vanishingradient's picture
Add deepagents_rl package sources
4d620d9 verified
Raw History Blame Contribute Delete
5.22 kB
"""Structured step logging for deepagents-rl pipelines.
Emits:
- Console (stderr) with timestamps + step tags
- Optional JSONL event file for full replay (`events.jsonl`)
Env:
DARL_LOG_LEVEL DEBUG|INFO|WARNING (default INFO)
DARL_LOG_FILE path to text log (optional)
DARL_EVENTS_JSONL path to JSONL events (optional; auto under artifacts if set via configure)
"""
from __future__ import annotations
import json
import logging
import os
import sys
import threading
import time
import traceback
from contextlib import contextmanager
from datetime import datetime, timezone
from pathlib import Path
from typing import Any, Iterator
_LOCK = threading.Lock()
_EVENTS_PATH: Path | None = None
_CONFIGURED = False
LOGGER_NAME = "deepagents_rl"
def _utc() -> str:
return datetime.now(timezone.utc).isoformat()
def get_logger(name: str | None = None) -> logging.Logger:
return logging.getLogger(name or LOGGER_NAME)
def configure_logging(
*,
level: str | None = None,
log_file: str | Path | None = None,
events_jsonl: str | Path | None = None,
force: bool = False,
) -> logging.Logger:
"""Idempotent logging setup for CLI scripts and harness modules."""
global _CONFIGURED, _EVENTS_PATH
log = get_logger()
lvl_name = (level or os.getenv("DARL_LOG_LEVEL") or "INFO").upper()
lvl = getattr(logging, lvl_name, logging.INFO)
if _CONFIGURED and not force:
log.setLevel(lvl)
if events_jsonl:
_EVENTS_PATH = Path(events_jsonl)
_EVENTS_PATH.parent.mkdir(parents=True, exist_ok=True)
return log
log.setLevel(lvl)
log.handlers.clear()
log.propagate = False
fmt = logging.Formatter(
fmt="%(asctime)s | %(levelname)-5s | %(name)s | %(message)s",
datefmt="%H:%M:%S",
)
sh = logging.StreamHandler(sys.stderr)
sh.setLevel(lvl)
sh.setFormatter(fmt)
log.addHandler(sh)
file_path = log_file or os.getenv("DARL_LOG_FILE")
if file_path:
fp = Path(file_path)
fp.parent.mkdir(parents=True, exist_ok=True)
fh = logging.FileHandler(fp, encoding="utf-8")
fh.setLevel(lvl)
fh.setFormatter(fmt)
log.addHandler(fh)
ev = events_jsonl or os.getenv("DARL_EVENTS_JSONL")
if ev:
_EVENTS_PATH = Path(ev)
_EVENTS_PATH.parent.mkdir(parents=True, exist_ok=True)
else:
_EVENTS_PATH = None
_CONFIGURED = True
log.debug("logging configured level=%s file=%s events=%s", lvl_name, file_path, _EVENTS_PATH)
return log
def _write_event(event: dict[str, Any]) -> None:
if _EVENTS_PATH is None:
return
with _LOCK:
with _EVENTS_PATH.open("a", encoding="utf-8") as f:
f.write(json.dumps(event, default=str) + "\n")
def log_event(
event_type: str,
message: str,
*,
level: int = logging.INFO,
**fields: Any,
) -> None:
"""Log a structured event to console + JSONL."""
log = get_logger()
payload = {
"ts": _utc(),
"event": event_type,
"message": message,
**fields,
}
# Compact console line
extras = {k: v for k, v in fields.items() if k not in ("traceback",)}
extra_s = ""
if extras:
try:
extra_s = " | " + json.dumps(extras, default=str)[:1500]
except Exception:
extra_s = f" | {extras!r}"[:1500]
log.log(level, f"[{event_type}] {message}{extra_s}")
_write_event(payload)
def step(name: str, message: str = "", **fields: Any) -> None:
log_event("step", message or name, step=name, **fields)
def call(name: str, message: str = "", **fields: Any) -> None:
log_event("call", message or name, call=name, **fields)
def result(name: str, message: str = "", **fields: Any) -> None:
log_event("result", message or name, call=name, **fields)
def error(name: str, message: str, **fields: Any) -> None:
log_event("error", message, level=logging.ERROR, call=name, **fields)
@contextmanager
def span(name: str, message: str = "", **fields: Any) -> Iterator[dict[str, Any]]:
"""Timed span: logs start + end with duration_ms."""
t0 = time.perf_counter()
ctx: dict[str, Any] = {"span": name, **fields}
log_event("span_start", message or name, **ctx)
try:
yield ctx
except Exception as e:
dt = int((time.perf_counter() - t0) * 1000)
error(
name,
f"{type(e).__name__}: {e}",
duration_ms=dt,
traceback=traceback.format_exc()[-3000:],
**{k: v for k, v in ctx.items() if k != "span"},
)
raise
else:
dt = int((time.perf_counter() - t0) * 1000)
log_event(
"span_end",
message or name,
duration_ms=dt,
ok=True,
**ctx,
)
def truncate(s: Any, n: int = 800) -> str:
text = s if isinstance(s, str) else json.dumps(s, default=str)
if len(text) <= n:
return text
return text[:n] + f"...(+{len(text)-n} chars)"