[2023-03-20] Library and Engine Team Update

[jdunkerley]

• James: Work on integrating ag-grid in place of JS table viz. Use AG Grid instead of the old table.js rendering #5915
• James: Design requirement for data grid view longer term. Table Visualization Requirements #5965
• James: Design for execution modes and the play button. Execution Modes, Configuration Modes and Play Buttons #5964 (awaiting review)
• James: Aligned type signatures for method. Fix up some type signatures... #5979
• Dmitry: Fixed tagValues for constructors so correct FQN. Fix qualified names in tag values #5978
• James: Resolved some issues with drop downs on methods with warnings attached. Deal with warnings attached to value when making widgets. #5994 (passed remaining issue to Hubert as low level in engine)
• Radek: Integrated new Value Type into the in-memory storage types and repaired all the failing tests.
• Radek: Working on deriving Value Type from the results of JDBC queries.
• Greg: Aligned the various accessor methods for groups on the Match_2 type. Truffle regex: finish group accessors #5883
• Greg: Working on the replace and replace all functions. Implement Regular Expression replace and update Text.replace to the new API #5959
• Jaroslav: Meta.get_constructor_fields works for a Type/ Meta.get_constructor_fields works for Type #5869
• Jaroslav: Defer serialization of IR.
• Jaroslav: Avoid UUIDs in IR files. Don't persist UUIDs in IR caches #5918
• Jaroslav: Adding Meta.Type. Introducing Meta.Type #5956
• Jaroslav: Fix for UpsertVisualizationJob performance. Dedicated Module.containsUUID method with a cache #5945
• Hubert: Avoid GraalVM installing component multiple times. Don't log installed engines and runtimes in prod #5900
• Hubert: Resolving issue with exports causing conflicts with FQNs. Fully qualified names including same type and module names are not resolved properly #5902
• Hubert: Adding indexing to buildStdLib. buildStdLib*** tasks do not re-generate the corresponding stdlib index #5999
• Hubert: Looking at the current execution context code.
• Hubert: Investigating issue in instrumentation such that callMethod is unset.
• Pavel M: Finished the benchmark analysis tool. Add engine benchmark analysis tool #5852
• Pavel M: Implement Enso-specific asserts. Implement Enso-specific asserts #5973
• Pavel M: Sort handles Incomparable types by default. sort handles incomparable values #5998
• Dmitry: Spurious visualization request timeouts. Do not cancel detach visualization jobs #5913
• Dmitry: Writing down IR caches slows IDE startup. Delay writing IR caches #5957
• Dmitry: Parallelize the SQL DB initialization. Delay initialization of SQL database #5970
• Dmitry: Parallelize the JSON protocol init. Delay initialization of JSON-RPC protocol #5997
• Pawel G: Drop down display labels from Single_Choice. support custom widget entry labels #5705
• Pawel G: Simplify Node Edit mode.
• Pawel G: Planning and tasking for widgets on nodes.

[radeusgd]

Progress update on Value_Types - we have type mappings between Value_Type and SQL types for SQLite and Postgres along with unit tests, although the mappings require some further care to be considered finished.

Instead of implementing our own logic for inferring result types which may easily get out of sync if Database versions differ, we now rely on the Database itself to tell us the result type:

(not doing a screenshot as my build is currently logging tons of debug information that is just noise for presenting)

> c2.execute_update 'CREATE TABLE t123 (x int2, y int4, z int8, w float);'
> t123 = c2.query (SQL_Query.Table_Name "t123")
> t123.info.print
   | Column | Items Count | Value Type       
---+--------+-------------+-------------------
 0 | x      | 0           | (Integer 16 bits)
 1 | y      | 0           | (Integer 32 bits)
 2 | z      | 0           | (Integer 64 bits)
 3 | w      | 0           | (Float 64 bits)

> (t123.at 'x' + t123.at 'y') . value_type
>>> (Integer 32 bits)

> (t123.at 'x' + t123.at 'z') . value_type
>>> (Integer 64 bits)

> (t123.at 'x' + t123.at 'w') . value_type
>>> (Float 64 bits)

> t123.aggregate [Aggregate_Column.Group_By 'z', Aggregate_Column.Sum 'x', Aggregate_Column.Maximum 'x'] . info . print
   | Column    | Items Count | Value Type       
---+-----------+-------------+-------------------
 0 | z         | 0           | (Integer 64 bits)
 1 | Sum x     | 0           | (Integer 64 bits)
 2 | Maximum x | 0           | (Integer 16 bits)