File size: 4,377 Bytes
101858b | 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 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 | from __future__ import annotations
import json
import re
import time
from dataclasses import dataclass
from pathlib import Path
from typing import Any, Dict, List, Optional
def _normalize(text: str) -> str:
return re.sub(r"\s+", " ", (text or "")).strip()
def _tokens(text: str) -> List[str]:
return re.findall(r"[a-z0-9_]+", (text or "").lower())
@dataclass
class MemoryItem:
timestamp: float
query: str
text: str
source: str
reward: float
tags: List[str]
metadata: Dict[str, Any]
class PersistentMemoryPool:
def __init__(self, path: str | Path):
self.path = Path(path)
self.path.parent.mkdir(parents=True, exist_ok=True)
self.items: List[MemoryItem] = []
self._load()
def _load(self) -> None:
self.items = []
if not self.path.exists():
return
for line in self.path.read_text(encoding="utf-8").splitlines():
line = line.strip()
if not line:
continue
try:
payload = json.loads(line)
except json.JSONDecodeError:
continue
self.items.append(
MemoryItem(
timestamp=float(payload.get("timestamp", 0.0) or 0.0),
query=str(payload.get("query", "")),
text=str(payload.get("text", "")),
source=str(payload.get("source", "")),
reward=float(payload.get("reward", 0.0) or 0.0),
tags=[str(tag) for tag in payload.get("tags", [])],
metadata=dict(payload.get("metadata", {}) or {}),
)
)
def add(
self,
*,
query: str,
text: str,
source: str,
reward: float = 0.0,
tags: Optional[List[str]] = None,
metadata: Optional[Dict[str, Any]] = None,
) -> None:
item = MemoryItem(
timestamp=time.time(),
query=_normalize(query),
text=_normalize(text),
source=_normalize(source),
reward=float(reward),
tags=[str(tag) for tag in (tags or [])],
metadata=dict(metadata or {}),
)
self.items.append(item)
with self.path.open("a", encoding="utf-8") as handle:
handle.write(
json.dumps(
{
"timestamp": item.timestamp,
"query": item.query,
"text": item.text,
"source": item.source,
"reward": item.reward,
"tags": item.tags,
"metadata": item.metadata,
},
ensure_ascii=False,
)
+ "\n"
)
def search(self, query: str, max_results: int = 5) -> List[Dict[str, Any]]:
query_terms = set(_tokens(query))
ranked: List[tuple[float, MemoryItem]] = []
for item in self.items:
haystack_terms = set(_tokens(item.query + " " + item.text + " " + " ".join(item.tags)))
overlap = len(query_terms.intersection(haystack_terms))
if overlap == 0 and query_terms:
continue
score = float(overlap) + (item.reward * 0.25)
ranked.append((score, item))
ranked.sort(key=lambda pair: (pair[0], pair[1].timestamp), reverse=True)
results: List[Dict[str, Any]] = []
for score, item in ranked[:max_results]:
results.append(
{
"score": round(score, 4),
"query": item.query,
"text": item.text[:400],
"source": item.source,
"reward": item.reward,
"tags": item.tags,
}
)
return results
def build_context(self, query: str, max_results: int = 5, max_chars: int = 1200) -> str:
entries = self.search(query, max_results=max_results)
lines: List[str] = []
total = 0
for item in entries:
line = f"- [{item['source']}] {item['text']}"
total += len(line)
if total > max_chars:
break
lines.append(line)
return "\n".join(lines).strip()
|