#!/usr/bin/env python3
"""Bootstrap script: populates the `code` LanceDB collection from the
local repo tree, so the vector store isn't empty on first deploy.

Usage: python bootstrap_vector_store.py [repo_root]

Chunking is deliberately simple (fixed line-count windows) rather than
AST-aware — good enough to make code semantically searchable without
pulling in a language-specific parser for this first pass.
"""
import sys
from pathlib import Path

import vector_store

INCLUDE_SUFFIXES = {".rs", ".py", ".toml", ".md"}
EXCLUDE_DIRS = {"target", "__pycache__", ".git", "node_modules"}
CHUNK_LINES = 60
CHUNK_OVERLAP = 10


def iter_source_files(root: Path):
    for path in root.rglob("*"):
        if path.is_dir():
            continue
        if any(part in EXCLUDE_DIRS for part in path.parts):
            continue
        if path.suffix in INCLUDE_SUFFIXES:
            yield path


def chunk_lines(lines: list[str]) -> list[str]:
    if not lines:
        return []
    chunks = []
    step = CHUNK_LINES - CHUNK_OVERLAP
    for start in range(0, len(lines), step):
        chunk = "".join(lines[start : start + CHUNK_LINES])
        if chunk.strip():
            chunks.append(chunk)
        if start + CHUNK_LINES >= len(lines):
            break
    return chunks


def main():
    root = Path(sys.argv[1]) if len(sys.argv) > 1 else Path(__file__).resolve().parents[1]
    if not root.exists():
        print(f"error: {root} does not exist")
        sys.exit(1)

    total_files = 0
    total_chunks = 0
    for path in iter_source_files(root):
        try:
            lines = path.read_text(encoding="utf-8", errors="ignore").splitlines(keepends=True)
        except OSError as exc:
            print(f"skip {path}: {exc}")
            continue

        rel_path = str(path.relative_to(root))
        for i, chunk in enumerate(chunk_lines(lines)):
            vector_store.ingest_code_chunk(rel_path, i, chunk)
            total_chunks += 1
        total_files += 1
        print(f"indexed {rel_path} ({len(lines)} lines)")

    print(f"done: {total_files} files, {total_chunks} chunks indexed into 'code' collection")


if __name__ == "__main__":
    main()