Merge branch 'libarchive' of https://github.com/yorickvP/nix

17 files changed, 129 insertions, 183 deletions

diff --git a/Makefile.config.in b/Makefile.config.in
index 7e3b35b98..fe609ce06 100644
--- a/Makefile.config.in
+++ b/Makefile.config.in
@@ -18,6 +18,7 @@ SODIUM_LIBS = @SODIUM_LIBS@
 LIBLZMA_LIBS = @LIBLZMA_LIBS@
 SQLITE3_LIBS = @SQLITE3_LIBS@
 LIBBROTLI_LIBS = @LIBBROTLI_LIBS@
+LIBARCHIVE_LIBS = @LIBARCHIVE_LIBS@
 EDITLINE_LIBS = @EDITLINE_LIBS@
 bash = @bash@
 bindir = @bindir@
diff --git a/configure.ac b/configure.ac
index 99248b90c..b868343d2 100644
--- a/configure.ac
+++ b/configure.ac
@@ -178,6 +178,8 @@ AC_CHECK_LIB([bz2], [BZ2_bzWriteOpen], [true],
   [AC_MSG_ERROR([Nix requires libbz2, which is part of bzip2.  See https://web.archive.org/web/20180624184756/http://www.bzip.org/.])])
 AC_CHECK_HEADERS([bzlib.h], [true],
   [AC_MSG_ERROR([Nix requires libbz2, which is part of bzip2.  See https://web.archive.org/web/20180624184756/http://www.bzip.org/.])])
+# Checks for libarchive
+PKG_CHECK_MODULES([LIBARCHIVE], [libarchive >= 3.1.2], [CXXFLAGS="$LIBARCHIVE_CFLAGS $CXXFLAGS"])
 
 # Look for SQLite, a required dependency.
 PKG_CHECK_MODULES([SQLITE3], [sqlite3 >= 3.6.19], [CXXFLAGS="$SQLITE3_CFLAGS $CXXFLAGS"])
diff --git a/nix-rust/Cargo.lock b/nix-rust/Cargo.lock
index 6e5ecf45d..957c01e5a 100644
--- a/nix-rust/Cargo.lock
+++ b/nix-rust/Cargo.lock
@@ -55,17 +55,6 @@ dependencies = [
 ]
 
 [[package]]
-name = "filetime"
-version = "0.2.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "cfg-if 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)",
- "libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)",
- "redox_syscall 0.1.56 (registry+https://github.com/rust-lang/crates.io-index)",
- "winapi 0.3.8 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
 name = "fnv"
 version = "1.0.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -109,7 +98,6 @@ dependencies = [
  "lazy_static 1.4.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)",
  "proptest 0.9.4 (registry+https://github.com/rust-lang/crates.io-index)",
- "tar 0.4.26 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
 [[package]]
@@ -320,17 +308,6 @@ dependencies = [
 ]
 
 [[package]]
-name = "tar"
-version = "0.4.26"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "filetime 0.2.8 (registry+https://github.com/rust-lang/crates.io-index)",
- "libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)",
- "redox_syscall 0.1.56 (registry+https://github.com/rust-lang/crates.io-index)",
- "xattr 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
 name = "tempfile"
 version = "3.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -375,14 +352,6 @@ name = "winapi-x86_64-pc-windows-gnu"
 version = "0.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
-[[package]]
-name = "xattr"
-version = "0.2.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
 [metadata]
 "checksum assert_matches 1.3.0 (registry+https://github.com/rust-lang/crates.io-index)" = "7deb0a829ca7bcfaf5da70b073a8d128619259a7be8216a355e23f00763059e5"
 "checksum autocfg 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)" = "1d49d90015b3c36167a20fe2810c5cd875ad504b39cff3d4eae7977e6b7c1cb2"
@@ -393,7 +362,6 @@ dependencies = [
 "checksum c2-chacha 0.2.3 (registry+https://github.com/rust-lang/crates.io-index)" = "214238caa1bf3a496ec3392968969cab8549f96ff30652c9e56885329315f6bb"
 "checksum cfg-if 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)" = "4785bdd1c96b2a846b2bd7cc02e86b6b3dbf14e7e53446c4f54c92a361040822"
 "checksum cloudabi 0.0.3 (registry+https://github.com/rust-lang/crates.io-index)" = "ddfc5b9aa5d4507acaf872de71051dfd0e309860e88966e1051e462a077aac4f"
-"checksum filetime 0.2.8 (registry+https://github.com/rust-lang/crates.io-index)" = "1ff6d4dab0aa0c8e6346d46052e93b13a16cf847b54ed357087c35011048cc7d"
 "checksum fnv 1.0.6 (registry+https://github.com/rust-lang/crates.io-index)" = "2fad85553e09a6f881f739c29f0b00b0f01357c743266d478b68951ce23285f3"
 "checksum fuchsia-cprng 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)" = "a06f77d526c1a601b7c4cdd98f54b5eaabffc14d5f2f0296febdc7f357c6d3ba"
 "checksum getrandom 0.1.13 (registry+https://github.com/rust-lang/crates.io-index)" = "e7db7ca94ed4cd01190ceee0d8a8052f08a247aa1b469a7f68c6a3b71afcf407"
@@ -423,11 +391,9 @@ dependencies = [
 "checksum regex-syntax 0.6.12 (registry+https://github.com/rust-lang/crates.io-index)" = "11a7e20d1cce64ef2fed88b66d347f88bd9babb82845b2b858f3edbf59a4f716"
 "checksum remove_dir_all 0.5.2 (registry+https://github.com/rust-lang/crates.io-index)" = "4a83fa3702a688b9359eccba92d153ac33fd2e8462f9e0e3fdf155239ea7792e"
 "checksum rusty-fork 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)" = "3dd93264e10c577503e926bd1430193eeb5d21b059148910082245309b424fae"
-"checksum tar 0.4.26 (registry+https://github.com/rust-lang/crates.io-index)" = "b3196bfbffbba3e57481b6ea32249fbaf590396a52505a2615adbb79d9d826d3"
 "checksum tempfile 3.1.0 (registry+https://github.com/rust-lang/crates.io-index)" = "7a6e24d9338a0a5be79593e2fa15a648add6138caa803e2d5bc782c371732ca9"
 "checksum wait-timeout 0.2.0 (registry+https://github.com/rust-lang/crates.io-index)" = "9f200f5b12eb75f8c1ed65abd4b2db8a6e1b138a20de009dacee265a2498f3f6"
 "checksum wasi 0.7.0 (registry+https://github.com/rust-lang/crates.io-index)" = "b89c3ce4ce14bdc6fb6beaf9ec7928ca331de5df7e5ea278375642a2f478570d"
 "checksum winapi 0.3.8 (registry+https://github.com/rust-lang/crates.io-index)" = "8093091eeb260906a183e6ae1abdba2ef5ef2257a21801128899c3fc699229c6"
 "checksum winapi-i686-pc-windows-gnu 0.4.0 (registry+https://github.com/rust-lang/crates.io-index)" = "ac3b87c63620426dd9b991e5ce0329eff545bccbbb34f3be09ff6fb6ab51b7b6"
 "checksum winapi-x86_64-pc-windows-gnu 0.4.0 (registry+https://github.com/rust-lang/crates.io-index)" = "712e227841d057c1ee1cd2fb22fa7e5a5461ae8e48fa2ca79ec42cfc1931183f"
-"checksum xattr 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)" = "244c3741f4240ef46274860397c7c74e50eb23624996930e484c16679633a54c"
diff --git a/nix-rust/Cargo.toml b/nix-rust/Cargo.toml
index cd388bed6..1372e5a73 100644
--- a/nix-rust/Cargo.toml
+++ b/nix-rust/Cargo.toml
@@ -9,7 +9,6 @@ name = "nixrust"
 crate-type = ["cdylib"]
 
 [dependencies]
-tar = "0.4"
 libc = "0.2"
 #futures-preview = { version = "=0.3.0-alpha.19" }
 #hyper = "0.13.0-alpha.4"
diff --git a/nix-rust/src/c.rs b/nix-rust/src/c.rs
index 8d2507d37..c1358545f 100644
--- a/nix-rust/src/c.rs
+++ b/nix-rust/src/c.rs
@@ -1,22 +1,4 @@
-use super::{
-    error,
-    foreign::{self},
-    store::path,
-    store::StorePath,
-    util,
-};
-
-#[no_mangle]
-pub unsafe extern "C" fn unpack_tarfile(
-    source: foreign::Source,
-    dest_dir: &str,
-    out: *mut Result<(), error::CppException>,
-) {
-    out.write(
-        util::tarfile::unpack_tarfile(source, std::path::Path::new(dest_dir))
-            .map_err(|err| err.into()),
-    );
-}
+use super::{error, store::path, store::StorePath, util};
 
 #[no_mangle]
 pub unsafe extern "C" fn ffi_String_new(s: &str, out: *mut String) {
diff --git a/nix-rust/src/foreign.rs b/nix-rust/src/foreign.rs
deleted file mode 100644
index 7bce7753c..000000000
--- a/nix-rust/src/foreign.rs
+++ /dev/null
@@ -1,14 +0,0 @@
-/// A wrapper around Nix's Source class that provides the Read trait.
-#[repr(C)]
-pub struct Source {
-    fun: extern "C" fn(this: *mut libc::c_void, data: &mut [u8]) -> usize,
-    this: *mut libc::c_void,
-}
-
-impl std::io::Read for Source {
-    fn read(&mut self, buf: &mut [u8]) -> std::result::Result<usize, std::io::Error> {
-        let n = (self.fun)(self.this, buf);
-        assert!(n <= buf.len());
-        Ok(n)
-    }
-}
diff --git a/nix-rust/src/lib.rs b/nix-rust/src/lib.rs
index 9935cd27a..e62613ba8 100644
--- a/nix-rust/src/lib.rs
+++ b/nix-rust/src/lib.rs
@@ -12,7 +12,6 @@ extern crate proptest;
 #[cfg(not(test))]
 mod c;
 mod error;
-mod foreign;
 #[cfg(unused)]
 mod nar;
 mod store;
diff --git a/nix-rust/src/util/mod.rs b/nix-rust/src/util/mod.rs
index cd852c55f..eaad9d406 100644
--- a/nix-rust/src/util/mod.rs
+++ b/nix-rust/src/util/mod.rs
@@ -1,2 +1 @@
 pub mod base32;
-pub mod tarfile;
diff --git a/nix-rust/src/util/tarfile.rs b/nix-rust/src/util/tarfile.rs
deleted file mode 100644
index 74d60692c..000000000
--- a/nix-rust/src/util/tarfile.rs
+++ /dev/null
@@ -1,56 +0,0 @@
-use crate::{foreign::Source, Error};
-use std::fs;
-use std::io;
-use std::os::unix::fs::OpenOptionsExt;
-use std::path::{Component, Path};
-use tar::Archive;
-
-pub fn unpack_tarfile(source: Source, dest_dir: &Path) -> Result<(), Error> {
-    fs::create_dir_all(dest_dir)?;
-
-    let mut tar = Archive::new(source);
-
-    for file in tar.entries()? {
-        let mut file = file?;
-
-        let path = file.path()?;
-
-        for i in path.components() {
-            if let Component::Prefix(_) | Component::RootDir | Component::ParentDir = i {
-                return Err(Error::BadTarFileMemberName(
-                    file.path()?.to_str().unwrap().to_string(),
-                ));
-            }
-        }
-
-        let dest_file = dest_dir.join(path);
-
-        fs::create_dir_all(dest_file.parent().unwrap())?;
-
-        match file.header().entry_type() {
-            tar::EntryType::Directory => {
-                fs::create_dir(dest_file)?;
-            }
-            tar::EntryType::Regular => {
-                let mode = if file.header().mode()? & (libc::S_IXUSR as u32) == 0 {
-                    0o666
-                } else {
-                    0o777
-                };
-                let mut f = fs::OpenOptions::new()
-                    .create(true)
-                    .write(true)
-                    .mode(mode)
-                    .open(dest_file)?;
-                io::copy(&mut file, &mut f)?;
-            }
-            tar::EntryType::Symlink => {
-                std::os::unix::fs::symlink(file.header().link_name()?.unwrap(), dest_file)?;
-            }
-            tar::EntryType::XGlobalHeader | tar::EntryType::XHeader => {}
-            t => return Err(Error::Misc(format!("unsupported tar entry type '{:?}'", t))),
-        }
-    }
-
-    Ok(())
-}
diff --git a/release-common.nix b/release-common.nix
index 3765e61b3..60c8849d1 100644
--- a/release-common.nix
+++ b/release-common.nix
@@ -49,6 +49,7 @@ rec {
     [ curl
       bzip2 xz brotli zlib editline
       openssl pkgconfig sqlite boehmgc
+      libarchive
       boost
       nlohmann_json
       rustc cargo
diff --git a/src/libstore/download.cc b/src/libstore/download.cc
index af058ba42..0c231ca1d 100644
--- a/src/libstore/download.cc
+++ b/src/libstore/download.cc
@@ -908,7 +908,7 @@ CachedDownloadResult Downloader::downloadCached(
             printInfo("unpacking '%s'...", url);
             Path tmpDir = createTempDir();
             AutoDelete autoDelete(tmpDir, true);
-            unpackTarfile(store->toRealPath(store->printStorePath(*storePath)), tmpDir, std::string(baseNameOf(url)));
+            unpackTarfile(store->toRealPath(store->printStorePath(*storePath)), tmpDir);
             auto members = readDirectory(tmpDir);
             if (members.size() != 1)
                 throw nix::Error("tarball '%s' contains an unexpected number of top-level files", url);
diff --git a/src/libutil/local.mk b/src/libutil/local.mk
index 35c1f6c13..16c1fa03f 100644
--- a/src/libutil/local.mk
+++ b/src/libutil/local.mk
@@ -6,6 +6,6 @@ libutil_DIR := $(d)
 
 libutil_SOURCES := $(wildcard $(d)/*.cc)
 
-libutil_LDFLAGS = $(LIBLZMA_LIBS) -lbz2 -pthread $(OPENSSL_LIBS) $(LIBBROTLI_LIBS) $(BOOST_LDFLAGS) -lboost_context
+libutil_LDFLAGS = $(LIBLZMA_LIBS) -lbz2 -pthread $(OPENSSL_LIBS) $(LIBBROTLI_LIBS) $(LIBARCHIVE_LIBS) $(BOOST_LDFLAGS) -lboost_context
 
 libutil_LIBS = libnixrust
diff --git a/src/libutil/rust-ffi.cc b/src/libutil/rust-ffi.cc
index 8b8b7b75d..6f36b3192 100644
--- a/src/libutil/rust-ffi.cc
+++ b/src/libutil/rust-ffi.cc
@@ -19,15 +19,4 @@ std::ostream & operator << (std::ostream & str, const String & s)
     return str;
 }
 
-size_t Source::sourceWrapper(void * _this, rust::Slice<uint8_t> data)
-{
-    try {
-        // FIXME: how to propagate exceptions?
-        auto n = ((nix::Source *) _this)->read((unsigned char *) data.ptr, data.size);
-        return n;
-    } catch (...) {
-        abort();
-    }
-}
-
 }
diff --git a/src/libutil/rust-ffi.hh b/src/libutil/rust-ffi.hh
index 3b51661c2..469a5fba3 100644
--- a/src/libutil/rust-ffi.hh
+++ b/src/libutil/rust-ffi.hh
@@ -131,18 +131,6 @@ struct String : Vec<char, ffi_String_drop>
 
 std::ostream & operator << (std::ostream & str, const String & s);
 
-struct Source
-{
-    size_t (*fun)(void * source_this, rust::Slice<uint8_t> data);
-    nix::Source * _this;
-
-    Source(nix::Source & _this)
-        : fun(sourceWrapper), _this(&_this)
-    {}
-
-    static size_t sourceWrapper(void * _this, rust::Slice<uint8_t> data);
-};
-
 /* C++ representation of Rust's Result<T, CppException>. */
 template<typename T>
 struct Result
diff --git a/src/libutil/tarfile.cc b/src/libutil/tarfile.cc
index 1be0ba24c..68e918891 100644
--- a/src/libutil/tarfile.cc
+++ b/src/libutil/tarfile.cc
@@ -1,38 +1,132 @@
-#include "rust-ffi.hh"
-#include "compression.hh"
+#include <archive.h>
+#include <archive_entry.h>
 
-extern "C" {
-    rust::Result<std::tuple<>> *
-    unpack_tarfile(rust::Source source, rust::StringSlice dest_dir, rust::Result<std::tuple<>> & out);
-}
+#include "serialise.hh"
 
 namespace nix {
 
+struct TarArchive {
+    struct archive *archive;
+    Source *source;
+    std::vector<unsigned char> buffer;
+
+    void check(int err, const char *reason = "Failed to extract archive (%s)") {
+        if (err == ARCHIVE_EOF)
+            throw EndOfFile("reached end of archive");
+        else if (err != ARCHIVE_OK)
+            throw Error(reason, archive_error_string(this->archive));
+    }
+
+    TarArchive(Source& source) : buffer(4096) {
+        this->archive = archive_read_new();
+        this->source = &source;
+
+        archive_read_support_filter_all(archive);
+        archive_read_support_format_all(archive);
+        check(archive_read_open(archive, (void *)this, TarArchive::callback_open, TarArchive::callback_read, TarArchive::callback_close), "Failed to open archive (%s)");
+    }
+
+    TarArchive(const Path &path) {
+        this->archive = archive_read_new();
+
+        archive_read_support_filter_all(archive);
+        archive_read_support_format_all(archive);
+        check(archive_read_open_filename(archive, path.c_str(), 16384), "Failed to open archive (%s)");
+    }
+
+    // disable copy constructor
+    TarArchive(const TarArchive&) = delete;
+
+    void close() {
+        check(archive_read_close(archive), "Failed to close archive (%s)");
+    }
+
+    ~TarArchive() {
+        if (this->archive) archive_read_free(this->archive);
+    }
+
+private:
+    static int callback_open(struct archive *, void *self) {
+        return ARCHIVE_OK;
+    }
+
+    static ssize_t callback_read(struct archive *archive, void *_self, const void **buffer) {
+        TarArchive *self = (TarArchive *)_self;
+        *buffer = self->buffer.data();
+
+        try {
+            return self->source->read(self->buffer.data(), 4096);
+        } catch (EndOfFile &) {
+            return 0;
+        } catch (std::exception &err) {
+            archive_set_error(archive, EIO, "Source threw exception: %s", err.what());
+
+            return -1;
+        }
+    }
+
+    static int callback_close(struct archive *, void *self) {
+        return ARCHIVE_OK;
+    }
+};
+
+struct PushD {
+    char * oldDir;
+
+    PushD(const std::string &newDir)  {
+        oldDir = getcwd(0, 0);
+        if (!oldDir) throw SysError("getcwd");
+        int r = chdir(newDir.c_str());
+        if (r != 0) throw SysError("changing directory to tar output path");
+    }
+
+    ~PushD() {
+        int r = chdir(oldDir);
+        free(oldDir);
+        if (r != 0)
+            std::cerr << "warning: failed to change directory back after tar extraction";
+        /* can't throw out of a destructor */
+    }
+};
+
+static void extract_archive(TarArchive &archive, const Path & destDir) {
+    // need to chdir back *after* archive closing
+    PushD newDir(destDir);
+    struct archive_entry *entry;
+    int flags = ARCHIVE_EXTRACT_FFLAGS
+        | ARCHIVE_EXTRACT_PERM
+        | ARCHIVE_EXTRACT_SECURE_SYMLINKS
+        | ARCHIVE_EXTRACT_SECURE_NODOTDOT
+        | ARCHIVE_EXTRACT_SECURE_NOABSOLUTEPATHS;
+
+    for(;;) {
+        int r = archive_read_next_header(archive.archive, &entry);
+        if (r == ARCHIVE_EOF) break;
+        else if (r == ARCHIVE_WARN)
+            std::cerr << "warning: " << archive_error_string(archive.archive) << std::endl;
+        else
+            archive.check(r);
+
+        archive.check(archive_read_extract(archive.archive, entry, flags));
+    }
+
+    archive.close();
+}
+
 void unpackTarfile(Source & source, const Path & destDir)
 {
-    rust::Source source2(source);
-    rust::Result<std::tuple<>> res;
-    unpack_tarfile(source2, destDir, res);
-    res.unwrap();
+    auto archive = TarArchive(source);
+
+    createDirs(destDir);
+    extract_archive(archive, destDir);
 }
 
-void unpackTarfile(const Path & tarFile, const Path & destDir,
-    std::optional<std::string> baseName)
+void unpackTarfile(const Path & tarFile, const Path & destDir)
 {
-    if (!baseName) baseName = std::string(baseNameOf(tarFile));
-
-    auto source = sinkToSource([&](Sink & sink) {
-        // FIXME: look at first few bytes to determine compression type.
-        auto decompressor =
-            hasSuffix(*baseName, ".bz2") ? makeDecompressionSink("bzip2", sink) :
-            hasSuffix(*baseName, ".gz") ? makeDecompressionSink("gzip", sink) :
-            hasSuffix(*baseName, ".xz") ? makeDecompressionSink("xz", sink) :
-            makeDecompressionSink("none", sink);
-        readFile(tarFile, *decompressor);
-        decompressor->finish();
-    });
-
-    unpackTarfile(*source, destDir);
+    auto archive = TarArchive(tarFile);
+
+    createDirs(destDir);
+    extract_archive(archive, destDir);
 }
 
 }
diff --git a/src/libutil/tarfile.hh b/src/libutil/tarfile.hh
index ce0911e2a..89a024f1d 100644
--- a/src/libutil/tarfile.hh
+++ b/src/libutil/tarfile.hh
@@ -4,7 +4,6 @@ namespace nix {
 
 void unpackTarfile(Source & source, const Path & destDir);
 
-void unpackTarfile(const Path & tarFile, const Path & destDir,
-    std::optional<std::string> baseName = {});
+void unpackTarfile(const Path & tarFile, const Path & destDir);
 
 }
diff --git a/src/nix-prefetch-url/nix-prefetch-url.cc b/src/nix-prefetch-url/nix-prefetch-url.cc
index 5040b6fb8..2ec0b07ac 100644
--- a/src/nix-prefetch-url/nix-prefetch-url.cc
+++ b/src/nix-prefetch-url/nix-prefetch-url.cc
@@ -190,10 +190,7 @@ static int _main(int argc, char * * argv)
                 printInfo("unpacking...");
                 Path unpacked = (Path) tmpDir + "/unpacked";
                 createDirs(unpacked);
-                if (hasSuffix(baseNameOf(uri), ".zip"))
-                    runProgram("unzip", true, {"-qq", tmpFile, "-d", unpacked});
-                else
-                    unpackTarfile(tmpFile, unpacked, std::string(baseNameOf(uri)));
+                unpackTarfile(tmpFile, unpacked);
 
                 /* If the archive unpacks to a single file/directory, then use
                    that as the top-level. */