v0.21.0

• Update vecorel-cli
• Make the library better usable as a Python library
• Add support for Python 3.14, remove support for Python 3.10
• Support HCAT mapping CSV files without crop_code
• Split Germany BB and NDS in block dataset and crop fields
• Fix the column additions of the determination fields in the AI4SF converter
• Add HCAT to datasets where possible
• Updated years & variants for at_crop, be_vlg, es_an, es_cl, es_pv, ie, pt, se
• Extend create_stac, include include fiboa data
• Publish command; skip hidden files, generate better texts
• Fix to vecorel: converter.license and provider should be string
• Added a Dockerfile to simplify working with fiboa
• Command fiboa publish to automate source coop publication process
  • Run the converter to get the parquet file
  • Validate parquet file
  • Check for README.md, if missing generates one based on data-survey (if available) and converter
  • Check for LICENSE.txt, if missing generate one based on the converter file
  • Generate pmtiles file
  • Check AWS-environment vars
  • Synchronize parquet + pmtiles + README/LICENSE to source coop repo
• Check for license validity, either SPDX string or custom with url
• Seperate concerns for HCAT utility classes:
  • AddHCATMixin assures hcat-extension validity and csv-based data-conversion if required
  • EuroCropsConverterMixin is a BaseClass for EuroCrops-provided datasets
  • EuroLandBaseConverter is a BaseClass for Euroland-provided datasets
• Avoid base property schema override
• Add Converter for Bulgaria
• Remove unintended CommonMark formatting (indentation) from descriptions in converters
• Fibo improve command:
  • Upgrades from fiboa-0.2 if required
  • Adds HCAT if specified
• Various minor bug fixes

v0.20.2

• Reimplementation tests and CI
• Update extensions namespace to fiboa.org, use vecorel schemas
• Calibrate ha and m2 calculation in converters
• Adapt create-stac-collection to fiboa

v0.20.1

• Reimplementation some modules (e.g. converters)
• Upgrade vecorel-cli dependency

v0.20.0

• Reimplementation based on vecorel-cli
• Upgrade to support fiboa >= 0.3
• Removed support for fiboa < 0.3

v0.10.0

Added

• Converter for Austrian crop fields
• Converters for Spain: Aragon, Andalusia, Balearic Islands, Basque Country, Catalonia, Cantabria, Castilla y León, Castilla-La Mancha, Canary Islands, Extremadura, Galicia, Madrid, Navarra, Valencia
• Use ruff format + linting for a uniform code style

Changed

• Refactored converters to use a class-based approach
• Start to use https://fiboa.org/code/ prefixed codes for our own code lists
• Use only unix line-endings in source files
• Use set instead of list for Converter.extensions
• Converted datasets are hilbert-curve sorted

v0.9.0

Added

• Command fiboa improve with helpers to
  • change the CRS
  • change the GeoParquet version and compression
  • fill missing perimeter/area values
  • fix invalid geometries
  • rename columns
• Converter for Lithuania (EuroCrops)
• Converter for Slovenia
• Converter for Slovakia
• Converter for Switzerland
• Converter for Czech
• Converter for US Department of Agriculture Crop Sequence Boundaries
• Converter for California (US) Statewide Crop Mapping
• Converter for Latvia (from original source)
• Converter for Japan, currently based on supplied (non-fiboa) parquet files
• fiboa convert: New parameter --original-geometries / -og to keep the original geometries

Changed

• fiboa convert:
  • Writes custom schemas to collection metadata
  • Geometries are made valid using GeoPanda's make_valid method by default
  • MultiPolygons are converted to Polygons by default
• fiboa validate uses custom schemas for validation
• fiboa merge keeps custom schemas when needed
• Extended converter for Croatia; with crop_code and crop_name
• Many converters implement the admin extension

Removed

• fiboa convert: Removed the explicit parameter explode_multipolygon from the converter

Fixed

• Fix converter for Estland to use the id ec_ee instead of ec_es
• Assure tests don't download external sources

v0.8.0

Added

• Merge command: fiboa merge
• Converter for Croatia
• Converter for Germany, Mecklenburg-Western Pomerania
• Converter for Germany, Saarland
• Converter for Germany, Saxony
• Converter for Estonia (EuroCrops)
• Converter for Sweden
• Converter for Luxembourg
• Converter for Ireland
• Converter for Lacuna Labels (A region-wide, multi-year set of crop field boundary labels for Africa)

Changed

• fiboa convert: Default compression changed from zstd to brotli
• The default row group size of exported parquet files was changed from ~1.000.000 to 25.000

Fixed

• Datatype conversion from pandas to pyarrow fixed

v0.7.0

Added

• Converter for Luís Eduardo Magalhães (LEM) and other municipalities in the west of Bahia state, Brazil (br_ba_lem)
• Converter for Denmark (dk)
• Additional converters for EuroCrops datasets: ec_be_vlg, ec_nl_crop
• New parameter --geoparquet1 to generate GeoParquet 1.0 without bbox column instead of GeoParquet 1.1 with bbox column

Changed

• CLI creates GeoParquet 1.1 with bbox column by default
• The function signature in the convert function of the converters has changed to a simpler more future-proof variant.
• The EuroCrops converters extend the original converters

Fixed

• fiboa convert: Create output folder if it doesn't exist
• Strip whitespaces/newlines from created STAC collections
• fiboa create-geojson: Don't write FeatureCollections to folder if a filename is given

v0.6.0

Added

• Added a SHORT_NAME variable to the converter template
• Added a FILE_MIGRATION variable to the converter template for per-file migrations
• Added a LAYER_FILTER variable to the converter template for loading specific layers from a file
• Added -i parameter to specify input files for converters
• fiboa converters output can be customized with options -p, -s and -v.
• fiboa convert reads JSON file with custom logic, which allows to access nested objects through dot notation
• Converter for Slovenia via EuroCrops (ec_lv)
• Converter for Planet's Automated Field Boundary (planet_afb)
• Converter for Portugal (pt)
• Converter for DigiFarm (digifarm)
• Converter for AI4SmallFarms in Cambodia and Vietnam (ai4sf)
• Further tests

Changed

• The BBOX is optional in the converter template as it will be computed automatically from the data.
• The PROVIDER_NAME and PROVIDER_URL variables in the converter template were replaced by PROVIDERS
• fiboa converters is more readable by default
• Upgraded to geopandas 1.0.0, which migrates from fiona to pyogrio for data loading
• The EuroCrops converters (prefix: ec_) use the HCAT fiboa extension

Fixed

• Fixed schema issue for the tk10 column in de_bb converter
• jsonschema library doesn't warn against external references any longer

v0.5.0

Added

• Basic support for patternProperties in GeoParquet creation
• The converter template accepts multiple input URLs
• Added parameter to explode multipolygons to polygons (explode_multipolygon, default: False)
• Converter for Belgium, Flanders (be_vlg)
• Converter for Belgium, Wallonia (be_wa)
• Converter for Finland (fi)
• Converter for France (fr)
• Converter for The Netherlands (nl and nl_crops)
• Converter for Slovenia via EuroCrops (ec_si)

Changed

• The --cache option for the convert command asks for a folder instead of a file
• The cache_file parameter in converters has been renamed to cache (requires changes in the converter templates)
• The converter template allows for more detailed source information
• The URI constant in the template was renamed to SOURCES (requires changes in the converter templates)

Fixed

• Extensions were not correctly displayed in describe and validate command
• Fixed regular expressions for email and uuid in data validation