grufino commented on code in PR #22419:
URL: https://github.com/apache/beam/pull/22419#discussion_r930043048
##########
sdks/python/apache_beam/io/filesystem.py:
##########
@@ -166,6 +177,9 @@ def _initialize_decompressor(self):
self._decompressor = bz2.BZ2Decompressor()
elif self._compression_type == CompressionTypes.DEFLATE:
self._decompressor = zlib.decompressobj()
+ elif self._compression_type == CompressionTypes.ZSTD:
+ self._decompressor = zstandard.ZstdDecompressor(
+ max_window_size=2147483648).decompressobj()
Review Comment:
I know this seems arbitrary (and probably is), but without this param the
decompressor fails with the following error when reading big files:
```
zstd.ZstdError: zstd decompress error: Frame requires too much memory for
decoding
```
I thought about using `DEFAULT_READ_BUFFER_SIZE` but it seems too low for
this param. I can either create a constant or a new parameter if necessary, let
me know what you find best.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
