I have a large .xz file (few gigabytes). It's full of plain text. I want to process the text to create custom dataset. I want to read it line by line because it is too big. Anyone have an idea how to do it ?

I already tried this How to open and read LZMA file in-memory but it's not working.

EDIT: i got this error 'ascii' codec can't decode byte 0xfd in position 0: ordinal not in range(128)

on the line for line in uncompressed: from the link

EDIT2: My code (using python 3.5)

with open(filename) as compressed:with lzma.LZMAFile(compressed) as uncompressed:for line in uncompressed:print(line)

I was faced to the same question some weeks ago. This snippet worked for me:

import lzma
with lzma.open('filename.xz', mode='rt') as file:for line in file:print(line)

This assumes that the text data in the compressed file was encoded in utf-8 (which was the case for my data). There is an encoding argument in function lzma.open() which allows you to set another encoding if needed

EDIT (after you own edit): try to force encoding='utf-8' in lmza.open()