oclaw/runtime/tools/public/search_files_context_tool.py

103 lines
3.7 KiB
Python

from __future__ import annotations
import re
from pathlib import Path
from typing import Any
from oclaw.runtime.tools.base import ToolSpec
from oclaw.runtime.tools.path_guard import resolve_workspace_path, truncate_text
def search_files_context_tool() -> ToolSpec:
def _iter_files(root: Path, file_glob: str) -> list[Path]:
if root.is_file():
return [root]
try:
return [p for p in root.rglob(file_glob) if p.is_file()]
except Exception:
return []
def _handler(args: dict[str, Any]) -> dict[str, Any]:
root_raw = str(args.get("root") or ".").strip()
file_glob = str(args.get("file_glob") or "**/*").strip() or "**/*"
pattern = str(args.get("pattern") or "").strip()
regex = bool(args.get("regex", True))
max_matches = int(args.get("max_matches") or 200)
context_lines = int(args.get("context_lines") or 2)
if max_matches <= 0:
max_matches = 1
if context_lines < 0:
context_lines = 0
try:
root = resolve_workspace_path(root_raw)
except Exception as exc:
return {"ok": False, "error": "invalid_root", "detail": str(exc)}
if not pattern:
return {"ok": False, "error": "pattern_required"}
matcher = re.compile(pattern) if regex else None
out: list[dict[str, Any]] = []
scanned = 0
for fp in _iter_files(root, file_glob):
scanned += 1
if len(out) >= max_matches:
break
try:
lines = fp.read_text(encoding="utf-8", errors="replace").splitlines()
except Exception:
continue
for idx, line in enumerate(lines):
hit = matcher.search(line) if matcher else (pattern in line)
if not hit:
continue
start = max(0, idx - context_lines)
end = min(len(lines), idx + context_lines + 1)
out.append(
{
"path": str(fp),
"line": idx + 1,
"match": line,
"context": [f"{i + 1}|{lines[i]}" for i in range(start, end)],
}
)
if len(out) >= max_matches:
break
return {
"ok": True,
"root": str(root),
"file_glob": file_glob,
"pattern": pattern,
"regex": regex,
"scanned_files": scanned,
"matches": out,
}
return ToolSpec(
name="search_files_context",
description="Search files under a root and return matched lines with surrounding context.",
parameters={
"type": "object",
"properties": {
"root": {"type": "string", "description": "Directory to search under."},
"file_glob": {"type": "string", "description": "File glob under root.", "default": "**/*"},
"pattern": {"type": "string", "description": "Regex or substring pattern."},
"regex": {"type": "boolean", "description": "Treat pattern as regex if true.", "default": True},
"max_matches": {"type": "integer", "description": "Max matches to return.", "default": 200},
"context_lines": {"type": "integer", "description": "Number of lines of context on each side.", "default": 2},
},
"required": ["pattern"],
"additionalProperties": False,
},
handler=_handler,
tags=frozenset({"public", "search", "workspace"}),
read_only=True,
risk_level="low",
)
__all__ = ["search_files_context_tool"]