Currently, we attempt to hide the nvd3 tooltips (if any were on screen) before we draw a new viz after rerunning a query. The hiding is done by selecting the first nvtooltip element and setting the opacity to 0.
This somtimes leave behind a trail of old tooltips if a tooltip is left behind by this nvd3 bug https://github.com/novus/nvd3/issues/1262. This PR modifies the behavior of how we clean up tooltips between rerun of queries by selecting all nvd3 tooltips and removing them all from the DOM before redrawing nvd3 vizzes.
* Revert "creating new circular-json safe stringify and replacing one call (#6772)"
This reverts commit 11a7ad00b7.
* Revert "Improve Unicode support for MSSQL (#6690)"
This reverts commit c44ae612df.
* Revert "Fix uniqueness constraints on tables table (#6718)"
This reverts commit c4fb7a0a87.
- Rename the default feature flags key in `config.py` to DEFAULT_FEATURE_FLAGS
- Merge default feature flags with user defined ones allowing the latter to overwrite the former
- Expose feature_flags for both server and client to use
- Add a utility method for checking whether a feature flag is on on server side
Summary: Superset code enforces (in Table crud view pre_add) that the
table is unique within <database, schema, table_name). Indeed in commit
15b67b2c6c (in 2016), the model was
updated to reflect that. However, it was never ported over to a
migration.
I am fixing that in this diff. I am choosing to make this be a new
migration instead of fixing an existing one since I want to fix existing
installations also cleanly.
I also considered removing the uniqueness constraint, but that won't
work: First because anyway there are other places where the <database,
schema, table> uniqueness is enforced in code. But also, the .sql field
isn't a first citizen yet: The schema of the table is picked up from the
table-name and the sql part is only used when creating the explore
query. So indeed we want this uniqueness constraint. (Also it breaks the
unit tests in dict_import_export_tests.py)
[Perhaps it can be removed when we have true .sql support, but for now
the user would have to create a database view and he can use that as the
'table name'. That way he gets schema inference also]
Also added INFO logging to the alembic migration.
* Add licenses
* More licenses
* Ignore messages.json as they are generated
* More licenses
* Also typescript
* Fix alignment
* Add to svg
* Many more licenses
* more licenses
* Better excludes
* Add licenses to docs and md files
* Pre-finalize licenses
* Fix newlines
* Finalize all sourde licenses
* Fix lint
* Making thrift, pyhive and tableschema as extra_requires
Looking at the dependency tree for license related questions, I noticed
that tableschema had a huge tree, and only people running Hive really
need it. Making this as well as pyhive and thrift optional.
Also bumping some python dependencies
* Run pip-compile
* Removing refs to past.builtins (from future lib)
* Add thrift
* update contributing based on SIP-13
* update link
* add labels instruction
* minor edit
* minor edit
* update table of content
* Add ASF header and update insructions
* update table
* update text
* Conditionally mutate and quote sqla labels decouple sqla logic from viz.py
* Prefix hashed label with underscore if bigquery label exceeds 128 chars
* Add comments for label cache
* Rename to mutated_labels and simply
* Rename mutated_label to get_label and simplify make_label_compatible in db_engine_specs
* Add note about deterministic and unique mutated labels
* add hash to label that has been prefixed with underscore
* Fix PEP8 escape warning
* Fix DeckPathViz get_metric_label call
* [cosmetic] remove margin-bottom on dropdowns
react-select components have been added 15px bottom-margin without a
good reason. Other controls like text or checkboxes don't have that,
making Selects look non-standard, add making the control list too
sparse.
My perspective is that all controls should have the same
padding/margins.
* Adding a few licenses
* [cosmetic] CRUD's list from 'Add Filter' to 'Filter Results'
* Add global new button
* A better viz type selector
* lint
* Move view to new module, add unit test
* 'Filter Results' -> 'Filter'
* db merge
* Filter -> 'Filter List'
* [WiP] deprecate auto-generated metrics & fix examples
Picking up some leftover pieces not needed anymore since the
MetricsControl
* db merge
* db merge
* fix migration
* Creating metrics required for tests
User @ravi on Slack reported issues trying to use url_param jinja macro.
explore_json would raise a "KeyError: 'url_params'".
The feature works when a parameter is passed, but fails hard when none
are passed.
I improved the docstring on the macro as well.
The "altered" tag in the explore view shows up more often than it
should. By treating null, [] {}, undefined as identical will help reporting
only the differences that matter.
- Clean up css and html of the sql editor button area (removing unused styles)
- Restyle the button area and the resizer to be more like the one proposed in issue 6565
* resolved merge conflicts with upstream
* changed the key in a spot I missed
* linting
* refactored my code out into a function
* update util
* simplified the code logic
* cleaned up code
Refactored deprecated functionalities:
- Used 'celery worker' command instead of 'superset worker' which is
deprecated since 0.26.0
- Used 'gunicorn' command instead of 'superset runserver'
Changed margin for the slices description in dashboard view, previously the margin was 5px for all the sides but for improve the view just apply margin for top and bottom.
* Fix: updated required cryptography version to 2.4.2 to resolve#6509
* Fix: updated cryptography version to 2.4.2 to resolve#6509
* Fix: updated cryptography version to 2.4.2 to resolve#6509
* Revert "Fix: updated required cryptography version to 2.4.2 to resolve#6509"
This reverts commit b57b08c8e5.
* Fix: updated cryptography version to 2.4.2
* [SIP-14] remove dependency on yarn in favor of npm
Find more details here
https://github.com/apache/incubator-superset/issues/6217
* Add core-js to dev deps
* lint
* fix cypress
* break down cypress_build.sh
* fix cypress
* Use 'npm ci' in place of 'npm install'
* Bump npm req to >=6.5.0
* Enforcing npm@'>=6.5.0'
* Using latest lts of nvm
* Make cache settings global
* remove uneeded 'npm run test' as 'npm run cover' takes care of that
* Prefix with 'time'
* Trying to upgrade babel-eslint instead of downgrading
* upgrading babel-polyfill to '@babel/polyfill'
* Moving to babel 7
* remove unused packages
When a URL gets too long (usually happens with controls that allow
free form text), it creates an array of bugs.
* bug when creating a short URL, because the POST request's referrer
field is too long
* a bug when saving a chart (when the modal shows up) also because of
the referrer being too long
Some work has been done in the past to mitigate this, and I'm unclear if
it's some sort of regression because of the Flask upgrade or some
browser change, or whether these bugs have always been around.
This is a request for comment on an approach that works. Let me know if
you can think of better ideas as to how to manage this.
My current solution looks for 8000+ characters URLs and shrinks them to
only `/superset/explore/?URL_IS_TOO_LONG_TO_SHARE&form_data=...` and we
only keep the formData keys for `datasource` and `viz_type`. Not super
elegant but does the trick.
* [scheduled reports] Add support for scheduled reports
* Scheduled email reports for slice and dashboard visualization
(attachment or inline)
* Scheduled email reports for slice data (CSV attachment on inline table)
* Each schedule has a list of recipients (all of them can receive a single mail,
or separate mails)
* All outgoing mails can have a mandatory bcc - for audit purposes.
* Each dashboard/slice can have multiple schedules.
In addition, this PR also makes a few minor improvements to the celery
infrastructure.
* Create a common celery app
* Added more celery annotations for the tasks
* Introduced celery beat
* Update docs about concurrency / pools
* [scheduled reports] - Debug mode for scheduled emails
* [scheduled reports] - Ability to send test mails
* [scheduled reports] - Test email functionality - minor improvements
* [scheduled reports] - Rebase with master. Minor fixes
* [scheduled reports] - Add warning messages
* [scheduled reports] - flake8
* [scheduled reports] - fix rebase
* [scheduled reports] - fix rebase
* [scheduled reports] - fix flake8
* [scheduled reports] Rebase in prep for merge
* Fixed alembic tree after rebase
* Updated requirements to latest version of packages (and tested)
* Removed py2 stuff
* [scheduled reports] - fix flake8
* [scheduled reports] - address review comments
* [scheduled reports] - rebase with master
* Use py3's f-strings instead of s.format(**locals())
In light of the bug reported here
https://github.com/apache/incubator-superset/issues/6347, which seems
like an odd `.format()` issue in py3, I greped and replaced all
instances of `.format(**locals())` using py3's f-strings
* lint
* fix tests
* [SIP-5] Build metrics in query_object in the client
- Unify the metric interface (absorb the current plain string metric for built-in metric keys into the format used by adhoc metric)
- Port the logic in adhocMetric on the client and process_metrics in the backend to the new typed Metrics class
- Omit hasCustomLabel and formFromData properties from the new metric interface as their value can be inferred from label and optionName
- Expose from the Metrics class both metrics and their labels as public methods to match the all_metrics and metric_labels fields in the backend code
- Provide defaut values for filters, metrics and groupby in the backend
* addressing PR comments
* Adding a comment for metrictype values
* Revise LoadableRenderer (+3 squashed commits)
Squashed commits:
[f1614c7c] extract createLoadableRenderer into a separate function
[8689bc94] extend LoadableRenderer
[3d04ff2b] remove skipRendering
* add number of times function was called to the test
- Unify the metric interface (absorb the current plain string metric for built-in metric keys into the format used by adhoc metric)
- Port the logic in adhocMetric on the client and process_metrics in the backend to the new typed Metrics class
- Omit hasCustomLabel and formFromData properties from the new metric interface as their value can be inferred from label and optionName
- Expose from the Metrics class both metrics and their labels as public methods to match the all_metrics and metric_labels fields in the backend code
- Provide defaut values for filters, metrics and groupby in the backend
* Deprecate database attribute allow_run_sync
There's really 2 modes of operations in SQL Lab, sync or async
though currently there are 2 boolean flags: allow_run_sync and
allow_run_async, leading to 4 potential combinations, only 2 of which
actually makes sense.
The original vision is that we'd expose the choice to users and they
would decide which `Run` or `Run Async` button to hit.
Later on we decided to have a
single button and for each database to be either sync or async.
This PR cleans up allow_run_sync by removing references to it.
* Fix build
* Add db migration
* using batch_op
- Improve Docker image
- smaller
- faster to build
- deterministict dependencies (see #5958)
- Rework process to simplify setting things up
- updated documentation
- less commands to type
- no files to move and modify
- optional loading of samples
- Still working in standalone mode (without volumes for superset)
* [bugfix] deckgl legend is not interactive
Somehow when clicking categories in the legend, nothing happened. The
problem seemed to be around the `getDerivedStateFromProps` override
which gets triggered after each render and would alter the state and
reset the categories state to their origin.
Changed the method name to not be an override and just be called once
in the constructor
https://reactjs.org/docs/react-component.html#static-getderivedstatefromprops
* lint
* Minor improvements to Histogram viz
* prevent legend overflow (spill) by simply hiding it
* legend title only includes metric name when necessary
* control validator asking forcing at least one numeric column to be
selected
* Removing print()
* Adding cypress tests for showing erros, time range filter, and verbose name
* Adding tests for time range without ago
* Adding test to verify error message showing
* [bugfix] deck.gl grid&hex don't respond to zoom&drag
Following recent changes, Hex & Grid component wouldn't respond to
viewport interactions. This fixes the issue.
This PR also greatly improve rendering performance by only recomputing
the deck.gl layers when needed, as opposed to every time the viewport
changes, which is very inneficient.
Note that most other deck.gl layers seem to be systematically
recomputing layers, we'll have to fix this in another PR.
* Addressing comments
* [SIP-5] QueryBuilder in the client for groupby field
- Lay the structure of the QueryContext builder in the client
- QueryContext builder composes different portions of the queryContext object (to be sent to server from the client) from the datasourceBuilder and the queryObjectBuilder.
- The datasourceBuilder builds the datasource id and type by parsing them from the datasource field in formdata
- The queryObjectBuilder currently only builds the groupby field. It will further compose the queryObject from sub query builders in future PRs.
- Create a buildQuery method for WordCloud and override the groupby value with the value of series from formdata.
* Addressing PR comments
- Rename query builder files and their default exports
- Move tests into spec/javascripts/superset-ui for easy mass migration to superset-uiin the future
- Define viz specific formData and export formData for each viz type as intersection type of the generic formData and the viz specific one
- Specify the type signature for sqla and druid based data sources
* Addressing additional PR comments.
- Introduce a Datasource class and leverage Typescript's declaration merging of the interface
by the same name.
- Let Typescript infer the return type of various builders instead of specifying them explicitly
* Further tweaking the generic buildQueryContext method and viz speicific buildQuery methodes per PR comments.
* git mv a renamed file.
* addressing additional pr comments
* Update the description based on Ubuntu 16.04 with 18.04,
since Python version bundled with the former is 3.5,
which is not already supported
* Remove obsolete descriptions based on Python <= 3.5
Since https://github.com/apache/incubator-superset/pull/6287 and
effectively moving to a new version of d3, d3-format and d3-time-format
raises when receiving invalid input strings.
This code wraps the potential issues inside `try` blocks that will
effectively return an `ERROR` string as output to the formatting
function.
When hitting the Query button in the explore view, the previous chart
gets shown (flaskhed quickly) between the moment where the loading
spinner goes away and the new chart show up. This fix seems to prevent
this.
* Fix EventFlow transform is not a function
transformProps was being passed into EventFlowChartPlugin as object (module) as opposed to function. This commit changes the implementation to be consistent with the other ChartPlugins and fixes the typeError that causes EventFlow Chart to not display.
* Revert to lazy-loading transformProps
- Revert to lazy-loading transformProps
- Fix in Superchart will default to transformProps.default if available to allow for the lazy-loading
* Fix for lazy-loaded imports in ChartPlugins
* Fix lint error
* Add d3 micro packages
* Replace d3 imports with specific modules import
* Define d3 colors
* import specific d3 submodules instead of entire d3
* update function name
* move function location and fix small bug
* Move primary color to control
* remove colorscalefactory usage
* remove unused d3
* fix unit test
* fix color picker
* use @superset-ui/color
* update package version
* remove files that are extracted
* replace all references
* fix two files
* Revert some changes to split to another PR
* remove adaptor
* Address Christine's comment
* remove d3 v3 from calendar
* remove d3.scale.threshold
* Get rid of colorScalerFactory and revise hexToRGB
* fix color cleaning
* fix lint
- Datasource ids are being serialized into strings on the client side and python's default json.load doesn't deserialize strings representing integers as integers. Adding an integer parsing step when initializing QueryContext.
* Wrap <LoadableRenderer /> with <ErrorBoundary />
It appears that since the introduction of <SuperChart />, errors in the
visualization javascript (which are somewhat common and expected,
especially as we'll support plugins) were not handled and the whole
page would throw and go missing.
Here I'm introducing a new <ErrorBoundary /> component that elegantly
wraps other
components and handles errors. It's inspired by:
https://reactjs.org/docs/error-boundaries.html
The default behavior of the component is to simply surface the error
as an <Alert bsStyle="danger" /> and exposes the React stacktrace
when clicking on the error.
It's also possible to use component and pass an onError handler and not
show the default message.
This also fixes some minor bugs in TimeTable.
* Addressing comments
* Getting more stack traces
* Adressing comments
If you follow the documentation, the current docker instructions return an error when trying to connect via localhost. This seems to be due to the fact that it's running on port 8080, but forwarding port 8088. I suppose the cleaner solution is to do this, as opposed to changing both the compose file and the instructions... Hope this helps!!
* [SIP-5] Open a new /api/v1/query endpoint that takes query_obj
- Introduce a new handle_superset_exception decorator to avoid repeating the logic for catching SupersetExceptions
- Create a query_obj_backfill method that takes form_data and constructs a query_obj that will be constructed in the client in the future. Use the backfill in explore_json.
- Create a new /api/v1/query endpoint that takes query_obj only and returns the payload data. Note the query_obj is constructed in the client. The endpoint currently only handles query_obj for table view viz (we'll be adding support to new viz types as we go).
- Unit test to verify the new endpoint for table view
* fix tests and lint errors
* - Move the new query endpoint into its own api.py view.
- Create QueryObject and QueryContext class to encapsulate query_object to be built from the client and additional info (e.g. datasource) needed to get the data payload for a given query
- Remove the query_obj_backfill as we'll start building the first query_object on the client so it no longer makes sense to have a short-lived backfill for the matter of days.
* Fixing lint and test errors
* Fixing additional lint error from the previous rebase.
* fixing additional lint error
* addressing additional pr comments
* Make /query accept a list of queries in the query_context object.
* fixing a lint error
* - Move time_shift based calculation and since, until check into util
- Add typing info for get_since_until
- Add new unit tests to verify time_shift calculation and the since until check
* Wrap <LoadableRenderer /> with <ErrorBoundary />
It appears that since the introduction of <SuperChart />, errors in the
visualization javascript (which are somewhat common and expected,
especially as we'll support plugins) were not handled and the whole
page would throw and go missing.
Here I'm introducing a new <ErrorBoundary /> component that elegantly
wraps other
components and handles errors. It's inspired by:
https://reactjs.org/docs/error-boundaries.html
The default behavior of the component is to simply surface the error
as an <Alert bsStyle="danger" /> and exposes the React stacktrace
when clicking on the error.
It's also possible to use component and pass an onError handler and not
show the default message.
This also fixes some minor bugs in TimeTable.
* Addressing comments
* Add separate limit setting for SqlLab
Use separate param for wrap sql
Get query limit from config
unit tests for limit control rendering in sql editor
py unit test
pg tests
Add max rows limit
Remove concept of infinity, always require defined limits
consistency
Assert on validation errors instead of tooltip
fix unit tests
attempt persist state
pr comments and linting
* load configs in via common param
* default to 1k
* change because error: E: The method driver /usr/lib/apt/methods/https could not be found.
要提交的变更:
修改: contrib/docker/Dockerfile
* 修改: contrib/docker/Dockerfile
* [build] fix pip install issues on OSX High Sierra
I think requirements.txt was out-of-sync as well.
Also had to:
export
LDFLAGS="-L/usr/local/opt/openssl/lib"
export
CPPFLAGS="-I/usr/local/opt/openssl/include"
export
PKG_CONFIG_PATH="/usr/local/opt/openssl/lib/pkgconfig"
* Fix click
* setup plugin and add SuperChart
* Integrate SuperChart into Chart.jsx
* Add vizType as class name
* Remove .slice_container
* add snakeCase to sanitize class name
* Remove old code to load charts
* Fix supportedAnnotationTypes
* Update AnnotationTypes, remove unnecessary imports and dependency from VIZ_TYPES
* remove index.js and update unit test
* resolve tree map issue
* fix issue with annotation types
* fix proptypes
* add )
* address a few comments
* create bound functions
* bound more functions
* add reselect
* use reselect for chartprops
* improve performance with reselect
* remove unused props
* Remove getFilters() and update table test
* Remove unused code
* Delete adaptors
* switch to react-loadable
* Rewrite with reloadable
* Add timeout back
* remove loading
* update table unit test
* remove pastDelay
* WIP
* WIP
* WIP
* WIP
* Fix color bucketing
* Fixed colors
* Fix no num categories selected
* Colors working
* Fix no metric selected
* Visual cues for selection
* Add unit tests
* Remove jest from deps
* Rename category to bucket
* Small fixes
* Fix lint
* Fix unit tests
* Remove duplicate hexToRGB
* Fix import
* Change order of arguments in getBuckets
* Refactor function signature
* [Dashboard] disable force refresh when chart is still loading
- disable force refresh for chart, when chart is still loading
- disable force refresh for dashboard, when any one of chart is loading
* remove css change (introduced by mistake)
* add integration tests for disable/enable force refresh
* [SIP-9] Introduce TypeScript
- Introduce TypeScript and co to both source and tests
- Define alias for src directory in both webpack config and jest config so we can avoid using long relative paths like ../../src in both source and tests
- Type check feature flags system to prevent typos of flag names
- Change the feature flags system and the flags on window instead of populating them through the state tree. When introducing the first SCOPED_FILTER feature flag, it became too difficult to pipe the flags through the state initializers and layers of components and containers (the resulting code is hard to read and has a handful of methods taking an additional feature flag map parameter). Given that feature flags don't change throughout the life time of the app, it is better to leave them on window for easy access than piping them through the global state tree, which is meant to store the state of the app which changes frequently.
- Add a barebone filter panel that only shows when the SCOPED_FILTER feature flag is on
* Remove unnecessary dev-dependency on gl
* - Adding linting for TypeScript files via tslint.
- Fixing linting for Javascript files importing Typscript files
- Also fix linting for Javascript files that now leverage the webpack alias for the src directory
- up Typescript and type def versions
* Rename src directory's webpack alias from @ to src to be more explicit.
* refactor color scheme
* Update data structure
* Update color scheme files
* wip
* convert all sequential schemes
* Update how color schemes are managed. Extend it for sequential schemes
* extract color setup into separate file
* Update imports
* update imports
* Add new functions to Registry
* Update ColorSchemeManager to extends Registry and update unit tests
* Add test for Registry
* Rename ColorSchemeManager to ColorSchemeRegistry
* Fix unit tests
* Update API
* Fix imports
* Add label field
* Fix reference to colors
* update SequentialScheme contructor
* Fix controls
* rename manager to registry
* Split sequential schemes into multiple files
* update sequential color labels
* add values and valuesAsPromise()
* use .values()
* rename basicChartInput to chartProps
* replace BasicChartInput with ChartProps class
* Add transformProps to deck.gl plugins
* Add space
* delete BasicChartInput
* add width and height as part of ChartProps and update transformProps
* remove width and height
* remove adaptors from deck.gl components
* remove deck.gl creataAdaptor
* use parent size to compute width
* simplify this.width() and this.height()
* remove unnecessary jquery
* fix indent and import
* adjust headerheight from 100 to 80
* fix test
* [utils] gathering/refactoring into a "utils/" folder
Moving current utils.py into utils/core.py and moving other *util*
modules under this new "utils/" as well.
Following steps include eroding at "utils/core.py" and breaking it down
into smaller modules.
* Improve tests
* Make loading examples in scope for tests
* Remove test class attrs examples_loaded and requires_examples
* [refactor] Migrate from Mocha+Chai to Jest
This change migrates all the existing unit tests
- to Jest's global expect and matchers from chai's imported expect, asserts and matchers.
- to Jest's describe/test from mocha's describe/it
The majority of the mechanical changes to tests are achieved through running jest-codemods. The only two note-worthy manual tweaks:
1. Setting a testURL of http://localhost in jest config and adjusting a few tests to leverage this value instead of relying on about:blank.
2. Re-enabling ExploreChartPanel_spec which was previously commented out as we cannot have empty tests with nothing in it with Jest. :)
This change also removes dependencies to Mocha and Chai.
* Remove the test:one command as it now does the same thing as test.
* Fixing lint errors. The diff looks large but is large done through `yarn run lint --fix`
The only noteworthy change is the one in eslintrc for tests. The env has been updated from mocha to jest.
* Adding eslint-plugin-jest and further modify tests.
- One small fix in sqllab's Timer Spec for a test that is not using the spy it created for testing.
- Deletion of a duplicated test caught by eslint-plugin-jest.
* - Make istanbul coverage work with Jest.
- Remove dependency on stand-alone istanbul and babel-istanbul as they're built-into jest. Yes!
* Attempt to fix dynamic imports in tests.
* run sequentially and log heap usage
* - tweaking maxworkers for travis and specifying coverageDirectory for codecov
- remove dynamic import in shim.js now that it is set in babelrc for tests only.
* add credit field
* change image size limit for url-loader
* create chord ChartPlugin
* Update config to include original file name in the output
* add calendar
* add country map
* add eventflow
* add thumbnails
* add eventflow, filterbox, force-directed and heatmap
* Add histogram, horizon, mapbox, paired t-test, parallel coordinates
* Create plugins for the rest one-off charts
* ignore chartplugin files from istanbul
* feat: import/export dashboards via cli
* style: fixed lint error
* test: added test for import and export util
* test: removing import test as it is causing integrity issues
Import is a wrapper around exist functionality so we can go ahead without a test or mock the actual db operation using https://docs.python.org/3/library/unittest.mock.html
And validate the wrapper operations only.
* test: remove test data file
* test: removed usage of reserved keyword id
* Fix: remove test that is causing unintended impact
* add unit tests
* add test structure
* add unit tests for Registry
* add LoaderRegistry unit test
* add unit test for makeSingleton
* add type check
* add plugin data structures
* simplify API
* add preset tests
* update test message
* fix lint
* update makeSingleton
* update Plugin, Preset and unit test
* revise Registry code
* update unit test, add remove function
* update test
* update unit test
* update plugin unit test
* add .keys(), .entries() and .entriesAsPromise()
* update test description
* Allow user to force refresh metadata
* fix javascript test error
* nit
* fix styling
* allow custom cache timeout configuration on any database
* minor improvement
* nit
* fix test
* nit
* preserve the old endpoint
* Deprecate support for Python < 3.6
This is a first step, beyond this we can:
* remove all from future imports
* remove 'six' lib as a dependency
* start using f-strings
* enjoy ourselves
* fix tox
* Rebasing
* fix
* Add integration test for world map
* add pie chart
* add area
* use should for assertion
* update area test
* update it message
* remove null params
* add pivot tests
* remove urlparams
* add dual_line
* add sunburst test
* add big number
* add sankey
* add bubble
* add box plot
* add treemap tests
* combine all vis under single test
* feat: import/export dashboards via cli
* style: fixed lint error
* test: added test for import and export util
* test: removing import test as it is causing integrity issues
Import is a wrapper around exist functionality so we can go ahead without a test or mock the actual db operation using https://docs.python.org/3/library/unittest.mock.html
And validate the wrapper operations only.
* test: remove test data file
* test: removed usage of reserved keyword id
* [feat] Feature flag system via config
Adding a feature flag system that is driven by superset_config.py. This change includes:
- Server side changes to specify a dedicated FEATURE_FLAG dictionary for listing feature flags. E.g.
```
FEATURE_FLAGS = { 'SCOPED_FILTER': true }
```
- Pass the new feature flags to client via bootstrap-data
- Client side changes to inject feature flags into the redux state tree for dashboard, explore view and SqlLab
- Client side refactor/clean up so the feature flags can be properly tested. Also avoid modifying incoming bootstrap data when creating initial state for the redux state tree
- Re-enable tests that were previously disabled for ExploreViewContainer
* Fix lint errors.
* Remove the partial attempt to get reference to src working in tests (so we don't have to write ../../../src and such in tests). This will in a separate PR.
* add configuration for optimizing webpack
* resolve webpack cli issue
* Update outdated packages
* extract brace as its own chunk
* update comments
* add list of libs to skip parsing
* remove some libs from list
* remove noParse
* remove lodash.throttle from dependency
* add babel-plugin-lodash'
* use lodash instead of underscore for isFunction
* switch underscore to lodash
* switch from underscore to lodash flatten
* Remove slugify and use kebabCase from lodash instead
* [explore] add "View samples" modal to action buttons
Also broke down the `View query` and `View results` as different
request so that viewing the query does not require fetching the results
anymore
* fix js tests
* lint
* [docs] iterating to get to ASF compliance on project site
https://whimsy.apache.org/pods/project/superset is a place to check on
whether we match the ASF's requirements
* Minor adjustments
* Add schema level access control on csv upload
* add db migrate merge point
* fix flake 8
* fix test
* remove unnecessary db migration
* fix flake
* nit
* fix test for test_schemas_access_for_csv_upload_endpoint
* fix test_csv_import test
* use security_manager to check whether schema is allowed to be accessed
* bring security manager to the party
* flake8 & repush to retrigger test
* address comments
* remove trailing comma
- Reorganize sections for better navigability
- Add table of contents
- Rework frontend assets section for clarity and DRY
- Rework translating section for clarity, add "Add Translations" contribution type
- Move release docs only useful for maintainers to RELEASING.md
- Other miscellaneous improvements
* reactify world map
* add createAdaptor
* fix typo
* add schema
* move directory
* remove unnecessary lines
* make setRef a function
* convert keys to camelcase
* add unit test
* update formatting
* add check for displayName
* pass width and height as separate inputs
Ignoring history.pushState handling errors as they are impredictable and
crash the application hard.
I'm thinking this may be because the payload object is too big and the
browser is unhappy about it. I'm not sure if the payload/state is ever
used (when hitting back?). The error seem to stop when replacing payload
by an empty object.
* [refactor] Remove dependency to personal fork of supercluster from mapbox visualizations
- Update dependency to reference the vanilla supercluster
- Clean up backend api call for mapbox vizzes to ensure a boolean is sent to indicate whether the viz includes custom metric for clustering
- Refactor of mapbox and its cluster overlay components to use vanilla supercluster and its recommeded way for handling clustering based on custom aggregations.
- Allow reclustering within the initial bounds on render in mapbox visualizations (stay true to old behaviors).
- Remove the median aggregation from available cluster label aggregators as there is no memory efficient way to implement this and it is unknown how often this feature is used
- Updating doc to mention the backward incompatible change re median
* Perform the check for has_custom_metric through `not None` to produce a boolean and rename the field reflect it is a boolean.
* Adds new macro to get filter values from "filters" and "extra_filters"
Adds test for filter_values macro
Adds doco for filter_values
Changes filter_values return type to be a list rather than string
* Makes return value type consistent
- filter_values always return a list
* move into folder and scaffold adaptor
* extract width and height
* remove jquery
* extract showBrush
* extract lineInterpolation
* extract xAxisFormat and yAxisFormat
* extract annotationData
* extract xTicksLayout and colorScheme
* extract showXXX
* extract x and y axis labels
* extract showminmax
* extract pie chart props
* extract area chart props
* extract logscale and yAxisBounds
* extract margin
* extract bubble props x,y,size
* extract contribution, comparisonType and color_picker
* remove the last fd.xxx
* remove unnecessary variables
* remove slice.container
* fix unit test reference
* Rewrite logic to compute max label lengths to use only d3, not jquery.
* extract annotationLayers and no more slice.xxx in nvd3vis
* rename x, y, size to xField, yField, sizeField
* use arrow function
* move tooltip function
* extract helper functions into utils
* remove unused argument
* fix height calculation and show bar labels
* rename function
* update unit test
* organize tooltip generator
* update line_multi
* Add proptypes for data
* list proptypes for data
* extract tooltip function for bubble chart
* rename variables
* remove console.log
* enumerate vizTypes and pieLabelType
* parse maxBubble
* use new color scale
* fix import"
* remove line
Somehow checking the type of a JSX-defined React component is not
straightforward and changes through versions of React.
Using the `name` attr addresses the issue for now.
* Break TimeTable into smaller pieces
* extract function to compute color
* Handle height and scrollbar
* sort out isGroupBy
* Set default values
* Specify proptypes for data
* rename fields and update proptypes
* Add default props
* remove commented line
* swap import
* Enable Teradata
New DB engine spec for Teradata:
- LimitMethod should be WRAP_SQL since Teradata does not supporting "LIMIT" clause (TOP)
- Timegrains for Teradata is added
* Update formatting to pass flake8 tests
Since the recent PRs around webpack 4 and reloading, FAB pages have been
flickering on load, where a themeless Superset is shown for a fraction
of a second until the bootstrap theme gets loaded up.
This addresses it by moving the theme JS to the head section of the html
page.
* Create new classes for handling categorical colors
* verify to pass existing unit tests
* separate logic for forcing color and getting color
* replace getColorFromScheme with CategoricalColorManager
* organize static functions
* migrate to new function
* Remove ALL_COLOR_SCHEMES
* move sequential colors to another file
* extract categorical colors to separate file
* move airbnb and lyft colors to separate files
* fix missing toFunction()
* Rewrite to support local and global force items, plus namespacing.
* fix references
* revert nvd3
* update namespace api
* Update the visualizations
* update usage with static functions
* update unit test
* add unit test
* rename default namespace
* add unit test for color namespace
* add unit test for namespace
* start unit test for colorschememanager
* add unit tests for color scheme manager
* check returns for chaining
* complete unit test for the new classes
* fix color tests
* update unit tests
* update unit tests
* move color scheme registration to common
* update unit test
* rename sharedForcedColors to parentForcedColors
* remove import
* Pan mapbox map on query
- Modify the backend to send back the bounds of the points to cluster
- Clean up of mapbox component to derive lat, lon and zoom from the mercator that fitsbounds based on the bounds passed from the backend. This and the computation of clusters are now only done on initial load b/c we are not making queries to the backend on pan/zoom
- A minor clean up of unused props and also making the visual properties panel open by default
* Remove default latitude, longitude and zoom props which are unnecessary for computing the viewport now that it is driven by the bounds from the backend.
* a few small tweaks (rename vars to be sneak case and avoid two extra hash lookups)
* Bumping react==16.4.1 & enzyme==3.3.0
The upgrade was pretty smooth except for a cryptic message coming
out of react-select around running multiple copies of React. It turns
out the `common` bundle had React and was conflicting with explore and
dashboard apps, only in 16.x. This somehow wasn't a problem in 15.x
outside of the wasted resources.
Running 16.4 should bring in all sorts of perf improvements and features
we've all been waiting for.
https://reactjs.org/blog/2017/09/26/react-v16.0.html
TODO: react-bootstrap-datetimepicker isn't compatible with React 16
* Trying to deprecate react-bootstrap-datetime
* Moving forward
* Reintroducing tests
* [sql_json] allow not specifying client_id
We're opening the sql_json endpoint at Lyft to other apps leveraging
Superset as a data-access layer that enforces authentication and our data
access policy.
Currently sql_json requires the client to pass a `client_id` parameter
that uniquely identifies the query, that can then be used for polling
when in async mode. This PR makes it such that you don't have to define
a client_id anymore. It just gets generated when not passed.
* adressing comments
* Break MapBox into smaller pieces
* Replace React.createElement with regular jsx
* detach setControlValue
* enable render trigger
* Pass explicit props rather than pass all that exists in payload.data. Also use formData when possible.
* Rename sliceWidth, sliceHeight to width, height. Use deconstructor. Extract function.
* use arrow function
* fix linting and remove css
* Extract slice and formData
* indent
* update data proptype
* enable theme
* remove legacy code
* rename file
* Add legend
* Implement WithLegend
* align legend items to the right for bottom position
* add line at end of file
* fix linting issues
* bug: don't show query overlay when panning mapbox maps
Since we don't want to prompt user to rerun query every time they pan a map, prevent the query overlay from showing by setting dontRefreshOnChange for viewport_latitude and viewport_longitude controls.
* Prevent query overlay on viewport zoom also.
* setup devserver with correct proxy
* dev server wroking with written manifest
* add comments
* Change webpack to default port 9000 and minor js formatting
* Use hash in development.
* write to disk = true
* Take ports as argument for dev-server
* update instructions
* update instructions
* add devtools
* use mode instead of NODE_ENV
* use minicssextract for prod (like before)
* remove empty line
* Replace dataframe label override logic with table column override
* Add mutation to any_date_col
* Linting
* Add mutation to oracle and redshift
* Fine tune how and which labels are mutated
* Implement alias quoting logic for oracle-like databases
* Fix and align column and metric sqla_col methods
* Clean up typos and redundant logic
* Move new attribute to old location
* Linting
* Replace old sqla_col property references with function calls
* Remove redundant calls to mutate_column_label
* Move duplicated logic to common function
* Add db_engine_specs to all sqla_col calls
* Add missing mydb
* Add note about snowflake-sqlalchemy regression
* Make db_engine_spec mandatory in sqla_col
* Small refactoring and cleanup
* Remove db_engine_spec from get_from_clause call
* Make db_engine_spec mandatory in adhoc_metric_to_sa
* Remove redundant mutate_expression_label call
* Add missing db_engine_specs to adhoc_metric_to_sa
* Rename arg label_name to label in get_column_label()
* Rename label function and add docstring
* Remove redundant db_engine_spec args
* Rename col_label to label
* Remove get_column_name wrapper and make direct calls to db_engine_spec
* Remove unneeded db_engine_specs
* Rename sa_ vars to sqla_
* update indent
* extract formData and data.
* take filter
* remove dependencies
* remove removeFilter
* add comment
* remove columnFormats and verboseMap from props. clarify a few more props
* fix linting issue
* minor syntax
* syntax fix
* Move check to adaptor
* update unit test
* remove code related to .widget
* rename variables for clarity
* move Option fix to browser.js
* add column to annotation model
* update migration file
* change to Text
* remove old migration file
* add migration file
* remove JSON
* add new column to view
* add comma back
* linting
* lint some more
* missing comma
* rename columns
* add degrade and new migration file
* update version
* fixe changed name
* remove json from list columns
* Bug: fixing async syntax for python 3.7
Rename async to async_ so superset installs for python 3.7.
* Addressing PR comments. Use kwargs instead of explicitly specifying async_ so downstream engines (e.g. PyHive) that supports async can choose to use the async_ in pythonwq3.7 and async in <=python3.6
* addressing additional pr comments
* allow user to choose color
* add color control
* remove rgbToHex reference
* change to color picker control
* add d3-color
* remove module
* use d3-colors
* add semicolon
* [bugfix] geohash lat/long is reversed
This allows support for reversing geohashes. Note that the default option
was the wrong way.
* addressing comments
* make reverse_geohash_decode a staticmethod
* Field names in big query can contain only alphanumeric and underscore
* bad quote
* better place for mutating labels
* lint
* bug fix thanks to mistercrunch
* lint
* lint again
I noticed that `secondary_metric` was required in the chart that users
it, namely sunburst, parallel coordinates, and world_map.
I set the control as allowing null, and found related bugs in the
process:
* parallel coordinates did not have support for the new MetricControl
* added color scheme support for parallel coordinates
* added option to set row_limit on these viz types
* sunburst to support numeric columns (number would show as blank)
* Remove "for"
* add space
* Separate control to select table from database and schema.
* Adjust schema displays
* Fix caret and arrow position in Select and Tab
* Reduce space after caret in tab header
* Use translator
* Align icons in the pop-up menu in Sql Lab
* Add new table in front of the list (so it will appear on top)
* shorten message
* reduce line
Somehow we have a "IN" filter where the filter is a string, not an
array. While this may need to get fixed upstream as well, this prevents
the explore view from completely crashing.
Side note: this function looked somewhat brittle, I'm assuming it's used to keep
the free form SQL tab in sync and not used to generate the actual SQL
being executed.
* Fixing issue with extra params in formData
* Pass in param use_slice_data to decide whether to use slice data
* Fixing core_tests to not use explore_json overrides
I think that the only place where we want db connection pooling would be
to talk to the metadata database. SQL Lab should close its connections
and never pool them.
Given that each Gunicorn worker will create its own pool that can lead
to way too many connections opened.
closes https://github.com/apache/incubator-superset/issues/4666
* got skeleton started
* added d3-scale-chromatic to package.json
* got hex values instead of calling from a function
* got rid of d3-scale-chromatic - no longer needed
* added schemas to controls
* damn editor broken some line spacing
* commit
* fix style issues
* whyyyyy won't this build
* whyyyyy won't this build
* damn typo
* hahaha got editor to deal with style configs
* no i guess i didn't
* gotta get them all
* again
* trying to get docker build ot work
* updated installation docs with some osx instructions
* restoring yarn.lock not sure why it changed
* trying to fix indent
* trying again
* CODE STYLE CHANGES WORK
* removing some colors that are too close to white
* human readable labels for names
* human readable labels for names
* Allow min/max value for the sparkline in time series table
* show bound lines
* User can choose to show y-axis bounds
* show label for the bounds
* compute necessary padding for the bound label
* extract sparkline code to another component
* can show y-axis in sparkline without setting bounds
* reorder option rows
* Template dashboard
* Fix MySQL test
* Model for user attributes
* Redirect to welcome dash
* Fix lint
* Add missing file
* Add migration script
* Fix lint
* Fix more lint
* update python version for testing to 3.6
* CONTRIBUTING.md states pull requests should work on py27, py36+
* CONTRIBUTING.md states pull requests should work on py27 and py36
* Improve URLs for Chart and Dashboard ModelViews
Prior to this, the ModelView for Chart and Dashboard would be
at `/slicemodelview/list/` and `/dashboardmodelview/list/`.
Now we have cleaner URLs at `/chart/list/` and `/dashboard/list/`
* Fix unrelated js lint
* addressing comments
* Add function to fix dataframe column case
* Fix broken handle_nulls method
* Add case sensitivity option to dedup
* Refactor function definition and call location
* Remove added blank line
* Move df column rename logit to db_engine_spec
* Remove redundant variable
* Update comments in db_engine_specs
* Tie df adjustment to db_engine_spec class attribute
* Fix dedup error
* Linting
* Check for db_engine_spec attribute prior to adjustment
* Rename case sensitivity flag
* Linting
* Remove function that was moved to db_engine_specs
* Get metrics names from utils
* Remove double import and rename dedup variable
* Re-add dashboard short links
* Make the short link available to all users
* Include filters in the short link for dashboards
* Remove duplicate key causing linter error
* Change URL Short link button into a menu item with Modal
* Split out tests, bind URL short link in constructor
* [sql lab] simplify the visualize flow
The "visualize flow" linking SQL Lab to the "explore view" has never
worked so great for people, here's a list of issues:
* it's not really clear to users that their query is wrapped as a
subquery, and the explore view runs queries on top of it
* lint + fix tests
* Addressing comments
support toggling trend line
experiment using svg to measure size instead of canvas
refactor BigNumber with chart sticked to the bottom
made header line stick to bottom and css adjustment
remove commented code
fix svg text size estimation
remove unused code and round dimensions
handle missing getbbox
remove vx/text from dependency
ensure svg deletion after measurement
add comment to css file.
Add `positive` and `negative` class based on diff.
Add default props.
rename variable
accept container as argument and remove redundant font family
refactor visutils for consistent api
update xy-chart
update xy-chart to alpha1.0
fix pointseries
remove points
update yarn.lock
[4c9c01d7] update xy-chart version
remove unused import
* Add interim grains
* Refactor and add blacklist
* Change PT30M to PT0.5H
* Linting
* Linting
* Add time grain addons to config.py and refactor engine spec logic
* Remove redundant import and clean up config.py
* Fix bad rebase
* Implement changes proposed by @betodealmeida
* Revert removal of name from Grain
* Linting
* [sql lab] extract Hive error messages
So pyhive returns an exception object with a stringified thrift error
object. This PR uses a regex to extract the errorMessage portion of that
string.
* Unit test
* Visualization Unicode bug fix
* Fix the build (#5403)
The travis build has been failing for 2 reasons recently
* pylint takes > 10 minutes without outputing
* bad merge confict auto resolve in controls.jsx
* Visualization Unicode bug fix
* Move flake8-related packages deps to reqs-dev.txt
My VIM which is integrated with flake8 wouldn't match the output from
travis and would often miss things related to the flake8 plugins
installed using Tox.
By moving this to requirements-dev.txt, we can expect developers would
have the proper configuration locally and get matching results with
Travis when running flake8 or in their IDEs if its integrated with
flake8..
* merging migratinos
* sorting packages
* Specify folder for flake8 processing
* pin pycodestyle==2.3.1
* merge db migrations
Somehow it's not possible to set the row_limit for a line chart,
resulting to a 10k cap in our environment. In some cases the user may
want more than that.
The flow to import a table definition in Superset is confusing, user may
think they are creating a table or what not. This makes the flow a bit
more clear.
* Migrate flask_script to the Flask built-in click.
Flask 0.11 is the built-in integration of the click command line interface.
Flask-Migrate support for the new Flask CLI based on Click after Release 2.0.0.
* Resolved merge conflicts.
* Fixed issue introduced from bad merge.
* Fixed flake8 errors, added build to excluded flake8 stuff.
* * Moved the FlaskGroup declaration to the driver script.
* Moved shell context definition to cli.py
* Switched shell context definition to use decorator.
* Moved create_app definition to cli.py
* Fixed InvocationError with a wrapped function
* Added extra newlines between functions
* Removed flask-script dependency.
The dashboard page bootstrap data currently attempts to generate the
`SELECT` statement with column name details for each table represented
in the dash. This means it calls the database(s) and waits for column
details prior to returning any HTML.
This makes the default select_star property generate a simple
`SELECT *` with no column details instead, which doesn't require
interogating the DBs.
* [perf] add webpack 4 + SplitChunks + lazy load visualizations (#5240)
* [webpack] setup lazy loading for all visualizations
* [lazy-load] push renderVis function to <Chart /> state
* no mapbox token
* [lazy loading] use native webpack import func to fix chunk names, add babel-plugin-syntax-dynamic-import, fix rebase bug.
* fix geojson import, undefined t, and fix async css bug
* [lazy load] actually add babel-plugin-syntax-dynamic-import
* [webpack] working dev version of webpack v4
* [webpack 4] fix url issues, use mini-css-extract-plugin and webpack-assets-manifest plugins
* [webpack 4] use splitchunks for all files, update templates to multi-file entrypoints
* [webpack 4] multiple theme entry files for markup vis css, don't uglify mapbox
* [webpack 4] lint python manifest changes, update yarn lock.
* [webpack 4] fix tests with babel-plugin-dynamic-import-node
* [webpack 4] only use 'dynamic-import-node' plugin in tests, update <Chart /> vis promise when vis type changes
* [webpack 4] clean up package.json and yarn.lock after rebase
* [webpack 4] lint?
* [webpack 4] lint for real
* [webpack 4][istanbul] ignore visualizations/index.js
* [webpack 4] fix rebase bug, update dashboard/deprecated/chart/Chart.jsx to use vis promises.
* [webpack 4] lint and test
* [webpack 4] yarn.lock
* WIP
* Working version
* Clean code
* Fix lint
* Fix unit test; show only for sqla
* Working on UX
* Dropdown working 66%
* Working but needs CSS
* Fixed table
* Fix lint
* Fix unit test
* Fix languages path
* Fixes
* Fix Javascript lint
Moving to using Twine to upload to pypi and fixing up the markdown
support so that the page on Pypi looks like the README on Github.
This has been tested on the 0.26 branch starting 0.26.3
* Fix display limit in sql lab
* Add redundant variable from config.py
* Fix test config
* Change to SQL_MAX_ROW
* Remove last remaining occurence of DISPLAY_SQL_MAX_ROW
The travis build has been failing for 2 reasons recently
* pylint takes > 10 minutes without outputing
* bad merge confict auto resolve in controls.jsx
* [toasts] get rid of notify globals, refactor messageToasts for use by entire app
* [remove notify] use arrow func in ajax call
* fix lint + tests
* actually fix tests from messageToast refactor
* add 'test:one' npm script
* debugger
* [toasts] convert bootstrap flash messages to toasts in explore + sqllab
* [toasts][tests] import from right file
Pie chart was using `limit` instead of `row_limit` which didn't work.
Also set a lower default limit to 25 to prevent the chart from getting
super crowded.
Unrelated drive-by fix of setting "percentage metric" to default to `[]`
* Fix how the annotation layer interpretes the timestamp string without timezone info; use it as UTC
* [Bug fix] Fixed/Refactored annotation layer code so that non-timeseries annotations are applied based on the updated chart object after adding all data
* [Bug fix] Fixed/Refactored annotation layer code so that non-timeseries annotations are applied based on the updated chart object after adding all data
* Fixed indentation
* Fix the key string value in case series.key is a string
* Fix the key string value in case series.key is a string
* [Bug fix] Divide by 1000.000 in epoch_ms_to_dttm() to not lose precision in Presto
* [Bug fix] Divide by 1000.000 in epoch_ms_to_dttm() to not lose precision in Presto
* Improve database type inference
Python's DBAPI isn't super clear and homogeneous on the
cursor.description specification, and this PR attempts to improve
inferring the datatypes returned in the cursor.
This work started around Presto's TIMESTAMP type being mishandled as
string as the database driver (pyhive) returns it as a string. The work
here fixes this bug and does a better job at inferring MySQL and Presto types.
It also creates a new method in db_engine_specs allowing for other
databases engines to implement and become more precise on type-inference
as needed.
* Fixing tests
* Adressing comments
* Using infer_objects
* Removing faulty line
* Addressing PrestoSpec redundant method comment
* Fix rebase issue
* Fix tests
When receiving a VARBINARY field out of Presto, it shows up as type
`bytes` out of the pyhive driver. Then the pre 3.15 version of
simplejson attempts to convert it to utf8 by default and it craps out.
I bumped to simplejson>=3.25.0 and set `encoding=None` as documented
here
https://simplejson.readthedocs.io/en/latest/#basic-usage so that we can
handle bytes on our own.
As Superset extends flask SecurityManager with its own implementation, it's not obvious how to connect Superset with OAuth2 authorization servers that are not covered under flask.
* [webpack] setup lazy loading for all visualizations
* [lazy-load] push renderVis function to <Chart /> state
* no mapbox token
* [lazy loading] use native webpack import func to fix chunk names, add babel-plugin-syntax-dynamic-import, fix rebase bug.
* fix geojson import, undefined t, and fix async css bug
* [lazy load] actually add babel-plugin-syntax-dynamic-import
* [webpack] working dev version of webpack v4
* [webpack 4] fix url issues, use mini-css-extract-plugin and webpack-assets-manifest plugins
* [webpack 4] use splitchunks for all files, update templates to multi-file entrypoints
* [webpack 4] multiple theme entry files for markup vis css, don't uglify mapbox
* [webpack 4] lint python manifest changes, update yarn lock.
* [webpack 4] fix tests with babel-plugin-dynamic-import-node
* [webpack 4] only use 'dynamic-import-node' plugin in tests, update <Chart /> vis promise when vis type changes
* [webpack 4] clean up package.json and yarn.lock after rebase
* [webpack 4] lint?
* [webpack 4] lint for real
* [webpack 4][istanbul] ignore visualizations/index.js
* [explore] fix autocomplete on verbose names
Currently when searching for metrics or groupbys, the autocomplete
search functionality only matches based on the metric_name, though in
some cases the verbose_name is displayed and disregarded for search
purposes.
Also another issue is that not all pre-defined metrics show up in the
drop down which is confusing. Users may have simple metrics for which
they setup a nice verbose name and/or description and expect to see
those in the dropdown.
This PR addresses it for metric and column-related dropdowns.
* Add unit test
* [webpack] setup lazy loading for all visualizations
* [lazy-load] push renderVis function to <Chart /> state
* no mapbox token
* [lazy loading] use native webpack import func to fix chunk names, add babel-plugin-syntax-dynamic-import, fix rebase bug.
* fix geojson import, undefined t, and fix async css bug
* [lazy load] actually add babel-plugin-syntax-dynamic-import
* [webpack] working dev version of webpack v4
* [webpack 4] fix url issues, use mini-css-extract-plugin and webpack-assets-manifest plugins
* [webpack 4] use splitchunks for all files, update templates to multi-file entrypoints
* [webpack 4] multiple theme entry files for markup vis css, don't uglify mapbox
* [webpack 4] lint python manifest changes, update yarn lock.
* [webpack 4] fix tests with babel-plugin-dynamic-import-node
* [webpack 4] only use 'dynamic-import-node' plugin in tests, update <Chart /> vis promise when vis type changes
* [webpack 4] clean up package.json and yarn.lock after rebase
* [webpack 4] lint?
* [webpack 4] lint for real
* fetch datasources from broker endpoint when refresh new datasources
* remove get_base_coordinator_url as out of use
* add broker_endpoint in get_test_cluster_obj
This commit will try to dockerize superset in local development
environment.
The basic design is:
- Enable superset, redis and postgres service instead of using sqlite,
just want to simulate production environment settings
- Use environment variables to config various app settings. It's easy to
run and config superset to any environment if we use environment than
traditional config files
- For local development environment, we just expose postgres and redis
to local host machine thus you can connect local port via `psql` or
`redis-cli`
- Wrap start up command in a standard `docker-entrypoint.sh`, and use
`tail -f /dev/null` combined with manually `superset runserver -d` to
make sure that code error didn't cause the container to fail.
- Use volumes to share code between host and container, thus you can use
your favourite tools to modify code and your code will run in
containerized environment
- Use volumes to persistent postgres and redis data, and also
`node_modules` data.
- If we don't cache `node_modules` in docker volume, then every time
run docker build, the `node_modules` directory, will is about 500 MB
large, will be sent to docker daemon, and make the build quite slow.
- Wrap initialization commands to a single script `docker-init.sh`
After this dockerize setup, any developers who want to contribute to
superset, just follow three easy steps:
```
git clone https://github.com/apache/incubator-superset/
cd incubator-superset
cp contrib/docker/{docker-build.sh,docker-compose.yml,docker-entrypoint.sh,docker-init.sh,Dockerfile} .
cp contrib/docker/superset_config.py superset/
bash -x docker-build.sh
docker-compose up -d
docker-compose exec superset bash
bash docker-init.sh
```
* Improve time shift color and pattern
* Revert change
* Fix js unit test
* Move code to better place, add unit test
* Move classed code to backend
* Remove console.log
* Remove 1 hour time compare
* Remove unused import
By stop polling when presto query already finished.
When user make queries to Presto via SQL Lab, presto will run the query
and then it can return all data back to superset in one shot.
However, the default implementation of superset has enabled a default
polling for presto to:
- Get the fancy progress bar
- Get the data back when the query finished.
However, the polling implementation of superset is not right.
I've done a profiling with a table of 1 billion rows, here're some data:
- Total number of rows: 1.02 Billion
- SQL Lab query limit: 1 million
- Output Data: 1.5 GB
- Superset memory consumed: about 10-20 GB
- Time: 7 minutes to finish in Presto, takes additional 15 minutes for
superset to get and store data.
The problems with default issue is, even if presto has finished the
query (7 minutes with above profiling), superset still do lots of wasted
polling, in above profiling, superset sent about 540 polling in total,
and at half of the polling is not necessary.
Part of the simplied polling response:
```
{
"infoUri": "http://10.65.204.39:8000/query.html?20180525_042715_03742_nza9u",
"id": "20180525_042715_03742_nza9u",
"nextUri": "http://10.65.204.39:8000/v1/statement/20180525_042715_03742_nza9u/11",
"stats": {
"state": "FINISHED",
"queuedSplits": 21701,
"progressPercentage": 35.98264191882267,
"elapsedTimeMillis": 1029,
"nodes": 116,
"completedSplits": 15257,
"scheduled": true,
"wallTimeMillis": 2571904,
"peakMemoryBytes": 0,
"processedBytes": 40825519532,
"processedRows": 47734066,
"queuedTimeMillis": 0,
"queued": false,
"cpuTimeMillis": 849228,
"rootStage": {
"state": "FINISHED",
"queuedSplits": 0,
"nodes": 1,
"totalSplits": 17,
"processedBytes": 16829644,
"processedRows": 11495,
"completedSplits": 17,
"stageId": "0",
"done": true,
"cpuTimeMillis": 69,
"subStages": [
{
"state": "CANCELED",
"queuedSplits": 21701,
"nodes": 116,
"totalSplits": 42384,
"processedBytes": 40825519532,
"processedRows": 47734066,
"completedSplits": 15240,
"stageId": "1",
"done": true,
"cpuTimeMillis": 849159,
"subStages": [],
"wallTimeMillis": 2570374,
"userTimeMillis": 730020,
"runningSplits": 5443
}
],
"wallTimeMillis": 1530,
"userTimeMillis": 50,
"runningSplits": 0
},
"totalSplits": 42401,
"userTimeMillis": 730070,
"runningSplits": 5443
}
}
}
```
Superset will terminate the polling when it finds that `nextUri`
becomes none, but actually, when `["stats"]["state"] == "FINISHED"`,
it means that presto has already finished the query and superset can stop
polling and get the data back.
After this simple optimization, we get a 2-5x performance boost for
Presto SQL Lab queries.
* Bump celery to 4.1.1
Docs reference `celery worker --app=superset.sql_lab:celery_app
--pool=gevent -Ofair` command which seems only to work with Celery 4.1.1
* Add UPDATING.md message
* Added support for URLShortLinkButton to work for the dashboard case
* Fix lint errors and test
* Change references to 'slice' to 'chart'.
* Add unit tests to improve coverage
* Fixing lint errors
* Refactor to make URLShortLink more generic. Remove history modification code, redirect should be handling this.
* Remove history modification code, redirect should be handling this
* Generate a shorter link without the directory, and delegate default linked to the contents of window.location
* Fix lint errors
* Fix test_shortner test to check for new pattern
* Remove usage of addHistory to manipulate explore shortlink redirection
* Address build failure and using better practices for shortlink defaults
* Fixing alphabetical order
* More syntax mistakes
* Revert explore view history changes
* Fix use of component props, & rebase
Currently we assign release version number in release branches and
master was still pointing to some old version number from when the
process was different. We need a dummy version number that both setuptools
and npm are ok with.
* Initial test
* Save
* Working version
* Use since/until from payload
* Option to prefix metric name
* Rename LineMultiLayer to MultiLineViz
* Add more styles
* Explicit nulls
* Add more x controls
* Refactor to reuse nvd3_vis
* Fix x ticks
* Fix spacing
* Fix for druid datasource
* Rename file
* Small fixes and cleanup
* Fix margins
* Add proper thumbnails
* Align yaxis1 and yaxis2 ticks
* Improve code
* Trigger tests
* Move file
* Small fixes plus example
* Fix unit test
* Remove SQL and Filter sections
* Fix percent_metrics ZeroDivisionError and can not get metrics with label issue
* convert iterator to list
* get percentage metrics with get_metric_label method
* Adding tests case for expression type metrics
* Simplify expression
* Make port number optional in superset for druid
It appears that urllib throws error with ssl if port number is provided
```
url = "https://example.com:443/druid/v2"
req = urllib.request.Request(url, druid_query_str, headers)
res = urllib.request.urlopen(req)
```
The above call fails with the following error:
```
urllib2.HTTPError: HTTP Error 404: Not Found
```
If url is set to https://example.com/druid/v2 it works, this change
makes the port number optional.
* Rewrite if/else in concisely python way
* [WiP] make MetricsControl the standard across visualizations
This spreads MetricsControl across visualizations.
* Addressing comments
* Fix deepcopy issue using shallow copy
* Fix tests
* [sql lab] a better approach at limiting queries
Currently there are two mechanisms that we use to enforce the row
limiting constraints, depending on the database engine:
1. use dbapi's `cursor.fetchmany()`
2. wrap the SQL into a limiting subquery
Method 1 isn't great as it can result in the database server storing
larger than required result sets in memory expecting another fetch
command while we know we don't need that.
Method 2 has a positive side of working with all database engines,
whether they use LIMIT, ROWNUM, TOP or whatever else since sqlalchemy
does the work as specified for the dialect. On the downside though
the query optimizer might not be able to optimize this as much as an
approach that doesn't use a subquery.
Since most modern DBs use the LIMIT syntax, this adds a regex approach
to modify the query and force a LIMIT clause without using a subquery
for the database that support this syntax and uses method 2 for all
others.
* Fixing build
* Fix lint
* Added more tests
* Fix tests
* Force lowercase column names for Snowflake and Oracle
* Force lowercase column names for Snowflake and Oracle
* Remove lowercasing of DB2 columns
* Remove DB2 lowercasing
* Fix test cases
* use session context manager
* contextlib2 added to requirements.txt
* Fixing error: Import statements are in the wrong order. from contextlib2 import contextmanager should be before import sqlalchemy
* Fixing return inside generator
* fixed C812 missing trailing comma
* E501 line too long
* fixed E127 continuation line over-indented for visual indent
* E722 do not use bare except
* reorganized imports
* added context manager contextlib2.contextmanager
* fixed import ordering
* Changes "Import the dashboards." to "Import dashboards"
* Cleans up the HTML to add quotes, self close tags, etc.
* Adds a class to the `<submit>` button to utilize bootstrap style
* Remove the `<title>` tag in body as it's not vaild HTML and redundant with `{% block %}`
* add extraction fn support for Druid queries
* bump pydruid version to get extraction fn commits
* update and add tests for druid for filters with extraction fns
* conform to flake8 rules
* fix flake8 issues
* bump pyruid version for extraction function features
It appears the officially maintained fork of flask-cache is
flask-caching https://github.com/sh4nks/flask-caching . It is fully
compatible with flask-cache.
fixes https://github.com/apache/incubator-superset/issues/4926
In rare cases where the query is stopped before it is started, SQL Lab
returns an unexpected string payload instead of a normal dictionary.
This aligns the process to handle the error in a homogeneous fashion.
* Add helm chart to install superset in kubernetes
* set resources into unlimited
* Add descriptions to Chart.yaml
* add an entry in docs/installation.rst
if a filter is created on a chart, and the column is removed from the
dataset, you get a "'NoneType' object has no attribute
'is_num'" or something to that
effect. This fix disregards the filter.
Also error messages were HTML escaped which React does already anyways
so that's not necessary [anymore] here.
* [bugfix] temporal columns with expression fail
error msg: "local variable 'literal' referenced before assignment"
Error occurs [only] when using temporal column defined as a SQL
expression.
Also noticed that examples were using `granularity` instead of using
`granularity_sqla` as they should. Fixed that here.
* Add tests
* Fix 'pip install .'
Fix error :
> flask-appbuilder 1.10.0 has requirement Flask-SQLAlchemy==2.1,
> but you'll have flask-sqlalchemy 2.3.2 which is incompatible.
> botocore 1.10.5 has requirement python-dateutil<2.7.0,>=2.1, but you'll
> have python-dateutil 2.7.2 which is incompatible.
* remove flask-sqlalchemy==2.1 from reqs.txt
* [formats] add better defaults for time + number formatting
* [formatDate] add tests for concise formatDate
* [nvd3] use verbose time format in tooltips
* [number format] improve number format description
* [formats] revert to .3s defaults, tweak number format preview
* [formats] default number vis to .3s
* [WiP] [explore] proper filtering of NULLs and ''
TODO: handling of Druid equivalents
* Unit tests
* Some refactoring
* [druid] fix 'Unorderable types' when col has nuls
Error "unorderable types: str() < int()" occurs when grouping by a
numerical Druid colummn that contains null values.
* druid/pydruid returns strings in the datafram with NAs for nulls
* Superset has custom logic around get_fillna_for_col that fills in the
NULLs based on declared column type (FLOAT here), so now we have a mixed
bag of type in the series
* pandas chokes on pivot_table or groupby operations as it cannot sorts
mixed types
The approach here is to stringify and fillna('<NULL>') to get a
consistent series.
* typo
* Fix druid_func tests
* Addressing more comments
* last touches
* fixed RSTs so images will show up on github
* fresh screenshots on main page
* removing irrelevant portions
* moved a set of sections under 'Misc'
* rebuilt the Gallery with all screenshots
* Filtering out SQLLab views out of table list view by default
This adds a `is_sqllab_view` flag in the "tables" table, and makes the
filters those out in the Tables list view.
The filter showing on top may be a bit confusing, but certainly less
than seeing lots of user generated views.
* flake
* Addressing comments
The default setting which would look for either `count` first and then
any metric stopped working when we landed MetricsControl. This mimics
the previous behavior
* Fix how the annotation layer interpretes the timestamp string without timezone info; use it as UTC
* [Bug fix] Fixed/Refactored annotation layer code so that non-timeseries annotations are applied based on the updated chart object after adding all data
* [Bug fix] Fixed/Refactored annotation layer code so that non-timeseries annotations are applied based on the updated chart object after adding all data
* Fixed indentation
* Fix the key string value in case series.key is a string
* Fix the key string value in case series.key is a string
* Add ISO duration to time grains
* Use ISO duration
* Remove debugging code
* Add module to yarn.lock
* Remove autolint
* Druid granularity as ISO
* Remove dangling comma
* [sql lab] preserve schema through visualize flow
https://github.com/apache/incubator-superset/pull/4696 got tangled
into refactoring views out of views/core.py and onto views/sql_lab.py
This is the same PR without the refactoring.
* Fix lint
* Disabled run query button if sql query editor is empty
* Removing unnecessary white space
* Fix failing test for sql props
* Adding sql variable into propTypes and defaultProps
* [bugfix] convert metrics to numeric in dataframe
It appears sometimes the dbapi driver and pandas's read_sql fail at
returning the proper numeric types for metrics and they show up as
`object` in the dataframe. This results in "No numeric types to
aggregate" errors when trying to perform aggregations or pivoting in
pandas.
This PR looks for metrics in dataframes that are typed as "object"
and uses pandas' to_numeric to convert.
* Fix tests
* Remove all iteritems
superset appends DRUID_TZ info to intervals but not to granularity which causes one day's data return as 2 days. This fix is also pass DRUID_TZ to granularity.
* Allowing limit ordering by post-aggregation metrics
* don't overwrite og dictionaries
* update tests
* python3 compat
* code review comments, add tests, implement it in groupby as well
* python 3 compat for unittest
* more self
* Throw exception when get aggregations is called with postaggs
* Treat adhoc metrics as another aggregation
* Fix deepequality logic
Zeroed-in on this while a Deck Scatter Plot chart was prompting for
refresh on-load.
I'm pretty sure using JSON.stringify as a proxy for deep equality is
wrong.
Not sure how to handle yarn.lock changes here. The changes on yarn.lock
are the result of "only" running `yarn add deep-equal`
* Adressing comments
* [sqllab] fix data grid's instant search function
It looks like any non-string type would break the search feature.
of `FilterableTable`
* Addressing comments
Turns out the SQL would only be committed to the redux store `onBlur`
event to avoid the laggy typing. The delay come from the localStorage
binding that add enough millisecs of delay to feel odd while typing.
I now store the most recent SQL in the local and use that instead.
* move access permissions methods to security manager
* consolidate all security methods into SupersetSecurityManager
* update security method calls
* update calls from tests
* move get_or_create_main_db to utils
* raise if supersetsecuritymanager is not extended
* rename sm to security_manager
* Fix how the annotation layer interpretes the timestamp string without timezone info; use it as UTC
* [Bug fix] Fixed/Refactored annotation layer code so that non-timeseries annotations are applied based on the updated chart object after adding all data
* [Bug fix] Fixed/Refactored annotation layer code so that non-timeseries annotations are applied based on the updated chart object after adding all data
* Fixed indentation
* Use 'count' as the default metric when available
Count is a much better default than the current behavior which is to use
whatever the first metric in the list happens to be.
* Addressing nits
* Adding column type label to dropdowns
* Changing the style of column type label
* Adding tests for ColumnTypeLabel
* Adding tests for time and fixing if statement order
* Changing location of ColumnTypeLabel tests
* Updating ColumnTypeLabel structure
* [bugs] account for annotations in nvd3 x scale domain, fix dynamic width explore charts, allow 0 in text control
* tweak TextControl casting
* [annotations] filter separately from finding data extent
* Initial work
* Working version
* Specify legend position
* Max height with scroll
* Fix lint
* Better compatibility with nvd3
* Fix object.keys polyfill version
* Fix lint
* [sql lab] disable cross schema search
This is killing our metastore as people type it emits large
all-table-dump as they hit the keystroke. It never returns as it times
out and hammers the poor metastore.
Also some improvements around the disabling the table select on the left
panel and having the table name not be sticky.
* typo
* Adding option to visualize negative values in Table view
* Adding option for highlighting and right aligning
* Fixed typo
* Fixed case and condition
* Formatting
* Aligning left and default changes
* Changing default
* [Explore] Save url parameters when user save slices
* remove print
(cherry picked from commit bd9ecbe)
* add unit test
(cherry picked from commit 0f350ad)
* wrapping all request params into url_params
(cherry picked from commit 17197c1)
* Fix how the annotation layer interpretes the timestamp string without timezone info; use it as UTC
* Fix how the Interval annotation layer interpretes the timestamp string without timezone info; use it as UTC
* Expose hook to inject database connection logic on the fly
This environment configuration setting hook allows administrators to
alter the database connection parameters on the fly based on user
information. This can be use for a variety of purposes:
* rewire a subset of users to use different database user accounts
* pass user related information to the database for logging or QoS
purposes
* ...
* Fixes
* adding refresh chart overlay when chart is out of sync with control panel
* fading the visualization when stale
* addressing comments from team on layout of UI
* Introduce an onInit method for when a new viz_type is selected
This allows for clearing certain controls where/when needed. For
instance here, when loading deck_scatter, even if there was a time
granularity picked for the previous viz_type, we want to unselect it.
* making it functional
* Make filterbox i18n
* Change double-quote to single-quote in localization function t() to pass
lint test
* Updated .po file with italian translation. New strings generated and translated via babel-extract
* Add to history on instant control change
* Update latestQueryFormData on render triggered
* Add new message type
* Update latestQueryFormData in UPDATE_QUERY_FORM_DATA
* Remove comments from queries in SQL Lab that break Explore view
This fixes an issue where comments on the last line of the source query comment out the closing parenthesis of the subquery.
* Add test case for SqlaTable with query with comment
This test ensures that comments in the query are removed when calling SqlaTable.get_from_clause().
* Add missing blank line class definition (PEP8)
* Initial working prototype
* Small fixes
* Refactoring dekgl
* Show all data when no time grain is selected
* Refactor layers
* Standardize function name
* Fix exports
* Fix require
* Initial working prototype
* Small fixes
* Show all data when no time grain is selected
* Moving play bar to correct location
* Split component
* Working on CSS
* Remove control
* Positioning the play slider
* Fix refresh of slider state
* Fix lint
* Small fixes
* Smoother animation for scans
* Fix versions
* Play/pause with spacebar.
* Small fixes
* Clean stuff that went to other PRs
* Address issues
* Refactor scatter animation
Moving to a grid that is 4 times more granular. 48 columns.
Shipping a db migration script which will upgrade the position metadata
to reflect this change.
Also adapting the examples to the new grid parameters.
* [Explore view] Use POST method for charting requests
* fix per code review comments
* more code review fixes
* code review fix: remove duplicated calls for getting values from request
* [Explore view] Use POST method for charting requests
* fix per code review comments
* more code review fixes
* code review fix: remove duplicated calls for getting values from request
* [geo] introduce "Auto Zoom" control
On geospatial visualization, checking the "Auto Zoom" control makes it
such that the viewport is fitted to the data upon rendering the chart.
For dashboards with region filters, the map should jump to the right
position.
Eventually we should enhance this to fly and ease to the position in an
animated way.
* Added TODO notes
This fixes an issue with the newer Google's DBAPI library, which moved Row from google.cloud.bigquery._helper.Row to google.cloud.bigquery.Row, causing ImportError
* Switched yaml.load to yaml.safe_load to prevent code execution via crafted yaml files
Python's yaml.laod can lead to code execution via crafted yaml files such as:
```
code_exec: !!python/object/apply:subprocess.check_output ['ls']
```
* Fixed XSS via bleach
It was possible to get an XSS via the markdown library via simply setting a description containing arbitary HTML tags.
It was also possible to create links that went to the `javascript:` link handler (eg `[example](javascript:alert(0)`)
Using bleach to sanitize it solves both of these.
* Added XFO header by default to prevent clickjacking attacks
Note that with this application clickjacking can be relatively severe via the SQLLab functionality
which allows executing arbitary SQL.
* Added justification for dangerouslySetInnerHTML
* Fixed linting errors
* Fixed linting errors
* Add permissions checks for save_or_overwrite_slice
* Change is_owner to check_ownership
* Add translation to chart and dashboard error messages
* Appease the python linter by using single quotes
Druid sometimes returns error message that are contained in "<>", as in
`<urlopen error [Errno 61] Connection refused>`. Since Superset's
approach is often to bubble up messages coming from external library,
it's impossible to predict whether it will contain special characters.
There are some cases where our error handling does return some html
(presto?),
but we should manage that upstream. Plus the current setup has security concerns,
so let's move away from that.
fillna would miss out on identifying STRING columns for Druid and
replace None in string columns with a numeric `0`. This
mixed type column would confuse
pandas down the line on some operations like `df.pivot_table`.
Though flask has a builtin web debugger, ipdb some times still work
better. So I think add ipdb to dev dependencies is a good option for
people who prefer CLI debugging.
* cast postgresql date_trunc() to timestamp without time zone to prevent pandas error
* fix formatting for flake8
* change cast to timezone conversion instead
* Update installation.rst for Ubuntu 16.04 LTS
Ubuntu 16.04 by default install python2.7 alongside with python 3.5 and set python2.7 as default. If you have created a virtualenv with python3.5 compilation fails due to wrong python-dev library installed.
If you install ``python3.5-dev`` the build for the wheel package of ``cryptography`` run fine.
* Add options in D3_TIME_FORMAT_OPTIONS for non-english Time Format.
Added '%d/%m/%Y' and '%d/%m/%Y %H:%M:%S' as valid drop-down option for Axis Format
* [cli] permission cleanup on 'superset init'
FAB sometimes creates NULL/None permissions in the database,
presumably a race condition when multiple gunicorn workers start at the
same time, and those create issues raising "AttributeError: 'NoneType'
object has no attribute 'name'"
* Linting
Ubuntu 16.04 by default install python2.7 alongside with python 3.5 and set python2.7 as default. If you have created a virtualenv with python3.5 compilation fails due to wrong python-dev library installed.
If you install ``python3.5-dev`` the build for the wheel package of ``cryptography`` run fine.
* Use the query_obj as the basis for the cache key
When we recently moved from hashing form_data to define the cache_key
towards using the rendered query instead,
it made is such that non deterministic form
control values like relative times specified in "from" and "until" time
bound resulted in making those miss cache 100% of the time.
Here we move away from using the rendered query and using the query_obj
instead.
* Deprecating using form_data in templates
Filters applied to deck_multi will be passed down to layers as.
If the column isn't set as "filterable", the filter is ignored.
Also note that Dashboard configuration in regards to
"filter_immune_slices" and such will be disregarded in this context as
it isn't the dashboard controller passing down the filter and that
context is not easily accessible here.
Moving from having the user define an interceptor function that operates
on one object at a time.
By passing the entire array, it's possible to do multiple pass where
needed. A common pattern might be to figure out the max value in order
to define a scaler function. That's only possible if dealing with the
whole array.
* conditional check on datatype of results before converting to df
fix type checking
fix conditional checks
remove trailing whitespace and fix df_data fallback def
actually remove trailing whitespace
generalized type check to check all columns for dict
refactor dict col check
* move df conversion to helper and add unit test
add missing newlines
another missing newline
fix quotes
more quote fixes
Currently, even though `get_sqla_engine` calls get memoized, engines are
still short lived since they are attached to an models.Database ORM
object. All engines created through this method have the scope of a web
request.
Knowing that the SQLAlchemy objects are short lived means that
a related connection pool would also be short lived and mostly useless.
I think it's pretty rare that connections get reused within the context
of a view or Celery worker task.
We've noticed on Redshift that Superset was leaving many connections
opened (hundreds). This is probably due to a combination of the current
process not garbage collecting connections properly, and perhaps the
absence of connection timeout on the redshift side of things. This
could also be related to the fact that we experience web requests timeouts
(enforced by gunicorn) and that process-killing may not allow SQLAlchemy
to clean up connections as they occur (which this PR may not help
fixing...)
For all these reasons, it seems like the right thing to do to use
NullPool for external connection (but not for our connection to the metadata
db!).
Opening the PR for conversation. Putting this query into our staging
today to run some tests.
* Working polygon layer for deckGL
* add js controls
* add thumbnail
* better description
* refactor to leverage line_column controls
* templates: open code and documentation on a new tab (#4217)
As they are external resources.
* Fix tutorial doesn't match the current interface #4138 (#4215)
* [bugfix] markup and iframe viz raise 'Empty query' (#4225)
closes https://github.com/apache/incubator-superset/issues/4222
Related to: https://github.com/apache/incubator-superset/pull/4016
* [bugfix] time_pivot entry got missing in merge conflict (#4221)
PR here https://github.com/apache/incubator-superset/pull/3518 missed a
line of code while merging conflicts with time_pivot viz
* Improve deck.gl GeoJSON visualization (#4220)
* Improve geoJSON
* Addressing comments
* lint
* refactor to leverage line_column controls
* refactor to use DeckPathViz
* oops
Funky datatypes in some databases like BLOBs will have the DBAPI return
python types that can't be serialized to JSON out of the box.
Currently, when this happens SQL Lab fails in a bad way with a gigantic
HTML error message.
This allows specifying a pessimistic JSON serializer handler that will
simply show "Unserializable [type]"
* Using JS to customize spatial viz and tooltips
* Add missing deck_multi.png
* Improve GeoJSON layer with JS support and extra controls
* Addressing comments
When generating an auto SUM() metric on a column, Superset currently
will go `SUM(table_name.column_name)`. This is an issue when moving to
point to another table. It's common to work on some temporary table or
work table and eventually need to point Superset to an alternate table.
* Make Welcome page into a simple React app
This removes a dependency on datatables, we should be able to get rid
of it as we re-write the Table and PivotTable viz
* tests/lint
* Bump node version to latest
Some valid sqlalchemy uri's return a URL object with database=None, which causes the following error:
```
2018-01-05 17:59:47,560:ERROR:root:argument of type 'NoneType' is not iterable
Traceback (most recent call last):
File "/opt/incubator-superset/superset/sql_lab.py", line 186, in execute_sql
user_name=user_name,
File "/opt/incubator-superset/superset/utils.py", line 124, in __call__
return self.func(*args, **kwargs)
File "/opt/incubator-superset/superset/models/core.py", line 644, in get_sqla_engine
url = self.db_engine_spec.adjust_database_uri(url, schema)
File "/opt/incubator-superset/superset/db_engine_specs.py", line 505, in adjust_database_uri
if '/' in database:
TypeError: argument of type 'NoneType' is not iterable
```
This patch corrects that problem.
Delay creating DATA_DIR until config is fully imported.
This allows superset_config.py to override DATA_DIR before superset
attempts to create it in a potentially unwriteable location.
* Fix USA's state geojson for 'Country Map' visualization
Turns out the ISO codes were missing from the geojson file, this adds it
and uses human-readable indents.
* using proper ISO codes
* Linting
New linting rules started applying, I'm guessing a new version of
pylint?
* added deckgl geojson layer
* linting
* fixed comments
* addressed comments
* added override with controls.color_picker > 0
* set var properly
* set colors if property doesnt exist at all
* refacator on property mapping
* Introduce Javascript controls
This allows power-users to perform intricate transformations on data and
objects using javascript code.
The operations allowed are "sanboxed" or limited using node's vm
`runInNewContext`
https://nodejs.org/api/vm.html#vm_vm_runinnewcontext_code_sandbox_options
For now I'm only enabling in the line chart visualization, but the plan
would be to go towards offering more power to people who can write some
JS moving forward.
* Not applied
* Adding full Annotation Framework
* Viz types
* Re organizing native annotations
* liniting
* Bug fix
* Handle no data
* Cleanup
* Refactor slice form_data to data
For context, the viewport gets updated dynamically from the user
actions on the map. This is done on a timer every second or so to keep
the form data updated with the viewport settings.
With renderTrigger=true on that control that generates re-renders which
introduces glitches while zooming/panning. So turning it off as we don't
really expect users to input viewport info directly in the control
anyways.
Before this PR the only way to query lat/long is in the shape of 2
columns that contains lat and long.
Now we're adding 2 more options:
* a single column that has lat and long with a delimiter in between
* support for geohashes - geohashes are cool
* change reference for slices to chart
* change profile page reference
* change reference for Associated Slices
* change back to single quotes
* fix other single quotes
* linting
* last one
* fix test
The nvd3 docs say that it if the fastdom library is present it makes use
of it.
"Including Fastdom in your project can greatly increase the performance
of the line chart (particularly in Firefox and Internet Explorer) by
batching DOM read and write operations to avoid layout thrashing. NVD3
will take advantage of Fastdom if present."
* [sql lab] fix position of 'save query' Popover
The "Save Query" popover renders on the upper left corner as opposed to
where it should (relative to the Save Query button). After a bit of
research, it seems like Popover won't render in the right place when
parents are absolute.
I'm guessing this stopped working properly when I added the resizable
panes.
Anyhow, the solution here is to use a modal instead.
* Fixing tests
* fix 'superset db history'
Related Error msg when running `superset db history`:
"NameError: Can't invoke function 'get_bind', as the proxy object has not
yet been established for the Alembic 'Operations' class. Try placing
this code inside a callable."
* Lint
A bug in to_dict(orient="records") in pandas/core/frame.py prevents
datetimes with time zones to be worked with. This works around the
issue in superset by re-implementing the logic of pandas in the
correct way. Until pandas fixes the issue this code should stay.
https://github.com/pandas-dev/pandas/issues/18372
This closes#1929
Users can define `label_colors` in a dashboard's JSON metadata that
enforces a label to color mapping.
This also makes the function that maps labels to colors case insensitive.
* DECKGL integration
Adding a new set of geospatial visualizations building on top of the
awesome deck.gl library. https://github.com/uber/deck.gl
While the end goal it to expose all types of layers and let users bind
their data to control most props exposed by the deck.gl API, this
PR focusses on a first set of visualizations and props:
* ScatterLayer
* HexagonLayer
* GridLayer
* ScreenGridLayer
* Addressing comments
* lint
* Linting
* Addressing chri's comments
I was wondering what was left to do in order to remove Dashboard.jsx
and superset.js, and it looks like they can just be pulled out.
I am so happy to get rid of what used to be the messiest JS files in the
whole repo.
Thanks @graceguo!
* README: add the mailing list
And mark the google group as deprecated
* README: move the screenshots at the end
So hopefully it'll be easier to find the resources
I was getting some "Could not connect to server" when there was
a proper json payload with an `error` key, the change here makes sure to
prioritize those messages over the generic one.
Making it clear that master has a `dev`-suffixed version. Proper release
number will be created in release branches.
There are constraints as to what npm and setuptools will accept as a
proper version number and N.N.Ndev seems to work so I'm rolling with it
Open to suggestion as to how to tag `master`
I was having issues getting an RPM to build. My error was
Processing files: superset-0.20.1-1.noarch
error: File must begin with "/": Lockup
error: File must begin with "/": With
error: File must begin with "/": Text.svg
error: File must begin with "/": Lockup
error: File must begin with "/": With
error: File must begin with "/": Text@2x.png
error: File must begin with "/": Lockup
error: File must begin with "/": Without
error: File must begin with "/": Text@1x.svg
error: File must begin with "/": Lockup
error: File must begin with "/": Without
error: File must begin with "/": Text@2x.png
error: File must begin with "/": Mark.png
error: File must begin with "/": Mark@1x.svg
for ref https://github.com/pypa/setuptools/issues/767
File renaming fixes this issue
In `assets/static` we have a symlink pointing to `docs/_build/html`
which is .gitignored and missing in the repo since we don't want the
html-generated docs in the repo, meaning the symlink is broken
Having the symlink broken creates issues with pip-compile somehow.
Adding a dummy file here addresses that
fixes https://github.com/apache/incubator-superset/issues/3507
This prevents malformed JSON from getting saved in a slice's params. It
also prevents the issue described in #3507 from happening though as a
result malformed slices will render using default control values.
* Fix the ISO code description of region/province/department
* Missing the data of one province and two regions of China
* Revert " Missing the data of one province and two regions of China"
This reverts commit abff4555cd.
* Add srcdoc-polyfill tosupport Internet Explorer iframes in markup slices. Add allow-top-navigation and allow-popups to support links within iframes
* Remove jquery from markup.js
* ColorSchemeControl: fixing bad use of PropTypes
Accessing PropTypes via the main React package is deprecated, and will be
removed in React v16.0. Use the latest available v15.* prop-types package from
npm instead. For info on usage, compatibility, migration and more, see
https://fb.me/prop-types-docs
* Control: adding PropTypes.func in types allowed inside `value` prop
This removes a warning during yarn tests
Fix#3589
* tests(QueryStateLabel): removing missing prop warning
```
Warning: Failed prop type: The prop `query` is marked as required in
`QueryStateLabel`, but its value is `undefined`.
in QueryStateLabel
```
* SaveQuery: removing invalid prop `target` supplied to `Overlay`.
This removes a warning during yarn tests:
```
Warning: Failed prop type: Invalid prop `target` supplied to `Overlay`.
```
* RunQueryActionButton: removing `isRequired` from queryState props
This removes a warning during yarn tests:
```
Warning: Failed prop type: The prop `queryState` is marked as required in
`RunQueryActionButton`, but its value is `null`.
```
When the database impersonation flag is on, a query using a template
fails. It has to do with templating using a database connection without
a username being specified by the caller, along with the fact that the
work is taking place on a worker, outside a web request, where
referencing g.user raises this exception.
Somehow the nature of `g` in Flask has changed where `g.user` used to
be provided outside the web request scope and its not anymore.
The fix here should address that.
* Feature: added datepicker and time granularity options to dashboard filter
* Added option for Druid datasource time filters
* added more checkbox control over dashboard time filters
it was downgraded in commit 7045018d86.
The old version used React.PropTypes and it was deprecated.
This version is compatible with React 15+.
Fix#3588
* removing deprecated babel-preset-es2015 in favor of babel-preset-env
This removes this yarn warning during install:
```
warning babel-preset-es2015@6.24.1:
Thanks for using Babel: we recommend using babel-preset-env now:
please read babeljs.io/env to update!
```
* updating sinon to 4.0.0
This removes wrench from its dependencies and this yarn warning during install:
```
warning wrench@1.3.9: wrench.js is deprecated!
You should check out fs-extra (https://github.com/jprichardson/node-fs-extra)
for any operations you were using wrench for.
Thanks for all the usage over the years.
```
* Adding annotations to backend
* Auto fetching Annotations on the backend
* Closing the loop
* Adding missing files
* annotation layers UI
for https://github.com/apache/incubator-superset/issues/3502
* a few fixes per code review.
- add annotation input sanity check before add and before update.
- make SelectAsyncControl component statelesis, and generic
- add annotation description in d3 tool tip
- use less variable to replace hard-coded color
* exposed API endpoint to get querystring for a slice
* Added unit tests for endpoint
* fixed test case for python3
* moved get querystring logic into its own func
* renamed query string endpoint
* 1. after user make sql query and visualize, allow user click title to update slice title, and create a new slice at the same time.
2. don't save new title if it is empty. Will still show old title.
* change saveSlice call response and update explore view
- if current user is allowed to edit dashboard, we will allow this user to edit slice name.
- show different tooltip given allowed/not-allowed to update slice name.
- user will click slice name and update.
- after user submit edit, if he doesn't have right to alter slice, server-side will return error message to client-side. Slice name will not be changed or saved.
- will show notification after save slice name.
* Chinese page
* Using react-intl-universal to improve multi language in react page
* Using react-intl-universal to improve multi language in react page
* react_intl_universal
* change
* change
* change
* change
* change
* change
* change
* merge
* multiple page in js
* merge
* merge
* merge
* merge
* Js Translations
* JS Translation
* JS Translations
* Js translation
* JS translations
* JS translations
* Js translaion
* JS en Translation
* JS Translation
* upgrade document
Fixing the damn build (#3179)
* Fixing the build
* Going deeper
[bugfix] only filterable columns should show up in FilterBox list (#3105)
* [bugfix] only filterable columns should show up in FilterBox list
* Touchups
Datasource cannot be empty (#3035)
add title description to model view (#3045)
* add title description to model view
* add missing import
Add 'show/hide totals' option to pivot table vis (#3101)
[bugfix] numeric value for date fields in table viz (#3036)
Bug was present only when using the NOT GROUPED BY option
fixes https://github.com/ApacheInfra/superset/issues/3027
fix hive.fetch_logs (#2968)
add Zalando to the list of organizations (#3171)
docs: fixup installation examples code indentation (#3169)
[bugfix] fix bar order (#3180)
[bugfix] visualize flow error: 'Metric x is not valid' (#3181)
The metric name in the frontend doesn't match the one generated on the
backend. It turns out the explore view will default to the first
metric so specifying one isn't needed.
Fix the segment interval for pulling metadata (#3174)
The end of the interval would be on the truncated today date, which
means that you will exclude today. If your realtime ingestion job
runs shorter than a day, the metadata cannot be pulled from the
druid cluster.
Bump cryptography to 1.9 (#3065)
As 1.7.2 doesn't compile here with openssl 1.1.0f
Escaping the user's SQL in the explore view (#3186)
* Escaping the user's SQL in the explore view
When executing SQL from SQL Lab, we use a lower level API to the
database which doesn't require escaping the SQL. When going through
the explore view, the stack chain leading to the same method may need
escaping depending on how the DBAPI driver is written, and that is the
case for Presto (and perhaps other drivers).
* Using regex to avoid doubling doubles
[sqllab] improve Hive support (#3187)
* [sqllab] improve Hive support
* Fix "Transport not open" bug
* Getting progress bar to show
* Bump pyhive to 0.4.0
* Getting [Track Job] button to show
* Fix testzz
Add BigQuery engine specifications (#3193)
As contributed by @mxmzdlv on issue #945
[bugfix] fix merge conflict that broke Hive support (#3196)
Adding 'apache' to docs (#3194)
[druid] Allow custom druid postaggregators (#3146)
* [druid] Allow custom druid postaggregators
Also, fix the postaggregation for approxHistogram quantiles so it adds
the dependent field and that can show up in the graphs/tables.
In general, postAggregators add significant power, we should probably
support including custom postAggregators. Plywood has standard
postAggregators here, and a customAggregator escape hatch that allows
you to define custom postAggregators.
This commit adds a similar capability for Superset and a additional
field/fields/fieldName breakdown of the typical naming for dependent
aggregations, which should make it significantly easier to develop
approxHistogram and custom postAggregation-required dashboards.
* [druid] Minor style cleanup in tests file.
* [druid] Apply code review suggestions
* break out CustomPostAggregator into separate class. This just cleans
up the creation of the postaggregator a little bit.
* minor style issues.
* move the function around so the git diff is more readable
add combine config for metrics in pivot table (#3086)
* add combine config for metrics in pivot table
* change method to stack/unstack
* update backendSync
Autofocus search input in VizTypeControl modal onEnter (#2929)
Speed up JS build time (#3203)
Also bumping a few related libs
JS Translation
JS translations
js translation
fix issue 3204 (#3205)
[bugfix] capture Hive job_id pre-url transformation (#3213)
js translation
fix issue 3204 (#3205)
[bugfix] capture Hive job_id pre-url transformation (#3213)
[docs] update url in CONTRIBUTING.md (#3212)
[sqllab/cosmetics] add margin-top for labels in query history (#3222)
[explore] nvd3 sort values in rich tooltip (#3197)
[sqllab] fix UI shows 'The query returned no results' momentarily (#3214)
this is visible when running async queries between the fetching and
success state as the rows are getting cached in the component
[explore] DatasourceControl to pick datasource in modal (#3210)
* [explore] DatasourceControl to pick datasource in modal
Makes it easier to change datasource, also makes it such that the list
of all datasources doesn't need to be loaded upfront.
* Adding more metadata
* Js translation
* js tran
* js trans
* js trans
* js tran
* js trans
* js trans
* js tran
* js translation
* js trans
* js translation
* try load language pack async
* Backend translations things
* create language pack inside common data
* performance improvement for js i18n.
- js bundle should not contain localized content
- we populate translation content from server-side, in boostrap.common.language_pack
- in client-side, use promise to wrap around translation content. text will be translated after translation content arrived/parsed.
- fix linting
* fix Timer unit test
* 1. add global hook for all tests, to make translation pack avaialble before each test starts.
2. fix unit test for Timer component
3. remove noused method get_locale, and modules
4. fix page reload after user change page language
* parse and build i18n dictionary as a module
* fix sync-backend task, which should run without DOM
Since the latest pyhive, we don't need pythrifthiveapi as they ship with
the latest version.
There's actually a conflict between the new pyhive and pythrifthiveapi
and this fixes it.
* Add "Impersonate user" setting to Datasource
* Add tests
* Use g.user.username for all the sync cases
* use uri.username instead of uri.user
* Small refactoring
* Fixed bugs when removing filter, switching operators, and switching columns
* Fixed lint errors for code style
* Added more unit tests for FilterControl
* Code format changes to meet standards
* add lanscape?
* prevent empty queries from being sent to db
* Revert "add lanscape?"
This reverts commit b3d6e80af4.
* improve code style
* nit
* unnit
* renit
* [heatmap] numerous improvements
* flexibility as to how to sort X and Y axis (alpha/value, desc/asc)
* option to show a legend
* fixed margins, maximize real estate
* allowed users to define bounds
* Tunning
Currently the heatmap axis are sorted randomly, this PR makes it such
that it's properly sorted.
Also allowing for specifying the left and bottom margins
* Bumping npm
* Fixing the build
This addresses long standing issues and hacks around react-map-gl
that have been creating all sorts of build issues over time.
It appears that recent changes broke things further.
This PR upgrades to the latest `react-map-gl` and removes related hacks.
Previously all 'since' date expression evaluated in the future like
`30 days` would be reassigned to the past (now - `30 days`). It would
extend to fixed dates which is a bad thing and was removed.
Now we have reports and dashboards in the wild that use things like `30
days` and we'd like to not break those as we roll out the next version.
This fix should allow for that.
* [explore] fixed padding bug on filter section
* fix slice_id logging issue
* [superset-sqllab] fix slice_id population in appropriate column
* [explore-logging] test the slice_id logging fix
* fix travis errors
* fix nits pointed out in PR comments
* cleanup tests
* made python more beautiful
* made python even more beautiful
* made python even more more beautiful
* made python even more more more beautiful
* fix lint error
* make exception handling more specific
* fixed silly error
* fixed argument indentation
The left panel of the explore view has become crowded and overwhelming
overtime. This PR adds functionality to collapse the control sections,
and sets most sections to be collapse by default as the explore view
opens up.
* breakdown `Query` section for most viz
* bring filters to the top, under Query section
* collapse most sections by default
* removed confusing outdated description for Filter section
* Update to the version of react-ace with the fixed sizing issues
* Make ace editor resizable
* Use small util method for offset calculation instead of $
* Test ResizableAceEditor
* Make the right pane of the Sql Lab scrollable
* Add default and min height to the ResizableAceEditor
* Implement SplitPane
* Make Splitter fullscreen
* React on resize of the window
* Implement min and max
* Get rid of a magic number + add margin
* Handle resize event with delay + cleanup the code
* Make ResultSet adjustable
* Make QueryHistory adjustable
* Remove ResizableAceEditor
* Make linter happy
* Test SplitPane
* Init sizes properly
number of issues are addressed here:
* x axis formatting doesn't apply in bubble chart
* adding option to show/hide bounds of x and y axis
* with a flaky 'auto' margin mode, allow user to specify hard values or
auto
* x label font-size was different than y axis
* show more options and reorg Control panels for Axes in `line` viz
for chart type: bubble, area, treemap and country map, color scheme change cannot take effect immediately. So have to disable immediately re-render function and hide lightening icon for these chart type.
* Allow for multiple color schemes
1. create ColorSchemeControl component
2. using the same new control component for linear colors
* add color spectum for linear color scheme
* remove dup css
* fix controls setting for linear color scheme
* minor fix by code review comment
Style wasn't working right for separator widget since the iframe sandboxing of
the markup widget. This addresses this small issue and also now allows
for html in the separator widget
* [mapbox] fix viewport alterations
Since explorev2 it appears that altering the viewport hasn't been
changing the controls as it used to. This PR addresses it.
* lint
Somehow Ace's "changeSelection" event is triggered when hitting
backspace (and shouldn't!).
changeSelection on our side triggers enough work to make the
holding backspace sluggish and laggy.
This fix ignores selection with a length of 1, avoiding mutating the
state altogether when hitting/holding backspace.
* improve superset list view page layout
- less header spaces and stacks
- move pagination down to bottom
- apply material design style to 'add' action button
- will apply to all superset list view, like slices list, security tab lists etc.
* improve superset list view page layout
- less header spaces and stacks
- move pagination down to bottom
- apply material design style to 'add' action button
- will apply to all superset list view, like slices list, security tab lists etc.
* improve superset list view page layout
- less header spaces and stacks
- move pagination down to bottom
- apply material design style to 'add' action button
- will apply to all superset list view, like slices list, security tab lists etc.
* improve superset list view page layout
- less header spaces and stacks
- move pagination down to bottom
- apply material design style to 'add' action button
- will apply to all superset list view, like slices list, security tab lists etc.
* improve superset list view page layout
- less header spaces and stacks
- move pagination down to bottom
- apply material design style to 'add' action button
- will apply to all superset list view, like slices list, security tab lists etc.
* remove tabs from indentation
* fix merge conflicts
* adjust css after code merge
parse_human_datetime parses date-only strings, e.g. "today", returning the correct date but time set at 9am. This is an internal implementation in parsedatetime. This patch resets to midnight. If time is specified and parsed, it is correctly returned.
* add translatable columns in label_columns of the view
* display the verbose_name of columns in list view, just like in the metrics list
* Revert "display the verbose_name of columns in list view, just like in the metrics list"
This reverts commit f815d3b3ed.
* [add] Save filters to dashboard
* format code
* fix CI error
* add semicolon semi
* fix none object
* add test data
optimize the js code
fix the compatibility issue
* fix urllib to urllib.parse
* add space
* update test case
* remove 'return'
* fix error
* update test case
* split reducer logic for ExploreViewContainer
* fix saveModal component and unit tests
* revert changes in SaveModal_spec.
will make another commit just to improve test coverage for SaveModal component.
* improve test coverage for explore view components:
- SaveModal component
- URLShortLinkButton
* remove comment-out code
* [bugfix] wrong 'Cant have overlap between Series and Breakdowns' (#3254)
* [explore] make edit datasource a basic link (#3244)
* Relying on FAB for font-awesome.min.css (#3261)
* Modernize SQLA pessimistic handling (#3256)
Looks like SQLAlchemy has redefined the best practice around
pessimistic connection handling.
* [webpack] break CSS and JS files while webpackin' (#3262)
* [webpack] break CSS and JS files while webpackin'
* cleaning up some templates
* Fix pylint issue
* import logging (#3264)
* [bugfix] preserve order in groupby (#3268)
Recently in
https://github.com/apache/incubator-superset/commit/4c3313b01cb508ced8519a68f6479db423974929
I introduced an issue where the order of groupby fields might change.
This addresses this issue and will preserve ordering.
* Explicitly add Flask as dependancy (#3252)
* Use sane Celery defaults to prevent tasks from being delayed (#3267)
* Improve the chart type of Visualize in sqllab (#3241)
* Improve the chart type of Visualize in sqllab & Add some css & Fix the link address in the navbar
* add vizTypes filter
* Set default ports Druid (#3266)
For Druid set the default port for the broker and coordinator.
* [explore] Split large reducer logic in ExploreViewContainer (#3088)
* split reducer logic for ExploreViewContainer
* fix saveModal component and unit tests
* revert changes in SaveModal_spec.
will make another commit just to improve test coverage for SaveModal component.
* remove comment-out code
* fix merge confilicts
* split reducer logic for ExploreViewContainer
* fix saveModal component and unit tests
* revert changes in SaveModal_spec.
will make another commit just to improve test coverage for SaveModal component.
* remove comment-out code
* fix merge confilicts
* [explore] DatasourceControl to pick datasource in modal
Makes it easier to change datasource, also makes it such that the list
of all datasources doesn't need to be loaded upfront.
* Adding more metadata
* [druid] Allow custom druid postaggregators
Also, fix the postaggregation for approxHistogram quantiles so it adds
the dependent field and that can show up in the graphs/tables.
In general, postAggregators add significant power, we should probably
support including custom postAggregators. Plywood has standard
postAggregators here, and a customAggregator escape hatch that allows
you to define custom postAggregators.
This commit adds a similar capability for Superset and a additional
field/fields/fieldName breakdown of the typical naming for dependent
aggregations, which should make it significantly easier to develop
approxHistogram and custom postAggregation-required dashboards.
* [druid] Minor style cleanup in tests file.
* [druid] Apply code review suggestions
* break out CustomPostAggregator into separate class. This just cleans
up the creation of the postaggregator a little bit.
* minor style issues.
* move the function around so the git diff is more readable
* [sqllab] improve Hive support
* Fix "Transport not open" bug
* Getting progress bar to show
* Bump pyhive to 0.4.0
* Getting [Track Job] button to show
* Fix testzz
* Escaping the user's SQL in the explore view
When executing SQL from SQL Lab, we use a lower level API to the
database which doesn't require escaping the SQL. When going through
the explore view, the stack chain leading to the same method may need
escaping depending on how the DBAPI driver is written, and that is the
case for Presto (and perhaps other drivers).
* Using regex to avoid doubling doubles
The end of the interval would be on the truncated today date, which
means that you will exclude today. If your realtime ingestion job
runs shorter than a day, the metadata cannot be pulled from the
druid cluster.
The metric name in the frontend doesn't match the one generated on the
backend. It turns out the explore view will default to the first
metric so specifying one isn't needed.
Related Msg:
sqlalchemy.orm.exc.ObjectDeletedError: Instance '<PermissionView at
0x7f10306b0e90>' has been deleted, or its row is otherwise not present.
* upgrade celery to 4.0.2
* using Redis for unit tests (sqla broker not supported in Celery 4)
* Setting Celery's soft_time_limit based on `SQLLAB_ASYNC_TIME_LIMIT_SEC` config
* Better error handling in async tasks
* Better statsd logging in async tasks
* show [pending/running] query status in Results tab
* systematically using sqla NullPool on worker (async) to limit number
of database connections
* move to explore folder and delete explorev2 folder
* add tests for fetchDatasourceMetadata
* tests for fetchDatasources, fetchDatasourceMetadata
* use $.ajax for fetch dashboards and write test
* sort explicitly on label
* add simple test for /slicemodelview/add endpoint
* make comments and method names more clear
* fix test name
* be more explicit, test status_code
* get compiled js file names
* make manifest available as template var
* use script src directly to avoid flash of unstyled content in the case of csstheme.js
* linting
* attempt to fix tests
* exception
* print the path when no manifest file found
* handle case when manifest.json is not present for some reason, or in the case of tests
* Fix handling of Chunked requests
Add fix for handling chunk encoding requests.
If ENABLE_CHUNK_ENCODING is set to true, for requests with transfer
encoding set to true. It will set wsgi.input_terminated to true which
tells werkzeug to ignore content-length and read the stream till the
end.
break comment in multiple lines
* remove debug print logging
* dont set first datasource as default
* not used
* add a disabled class if datasource is not selected
* sort datasources alphabettically
* make btn disabled is no datasource is selected
* fix linting
* [big number] various improvements
* dynamic number of X axis ticks based on width to prevent label overlap
* corrected overflow on the x axis
* improved tooltips (precise arrow and visible data point circle on hover)
* Fixing tooltips in heatmap viz
* initial structure for add new slice page
* simplify add slice form
* add a test
* fix long line
* use underscore for template name
* fix controls path
* fix vis types select
Some python developers use Python version manager to create and switch Python
enviroments, pyenv (http://github.com/yyuu/pyenv) is one of the version managers.
When you use `pyenv local x.x.x`, `pyenv` will put a `.python-version` file in
project directory, which, in general, should not be commited to git.
Same as explore view, if query takes > 45 seconds not returned, we will show Query timeout warning message. Otherwise user will see 504 Gateway timeout error.
* CONTRIBUTING: document starter_task label
* CONTRIBUTING: PR should work travis python versions
* CONTRIBUTING: move PR guidelines at the top
* CONTRIBUTING: no need to open a new pr after review
in SqlLab view, if query takes over 45 seconds, we will show advise to store a summarized data set before user clicks on Visualize button.
This advise will not block Visualize button.
fixes https://github.com/airbnb/superset/issues/2733
* rename spec folder
* remove special handling for viz_type control since it now uses VizTypeControl
* add test for getOptions
* linting
* add test for cwp
* linting
The reason is that it should be able to create arbitrary filters
over attributes with very high cardinality, where loading all
possible combinations into the filter box is infeasible,
e.g. IPv6 addresses.
* feat(visualization): Create new visualization's tools that display country with geojson file. Currently only france and Spain are loaded
* refacto(visualization): fix issues created by codeclimate
* refacto(visualization): fix issues created by travis for mysql databases
* refacto(visualization): fix issues created by codeclimate
* refacto(visualization): fix issues created by codeclimate
* refacto(visualization): fix issues created by codeclimate
* refacto(visualization): fix issues created by codeclimate
* fix(visualization): fix issues
* doc(visualization): Init visualization documentation
* doc(visualization): init visualization documentation for country
* fix(visualization): implement iso 3166-2 for id of country
* fix(visualization): implement iso 3166-2 for id of country
* doc(visualization): init visualization documentation for country
* fix(visualization): implement iso 3166-2 for id of country
* feat(country_map): add ukraine map in component
* fix(visualization): Test dont working and add color based on metrics
* refacto(visualization): fix issues for es6 and color component
* Show clear and actionable query timeout error message
1. Instead of waiting for a long time or server-side response 504 Gateway timeout, explore view now add a query timeout threshold. After timeout it will show specific querytimeout message.
2. fix alert box close button position.
* Show clear and actionable query timeout error message
1. Instead of waiting for a long time or server-side response 504 Gateway timeout, explore view now add a query timeout threshold. After timeout it will show specific querytimeout message.
2. fix alert box close button position.
3. fix a linting error.
Load times on list view pages like Slices, Dashboards, Tables and Database
have grown to be terrible over time.
After a bit of digging, I found that the not specifying `search_columns`
in ModelViews actually means "all columns" and that for each filter,
FAB goes and fetches a list of all values to prepopulate the
filter dropdowns. That means that the list of tables would fetch all
slices and all users upfront which is horrible. Worse, database list
view would fetch all queries with is insane.
This picks a subset of columns for search/filters and changes the
default to show only 100 elements per page instead of 500
auto select should only apply to cases where select control doesn't have pre-selected value prop. If select control has pre-selected value (passed in from value prop), auto select first avaliable options should not apply.
* [sql lab] improvements to the left panel
* better error handling with error notifications
* table is added instantly with a loading image
* Fixing tests
* Fixed tests
* Decimal is a valid numeric type
This will allow automatic creation of sum__ and avg__ metrics
* formatting for line length at 80 chars
* Reformating again, for the mandatory 79 chars maximum.
Somehow the CSS trick we use to display histogram-type visual elements
in table views started showing some odd glitches since
what I believe was a not-so-recent version of webkit.
The bug is around the `linear-gradient` call under `background-image`
prop being used to show non-gradient or "hard" colors, while using
transparency.
I was able to find a workaround that addresses things in chrome by using
a fake [minimal] gradient instead of a flat color.
* db_engine_specs: added engine spec for clickhouse
* models: group by time without join on subquery
if this is enabled, it will remove the subquery when grouping by time,
and group by time will be rewritten as group by timeslot, ...
this particularly fixes clickhouse backend
* docs: added clickhouse
* models: allow using secondary dttm columns as constraints
* superset/forms: configurable default slice since time
* superset/models: respect dialect quoting
* dttm handling, codeclimate tweaks
* react: using prop-types package to fix deprecated React.PropTypes property warning
https://facebook.github.io/react/warnings/dont-call-proptypes.html
* removing babel devDependency because has been deprecated in favor of babel-cli
this fixes a warning during `npm install`:
```
npm WARN deprecated babel@6.23.0:
In 6.x, the babel package has been deprecated in favor of babel-cli.
Check https://opencollective.com/babel to support the Babel maintainers
```
* js: setting ExploreActionButtons.queryEndpoint PropType as required
because it's required in the child component DisplayQueryButton
* js(tests): using object in expandedSlices prop type of SliceCell tests
* js(tests): adding required props to SqlEditor mockedProps
* js(tests): adding required prop editorHeight to TabbedSqlEditors mockedProps
* js: removing unused moments dependency
* Fixing PropTypes warning message
React recently started warning on the upcoming deprecation of
React.PropTypes, the new approach is to use the `prop-types`
npm package instead.
* Fixing the tests
The standalone view doesn't need to fetch the datasources asynchronously
to fill in the dataources dropdown, it also doesn't need to render any
of the controls.
* fix scrolling for query search, getHeight was returning NAN
* add styling for query search pagination
* remove console.log
* make pagination buttons match bootstrap theme small buttons
At Airbnb we use these WARNING_MSG configuration element to make it
clear that we're in the staging environment.
Before this PR it would render under the navbar and mess up some of the
heights configurations, where you wouldn't be able to scroll all the way
to the bottom of the page. This fixes it.
* make react-virtualized table work
use dynamic sizing for cell width
enable filtering
require height prop for result set component
* fix tests and linting
* move some state to props
* move getTextWidth to visUtils
* make striped rows optional
* fix striped proptype
* update name to FilterableTable
* add basic test and fix linting
* accept array of columns keys rather than an array of objects that needs to be mapped
* move container div inside the component
* rename styles
* fit table component to width if it's smaller than parent container
* move stylesheet to javascript folder otherwise it throws an error on npm run prod
* move css to index.jsx
* fix result set spec
* fix linting and test
* fix result set props
* keep list immutable
Introducing a nice component as a label that show when data was
loaded from cache, when the cache was taken (in humanize duration as in
`a few minutes ago`) in a tooltip, and it can act as a button that
can trigger a force-refresh.
While working on it, it became clear that it was going to be hard to
use this component in the Dashboard view since it's not pure React.
I'm planning on refactoring the dashboard view with proper React/Redux
and introducing the CachedLabel component at that point.
While digging around in the Dashboard view I realized that there was a
bunch on unused code around managing timers that was used in explorev1
and decided to rip it out.
* Add UNIX socket option to runserver
Added an optional parameter to runserver to server from a UNIX socket instead of an address:port. I believe it is fairly common to server from sockets when using a web server like Nginx on the same host.
* Collapsed if/else logic for address or socket
Also wrapped help description for socket parameter
* [explore] fix and clean
Currently it's not possible to view queries while they are running, the
spinner appears endlessly. I decided to rearrange things a bit while
debugging so I could see clearly through it.
* Adding NotImplemented methods to base classes
* Fixing
* Piling up
Column functions are currently escaped with the
same dialect that is being used by Superset for its
metadata storage. Instead the dialect of the target
table should be used when escaping column names.
* [bugfix] clarifying how to create a slice
First time users [rightfully] think that the proper way to create a
slice is to click the `+` button on the `Slices` CRUD page.
Changing / simplifying the redirect logic and message flashing that was
failing here somehow.
* Changing message from danger to info
* Revert "[big num] make sure scatterplot dots align properly (#2559)"
This reverts commit f24ddfd467.
* Revert "Revert "measure x axis labels too and use the longest to determine margins" (#2550)"
This reverts commit fe68bc31c3.
* Revert "[bug num vis] fix sizing for single digits (#2548)"
This reverts commit 3d2c791ff1.
* revert all big num changes.
* fix linting
* make document global, fix linting
* sql_lab.py: compress via utils
* utils.py: added zlib_compress and zlib_compress_to_string
* core.py: converted to use zlib_decompress_to_string; renamed uncompress to decompress in utils.py
* utils_tests.py: added test for compress/decompress
* fixed broken utils test; removed redundant code and empty lines from utils.py
* utils.py: corrected docstrings, removed unnecessary 'else'
* removed yet another superfluous else
* fix js error, make should render viz more readable, add tooltips, style axis and add min/max to xaxis
* put getTextWidth in utils
* add document to eslint globals config
* Initial commit to support the athena DB
This work was done at tobii.com (in collaboration with knowit.se), and depends on:
- A patched version of PyAthenaJDBC (https://github.com/dwa/PyAthenaJDBC/tree/dwa-tobii-dict_formatter)
- A patched version of PyHive (https://github.com/dwa/PyHive/tree/dwa-tobii-sqlalchemy-athena)
And can be used like so:
athena://<user>:<password>@athena.us-east-1.amazonaws.com/?region_name=<region>&s3_staging_dir=s3%3A//<staging_bucket_of_choice>
* Rebased, and fixed two lint issues
* rename athena engine: athena -> awsathena
* remove network status feature
* only fetch queries if there are started or running queries
* don't use local storage
* remove last network status from actions
* don't remove support for local storage
* address pr comments and linting
* use .some rather than .forEach
* [sqllab] assign types for visualize flow
Somehow when using the visualize flow, the types were not
assigned at all, creating some bugs downstream. This PR attempts to get
the information required based on what pandas is knows and the types in
the data itself.
* Fixing tests
* Fixing tests
* Fixing more tests
* Fixing the last py3 tests
* use fast filter for tables select
* react-virtualized needs this package, throws error otherwise...
* continue to use async select and this.getTableNamesBySubStr when a schema is not selected
* import styles in same place as react-select css import
When adding a Redshift database, the docs said to use a connection
string that started with `redshift+psycopg2://...`, but this results
in a Python error from how SQLAlchemy executes the `get_schema_names`
method (first reported [here](https://github.com/airbnb/superset/issues/2364#issuecomment-284705179)).
The fix is to use `postgresql+psycog2://...` for the Redshift connection
string.
* Support more filter operators
* more filter operators [>, <, >=, <=, ==, !=, LIKE]
* Fix need to escape/double `%` in LIKE clauses
* spinner while loading values when changing column
* datasource config elements to allow to applying predicates when
fetching filter values
* refactor
* Removing doubling parens
* rebasing
* Merging migrations
* [table viz] allow showing time granularity in table (#2284)
The time granularity currently does not show up in table viz. Now
defining a granularity will add an extra ISO-formatted time column in
the table. The column will be added for both grouped by and not grouped
by.
* Changes based on comments
The time granularity currently does not show up in table viz. Now
defining a granularity will add an extra ISO-formatted time column in
the table. The column will be added for both grouped by and not grouped
by.
The time granularity currently does not show up in table viz. Now
defining a granularity will add an extra ISO-formatted time column in
the table. The column will be added for both grouped by and not grouped
by.
Caching wasn't working after deprecate_v1, this addresses it. Also surfacing
whether the data is served from cache in explore view and a way to force
run the query bypassing the cache.
on startup, FAB spits out a bunch of logging messages that aren't useful
in most cases. This shuts them down by default. They can be turned back
on with `config.SILENCE_FAB = True`
Also shushing a flask-cache warning around setting up a null (default) cache
* Simplifying the viz interface (#2005)
* Working on dashes
* Making this a collaborative branch
* Fixing some bugs
* Fixing bugs
* More improvements
* Add datasource back in bootstrap data
* Decent state
* Linting
* Moving forward
* Some more linting
* Fix the timer
* Triggering events through state
* Lingint
* Put filters in an array instead of flt strings (#2090)
* Put filters in an array instead of flt strings
* Remove query_filter(), put opChoices into Filter
* Update version_info.json
* Fix migrations
* More renderTrigger=true
* Fixing bugs
* Working on standalone
* getting standalone to work
* Fixed forcedHeight for standalone =view
* Linting
* Get save slice working in v2 (#2106)
* Filter bugfix
* Fixing empty series limit bug
* Fixed dashboard view
* Fixing short urls
* Only allow owners to overwrite slice (#2142)
* Raise exception when date range is wrong
* Only allow owner to overwrite a slice
* Fix tests for deprecate v1 (#2140)
* Fixed tests for control panels container and filters
* Fixed python tests for explorev2
* Fix linting errors
* Add in stop button during slice querying/rendering (#2121)
* Add in stop button during slice querying/rendering
* Abort ajax request on stop
* Adding missing legacy module
* Removing select2.sortable.js because of license
* Allow query to display while slice is loading (#2100)
* Allow query to display while slice is loading
* Put latestQueryFormData in store
* Reorganized query function, got rid of tu[le return values
* Merging migrations
* Wrapping up shortner migration
* Fixing tests
* Add folder creation to syncBackend
* Fixing edit URL in explore view
* Fix look of Stop button
* Adding syntax highlighting to query modal
* Fix cast_form_data and flase checkbox on dash
* Bugfix
* Going deeper
* Fix filtering
* Deleing invalid filters when changing datasource
* Minor adjustments
* Fixing calendar heatmap examples
* Moving edit datasource button to header's right side
* Fixing mapbox example
* Show stack trace when clicking alert
* Adding npm sync-backend command to build instruction
* Bumping up JS dependencies
* rm dep on select2
* Fix py3 urlparse
* rm superset-select2.js
* Improving migration scripts
* Bugfixes on staging
* Fixing Markup viz
* Fix werkzeug instance was created twice in Debug Mode (#2135)
* add reloader option for flask (#2136)
* using --no-reload option for flask
* divide a line of code into two lines for PEP8
* Rename rv => o in the decorator.
* Address comments.
* Permissions cleanup: remove none and duplicates. (#1967)
* Updates
* Rename var and dropdown text
* Cleanup
* Resolve comments.
* Add user to the perm check.
* fix axis label size bug and accommodate dual axis chart
* don't adjust margins for bar time series
* handling this below now
* apply different margin padding if explore or dashboard
* fix linting
* re-render chart after adjusting for long axis labels
* measure all of the labels and take the max height
* add missing isTimeSeries var
* fix linting
* use jQuery to get text ticks
* rotate 45 rather than 90
* Add KeyValue model for storing id-value pairs
use it for storing shared queries
* Change string to text and added test
* Put getQueryLink in one place
* Changed migration down version
* Changes based on comments
* Update bcf3126872fc_add_keyvalue.py
In Vertica, inner queries require different aliases than the main query.
This is an example of query generated before this patch:
SELECT chain AS chain,
weekstartday AS __timestamp,
SUM(inventory) AS "Inventory"
FROM mytable
JOIN
(SELECT chain AS chain__,
SUM(inventory) AS "Inventory"
FROM mytable
WHERE weekstartday >= '2016-01-24 00:00:00'
AND weekstartday <= '2017-01-17 00:00:00'
GROUP BY chain
ORDER BY "Inventory" DESC LIMIT 50) AS anon_1 ON chain = chain__
WHERE weekstartday >= '2016-01-24 00:00:00'
AND weekstartday <= '2017-01-17 00:00:00'
GROUP BY chain,
weekstartday
ORDER BY "Inventory" DESC LIMIT 50000
Which in Vertica produces the error:
Error: ('42702', '[42702] ERROR 2671: Column reference "inventory" is ambiguous\n (2671) (SQLExecDirectW)')
And this is the same example after the patch:
SELECT chain AS chain,
weekstartday AS __timestamp,
SUM(inventory) AS "Inventory"
FROM mytable
JOIN
(SELECT chain AS chain__,
SUM(inventory) AS mme_inner__
FROM mytable
WHERE weekstartday >= '2016-01-24 00:00:00'
AND weekstartday <= '2017-01-17 00:00:00'
GROUP BY chain
ORDER BY mme_inner__ DESC LIMIT 50) AS anon_1 ON chain = chain__
WHERE weekstartday >= '2016-01-24 00:00:00'
AND weekstartday <= '2017-01-17 00:00:00'
GROUP BY chain,
weekstartday
ORDER BY "Inventory" DESC LIMIT 50000
Related PR:
https://github.com/airbnb/superset/commit/19f5371787df8d65137b7cf2ceb6c2a6b764cdca
* Cleanup fulfilled requests after approve
* Modified tests
* Moved to separate test, add user to access functions
* Moved to separate test and added test cases
* Fixed issue with dryrun
* More changes based on comments
I had issues compiling the src, I think it's because the syntax is different than the one used here?
With the dist it works fine.
I also saw there were other related issues in the past, so I'm suggesting to switch to the dist file.
* [explorev2] using label in 'Visualization Type' Select
* moved url related logic upstream in the caller
* moved option creation logic from render method to the constructor as
it only needs to be executed once
* Refactoring out getOptions
Fix cache set after c14c7ed update json serializer behaviour
but forget to update one call site.
While at it move payload serialization outside of catch all try
block since we want to see explosions if any.
Fix#1910
* Druid dashboard import
* Fix tests.
* Parse params with trailing commas
* Resolved issue with datasource is not in DB yet when slice perms are set
* Finish rebase
* Fix heads alembic problem.
* Druid dashboard import
* Fix tests.
* Parse params with trailing commas
* Resolved issue with datasource is not in DB yet when slice perms are set
* Finish rebase
* Add #views and #distinct users to created dashboard on profile page
* Added index on logs to speed up query
* Added #views and #users for slice table
* Add a views column to dashboards and slices, prepopulate
them with Log data
* Remove index on Log model
* Remove unused index
* Update 1b2c3f7c96f9_.py
fix multiple heads
* Exclude postgres in prepopulating views column
* Add dual-axis line chart to viz
- Add a new viz using two y axis for multi-chart in nvd3
- Option to set metric and axis format for two y axis
* Resolve conflicts
* Fixed x axis and resized thumbnail
* Added example to __init__.py
* Change get_df to match with format
* only show single run query button, allow async if possible
* only pass the needed props, rather than entire objects to the component
* add simple test
* fix linting
* Initial
* rewrite some line to make it short and setting merge variable temporarily
* rewrite commit author
* add emitted attribute
* Fix typo
* fix test error
* fix typo
* test added
* Improving logging with duration and referrer
* Handling case where referrer is None
* Providing log_this with its own session
* Attempting to fix tests
* Fixing tests
* Enable freeform-select with fetched column values for filter values
- db migration to add filter_select_enabled
- add freeform-multi option for Selectfield
- modify formatFilter() function on query to accomodate filter-select
* Fix js tests
* Fix codeclimate issue
* Changes based on comments
* Add test for filter endpoint
* Extract out renderFilterFormField function from render
* Fix landscape issues
* Stop ChartContainer from rendering twice on chartStatus change
* Make spinner overlay and dim chart while laoding
* Added timeout to make render more stable on resize
* Put viz in state and call resize at browser size change
* add render after height change since resize depends on render called on the same object
* Change name of renderVis to renderVisOnChange
* Only call resize at height change, persist viz in state
* Call resize wihout render at window size change
* Sort searched queries by recency
* Fixed sqllab tests
* Add one more tr to QueryTable spec for pagination
* Change desc to asc as we reverse queries in component
This should fix issue #1339.
IE 11 and lower has a long standing issue: out-of-document element's
pathname has no leading '/'. See
https://connect.microsoft.com/IE/feedbackdetail/view/1002846/pathname-incorrect-for-out-of-document-elements
And Superset's Slice.jsonEndpoint() method relies on pathname() to build
JSON API URL for slices:
```javascript
jsonEndpoint() {
const parser = document.createElement('a');
parser.href = data.json_endpoint;
let endpoint = parser.pathname + this.querystring();
endpoint += '&json=true';
endpoint += '&force=' + this.force;
return endpoint;
},
```
`parser` above is exactly an out-of-document element. Therefore when
running in IE <= 11, Superset would build wrong JSON endpoint URLs,
hence the 404 errors for loading data for slices.
This commit adds a simple workaround when leading '/' is missing in the
value returned by pathname().
* add config option for explore v2 beta users, and send through v2 path
* fix long lines
* use role rather than user ids in config
* add test
* simplify role check
* remove extra line
* use different test user for v2 tests
* Not working errors
* Remove update_explore endpoint, only update explore endpoints in reducer on query
* Change scroll to auto and make container reponse to height:
* Remove update_explore endpoint
* Remove can_update_explore perm
* Add NVD3's bullet chart.
* Add empty lines before nested function definitions.
* Add thumbnail for bullet chart.
* Add bullet chart to gallery.rst.
* Add "requiresTime: false", fix indentation.
* Avoid scaling bullet chart vertically.
* Use a default if no range is specified.
* Fix coloring of bullet chart.
* Add http to copied url and move function to componentWillReceiveProps
* Added getText() to CopyToClipbaord to enable ajax calls for getting copy text
* Set ajax call to synchronous (document.execCommand only works in synchronous mode
* Make cell-click filter in table viz optional
- Added table_filter checkbox in table viz
- If set to false, clicking on a cell in table will not apply filter to
dashboard
* Fix codeclimate issue
* Change default to false
* Fixed js error when results are not available
* Flush data and query in results when running new query, keeping columns
* add exception for columns
* Move setState from componentWillMount to componentWillReceiveProps
* Nit
* Fix superset cli for python3
dict.iteritems() has been removed since dict.items() returns an
iterable in python3. Shouldn't be a big deal for python2 to load
all the data into a list.
Fix#1756
* bin/superset: avoid some work when reading config
We don't need to unpack and then pack again a dictionary.
* Support running superset via pex
* [superset] Update default port in superset/bin/superset
* Fix codeclimate line length issues
* Fix another line length issue, in config.py
* Add trivial utils test to increase test coverage
* Clean up runserver handling
* Make chart container more responsive
**Leave chart empty when theres error
**Make all boolean fields auto-query chart when changed
* Replace forEach with some
* Added fields to autoQueryFields
* Bug fixes in Save Modal
Issues solved:
- Save button doesn't pass in gotodash
- slice_name was passed in from store as original slice_name instead of
new one in 'saveas' action
- datasource_type wasn't passed in to defaultViz and defaultForm
function
* Change css filename to exploreV2
* Moved out utils
* Implement table name extraction tests.
* Address comments.
* Fix tests and reimplement the token processing.
* Exclude aliases.
* Clean up print statements and code.
* Reverse select test.
* Fix failing test.
* Test JOINs
* refactore as a class
* Check for permissions in SQL Lab.
* Implement permissions check for the datasources in sql_lab
* Address comments.
* Add per schema permissions.
* Address comments.
* Add schema_access perms to the alpha and admin
* Create permissions on addition databases and datasources.
* Remove hybrid_property. Linter complains.
* Fixed table_name does not exist in druid
* Make Chart container scrollable for large chart
* Fixed bug of action buttons not clickable in heatmap
* Solve codeclimate issue
* Limit overflow to x
* render table name if in table view
* only render fave star and edit button if slice, not table
* fix error when table view
* use [table-name] - untitled format
* remove extra fave star
* Added filter in ControlPanelsContainer
* Move function for getting url params object to utils
* Fixed python test
* Move Filter to separate component
* Added specs and made changes based on comments
* Moved specs to right folder
* For some fields we would like to re-render chart once field is
* changed, saving user the time to click query button
* Such fields are stored in an array in store, could add more fields in
* the future
* Get sections to render when switching datasource
- Move sectionsToRender in store and use is for defaultFormData
- Change some SelectField to FreeFormSelect according to forms.py
* Solved the css not found problem in staging
* Fixed js tests
- Keep the scope as narrow as possible, to make it easier to implement.
- Remember that this is a volunteer-driven project, and that contributions are welcome :)
For large features or major changes to codebase, please create **Superset Improvement Proposal (SIP)**. See template from [SIP-0](https://github.com/apache/incubator-superset/issues/5602)
### Fix Bugs
Look through the GitHub issues for bugs. Anything tagged with "bug" is
Look through the GitHub issues. Issues tagged with `#bug` is
open to whoever wants to implement it.
### Implement Features
Look through the GitHub issues for features. Anything tagged with
"feature" is open to whoever wants to implement it.
Look through the GitHub issues. Issues tagged with
`#feature` is open to whoever wants to implement it.
### Documentation
### Improve Documentation
Superset could always use better documentation,
whether as part of the official Superset docs,
in docstrings, `docs/*.rst` or even on the web as blog posts or
articles.
articles. See [Documentation](#documentation) for more details.
### Submit Feedback
### Add Translations
The best way to send feedback is to file an issue on GitHub.
If you are proficient in a non-English language, you can help translate text strings from Superset's UI. You can jump in to the existing language dictionaries at `superset/translations/<language_code>/LC_MESSAGES/messages.po`, or even create a dictionary for a new language altogether. See [Translating](#translating) for more details.
If you are proposing a feature:
### Ask Questions
- Explain in detail how it would work.
- Keep the scope as narrow as possible, to make it easier to
implement.
- Remember that this is a volunteer-driven project, and that
contributions are welcome :)
There is a dedicated [`apache-superset` tag](https://stackoverflow.com/questions/tagged/apache-superset) on [StackOverflow](https://stackoverflow.com/). Please use it when asking questions.
## Latest Documentation
## Pull Request Guidelines
Latest documentation and tutorial are available [here](http://airbnb.io/superset)
A philosophy we would like to strongly encourage is
## Setting up a Python development environment
> Before creating a PR, create an issue.
Check the [OS dependencies](http://airbnb.io/superset/installation.html#os-dependencies) before follows these steps.
The purpose is to separate problem from possible solutions.
# fork the repo on GitHub and then clone it
# alternatively you may want to clone the main repo but that won't work
# so well if you are planning on sending PRs
# git clone git@github.com:airbnb/superset.git
**Bug fixes:** If you’re only fixing a small bug, it’s fine to submit a pull request right away but we highly recommend to file an issue detailing what you’re fixing. This is helpful in case we don’t accept that specific fix but want to keep track of the issue. Please keep in mind that the project maintainers reserve the rights to accept or reject incoming PRs, so it is better to separate the issue and the code to fix it from each other. In some cases, project maintainers may request you to create a separate issue from PR before proceeding.
# [optional] setup a virtual env and activate it
virtualenv env
source env/bin/activate
**Refactor:** For small refactors, it can be a standalone PR itself detailing what you are refactoring and why. If there are concerns, project maintainers may request you to create a `#SIP` for the PR before proceeding.
# install for development
python setup.py develop
**Feature/Large changes:** If you intend to change the public API, or make any non-trivial changes to the implementation, we requires you to file a new issue as `#SIP` (Superset Improvement Proposal). This lets us reach an agreement on your proposal before you put significant effort into it. You are welcome to submit a PR along with the SIP (sometimes necessary for demonstration), but we will not review/merge the code until the SIP is approved.
# Create an admin user
fabmanager create-admin --app superset
In general, small PRs are always easier to review than large PRs. The best practice is to break your work into smaller independent PRs and refer to the same issue. This will greatly reduce turnaround time.
# Initialize the database
superset db upgrade
Finally, never submit a PR that will put master branch in broken state. If the PR is part of multiple PRs to complete a large feature and cannot work on its own, you can create a feature branch and merge all related PRs into the feature branch before creating a PR from feature branch to master.
# Create default roles and permissions
superset init
### Protocol
# Load some data to play with
superset load_examples
#### Authoring
# start a dev web server
superset runserver -d
- Fill in all sections of the PR template.
- Add prefix `[WIP]` to title if not ready for review (WIP = work-in-progress). We recommend creating a PR with `[WIP]` first and remove it once you have passed CI test and read through your code changes at least once.
- **Screenshots/GIFs:** Changes to user interface require before/after screenshots, or GIF for interactions
- If no screenshot is provided, the committers will mark the PR with `need:screenshot` label and will not review until screenshot is provided.
- **Dependencies:** Be careful about adding new dependency and avoid unnecessary dependencies.
- For Python, include it in `setup.py` denoting any specific restrictions and in `requirements.txt` pinned to a specific version which ensures that the application build is deterministic.
- For Javascript, include new libraries in `package.json`
- **Tests:** The pull request should include tests, either as doctests, unit tests, or both. Make sure to resolve all errors and test failures. See [Testing](#testing) for how to run tests.
- **Documentation:** If the pull request adds functionality, the docs should be updated as part of the same PR. Doc string are often sufficient, make sure to follow the sphinx compatible standards.
- **CI:** Reviewers will not review the code until all CI tests are passed. Sometimes there can be flaky tests. You can close and open PR to re-run CI test. Please report if the issue persists. After the CI fix has been deployed to `master`, please rebase your PR.
- **Code coverage:** Please ensure that code coverage does not decrease.
- Remove `[WIP]` when ready for review. Please note that it may be merged soon after approved so please make sure the PR is ready to merge and do not expect more time for post-approval edits.
- If the PR was not ready for review and inactive for > 30 days, we will close it due to inactivity. The author is welcome to re-open and update.
#### Reviewing
## Setting up the node / npm javascript environment
- Use constructive tone when writing reviews.
- If there are changes required, state clearly what needs to be done before the PR can be approved.
- If you are asked to update your pull request with some changes there's no need to create a new one. Push your changes to the same branch.
- The committers reserve the right to reject any PR and in some cases may request the author to file an issue.
`superset/assets` contains all npm-managed, front end assets.
Flask-Appbuilder itself comes bundled with jQuery and bootstrap.
While these may be phased out over time, these packages are currently not
managed with npm.
#### Merging
### Node/npm versions
Make sure you are using recent versions of node and npm. No problems have been found with node>=5.10 and npm>=3.9.
- At least one approval is required for merging a PR.
- PR is usually left open for at least 24 hours before merging.
- After the PR is merged, [close the corresponding issue(s)](https://help.github.com/articles/closing-issues-using-keywords/).
### Using npm to generate bundled files
#### Post-merge Responsibility
#### npm
First, npm must be available in your environment. If it is not you can run the following commands
(taken from [this source](https://gist.github.com/DanHerbert/9520689))
```
brew install node --without-npm
echo prefix=~/.npm-packages >> ~/.npmrc
curl -L https://www.npmjs.com/install.sh | sh
- Project maintainers may contact the PR author if new issues are introduced by the PR.
- Project maintainers may revert your changes if a critical issue is found, such as breaking master branch CI.
## Managing Issues and PRs
To handle issues and PRs that are coming in, committers read issues/PRs and flag them with labels to categorize and help contributors spot where to take actions, as contributors usually have different expertises.
Triaging goals
- **For issues:** Categorize, screen issues, flag required actions from authors.
- **For PRs:** Categorize, flag required actions from authors. If PR is ready for review, flag required actions from reviewers.
First, add **Category labels (a.k.a. hash labels)**. Every issue/PR must have one hash label (except spam entry). Labels that begin with `#` defines issue/PR type:
| Label | for Issue | for PR |
|-------------------|-----------|--------|
| `#bug` | Bug report | Bug fix |
| `#code-quality` | Describe problem with code, architecture or productivity | Refactor, tests, tooling |
| `#feature` | New feature request | New feature implementation |
| `#refine` | Propose improvement that does not provide new features and is also not a bug fix nor refactor, such as adjust padding, refine UI style. | Implementation of improvement that does not provide new features and is also not a bug fix nor refactor, such as adjust padding, refine UI style. |
| `#doc` | Documentation | Documentation |
| `#question` | Troubleshooting: Installation, Running locally, Ask how to do something. Can be changed to `#bug` later. | N/A |
| `#SIP` | Superset Improvement Proposal | N/A |
| `#ASF` | Tasks related to Apache Software Foundation policy | Tasks related to Apache Software Foundation policy |
Then add other types of labels as appropriate.
- **Descriptive labels (a.k.a. dot labels):** These labels that begin with `.` describe the details of the issue/PR, such as `.ui`, `.js`, `.install`, `.backend`, etc. Each issue/PR can have zero or more dot labels.
- **Need labels:** These labels have pattern `need:xxx`, which describe the work required to progress, such as `need:rebase`, `need:update`, `need:screenshot`.
- **Risk labels:** These labels have pattern `risk:xxx`, which describe the potential risk on adopting the work, such as `risk:db-migration`. The intention was to better understand the impact and create awareness for PRs that need more rigorous testing.
- **Status labels:** These labels describe the status (`abandoned`, `wontfix`, `cant-reproduce`, etc.) Issue/PRs that are rejected or closed without completion should have one or more status labels.
- **Version labels:** These have the pattern `vx.x` such as `v0.28`. Version labels on issues describe the version the bug was reported on. Version labels on PR describe the first release that will include the PR.
Committers may also update title to reflect the issue/PR content if the author-provided title is not descriptive enough.
If the PR passes CI tests and does not have any `need:` labels, it is ready for review, add label `review` and/or `design-review`.
If an issue/PR has been inactive for >=30 days, it will be closed. If it does not have any status label, add `inactive`.
## Setup Local Environment for Development
First, [fork the repository on GitHub](https://help.github.com/articles/about-forks/), then clone it. You can clone the main repository directly, but you won't be able to send pull requests.
aren't actually stored in that directory. Instead, you should add and commit
images (and any other static assets) to the `superset/assets/images` directory.
When the docs are deployed to https://superset.incubator.apache.org/, images
are copied from there to the `_static/img` directory, just like they're referenced
in the docs.
For example, the image referenced above actually lives in `superset/assets/images/tutorial`. Since the image is moved during the documentation build process, the docs reference the image in `_static/img/tutorial` instead.
#### API documentation
Generate the API documentation with:
```bash
pip install -r docs/requirements.txt
python setup.py build_sphinx
```
### Flask server
Make sure your machine meets the [OS dependencies](https://superset.incubator.apache.org/installation.html#os-dependencies) before following these steps.
```bash
# Create a virtual environemnt and activate it (recommended)
virtualenv -p python3 venv # setup a python3.6 virtualenv
source venv/bin/activate
# Install external dependencies
pip install -r requirements.txt
pip install -r requirements-dev.txt
# Install Superset in editable (development) mode
pip install -e .
# Create an admin user
fabmanager create-admin --app superset
# Initialize the database
superset db upgrade
# Create default roles and permissions
superset init
# Load some data to play with
superset load_examples
# Start the Flask dev web server from inside the `superset` dir at port 8088
# Note that your page may not have css at this point.
# See instructions below how to build the front-end assets.
cd superset
FLASK_ENV=development flask run -p 8088 --with-threads --reload --debugger
```
#### Logging to the browser console
This feature is only available on Python 3. When debugging your application, you can have the server logs sent directly to the browser console:
```bash
FLASK_ENV=development flask run -p 8088 --with-threads --reload --debugger --console-log
```
You can log anything to the browser console, including objects:
```python
fromsupersetimportapp
app.logger.error('An exception occurred!')
app.logger.info(form_data)
```
### Frontend Assets
Frontend assets (JavaScript, CSS, and images) must be compiled in order to properly display the web UI. The `superset/assets` directory contains all NPM-managed front end assets. Note that there are additional frontend assets bundled with Flask-Appbuilder (e.g. jQuery and bootstrap); these are not managed by NPM, and may be phased out in the future.
First, be sure you are using recent versions of NodeJS and npm. Using [nvm](https://github.com/creationix/nvm) to manage them is recommended.
#### Prerequisite
#### Installing Dependencies
Install third-party dependencies listed in `package.json`:
```bash
# From the root of the repository
cd superset/assets
# Install dependencies from `package-lock.json`
npm ci
```
#### Building
You can run the Webpack dev server (in a separate terminal from Flask), which runs on port 9000 and proxies non-asset requests to the Flask server on port 8088. After pointing your browser to `http://localhost:9000`, updates to asset sources will be reflected in-browser without a refresh.
```bash
# Run the dev server
npm run dev-server
# Run the dev server on a non-default port
npm run dev-server -- --port=9001
# Run the dev server proxying to a Flask server on a non-default port
npm run dev-server -- --supersetPort=8081
```
Alternatively you can use one of the following commands.
```bash
# Start a watcher that recompiles your assets as you modify them (but have to manually reload your browser to see changes.)
npm run dev
# Compile the Javascript and CSS in production/optimized mode for official releases
npm run prod
# Start a web server that manages and updates your assets as you modify them
npm run dev
# Copy a conf file from the frontend to the backend
npm run sync-backend
```
For every development session you will have to start a flask dev server
as well as an npm watcher
#### Updating NPM packages
Use npm in the prescribed way, making sure that
`superset/assets/package-lock.json` is updated according to `npm`-prescribed
best practices.
#### Feature flags
Superset supports a server-wide feature flag system, which eases the incremental development of features. To add a new feature flag, simply modify `superset_config.py` with something like the following:
```
superset runserver -d -p 8081
npm run dev
FEATURE_FLAGS = {
'SCOPED_FILTER': True,
}
```
If you want to use the same flag in the client code, also add it to the FeatureFlag TypeScript enum in `superset/assets/src/featureFlags.ts`. For example,
```
export enum FeatureFlag {
SCOPED_FILTER = 'SCOPED_FILTER',
}
```
## Testing
Python tests can be run with:
./run_tests.sh
We use [Mocha](https://mochajs.org/), [Chai](http://chaijs.com/) and [Enzyme](http://airbnb.io/enzyme/) to test Javascript. Tests can be run with:
cd /superset/superset/assets/javascripts
npm i
npm run test
`superset/config.py` contains `DEFAULT_FEATURE_FLAGS` which will be overwritten by
those specified under FEATURE_FLAGS in `superset_config.py`. For example, `DEFAULT_FEATURE_FLAGS = { 'FOO': True, 'BAR': False }` in `superset/config.py` and `FEATURE_FLAGS = { 'BAR': True, 'BAZ': True }` in `superset_config.py` will result
in combined feature flags of `{ 'FOO': True, 'BAR': True, 'BAZ': True }`.
## Linting
Lint the project with:
# for python changes
flake8 changes tests
flake8 changes superset
```bash
# for python
tox -e flake8
# for javascript
npm run lint
## Linting with codeclimate
Codeclimate is a service we use to measure code quality and test coverage. To get codeclimate's report on your branch, ideally before sending your PR, you can setup codeclimate against your Superset fork. After you push to your fork, you should be able to get the report at http://codeclimate.com . Alternatively, if you prefer to work locally, you can install the codeclimate cli tool.
All python tests are carried out in [tox](http://tox.readthedocs.io/en/latest/index.html)
a standardized testing framework.
All python tests can be run with any of the tox [environments](http://tox.readthedocs.io/en/latest/example/basic.html#a-simple-tox-ini-default-environments), via,
```bash
tox -e <environment>
```
docker-machine start
eval "$(docker-machine env default)”
codeclimate analyze
For example,
```bash
tox -e py36
```
More info can be found here: https://docs.codeclimate.com/docs/open-source-free
Alternatively, you can run all tests in a single file via,
As part of the npm build process, CSS for Superset is compiled from `Less`, a dynamic stylesheet language.
Note that the test environment uses a temporary directory for defining the
SQLite databases which will be cleared each time before the group of test
commands are invoked.
It's possible to customize or add your own theme to Superset, either by overriding CSS rules or preferably
by modifying the Less variables or files in `assets/stylesheets/less/`.
### JavaScript Testing
The `variables.less` and `bootswatch.less` files that ship with Superset are derived from
[Bootswatch](https://bootswatch.com) and thus extend Bootstrap. Modify variables in these files directly, or
swap them out entirely with the equivalent files from other Bootswatch (themes)[https://github.com/thomaspark/bootswatch.git]
We use [Jest](https://jestjs.io/) and [Enzyme](http://airbnb.io/enzyme/) to test Javascript. Tests can be run with:
## Pull Request Guidelines
```bash
cd superset/assets
npm run test
```
Before you submit a pull request from your forked repo, check that it
meets these guidelines:
### Integration Testing
1. The pull request should include tests, either as doctests,
unit tests, or both.
2. If the pull request adds functionality, the docs should be updated
as part of the same PR. Doc string are often sufficient, make
sure to follow the sphinx compatible standards.
3. The pull request should work for Python 2.6, 2.7, and ideally python 3.3.
``from __future__ import`` will be required in every `.py` file soon.
4. Code will be reviewed by re running the unittests, flake8 and syntax
should be as rigorous as the core Python project.
5. Please rebase and resolve all conflicts before submitting.
We use [Cypress](https://www.cypress.io/) for integration tests. Tests can be run by `tox -e cypress`. To open Cypress and explore tests first setup and run test server:
```bash
exportSUPERSET_CONFIG=tests.superset_test_config
superset db upgrade
superset init
superset load_test_users
superset load_examples
superset runserver
```
## Translations
Run Cypress tests:
We use [Babel](http://babel.pocoo.org/en/latest/) to translate Superset. The
key is to instrument the strings that need translation using
`from flask_babel import lazy_gettext as _`. Once this is imported in
a module, all you have to do is to `_("Wrap your strings")` using the
underscore `_` "function".
```bash
cd /superset/superset/assets
npm run build
npm run cypress run
```
To enable changing language in your environment, you can simply add the
`LANGUAGES` parameter to your `superset_config.py`. Having more than one
options here will add a language selection dropdown on the right side of the
navigation bar.
## Translating
LANGUAGES = {
'en': {'flag': 'us', 'name': 'English'},
'fr': {'flag': 'fr', 'name': 'French'},
'zh': {'flag': 'cn', 'name': 'Chinese'},
}
We use [Babel](http://babel.pocoo.org/en/latest/) to translate Superset. In Python files, we import the magic `_` function using:
As per the [Flask AppBuilder documentation] about translation, to create a
new language dictionary, run the following command:
```python
fromflask_babelimportlazy_gettextas_
```
pybabel init -i ./babel/messages.pot -d superset/translations -l es
then wrap our translatable strings with it, e.g. `_('Translate me')`. During extraction, string literals passed to `_` will be added to the generated `.po` file for each language for later translation.
At runtime, the `_` function will return the translation of the given string for the current language, or the given string itself if no translation is available.
Then it's a matter of running the statement below to gather all stings that
need translation
In JavaScript, the technique is similar: we import `t` (simple translation), `tn` (translation containing a number).
If you get errors running `po2json`, you might be running the Ubuntu package with the same
name, rather than the NodeJS package (they have a different format for the arguments). If
there is a conflict, you may need to update your `PATH` environment variable or fully qualify
the executable path (e.g. `/usr/local/bin/po2json` instead of `po2json`).
If you get a lot of `[null,***]` in `messages.json`, just delete all the `null,`.
For example, `"year":["年"]` is correct while `"year":[null,"年"]`is incorrect.
### Creating a new language dictionary
To create a dictionary for a new language, run the following, where `LANGUAGE_CODE` is replaced with
the language code for your target language, e.g. `es` (see [Flask AppBuilder i18n documentation](https://flask-appbuilder.readthedocs.io/en/latest/i18n.html) for more details):
[](https://gitter.im/airbnb/superset?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
[](https://gitter.im/airbnb/superset?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you under the Apache License, Version 2.0 (the
"License"); you may not use this file except in compliance
with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.
-->
# Updating Superset
This file documents any backwards-incompatible changes in Superset and
assists people when migrating to a new version.
## Superset 0.32.0
* If you use `Hive` or `Presto`, we've moved some dependencies that were
in the main package as optional now. To get these packages,
run `pip install superset[presto]` and/or `pip install superset[hive]` as
required.
## Superset 0.31.0
* boto3 / botocore was removed from the dependency list. If you use s3
as a place to store your SQL Lab result set or Hive uploads, you may
have to rely on an alternate requirements.txt file to install those
dependencies.
* From 0.31.0 onwards, we recommend not using the npm package `yarn` in
favor of good old `npm install`. While yarn should still work just fine,
you should probably align to guarantee builds similar to the ones we
use in testing and across the community in general.
## Superset 0.30.0
* 0.30.0 includes a db_migration that removes allow_run_sync. This may
require downtime because during the migration if the db is migrated first,
superset will get 500 errors when the code can't find the field (until
the deploy finishes).
## Superset 0.29.0
* India was removed from the "Country Map" visualization as the geojson
file included in the package was very large
## Superset 0.28.0
* Support for Python 2 is deprecated, we only support >=3.6 from
`0.28.0` onwards
* Superset 0.28 deprecates the previous dashboard layout. While 0.27
offered a migration workflow to users and allowed them to validate and
publish their migrated dashboards individually, 0.28 forces
the migration of all
dashboards through an automated db migration script. We
do recommend that you take a backup prior to this migration.
* Superset 0.28 deprecates the `median` cluster label aggregator for mapbox visualizations. This particular aggregation is not supported on mapbox visualizations going forward.
* Superset 0.28 upgrades `flask-login` to `>=0.3`, which includes a
Yes, using the ``Upload a CSV`` button under the ``Sources`` menu item.
This brings up a form that allows you specify required information.
After creating the table from CSV, it can then be loaded like any
other on the ``Sources -> Tables`` page.
Why are my queries timing out?
------------------------------
If you are seeing timeouts (504 Gateway Time-out) when running queries,
it's because the web server is timing out web requests. If you want to
increase the default (50), you can specify the timeout when starting the
web server with the ``-t`` flag, which is expressed in seconds.
There are many reasons may cause long query timing out.
``superset runserver -t 300``
- For running long query from Sql Lab, by default Superset allows it run as long as 6 hours before it being killed by celery. If you want to increase the time for running query, you can specify the timeout in configuration. For example:
``SQLLAB_ASYNC_TIME_LIMIT_SEC = 60 * 60 * 6``
- Superset is running on gunicorn web server, which may time out web requests. If you want to increase the default (50), you can specify the timeout when starting the web server with the ``-t`` flag, which is expressed in seconds.
``superset runserver -t 300``
- If you are seeing timeouts (504 Gateway Time-out) when loading dashboard or explore slice, you are probably behind gateway or proxy server (such as Nginx). If it did not receive a timely response from Superset server (which is processing long queries), these web servers will send 504 status code to clients directly. Superset has a client-side timeout limit to address this issue. If query didn't come back within clint-side timeout (60 seconds by default), Superset will display warning message to avoid gateway timeout message. If you have a longer gateway timeout limit, you can change the timeout settings in ``superset_config.py``:
``SUPERSET_WEBSERVER_TIMEOUT = 60``
Why is the map not visible in the mapbox visualization?
@@ -70,7 +105,7 @@ It's easy: use the ``Filter Box`` widget, build a slice, and add it to your
dashboard.
The ``Filter Box`` widget allows you to define a query to populate dropdowns
that can be use for filtering. To build the list of distinct values, we
that can be used for filtering. To build the list of distinct values, we
run a query, and sort the result by the metric you provide, sorting
descending.
@@ -78,6 +113,11 @@ The widget also has a checkbox ``Date Filter``, which enables time filtering
capabilities to your dashboard. After checking the box and refreshing, you'll
see a ``from`` and a ``to`` dropdown show up.
By default, the filtering will be applied to all the slices that are built
on top of a datasource that shares the column name that the filter is based
on. It's also a requirement for that column to be checked as "filterable"
in the column tab of the table editor.
But what about if you don't want certain widgets to get filtered on your
dashboard? You can do that by editing your dashboard, and in the form,
edit the ``JSON Metadata`` field, more specifically the
@@ -85,15 +125,16 @@ edit the ``JSON Metadata`` field, more specifically the
never be affected by any dashboard level filtering.
..code::
..code-block::json
{
"filter_immune_slices":[324,65,92],
"expanded_slices":{},
"filter_immune_slice_fields":{
"177": ["country_name", "__from", "__to"],
"32": ["__from", "__to"]
}
"177":["country_name","__time_range"],
"32":["__time_range"]
},
"timed_refresh_immune_slices":[324]
}
In the json blob above, slices 324, 65 and 92 won't be affected by any
@@ -103,19 +144,174 @@ Now note the ``filter_immune_slice_fields`` key. This one allows you to
be more specific and define for a specific slice_id, which filter fields
should be disregarded.
Note the use of the ``__from`` and ``__to`` keywords, those are reserved
for dealing with the time boundary filtering mentioned above.
Note the use of the ``__time_range`` keyword, which is reserved for dealing
with the time boundary filtering mentioned above.
But what happens with filtering when dealing with slices coming from
different tables or databases? If the column name is shared, the filter will
be applied, it's as simple as that.
How to limit the timed refresh on a dashboard?
----------------------------------------------
By default, the dashboard timed refresh feature allows you to automatically re-query every slice
on a dashboard according to a set schedule. Sometimes, however, you won't want all of the slices
to be refreshed - especially if some data is slow moving, or run heavy queries. To exclude specific
slices from the timed refresh process, add the ``timed_refresh_immune_slices`` key to the dashboard
``JSON Metadata`` field:
..code-block::json
{
"filter_immune_slices":[],
"expanded_slices":{},
"filter_immune_slice_fields":{},
"timed_refresh_immune_slices":[324]
}
In the example above, if a timed refresh is set for the dashboard, then every slice except 324 will
be automatically re-queried on schedule.
Slice refresh will also be staggered over the specified period. You can turn off this staggering
by setting the ``stagger_refresh`` to ``false`` and modify the stagger period by setting
``stagger_time`` to a value in milliseconds in the ``JSON Metadata`` field:
..code-block::json
{
"stagger_refresh":false,
"stagger_time":2500
}
Here, the entire dashboard will refresh at once if periodic refresh is on. The stagger time of
2.5 seconds is ignored.
Why does fabmanager or superset freezed/hung/not responding when started (my home directory is NFS mounted)?
superset creates and uses an sqlite database at ``~/.superset/superset.db``. Sqlite is known to `don't work well if used on NFS`__ due to broken file locking implementation on NFS.
By default, superset creates and uses an sqlite database at ``~/.superset/superset.db``. Sqlite is known to `don't work well if used on NFS`__ due to broken file locking implementation on NFS.
__ https://www.sqlite.org/lockingv3.html
One work around is to create a symlink from ~/.superset to a directory located on a non-NFS partition.
You can override this path using the ``SUPERSET_HOME`` environment variable.
Another work around is to change where superset stores the sqlite database by adding ``SQLALCHEMY_DATABASE_URI = 'sqlite:////new/location/superset.db'`` in superset_config.py (create the file if needed), then adding the directory where superset_config.py lives to PYTHONPATH environment variable (e.g. ``export PYTHONPATH=/opt/logs/sandbox/airbnb/``).
What if the table schema changed?
---------------------------------
Table schemas evolve, and Superset needs to reflect that. It's pretty common
in the life cycle of a dashboard to want to add a new dimension or metric.
To get Superset to discover your new columns, all you have to do is to
go to ``Menu -> Sources -> Tables``, click the ``edit`` icon next to the
table who's schema has changed, and hit ``Save`` from the ``Detail`` tab.
Behind the scene, the new columns will get merged it. Following this,
you may want to
re-edit the table afterwards to configure the ``Column`` tab, check the
appropriate boxes and save again.
How do I go about developing a new visualization type?
To clarify, the *database backend* is an OLTP database used by Superset to store its internal
information like your list of users, slices and dashboard definitions.
Superset is tested using Mysql, Postgresql and Sqlite for its backend. It's recommended you
install Superset on one of these database server for production.
Using a column-store, non-OLTP databases like Vertica, Redshift or Presto as a database backend simply won't work as these databases are not designed for this type of workload. Installation on Oracle, Microsoft SQL Server, or other OLTP databases may work but isn't tested.
Please note that pretty much any databases that have a SqlAlchemy integration should work perfectly fine as a datasource for Superset, just not as the OLTP backend.
How can i configure OAuth authentication and authorization?
interface. This interface allows for defining database-specific configurations
and logic
that go beyond the SQLAlchemy and DBAPI scope. This includes features like:
* date-related SQL function that allow Superset to fetch different
time granularities when running time-series queries
* whether the engine supports subqueries. If false, Superset may run 2-phase
queries to compensate for the limitation
* methods around processing logs and inferring the percentage of completion
of a query
* technicalities as to how to handle cursors and connections if the driver
is not standard DBAPI
* more, read the code for more details
Beyond the SQLAlchemy connector, it's also possible, though much more
involved, to extend Superset and write
your own connector. The only example of this at the moment is the Druid
connector, which is getting superseded by Druid's growing SQL support and
the recent availability of a DBAPI and SQLAlchemy driver. If the database
you are considering integrating has any kind of of SQL support, it's probably
preferable to go the SQLAlchemy route. Note that for a native connector to
be possible the database needs to have support for running OLAP-type queries
and should be able to things that are typical in basic SQL:
- aggregate data
- apply filters (==, !=, >, <, >=, <=, IN, ...)
- apply HAVING-type filters
- be schema-aware, expose columns and types
Another work around is to change where superset stores the sqlite database by adding ``SQLALCHEMY_DATABASE_URI = 'sqlite:////new/localtion/superset.db'`` in superset_config.py (create the file if needed), then adding the directory where superset_config.py lives to PYTHONPATH environment variable (e.g. ``export PYTHONPATH=/opt/logs/sandbox/airbnb/``).
**Windows** isn't officially supported at this point, but if you want to
attempt it, download `get-pip.py <https://bootstrap.pypa.io/get-pip.py>`_, and run ``python get-pip.py`` which may need admin access. Then run the following: ::
@@ -47,17 +129,18 @@ attempt it, download `get-pip.py <https://bootstrap.pypa.io/get-pip.py>`_, and r
Python virtualenv
-----------------
It is recommended to install Superset inside a virtualenv. Python 3 already ships virtualenv, for
Python 2 you need to install it. If it's packaged for your operating systems install it from there
otherwise you can install from pip: ::
It is recommended to install Superset inside a virtualenv. Python 3 already ships virtualenv.
But if it's not installed in your environment for some reason, you can install it
via the package for your operating systems, otherwise you can install from pip: ::
pip install virtualenv
You can create and activate a virtualenv by: ::
# virtualenv is shipped in Python 3 as pyvenv
virtualenv venv
. ./venv/bin/activate
# virtualenv is shipped in Python 3.6+ as venv instead of pyvenv.
# See https://docs.python.org/3.6/library/venv.html
python3 -m venv venv
. venv/bin/activate
On windows the syntax for activating it is a bit different: ::
@@ -80,7 +163,7 @@ Follow these few simple steps to install Superset.::
# Install superset
pip install superset
# Create an admin user
# Create an admin user (you will be prompted to set a username, first and last name before setting a password)
fabmanager create-admin --app superset
# Initialize the database
@@ -92,11 +175,8 @@ Follow these few simple steps to install Superset.::
# Create default roles and permissions
superset init
# Start the web server on port 8088
superset runserver -p 8088
# To start a development web server, use the -d switch
# superset runserver -d
# To start a development web server on port 8088, use -p to bind to another port
superset runserver -d
After installation, you should be able to point your browser to the right
@@ -106,10 +186,61 @@ the credential you entered while creating the admin account, and navigate to
your datasources for Superset to be aware of, and they should show up in
`Menu -> Datasources`, from where you can start playing with your data!
Please note that *gunicorn*, Superset default application server, does not
work on Windows so you need to use the development web server.
The development web server though is not intended to be used on production systems
so better use a supported platform that can run *gunicorn*.
A proper WSGI HTTP Server
-------------------------
While you can setup Superset to run on Nginx or Apache, many use
Gunicorn, preferably in **async mode**, which allows for impressive
concurrency even and is fairly easy to install and configure. Please
refer to the
documentation of your preferred technology to set up this Flask WSGI
application in a way that works well in your environment. Here's an **async**
Make sure the user has privileges to access and use all required
databases/schemas/tables/views/warehouses, as the Snowflake SQLAlchemy engine does
not test for user rights during engine creation.
See `Snowflake SQLAlchemy <https://github.com/snowflakedb/snowflake-sqlalchemy>`_.
Teradata
---------
The connection string for Teradata looks like this ::
teradata://{user}:{password}@{host}
*Note*: Its required to have Teradata ODBC drivers installed and environment variables configured for proper work of sqlalchemy dialect. Teradata ODBC Drivers available here: https://downloads.teradata.com/download/connectivity/odbc-driver/linux
Note that the above command will install Superset into ``default`` namespace of your Kubernetes cluster.
Custom OAuth2 configuration
---------------------------
Beyond FAB supported providers (github, twitter, linkedin, google, azure), its easy to connect Superset with other OAuth2 Authorization Server implementations that support "code" authorization.
The first step: Configure authorization in Superset ``superset_config.py``.
..code-block::python
AUTH_TYPE=AUTH_OAUTH
OAUTH_PROVIDERS=[
{'name':'egaSSO',
'token_key':'access_token',# Name of the token in the response of access_token_url
'icon':'fa-address-card',# Icon for the provider
'remote_app':{
'consumer_key':'myClientId',# Client Id (Identify Superset application)
'consumer_secret':'MySecret',# Secret for this Client Id (Identify Superset application)
'request_token_params':{
'scope':'read'# Scope for the Authorization
},
'access_token_method':'POST',# HTTP Method to call access_token_url
'access_token_params':{# Additional parameters for calls to access_token_url
'client_id':'myClientId'
},
'access_token_headers':{# Additional headers for calls to access_token_url
As mentioned in the `Installation & Configuration <https://superset.incubator.apache.org/installation.html#installation-configuration>`_ documentation,
it's possible for administrators to expose more more macros in their
environment using the configuration variable ``JINJA_CONTEXT_ADDONS``.
All objects referenced in this dictionary will become available for users
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.