-# --------------------------
-
-class WebSearch:
-    """ Real Web Search helper for CrystalWindow using Google (Serper.dev). """
-
-    # --- DuckDuckGo Instant Answer (kept for fallback) ---
-    DUCKDUCKGO_INSTANT = "https://api.duckduckgo.com/"
-
-    # --- Google Search (Serper.dev) ---
+class SearchResult:
+    summary: str
+    links: List[Tuple[str, str]]     # (title, url)
+    raw: Any = None                  # raw API payload
+
+
+# ================================
+# Main WebBrowse Search Handler
+# ================================
+
+class WebBrowse:
+    """
+    Lightweight universal web search interface.
+    Supports:
+      - Google Search via Serper.dev (primary)
+      - DuckDuckGo Instant Answer (fallback)
+
+    Designed to be extendable and safe.
+    """
+
+    # --------------------------
+    # API Endpoints
+    # --------------------------
+    DUCK_URL = "https://api.duckduckgo.com/"
     SERPER_URL = "https://google.serper.dev/search"
-    SERPER_API_KEY = "a8e02281a0c3b58bb29a3731b6a2aec1d8d8a487"  # <<<<<<<< REPLACE THIS
 
-    # --------------------------------------------------
-    # DuckDuckGo Instant Answer (old API)
-    # --------------------------------------------------
+    # --------------------------
+    # Set your Serper.dev key
+    # --------------------------
+    SERPER_API_KEY: str = ""     # <-- user must set manually
+
+    # =====================================================
+    # DuckDuckGo Instant Answer (fallback, small summaries)
+    # =====================================================
     @classmethod
-    def _duckduckgo_instant(cls, query: str) -> WebSearchResult:
+    def _search_duckduckgo(cls, query: str) -> SearchResult:
         params = {
             "q": query,
             "format": "json",
             "no_html": 1,
-            "no_redirect": 1,
+            "no_redirect": 1
         }
+
         try:
-            r = requests.get(cls.DUCKDUCKGO_INSTANT, params=params, timeout=8)
+            r = requests.get(cls.DUCK_URL, params=params, timeout=8)
             r.raise_for_status()
         except Exception as e:
-            return WebSearchResult(text=f"DuckDuckGo error: {e}", links=[])
+            return SearchResult(f"DuckDuckGo error: {e}", [], None)
 
         data = r.json()
-        abstract = data.get("AbstractText") or data.get("Abstract") or ""
+        summary = data.get("AbstractText") or data.get("Abstract") or ""
         links: List[Tuple[str, str]] = []
 
+        # Abstract main URL
         if data.get("AbstractURL"):
             links.append((data.get("Heading", "Result"), data["AbstractURL"]))
 
-        def extract_related(rt):
-            if isinstance(rt, dict):
-                if rt.get("FirstURL") and rt.get("Text"):
-                    links.append((rt["Text"], rt["FirstURL"]))
-                for key in ("Topics", "RelatedTopics"):
-                    if isinstance(rt.get(key), list):
-                        for sub in rt[key]:
-                            extract_related(sub)
-            elif isinstance(rt, list):
-                for item in rt:
-                    extract_related(item)
-
-        extract_related(data.get("RelatedTopics", []))
-        if not abstract and links:
-            abstract = "\n".join([t for t, _ in links[:3]])
-
-        return WebSearchResult(text=abstract.strip(), links=links, raw=data)
-
-    # --------------------------------------------------
-    # Google Search via Serper.dev (NEW)
-    # --------------------------------------------------
+        # Recursive topic extractor
+        def extract_topics(node):
+            if isinstance(node, dict):
+                if node.get("FirstURL") and node.get("Text"):
+                    links.append((node["Text"], node["FirstURL"]))
+                for k in ("Topics", "RelatedTopics"):
+                    if isinstance(node.get(k), list):
+                        for sub in node[k]:
+                            extract_topics(sub)
+            elif isinstance(node, list):
+                for sub in node:
+                    extract_topics(sub)
+
+        extract_topics(data.get("RelatedTopics", []))
+
+        # If summary is empty, generate from first few topics
+        if not summary and links:
+            summary = "\n".join(t for t, _ in links[:3])
+
+        return SearchResult(summary.strip(), links, data)
+
+    # =====================================================
+    # Google Search via Serper.dev
+    # =====================================================
     @classmethod
-    def _google_serper(cls, query: str) -> WebSearchResult:
+    def _search_google(cls, query: str) -> SearchResult:
+        if not cls.SERPER_API_KEY:
+            return SearchResult("Google engine requires SERPER_API_KEY.", [], None)
+
         headers = {
             "X-API-KEY": cls.SERPER_API_KEY,
-            "Content-Type": "application/json"
+            "Content-Type": "application/json",
         }
         body = {"q": query}
 
@@ -103,12 +111,11 @@ class WebSearch:
             r = requests.post(cls.SERPER_URL, headers=headers, json=body, timeout=8)
             r.raise_for_status()
         except Exception as e:
-            return WebSearchResult(text=f"Google search error: {e}", links=[])
+            return SearchResult(f"Google search error: {e}", [], None)
 
         data = r.json()
-
         links = []
-        summary_parts = []
+        summaries = []
 
         for item in data.get("organic", []):
             title = item.get("title", "Untitled")
@@ -117,117 +124,36 @@ class WebSearch:
 
             links.append((title, url))
             if snippet:
-                summary_parts.append(snippet)
+                summaries.append(snippet)
 
-        summary = "\n".join(summary_parts[:3]).strip()
-        if not summary:
-            summary = "(no summary)"
+        summary = "\n".join(summaries[:3]).strip() or "(no summary)"
 
-        return WebSearchResult(summary, links, data)
+        return SearchResult(summary, links, data)
 
-    # --------------------------------------------------
-    # Main public search()
-    # --------------------------------------------------
+    # =====================================================
+    # Public search() method
+    # =====================================================
     @classmethod
-    def search(cls, query: str, engine: str = "google") -> WebSearchResult:
-        if not query or not query.strip():
-            return WebSearchResult(text="", links=[])
+    def search(cls, query: str, engine: str = "google") -> SearchResult:
+        if not query.strip():
+            return SearchResult("", [], None)
 
-        q = query.strip().lower()
-        engine = engine.lower()
+        engine = engine.lower().strip()
 
         if engine == "google":
-            return cls._google_serper(q)
+            return cls._search_google(query)
 
         if engine == "duckduckgo":
-            return cls._duckduckgo_instant(q)
+            return cls._search_duckduckgo(query)
 
-        return WebSearchResult(text=f"Unknown engine '{engine}'", links=[])
+        return SearchResult(f"Unknown search engine '{engine}'", [], None)
 
-    # --------------------------------------------------
-    # URL opener
-    # --------------------------------------------------
+    # =====================================================
+    # URL Opener
+    # =====================================================
     @staticmethod
     def open_url(url: str) -> None:
         try:
             webbrowser.open(url)
         except Exception:
             pass
-
-    # --------------------------------------------------
-    # Optional PySide6 GUI
-    # --------------------------------------------------
-    if PYSIDE6_AVAILABLE:
-        class _CWWebTab(QWidget):
-            def __init__(self, parent=None):
-                super().__init__(parent)
-                self.setWindowTitle("WebSearch")
-
-                layout = QVBoxLayout(self)
-
-                # Search bar
-                bar = QHBoxLayout()
-                self.qbox = QLineEdit()
-                self.qbox.setPlaceholderText("Search the web...")
-                btn = QPushButton("Search")
-                btn.clicked.connect(self._do_search)
-                bar.addWidget(self.qbox)
-                bar.addWidget(btn)
-                layout.addLayout(bar)
-
-                # Summary text
-                self.summary = QTextEdit()
-                self.summary.setReadOnly(True)
-                self.summary.setFixedHeight(150)
-                layout.addWidget(self.summary)
-
-                # Links list
-                self.links = QListWidget()
-                self.links.itemActivated.connect(self._open_item)
-                layout.addWidget(self.links)
-
-                self.note = QLabel("Powered by Google (Serper.dev) API")
-                self.note.setAlignment(Qt.AlignCenter)
-                layout.addWidget(self.note)
-
-            def _populate(self, result: WebSearchResult):
-                self.summary.setPlainText(result.text)
-                self.links.clear()
-                for t, u in result.links:
-                    item = QListWidgetItem(f"{t}\n{u}")
-                    item.setData(Qt.UserRole, u)
-                    self.links.addItem(item)
-
-            def _open_item(self, item: QListWidgetItem):
-                url = item.data(Qt.UserRole)
-                if url:
-                    QDesktopServices.openUrl(QUrl(url))
-
-            def _do_search(self):
-                q = self.qbox.text().strip()
-                if not q:
-                    self.summary.setPlainText("Type something to search.")
-                    return
-                self.summary.setPlainText("Searching...")
-                res = WebSearch.search(q, engine="google")
-                self._populate(res)
-
-        @classmethod
-        def register_cw_tab(cls, window):
-            try:
-                tab = cls._CWWebTab()
-                if hasattr(window, "add_tab"):
-                    window.add_tab("WebSearch", tab)
-                elif hasattr(window, "addTab"):
-                    window.addTab(tab, "WebSearch")
-                else:
-                    if hasattr(window, "setCentralWidget"):
-                        window.setCentralWidget(tab)
-                return tab
-            except Exception:
-                return None
-
-    else:
-        @classmethod
-        def register_cw_tab(cls, window):
-            return None