splitzip

Create split ZIP archives compatible with standard tools.

No 7-Zip required. Archives created with splitzip can be extracted using Windows Explorer, WinZip, macOS Archive Utility, and standard unzip on Linux.

Features

• Pure Python, no external dependencies
• Compatible with standard ZIP tools (no proprietary formats)
• Human-friendly size specifications ("100MB", "700MiB", "4.7GB")
• Progress callbacks for large files
• Context manager support
• CLI tool included

Installation

pip install splitzip

Quick Start

Simple Usage

import splitzip

# Create a split archive from files
splitzip.create(
    "backup.zip",
    ["documents/", "photos/", "important.pdf"],
    split_size="650MB"
)

Output files: backup.z01, backup.z02, ..., backup.zip

Context Manager

from splitzip import SplitZipWriter

with SplitZipWriter("backup.zip", split_size="100MB") as zf:
    # Add files
    zf.write("document.pdf")
    zf.write("photos/", recursive=True)
    
    # Add with custom name
    zf.write("secret.txt", arcname="data/renamed.txt")
    
    # Add content directly
    zf.writestr("hello.txt", b"Hello, World!")
    zf.writestr("config.json", '{"key": "value"}')

Advanced Options

from splitzip import SplitZipWriter, STORED, DEFLATED

def on_progress(filename, bytes_done, total_bytes):
    pct = (bytes_done / total_bytes) * 100
    print(f"\r{filename}: {pct:.1f}%", end="")

def on_volume(volume_num, path):
    print(f"Created volume: {path}")

with SplitZipWriter(
    "backup.zip",
    split_size="700MiB",          # DVD size
    compression=DEFLATED,          # or STORED for no compression
    compresslevel=9,               # 1-9 (default: 6)
    on_volume=on_volume,           # Volume creation callback
    on_progress=on_progress,       # Progress callback
) as zf:
    zf.write("large_file.bin")

Streaming from File Objects

import io
from splitzip import SplitZipWriter

data = get_data_from_somewhere()

with SplitZipWriter("archive.zip", split_size="100MB") as zf:
    zf.write_fileobj(
        io.BytesIO(data),
        arcname="streamed.bin",
        size=len(data)  # Optional, enables progress callback
    )

Size Specifications

splitzip accepts sizes in multiple formats:

Format	Example	Bytes
Integer	104857600	104,857,600
Bytes	"100B"	100
Kilobytes (decimal)	"100KB"	100,000
Megabytes (decimal)	"100MB"	100,000,000
Gigabytes (decimal)	"1GB"	1,000,000,000
Kibibytes (binary)	"100KiB"	102,400
Mebibytes (binary)	"700MiB"	734,003,200
Gibibytes (binary)	"1GiB"	1,073,741,824

Common split sizes:

• CD-ROM: "650MB" or "700MB"
• FAT32 limit: "4GiB" (minus 1 byte)
• Email attachment: "25MB"

Command Line Interface

# Create a split archive
splitzip create -o backup.zip -s 100MB file1.txt directory/

# With options
splitzip create -o backup.zip -s 700MiB \
    --level 9 \           # Max compression
    --verbose \           # Show progress
    documents/ photos/

# Store without compression
splitzip create -o backup.zip -s 100MB --store largefile.bin

Output File Naming

splitzip follows the standard ZIP split archive convention:

backup.z01  (first volume)
backup.z02  (second volume)
backup.z03  (third volume)
...
backup.zip  (final volume, contains central directory)

All files must be present in the same directory for extraction.

Compatibility

Tool	Single Volume	Split Archive
Windows Explorer	✅	✅
WinZip	✅	✅
7-Zip	✅	✅
macOS Archive Utility	✅	✅*
Linux unzip	✅	✅**
Python zipfile	✅	❌

* May require all files to be selected and opened together
** May require unzip -F flag for split archives

API Reference

SplitZipWriter(path, split_size, compression=DEFLATED, compresslevel=6, on_volume=None, on_progress=None)

• path (str | Path): Path for the final .zip file.
• split_size (int | str): Maximum size per volume (bytes or human-readable string).
• compression (int): DEFLATED (default) or STORED.
• compresslevel (int): DEFLATE level 1–9 (default 6).
• on_volume ((int, Path) -> None | None): Called when a volume is created.
• on_progress ((str, int, int) -> None | None): Called with (filename, bytes_done, total_bytes).

Methods

• write(path, arcname=None, recursive=True, compression=None, compresslevel=None) — Add a file or directory. Set recursive=False to add only the directory entry without contents. Symlinks are skipped.
• writestr(arcname, data, compression=None, compresslevel=None) — Write bytes/str directly.
• write_fileobj(fileobj, arcname, size=None, compression=None, compresslevel=None) — Write from a file-like object.
• close() — Finalize the archive. Returns list of volume paths.

Exceptions

Exception	When raised
SplitZipError	Base exception for all splitzip errors
VolumeError	Volume management errors
VolumeTooSmallError	Split size too small for headers
UnsafePathError	Path traversal detected (zip slip)
CompressionError	Compression failure
IntegrityError	CRC mismatch

Limitations

• Minimum volume size: 64KB
• No ZIP64 support: Individual files must be under 4GB, total entries under 65,535
• No encryption: Use filesystem encryption for sensitive data
• No reading/extraction: This is a write-only library (use standard tools to extract)
• No ZSTD/LZMA: Only DEFLATE and STORED compression (for compatibility)
• Symlinks are skipped: Symbolic links are ignored with a warning

Development

# Clone and install in development mode
git clone https://github.com/twwat/splitzip
cd splitzip
pip install -e ".[dev]"

# Run tests
pytest

# Run with coverage
pytest --cov=splitzip

# Type checking
mypy src/splitzip

# Linting
ruff check src/splitzip

License

MIT License. See LICENSE for details.

Contributing

Contributions welcome! Please open an issue to discuss major changes before submitting a PR.