This commit is contained in:
Neo
2026-01-27 22:45:50 +08:00
parent a6ad343092
commit 4c192e921c
476 changed files with 381543 additions and 5819 deletions

View File

@@ -0,0 +1,91 @@
# -*- coding: utf-8 -*-
from __future__ import annotations
import logging
import sys
from contextlib import contextmanager
from datetime import datetime
from pathlib import Path
from typing import Iterator, TextIO
class TeeStream:
def __init__(self, *streams: TextIO) -> None:
self._streams = streams
def write(self, data: str) -> int:
for stream in self._streams:
stream.write(data)
return len(data)
def flush(self) -> None:
for stream in self._streams:
stream.flush()
def isatty(self) -> bool:
return False
def fileno(self) -> int:
return self._streams[0].fileno()
def build_log_path(log_dir: Path, prefix: str, tag: str = "") -> Path:
suffix = f"_{tag}" if tag else ""
stamp = datetime.now().strftime("%Y%m%d_%H%M%S")
return log_dir / f"{prefix}{suffix}_{stamp}.log"
@contextmanager
def configure_logging(
name: str,
log_file: Path | None,
*,
level: str = "INFO",
console: bool = True,
tee_std: bool = True,
) -> Iterator[logging.Logger]:
logger = logging.getLogger(name)
logger.handlers.clear()
logger.setLevel(getattr(logging, level.upper(), logging.INFO))
logger.propagate = False
formatter = logging.Formatter(
"%(asctime)s [%(levelname)s] %(name)s: %(message)s",
"%Y-%m-%d %H:%M:%S",
)
original_stdout = sys.stdout
original_stderr = sys.stderr
log_fp: TextIO | None = None
try:
if log_file:
log_file.parent.mkdir(parents=True, exist_ok=True)
log_fp = open(log_file, "a", encoding="utf-8", buffering=1)
if tee_std:
if console:
sys.stdout = TeeStream(original_stdout, log_fp)
sys.stderr = TeeStream(original_stderr, log_fp)
else:
sys.stdout = log_fp
sys.stderr = log_fp
file_handler = logging.StreamHandler(log_fp)
file_handler.setFormatter(formatter)
logger.addHandler(file_handler)
if console:
console_handler = logging.StreamHandler(original_stdout)
console_handler.setFormatter(formatter)
logger.addHandler(console_handler)
yield logger
finally:
for handler in list(logger.handlers):
handler.flush()
handler.close()
logger.removeHandler(handler)
if log_fp:
log_fp.flush()
log_fp.close()
sys.stdout = original_stdout
sys.stderr = original_stderr

View File

@@ -0,0 +1,97 @@
# -*- coding: utf-8 -*-
"""Time window helpers for ETL and validation tasks."""
from __future__ import annotations
from datetime import datetime, timedelta, time
from typing import List, Tuple
from zoneinfo import ZoneInfo
def _ensure_tz(dt: datetime, tz: ZoneInfo | None) -> datetime:
if tz is None:
return dt
if dt.tzinfo is None:
return dt.replace(tzinfo=tz)
return dt.astimezone(tz)
def _next_month_start(dt: datetime, tz: ZoneInfo | None) -> datetime:
year = dt.year
month = dt.month
if month == 12:
year += 1
month = 1
else:
month += 1
return datetime(year, month, 1, tzinfo=tz)
def calc_window_minutes(start: datetime, end: datetime) -> int:
if end <= start:
return 0
return max(1, int((end - start).total_seconds() // 60))
def split_window(
start: datetime,
end: datetime,
*,
tz: ZoneInfo | None,
split_unit: str | None,
compensation_hours: int | float | None,
) -> List[Tuple[datetime, datetime]]:
start = _ensure_tz(start, tz)
end = _ensure_tz(end, tz)
comp = int(compensation_hours or 0)
if comp:
start = start - timedelta(hours=comp)
end = end + timedelta(hours=comp)
if end <= start:
return []
unit = (split_unit or "").strip().lower()
if unit in ("", "none", "off", "false", "0"):
return [(start, end)]
if unit not in ("month", "monthly"):
return [(start, end)]
windows: List[Tuple[datetime, datetime]] = []
cur = start
while cur < end:
boundary = _next_month_start(cur, tz)
nxt = boundary if boundary < end else end
if nxt <= cur:
break
windows.append((cur, nxt))
cur = nxt
return windows
def build_window_segments(
cfg,
start: datetime,
end: datetime,
*,
tz: ZoneInfo | None,
override_only: bool,
) -> List[Tuple[datetime, datetime]]:
split_unit = cfg.get("run.window_split.unit", "month")
compensation_hours = cfg.get("run.window_split.compensation_hours", 0)
if override_only:
override_start = cfg.get("run.window_override.start")
override_end = cfg.get("run.window_override.end")
if not (override_start and override_end):
split_unit = "none"
compensation_hours = 0
return split_window(
start,
end,
tz=tz,
split_unit=split_unit,
compensation_hours=compensation_hours,
)