sync from accepted repo: init_knowledge.py

This commit is contained in:
2026-05-27 12:36:17 +00:00
parent 5f93cc403f
commit 808a916782
+21 -20
View File
@@ -4,28 +4,30 @@ from __future__ import annotations
import argparse
from pathlib import Path
try:
from knowledge_base import build_knowledge_base
except ModuleNotFoundError:
from .knowledge_base import build_knowledge_base
from knowledge_base import KnowledgeBase
SUPPORTED_SUFFIXES = {".txt", ".md", ".markdown"}
def load_directory(directory: str | Path) -> int:
root = Path(directory)
if not root.exists():
raise FileNotFoundError(f"Директория не найдена: {root}")
def load_directory(knowledge_dir: Path, kb: KnowledgeBase | None = None) -> int:
kb = kb or KnowledgeBase()
if not knowledge_dir.exists():
knowledge_dir.mkdir(parents=True, exist_ok=True)
print(f"Создана пустая директория: {knowledge_dir}")
return 0
kb = build_knowledge_base()
total_chunks = 0
for path in sorted(root.rglob("*")):
if not path.is_file() or path.suffix.lower() not in SUPPORTED_SUFFIXES:
continue
content = path.read_text(encoding="utf-8")
title = str(path.relative_to(root))
total_chunks += kb.add_document(content, title)
print(f"Загружен: {title} ({len(content)} символов)")
files = [p for p in knowledge_dir.rglob("*") if p.suffix.lower() in SUPPORTED_SUFFIXES]
if not files:
print(f"Нет файлов {SUPPORTED_SUFFIXES} в {knowledge_dir}")
return 0
for file_path in sorted(files):
chunks = kb.add_file(file_path)
total_chunks += chunks
print(f" + {file_path.name}: {chunks} чанк(ов)")
print(f"Готово: {len(files)} файл(ов), {total_chunks} чанк(ов).")
return total_chunks
@@ -33,12 +35,11 @@ def main() -> None:
parser = argparse.ArgumentParser(description="Инициализация базы знаний из папки")
parser.add_argument(
"--dir",
default="./docs",
help="Путь к директории с .txt/.md файлами",
default="./knowledge",
help="Путь к директории с документами (.txt, .md)",
)
args = parser.parse_args()
chunks = load_directory(args.dir)
print(f"Готово. Всего чанков: {chunks}")
load_directory(Path(args.dir))
if __name__ == "__main__":