Update tools/quran_search.py

tools/quran_search.py

@@ -9,7 +9,7 @@ import time
 
 class QuranSearchEngine:
     def __init__(self):
-        self.full_quran_url = "https://cdn.jsdelivr.net/npm/[email protected]/dist/quran.json"
+        self.api_url = "https://quranapi.pages.dev/api/"
         self.logger = logging.getLogger(__name__)
         self.full_quran = None
         self.surahs = None
@@ -25,17 +25,38 @@ class QuranSearchEngine:
         max_retries = 3
         for attempt in range(max_retries):
             try:
-                response = requests.get(self.full_quran_url, timeout=10)
+                response = requests.get(f"{self.api_url}surah.json", timeout=10)
                 response.raise_for_status()
-                self.full_quran = response.json()  # Array of surah dicts
-                self.surahs = self.full_quran
+                self.surahs = response.json()  # Array of surah metadata
+                # Fetch full verses by iterating over surah IDs
+                self.full_quran = []
+                for surah_id in range(1, 115):
+                    surah_response = requests.get(f"{self.api_url}{surah_id}.json", timeout=10)
+                    surah_response.raise_for_status()
+                    self.full_quran.append(surah_response.json())
                 break
             except Exception as e:
-                self.logger.error(f"Attempt {attempt + 1}/{max_retries} failed to fetch full Quran: {e}")
+                self.logger.error(f"Attempt {attempt + 1}/{max_retries} failed to fetch full Quran from API: {e}")
                 if attempt == max_retries - 1:
-                    self.surahs = self._load_fallback_surahs()
+                    self._load_fallback_quran()
                 time.sleep(2 ** attempt)
     
+    def _load_fallback_quran(self):
+        max_retries = 3
+        for source in QURAN_DATA_SOURCES:
+            for attempt in range(max_retries):
+                try:
+                    response = requests.get(source, timeout=10)
+                    response.raise_for_status()
+                    self.full_quran = response.json()  # Array of surah dicts
+                    self.surahs = self.full_quran
+                    break
+                except Exception as e:
+                    self.logger.error(f"Attempt {attempt + 1}/{max_retries} failed to fetch fallback from {source}: {e}")
+                    if attempt == max_retries - 1 and source == QURAN_DATA_SOURCES[-1]:
+                        self.surahs = self._load_fallback_surahs()
+                    time.sleep(2 ** attempt)
+    
     def _load_all_verses_and_embeddings(self):
         if not self.full_quran:
             self.logger.error("No full Quran loaded, skipping verse loading")
@@ -45,12 +66,15 @@ class QuranSearchEngine:
             return
         
         for surah in self.full_quran:
-            surah_id = surah['id']
-            for verse in surah['verses']:
+            surah_id = surah.get('id', surah.get('surah_id', 1))  # Handle API vs fallback structure
+            verses = surah.get('verses', surah.get('arabic1', []))  # API uses 'arabic1', fallback uses 'verses'
+            for verse in verses:
+                verse_num = verse.get('id', verse.get('verse_id', 1))
+                text = verse.get('text', verse)  # Handle different key structures
                 self.all_verses.append({
                     'surah_id': surah_id,
-                    'verse_num': verse['id'],
-                    'text': verse['text']
+                    'verse_num': verse_num,
+                    'text': text
                 })
         
         try:
@@ -77,17 +101,17 @@ class QuranSearchEngine:
     def get_surah_text(self, surah_id):
         if self.full_quran:
             try:
-                surah = self.full_quran[surah_id - 1]
-                verses = surah['verses']
-                return "\n\n".join(f"آية {v['id']}: {v['text']}" for v in verses)
-            except IndexError:
-                self.logger.error(f"Surah {surah_id} not found in cached data")
+                surah = next((s for s in self.full_quran if s.get('id', s.get('surah_id', 1)) == surah_id), None)
+                if surah:
+                    verses = surah.get('verses', surah.get('arabic1', []))
+                    return "\n\n".join(f"آية {v.get('id', i + 1)}: {v.get('text', v)}" for i, v in enumerate(verses))
+            except Exception as e:
+                self.logger.error(f"Error processing cached surah {surah_id}: {e}")
         
-        # Fallback if cache failed
         max_retries = 3
         for attempt in range(max_retries):
             try:
-                response = requests.get(f"https://quranapi.pages.dev/api/{surah_id}.json", timeout=10)
+                response = requests.get(f"{self.api_url}{surah_id}.json", timeout=10)
                 response.raise_for_status()
                 data = response.json()
                 verses = data['arabic1']
@@ -110,7 +134,7 @@ class QuranSearchEngine:
             results = []
             for idx in top_indices:
                 verse = self.all_verses[idx]
-                surah_name = self.surahs[verse['surah_id'] - 1]['name']
+                surah_name = next((s['name'] for s in self.surahs if s.get('id', s.get('surah_id', 1)) == verse['surah_id']), f"سورة {verse['surah_id']}")
                 results.append(
                     f"سورة {surah_name} - آية {verse['verse_num']}:\n{verse['text']}"
                 )
@@ -124,21 +148,20 @@ class QuranSearchEngine:
         matches = []
         for verse in self.all_verses:
             if query_lower in verse['text'].lower():
-                surah_name = self.surahs[verse['surah_id'] - 1]['name'] if self.surahs else f"سورة {verse['surah_id']}"
+                surah_name = next((s['name'] for s in self.surahs if s.get('id', s.get('surah_id', 1)) == verse['surah_id']), f"سورة {verse['surah_id']}")
                 matches.append(f"سورة {surah_name} - آية {verse['verse_num']}:\n{verse['text']}")
         return "\n\n".join(matches[:top_k]) or "لا توجد نتائج مطابقة."
 
-    # Fallback methods (unchanged)
     def _load_fallback_surahs(self):
         try:
             for source in QURAN_DATA_SOURCES:
                 try:
-                    df = pd.read_csv(source)
-                    return [
-                        (row['name_arabic'], row['surah_id'])
-                        for _, row in df.drop_duplicates(subset=['surah_id']).iterrows()
-                    ]
-                except:
+                    response = requests.get(source, timeout=10)
+                    response.raise_for_status()
+                    data = response.json()
+                    return [(s['name'], s['id']) for s in data]
+                except Exception as e:
+                    self.logger.error(f"Failed to load fallback surahs from {source}: {e}")
                     continue
             return [
                 ("الفاتحة", 1),