Releases: wmo-im/synop2bufr
v0.6.2
v0.6.1
Bug Fixes
- The return objects now always have the same structue for the
_meta
key, which prevents the bugError: '_meta'
when stations can't be found in the metadata or a report can't be parsed. This now means the warning and error messages have been moved toresult["_meta"]["result"]["warnings"]
andresult["_meta"]["result"]["errors"]
respectively. - Updated
cli.py
to handleNoneType
objects returned in the_meta
key correctly so that no CSV file or BUFR file is written for unsuccessfully decoded reports.
v0.6.0
Main Changes
- Pymetdecoder and csv2bufr logs are now captured and stored appropriately in warning and error arrays which can be accessed in the webapp.
- Valid min and max values have been implemented directly into the 307080 and 307096 mapping files so that out of range errors are handled before ecCodes and are more informative for the user. Many of these valid ranges have been taken from the WMO Guide to Instruments and Methods.
- A verification that, if both reported, the dewpoint temperature is less than or equal to the air temperature has been added.
- Unit tests have been added for the two quality control steps above.
- NIL reports are no longer written as an empty BUFR and the user is informed of this through a warning.
- Test data has been updated, most notably the wis2box-data now has a
barometer_height
column to meet the current requirements of synop2bufr.
Bug Fixes
- When
iR = 3
, absent precipitation groups represent zero preciptation, not missing data. Previously this was not supported and precipitation levels were set to missing. - Section 3 and section 4 mappings were not being updated correctly which caused the incorrect error
Missing station height
.
v0.5.1
v0.5.0
This release provides the following updates:
- Dynamic use of the 307080 and 307096 templates depending on the method of wind measurement.
- Encoding of the barometer height obtained from the station list file.
- Support for
nil
reports. - Support for transformation of multiple GTS messages in one file by partitioning according to the end of message signal
nnnn
. - A CSV file is output during transformation, containing the decoded data of the input file where each row corresponds to each successfully converted report.
- Restructuring of the example data, addition of Cuban GTS data and synthetic data (which contains almost all possible groups that can be written in a synop report).
- Updated unit tests.
- Various bug fixes found by testing with different GTS data.
v0.4.1
v0.4.0
Merge pull request #22 from wmo-im/debug Fixes #21
v0.3.2
Bug fix: empty lines in metadata file no longer causes error.
v0.3.1
note: minor update due to failed release.
This release provides the following small fixes:
Default value of the --metadata option is now station_list.csv rather than metadata.csv to be consistent with the exercises and other modules.
The logging level in cli.py determined by an environment variable has now been given a default value "ERROR" in the case that no environment variable is present (for example, when pip installing the module rather than running from the Docker image).
Potential breaking changes
Update to pinned version of csv2bufr and wmo-im fork of pymetdecoder. Stricter error checking in pymetdecoder may break using previously working data files.
v0.3.0
This release provides the following small fixes:
- Default value of the --metadata option is now station_list.csv rather than metadata.csv to be consistent with the exercises and other modules.
- The logging level in cli.py determined by an environment variable has now been given a default value "ERROR" in the case that no environment variable is present (for example, when pip installing the module rather than running from the Docker image).
Potential breaking changes
Update to pinned version of csv2bufr and wmo-im fork of pymetdecoder. Stricter error checking in pymetdecoder may break using previously working data files.