1.2.3

• Handle consolidated Zarrs correctly when appending data (fixes #47).

• Add configurations for SST_MED_SST_L4_NRT_OBSERVATIONS_010_004

• Update LWQ300 configuration

• Improve validation script

1.2.2

• Fix incorrect version number.

1.2.1

• Fix an occasionally failing test and update an example configuration.

1.2.0

• Introduce the new parameter output.append_mode to select behaviour
  when appending new data which overlaps in the append dimension with the
  existing data.

• Adjusted the built-in pre-processing to handle more cases:
  Bounds dimensions of concatenation coordinates are extended by their
  dimension and transformed to variables if necessary.

• Whether to adjust output metadata (global attributes) after the last
  write/append can now be forced by the new output.adjust_metadata
  setting whose default is false. If set to true, this will adjust
  the following metadata attributes:

  • "history"
  • "source"
  • "time_coverage_start"
  • "time_coverage_end"

  In addition, extra metadata (global attributes) can now be added
  after the last write/append using the new setting
  output.metadata whose value is a mapping from attribute
  names to values.

  The above functionality is also reflected in two new CLI options
  --adjust-metadata and --finalize-only. In combination, they
  can be used to later adjust metadata in already existing Zarr
  datasets. (#20, #34)

• Fixed a problem that avoided appending datasets that contained variables
  with dtype = "|S1" (ESA SST CCI). Such variables are written initially,
  but then no longer appended at all given that they do not contain the
  dimension along we append. Otherwise nc2zarr will still fail. (#38)

• Fixed a severe problem where data was "encoded" twice before
  being appended to existing data. The bug became apparent
  with input.decode_cf = false (the default) and for variables
  that defined a scale_factor and/or add_offset encoding
  attribute. (#35)

• Added a CLI tool nc2zarr-batch.

• Add and update some sample configurations and scripts.

1.1.1

• Fix an incompatibility with version 8 of the click library which
  was causing command-line parsing errors.

1.1.0

• Ensure attributes are maintained when missing dimensions are added to
  concatenation dimension variable. (#32)

• Added some basic batch utilities that help spawning multiple concurrent
  nc2zarr jobs, see new module nc2zarr.batch. (#19)

• Local input and output paths may now include tilde '~' which will expand
  to the current user's home, and '~username' for the home directory
  of a specified user. (#26)

• Fixed a problem when input/sort_by was "name" and one of input/paths
  ended with "/". In these cases sorting did not work. (#29)

• Fixed problem where appending unnecessarily required a coordinate
  variable (#27)

• Input path that were no wildcards have been ignored if the input path did
  not exist. Now an error is raised in such cases. (#25)

• Fixed exception TypeError: 'int' object is not iterable
  raised with some process/rechunk configurations. (#23)

• You can now provide your Python code for customization
  of the datasets read and written. (#16)

  • input/custom_preprocessor: "module:function" is called on
    each input after optional variable selection.
  • process/custom_processor: "module:function" is called after
    optional renaming and before optional rechunking.
  • output/custom_postprocessor: "module:function" is called before
    the final dataset is written. (#21)

  All three functions are expected to receive an xarray.Dataset object
  as only argument and return the same or modified xarray.Dataset object.
  Note: to let Python import "module" that is not a user package,
  you can extend the PYTHONPATH environment variable before
  calling nc2zarr:

  export PYTHONPATH=${PYTHONPATH}:path/to/my/modules
  nc2zarr ...
  

• Add more example configurations and scripts.

1.0.0

• Fixed some issues with Zarr (re-)chunking given by process parameter
  process/rechunk:

  • Fix: New chunking can now overlap multiple chunks in input
    (Dask chunks). (#12)

  • Introduced new chunk size value "input" which forces using chunk sizes
    from input. (#12)

  • Fixed problems that occurred when using the process/rechunk
    configuration

    • if a dimension was not configured at all;
    • if the dataset contained 0-dimension variables, which is
      typical for variables used to describe the spatial reference system,
      e.g. crs.

    In these cases the error message from Zarr was
    AssertionError: We should never get here.". (#14)

• Introduced new input parameter prefetch_chunks. (#10)

• Add an AppVeyor CI configuration file.

• Add the --sort-by command-line option.

• Add example configuration files for the following datasets:

  • C_GLS_LWQ300_GLOBE_OLCI_V1
  • OCEANCOLOUR_BAL_CHL_L3_NRT_OBSERVATIONS_009_049
  • OCEANCOLOUR_BS_CHL_L4_NRT_OBSERVATIONS_009_045
  • OCEANCOLOUR_MED_CHL_L4_NRT_OBSERVATIONS_009_041

• Add demo Jupyter notebooks which open and plot converted data from object
  storage.

0.1.0

Initial version.