File size: 2,431 Bytes
fbf3c28 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 | """
id: rag
title: Lightweight RAG
author: admin
description: Store documents and retrieve by simple keyword scoring.
version: 0.1.0
license: Proprietary
"""
import json
import re
import time
from pathlib import Path
ROOT = Path("/data/adaptai/rag")
ROOT.mkdir(parents=True, exist_ok=True)
INDEX = ROOT / "index.json"
class Tools:
def add_text(self, text: str, doc_id: str = "") -> dict:
doc_id = doc_id or f"doc_{int(time.time())}"
(ROOT / f"{doc_id}.txt").write_text(text, encoding="utf-8")
self._reindex()
return {"ok": True, "doc_id": doc_id}
def add_file(self, path: str, doc_id: str = "") -> dict:
p = Path(path)
if not p.exists():
return {"ok": False, "error": "file not found"}
text = p.read_text(encoding="utf-8", errors="replace")
return self.add_text(text, doc_id)
def query(self, q: str, top_k: int = 5) -> dict:
idx = self._load_index()
tokens = self._tok(q)
scored = []
for doc_id, dt in idx.items():
score = sum(dt.get(t, 0) for t in tokens)
if score:
scored.append((score, doc_id))
scored.sort(reverse=True)
hits = []
for score, doc_id in scored[:top_k]:
text = (ROOT / f"{doc_id}.txt").read_text(
encoding="utf-8", errors="replace"
)
hits.append({"doc_id": doc_id, "score": score, "preview": text[:400]})
return {"query": q, "hits": hits}
def list(self) -> dict:
idx = self._load_index()
return {"docs": list(idx.keys())}
def remove(self, doc_id: str) -> dict:
try:
(ROOT / f"{doc_id}.txt").unlink(missing_ok=True)
self._reindex()
return {"ok": True}
except Exception as e:
return {"ok": False, "error": str(e)}
def _tok(self, s: str):
s = s.lower()
return re.findall(r"[a-z0-9]+", s)
def _reindex(self):
idx = {}
for f in ROOT.glob("*.txt"):
counts = {}
for t in self._tok(f.read_text(encoding="utf-8", errors="replace")):
counts[t] = counts.get(t, 0) + 1
idx[f.stem] = counts
INDEX.write_text(json.dumps(idx), encoding="utf-8")
def _load_index(self):
if not INDEX.exists():
return {}
return json.loads(INDEX.read_text(encoding="utf-8"))
|