Make use of common resolution and period

• Create general Resolution and Period enumerations that can be used anywhere
• Create a full dataframe even if no values exist at requested time
• Add further attributes to the class structure
• Make dates timezone aware
• Restrict dates to isoformat

Fix 10minutes fileindex intervals

• Fix 10minutes file index interval range by adding timezone information

Move functionality to core components, use timezones for dates

• Move more functionality into core classes
• Add more attributes to the core e.g. source and timezone
• Make dates of internal data timezone aware, set start date and end date to UTC
• Add issue date to Mosmix class that actually refers to the Mosmix run instead of start
  date and end date
• Use Result object for every data related return
• In accordance with typical naming conventions, DWDObservationSites is renamed to
  DWDObservationStations, the same is applied to DWDMosmixSites
• The name ELEMENT is removed and replaced by parameter while the acutal parameter set
  e.g. CLIMATE_SUMMARY is now found under PARAMETER_SET
• Remove StorageAdapter and its dependencies
• Methods self.collect_data() and self.collect_safe() are replaced by self.query() and
  self.all() and will deprecate at some point

v0.11.1

• Bump h5py to version 3.1.0 in order to satisfy installation on Python 3.9

Fix InfluxDB exports and increase efficiency and performance

• InfluxDB export: Fix export in non-tidy format (#230). Thanks, @wetterfrosch!
• InfluxDB export: Use "quality" column as tag (#234). Thanks, @wetterfrosch!
• InfluxDB export: Use a batch size of 50000 to handle larger amounts of data (#235). Thanks, @wetterfrosch!
• Update radar examples to use wradlib>=1.9.0. Thanks, @kmuehlbauer!
• Change wherever possible column type to category
• Increase efficiency by downloading only historical files with overlapping dates if start_date and end_date are given
• Use periods dynamically depending on start and end date
• Fix inconsistency within 1 minute precipitation data where historical files have more columns
• Improve DWD PDF parser to extract quality information and select language.
  Also, add an example at example/dwd_describe_fields.py as well as
  respective documentation.

v0.10.1: Radar tests: Switch to ASR Borkum.

• Upgrade to dateparser-1.0.0. Thanks, @steffen746, @noviluni and @Gallaecio!
  This fixes a problem with timezones on Windows. The reason is that
  Windows has no zoneinfo database and tzlocal switched from pytz to tzinfo.
  See also #222.

Improve MOSMIX subsystem and various refactorings

• CLI: Obtain "--tidy" argument from command line
• Extend MOSMIX support to equal the API of observations
• DWDObservationSites now filters for those stations which have a file on the server
• DWDObservationData now also takes an individual parameter
  independent of the pre-configured DWD datasets by using DWDObservationParameter or
  similar names e.g. "precipitation_height"
• Newly introduced coexistence of DWDObservationParameter and DWDObservationParameterSet
  to address parameter sets as well as individual parameters
• Imports are changed to submodule thus now one has to import everything from
  wetterdienst.dwd
• Renaming of time_resolution to resolution, period_type to period, several other
  relabels

Large refactoring and full coverage of DWD radar data

• Large refactoring
• Make period type in DWDObservationData and cli optional
• Activate SQL querying again by using DuckDB 0.2.2.dev254. Thanks, @Mytherin!
• Fix coercion of integers with nans
• Fix problem with storing IntegerArrays in HDF
• Rename DWDStationRequest to DWDObservationData
• Add DWDObservationSites API wrapper to acquire station information
• Move discover_climate_observations to DWDObservationMetadata.discover_parameters
• Add PDF-based DWDObservationMetadata.describe_fields()
• Upgrade Docker images to Python 3.8.6
• Move intermediate storage of HDF out of data collection
• Fix bug with date filtering for empty/no station data for a given parameter
• Radar data: Add non-RADOLAN data acquisition

Add multiple storage pipelines, MOSMIX forecast and API

• Add TTL-based persistent caching using dogpile.cache
• Add example/radolan.py and adjust documentation
• Export dataframe to different data sinks like SQLite, DuckDB, InfluxDB and CrateDB
• Query results with SQL, based on in-memory DuckDB
• Split get_nearby_stations into two functions, get_nearby_stations_by_number and
  get_nearby_stations_by_distance
• Add MOSMIX client and parser. Thanks, @jlewis91!
• Add basic HTTP API

Enhance documentation, add test for example notebook, fix bug with TimeResolution enumeration

• Add test for Jupyter notebook
• Add function to discover available climate observations (time resolution, parameter, period type)
• Make the CLI work again and add software tests to prevent future havocs
• Use Sphinx Material theme for documentation
• Fix typo in enumeration for TimeResolution.MINUTES_10