ZIP Archive Format — Binary Internals Parser

ZIP is one of the most widely used archive formats in the world. Beyond .zip files themselves, many common formats are ZIP archives in disguise: Java .jar files, Android .apk packages, Microsoft Office .docx and .xlsx documents, and OpenDocument .odt files are all valid ZIP archives that Binary Internals can open and parse. The ZIPFile parser locates the End of Central Directory Record at the end of the file, uses it to find the Central Directory, and then maps each Central Directory entry back to its Local File Header at the front of the file. All three structures are exposed as annotated, interactive tree nodes.

File Extension

.zip — registered by the ZIP plugin. Any file opened with a .zip extension will be dispatched to ZIPFile.

Files with extensions .jar, .apk, .docx, .xlsx, .odt, and similar formats that use ZIP as their container can also be opened by renaming them to .zip or by using the HTTP API with ?name=file.zip.

ZIP File Structure Overview

A standard ZIP file has three main regions, processed in reverse order by the parser:

[Local File Header 1] [File Data 1] [Data Descriptor 1?]
[Local File Header 2] [File Data 2] [Data Descriptor 2?]
...
[Central Directory Structure 1]
[Central Directory Structure 2]
...
[End of Central Directory Record]

The parser begins at the end of the file, scans backward for the PK\x05\x06 signature (bytes 0x50 0x4B 0x05 0x06) to find the End of Central Directory Record, then jumps to the Central Directory offset, and finally seeks to each Local File Header using the RelativeOffsetOfLocalHeader stored in the Central Directory.

End of Central Directory Record

The End of Central Directory Record is always located at or near the end of the file. It provides the information needed to parse everything else:

Field	Size	Description
Signature	4 bytes	`PK\x05\x06` = `0x50 0x4B 0x05 0x06`
Disk number	2 bytes	Disk number of this record
Central dir start disk	2 bytes	Disk where central directory starts
Entries on this disk	2 bytes	Number of central directory entries on this disk
Total entries	2 bytes	Total entries in the central directory (`EntryTotalNumber`)
Central dir size	4 bytes	Size of the central directory in bytes
Central dir offset	4 bytes	Offset of central directory from start of first disk (`CentralDirectoryOffset`)
Comment length	2 bytes	Length of the archive comment
Comment	variable	Archive comment (optional)

The EntryTotalNumber and CentralDirectoryOffset fields drive all subsequent parsing. If EntryTotalNumber is zero, the file is a valid but empty archive and no further structures are parsed.

Central Directory Structure

One CentralDirectoryStructure entry exists per file in the archive. It contains all the metadata about the entry as well as a pointer back to the Local File Header.

FileHeader fields

Field	Size	Description
`Signature`	4 bytes	`PK\x01\x02` = `0x50 0x4B 0x01 0x02`
`VersionMadeBy`	2 bytes	Version and OS that created the entry
`VersionNeededToExtract`	2 bytes	Minimum version needed to extract
`GeneralPurposeBitFlag`	2 bytes	16-bit flags (encryption, data descriptor, UTF-8 name, etc.)
`CompressionMethod`	2 bytes	Compression algorithm (0 = stored, 8 = deflate, etc.)
`LastModFileTime`	2 bytes	Last modification time in MS-DOS format
`LastModFileDate`	2 bytes	Last modification date in MS-DOS format
`CRC32`	4 bytes	CRC-32 checksum of uncompressed data
`CompressedSize`	4 bytes	Size of the compressed file data
`UncompressedSize`	4 bytes	Size of the original uncompressed file
`FileNameLength`	2 bytes	Length of the filename field
`ExtraFieldLength`	2 bytes	Length of the extra field
`FileCommentLength`	2 bytes	Length of the per-file comment
`DiskNumberStart`	2 bytes	Disk number where the file starts
`InternalFileAttributes`	2 bytes	Internal attributes (e.g., text vs binary flag)
`ExternalFileAttributes`	4 bytes	External (OS-specific) attributes
`RelativeOffsetOfLocalHeader`	4 bytes	Byte offset from the start of the first disk to the Local File Header
`FileName`	variable	UTF-8 encoded filename
`ExtraField`	variable	Extension data (e.g., ZIP64, NTFS timestamps)
`FileComment`	variable	Per-file comment (optional)

Local File Header

Each file’s data is preceded by a LocalFileHeader immediately before the compressed bytes. The layout is similar to the Central Directory entry but without the VersionMadeBy, DiskNumberStart, per-file comment, and internal/external attributes fields.

Field	Size	Description
`Signature`	4 bytes	`PK\x03\x04` = `0x50 0x4B 0x03 0x04`
`VersionNeededToExtract`	2 bytes	Minimum version needed to extract
`GeneralPurposeBitFlag`	2 bytes	16-bit flags
`CompressionMethod`	2 bytes	Compression algorithm
`LastModFileTime`	2 bytes	MS-DOS time
`LastModFileDate`	2 bytes	MS-DOS date
`CRC32`	4 bytes	CRC-32 checksum
`CompressedSize`	4 bytes	Compressed size
`UncompressedSize`	4 bytes	Uncompressed size
`FileNameLength`	2 bytes	Filename length
`ExtraFieldLength`	2 bytes	Extra field length
`FileName`	variable	UTF-8 filename
`ExtraField`	variable	Extension data

The LocalFileHeader total length is 30 + FileNameLength + ExtraFieldLength bytes, followed immediately by CompressedSize bytes of compressed file data.

General Purpose Bit Flag

The 16-bit GeneralPurposeBitFlag is expanded in the tree into individual bit nodes. Both the LocalFileHeader and CentralDirectoryStructure.FileHeader expose this flag. Some notable bits:

Bit	Meaning
0	File is encrypted
3	Data descriptor follows the compressed data (sizes/CRC are zero in the local header)
6	Strong encryption
11	Filename and comment are encoded in UTF-8

Known Signatures

// From ZIPFile.java
LOCAL_FILE_HEADER      = { 0x50, 0x4B, 0x03, 0x04 }  // PK♥♦
CENTRAL_FILE_HEADER    = { 0x50, 0x4B, 0x01, 0x02 }  // PK☺☻
CENTRAL_END            = { 0x50, 0x4B, 0x05, 0x06 }  // PK♣♠ (EOCD)
DIGITAL_SIG_HEADER     = { 0x50, 0x4B, 0x05, 0x05 }
CENTRAL_ZIP64_RECORD   = { 0x50, 0x4B, 0x06, 0x06 }
CENTRAL_ZIP64_LOCATOR  = { 0x50, 0x4B, 0x06, 0x07 }
ARCHIVE_EXTRA_DATA     = { 0x50, 0x4B, 0x06, 0x08 }

The minimum valid ZIP file length is 22 bytes (the bare minimum for an empty archive containing only the End of Central Directory Record). Files shorter than 22 bytes are rejected with a FileFormatException.

Getting Started

Supported Formats

Web Server & API

Developer Guide

ZIP Archive Format — Binary Internals Parser

File Extension

ZIP File Structure Overview

End of Central Directory Record

Central Directory Structure

FileHeader fields

Local File Header

General Purpose Bit Flag

Known Signatures

​File Extension

​ZIP File Structure Overview

​End of Central Directory Record

​Central Directory Structure

​FileHeader fields

​Local File Header

​General Purpose Bit Flag

​Known Signatures

File Extension

ZIP File Structure Overview

End of Central Directory Record

Central Directory Structure

FileHeader fields

Local File Header

General Purpose Bit Flag

Known Signatures