Open-source AI podcast clipper. Generate vertical clips with face tracking and burned-in captions. CLI, MCP server, web app.
Python ai-toolsauto-captionscaptionscliffmpegmcpmcp-serveropusclip-alternativepodcast-clipspodcastsshortform-videovertical-videowhisper
"""Single, greppable structured log line for backend events.
The render pipeline previously scattered `print(..., file=sys.stderr)` calls
with inconsistent prefixes, which made it hard to answer the most common
question: "why was this clip framed/encoded this way?". Routing those through
`log_event` gives every line a `[category]` prefix and `key=value` fields so the
chosen path is always visible (and easy to grep) without a debugger.
"""
import os
import sys
import time
from contextlib import contextmanager
_VERBOSE = os.environ.get("PODCLI_LOG_VERBOSE", "").lower() in ("1", "true", "yes")
_RESERVED_FIELDS = frozenset({"category", "message", "level", "stage", "ms"})
def log_event(category: str, message: str, *, level: str = "info", **fields) -> None:
"""Emit one structured line: `[category] message k=v k=v`.
`level="debug"` lines are suppressed unless PODCLI_LOG_VERBOSE is set.
"""
if level == "debug" and not _VERBOSE:
return
parts = [f"[{category}]"]
if level in ("warn", "error"):
parts.append(f"{level.upper()}:")
parts.append(message)
extras = " ".join(f"{k}={v}" for k, v in fields.items() if v is not None)
if extras:
parts.append(extras)
print(" ".join(parts), file=sys.stderr, flush=True)
def info(category: str, message: str, **fields) -> None:
log_event(category, message, level="info", **fields)
def warn(category: str, message: str, **fields) -> None:
log_event(category, message, level="warn", **fields)
def debug(category: str, message: str, **fields) -> None:
log_event(category, message, level="debug", **fields)
@contextmanager
def timed(category: str, stage: str, **fields):
"""Time a block and emit `[category] timing stage=... ms=...` on exit.
Debug level, so stage timings stay silent unless PODCLI_LOG_VERBOSE is set.
Emits on failure too: a stage that blows up after 40s is exactly the one
worth seeing. Yields a dict the caller can add fields to before the line is
written, for counts that are only known once the block has run.
"""
extra: dict = {}
start = time.perf_counter()
try:
yield extra
finally:
elapsed_ms = int((time.perf_counter() - start) * 1000)
# A caller field named like a log_event parameter would raise TypeError
# here and mask whatever exception is already unwinding.
payload = {
key: value
for key, value in {**fields, **extra}.items()
if key not in _RESERVED_FIELDS
}
log_event(
category, "timing", level="debug",
stage=stage, ms=elapsed_ms, **payload,
)