Fiona

Fiona is OGR’s neat, nimble, no-nonsense API for Python programmers.

https://travis-ci.org/Toblerity/Fiona.png?branch=master https://coveralls.io/repos/Toblerity/Fiona/badge.png

Fiona is designed to be simple and dependable. It focuses on reading and writing data in standard Python IO style and relies upon familiar Python types and protocols such as files, dictionaries, mappings, and iterators instead of classes specific to OGR. Fiona can read and write real-world data using multi-layered GIS formats and zipped virtual file systems and integrates readily with other Python GIS packages such as pyproj, Rtree, and Shapely.

For more details, see:

Usage

Collections

Records are read from and written to file-like Collection objects returned from the fiona.open() function. Records are mappings modeled on the GeoJSON format. They don’t have any spatial methods of their own, so if you want to do anything fancy with them you will probably need Shapely or something like it. Here is an example of using Fiona to read some records from one data file, change their geometry attributes, and write them to a new data file.

import fiona

# Register format drivers with a context manager

with fiona.drivers():

    # Open a file for reading. We'll call this the "source."

    with fiona.open('tests/data/coutwildrnp.shp') as source:

        # The file we'll write to, the "sink", must be initialized
        # with a coordinate system, a format driver name, and
        # a record schema.  We can get initial values from the open
        # collection's ``meta`` property and then modify them as
        # desired.

        meta = source.meta
        meta['schema']['geometry'] = 'Point'

        # Open an output file, using the same format driver and
        # coordinate reference system as the source. The ``meta``
        # mapping fills in the keyword parameters of fiona.open().

        with fiona.open('test_write.shp', 'w', **meta) as sink:

            # Process only the records intersecting a box.
            for f in source.filter(bbox=(-107.0, 37.0, -105.0, 39.0)):

                # Get a point on the boundary of the record's
                # geometry.

                f['geometry'] = {
                    'type': 'Point',
                    'coordinates': f['geometry']['coordinates'][0][0]}

                # Write the record out.

                sink.write(f)

    # The sink's contents are flushed to disk and the file is
    # closed when its ``with`` block ends. This effectively
    # executes ``sink.flush(); sink.close()``.

# At the end of the ``with fiona.drivers()`` block, context
# manager exits and all drivers are de-registered.

The fiona.drivers() function and context manager are new in 1.1. The example above shows the way to use it to register and de-register drivers in a deterministic and efficient way. Code written for Fiona 1.0 will continue to work: opened collections may manage the global driver registry if no other manager is present.

Reading Multilayer data

Collections can also be made from single layers within multilayer files or directories of data. The target layer is specified by name or by its integer index within the file or directory. The fiona.listlayers() function provides an index ordered list of layer names.

with fiona.drivers():

    for layername in fiona.listlayers('tests/data'):
        with fiona.open('tests/data', layer=layername) as src:
            print(layername, len(src))

# Output:
# (u'coutwildrnp', 67)

Layer can also be specified by index. In this case, layer=0 and layer='test_uk' specify the same layer in the data file or directory.

with fiona.drivers():

    for i, layername in enumerate(fiona.listlayers('tests/data')):
        with fiona.open('tests/data', layer=i) as src:
            print(i, layername, len(src))

# Output:
# (0, u'coutwildrnp', 67)

Writing Multilayer data

Multilayer data can be written as well. Layers must be specified by name when writing.

with fiona.drivers():

    with open('tests/data/cowildrnp.shp') as src:
        meta = src.meta
        f = next(src)

    with fiona.open('/tmp/foo', 'w', layer='bar', **meta) as dst:
        dst.write(f)

    print(fiona.listlayers('/tmp/foo'))

    with fiona.open('/tmp/foo', layer='bar') as src:
        print(len(src))
        f = next(src)
        print(f['geometry']['type'])
        print(f['properties'])

    # Output:
    # [u'bar']
    # 1
    # Polygon
    # OrderedDict([(u'PERIMETER', 1.22107), (u'FEATURE2', None), (u'NAME', u'Mount Naomi Wilderness'), (u'FEATURE1', u'Wilderness'), (u'URL', u'http://www.wilderness.net/index.cfm?fuse=NWPS&sec=wildView&wname=Mount%20Naomi'), (u'AGBUR', u'FS'), (u'AREA', 0.0179264), (u'STATE_FIPS', u'49'), (u'WILDRNP020', 332), (u'STATE', u'UT')])

A view of the /tmp/foo directory will confirm the creation of the new files.

$ ls /tmp/foo
bar.cpg bar.dbf bar.prj bar.shp bar.shx

Collections from archives and virtual file systems

Zip and Tar archives can be treated as virtual filesystems and Collections can be made from paths and layers within them. In other words, Fiona lets you read and write zipped Shapefiles.

with fiona.drivers():

    for i, layername in enumerate(
            fiona.listlayers(
                '/',
                vfs='zip://tests/data/coutwildrnp.zip')):
        with fiona.open(
                '/',
                vfs='zip://tests/data/coutwildrnp.zip',
                layer=i) as src:
            print(i, layername, len(src))

# Output:
# (0, u'coutwildrnp', 67)

Fiona CLI

Fiona’s command line interface, named “fio”, is documented at docs/cli.rst. Its fio info pretty prints information about a data file.

$ fio info --indent 2 tests/data/coutwildrnp.shp
{
  "count": 67,
  "crs": "EPSG:4326",
  "driver": "ESRI Shapefile",
  "bounds": [
    -113.56424713134766,
    37.0689811706543,
    -104.97087097167969,
    41.99627685546875
  ],
  "schema": {
    "geometry": "Polygon",
    "properties": {
      "PERIMETER": "float:24.15",
      "FEATURE2": "str:80",
      "NAME": "str:80",
      "FEATURE1": "str:80",
      "URL": "str:101",
      "AGBUR": "str:80",
      "AREA": "float:24.15",
      "STATE_FIPS": "str:80",
      "WILDRNP020": "int:10",
      "STATE": "str:80"
    }
  }
}

Installation

Fiona requires Python 2.6, 2.7, 3.3, or 3.4 and GDAL/OGR 1.8+. To build from a source distribution you will need a C compiler and GDAL and Python development headers and libraries (libgdal1-dev for Debian/Ubuntu, gdal-dev for CentOS/Fedora).

To build from a repository copy, you will also need Cython to build C sources from the project’s .pyx files. See the project’s requirements-dev.txt file for guidance.

The Kyngchaos GDAL frameworks will satisfy the GDAL/OGR dependency for OS X, as will Homebrew’s GDAL Formula (brew install gdal).

Python Requirements

Fiona depends on the modules six, cligj, argparse, and ordereddict (the two latter modules are standard in Python 2.7+). Pip will fetch these requirements for you, but users installing Fiona from a Windows installer must get them separately.

Unix-like systems

Assuming you’re using a virtualenv (if not, skip to the 4th command) and GDAL/OGR libraries, headers, and gdal-config program are installed to well known locations on your system via your system’s package manager (brew install gdal using Homebrew on OS X), installation is this simple.

$ mkdir fiona_env
$ virtualenv fiona_env
$ source fiona_env/bin/activate
(fiona_env)$ pip install Fiona

If gdal-config is not available or if GDAL/OGR headers and libs aren’t installed to a well known location, you must set include dirs, library dirs, and libraries options via the setup.cfg file or setup command line as shown below (using git).

(fiona_env)$ git clone git://github.com/Toblerity/Fiona.git
(fiona_env)$ cd Fiona
(fiona_env)$ python setup.py build_ext -I/path/to/gdal/include -L/path/to/gdal/lib -lgdal install

Or specify that build options should be provided by a particular gdal-config program.

(fiona_env)$ GDAL_CONFIG=/path/to/gdal-config pip install .

Windows

Binary installers are available at http://www.lfd.uci.edu/~gohlke/pythonlibs/#fiona and coming eventually to PyPI.

You can download a binary distribution of GDAL from here. You will also need to download the compiled libraries and headers (include files).

When building from source on Windows, it is important to know that setup.py cannot rely on gdal-config, which is only present on UNIX systems, to discover the locations of header files and libraries that Fiona needs to compile its C extensions. On Windows, these paths need to be provided by the user. You will need to find the include files and the library files for gdal and use setup.py as follows.

$ python setup.py build_ext -I<path to gdal include files> -lgdal_i -L<path to gdal library>
$ python setup.py install

Note: The GDAL dll (gdal111.dll) and gdal-data directory need to be in your Windows PATH otherwise Fiona will fail to work.

Development and testing

Building from the source requires Cython. Tests require Nose. If the GDAL/OGR libraries, headers, and gdal-config program are installed to well known locations on your system (via your system’s package manager), you can do this:

(fiona_env)$ git clone git://github.com/Toblerity/Fiona.git
(fiona_env)$ cd Fiona
(fiona_env)$ pip install -e .
(fiona_env)$ nosetests

If you have a non-standard environment, you’ll need to specify the include and lib dirs and GDAL library on the command line:

(fiona_env)$ python setup.py build_ext -I/path/to/gdal/include -L/path/to/gdal/lib -lgdal develop
(fiona_env)$ nosetests

Changes

1.6.0 (2015-07-21)

  • Upgrade Cython requirement to 0.22 (#214).
  • New BytesCollection class (#215).
  • Add GDAL’s OpenFileGDB driver to registered drivers (#221).
  • Implement CLI commands as plugins (#228).
  • Raise click.abort instead of calling sys.exit, preventing suprising exits (#236).

1.5.1 (2015-03-19)

  • Restore test data to sdists by fixing MANIFEST.in (#216).

1.5.0 (2015-02-02)

  • Finalize GeoJSON feature sequence options (#174).
  • Fix for reading of datasets that don’t support feature counting (#190).
  • New test dataset (#188).
  • Fix for encoding error (#191).
  • Remove confusing warning (#195).
  • Add data files for binary wheels (#196).
  • Add control over drivers enabled when reading datasets (#203).
  • Use cligj for CLI options involving GeoJSON (#204).
  • Fix fio-info –bounds help (#206).

1.4.8 (2014-11-02)

  • Add missing crs_wkt property as in Rasterio (#182).

1.4.7 (2014-10-28)

  • Fix setting of CRS from EPSG codes (#149).

1.4.6 (2014-10-21)

  • Handle 3D coordinates in bounds() #178.

1.4.5 (2014-10-18)

  • Add –bbox option to fio-cat (#163).
  • Skip geopackage tests if run from an sdist (#167).
  • Add fio-bounds and fio-distrib.
  • Restore fio-dump to working order.

1.4.4 (2014-10-13)

  • Fix accidental requirement on GDAL 1.11 introduced in 1.4.3 (#164).

1.4.3 (2014-10-10)

  • Add support for geopackage format (#160).
  • Add -f and –format aliases for –driver in CLI (#162).
  • Add –version option and env command to CLI.

1.4.2 (2014-10-03)

  • –dst-crs and –src-crs options for fio cat and collect (#159).

1.4.1 (2014-09-30)

  • Fix encoding bug in collection’s __getitem__ (#153).

1.4.0 (2014-09-22)

  • Add fio cat and fio collect commands (#150).
  • Return of Python 2.6 compatibility (#148).
  • Improved CRS support (#149).

1.3.0 (2014-09-17)

  • Add single metadata item accessors to fio inf (#142).
  • Move fio to setuptools entry point (#142).
  • Add fio dump and load commands (#143).
  • Remove fio translate command.

1.2.0 (2014-09-02)

  • Always show property width and precision in schema (#123).
  • Write datetime properties of features (#125).
  • Reset spatial filtering in filter() (#129).
  • Accept datetime.date objects as feature properties (#130).
  • Add slicing to collection iterators (#132).
  • Add geometry object masks to collection iterators (#136).
  • Change source layout to match Shapely and Rasterio (#138).

1.1.6 (2014-07-23)

  • Implement Collection __getitem__() (#112).
  • Leave GDAL finalization to the DLL’s destructor (#113).
  • Add Collection keys(), values(), items(), __contains__() (#114).
  • CRS bug fix (#116).
  • Add fio CLI program.

1.1.5 (2014-05-21)

  • Addition of cpl_errs context manager (#108).
  • Check for NULLs with ‘==’ test instead of ‘is’ (#109).
  • Open auxiliary files with encoding=’utf-8’ in setup for Python 3 (#110).

1.1.4 (2014-04-03)

  • Convert ‘long’ in schemas to ‘int’ (#101).
  • Carefully map Python schema to the possibly munged internal schema (#105).
  • Allow writing of features with geometry: None (#71).

1.1.3 (2014-03-23)

  • Always register all GDAL and OGR drivers when entering the DriverManager context (#80, #92).
  • Skip unsupported field types with a warning (#91).
  • Allow OGR config options to be passed to fiona.drivers() (#90, #93).
  • Add a bounds() function (#100).
  • Turn on GPX driver.

1.1.2 (2014-02-14)

  • Remove collection slice left in dumpgj (#88).

1.1.1 (2014-02-02)

  • Add an interactive file inspector like the one in rasterio.
  • CRS to_string bug fix (#83).

1.1 (2014-01-22)

  • Use a context manager to manage drivers (#78), a backwards compatible but big change. Fiona is now compatible with rasterio and plays better with the osgeo package.

1.0.3 (2014-01-21)

  • Fix serialization of +init projections (#69).

1.0.2 (2013-09-09)

  • Smarter, better test setup (#65, #66, #67).
  • Add type=’Feature’ to records read from a Collection (#68).
  • Skip geometry validation when using GeoJSON driver (#61).
  • Dumpgj file description reports record properties as a list (as in dict.items()) instead of a dict.

1.0.1 (2013-08-16)

  • Allow ordering of written fields and preservation of field order when reading (#57).

1.0 (2013-07-30)

  • Add prop_type() function.
  • Allow UTF-8 encoded paths for Python 2 (#51). For Python 3, paths must always be str, never bytes.
  • Remove encoding from collection.meta, it’s a file creation option only.
  • Support for linking GDAL frameworks (#54).

0.16.1 (2013-07-02)

  • Add listlayers, open, prop_width to __init__py:__all__.
  • Reset reading of OGR layer whenever we ask for a collection iterator (#49).

0.16 (2013-06-24)

  • Add support for writing layers to multi-layer files.
  • Add tests to reach 100% Python code coverage.

0.15 (2013-06-06)

  • Get and set numeric field widths (#42).
  • Add support for multi-layer data sources (#17).
  • Add support for zip and tar virtual filesystems (#45).
  • Add listlayers() function.
  • Add GeoJSON to list of supported formats (#47).
  • Allow selection of layers by index or name.

0.14 (2013-05-04)

  • Add option to add JSON-LD in the dumpgj program.
  • Compare values to six.string_types in Collection constructor.
  • Add encoding to Collection.meta.
  • Document dumpgj in README.

0.13 (2013-04-30)

  • Python 2/3 compatibility in a single package. Pythons 2.6, 2.7, 3.3 now supported.

0.12.1 (2013-04-16)

  • Fix messed up linking of README in sdist (#39).

0.12 (2013-04-15)

  • Fix broken installation of extension modules (#35).
  • Log CPL errors at their matching Python log levels.
  • Use upper case for encoding names within OGR, lower case in Python.

0.11 (2013-04-14)

  • Cythonize .pyx files (#34).
  • Work with or around OGR’s internal recoding of record data (#35).
  • Fix bug in serialization of int/float PROJ.4 params.

0.10 (2013-03-23)

  • Add function to get the width of str type properties.
  • Handle validation and schema representation of 3D geometry types (#29).
  • Return {‘geometry’: None} in the case of a NULL geometry (#31).

0.9.1 (2013-03-07)

  • Silence the logger in ogrext.so (can be overridden).
  • Allow user specification of record field encoding (like ‘Windows-1252’ for Natural Earth shapefiles) to help when OGR can’t detect it.

0.9 (2013-03-06)

  • Accessing file metadata (crs, schema, bounds) on never inspected closed files returns None without exceptions.
  • Add a dict of supported_drivers and their supported modes.
  • Raise ValueError for unsupported drivers and modes.
  • Remove asserts from ogrext.pyx.
  • Add validate_record method to collections.
  • Add helpful coordinate system functions to fiona.crs.
  • Promote use of fiona.open over fiona.collection.
  • Handle Shapefile’s mix of LineString/Polygon and multis (#18).
  • Allow users to specify width of shapefile text fields (#20).

0.8 (2012-02-21)

  • Replaced .opened attribute with .closed (product of collection() is always opened). Also a __del__() which will close a Collection, but still not to be depended upon.
  • Added writerecords method.
  • Added a record buffer and better counting of records in a collection.
  • Manage one iterator per collection/session.
  • Added a read-only bounds property.

0.7 (2012-01-29)

  • Initial timezone-naive support for date, time, and datetime fields. Don’t use these field types if you can avoid them. RFC 3339 datetimes in a string field are much better.

0.6.2 (2012-01-10)

  • Diagnose and set the driver property of collection in read mode.
  • Fail if collection paths are not to files. Multi-collection workspaces are a (maybe) TODO.

0.6.1 (2012-01-06)

  • Handle the case of undefined crs for disk collections.

0.6 (2012-01-05)

  • Support for collection coordinate reference systems based on Proj4.
  • Redirect OGR warnings and errors to the Fiona log.
  • Assert that pointers returned from the ograpi functions are not NULL before using.

0.5 (2011-12-19)

  • Support for reading and writing collections of any geometry type.
  • Feature and Geometry classes replaced by mappings (dicts).
  • Removal of Workspace class.

0.2 (2011-09-16)

  • Rename WorldMill to Fiona.

0.1.1 (2008-12-04)

  • Support for features with no geometry.

Credits

Fiona is written by:

Fiona would not be possible without the great work of Frank Warmerdam and other GDAL/OGR developers.

Some portions of this work were supported by a grant (for Pleiades) from the U.S. National Endowment for the Humanities (http://www.neh.gov).