TAR Archive & Extract

The TAR Format: From Tape to Cloud

TAR stands for Tape Archive, a format originally designed in 1979 for writing sequential backups to magnetic tape drives on Unix systems. Despite its age, TAR remains the dominant archiving format in Linux, macOS, and the open-source ecosystem. Every source code release on GitHub, every Docker image layer, and every Node.js package published to npm uses TAR as its packaging format.

The 512-Byte Block Architecture

TAR's design reflects its tape heritage. Every piece of data is aligned to 512-byte blocks — the natural sector size of magnetic tape. Each file in the archive begins with a 512-byte header block that contains the filename (100 bytes, null-terminated), file permissions in octal ASCII (8 bytes), owner and group IDs (8 bytes each), the file size in octal (12 bytes), a Unix timestamp (12 bytes), a checksum (8 bytes), and a type flag (1 byte). After the header, the file's raw data follows, padded with zero bytes to the next 512-byte boundary. The archive ends with two consecutive zero-filled blocks as an end-of-archive marker.

The ustar Standard (POSIX.1-1988)

The original TAR format limited filenames to 100 characters and had no standardized magic number. The ustar (Unix Standard TAR) extension, codified in POSIX.1-1988, added a 6-byte magic field at byte 257 (ustar) and a 155-byte prefix field at byte 345. By combining the prefix and name fields, ustar supports filenames up to 256 characters. This tool implements ustar for both reading and writing, ensuring compatibility with archives produced by GNU tar, BSD tar, and other modern implementations.

.tar.gz vs .tar: Archive Then Compress

TAR itself applies no compression — it simply concatenates files sequentially. Compression is handled by a separate tool, most commonly GZIP. The convention .tar.gz (or .tgz) indicates a TAR archive compressed with GZIP. This two-stage approach is architecturally superior to formats like ZIP, where each file is compressed individually. Because GZIP sees the entire concatenated stream, it can exploit redundancy across files — common headers in log files, shared boilerplate in source code — achieving significantly better compression ratios. The trade-off is that you cannot extract a single file without decompressing the entire stream, but for most use cases the improved compression outweighs this limitation.

The TAR Checksum Algorithm

Each TAR header includes an 8-byte checksum field at bytes 148-155. The checksum is computed by summing all 512 bytes of the header, with the checksum field itself treated as eight space characters (0x20). This produces an unsigned integer stored as a 6-digit octal string followed by a null byte and a space. The checksum provides basic integrity verification — it detects single-byte corruption but is not a cryptographic hash. For stronger integrity guarantees, pair TAR with a file checksum tool.

Browser-Native Implementation

This tool implements a custom TAR parser in pure JavaScript — no external libraries, no WebAssembly, no server-side processing. The parser reads the raw bytes of the archive using the File.arrayBuffer() API, walks through 512-byte blocks, validates checksums, decodes octal ASCII fields, and extracts file data into Blob objects. For .tar.gz archives, the browser's native DecompressionStream('gzip') handles decompression before the TAR parser processes the uncompressed stream.

When to Use TAR Archives

TAR is the standard archiving format across the Unix, Linux, and open-source ecosystem. Here are the most common scenarios where you will encounter or create TAR archives:

• Source code distribution — GitHub release tarballs, npm packages, Python sdist packages, and Ruby gems all use .tar.gz as their primary distribution format. When you download a release from GitHub, the "Source code (tar.gz)" link provides a TAR archive.
• Docker and container images — Every Docker image layer is stored as a TAR archive. When you run docker save, the output is a TAR file containing all image layers. Understanding TAR helps when debugging container builds or inspecting image contents.
• System backups and migrations — System administrators use tar to create full directory backups that preserve file permissions, ownership, and timestamps. A typical backup command like tar czf backup.tar.gz /home creates a compressed archive of the entire home directory.
• Log file bundling — Collecting log files from multiple servers or services into a single .tar.gz for analysis or archival. The cross-file compression of GZIP over TAR is especially effective with log files because they share common patterns.
• Data science and ML datasets — Large datasets (ImageNet, Common Crawl, LAION) are distributed as TAR archives because the format handles millions of files without the directory-entry overhead of ZIP. Streaming TAR extraction is also more memory-efficient than ZIP for large archives.
• Cross-platform file transfer — When transferring files between macOS, Linux, and Windows systems, TAR preserves Unix file permissions and symbolic links that ZIP may not handle correctly. This matters for development environments and configuration files.

TAR vs ZIP: When to Choose Each

Use TAR + GZIP when you need the best compression ratio, when working in Unix/Linux environments, or when preserving file metadata matters. Use ZIP when sharing with Windows users who may not have tar utilities, when you need to extract individual files without decompressing the entire archive, or when the recipient expects a self-contained compressed archive. Both formats are lossless and widely supported.

Privacy and Security

Your archives are processed entirely within your browser. No data is transmitted to any server — the TAR parser runs as client-side JavaScript, and GZIP decompression uses the browser's native DecompressionStream API. This makes the tool safe for working with proprietary source code, configuration files containing secrets, database exports, or any other sensitive content. You can verify this by opening your browser's Network tab — no requests are made during processing.