[pbs-devel] [PATCH v4 proxmox-backup 14/26] fix #3174: extractor: impl seq restore from appendix

[Christian Ebner <c.ebner@proxmox.com>]

Restores the file payloads for all AppendixRef entries encountered
during the sequential restore of the pxar archive.
This is done by iterating over all the files listed in the corresponding
state variable, opening each of the parent directory while storing its
metadata for successive restore and creating the file, followed by
writing the contents to it.

When leaving the directories, their metatdata is restored.

Signed-off-by: Christian Ebner <c.ebner@proxmox.com>
---
Changes since version 3:
- Use BTreeMap and sorted insert instead of Vec and sorting afterwards

Changes since version 2:
- Sort entries by their appendix start offset for restore. Required
  since chunks are now normalized during upload.

Changes since version 1:
- Use the Encoder<SeqSink> to get encoded metadata byte size

 pbs-client/src/pxar/create.rs  |   4 +-
 pbs-client/src/pxar/extract.rs | 149 +++++++++++++++++++++++++++++++--
 pbs-client/src/pxar/tools.rs   |   1 +
 3 files changed, 144 insertions(+), 10 deletions(-)

diff --git a/pbs-client/src/pxar/create.rs b/pbs-client/src/pxar/create.rs
index 611d7421..50bba4e6 100644
--- a/pbs-client/src/pxar/create.rs
+++ b/pbs-client/src/pxar/create.rs
@@ -43,7 +43,7 @@ pub struct PxarCreateOptions {
     pub skip_lost_and_found: bool,
 }
 
-fn detect_fs_type(fd: RawFd) -> Result<i64, Error> {
+pub fn detect_fs_type(fd: RawFd) -> Result<i64, Error> {
     let mut fs_stat = std::mem::MaybeUninit::uninit();
     let res = unsafe { libc::fstatfs(fd, fs_stat.as_mut_ptr()) };
     Errno::result(res)?;
@@ -776,7 +776,7 @@ impl Archiver {
     }
 }
 
-fn get_metadata(
+pub fn get_metadata(
     fd: RawFd,
     stat: &FileStat,
     flags: Flags,
diff --git a/pbs-client/src/pxar/extract.rs b/pbs-client/src/pxar/extract.rs
index d2d42749..7bdcb595 100644
--- a/pbs-client/src/pxar/extract.rs
+++ b/pbs-client/src/pxar/extract.rs
@@ -1,6 +1,6 @@
 //! Code for extraction of pxar contents onto the file system.
 
-use std::collections::HashMap;
+use std::collections::{BTreeMap, HashMap};
 use std::ffi::{CStr, CString, OsStr, OsString};
 use std::io;
 use std::os::unix::ffi::OsStrExt;
@@ -74,7 +74,7 @@ struct ExtractorIterState {
     err_path_stack: Vec<OsString>,
     current_match: bool,
     end_reached: bool,
-    appendix_list: Vec<(PathBuf, u64, u64)>,
+    appendix_refs: BTreeMap<u64, (PathBuf, u64)>,
 }
 
 /// An [`Iterator`] that encapsulates the process of extraction in [extract_archive].
@@ -99,7 +99,7 @@ impl ExtractorIterState {
             err_path_stack: Vec::new(),
             current_match: options.extract_match_default,
             end_reached: false,
-            appendix_list: Vec::new(),
+            appendix_refs: BTreeMap::new(),
         }
     }
 }
@@ -313,6 +313,141 @@ where
 
                 res
             }
+            (_, EntryKind::Appendix { total }) => {
+                // Bytes consumed in decoder since encountering the appendix marker
+                let mut consumed = 0;
+                for (offset, (path, size)) in &self.state.appendix_refs {
+                    self.extractor.allow_existing_dirs = true;
+
+                    // Open dir path components, skipping the root component, get metadata
+                    for dir in path.iter().skip(1) {
+                        let parent_fd = match self.extractor.dir_stack.last_dir_fd(true) {
+                            Ok(parent_fd) => parent_fd,
+                            Err(err) => return Some(Err(err.into())),
+                        };
+                        let fs_magic =
+                            match crate::pxar::create::detect_fs_type(parent_fd.as_raw_fd()) {
+                                Ok(fs_magic) => fs_magic,
+                                Err(err) => return Some(Err(err.into())),
+                            };
+
+                        let mut fs_feature_flags = Flags::from_magic(fs_magic);
+                        let file_name = match CString::new(dir.as_bytes()) {
+                            Ok(file_name) => file_name,
+                            Err(err) => return Some(Err(err.into())),
+                        };
+                        let fd = match proxmox_sys::fd::openat(
+                            &parent_fd,
+                            file_name.as_ref(),
+                            OFlag::O_NOATIME,
+                            Mode::empty(),
+                        ) {
+                            Ok(fd) => fd,
+                            Err(err) => return Some(Err(err.into())),
+                        };
+
+                        let stat = match nix::sys::stat::fstat(fd.as_raw_fd()) {
+                            Ok(stat) => stat,
+                            Err(err) => return Some(Err(err.into())),
+                        };
+                        let metadata = match crate::pxar::create::get_metadata(
+                            fd.as_raw_fd(),
+                            &stat,
+                            fs_feature_flags,
+                            fs_magic,
+                            &mut fs_feature_flags,
+                        ) {
+                            Ok(metadata) => metadata,
+                            Err(err) => return Some(Err(err)),
+                        };
+
+                        match self.extractor.enter_directory(
+                            dir.to_os_string(),
+                            metadata.clone(),
+                            true,
+                        ) {
+                            Ok(()) => (),
+                            Err(err) => return Some(Err(err)),
+                        };
+                    }
+
+                    let skip = *offset - consumed;
+                    match self.decoder.skip_bytes(skip) {
+                        Ok(()) => (),
+                        Err(err) => return Some(Err(err.into())),
+                    };
+
+                    let entry = match self.decoder.next() {
+                        Some(Ok(entry)) => entry,
+                        Some(Err(err)) => return Some(Err(err.into())),
+                        None => return Some(Err(format_err!("expected entry"))),
+                    };
+
+                    let file_name_os = entry.file_name();
+                    let file_name_bytes = file_name_os.as_bytes();
+
+                    let file_name = match CString::new(file_name_bytes) {
+                        Ok(file_name_ref) => file_name_ref,
+                        Err(err) => return Some(Err(err.into())),
+                    };
+
+                    let metadata = entry.metadata();
+
+                    self.extractor.set_path(path.as_os_str().to_owned());
+
+                    let contents = self.decoder.contents();
+                    match contents {
+                        None => {
+                            return Some(Err(format_err!(
+                                "found regular file entry without contents in archive"
+                            )))
+                        }
+                        Some(mut contents) => {
+                            let result = self
+                                .extractor
+                                .extract_file(
+                                    &file_name,
+                                    metadata,
+                                    *size,
+                                    &mut contents,
+                                    self.extractor
+                                        .overwrite_flags
+                                        .contains(OverwriteFlags::FILE),
+                                )
+                                .context(PxarExtractContext::ExtractFile);
+                            if let Err(err) = result {
+                                return Some(Err(err.into()));
+                            }
+                        }
+                    }
+
+                    // Iter over all dir path components, skipping the root component, set metadata
+                    for _dir in path.iter().skip(1) {
+                        if let Err(err) = self.extractor.leave_directory() {
+                            return Some(Err(err.into()));
+                        }
+                    }
+
+                    let mut bytes = match pxar::encoder::Encoder::<pxar::encoder::SeqSink>::byte_len(
+                        file_name.as_c_str(),
+                        &metadata,
+                    ) {
+                        Ok(bytes) => bytes,
+                        Err(err) => return Some(Err(err.into())),
+                    };
+                    // payload header size
+                    bytes += std::mem::size_of::<pxar::format::Header>() as u64;
+
+                    consumed += skip + bytes + *size;
+                }
+
+                let skip = *total - consumed;
+                if let Err(err) = self.decoder.skip_bytes(skip) {
+                    return Some(Err(err.into()));
+                }
+
+                Ok(())
+            }
             (true, EntryKind::Symlink(link)) => {
                 self.callback(entry.path());
                 self.extractor
@@ -382,11 +517,9 @@ where
                     file_size,
                 },
             ) => {
-                self.state.appendix_list.push((
-                    entry.path().to_path_buf(),
-                    *appendix_offset,
-                    *file_size,
-                ));
+                self.state
+                    .appendix_refs
+                    .insert(*appendix_offset, (entry.path().to_path_buf(), *file_size));
                 Ok(())
             }
             (false, _) => Ok(()), // skip this
diff --git a/pbs-client/src/pxar/tools.rs b/pbs-client/src/pxar/tools.rs
index aac5a1e7..174a7351 100644
--- a/pbs-client/src/pxar/tools.rs
+++ b/pbs-client/src/pxar/tools.rs
@@ -156,6 +156,7 @@ pub fn format_multi_line_entry(entry: &Entry) -> String {
 
     let (size, link, type_name) = match entry.kind() {
         EntryKind::File { size, .. } => (format!("{}", *size), String::new(), "file"),
+        EntryKind::Appendix { total } => (format!("{total}"), String::new(), "appendix"),
         EntryKind::AppendixRef {
             appendix_offset,
             file_size,
-- 
2.39.2