package / tmp file usage ...

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Attila & Noel,

	Just moving this to the public dev list:

On 13/02/2023 07:43, Attila Szűcs wrote:
     I plan to make test for zip64 ...
    I plan to check if i could avoid temporal files creations during export/import.. to mke a unit test that have 4gb sized files, but not write 4gb to SSD in every test run.

Noel might be interested in this. Noel, Attila suggests that every zip file we write we have to stream the full un-compressed data to /tmp files before compressing it. And these can be huge =)

That is (I guess) necessary if we want to be able to seek backwards and edit eg. Excel binary streams or whatever - but for the vast majority of our one-pass, hyper-verbose XML generation it seems very silly - everything would be better if we could deflate compress that in another thread as we go I think ;-) perhaps then backed by a file but much smaller than the 4Gb+ of rather verbose ODS/ODF for a large spreadsheet that we need 64bit un-compressed stream sizes for.

Noel - I wondered if you'd looked at this around saving performance - COOL at least does a lot more saving than loading - re: autosave =) I forget what we did in the past here.

I imagine we could ask for a non-seekable zip stream somehow from the package code just for this case (?)

	Just a thought,
	
		Michael.

--
michael.meeks@xxxxxxxxxxxxx <><, GM Collabora Productivity
Hangout: mejmeeks@xxxxxxxxx, Skype: mmeeks
(M) +44 7795 666 147 - timezone usually UK / Europe



[Index of Archives]     [LARTC]     [Bugtraq]     [Yosemite Forum]     [Photo]

  Powered by Linux