gdal-utils

This gdal-utils Python package is a sub-set of the GDAL Osgeo Python package.

If you've installed the GDAL package you have these utils. However you may wish to get a newer version of the utilities without changing GDAL, that's what this package is for.

For using and developing the utils please refer to the osgeo documentation and to the GDAL API Tutorial.

We don't systematically test the utils against a different versions of gdal. Versioning of GDAL and gdal-utils are independent from each other, today's gdal-utils v3.3 works with GDAL v3.4 but it might not with v3.2. We recommend upgrading to the latest versions of each as general practice. See 'Packaging' below for how to test compatibility.

Dependencies

> - gdal (the osgeo package) > - numpy (1.0.0 or greater) and header files (numpy-devel) (not > explicitly required, but many examples and utilities will not work > without it)

Installation

gdal-utils can be installed from pypi.org:

$ python -m pip install gdal-utils

After install the utilities are in PYTHYONHOME\Scripts and can be invoked like regular programs, gdal_edit instead of gdal_edit.py or python path/to/gdal_edit.py for example.

Packaging

Starting March 2022 installing gdal-utils with pip will use Setuptools' _console_scripts, which turn the scripts into native platform executables that call the script using the appropriate platform interpreter. This means you no longer need to something similar as a post-install step. If this causes problems with your distribution please file an issue on Github.

Recipe for testing gdal-utils compatibility against your installed GDAL binaries:

# Get installed GDAL version
export _GDALVER=`gdal-config --version``

# verify python osgeo is present
python -c "from osgeo import gdal;print(f'Python Osgeo version: {gdal.__version__}')"

git clone https://github.com/OSGeo/gdal.git --depth=50
git remote set-branches origin "*"
git fetch -v --depth=50

# Install current version of gdal-utils
cd swig/python/gdal-utils
pip install .
# OR, to use published PyPi.org version:
# pip install gdal-utils

# set source code tree to match binary gdal version
git checkout v$_GDALVER

cd gdal/autotest
python install -r requirements.txt
pytest ./pyscripts

Coding best practices

Use existing scripts for style guidelines, generally newer ones are will be more pythonic than old.

Passes _flake8 lint test.

When called without arguments script should return a message prefixed 'Usage: and use return code 2`.:

» gdal_edit
Usage: gdal_edit [--help-general] [-ro] [-a_srs ]
    ...
    [-unsetmd] [-oo =]... [-mo =]... 

Edit in place various information of an existing GDAL dataset.