web-image-meta

A lightweight Rust library for manipulating JPEG and PNG metadata, optimized for web images.

Features

• JPEG Support

  • Clean metadata while preserving orientation information
  • Read and write JPEG comments
  • Preserve ICC profiles
  • Remove EXIF, XMP, IPTC and other metadata

• PNG Support

  • Remove non-critical chunks
  • Read and write text chunks (tEXt)
  • Preserve transparency and color information

Installation

Add this to your Cargo.toml:

[dependencies]
web-image-meta = "0.1.0"

Usage

JPEG Examples

use web_image_meta::jpeg;

// Clean JPEG metadata while preserving orientation
let input_data = std::fs::read("input.jpg")?;
let cleaned_data = jpeg::clean_metadata(&input_data)?;
std::fs::write("cleaned.jpg", cleaned_data)?;

// Read JPEG comment
let comment = jpeg::read_comment(&input_data)?;
if let Some(text) = comment {
    println!("Comment: {}", text);
}

// Write JPEG comment
let data_with_comment = jpeg::write_comment(&input_data, "Copyright 2024")?;
std::fs::write("commented.jpg", data_with_comment)?;

PNG Examples

use web_image_meta::png;

// Remove non-critical chunks from PNG
let input_data = std::fs::read("input.png")?;
let cleaned_data = png::clean_chunks(&input_data)?;
std::fs::write("cleaned.png", cleaned_data)?;

// Read PNG text chunks
let chunks = png::read_text_chunks(&input_data)?;
for chunk in chunks {
    println!("{}: {}", chunk.keyword, chunk.text);
}

// Add text chunk to PNG
let data_with_text = png::add_text_chunk(
    &input_data,
    "Copyright",
    "© 2024 Example Corp"
)?;
std::fs::write("tagged.png", data_with_text)?;

API Reference

JPEG Functions

clean_metadata(data: &[u8]) -> Result<Vec<u8>, Error>

Removes all metadata except EXIF orientation information.

• Preserves: JFIF, ICC profiles, essential JPEG markers, EXIF orientation (tag 0x0112)
• Removes: All other EXIF data, XMP, IPTC, comments, APP markers (except APP0, APP1 with orientation, APP2 with ICC)
• Returns: Cleaned JPEG data

read_comment(data: &[u8]) -> Result<Option<String>, Error>

Reads the COM (comment) segment from a JPEG file.

• Returns: Some(String) if a comment exists, None otherwise
• Encoding: UTF-8 (lossy conversion for non-UTF-8 data)

write_comment(data: &[u8], comment: &str) -> Result<Vec<u8>, Error>

Writes or replaces a comment in a JPEG file.

• Replaces any existing comment
• Places comment before SOS marker
• Maximum length: 65,533 bytes

PNG Functions

clean_chunks(data: &[u8]) -> Result<Vec<u8>, Error>

Removes all non-critical chunks from a PNG file.

• Preserves: IHDR, PLTE, IDAT, IEND, tRNS, gAMA, cHRM, sRGB, iCCP, sBIT, pHYs
• Removes: tEXt, zTXt, iTXt, tIME, bKGD, and all other ancillary chunks
• Returns: Cleaned PNG data

read_text_chunks(data: &[u8]) -> Result<Vec<TextChunk>, Error>

Reads all tEXt chunks from a PNG file.

• Returns: Vector of TextChunk structs
• Only reads uncompressed tEXt chunks (not zTXt or iTXt)

add_text_chunk(data: &[u8], keyword: &str, text: &str) -> Result<Vec<u8>, Error>

Adds a new tEXt chunk to a PNG file.

• Keyword: 1-79 Latin characters (letters, numbers, spaces)
• Text: UTF-8 string of any length
• Places new chunk before IEND

Types

#[derive(Debug, Clone, PartialEq, Eq)]
pub struct TextChunk {
    pub keyword: String,  // 1-79 character Latin keyword
    pub text: String,     // Text content
}

#[derive(Debug)]
pub enum Error {
    InvalidFormat(String),  // Invalid image format
    Io(std::io::Error),    // I/O error
    ParseError(String),    // Parsing error
}

What Gets Preserved

JPEG

• Essential image data and structure
• EXIF Orientation (tag 0x0112) when present
• ICC color profiles (APP2)
• JFIF markers (APP0)
• All SOF markers (image encoding parameters)
• Huffman tables (DHT)
• Quantization tables (DQT)

PNG

• Critical chunks: IHDR, PLTE, IDAT, IEND
• Transparency: tRNS
• Color space: gAMA, cHRM, sRGB, iCCP, sBIT
• Physical dimensions: pHYs

What Gets Removed

JPEG

• EXIF data (except orientation)
• XMP metadata
• IPTC data
• Comments (when using clean_metadata)
• Photoshop resources (APP13)
• Other APP markers (APP3-APP15, except APP2 with ICC)

PNG

• Text chunks: tEXt, zTXt, iTXt
• Time chunks: tIME
• Background: bKGD
• Histogram: hIST
• Suggested palette: sPLT
• Other ancillary chunks

Error Handling

The library provides detailed error types:

• InvalidFormat: The input is not a valid JPEG/PNG file
• ParseError: The file structure is corrupted or invalid
• Io: System I/O errors

All functions validate their outputs to ensure the resulting images can be decoded.

Performance

This library is designed for web image optimization:

• Fast metadata stripping for reducing file sizes
• Preserves only essential information needed for proper display
• Memory-efficient processing
• Validates output to ensure images remain viewable

Safety

The library validates all inputs and outputs:

• Checks for valid JPEG/PNG signatures
• Validates chunk structures and CRCs (PNG)
• Ensures output images can be decoded
• Safe handling of malformed images

Test Coverage

The library includes comprehensive tests:

• 53 test cases covering various scenarios
• Tests for different image formats, color spaces, and edge cases
• Validation of output images using decoder libraries
• Tests run on Linux, macOS, and Windows

License

This project is licensed under the MIT License - see the LICENSE-MIT file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Acknowledgments

This library uses the following excellent crates:

• jpeg-decoder for JPEG validation
• png for PNG validation
• crc32fast for CRC calculation