Repository navigation
System.IO.Compression should store NoCompression entries without wrapping them in Deflate streams #26185
Description
Activity
Sounds like a reasonable thing to do. @ianhays?
Totally reasonable, and not the first time I've seen this request either. I can't find the other one with a quick search, though.
I can't think of any real circumstances in which removing the nocompression entry's deflate header would cause issues. For testing we should try a zip produced by the new codeset with windows' zip library, unix
unzip, and 7zip. Also should probably add a unit test of some sort.Well. I basically need this right now. Anyone has a fork or so where this is working?
Sounds like @Marv51 has the code change, above. We would welcome a PR.
- As I‘ve not contributed here before, I didn‘t have the environment/tests setup. But I will try to get this working again tonight. Dan Moseley <[email protected]> schrieb am Do. 8. Nov. 2018 um 17:39:…Sounds like @Marv51 <https://github.com/Marv51> has the code change, above. We would welcome a PR. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <https://github.com/dotnet/corefx/issues/29735#issuecomment-437064348>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAcxKks1XQdSHquKp2c7zf6reyjL_hu2ks5utF49gaJpZM4UBAWp> .
Oh, very nice. Thank you. :)
Thanks @Marv51. Let us know if the documentation in this repo is insufficient.
- added a commit that references this issue
on Nov 8, 2018 - added a commit that references this issue
on Dec 16, 2018 - ghost locked as resolved and limited conversation to collaborators
on Dec 17, 2020
Currently when a ZipArchive is created and an entry is added with the CompressionLevel set to "no-compression" the file inside the archive is still wrapped in a deflate stream. The deflate stream correctly does not to compress the file.
However, zip does natively support uncompressed files. The compression method "stored" is made for this.
To create OpenDocument format files (ods, odt etc.) it is required to store the mimetype file without wrapping it in any compression stream. They do this, so that the mimetype can always found at a specific offset.
This change should also be a tiny performance and archive size improvement.
I have attached a diff of the changes necessary to implement this. I've run the tests, however I've not yet been able to build corefx in a way that I can use with my app to validate the patch is working.
diff.zip
If this is something you think is possible to merge, I can spend more time on this to create a proper PR.