Create build.py

build.py

@@ -0,0 +1,63 @@
+#!/usr/bin/env python3
+# build.py — Build FAISS index (CPU)
+
+import json
+import pickle
+from pathlib import Path
+
+import faiss
+import numpy as np
+from sentence_transformers import SentenceTransformer
+
+# ---------- CONFIG ----------
+ROOT = Path(__file__).parent
+FLATTENED_JSONL = ROOT / "data" / "processed" / "flattened_docs.jsonl"
+INDEX_FILE = ROOT / "law_index.faiss"
+META_FILE = ROOT / "law_meta.pkl"
+EMB_MODEL = "paraphrase-multilingual-MiniLM-L12-v2"
+# ----------------------------
+
+def load_docs():
+    if not FLATTENED_JSONL.exists():
+        raise FileNotFoundError(f"{FLATTENED_JSONL} not found")
+
+    docs = []
+    with FLATTENED_JSONL.open("r", encoding="utf-8") as f:
+        for line in f:
+            if not line.strip():
+                continue
+            docs.append(json.loads(line))
+
+    print(f"[i] Loaded {len(docs)} documents")
+    return docs
+
+def build():
+    docs = load_docs()
+
+    texts = [d.get("text", "") for d in docs]
+
+    print("[i] Loading embedder...")
+    embedder = SentenceTransformer(EMB_MODEL)
+
+    print("[i] Computing embeddings...")
+    embs = embedder.encode(
+        texts,
+        convert_to_numpy=True,
+        show_progress_bar=True
+    ).astype(np.float32)
+
+    faiss.normalize_L2(embs)
+
+    dim = embs.shape[1]
+    index = faiss.IndexFlatIP(dim)
+    index.add(embs)
+
+    faiss.write_index(index, str(INDEX_FILE))
+    with open(META_FILE, "wb") as f:
+        pickle.dump(docs, f)
+
+    print("[i] Saved index →", INDEX_FILE)
+    print("[i] Saved metadata →", META_FILE)
+
+if __name__ == "__main__":
+    build()