Zip the documents

[jflesch]

It could be useful to actually zip each document:

• It would bring Paperwork closer to the way OpenDocuments work
• It would make documents transfer easier
• It would reduce the stress on the filesystem

[jflesch]

Beware of #124

[jflesch]

Will be redundant with #124

[jflesch]

Yeah, #124 is not happening before a very long time, so let's go with this for now.

[jflesch]

Should only be used on small documents (< 20 pages I guess). It would make image modifications really CPU-intensive / disk-intensive on big documents.

[jflesch]

or maybe .tar.gz :)

[tYYGH]

I’m not sure this is a good idea:

• How will this put less stress on the filesystem? The overall size should be almost the same since PDF and JPEG images don’t compress very well.
• When a part of a document will be read, the whole zip/tgz will have to be read.
• Same for writing only a part (rotating a page, changing the labels…)
• It will make Paperwork even more slow than it is today.

[jflesch]

How will this put less stress on the filesystem? The overall size should be almost the same since PDF and JPEG images don’t compress very well.

Less files --> less inodes ; + less modification time to check when Paperwork starts.

When a part of a document will be read, the whole zip/tgz will have to be read.

I need to check but I believe zip or tgz (or both) have indexes.

Same for writing only a part (rotating a page, changing the labels…)

Yes, I know, this is the main problem :/

It will make Paperwork even more slow than it is today.

Yes and no. It would make the start time much faster actually.

[jflesch]

I had a look:

• .zip have an index at the end of the file : https://en.wikipedia.org/wiki/Zip_(file_format)#Structure
• .tar.gz don't have any index

So .zip might more indicated here. I gues it also explains why LibreOffice and Office both use them too.

[jflesch]

Note that it would also help reduce fragmentation, which could improve documents load time:

• Ext* file systems try to keep single files in one single row. But files individually may be placed randomly on the hard drive
• When opening a multiple-pages document, Paperwork usually loads the page sequentially at first

[jflesch]

However, I guess keeping the labels out of the .zip file could be a good idea. No need to rewrite X MB when all you want is just fix the labels that have been guessed.

[tYYGH]

That would be a reasonable compromise indeed.