Fix progress tracking for multibyte text files

This commit is contained in:
Rami Abdelrazzaq 2026-09-12 09:12:01 -05:00
commit b260002d61

View file

@ -208,15 +208,39 @@ class UpdateWrapper:
def __init__(self, wrapped: io.IOBase, update: Callable[[int], None]) -> None:
self._wrapped = wrapped
self._update = update
try:
self._position: int | None = wrapped.tell()
except (AttributeError, OSError):
self._position = None
def _update_progress(self, fallback_length: int) -> None:
if self._position is None:
self._update(fallback_length)
return
try:
position = self._wrapped.tell()
except (AttributeError, OSError):
self._position = None
self._update(fallback_length)
return
delta = position - self._position
self._update(delta if delta >= 0 else fallback_length)
self._position = position
def __iter__(self) -> Iterator[bytes]:
for line in self._wrapped:
self._update(len(line))
# readline() keeps TextIOWrapper.tell() available, unlike iterating the
# TextIOWrapper directly. Its position is in the underlying file's
# bytes, so multibyte encodings advance the progress bar correctly.
while True:
line = self._wrapped.readline()
if not line:
break
self._update_progress(len(line))
yield line
def read(self, size: int = -1) -> bytes:
data = self._wrapped.read(size)
self._update(len(data))
self._update_progress(len(data))
return data
@ -659,4 +683,4 @@ def flatten(row: dict[str, Any]) -> dict[str, Any]:
:param row: A Python dictionary, optionally with nested dictionaries
"""
return dict(_flatten(row))
return dict(_flatten(row))