Skip to content

Releases: atlanhq/atlan-java

v1.11.5

15 May 09:03
Compare
Choose a tag to compare

πŸŽ‰ New features

  • Adds latest CosmosDB asset model
  • (Experimental) Adds generation of baseline user interface code for new typedefs to the typedef toolkit

🐞 Bug fixes

  • Updates data domain and data product creators to latest model requirements (as this area approaches GA)

πŸ“¦ Packages

  • Adds a new custom package to extend custom metadata scope to additional connections' and/or glossaries' assets
  • Adds displayName as an enriched attribute for the asset export (basic) and enrichment migrator packages
  • Adds more detailed diagnostics information (debug-level logs) for custom packages, to aid in future troubleshooting
  • Fixes data products import idempotency in asset import package
  • Fixes data products import to allow an empty asset selection DSL
  • Restores the Always image pull policy for packages during their snapshot releases, to simplify pre-release testing

πŸ₯— QOL improvements

  • Bumps default page size for FluentLineage from 10 to 1000 to improve performance of cycling through complete end-to-end lineage for large lineage graphs

v1.11.4

02 May 19:36
Compare
Choose a tag to compare

πŸŽ‰ New features

  • Adds ability to manage SSO group mappings programmatically
  • Adds ability to retrieve the depth of an asset as part of lineage responses
  • Adds further connector types (Netezza, Azure Service Bus, Cognite, and Syndigo)
  • (Experimental) Adds generation of connector credentials configuration to the custom package toolkit
  • (Experimental) Adds widget for using connector credentials as part of inputs for a custom package to the custom package toolkit
  • (Experimental) Adds additional widget options to the custom package toolkit, for example default values and disabling inputs
  • (Experimental) Adds code generator to the custom package toolkit to produce strongly-typed configuration for Python
  • Updates models to include closed preview access to data contracts

πŸ“¦ Packages

  • Adds the ability to include archived assets to the asset export (basic) package and the enrichment migrator package
  • Adds the ability for the asset import package to handle data domains and data products
  • Fixes an issue in custom packages that send emails with attachments, where the attachment could be deleted before the email is sent
  • Fixes an issue in custom packages that process CSV files to avoid whitespace, byte order marks and zero-width spaces in the header from causing the input to be misinterpreted

(Note: this is effectively a re-release of v1.11.3, which was missing some dependencies to fix the email issue above.)

v1.11.3

02 May 17:18
Compare
Choose a tag to compare

πŸŽ‰ New features

  • Adds ability to manage SSO group mappings programmatically
  • Adds ability to retrieve the depth of an asset as part of lineage responses
  • Adds further connector types (Netezza, Azure Service Bus, Cognite, and Syndigo)
  • (Experimental) Adds generation of connector credentials configuration to the custom package toolkit
  • (Experimental) Adds widget for using connector credentials as part of inputs for a custom package to the custom package toolkit
  • (Experimental) Adds additional widget options to the custom package toolkit, for example default values and disabling inputs
  • (Experimental) Adds code generator to the custom package toolkit to produce strongly-typed configuration for Python
  • Updates models to include closed preview access to data contracts

πŸ“¦ Packages

  • Adds the ability to include archived assets to the asset export (basic) package and the enrichment migrator package
  • Adds the ability for the asset import package to handle data domains and data products
  • Fixes an issue in custom packages that send emails with attachments, where the attachment could be deleted before the email is sent
  • Fixes an issue in custom packages that process CSV files to avoid whitespace, byte order marks and zero-width spaces in the header from causing the input to be misinterpreted

v1.11.2

19 Apr 11:40
Compare
Choose a tag to compare

πŸŽ‰ New features

  • Adds creator() helpers for Airflow assets
  • (Experimental) Adds further elements for generating the UX from the typedef toolkit
  • (Experimental) Adds a truncated index option to typedef toolkit

🐞 Bug fixes

  • Ensures qualifiedName is accessible on assets, even if it is embedded in uniqueAttributes
  • Ensures GUID assignment is tracked during batch operations, even when the GUID is reused as-is from what was sent in the request
  • (Experimental) Corrects URL pointer in the custom package toolkit
  • (Experimental) Fixes attributeDefs in the typedef toolkit to ensure it is always included (even if empty)

πŸ₯— QOL improvements

  • Makes logging clearer in cases where custom metadata used in an asset does not actually exist, causing the asset save to fail

v1.11.1

12 Apr 13:17
Compare
Choose a tag to compare

πŸŽ‰ New features

  • Adds AssetFilterGroup to enumerate the options for persona personalization of asset filters
  • (Experimental) Allows you to specify multiple supertypes when defining new custom types using the typedef toolkit
  • (Experimental) Adds a new rawJsonObject to all objects, containing the raw API response, allowing you to access any information in the API response that may not (yet) be modeled

🐞 Bug fixes

  • Fixes the parameter used for lineage-only propagation (now called restrictPropagationThroughHierarchy)

v1.11.0

29 Mar 17:55
Compare
Choose a tag to compare

πŸŽ‰ New features

  • (Experimental) Refactors the custom package config toolkit to use Pkl, to be language-agnostic

🐞 Bug fixes

  • Fixes a missing icon that could be used for tags, and would cause deserialization exceptions if used
  • Fixes an issue where certain single-cardinality structs were treated as multi-cardinality, which could cause deserialization exceptions

πŸ“¦ Packages

  • Fixes an issue parsing CSV files that do not use , as a field separator in the lineage builder package

πŸ₯— QOL improvements

  • Refactors the Gradle build process to simplify releases and version consistency

v1.10.14

27 Mar 16:09
Compare
Choose a tag to compare

πŸŽ‰ New features

  • Adds ability to exclude users from search log requests (most viewed assets, most recent viewers, etc)
  • Adds method to update an existing enumeration with more valid values
  • (Experimental) Adds an option for propagating tags only through lineage (not via hierarchy)

🐞 Bug fixes

  • Fixes tracking of failed assets during batch operations

πŸ“¦ Packages

  • Fixes a NullPointerException that could occur when invalid connections exist in a tenant
  • Adds user-provided designation (role) to the admin export package

πŸ₯— QOL improvements

  • Adds task queue introspection to deletion blocking, for more consistent results from the block() call on deletion operations
  • Adds further options (methods) for looking up and retrieving user information, including returning users' details after creating them, if desired

v1.10.13

20 Mar 22:44
Compare
Choose a tag to compare

πŸŽ‰ New features

  • (Experimental) Further typedef toolkit evolution (structs, tests)
  • (Experimental) Refactors how qualifiedNames are composed for multidimensional dataset (cube) assets

🐞 Bug fixes

  • Fixes issues with (experimental) typedef toolkit's generated output

πŸ“¦ Packages

  • Fixes a bug where some deployments were unable to run the asset importer due to an error about a missing library
  • Fixes an issue where source tag attachments could cause errors in export packages, if the source tag no longer exists
  • Fixes issue in cube asset builder with detecting the initial load of a cube
  • Refactors how qualifiedNames are composed in the cube asset builder package

πŸ₯— QOL improvements

  • Adds utility methods to package toolkit to allow syncing entire S3 prefixes (all objects therein) as well as individual objects

v1.10.12

13 Mar 12:34
Compare
Choose a tag to compare

πŸŽ‰ New features

  • (Experimental) Typedef toolkit
  • (Experimental) Model for multi-dimensional datasets (cubes)
  • Adds latest model for Spark jobs
  • Adds AWS Green Grass and Site Wise connector types

🐞 Bug fixes

  • Reduces default verbosity of caching-related errors (still included at debug-level)

πŸ“¦ Packages

  • (Experimental) Adds a new cube asset builder, for loading and managing multi-dimensional datasets (cubes)
  • Fixes issues with sending emails from custom packages: admin export, asset export (basic), adoption export, metadata impact report
  • Fixes ClassNotFoundException and IndexOutOfBoundsException in the OpenAPI spec loader package
  • Enrichment migrator package now allows copying enrichment to one or many target set of assets

βš™οΈ Background

  • Increases max retries for automated and parallelized integration tests, and improves connection creation retry logic there

v1.10.11

05 Mar 17:14
Compare
Choose a tag to compare

πŸŽ‰ New features

  • Adds latest ThoughtSpot and Domo models

🐞 Bug fixes

  • Fixes enum generator to support space-separated names
  • Fixes entity and relationship typedef creation with sensible defaults

πŸ“¦ Packages

  • Optimizes asset removal within the relational assets builder package (parallelism, skipping already-deleted assets)
  • Allows the field separator and batch size to be configured for import packages: asset import, enrichment migrator, lineage builder, relational assets builder

βš™οΈ Background

  • Updates all dependencies to their latest versions