add two failing tests and make them pass

This commit is contained in:
João Marcos 2025-05-05 23:52:49 -03:00
parent e15cb83bd4
commit f04c916f7f
6 changed files with 174 additions and 19 deletions

34
Cargo.lock generated
View File

@ -519,6 +519,12 @@ dependencies = [
"powerfmt",
]
[[package]]
name = "diff"
version = "0.1.13"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "56254986775e3233ffa9c4d7d3faaf6d36a2c09d30b20687e9f88bc8bafc16c8"
[[package]]
name = "difflib"
version = "0.4.0"
@ -818,6 +824,15 @@ version = "1.70.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7943c866cc5cd64cbc25b2e01621d07fa8eb2a1a23160ee81ce38704e97b8ecf"
[[package]]
name = "itertools"
version = "0.14.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2b192c782037fadd9cfa75548310488aabdbf3d2da73885b31bd0abd03351285"
dependencies = [
"either",
]
[[package]]
name = "jiff"
version = "0.2.8"
@ -1081,17 +1096,20 @@ dependencies = [
"filetime_creation",
"flate2",
"fs-err",
"glob",
"gzp",
"ignore",
"infer",
"insta",
"is_executable",
"itertools",
"libc",
"linked-hash-map",
"lz4_flex",
"num_cpus",
"once_cell",
"parse-display",
"pretty_assertions",
"proptest",
"rand 0.8.5",
"rayon",
@ -1246,6 +1264,16 @@ dependencies = [
"termtree",
]
[[package]]
name = "pretty_assertions"
version = "1.4.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3ae130e2f271fbc2ac3a40fb1d07180839cdbbe443c7a27e1e3c13c5cac0116d"
dependencies = [
"diff",
"yansi",
]
[[package]]
name = "proc-macro2"
version = "1.0.93"
@ -2047,6 +2075,12 @@ dependencies = [
"lzma-sys",
]
[[package]]
name = "yansi"
version = "1.0.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "cfe53a6657fd280eaa890a3bc59152892ffa3e30101319d168b781ed6529b049"
[[package]]
name = "zerocopy"
version = "0.7.35"

View File

@ -19,7 +19,7 @@ brotli = "7.0.0"
bstr = { version = "1.10.0", default-features = false, features = ["std"] }
bytesize = "1.3.0"
bzip2 = "0.4.4"
bzip3 = { version = "0.9.0", features = ["bundled"] , optional = true }
bzip3 = { version = "0.9.0", features = ["bundled"], optional = true }
clap = { version = "4.5.20", features = ["derive", "env"] }
filetime_creation = "0.2"
flate2 = { version = "1.0.30", default-features = false }
@ -58,9 +58,12 @@ clap_mangen = "0.2.24"
[dev-dependencies]
assert_cmd = "2.0.14"
glob = "0.3.2"
infer = "0.16.0"
insta = { version = "1.40.0", features = ["filters"] }
itertools = "0.14.0"
parse-display = "0.9.1"
pretty_assertions = "1.4.1"
proptest = "1.5.0"
rand = { version = "0.8.5", default-features = false, features = [
"small_rng",
@ -88,4 +91,5 @@ inherits = "release"
lto = false
opt-level = 2
incremental = true
codegen-units = 16
codegen-units = 32
strip = false

View File

@ -132,7 +132,7 @@ pub fn decompress_file(options: DecompressOptions) -> crate::Result<()> {
Snappy => Box::new(snap::read::FrameDecoder::new(decoder)),
Zstd => Box::new(zstd::stream::Decoder::new(decoder)?),
Brotli => Box::new(brotli::Decompressor::new(decoder, BUFFER_CAPACITY)),
Tar | Zip | Rar | SevenZip => unreachable!(),
Tar | Zip | Rar | SevenZip => decoder,
};
Ok(decoder)
};

View File

@ -36,7 +36,6 @@ pub fn list_archive_contents(
let zip_archive = zip::ZipArchive::new(reader)?;
let files = crate::archive::zip::list_archive(zip_archive, password);
list::list_files(archive_path, files, list_options)?;
return Ok(());
}
@ -46,7 +45,7 @@ pub fn list_archive_contents(
// Grab previous decoder and wrap it inside of a new one
let chain_reader_decoder =
|format: &CompressionFormat, decoder: Box<dyn Read + Send>| -> crate::Result<Box<dyn Read + Send>> {
|format: CompressionFormat, decoder: Box<dyn Read + Send>| -> crate::Result<Box<dyn Read + Send>> {
let decoder: Box<dyn Read + Send> = match format {
Gzip => Box::new(flate2::read::GzDecoder::new(decoder)),
Bzip => Box::new(bzip2::read::BzDecoder::new(decoder)),
@ -62,16 +61,22 @@ pub fn list_archive_contents(
Snappy => Box::new(snap::read::FrameDecoder::new(decoder)),
Zstd => Box::new(zstd::stream::Decoder::new(decoder)?),
Brotli => Box::new(brotli::Decompressor::new(decoder, BUFFER_CAPACITY)),
Tar | Zip | Rar | SevenZip => unreachable!(),
Tar | Zip | Rar | SevenZip => unreachable!("should be treated by caller"),
};
Ok(decoder)
};
for format in formats.iter().skip(1).rev() {
let mut misplaced_archive_format = None;
for &format in formats.iter().skip(1).rev() {
if format.archive_format() {
misplaced_archive_format = Some(format);
break;
}
reader = chain_reader_decoder(format, reader)?;
}
let files: Box<dyn Iterator<Item = crate::Result<FileInArchive>>> = match formats[0] {
let archive_format = misplaced_archive_format.unwrap_or(formats[0]);
let files: Box<dyn Iterator<Item = crate::Result<FileInArchive>>> = match archive_format {
Tar => Box::new(crate::archive::tar::list_archive(tar::Archive::new(reader))),
Zip => {
if formats.len() > 1 {
@ -120,7 +125,7 @@ pub fn list_archive_contents(
Box::new(archive::sevenz::list_archive(archive_path, password)?)
}
Gzip | Bzip | Bzip3 | Lz4 | Lzma | Snappy | Zstd | Brotli => {
panic!("Not an archive! This should never happen, if it does, something is wrong with `CompressionFormat::is_archive()`. Please report this error!");
unreachable!("Not an archive, should be validated before calling this function.");
}
};

View File

@ -60,8 +60,8 @@ impl Extension {
/// Checks if the first format in `compression_formats` is an archive
pub fn is_archive(&self) -> bool {
// Safety: we check that `compression_formats` is not empty in `Self::new`
self.compression_formats[0].is_archive_format()
// Index Safety: we check that `compression_formats` is not empty in `Self::new`
self.compression_formats[0].archive_format()
}
}
@ -103,7 +103,7 @@ pub enum CompressionFormat {
impl CompressionFormat {
/// Currently supported archive formats are .tar (and aliases to it) and .zip
fn is_archive_format(&self) -> bool {
pub fn archive_format(&self) -> bool {
// Keep this match like that without a wildcard `_` so we don't forget to update it
match self {
Tar | Zip | Rar | SevenZip => true,
@ -147,14 +147,13 @@ fn to_extension(ext: &[u8]) -> Option<Extension> {
))
}
fn split_extension(name: &mut &[u8]) -> Option<Extension> {
fn split_extension_at_end(name: &[u8]) -> Option<(&[u8], Extension)> {
let (new_name, ext) = name.rsplit_once_str(b".")?;
if matches!(new_name, b"" | b"." | b"..") {
return None;
}
let ext = to_extension(ext)?;
*name = new_name;
Some(ext)
Some((new_name, ext))
}
pub fn parse_format_flag(input: &OsStr) -> crate::Result<Vec<Extension>> {
@ -188,7 +187,7 @@ pub fn parse_format_flag(input: &OsStr) -> crate::Result<Vec<Extension>> {
/// Extracts extensions from a path.
///
/// Returns both the remaining path and the list of extension objects
/// Returns both the remaining path and the list of extension objects.
pub fn separate_known_extensions_from_name(path: &Path) -> (&Path, Vec<Extension>) {
let mut extensions = vec![];
@ -196,9 +195,12 @@ pub fn separate_known_extensions_from_name(path: &Path) -> (&Path, Vec<Extension
return (path, extensions);
};
// While there is known extensions at the tail, grab them
while let Some(extension) = split_extension(&mut name) {
while let Some((new_name, extension)) = split_extension_at_end(&name) {
name = new_name;
extensions.insert(0, extension);
if extensions[0].is_archive() {
break;
}
}
if let Ok(name) = name.to_str() {
@ -365,4 +367,19 @@ mod tests {
"linux.pkg.info.tar.zst"
);
}
#[test]
fn test_extension_parsing_with_multiple_archive_formats() {
assert_eq!(
separate_known_extensions_from_name("file.tar.zip".as_ref()),
("file.tar".as_ref(), vec![Extension::new(&[Zip], "zip")])
);
assert_eq!(
separate_known_extensions_from_name("file.7z.zst.zip.lz4".as_ref()),
(
"file.7z.zst".as_ref(),
vec![Extension::new(&[Zip], "zip"), Extension::new(&[Lz4], "lz4")]
)
);
}
}

View File

@ -8,7 +8,9 @@ use std::{
};
use fs_err as fs;
use itertools::Itertools;
use parse_display::Display;
use pretty_assertions::assert_eq;
use proptest::sample::size_range;
use rand::{rngs::SmallRng, Rng, SeedableRng};
use tempfile::tempdir;
@ -17,7 +19,7 @@ use test_strategy::{proptest, Arbitrary};
use crate::utils::{assert_same_directory, write_random_content};
/// tar and zip extensions
#[derive(Arbitrary, Debug, Display)]
#[derive(Arbitrary, Clone, Copy, Debug, Display)]
#[display(style = "lowercase")]
enum DirectoryExtension {
#[display("7z")]
@ -865,3 +867,96 @@ fn unpack_multiple_sources_into_the_same_destination_with_merge(
assert_eq!(5, out_path.as_path().read_dir()?.count());
}
#[test]
fn reading_nested_archives_with_two_archive_extensions_adjacent() {
let archive_formats = ["tar", "zip", "7z"].into_iter();
for (first_archive, second_archive) in archive_formats.clone().cartesian_product(archive_formats.rev()) {
let temp_dir = tempdir().unwrap();
let in_dir = |path: &str| format!("{}/{}", temp_dir.path().display(), path);
fs::write(in_dir("a.txt"), "contents").unwrap();
let files = [
"a.txt",
&format!("b.{first_archive}"),
&format!("c.{first_archive}.{second_archive}"),
];
let transformations = [first_archive, second_archive];
let compressed_path = in_dir(files.last().unwrap());
for (window, format) in files.windows(2).zip(transformations.iter()) {
let [a, b] = [window[0], window[1]].map(in_dir);
crate::utils::cargo_bin()
.args(["compress", &a, &b, "--format", format])
.assert()
.success();
}
crate::utils::cargo_bin()
.args(["list", &compressed_path, "--yes"])
.assert()
.success();
crate::utils::cargo_bin()
.args(["decompress", &compressed_path, "--dir", &in_dir("out"), "--yes"])
.assert()
.success();
let decompressed_files = glob::glob(&format!("{}/*", in_dir("out")))
.unwrap()
.collect::<Result<Vec<_>, _>>()
.unwrap();
let decompression_expected = format!("{}/out/{}", temp_dir.path().display(), files[1]);
assert_eq!(decompressed_files, [Path::new(&decompression_expected)]);
}
}
#[test]
fn reading_nested_archives_with_two_archive_extensions_interleaved() {
let archive_formats = ["tar", "zip", "7z"].into_iter();
for (first_archive, second_archive) in archive_formats.clone().cartesian_product(archive_formats.rev()) {
let temp_dir = tempdir().unwrap();
let in_dir = |path: &str| format!("{}/{}", temp_dir.path().display(), path);
fs::write(in_dir("a.txt"), "contents").unwrap();
let files = [
"a.txt",
&format!("c.{first_archive}"),
&format!("d.{first_archive}.zst"),
&format!("e.{first_archive}.zst.{second_archive}"),
&format!("f.{first_archive}.zst.{second_archive}.lz4"),
];
let transformations = [first_archive, "zst", second_archive, "lz4"];
let compressed_path = in_dir(files.last().unwrap());
for (window, format) in files.windows(2).zip(transformations.iter()) {
let [a, b] = [window[0], window[1]].map(in_dir);
crate::utils::cargo_bin()
.args(["compress", &a, &b, "--format", format])
.assert()
.success();
}
// // TODO: uncomment after fixing the 7z BadSignature error [4, 34, 77, 24, 96, 64]
// crate::utils::cargo_bin()
// .args(["list", &compressed_path, "--yes"])
// .assert()
// .success();
crate::utils::cargo_bin()
.args(["decompress", &compressed_path, "--dir", &in_dir("out"), "--yes"])
.assert()
.success();
let decompressed_files = glob::glob(&format!("{}/*", in_dir("out")))
.unwrap()
.collect::<Result<Vec<_>, _>>()
.unwrap();
let decompression_expected = format!("{}/out/{}", temp_dir.path().display(), files[2]);
assert_eq!(decompressed_files, [Path::new(&decompression_expected)]);
}
}