From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from firstgate.proxmox.com (firstgate.proxmox.com [212.224.123.68]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits)) (No client certificate requested) by lists.proxmox.com (Postfix) with ESMTPS id A1E639D6D0 for ; Fri, 24 Nov 2023 08:57:31 +0100 (CET) Received: from firstgate.proxmox.com (localhost [127.0.0.1]) by firstgate.proxmox.com (Proxmox) with ESMTP id 805D090A9 for ; Fri, 24 Nov 2023 08:57:01 +0100 (CET) Received: from proxmox-new.maurer-it.com (proxmox-new.maurer-it.com [94.136.29.106]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits)) (No client certificate requested) by firstgate.proxmox.com (Proxmox) with ESMTPS for ; Fri, 24 Nov 2023 08:57:00 +0100 (CET) Received: from proxmox-new.maurer-it.com (localhost.localdomain [127.0.0.1]) by proxmox-new.maurer-it.com (Proxmox) with ESMTP id 67CD441BFD; Fri, 24 Nov 2023 08:57:00 +0100 (CET) Message-ID: <0f719e51-3407-4923-bd22-502efb4bca4d@proxmox.com> Date: Fri, 24 Nov 2023 08:56:59 +0100 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Beta Content-Language: en-US To: Proxmox Backup Server development discussion , Filip Schauer References: <20231123130652.416129-1-f.schauer@proxmox.com> <20231123130652.416129-2-f.schauer@proxmox.com> From: Dominik Csapak In-Reply-To: <20231123130652.416129-2-f.schauer@proxmox.com> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-SPAM-LEVEL: Spam detection results: 0 AWL 0.017 Adjusted score from AWL reputation of From: address BAYES_00 -1.9 Bayes spam probability is 0 to 1% DMARC_MISSING 0.1 Missing DMARC policy KAM_DMARC_STATUS 0.01 Test Rule for DKIM or SPF Failure with Strict Alignment SPF_HELO_NONE 0.001 SPF: HELO does not publish an SPF Record SPF_PASS -0.001 SPF: sender matches SPF record T_SCC_BODY_TEXT_LINE -0.01 - URIBL_BLOCKED 0.001 ADMINISTRATOR NOTICE: The query to URIBL was blocked. See http://wiki.apache.org/spamassassin/DnsBlocklists#dnsbl-block for more information. [self.target, zip.rs, cachefly.net] Subject: Re: [pbs-devel] [PATCH v2 proxmox 1/1] fix #4995: compression: Include symlinks in zip file restore X-BeenThere: pbs-devel@lists.proxmox.com X-Mailman-Version: 2.1.29 Precedence: list List-Id: Proxmox Backup Server development discussion List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 24 Nov 2023 07:57:31 -0000 a few high level comments (did not look too closely on the code): * if we change the function/struct parameters anyway wouldn't it make more sense to add a 'filetype' enum instead of having a 'is_file' bool and a symlink option? i used the bool because we only had files + dirs, but now we add a third type, but imho representing the types properly would be better we even could put the content into the various enum parts (or even make the ZipEntry an enum altogether?) i know this refactoring is much more work than slapping just a new parameter on, but it makes it easier to understand the code and expand it if we need it (honestly i probably should have done so initially when adding the code) * what i'm missing here a bit is the source on how to encode symlinks in zip. the "official" zip spec[0] only talks about (symbolic) links in the description of a "-UNIX Extra Field" but you simply encode it here into the content how did you arrive at that solution? (also generally a commit message is a good idea ;) ) * for these things i'd also like a short comment (does not have to be in the commit message) on which systems you did test this, e.g. zipinfo/zip/unar on linux explorer on windows, mac (?), etc. * if you want to go the extra mile, i guess this would be a good time to add tests that create a new zip from test data, to see if they don't break with your changes 0: https://pkware.cachefly.net/webdocs/APPNOTE/APPNOTE-6.3.9.TXT On 11/23/23 14:06, Filip Schauer wrote: > Signed-off-by: Filip Schauer > --- > proxmox-compression/src/zip.rs | 46 ++++++++++++++++++++++++++-------- > 1 file changed, 35 insertions(+), 11 deletions(-) > > diff --git a/proxmox-compression/src/zip.rs b/proxmox-compression/src/zip.rs > index d2d3fd8..e30f50a 100644 > --- a/proxmox-compression/src/zip.rs > +++ b/proxmox-compression/src/zip.rs > @@ -204,6 +204,7 @@ pub struct ZipEntry { > offset: u64, > is_file: bool, > is_utf8_filename: bool, > + symlink_target: Option, > } > > impl ZipEntry { > @@ -211,7 +212,13 @@ impl ZipEntry { > /// > /// if is_file is false the path will contain an trailing separator, > /// so that the zip file understands that it is a directory > - pub fn new>(path: P, mtime: i64, mode: u16, is_file: bool) -> Self { > + pub fn new>( > + path: P, > + mtime: i64, > + mode: u16, > + is_file: bool, > + symlink_target: Option

, > + ) -> Self { > let mut relpath = PathBuf::new(); > > for comp in path.as_ref().components() { > @@ -226,6 +233,7 @@ impl ZipEntry { > > let filename: OsString = relpath.into(); > let is_utf8_filename = filename.to_str().is_some(); > + let symlink_target_osstr = symlink_target.map(|x| x.as_ref().into()); > > Self { > filename, > @@ -237,6 +245,7 @@ impl ZipEntry { > offset: 0, > is_file, > is_utf8_filename, > + symlink_target: symlink_target_osstr, > } > } > > @@ -360,7 +369,9 @@ impl ZipEntry { > comment_len: 0, > start_disk: 0, > internal_flags: 0, > - external_flags: (self.mode as u32) << 16 | (!self.is_file as u32) << 4, > + external_flags: (self.mode as u32) << 16 > + | (self.symlink_target.is_some() as u32) << 5 > + | (!self.is_file as u32) << 4, > offset, > }, > ) > @@ -486,23 +497,30 @@ impl ZipEncoder { > .ok_or_else(|| format_err!("had no target during add entry"))?; > entry.offset = self.byte_count.try_into()?; > self.byte_count += entry.write_local_header(&mut target).await?; > - if let Some(content) = content { > - let mut reader = HashWrapper::new(content); > + > + if content.is_some() || entry.symlink_target.is_some() { > let mut enc = DeflateEncoder::with_quality(target, Level::Fastest); > > - enc.compress(&mut reader).await?; > + if let Some(content) = content { > + let mut reader = HashWrapper::new(content); > + enc.compress(&mut reader).await?; > + entry.crc32 = reader.finish().0; > + } else if let Some(symlink_target) = entry.symlink_target.as_ref() { > + let cursor = std::io::Cursor::new(symlink_target.as_bytes()); > + let mut reader = HashWrapper::new(cursor); > + enc.compress(&mut reader).await?; > + entry.crc32 = reader.finish().0; > + } > + > let total_in = enc.total_in(); > let total_out = enc.total_out(); > target = enc.into_inner(); > > - let (crc32, _reader) = reader.finish(); > - > self.byte_count += total_out as usize; > entry.compressed_size = total_out; > entry.uncompressed_size = total_in; > - > - entry.crc32 = crc32; > } > + > self.byte_count += entry.write_data_descriptor(&mut target).await?; > self.target = Some(target); > > @@ -658,10 +676,16 @@ where > > if entry.file_type().is_file() { > let file = tokio::fs::File::open(entry.path()).await?; > - let ze = ZipEntry::new(entry_path_no_base, mtime, mode, true); > + let ze = ZipEntry::new(entry_path_no_base, mtime, mode, true, None); > Ok(Some((ze, Some(file)))) > } else if entry.file_type().is_dir() { > - let ze = ZipEntry::new(entry_path_no_base, mtime, mode, false); > + let ze = ZipEntry::new(entry_path_no_base, mtime, mode, false, None); > + let content: Option = None; > + Ok(Some((ze, content))) > + } else if entry.file_type().is_symlink() { > + let target = std::fs::read_link(entry.path())?; > + let ze = > + ZipEntry::new(entry_path_no_base, mtime, mode, true, Some(target.as_ref())); > let content: Option = None; > Ok(Some((ze, content))) > } else {