v1.0.12

Added

• SQLReturnType enum (#21)

Changed

• move to arrow_v1 for sql queries (#21)

Removed

• FoundrySQLClient (#21)

v1.0.11

Changed

• better SQL exceptions (#20)

Fixed

• pip installation documentation for zsh users (#19)

v1.0.10

Changed

• use ruff and black instead of pylint and ufmt (#15)
• converted the subprocess calls with git to python only imitations (#15)

Fixed

• environment variables didn't take precedence (#15)

v1.0.9

Fixed

• file upload for files greater than 2GB (#16)

v1.0.8

Added

• python 3.11 support, as pyspark 3.4 gained support for it (#13)
• conda-forge badges to the README (#13)

Fixed

• typo in docs (#12)

Removed

• pandas<2 restriction, as pyspark 3.4 supports it (#13)

v1.0.7

Added

• skip_instance_cache kwarg to FoundryFileSystem, which gets passed to fsspec
  it should be set to True in a multithreaded environment (e.g. Streamlit),
  as the same filesystem instance gets reused by default,
  which resulted in weird behaviour. (#11)
• an extra check, which may prevent sending a bad request to foundry
  when skip_instance_cache is not used (#11)

Changed

• remove pandas from top level import, which should speed up the initial import (#11)
• restrict pandas to version less than 2,
  as pyspark is currently not compatible with version 2 (#11)

v1.0.6

Changed

• Include dataset_rid in DatasetAlreadyExistsError to make it easier to consume. (#10)

v1.0.5

Fixed

• Use gatekeeper:view-resource over compass:view (#8)

v1.0.4

Fixed

• Better exception handling when git executable is missing (#6)

v1.0.3

Added

• tracker and changelog url to setup.cfg (#4)
• more pypi classifiers to setup.cfg (#4)

Changed

• Changelog is now in the format of Keep a Changelog (#4)
• Updated fsspec example in README (#3)

Fixed

• README code block highlighting and Apache License link (#2)
• wrong styling of pipy batch in readme (#1)