Tags: google/zetasql
Tags
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: Migrate from deprecated with() field to entry() field in ASTWithClause -- Change by ZetaSQL Team <[email protected]>: update AST docs -- Change by John Morcos <[email protected]>: Support annotations in SQL UDFs and UDAs. -- Change by ZetaSQL Team <[email protected]>: Add conditional logic around window function description for aggregate functions. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: use http archive ICU instead of BCR ICU. -- Change by ZetaSQL Team <[email protected]>: Add additional Rules on Default Label Options -- Change by ZetaSQL Team <[email protected]>: Reference implementation for JSON subscript syntax. -- Change by Alex Hirschberg <[email protected]>: SQLBuilder generates arr[idx] instead of arr[OFFSET(idx)] in DML update when possible. -- Change by ZetaSQL Team <[email protected]>: LIKE operator documentation cleanup -- Change by ZetaSQL Team <[email protected]>: Rename planning const to immutable const. -- Change by Brandon Dolphin <[email protected]>: Clarify comments for when to use BuiltinFunctionRegistry. -- Change by ZetaSQL Team <[email protected]>: Update information schema to add options on label and property definitions -- Change by ZetaSQL Team <[email protected]>: Add support for builtin Table-Valued Functions in BuiltinFunctionCache -- Change by Alex Hirschberg <[email protected]>: Update value_expr.cc to use correct product mode when printing type name -- Change by ZetaSQL Team <[email protected]>: adds compliance testing for LIKE ANY/SOME/ALL with collation -- Change by Divyanshu Ranjan <[email protected]>: Fix alignment gaps in QueryExpression. -- Change by ZetaSQL Team <[email protected]>: Add custom getters for WithClauseEntry fields that validates only one of the fields are populated -- Change by ZetaSQL Team <[email protected]>: Added the reference impl and compliance tests for column-level row identity columns. -- Change by ZetaSQL Team <[email protected]>: Add the serialization and deserialization of MeasureColumnDef:: row_identity_column_indices. -- Change by ZetaSQL Team <[email protected]>: Add links to the `FROM_BASE64` function and the `TO_BASE64` function in the Bytes type docs. -- Change by ZetaSQL Team <[email protected]>: Add MONTHS_BETWEEN signatures -- Change by ZetaSQL Team <[email protected]>: Improve tests for ADD_MONTHS -- Change by Divyanshu Ranjan <[email protected]>: Add support for named TABLE arg in TVF call. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: More compliance tests for aggregate filtering -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Added the measure rewriter support for column-level row identity columns. -- Change by ZetaSQL Team <[email protected]>: Change `absl::flat_hash_set` to `absl::btree_set` for row identity column indices in sample catalog. -- Change by ZetaSQL Team <[email protected]>: Support simplified group rows grammar -- Change by ZetaSQL Team <[email protected]>: Add operator_keyword_location to SetOperationItem. -- Change by ZetaSQL Team <[email protected]>: Mark FEATURE_SQL_GRAPH_SET_OPERATION_PROPAGATION_MODE as not in development. -- Change by Alex Hirschberg <[email protected]>: Avoid array-specific language around subscripts in `UPDATE ... SET` error messages -- Change by ZetaSQL Team <[email protected]>: When stringifying GQL FOR statements always generate WITH OFFSET -- Change by ZetaSQL Team <[email protected]>: Generate WITH when stringifying CALL PER() tvf() -- Change by ZetaSQL Team <[email protected]>: Rename ASTSelectWith to ASTWithModifier -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Support annotations in SQL UDFs and UDAs (TVFs in the child CL). -- Change by ZetaSQL Team <[email protected]>: Remove redundant comments in lazy_resolution_catalog.cc. -- Change by John Morcos <[email protected]>: Support annotations in SQL UDFs and UDAs (TVFs in the child CL). -- Change by ZetaSQL Team <[email protected]>: Update execute_query script GitOrigin-RevId: 2020059fa8d4b8b31150fd73b86453cd3fa1c4c7 Change-Id: Id90dc54a3958a83f5ae5de07fe278e55356fce58
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: Added TVFScan as a new measure column source. -- Change by ZetaSQL Team <[email protected]>: Add UUID to the JSON encoding table. -- Change by ZetaSQL Team <[email protected]>: Remove 2 redundant conditions. -- Change by ZetaSQL Team <[email protected]>: Support OPTIONS clauses for default label and property definitions -- Change by ZetaSQL Team <[email protected]>: Remove in_development from FEATURE_TVF_PSEUDO_COLUMNS flag -- Change by ZetaSQL Team <[email protected]>: Refactored the measure rewriter. -- Change by ZetaSQL Team <[email protected]>: Add missing LIKE ANY/ALL validation for disallowed syntax/clauses. -- Change by ZetaSQL Team <[email protected]>: Update `TvfSignature` to support returning the output schema as a `const Table*`. -- Change by ZetaSQL Team <[email protected]>: Fix SQLBuilder for CALL PER () tvf() when there are quotable identifiers in working table -- Change by ZetaSQL Team <[email protected]>: Fix SQLBuilder for GQL CALL PER with named arguments. -- Change by ZetaSQL Team <[email protected]>: Clean up deprecated WITH GROUP ROWS syntax (part 3) -- Change by ZetaSQL Team <[email protected]>: Remove in_development tag from `FEATURE_ALTER_VIEWS_ALTER_COLUMN_SET_OPTIONS`. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Modify SimpleCatalog serialization to skip built-in TableValuedFunctions if ignore_builtin is true. -- Change by Shannon Bales <[email protected]>: Cleanup AlgebrizedFunctionCall by braking it into smaller more manageable functions. -- Change by ZetaSQL Team <[email protected]>: Add the interface support for column-level row identity columns. -- Change by John Morcos <[email protected]>: Regularize annotation propagation for FLATTEN(row_type) now that it is integrated in the ComponentTypes() framework. -- Change by ZetaSQL Team <[email protected]>: Add argument names to variadic MAP function key and value arguments. -- Change by ZetaSQL Team <[email protected]>: Detect overlapping paths in the analyzer for `SET ... UPDATE` paths which include JSON subscripts with simple literals. -- Change by ZetaSQL Team <[email protected]>: Implement RowType::ComponentTypes() so that ROW type works with annotation frameworks and rewrites orthogonally. -- Change by ZetaSQL Team <[email protected]>: Clean up deprecated WITH GROUP ROWS syntax (part 2) -- Change by ZetaSQL Team <[email protected]>: Remove in_development from FEATURE_TVF_PSEUDO_COLUMNS flag -- Change by Shannon Bales <[email protected]>: Cleanup Algebrizer::AlgebrizeExpression. -- Change by Mike Kruskal <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Add testdata for aml graph in sample catalog GitOrigin-RevId: 1a5041213d7ae8e478d9dd744e556a4161e708d8 Change-Id: I50703e7d7666d3a45e70d985ec0d870a9065b40d
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: update ZetaSQL README doc -- Change by ZetaSQL Team <[email protected]>: The feature will go public. -- Change by ZetaSQL Team <[email protected]>: Update ZetaSql documentation. -- Change by ZetaSQL Team <[email protected]>: `rewrite_utils.cc` should always match any ColRef it creates with the same AnnotatedType of the referenced column. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Remove deprecated functions -- Change by ZetaSQL Team <[email protected]>: Stronger compile-time errors for `node->Is<SUBTYPE>()` when invalid -- Change by ZetaSQL Team <[email protected]>: Stronger compile-time errors for `node->Is<SUBTYPE>()` when invalid -- Change by Jeff Shute <[email protected]>: Add Table and Resolved AST API for TableScans to read by Column -- Change by ZetaSQL Team <[email protected]>: `rewrite_utils.cc` should always match any ColRef it creates with the same AnnotatedType of the referenced column. -- Change by Jeff Shute <[email protected]>: Exclude huge parser alternations tests from asan/msan. -- Change by ZetaSQL Team <[email protected]>: update zetasql version. -- Change by ZetaSQL Team <[email protected]>: Enable collation for the all_rewrites drivers for compliance tests. -- Change by ZetaSQL Team <[email protected]>: clean up ZetaSql WORKSPACE files -- Change by ZetaSQL Team <[email protected]>: skip long double related tests when it has the same size as double. -- Change by ZetaSQL Team <[email protected]>: update AST docs -- Change by ZetaSQL Team <[email protected]>: Implement analyzer support for SQL Pipes `AGGREGATE WITH DIFFERENTIAL_PRIVACY`. -- Change by ZetaSQL Team <[email protected]>: Updates LinkfyNodeNames to not escape HTML link. -- Change by ZetaSQL Team <[email protected]>: NO-OP moves util functions from rewrite_utils to like_any_all_rewriter.cc as they are only used there. -- Change by ZetaSQL Team <[email protected]>: Removed unused variable -- Change by ZetaSQL Team <[email protected]>: Unify SQL result border style with prescribed formatting for consistency throughout docs. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by Jeff Shute <[email protected]>: Rewriter support to visit and rewrite Type fields -- Change by ZetaSQL Team <[email protected]>: Mark FEATURE_TIMESTAMP_PICOS and FEATURE_TIMESTAMP_PRECISION as no longer in_development. -- Change by ZetaSQL Team <[email protected]>: Implement truncating a picos value when assigning to a proto timestamp field in DML. -- Change by ZetaSQL Team <[email protected]>: Add a note to the `ERROR` function example -- Change by ZetaSQL Team <[email protected]>: Remove GCC dependency -- Change by ZetaSQL Team <[email protected]>: Add a new LanguageFeature `FEATURE_LEGACY_BINARY_CONCAT`. -- Change by ZetaSQL Team <[email protected]>: Record parse location for ResolvedSequence Node -- Change by Divyanshu Ranjan <[email protected]>: Set `uses_upper_case_sql_name` false when adding TVFs in `LazyResolutionCatalog`. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by Shannon Bales <[email protected]>: Add feature labels for more Pipe operators. -- Change by Brandon Dolphin <[email protected]>: Add conversion for UINT64 in struct positional access. -- Change by ZetaSQL Team <[email protected]>: Add picosecond-level granularity to TIMESTAMP and related docs. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by Christoph Dibak <[email protected]>: Remove k_threshold from DIFFERENTIAL_PRIVACY documentation. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Do not call ExecuteQueryWithResult in dry_run mode from ScriptExecutorImpl. -- Change by Jeff Shute <[email protected]>: Make join columns work as joins in array scans and FLATTEN. -- Change by ZetaSQL Team <[email protected]>: explicitly cast Value* type to void* to suppress warning -- Change by ZetaSQL Team <[email protected]>: fixes ExprResolutionInfo declaration error -- Change by Jeff Shute <[email protected]>: Bind Catalog callbacks into RowType so it can do field lookups. -- Change by Brandon Dolphin <[email protected]>: Add explicit crash warning comments to all Value getters. -- Change by ZetaSQL Team <[email protected]>: Add JSON_FLATTEN function -- Change by ZetaSQL Team <[email protected]>: Call evaluator when the validation is terminated in ScriptExecutorImpl. -- Change by ZetaSQL Team <[email protected]>: Clean up deprecated files after ZetaSql modernization. -- Change by ZetaSQL Team <[email protected]>: Remove the unused method TestDriver::PreloadTypesAndFunctions(). -- Change by ZetaSQL Team <[email protected]>: Adds an option to return all insert rows from INSERT OR IGNORE DML including the ignored rows from execution in reference implementation. -- Change by ZetaSQL Team <[email protected]>: Support Sequence serialization between Java and c++ -- Change by ZetaSQL Team <[email protected]>: Add variadic MAP function compliance tests. -- Change by Jesse Rosenstock <[email protected]>: No public description -- Change by Shannon Bales <[email protected]>: Add machinery that enables the Resolver to insert feature labels. -- Change by ZetaSQL Team <[email protected]>: supports GQL FULL UNION -- Change by Divyanshu Ranjan <[email protected]>: Replace `FullName()` with `SQLName(/*enable_case_conversion_for_non_builtins=*/false)` in TVF error messages. -- Change by Jesse Rosenstock <[email protected]>: Update ExactFloat -- Change by ZetaSQL Team <[email protected]>: Improve error messages for casting MAP types. -- Change by ZetaSQL Team <[email protected]>: Upgrade to C++20 & migrate to Bzlmod -- Change by ZetaSQL Team <[email protected]>: Reserve space for STRUCT fields in TypeDeserializer -- Change by ZetaSQL Team <[email protected]>: Avoid re-resolution of finalized columns. -- Change by ZetaSQL Team <[email protected]>: Start migrating TestDriver::CreateDatabase() to use a TestDatabaseProto, to ensure all descriptors and types are internalized. -- Change by ZetaSQL Team <[email protected]>: removes redundant like_some.test compliance test file -- Change by ZetaSQL Team <[email protected]>: Improve error messages for array element casting failures. -- Change by Jeff Shute <[email protected]>: Bind more info into RowType for MULTIROW and join columns. -- Change by ZetaSQL Team <[email protected]>: Refactor GraphPathOp to support path mode locally -- Change by ZetaSQL Team <[email protected]>: Apply a restriction on references to SELECT column aliases in arguments to aggregations in other SELECT columns or in the WHERE clause when there's grouping or aggregation. -- Change by Jeff Shute <[email protected]>: Attach TpchCatalog in analyzer tests, and add some very basic tests for join columns. -- Change by ZetaSQL Team <[email protected]>: extract components to SetOperationResolverBase -- Change by ZetaSQL Team <[email protected]>: analyzer and reference implementation fix for the NOT LIKE ANY/ALL (subquery) case -- Change by ZetaSQL Team <[email protected]>: Let ExprResolutionInfo subscribe to its NameScope, in order to be notified when it represents a SELECT expression and a subquery in it references a column laterally from the prior SELECT column. -- Change by ZetaSQL Team <[email protected]>: Improve error messaging for STRUCT casting. -- Change by Jeff Shute <[email protected]>: Support join columns in Table, and add them in TpchCatalog. -- Change by Jeff Shute <[email protected]>: Fix for missing SELECT * EXCEPT/REPLACE hanlding for graph elements. -- Change by Jeff Shute <[email protected]>: Get star expansion working on ROW types. -- Change by Jeff Shute <[email protected]>: Make columns under ROW types work. -- Change by ZetaSQL Team <[email protected]>: Minor fix on RET_CHECK error message -- Change by Jeff Shute <[email protected]>: Add basic ROW type. -- Change by ZetaSQL Team <[email protected]>: Refactor TVFComputeResultTypeCallback to return unique_ptr. -- Change by ZetaSQL Team <[email protected]>: Allow DOT_PRODUCT as a valid distance type in vector_search TVF. -- Change by ZetaSQL Team <[email protected]>: Internal Change (And 232 more changes) GitOrigin-RevId: c92777b9fbdf00330c807b30e2b431842954e320 Change-Id: I744ade708bdb552c9cf7b52285bbf0e908656745
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: Move analyzer DP basic tests with set operations into their own dedicated file. -- Change by ZetaSQL Team <[email protected]>: Move analyzer DP tests with set operations related to the privacy user ID column to their own file. -- Change by ZetaSQL Team <[email protected]>: Update one of the comment symbols in AS-pipe-operator.md -- Change by ZetaSQL Team <[email protected]>: Remove incorrect note about pipe operators from the 'Pipe syntax' section -- Change by ZetaSQL Team <[email protected]>: Move DP tests for userid consistency with set operations in their own file. -- Change by ZetaSQL Team <[email protected]>: Increase timeouts for long-running tests -- Change by ZetaSQL Team <[email protected]>: Prevent measure propagation through OUTER JOINs in ArrayScans. -- Change by ZetaSQL Team <[email protected]>: Ensure correct error message when signature does not match for REGEXP_EXTRACT_GROUPS -- Change by ZetaSQL Team <[email protected]>: Prevent measure propagation through OUTER JOINs in ArrayScans. -- Change by ZetaSQL Team <[email protected]>: Add documentation for CREATE PROCEDURE statement. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Update reference implementation to support LIMIT ALL/LIMIT <null> -- Change by ZetaSQL Team <[email protected]>: Add differential privacy report and tablesample analyzer test files to ZetaSQL allowlist. -- Change by ZetaSQL Team <[email protected]>: Move default language features to top of differential privacy analyzer test file. -- Change by ZetaSQL Team <[email protected]>: Update ResolvedAST docs -- Change by ZetaSQL Team <[email protected]>: Move Analyzer tests for differential privacy queries using ARRAY into a dedicated file. -- Change by ZetaSQL Team <[email protected]>: Split OPTIONS clause-focused Analyzer DIFFERENTIAL_PRIVACY tests into their own file. -- Change by ZetaSQL Team <[email protected]>: Move Analyzer tests for differential privacy with FROM clause subqueries into dedicated file. -- Change by Jeff Shute <[email protected]>: Add resolved AST markers for ResolvedColumn creation vs references. -- Change by Christoph Dibak <[email protected]>: Add compliance test for nested queries -- Change by Christoph Dibak <[email protected]>: Allow dp queries as public group joins -- Change by ZetaSQL Team <[email protected]>: Documentation for REGEXP_EXTRACT_GROUPS -- Change by Brandon Dolphin <[email protected]>: Set moderate timeout for .../analyzer:analyzer_test. -- Change by ZetaSQL Team <[email protected]>: Refactor quantified path algebrizer test and minor fix its algebrizer -- Change by ZetaSQL Team <[email protected]>: Allow custom array size limit in GenerateArrayHelper -- Change by ZetaSQL Team <[email protected]>: Move Analyzer tests for differential privacy report format into dedicated file. -- Change by ZetaSQL Team <[email protected]>: Analyzer support for `UPDATE ... SET` statements for the JSON subscript operator. -- Change by ZetaSQL Team <[email protected]>: Clarify comment in 'AS pipe operator' example -- Change by ZetaSQL Team <[email protected]>: Update the script generating the AST docs to also generate the parser AST docs. -- Change by ZetaSQL Team <[email protected]>: mark FEATURE_MULTI_GROUPING_SETS not in development. -- Change by ZetaSQL Team <[email protected]>: Move analyzer tests for TABLESAMPLE support with DP queries into their own dedicated file. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Remove obsolete SELECT WITH DIFFERENTIAL_PRIVACY test. -- Change by ZetaSQL Team <[email protected]>: Documentation clarification about `ARRAY_AGG`. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: update behaviors when FEATURE_GROUPING_SETS is not enabled while FEATURE_MULTI_GROUPING_SETS is enabled. -- Change by Jeff Shute <[email protected]>: Run pyformat on all of gen_resolved_ast.py. -- Change by Divyanshu Ranjan <[email protected]>: Add test where relational argument is passed to BuiltinTableValuedFunction::CreateCall -- Change by ZetaSQL Team <[email protected]>: Support required language features in TableValuedFunctionOptions. -- Change by ZetaSQL Team <[email protected]>: Refactor & Simplify the rule for braced_ctor_extension_expr to remove the %prec directive. -- Change by ZetaSQL Team <[email protected]>: Disallow TVFs from having both graph and scalar-only signatures. -- Change by ZetaSQL Team <[email protected]>: Support pseudo-columns in value tables passed into TVFs -- Change by ZetaSQL Team <[email protected]>: Only create column annotation for timestamp precision when the target type is also a timestamp -- Change by Shannon Bales <[email protected]>: No public description -- Change by Divyanshu Ranjan <[email protected]>: Change signature of `TableValuedFunction::CreateEvaluator` to remove const requirement on TvfEvaluatorArg. -- Change by ZetaSQL Team <[email protected]>: RQG support for generating WHERE filter in aggregate function calls. -- Change by ZetaSQL Team <[email protected]>: Updated the ReferenceDriver to support executing DDLs by default when not used as a reference. GitOrigin-RevId: 5bd0300a28d6d86e224393b084a373f7e55862a8 Change-Id: Ifae8e9284316d42639bfb42736438a5c4b597ecd
Export of internal ZetaSQL changes. -- Change by Brandon Dolphin <[email protected]>: Fix crash in differential privacy functions with report. -- Change by ZetaSQL Team <[email protected]>: Updated the resolved ast doc for the ZetaSQL release 2025.6.1 -- Change by ZetaSQL Team <[email protected]>: Fix outdated documentation for the --rewrites flag -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Fix a bug where SQLBuilder generated impl... -- Change by ZetaSQL Team <[email protected]>: Add warning comment about the dependency of the REWRITE_INLINE_SQL_FUNCTION rewrite on having AnalyzerOptions.column_id_sequence_number set. -- Change by ZetaSQL Team <[email protected]>: Fix a bug where SQLBuilder generated implicit aliases to function args for which the ResolvedAST specified had no alias. -- Change by ZetaSQL Team <[email protected]>: Support multiway_unnest rewrite with non-constant join_expr -- Change by ZetaSQL Team <[email protected]>: Implement `JsonFromTimestamp()` for TimestampPicosValue input. -- Change by ZetaSQL Team <[email protected]>: Refactor the very large function GetJSONFunctions() into smaller helper functions. -- Change by ZetaSQL Team <[email protected]>: split test cases to fix regression test in zetasql. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Account for collation making different strings compare as equals when computing IsUniquelyOrdered(), which is used to signal nondeterminism. -- Change by ZetaSQL Team <[email protected]>: Migrate ArrayAnnotationMap to StructAnnotationMap. -- Change by ZetaSQL Team <[email protected]>: Handle is_outer and join_expr in multiway_unnest_rewriter -- Change by ZetaSQL Team <[email protected]>: Fix error location for GROUP BY in graph queries -- Change by ZetaSQL Team <[email protected]>: Refactor `measure_expression` library. -- Change by ZetaSQL Team <[email protected]>: Refactor construction of tables with measures in the SampleCatalog. -- Change by Shannon Bales <[email protected]>: Simplify `graph_element_pattern_filler`. Provide a more detailed description of the grammar conflict at a slightly higher level. Add test cases for the second form of the grammar conflict including the true conflict. -- Change by ZetaSQL Team <[email protected]>: Add compliance tests with measure expressions containing subqueries. -- Change by ZetaSQL Team <[email protected]>: Support subqueries in measure expressions. -- Change by Jeff Shute <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by Shannon Bales <[email protected]>: Fix the location of Graph return clause lists so that they do not include leading whitespace. -- Change by Jeff Shute <[email protected]>: Make the unversioned LanguageFeature enum names the primary name. -- Change by ZetaSQL Team <[email protected]>: Clarify how timestamps work with DST -- Change by ZetaSQL Team <[email protected]>: Fix bug with standalone expression resolution involving aggregate functions within subqueries. -- Change by ZetaSQL Team <[email protected]>: Move `ExtractTopLevelAggregates` to a separate library. -- Change by ZetaSQL Team <[email protected]>: Make the unversioned LanguageFeature enum names the primary name. -- Change by ZetaSQL Team <[email protected]>: Generalize annotation maps for composite types. -- Change by ZetaSQL Team <[email protected]>: Refactor several very large function-registration functions t... -- Change by Jeff Shute <[email protected]>: Make the unversioned LanguageFeature enum names the primary name. -- Change by ZetaSQL Team <[email protected]>: Moved syntax and operators tests out of bytes.test to literals.test and operators.test. -- Change by ZetaSQL Team <[email protected]>: Clarify that ANY_VALUE only ignores nulls if there are other values to return. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Moving SQL tests from the compliance directory to the appropriate file in either the parser or analyzer directory depending on the error type. -- Change by ZetaSQL Team <[email protected]>: Release graph dynamic labels and properties -- Change by ZetaSQL Team <[email protected]>: Fix HTML errors -- Change by Shannon Bales <[email protected]>: Remove `BisonParser` class step 8 of 8: Merge whats left of the "bison_parser" library into "parser_internal". -- Change by Jeff Shute <[email protected]>: Update compliance tests to use shortened feature names. -- Change by Jeff Shute <[email protected]>: Make LanguageFeature->string converter use the shorter aliases. -- Change by Jeff Shute <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Unify SqlBuilder logic on CALL [PER()] tvf() to always use YIELD to properly alias output columns. -- Change by ZetaSQL Team <[email protected]>: Fix broken links. -- Change by ZetaSQL Team <[email protected]>: Refactor several very large function-registration functions that had huge (~20K) stack frames into families of smaller functions. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by Shannon Bales <[email protected]>: Remove `BisonParser` class step 7 or 8: Delete the `BisonParser` class and migrate its last two fields to the Textmapper generated `Parser` class. -- Change by Jeff Shute <[email protected]>: Update an error message to use the alias for a feature enum. -- Change by Jeff Shute <[email protected]>: Update analyzer code to use shortened feature names. -- Change by Jeff Shute <[email protected]>: Switch to unversioned enum names in all parser and analyzer tests. -- Change by ZetaSQL Team <[email protected]>: Refactor several very large function-registration functions that had huge (~20K) stack frames into families of smaller functions. -- Change by ZetaSQL Team <[email protected]>: add more descriptions in fuzzy search and tokenization docs -- Change by ZetaSQL Team <[email protected]>: Handle the deprecated timestamp_picos value in GetProtoFieldDefault(). -- Change by ZetaSQL Team <[email protected]>: Remove inaccurate comments on measure value serialization. -- Change by Jeff Shute <[email protected]>: Add another test to make sure feature aliases are defined as expected. -- Change by ZetaSQL Team <[email protected]>: Algebrization support for measures. -- Change by ZetaSQL Team <[email protected]>: Add supporting logic for measure evaluation. -- Change by ZetaSQL Team <[email protected]>: Refactor UDA algebrization logic. -- Change by Shannon Bales <[email protected]>: Remove `BisonParser` class step 6 of 8: Move the `Parse` function out of the class to set it free! -- Change by ZetaSQL Team <[email protected]>: Add methods for accessing meaningful data from Measure values. -- Change by Shannon Bales <[email protected]>: Remove `BisonParser` class step 5 of 8: Pull 'input' out of the parser class. -- Change by ZetaSQL Team <[email protected]>: Introduce internal error fix proto -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Fix a bug in SqlBuilder when referencing DML target columns through correlated references in GQL CALL () variable scope list. -- Change by Jeff Shute <[email protected]>: Add versionless aliases and more attributes on LanguageFeatures. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by Shannon Bales <[email protected]>: Remove `BisonParser` class step 4 of 8: Move ownership of `ParserRuntimeInfo` to the caller. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Support creating a TIMESTAMP literal with precision annotation -- Change by ZetaSQL Team <[email protected]>: Remove stale comment -- Change by Shannon Bales <[email protected]>: Remove `BisonParser` class step 3 or 8: Factor `ASTNodeFactory` out of `BisonParser`. -- Change by ZetaSQL Team <[email protected]>: Reference implementation for analytic functions in MATCH_RECOGNIZE's DEFINE clause. -- Change by Shannon Bales <[email protected]>: Cleanup prefix parsing by Textmapper's [native support](https://textmapper.org/documentation.html#prefix-parsing) and removing SetForceTerminate. (And 384 more changes) GitOrigin-RevId: 5e0ac4fd0c37dea5775d52ed12039f0c52e6fb72 Change-Id: I2d6853cd548f38cae1853d5121a13746156d6851
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Add another signature for `CastFormatTimestampToString(PicoTime)` that takes an `absl::TimeZone` parameter. -- Change by ZetaSQL Team <[email protected]>: Remove in_development for feature_tokenized_search. -- Change by Nevena Kotlaja <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Trivial changes to Graph docs produced by running `generate-all-docs.sh` -- Change by ZetaSQL Team <[email protected]>: Rewrite CTEs in reverse order in DP / Aggregation threshold queries -- Change by Christoph Dibak <[email protected]>: get group selection epsilon from FunctionEpsilonAssigner -- Change by Ilya Biryukov <[email protected]>: Remove a source file from the filegroup that contains only headers -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Create constant_utils and move constant detection helpers in -- Change by ZetaSQL Team <[email protected]>: Removed deprecated `AnalyzerOptions::Data::lookup_expression_column_callback`. -- Change by ZetaSQL Team <[email protected]>: Update resolved AST doc. -- Change by Brandon Dolphin <[email protected]>: Add mechanism in FunctionSignatureRewriteOptions to control validations for catalog lookups of builtins. -- Change by ZetaSQL Team <[email protected]>: Step 1 of updating pico representation in ValueProto, to be consistent with the approved design. -- Change by ZetaSQL Team <[email protected]>: Implement cast picotime to string with format string, i.e. YYYY-MM-DD style formatting. -- Change by ZetaSQL Team <[email protected]>: Don't manually generate copy of non-constructor arg. -- Change by Brandon Dolphin <[email protected]>: Remove in_development=true from FEATURE_DISABLE_VALIDATE_REWRITERS_REFER_TO_BUILTINS. -- Change by ZetaSQL Team <[email protected]>: Update graph query examples in reference guide -- Change by ZetaSQL Team <[email protected]>: Add IsAnalysisConst utility function -- Change by Brandon Dolphin <[email protected]>: Prevent map function rewriter from resolving non-builtin catalog objects. -- Change by Brandon Dolphin <[email protected]>: Clean up override to provide filtered catalog to TypeofFunctionRewriter. -- Change by Brandon Dolphin <[email protected]>: Filter catalog before calling builtin rewriters. -- Change by Brandon Dolphin <[email protected]>: No public description -- Change by Christoph Dibak <[email protected]>: Fix zero remaining epsilon -- Change by ZetaSQL Team <[email protected]>: Mark `FEATURE_V_1_4_MULTILEVEL_AGGREGATION_IN_UDAS` as no longer in_development -- Change by ZetaSQL Team <[email protected]>: Make ComputeOp preserve order if its input does. -- Change by ZetaSQL Team <[email protected]>: Add more analyzer tests for SQL UDFs, UDAs, and TVFs with multi level aggregates that use the inlining rewriters -- Change by ZetaSQL Team <[email protected]>: Removed redudant capture of this since Algebrizer::AlgebrizeUdaCall() is now static. -- Change by ZetaSQL Team <[email protected]>: Cleanup some includes and BUILD files. -- Change by ZetaSQL Team <[email protected]>: Removed `in_development` annotation from `FEATURE_V_1_4_PIPE_RECURSIVE_UNION`. -- Change by ZetaSQL Team <[email protected]>: Avoid deep copying of non-constructor fields twice. -- Change by ZetaSQL Team <[email protected]>: Clarify the behavior of %G and %g format elements, and ISO with non-ISO. -- Change by ZetaSQL Team <[email protected]>: Implemented the SQLBuilder support for pipe recursive union. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Fix up comments on recursive query related resolved AST nodes to include pipe RECURSIVE UNION as one of the forms. -- Change by ZetaSQL Team <[email protected]>: Support CREATE LOCALITY GROUP in parser -- Change by ZetaSQL Team <[email protected]>: Update Nullabel/Groupable Data Type Behavior for GRAPH_ELEMENT/GRAPH_PATH -- Change by Brandon Dolphin <[email protected]>: Add Bazel dependency for cc_rules, load cc_test from there. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Fix error message on RETURN/WITH DISTINCT for unsupported type -- Change by ZetaSQL Team <[email protected]>: Update sample queries. -- Change by ZetaSQL Team <[email protected]>: Marks the WINDOW pipe operator as deprecated; EXTEND is to be used instead. -- Change by ZetaSQL Team <[email protected]>: Add JSON_CONTAINS function. -- Change by ZetaSQL Team <[email protected]>: Fix regex matcher in runtime_expected_errors.cc for RQG. -- Change by Jeff Shute <[email protected]>: Mark pipe WINDOW operator as deprecated -- Change by ZetaSQL Team <[email protected]>: Refactor ProcessQuery() to return a unique_ptr instead of a raw pointer. -- Change by Brandon Dolphin <[email protected]>: Implement BuiltinOnlyCatalog for use in analyzer when only builtin objects should be referenced. -- Change by ZetaSQL Team <[email protected]>: Cleanup some includes and BUILD files. -- Change by Jeff Shute <[email protected]>: Compliance test for window functions and named windows in pipe operators. -- Change by ZetaSQL Team <[email protected]>: Enable macro expansion support in execute_query tool -- Change by ZetaSQL Team <[email protected]>: Enable prune_unused_columns in execute_query. -- Change by Jeff Shute <[email protected]>: Add a couple compliance tests for pipe WITH. -- Change by Jeff Shute <[email protected]>: Implement pipe WITH. -- Change by Brandon Dolphin <[email protected]>: Remove `in_development` for feature_tokenized_search. -- Change by Jeff Shute <[email protected]>: Implemented named windows support in pipe SELECT and EXTEND. -- Change by ZetaSQL Team <[email protected]>: Simplify the serialization code. -- Change by ZetaSQL Team <[email protected]>: Remove `in_development` for feature_tokenized_search. -- Change by ZetaSQL Team <[email protected]>: Support CTEs in SQL TVF bodies in the direct implementation. -- Change by ZetaSQL Team <[email protected]>: Direct implementation for SQL TVFs in the ref impl. -- Change by ZetaSQL Team <[email protected]>: Simplify the serialization code. -- Change by ZetaSQL Team <[email protected]>: Automated Code Change -- Change by Brandon Dolphin <[email protected]>: Correct FEATURE_DISABLE_VALIDATE_REWRITERS_REFER_TO_BUILTINS to be an unversioned feature. -- Change by ZetaSQL Team <[email protected]>: Refactor ResolvedInsertStmt in SQLBuilder to use ProcessQuery. -- Change by ZetaSQL Team <[email protected]>: DCHECK TypeFactory::AddDependency() if `other_type` is null -- Change by ZetaSQL Team <[email protected]>: Respect the new line characters in the file when fixing multiline comments. -- Change by ZetaSQL Team <[email protected]>: Add compliance test support for TVFs. -- Change by Brandon Dolphin <[email protected]>: Cleanup the BUILD for analyzer. -- Change by ZetaSQL Team <[email protected]>: Improve an error message for mismatched value table -- Change by Brandon Dolphin <[email protected]>: Add FEATURE_V_1_4_DISABLE_VALIDATE_REWRITERS_REFER_TO_BUILTINS. -- Change by ZetaSQL Team <[email protected]>: Fixed a compilation warning -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Ignore `FEATURE_TABLE_SYNTAX_RESOLVE_ARGUMENT_LAST` when resolving TVFs from modules. (And 585 more changes) GitOrigin-RevId: 58df110fb4601855918511d30ea3edb727ef879e Change-Id: Ib86b420f11c0015d8105e9fc0ea0bd535fd4e560
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: Update the resolved ast doc for ZetaSQL. -- Change by Brandon Dolphin <[email protected]>: Refactor out shared AddMeasureColumnsToTable function for working with measure columns. -- Change by Christoph Dibak <[email protected]>: Implement per-aggregation budgeting in the reference impl -- Change by Nevena Kotlaja <[email protected]>: No public description -- Change by Nevena Kotlaja <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Implement method TimestampPicoValue::FromString(). -- Change by ZetaSQL Team <[email protected]>: Fix standalone expression evaluation for multi-level aggregates. -- Change by ZetaSQL Team <[email protected]>: Add documentation for graph query syntax -- Change by ZetaSQL Team <[email protected]>: Fix a bug where information from `ExprResolutionInfo` is not propagated correctly up to parent `ExprResolutionInfos` for multi-level aggregation. -- Change by ZetaSQL Team <[email protected]>: Add SQL rewriter and missing analyzer tests for map_contains_key. -- Change by Brandon Dolphin <[email protected]>: Add deserialization for MeasureType. -- Change by ZetaSQL Team <[email protected]>: Add the parser implementation for pipe recursive union. -- Change by ZetaSQL Team <[email protected]>: Refactor ResolveMatchRecognize() into smaller methods. -- Change by ZetaSQL Team <[email protected]>: Add a RET_CHECK to ensure that PIVOT expressions in the reference implementation are not multi-level aggregations. -- Change by ZetaSQL Team <[email protected]>: Ensure that volatile grouping expressions are not considered grouping consts. -- Change by ZetaSQL Team <[email protected]>: Rename ResolvedMatchRecognizeEmptyPattern to ResolvedMatchRecognizePatternEmpty to have a consistent prefix with other nodes. -- Change by Brandon Dolphin <[email protected]>: Remove documented requirement for FEATURE_V_1_4_MULTILEVEL_AGGREGATION when analyzing measure expressions. -- Change by ZetaSQL Team <[email protected]>: Add more compliance tests for multi-level aggregation. -- Change by ZetaSQL Team <[email protected]>: Move the OPTIONS() list to the end of the MATCH_RECOGNIZE() syntax. -- Change by ZetaSQL Team <[email protected]>: Add a new option to AnalyzerOptions and ErrorMessageOptions to enable an enhanced form of error redaction which preserves some information about the error, while keeping the output stable. -- Change by ZetaSQL Team <[email protected]>: Coerce the predicate expression in MATCH_RECOGNIZE's DEFINE clause to BOOL before giving an error. -- Change by ZetaSQL Team <[email protected]>: Update the resolved ast documentation for the ZetaSQL release in Q4 2024. -- Change by ZetaSQL Team <[email protected]>: Add set operators to pipe syntax: UNION, INTERSECT, EXCEPT, CORRESPONDING -- Change by Yassin Ezbakhe <[email protected]>: `MatcherCollection`: Add `AddMatcher` method. -- Change by Jeff Shute <[email protected]>: Implement experimental terminal pipe operator for EXPORT DATA. -- Change by Brandon Dolphin <[email protected]>: Fix casing in table name extraction for `TABLE x` and `FROM x` referring to TVF args. -- Change by ZetaSQL Team <[email protected]>: Support pattern variable references in MATCH_RECOGNIZE's measures clause. -- Change by ZetaSQL Team <[email protected]>: Add SQL rewriter for map_get function. -- Change by Brandon Dolphin <[email protected]>: Update error message in JSON literal test. -- Change by ZetaSQL Team <[email protected]>: small changes -- Change by ZetaSQL Team <[email protected]>: Update COUNT and COUNTIF documentation. -- Change by Brandon Dolphin <[email protected]>: Fix typo in measure_expression target names. -- Change by Brandon Dolphin <[email protected]>: Add Java implementation of MeasureType. -- Change by ZetaSQL Team <[email protected]>: Add support for graph queries -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: internal -- Change by ZetaSQL Team <[email protected]>: Add MATCH_RECOGNIZE compliance test covering the scenario where a DEFINE expression evaluates to NULL rather than TRUE or FALSE. -- Change by ZetaSQL Team <[email protected]>: Remove unimplemented optimization-related items from the pattern matching api: -- Change by Brandon Dolphin <[email protected]>: Implement MeasureType::TypeNameWithModifiers. -- Change by Brandon Dolphin <[email protected]>: Fix typo in sql_test_base.cc variable name. -- Change by ZetaSQL Team <[email protected]>: Add ZetaSketch to the list of external applications that can be used to build sketches. -- Change by ZetaSQL Team <[email protected]>: Add anonymization threshold to the AnonymizationInfo class -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Fix struct construction syntax in examples of sparse vectors. -- Change by Jeff Shute <[email protected]>: Initial resolved AST and analyzer for pipe FORK. -- Change by ZetaSQL Team <[email protected]>: Internal fix. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Add INTERVAL to JSON encoding documentation. -- Change by ZetaSQL Team <[email protected]>: Internal changes. -- Change by ZetaSQL Team <[email protected]>: Validate composite measure. -- Change by Brandon Dolphin <[email protected]>: Add deprecation warning for `TABLE x` syntax cases which will be affected by resolution order change. -- Change by ZetaSQL Team <[email protected]>: Add SQL rewriter for MAP_CARDINALITY -- Change by Brandon Dolphin <[email protected]>: Add feature `FEATURE_TABLE_SYNTAX_RESOLVE_ARGUMENT_LAST` to preserve the old (incorrect) resolution order for `TABLE x` syntax in SQL TVF bodies. -- Change by ZetaSQL Team <[email protected]>: Analyzer support for ON CONFLICT clause in INSERT DML -- Change by ZetaSQL Team <[email protected]>: Fix a bug in the script executor where redeclaring a variable without type parameters in a session would cause an error -- Change by Lev Kandel <[email protected]>: Prepare code for breaking change in Protobuf C++ API. Protobuf 6.30.0 will change the return types of Descriptor::name() and other methods to absl::string_view. This makes the code work both before and after such a change. -- Change by ZetaSQL Team <[email protected]>: Fix an incorrect comment. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Change complexity limit for state machine algo from 15,000 states to 100,000 edges, as it's the number of edges, not states, that really determines the time and space requirements of the algorithm. Also, moved the check from NFABuilder into NFA itself so that it gets enforced inside of EpsilonRemover, which, in the worst case, can drastically increase the number of edges, while simultaneously reducing the number of states. -- Change by Jeff Shute <[email protected]>: Add some support for referencing Constants with values. -- Change by ZetaSQL Team <[email protected]>: Unify the creation of multi_level_aggregate_info QRI for all the paths that create a default. -- Change by John Fremlin <[email protected]>: Add RowIdentityColumns function to catalog Table -- Change by ZetaSQL Team <[email protected]>: Remove references to Bison from comments in gen_parse_tree.py. -- Change by ZetaSQL Team <[email protected]>: Remove "Bison" from the names of some functions and variables in keywords.h. -- Change by ZetaSQL Team <[email protected]>: Enable SELECT .. FOR UPDATE for external use. -- Change by ZetaSQL Team <[email protected]>: Fix the order of parameters in the Regexp::Extract function. -- Change by ZetaSQL Team <[email protected]>: Add support for UINT64 arguments to `TIMESTAMP_FROM_UNIX_SECONDS`, `TIMESTAMP_FROM_UNIX_MILLIS`, and `TIMESTAMP_FROM_UNIX_MICROS`. -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Enable multi-level aggregation in RQG. -- Change by ZetaSQL Team <[email protected]>: Fix a bug with deferred resolution for multi-level aggregates. -- Change by ZetaSQL Team <[email protected]>: Fix some formatting issues with ST_EQUALS. -- Change by ZetaSQL Team <[email protected]>: Remove "pipe operator" from pipe operator table to economize. -- Change by Laramie Leavitt <[email protected]>: Avoid creating a new PRNG on each test loop. -- Change by Jeff Shute <[email protected]>: Parser implementation for experimental pipe FORK. -- Change by ZetaSQL Team <[email protected]>: Internal Change (And 482 more changes) GitOrigin-RevId: 158e00ddbc5bb8b8ef75e6b6473daf4fcef6174a Change-Id: I6520749d9676a3fce2e84a49f3e1dd6804390ab7
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: Updated the instructions for running execute_query with docker on MacOS with M1/M2 chips. -- Change by ZetaSQL Team <[email protected]>: Add a note about MacOS users seeing the error `execute_query_macos cannot be opened because the developer cannot be verified.` -- Change by ZetaSQL Team <[email protected]>: Refactoring in preparation for UPDATE constructor. -- Change by ZetaSQL Team <[email protected]>: Change the ZetaSQL Dockerfile to support different build modes. -- Change by Jeff Shute <[email protected]>: Add tests that check that a sql file runs successfully in execute_query. -- Change by ZetaSQL Team <[email protected]>: add a new TO_JSON signature that supports arg `unsupported_fiels`. -- Change by Jeff Shute <[email protected]>: Add some more example queries in examples/pipe_queries. -- Change by Brandon Dolphin <[email protected]>: Begin adding Measure type to TypeProto. -- Change by ZetaSQL Team <[email protected]>: Add per column OPTIONS and WITH COLUMN OPTIONS to analyzer. -- Change by ZetaSQL Team <[email protected]>: Handle lambda functions directly in the BuiltinFunctionRegistry scalar function APIs. -- Change by ZetaSQL Team <[email protected]>: Add per column OPTIONS and WITH COLUMN OPTIONS to analyzer. -- Change by ZetaSQL Team <[email protected]>: Add optional_ref library. -- Change by John Fremlin <[email protected]>: Add a testcase for deeply nested structs and arrays in JSON -- Change by ZetaSQL Team <[email protected]>: Update the ZetaSQL documentation: -- Change by John Fremlin <[email protected]>: Truncate output for deeply nested array expressions in unparser -- Change by ZetaSQL Team <[email protected]>: Update pipe syntax docs with TW peer review edits -- Change by Jeff Shute <[email protected]>: Fix execute_query command line help. -- Change by ZetaSQL Team <[email protected]>: add a new named arg `unsupported_fiels` for the TO_JSON function. -- Change by John Fremlin <[email protected]>: Truncate output for deeply nested CASE expressions in unparser -- Change by ZetaSQL Team <[email protected]>: Add MAP_REPLACE signatures, and reference implementation for KV pairs version -- Change by ZetaSQL Team <[email protected]>: Remove unnecessarily explicit function registrations from reference_impl/function.cc -- Change by Jeff Shute <[email protected]>: Adjust text area size so results are more visible. -- Change by ZetaSQL Team <[email protected]>: Disable formatting of SQL inside non-multiline string literals. -- Change by ZetaSQL Team <[email protected]>: Fixed issue with formatting SQL inside string literals when input string contains \r\n line endings. -- Change by ZetaSQL Team <[email protected]>: Format textproto inside annotated string literal. -- Change by ZetaSQL Team <[email protected]>: Disambiguate between open and close brackets annotations for braced constructor syntax. -- Change by Jeff Shute <[email protected]>: Improve multi-statement output in execute_query web. -- Change by ZetaSQL Team <[email protected]>: add a new named arg `unsupported_fiels` for the TO_JSON function. -- Change by ZetaSQL Team <[email protected]>: add a new built-in enum `UnsupportedFields` to be used by TO_JSON. -- Change by ZetaSQL Team <[email protected]>: Record parse location for OrderByItem iff record type is not PARSE_LOCATION_RECORD_NONE. -- Change by ZetaSQL Team <[email protected]>: Unify Lambda and non-lambda AlgebrizeFunctionCall codepaths -- Change by ZetaSQL Team <[email protected]>: small formatting updates for named arguments -- Change by ZetaSQL Team <[email protected]>: Fix the example Docker image name in the ZetaSQL doc. -- Change by ZetaSQL Team <[email protected]>: Refactor the parse AST and the grammar to use postfix table operators (e.g. TABLESAMPLE) on ASTTableExpression. -- Change by ZetaSQL Team <[email protected]>: Fix ZetaSQL documentation. GitOrigin-RevId: a68e25b308dadf3e78c4d22ec41adf72f8b08e5b Change-Id: I586b6974dbdb4e2bb4c99ba641ef96916ec33ba6
Exported ZetaSQL changes. - Added the support for [SQL pipe syntax](https://research.google/pubs/pub1005959/) - Improved the `execute_query` with an interactive web UI and more functionality. - Added new and improved SQL language features. - Improved documentation. GitOrigin-RevId: 88446a33c3a4498dab3f5cf2a1fe92c9f56d9723 Change-Id: Ia8ba13c3131dfc37b8ca9c60c9fd92752eac941d
Export of internal ZetaSQL changes. -- Change by ZetaSQL Team <[email protected]>: Add "alwayslink=1" to the target compliance:sql_test_base. -- Change by ZetaSQL Team <[email protected]>: Docs for JSON_KEYS function. -- Change by ZetaSQL Team <[email protected]>: Clarify that INT64 is converted to a string in JavaScript -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: CL to regenerate artifacts. -- Change by ZetaSQL Team <[email protected]>: Add MAP_VALUES_SORTED_BY_KEY function -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: No public description -- Change by ZetaSQL Team <[email protected]>: Add map_keys_sorted, map_keys_unsorted, map_values_sorted, and map_values_unsorted functions. -- Change by ZetaSQL Team <[email protected]>: Update ZetaSQL docs GitOrigin-RevId: d6002ecc05e21d26aef5686de0d2188ff622e244 Change-Id: Ifd9836e733968d7cf6ac4f17d94d3748cd6e95fe
PreviousNext