All notable changes to this project will be documented in this file.
Do not edit this file manually, it is generated automatically by changie.
- Add status to RetryTimeoutError
- Add permissioning to release tag action"
- Adding EnvironmentInfo query
- Remove unneeded subquery, which creates invalid SQL for some dialects.
- Update github action perission
- Bump max message size to 512mb
- Fix multiple group by test
- add error, status, and query_id to QueryFailedError
- updates message limit to support large datasets
- Add
lazyparameter to clients which allows lazy loading of certain model fields. - Officially support Python 3.13
- Fix type of group_by in query method
- Fix TimeDimension in group_by
- Add support to pass in typed group by
- Make
pyrightrun onstrictmode and addmypy.
- Added
FlexibleEnumMetato make enums non-breaking when we add a new value to the API.
- Fix release CI tags pattern
- Add code quality check on merge to
mainand on publish to PyPI
ExponentialBackoff.timeout_msis now deprecated. Prefer settingtotal_timeoutin the globalTimeoutOptions.
- Added
timeoutparameter to Semantic Layer client and to the underlying GraphQL clients - Added new
Connect,ExecuteandRetrysubclasses forTimeoutError
- Create CODEOWNERS file
- Fix Github Actions workflow for tagging releases
- Double the max message size
- Support for adhoc queries with only groupby. This is equivalent to listing dimension values.
- Fixes issue with
orderByandwherewhen going via GraphQL
- Fix release PR action to use correct title for PR
order_byclause of queries using saved queries no longer support string inputs and require explicitOrderByMetricorOrderByGroupBy
- Deprecate
TimeGranularityenum and all other fields that used it
- Specifying order by with
OrderByMetricandOrderByGroupBy - Add support for custom time granularity
- Order by for
compile_sqlnow works as expected
- Improved internal representation of query parameters and added better validation
- Improve error display
- Add new mechanisms to deprecate fields and classes
compile_sqlmethod for getting the compiled SQL of a query- Allow saved queries in
queryandcompile_sql
- Fixed wrong annotation for
saved_queriesin sync client.
- Better error messages
- Client-side validation of query parameters
- Add
Exportlist toSavedQuery - Add
SavedQueryQueryParamlist toSavedQuery
- Lower minimum PyArrow version to
12.0.0
- Removed some files that were added accidentally
- Relaxed dependency constraints
- Added
saved_queryfetching via GraphQL - Added
entityfetching via GraphQL - Added more fields to
dimension - Added more fields to
metric - Full support for all query parameters
- Add dimension values query
- Changed how field aliases to make it easier to define new models
- Improved how GraphQL gets generated under the hood
- Fix the way we specify dependencies to avoid conflicts.
- Add
stage_fixto lefthook hooks - Fix github actions for release PR
- Separate tests into unit and integration tests
- Fix warnings and failing integration test
- Change how clients create and poll for query results
- Fix create release CI
- Add
query()to sync client
- Fix import error on 3.9
- Fix
backoffnot usingexp_factor - Dimensions query had wrong value in protocol
- Add
has_sessionto underlying API clients - Add
user-agentheader for telemetry
- Known first party was wrong for Ruff
- Add integration tests with a real server
- Sync and async implementation of GraphQL and ADBC clients.