clean up

fix/src/fetcher/archive.rs

@@ -0,0 +1,173 @@
+use std::fs;
+use std::io::Cursor;
+use std::os::unix::ffi::OsStrExt;
+use std::path::{Path, PathBuf};
+
+use flate2::read::GzDecoder;
+
+#[derive(Debug, Clone, Copy)]
+pub enum ArchiveFormat {
+    TarGz,
+    TarXz,
+    TarBz2,
+    Tar,
+}
+
+impl ArchiveFormat {
+    pub fn detect(url: &str, data: &[u8]) -> Self {
+        if url.ends_with(".tar.gz") || url.ends_with(".tgz") {
+            return ArchiveFormat::TarGz;
+        }
+        if url.ends_with(".tar.xz") || url.ends_with(".txz") {
+            return ArchiveFormat::TarXz;
+        }
+        if url.ends_with(".tar.bz2") || url.ends_with(".tbz2") {
+            return ArchiveFormat::TarBz2;
+        }
+        if url.ends_with(".tar") {
+            return ArchiveFormat::Tar;
+        }
+
+        if data.len() >= 2 && data[0] == 0x1f && data[1] == 0x8b {
+            return ArchiveFormat::TarGz;
+        }
+        if data.len() >= 6 && &data[0..6] == b"\xfd7zXZ\x00" {
+            return ArchiveFormat::TarXz;
+        }
+        if data.len() >= 3 && &data[0..3] == b"BZh" {
+            return ArchiveFormat::TarBz2;
+        }
+
+        ArchiveFormat::TarGz
+    }
+}
+
+pub fn extract_tarball(data: &[u8], dest: &Path) -> Result<PathBuf, ArchiveError> {
+    let format = ArchiveFormat::detect("", data);
+
+    let temp_dir = dest.join("_extract_temp");
+    fs::create_dir_all(&temp_dir)?;
+
+    match format {
+        ArchiveFormat::TarGz => extract_tar_gz(data, &temp_dir)?,
+        ArchiveFormat::TarXz => extract_tar_xz(data, &temp_dir)?,
+        ArchiveFormat::TarBz2 => extract_tar_bz2(data, &temp_dir)?,
+        ArchiveFormat::Tar => extract_tar(data, &temp_dir)?,
+    }
+
+    strip_single_toplevel(&temp_dir, dest)
+}
+
+fn extract_tar_gz(data: &[u8], dest: &Path) -> Result<(), ArchiveError> {
+    let decoder = GzDecoder::new(Cursor::new(data));
+    let mut archive = tar::Archive::new(decoder);
+    archive.unpack(dest)?;
+    Ok(())
+}
+
+fn extract_tar_xz(data: &[u8], dest: &Path) -> Result<(), ArchiveError> {
+    let decoder = xz2::read::XzDecoder::new(Cursor::new(data));
+    let mut archive = tar::Archive::new(decoder);
+    archive.unpack(dest)?;
+    Ok(())
+}
+
+fn extract_tar_bz2(data: &[u8], dest: &Path) -> Result<(), ArchiveError> {
+    let decoder = bzip2::read::BzDecoder::new(Cursor::new(data));
+    let mut archive = tar::Archive::new(decoder);
+    archive.unpack(dest)?;
+    Ok(())
+}
+
+fn extract_tar(data: &[u8], dest: &Path) -> Result<(), ArchiveError> {
+    let mut archive = tar::Archive::new(Cursor::new(data));
+    archive.unpack(dest)?;
+    Ok(())
+}
+
+fn strip_single_toplevel(temp_dir: &Path, dest: &Path) -> Result<PathBuf, ArchiveError> {
+    let entries: Vec<_> = fs::read_dir(temp_dir)?
+        .filter_map(|e| e.ok())
+        .filter(|e| e.file_name().as_os_str().as_bytes()[0] != b'.')
+        .collect();
+
+    let source_dir = if entries.len() == 1 && entries[0].file_type()?.is_dir() {
+        entries[0].path()
+    } else {
+        temp_dir.to_path_buf()
+    };
+
+    let final_dest = dest.join("content");
+    if final_dest.exists() {
+        fs::remove_dir_all(&final_dest)?;
+    }
+
+    if source_dir == *temp_dir {
+        fs::rename(temp_dir, &final_dest)?;
+    } else {
+        copy_dir_recursive(&source_dir, &final_dest)?;
+        fs::remove_dir_all(temp_dir)?;
+    }
+
+    Ok(final_dest)
+}
+
+fn copy_dir_recursive(src: &Path, dst: &Path) -> Result<(), std::io::Error> {
+    fs::create_dir_all(dst)?;
+
+    for entry in fs::read_dir(src)? {
+        let entry = entry?;
+        let path = entry.path();
+        let dest_path = dst.join(entry.file_name());
+        let metadata = fs::symlink_metadata(&path)?;
+
+        if metadata.is_symlink() {
+            let target = fs::read_link(&path)?;
+            #[cfg(unix)]
+            {
+                std::os::unix::fs::symlink(&target, &dest_path)?;
+            }
+            #[cfg(windows)]
+            {
+                if target.is_dir() {
+                    std::os::windows::fs::symlink_dir(&target, &dest_path)?;
+                } else {
+                    std::os::windows::fs::symlink_file(&target, &dest_path)?;
+                }
+            }
+        } else if metadata.is_dir() {
+            copy_dir_recursive(&path, &dest_path)?;
+        } else {
+            fs::copy(&path, &dest_path)?;
+        }
+    }
+
+    Ok(())
+}
+
+pub fn extract_tarball_to_temp(data: &[u8]) -> Result<(PathBuf, tempfile::TempDir), ArchiveError> {
+    let temp_dir = tempfile::tempdir()?;
+    let extracted_path = extract_tarball(data, temp_dir.path())?;
+    Ok((extracted_path, temp_dir))
+}
+
+#[derive(Debug)]
+pub enum ArchiveError {
+    IoError(std::io::Error),
+}
+
+impl std::fmt::Display for ArchiveError {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            ArchiveError::IoError(e) => write!(f, "I/O error: {}", e),
+        }
+    }
+}
+
+impl std::error::Error for ArchiveError {}
+
+impl From<std::io::Error> for ArchiveError {
+    fn from(e: std::io::Error) -> Self {
+        ArchiveError::IoError(e)
+    }
+}