Files
llm/scripts/build_1c_knowledge_base.py

115 lines
3.9 KiB
Python

from __future__ import annotations
import argparse
import subprocess
import sys
from pathlib import Path
ROOT = Path(__file__).resolve().parents[1]
DEFAULT_METADATA = ROOT / "plugins" / "1c" / "metadata" / "examples" / "metadata.example.json"
DEFAULT_BSL_MODULES = ROOT / "plugins" / "1c" / "metadata" / "examples" / "bsl-modules.example.json"
DEFAULT_SOURCE_DIR = ROOT / "plugins" / "1c" / "rag" / "sources"
DEFAULT_CORPUS = ROOT / "plugins" / "1c" / "datasets" / "prepared" / "rag_corpus.jsonl"
DEFAULT_MANIFEST = ROOT / "plugins" / "1c" / "datasets" / "prepared" / "rag_manifest.json"
DEFAULT_INDEX = ROOT / "plugins" / "1c" / "datasets" / "prepared" / "rag_index.json"
DEFAULT_VECTOR_INDEX = ROOT / "plugins" / "1c" / "datasets" / "prepared" / "rag_vector_index.sqlite"
def run(command: list[str]) -> None:
print(" ".join(command))
result = subprocess.run(command, cwd=ROOT, text=True, check=False)
if result.returncode != 0:
raise SystemExit(result.returncode)
def main() -> int:
parser = argparse.ArgumentParser(description="Build the 1C RAG knowledge base.")
parser.add_argument("--metadata", type=Path, default=DEFAULT_METADATA)
parser.add_argument("--bsl-modules", type=Path, help="Optional BSL module snapshot JSON.")
parser.add_argument("--source-dir", type=Path, default=DEFAULT_SOURCE_DIR)
parser.add_argument("--corpus", type=Path, default=DEFAULT_CORPUS)
parser.add_argument("--manifest", type=Path, default=DEFAULT_MANIFEST)
parser.add_argument("--index", type=Path, default=DEFAULT_INDEX)
parser.add_argument("--vector-index", type=Path, default=DEFAULT_VECTOR_INDEX)
parser.add_argument("--skip-vector-index", action="store_true")
parser.add_argument(
"--include-example-bsl",
action="store_true",
help="Use the bundled BSL example when --bsl-modules is not provided.",
)
args = parser.parse_args()
metadata_output = args.source_dir / f"{args.metadata.stem}.metadata.md"
run([sys.executable, "scripts/validate_1c_metadata_snapshot.py", str(args.metadata)])
run(
[
sys.executable,
"scripts/convert_1c_metadata_to_rag.py",
"--input",
str(args.metadata),
"--output",
str(metadata_output),
]
)
bsl_modules = args.bsl_modules
if bsl_modules is None and args.include_example_bsl:
bsl_modules = DEFAULT_BSL_MODULES
if bsl_modules:
bsl_output = args.source_dir / f"{bsl_modules.stem}.bsl.md"
run([sys.executable, "scripts/validate_1c_bsl_modules.py", str(bsl_modules)])
run(
[
sys.executable,
"scripts/convert_1c_bsl_modules_to_rag.py",
"--input",
str(bsl_modules),
"--output",
str(bsl_output),
]
)
run([sys.executable, "scripts/validate_1c_rag_sources.py", "--source-dir", str(args.source_dir)])
run(
[
sys.executable,
"scripts/prepare_1c_rag_corpus.py",
"--source-dir",
str(args.source_dir),
"--output",
str(args.corpus),
"--manifest",
str(args.manifest),
]
)
run(
[
sys.executable,
"scripts/build_1c_rag_index.py",
"--corpus",
str(args.corpus),
"--output",
str(args.index),
]
)
if not args.skip_vector_index:
run(
[
sys.executable,
"scripts/build_1c_rag_vector_index.py",
"--corpus",
str(args.corpus),
"--output",
str(args.vector_index),
]
)
print(f"Built 1C knowledge base: {args.index}")
if not args.skip_vector_index:
print(f"Built 1C vector index: {args.vector_index}")
return 0
if __name__ == "__main__":
raise SystemExit(main())