Update utils.py

utils.py

@@ -1,69 +1,15 @@
-# utils.py
-
-import os
-import fitz  # PyMuPDF
-import faiss
-import numpy as np
-from openai import OpenAI
-import hashlib
-
-# تنظیمات API
-openai_api_key = os.getenv("API_Key")
-
-# Embedder
-def get_embedding(text, model="text-embedding-ada-002"):
-    from openai import OpenAI
-    client = OpenAI(api_key=openai_api_key)
-    response = client.embeddings.create(input=[text], model=model)
-    return response.data[0].embedding
-
-# استخراج متن از PDF و بردارسازی
-def extract_text_and_vectors(files):
-    documents = []
-    for file in files:
-        doc = fitz.open(file.name)
-        for page_num, page in enumerate(doc):
-            text = page.get_text()
-            if text.strip():
-                vector = get_embedding(text)
-                documents.append({
-                    "file_name": file.name,
-                    "page_num": page_num + 1,
-                    "text": text,
-                    "vector": np.array(vector).astype("float32"),
-                })
-    return documents
-
-# ساخت ایندکس FAISS
-def build_faiss_index(documents, dim=1536):
-    index = faiss.IndexFlatL2(dim)
-    vectors = [doc["vector"] for doc in documents]
-    index.add(np.array(vectors))
-    return index
-
-# جستجو در FAISS
-def search_similar_content(query, documents, index, k=3):
-    query_vector = np.array(get_embedding(query)).astype("float32").reshape(1, -1)
-    D, I = index.search(query_vector, k)
-    results = [documents[i] for i in I[0]]
-    return results
-
-# فرمت‌دهی پاسخ
-def format_response(results):
-    formatted = []
-    for r in results:
-        snippet = r["text"][:500].strip().replace('\n', ' ')
-        formatted.append(f"""📄 **{r['file_name']}** | صفحه {r['page_num']}\n{text_shorten(snippet)}\n""")
-    return "\n---\n".join(formatted)
-
-# کمک‌تابع برای خلاصه کردن متن
-def text_shorten(text, max_chars=300):
-    return text if len(text) <= max_chars else text[:max_chars] + "..."
-
-# لاگ نمونه‌ای از اسناد پردازش‌شده
-def log_debug_info(documents, max_samples=2):
-    info = f"📦 مجموع اسناد پردازش‌شده: {len(documents)}\n\n"
-    for i, doc in enumerate(documents[:max_samples]):
-        info += f"📝 فایل: {doc['file_name']} | صفحه: {doc['page_num']}\n"
-        info += f"متن نمونه: {text_shorten(doc['text'])}\n\n"
-    return info
+import json
+
+def load_material_db(path="material_db.json"):
+    with open(path, "r", encoding="utf-8") as f:
+        return json.load(f)
+
+def filter_items(materials, pole_height, pole_power, conductor_size):
+    matched = []
+    for item in materials:
+        cond = item["conditions"]
+        if (cond["pole_height"] in ["-", pole_height]) and \
+           (cond["pole_power"] in ["-", pole_power]) and \
+           (cond["conductor_size"] in ["-", conductor_size]):
+            matched.append(item)
+    return matched