What system are these zip permissions from and what do they mean? - zip

I'm trying to implement some code to unzip a file and write its contents to the filesystem with the right permissions. For most zip files, the external attributes are set to 0x040755 or 0x010644 and everything works fine.
However, a user recently complained that a certain zip file wasn't working for them and I took a look at it. It looks like the zip has three different permissions it uses:
0x10 for directories.
0x20 for files.
0x26 for backup files (this is a guess, based on the filename).
I'm guessing this is Windows. Is there any documentation about what numbers correspond to what on Windows-created zip files (or whatever system this is from)?

Related

os.symlink and window's .lnk files are different

I am processing a big images dataset and I'm trying to reorder the files in classes, while at the same time keeping the original directory structure.
To do this, I make a second directory structure with symlinks to the files in the first one.
Everything works as it should but for one small detail: the symlinks created via os.symlink() do not show the image thumbnail, while if I make a link of the same file (e.g., via right click & send to Desktop) I do see the thumbnail.
I wanted to check how the two link files differ (note, the link files themselves, not the linked file), but if I try to drag the os.symlink-generated file in a text editor it opens the linked file instead (while this does not happen with the .lnk file generated via right-click).
What's the difference between the link files? Is os.symlink making something different than a .lnk file? If so, is there a way to get the thumbnail? And if there's no such way, how can I make a .lnk file instead?

How to create a copy of a directory on Linux with links

I have a series of directories on Linux and each directory contains lots of files and data. The data in those directories are automatically generated, but multiple users will need to perform more analysis on that data and generate more files, change the structure, etc.
Since these data directories are very large, I don't want several people to make a copy of the original data so I'd like to make a copy of the directory and link to the original from the new one. However, I'd like any changes to be kept only in the new directory, and leave the original read only. I'd prefer not to link only specific files that I define because the data in these directories is so varied.
So I'm wondering if there is a way to create a copy of a directory by linking to the original but keeping any changed files in the new directory only.
It turns out this is what I wanted to:
cp -al <origdir> <newdir>
It will copy an entire directory and create hard links to the original files. If the original file is deleted, the copied file still exists, and vice-versa. This will work perfectly, but I found newdir must not already exist. As long as the original files are read-only, you'll be able to create an identical, safe copy of the original directory.
However, since you are looking for a way that people can write back changes, UnionFS is probably what you are looking for. It provides means to combine read-only and read-write locations into one.
Unionfs allows any mix of read-only and read-write branches, as well as insertion and deletion of branches anywhere in the fan-out.
Originally I was going to recommend this (I use it a lot):
Assuming the permissions aren't an issue (e.g. only reading is required) I would suggest to bind-mount them into place.
mount -B <original> <new-location>
# or
mount --bind <original> <new-location>
<new-location> must exist as a folder.

How can I specify the version of zip to use in bash?

Background (not necessary to read)
I'm tinkering with MS office files for work (trying to figure out the quickest, easiest way to automate generation of arbitrary-length excel and powerpoint files). Since actual excel files are just zipped archives with .xlsx appended to the filename, I've been unzipping them, editing the xml, rezipping them, and seeing whether OpenOffice can still load them.
However, I've realized (after not too much such testing, thankfully) that, by default, the 'zip' command in bash (or, at least, on my mac) is zipping the files in a format that only requires unzip v1.0 to extract, but normal excel files are zipped in such a way that they require v2.0 to extract. I checked this is a problem by zipping and unzipping an excel file that I knew loaded normally, and then trying to load it. OpenOffice was displeased.
So, I know I need to make the file zip exactly the way excel does, but how to make that happen I'm not sure. I have zip version 3 on my computer, so hopefully if the zip/unzip release cycles are synchronized it should be possible, but I didn't see anything on the man page that immediately seemed to be the solution.
edit:
And zip -9 (which zip -h helpfully says instructs zip to 'zip better') still only requires v1.0 to extract.
Question:
How can I specify in bash that I want zip to zip a file in such a way that it would require unzip v2.0 to unzip?
Often, the reason for an incompatibility between compressed files produced by different versions is the compression algorithm used. If the files were compressed with an algorithm that didn't exist in zip 1.0, that would cause the incompatibility you're seeing.
Look at the man page for your zip utility, see if there's an option to specify the type of compression to use. If there is, look at the existing files created from Excel, and find out what type of compression algorithm they're compressed with, and use that.
On my Linux system, zip reports "This is Zip 2.31 (March 8th 2005), by Info-ZIP.", and it does not have an option for specifying the compression algorithm. On my Windows system, 7-zip does have the option, and it looks like they do have a Mac version available, so you could try that if your zip utility doesn't support that option.

Apply exact stat parameters from one file to another using C in Linux

I would like get stat parameters from one file and later apply it as is, to a copy of the same file (Including type, path, permissions, size, etc.).
The original file will be long gone from this directory and the copy will take his place, and should get the same exact properties.
How could this be done using C in Linux?
If I understand your question correctly, you don't need to write your own program to do that.
If your files reside on the same machine, you can preserve and restore times and permissions in tar archives. The p option handles permissions, and times are persisted by default (except atime, but --atime-preserve can work around that).
Alternatively, if you want to restore files from a remote server, you can use rsync with the -a option.

Can a Self Extracting Zip File read a registry entry?

I'm trying to get my website to talk to a friend's program. Think ITunes - one main program with hundreds of thousands of little things installed into it. We don't want to have to create an InstallShield install program for each of those hundreds of thousands of little things.
We have the files grouped into the folder sub-structure.
We have a .REG file for what registry entry needs to be added to see the new folder group.
But is there a way to do a self extracting zip file that reads a registry entry so we know where they installed the original program to be able to dump the new files there as well? I want them to double-click the EXE and click Finish and for everything to work.
(I've been looking into INF and CAB files through IExpress.exe, but haven't found the answer. I remember Package for the Web didn't have an option to read a registry entry, but did let you modify the suggested install path.)
Thanks so much.
Best wishes,
Andrea
But is there a way to do a self extracting zip file that reads a registry entry so we know where they installed the original program to be able to dump the new files there as well? I want them to double-click the EXE and click Finish and for everything to work
Well, yes and no. There are self-extractors that can run a program after extracting all files. DotNetZip, for example, can produce an SFX which can do this.
Just an aside: a normal SFX is just a zip file, with a "stub" executable merged with it. The stub exe can do anything it wants to do, but the most basic thing it does is extract the files in the zip. When you use DotNetZip to produce an SFX, it embeds its particular stub into the zip. That stub knows how to extract files, and also knows how to invoke a program after extracting. You can also produce your own stub that can do other more exotic things.
So you could use an SFX for your purpose. When run, it would extract, then invoke it's extra program. The program could look in the registry, then move or relocate the extracted files to the appropriate place. Then terminate.
For a different twist, the SFX might have just two files: the program-to-run (the one that reads the registry, and another embedded zip. Then when the SFX runs it generates 2 files. Then it invokes the program-to-run, which reads the registry, then unpacks the contained zip and puts the files into the desired place.
Ok, so you could do it.
Should you?
mmm, maybe. This really is an installer, so, you should decide whether you want to use a zip as an installer. Don't forget, if you use an SFX as an installer, there's no good way to uninstall.
Have you tried Inno Setup toolchain? It's a bit better than a bare Self-Extracting ZIP file, it's a setup creation utility. I'm convinced it has got something to put some entry in the Registry, look also at the plugins.
Basically, a self-extracting executable that alters the registry, it's a setup program. So why don't you go for a proper one?
Website: http://www.jrsoftware.org/isinfo.php

Resources