Article CRC check, sabnzbd 0.7.11
Posted: April 27th, 2013, 3:17 pm
I discovered a lot of INFO message in the log of sabnzbd, likes this one
It looks like the CRC result is stored in 64 bits, instead of 32 bits
Although the article is tagged as corrupted and missing, the downloaded file is valid, which is normal since the articles tagged as corrupted are valid.
For information, I am running on 32 bits PC
Edit: Well after code analysis, I think the problem is inside the header of the message. The solution could be to keep the 8 last characters of the CRC (decoder.py : 299)
Edit2: To fix it (tested and working), I added this line after line 299
Or maybe (not tested) replace line 299 by
Code: Select all
INFO::[decoder:121] Erreur CRC dans [email protected] (FFFFFFFF2F249386 -> 2F249386)Although the article is tagged as corrupted and missing, the downloaded file is valid, which is normal since the articles tagged as corrupted are valid.
For information, I am running on 32 bits PC
Code: Select all
Linux benjarobin 3.8.8-2-ARCH #1 SMP PREEMPT Tue Apr 23 11:35:43 CEST 2013 i686 GNU/LinuxEdit2: To fix it (tested and working), I added this line after line 299
Code: Select all
_partcrc = _partcrc[-8:]Code: Select all
_partcrc = '00000000' + yend[crcname].upper()
_partcrc = _partcrc[-8:]