Create StuffIt archives on Unix systems

Summary

sit is a command-line tool that can create compressed StuffIt 1.5.1-compatible archives. The archive may contain both files and folders. Resource forks and metadata are preserved, and the resulting archive file can be extracted with StuffIt or StuffIt Expander on a classic Mac OS system, as well as with The Unarchiver on a modern system. Run the tool in Terminal with no arguments to see its options.

Introduction

The sit program was originally written in 1988 and posted to Usenet's comp.sources.mac newsgroup by Tom Bereiter. The program assumed that it was running on a Unix system where each input file either had no resource information at all, or was split into 3 binary files named with .data, .rsrc, and .info extensions.

This modern update of sit takes into account the fact that macOS has been a Unix system since 2001, and that input files may have native resource forks. The program has been enhanced to accept either files or directories as input, and will archive the contents of directories recursively. Its primary purpose is to create a portable container that can be safely transferred from a modern system to a classic Mac computer or emulator.

Usage

sit [-v] [-u] [-T type] [-C creator] [-o dstfile] file ...

Creates a StuffIt 1.5.1-compatible archive from one or more files (or folders) specified as arguments. A combination of files and folders can be specified. The default output file is "archive.sit" if the -o option is not provided. Use -v, -vv, or -vvv to see increasingly verbose output.

Files without a resource fork are assigned the default type TEXT and creator KAHL, identifying them as a text file created by THINK C. You can override the default type and creator with the -T and -C options.

The -u option converts all linefeeds ('\n') to carriage returns ('\r') in the data fork of the file. This is really only useful when archiving plain Unix text files which you intend to open in a classic Mac application like SimpleText or MacWrite. In general, you should avoid this option, especially if you are archiving other types of documents or applications.

Examples

# create "archive.sit" containing three specified files
sit file1 file2 file3

# create "FolderArchive.sit" containing FolderToBeArchived
sit -o FolderArchive.sit FolderToBeArchived

# specify that untyped files are JPEG and open in GraphicConverter
sit -o jpgArchive.sit -T JPEG -C GKON *.jpg

Building

Build the tool by simply typing make while sit is the current directory.

This should build cleanly on any Unix system with developer tools installed. On macOS, you may be prompted to install Xcode's CLTools support when you first attempt to run make.

Limitations

Unlike StuffIt, this program does not currently offer a choice of compression algorithms to use. LZW is supported and used by default to compress archives. While LZW compression offers significant savings, Huffman compression was also supported by StuffIt 1.5.1, and may offer additional savings for some files. The possibility of adding Huffman encoding is being investigated for a future update.

This program is known to compile and run on macOS systems (Snow Leopard 10.6 or later). While it is intended that the software should be able to compile and run on any UNIX system, it currently may not obtain file creation dates properly on certain systems whose stat structure does not contain a st_birthtime field, as this is not yet part of a POSIX standard.

This software may contain bugs. Use at your own risk.

Linux and Cross-Platform Support

This version of sit has been enhanced to work on Linux and other Unix systems. On non-macOS platforms, resource forks and file metadata can be provided using the AppleDouble file format in both ._filename or .rsrc sidecars. They can also be provided with the .rsrc and .info files which are output by the xbin program.

Filter and reassemble BinHex files from Usenet

macbinfilt is a companion utility for processing Mac binary files posted to Usenet newsgroups, particularly comp.binaries.mac. In the 1980s and 1990s, Mac software was commonly distributed through Usenet in BinHex 4.0 format, often split across multiple article parts.

What it does:

• Filters Usenet article text to extract BinHex-encoded data
• Automatically reorders multi-part articles if parts arrive out of sequence
• Strips extraneous headers, signatures, and non-BinHex content
• Outputs clean BinHex data ready for decoding with xbin or similar tools

Usage:

# Process a single article or file
macbinfilt article.txt > output.hqx

# Process multiple parts (will reorder automatically)
macbinfilt part1.txt part3.txt part2.txt > complete.hqx

# Process from stdin (e.g., from a newsreader)
cat article.txt | macbinfilt > output.hqx

When to use it:

• You're working with historical Mac software archives from Usenet
• You need to extract BinHex data from articles with headers and footers
• You received multi-part articles out of order and need them reassembled
• You're processing comp.binaries.mac archives

Notes:

• macbinfilt looks for lines containing the text "part N of M" to identify multi-part files
• Missing parts will be reported as errors
• Output always begins with the BinHex signature: (This file must be converted with BinHex 4.0)
• The filtered output can then be decoded using BinHex decoders like xbin