Delete nexframe_brain.py

nexframe_brain.py

@@ -1,57 +0,0 @@
-# =========================================================
-#  NexFrame Brain — simplified version for Hugging Face
-# =========================================================
-import os, re, json, hashlib
-from pathlib import Path
-from datetime import datetime, timezone
-from sklearn.feature_extraction.text import TfidfVectorizer
-from sklearn.neighbors import NearestNeighbors
-
-def utc_now(): return datetime.now(timezone.utc).isoformat()
-def sha512_text(t): return hashlib.sha512(t.encode()).hexdigest()
-def clean_ws(s): return re.sub(r"\s+", " ", s).strip()
-
-class NexFrameBrain:
-    def __init__(self, base="./core"):
-        self.base = base
-        self.index = {}
-        self.corpus = []
-        self._load_files()
-        self._build_retriever()
-
-    # -------- load your txt/json/doc files ----------
-    def _load_files(self):
-        base = Path(self.base)
-        if not base.exists():
-            print("⚠️ core folder missing"); return
-        for f in base.rglob("*"):
-            if f.suffix.lower() in [".txt", ".json", ".md", ".py"]:
-                try:
-                    txt = open(f, "r", errors="ignore").read()
-                    self.corpus.append({"file": str(f.name), "text": clean_ws(txt)})
-                except Exception as e:
-                    print("skip", f, e)
-        print(f"🧠 Loaded {len(self.corpus)} core files")
-
-    # -------- build small retriever ----------
-    def _build_retriever(self):
-        texts = [d["text"] for d in self.corpus]
-        if not texts:
-            self.vectorizer=None; self.nn=None; return
-        self.vectorizer = TfidfVectorizer(stop_words="english", max_features=10000)
-        X = self.vectorizer.fit_transform(texts)
-        self.nn = NearestNeighbors(n_neighbors=5, metric="cosine").fit(X)
-
-    # -------- main chat ----------
-    def ask(self, query):
-        if not self.vectorizer: return "No data yet."
-        Xq = self.vectorizer.transform([query])
-        dists, idxs = self.nn.kneighbors(Xq)
-        parts=[]
-        for d,i in zip(dists[0],idxs[0]):
-            doc=self.corpus[i]
-            parts.append(f"({doc['file']}) {doc['text'][:400]}...")
-        ans="\n\n".join(parts)
-        return ans or "No match found."
-
-brain = NexFrameBrain()