Python XMP Toolkit

2.1.0 · active · verified Fri Apr 17

Python XMP Toolkit (version 2.1.0) provides Python bindings for Adobe's XMP Toolkit SDK by wrapping the Exempi C++ library. It enables reading, writing, and manipulating XMP (Extensible Metadata Platform) metadata in various file formats (e.g., JPEG, TIFF, PDF). The library maintains an active release cadence, with recent updates focusing on stability and compatibility with newer Exempi versions.

Common errors

```
ImportError: cannot import name 'XMPFiles' from 'xmp_toolkit'
```
cause Attempting to import from the old `xmp_toolkit` module after upgrading to version 2.0.0 or later.

fix Change your import statements from `from xmp_toolkit import ...` to `from libxmp import ...`.
```
libxmp.exempi.ExempiError: Exempi library not found.
```
cause The underlying C++ Exempi library is not installed or not discoverable in your system's library paths.

fix Install Exempi on your system. For Debian/Ubuntu: `sudo apt-get install libexempi-dev`. For macOS (Homebrew): `brew install exempi`. Ensure your system's dynamic linker can find the library.
```
libxmp.exempi.ExempiError: Could not open file (no such file or directory)
```
cause The file specified for `XMPFiles` does not exist, or the provided path is incorrect, or there are insufficient permissions to access it.

fix Verify the file path is correct and the file exists. Check file permissions for read/write access depending on your operation. Use absolute paths where possible.

Warnings

breaking Version 2.0.0 introduced significant breaking changes, primarily renaming the main module from `xmp_toolkit` to `libxmp` and overhauling the API. Code written for 1.x will not work directly with 2.x.
Fix: Update all imports from `xmp_toolkit` to `libxmp`. Review the new API for `XMPFiles`, `XMPMeta`, and constants (now `consts`) to adapt method calls.
gotcha This Python library requires the Exempi C++ library to be installed on your system. It is not bundled and will cause runtime errors if missing.
Fix: Install `libexempi` (or `libexempi-dev`) via your system's package manager (e.g., `sudo apt-get install libexempi-dev` on Debian/Ubuntu, `brew install exempi` on macOS) before using `python-xmp-toolkit`.
gotcha When working with `XMPFiles`, ensure the target file path exists and that your application has the necessary read/write permissions. Opening files for update (e.g., `open_for_update=True`) requires write access.
Fix: Always use `try-except` blocks to handle `libxmp.exempi.ExempiError` when dealing with file operations. Verify file paths and permissions before attempting to open or modify XMP data.

Install

pip install python-xmp-toolkit Install package

Imports

XMPFiles
wrong:
```
from xmp_toolkit import XMPFiles
```
correct:
```
from libxmp import XMPFiles
```
The primary module name changed from `xmp_toolkit` to `libxmp` in version 2.0.0.
XMPMeta
wrong:
```
from xmp_toolkit import XMPMeta
```
correct:
```
from libxmp import XMPMeta
```
The primary module name changed from `xmp_toolkit` to `libxmp` in version 2.0.0.
consts
wrong:
```
from xmp_toolkit import XMPConsts
```
correct:
```
from libxmp import consts
```
The constants object was renamed from `XMPConsts` to `consts` and moved to the `libxmp` module in version 2.0.0.

Quickstart

This quickstart demonstrates how to create, manipulate, and serialize XMP metadata using the `XMPMeta` class, which handles XMP data in memory. It covers setting properties, adding array items, retrieving values, and converting the XMP object to an RDF/XML string.

from libxmp import XMPMeta, consts

# Create a new XMP metadata object
xmp = XMPMeta()

# Set properties in the Dublin Core namespace
xmp.set_property(consts.NS_DC, "title", "My Awesome Photo")
xmp.set_property(consts.NS_DC, "creator", "Jane Doe")

# Add items to an array property (e.g., keywords)
xmp.append_array_item(consts.NS_DC, "subject", "landscape", None, consts.XMP_ARRAY_LAST_ITEM)
xmp.append_array_item(consts.NS_DC, "subject", "mountains", None, consts.XMP_ARRAY_LAST_ITEM)

# Get a property
title = xmp.get_property(consts.NS_DC, "title")
print(f"Title: {title}")

# Iterate over array items
print("Subjects:")
for i in range(xmp.count_array_items(consts.NS_DC, "subject")):
    # Array items are 1-indexed in XMP
    item = xmp.get_array_item(consts.NS_DC, "subject", i + 1) 
    print(f"- {item}")

# Serialize XMP to RDF/XML string
rdf_string = xmp.dump_as_rdf()
print("\n--- XMP as RDF ---")
print(rdf_string)
print("------------------")

view raw JSON →