v0.7.0

Changed

• Adds quotes to the attributes of PartiQL tuple's debug output so it can be read and transformed using Kotlin partiql-cli
• BREAKING: partiql-eval: Changes the interface to EvalPlan to accept an EvalContext
• BREAKING: partiql-eval: Changes EvaluationError to not implement Clone
• BREAKING: partiql-eval: Changes the structure of EvalPlan

Added

• partiql-extension-visualize: Add partiql-extension-visualize for visualizing AST and logical plan
• partiql-eval: Add a SessionContext containing both a system-level and a user-level context object usable by expression evaluation

Fixed

• partiql-logical-planner: Fixed ORDER BY's ability to see into projection aliases
• partiql-eval: Fixed errors in BaseTableExprs get added to the evaluation context
• partiql-eval: Fixed certain errors surfacing in Permissive evaluation mode, when they should only be present in Strict mode

v0.6.0

The main highlights of this release:

• >90% conformance tests passing
• Typing of simple SFW queries
• Initial steps to move builtin functions into the catalog
• Strict evaluation mode
• Performance improvements related to GROUP BY and aggregation evaluation

Changed

• BREAKING: partiql-value: BindingsName changed to hold Cow<str> rather than String
• BREAKING: partiql-eval: Construction of expression evaluators changed to separate binding from evaluation of expression. & implement strict eval
• BREAKING: partiql-value: Value trait's is_null_or_missing renamed to is_absent
• BREAKING: partiql-value: Value trait's coerce_to_tuple, coerece_to_bag, and coerce_to_list methods renamed to coerce_into_tuple, coerece_into_bag, and coerece_into_list.
• BREAKING: partiql-value: Tuple's pairs and into_pairs changed to return concrete Iterator types.
• BREAKING: partiql-eval: EvaluatorPlanner construction now takes an EvaluationMode parameter.
• BREAKING: partiql-eval: like_to_re_pattern is no longer public.
• BREAKING: partiql-value: Box Decimals in Value to assure Value fits in 16 bytes.
• BREAKING: partiql-logical-planner: moves NameResolver to partiql-ast-passes
• BREAKING: partiql-value: removes partiql from value macro_rules; e.g. partiql_bag renames to bag.
• BREAKING: partiql-ast: changed modeling of Query and SetExpr nodes to support ORDER BY, LIMIT, OFFSET in children of set operators
  • Affects the AST and visitor
• BREAKING: partiql-ast: rename of SetExpr to BagOpExpr and SetOp to BagOp
  • Affects the AST and visitor
• BREAKING: partiql-parser: Parsed struct's ast field is now an ast::AstNode<ast::TopLevelQuery>
• BREAKING: partiql-eval: Evaluable trait's update_input fn now also takes in an EvalContext
• BREAKING: partiql-logical: changed modeling of Project exprs to be a Vec<(String, ValueExpr)> rather than a HashMap<String, ValueExpr> to support multiple project items with the same alias
• BREAKING: partiql-logical: changed modeling of VarRef to include a VarRefType to indicate whether to do a local vs global binding lookup

Added

• Strict mode evaluation partial support added.
• Add interface for STRICT mode evalution to EvaluatorPlanner.
• Add ability for partiql-extension-ion extension encoding/decoding of Value to/from Ion Element
• Add partiql-types crate that includes data models for PartiQL Types.
• Add partiql_ast_passes::static_typer for type annotating the AST.
• Add ability to parse ORDER BY, LIMIT, OFFSET in children of set operators
• Add OUTER bag operator (OUTER UNION, OUTER INTERSECT, OUTER EXCEPT) implementation
• Add experimental partiql_logical_planner::typer for typing PartiQL queries with the initial support for simple SFW queries with SELECT and FROM clauses only with no operators, JOINs, etc.
• Add NullSortedValue to specify ordering null or missing values partiql_value::Values before or after all other values
• Implements the aggregation functions ANY, SOME, EVERY and their COLL_ versions
• Add COUNT(*) implementation
• Add to_vec method to List and Bag to convert to a Vec

Fixed

• Fixes parsing of multiple consecutive path wildcards (e.g. a[*][*][*]), unpivot (e.g. a.*.*.*), and path expressions (e.g. a[1 + 2][3 + 4][5 + 6])—previously these would not parse correctly.
• partiql-parser set quantifier for bag operators fixed to DISTINCT
• partiql-parser set quantifier for bag operators fixed to be DISTINCT when unspecified
• partiql-logical-planner add error for when a HAVING is included without GROUP BY
• Fixes variable resolution lookup order and excessive lookups
• Fixes variable resolution of some ORDER BY variables
• Fixes nested list/bag/tuple type ordering for when ASC NULLS LAST and DESC NULLS FIRST are specified
• partiql-value fix deep equality of list, bags, and tuples
• Fixes bug when using multiple aggregations without a GROUP BY
• Performance improvements to grouping/evaluation

v0.5.0

Changed

• BREAKING: partiql-eval: evaluate on Evaluable returns a Value rather than an Option<Value>
• BREAKING: partiql-ast: changes the modeling of Bag/List/Tuple literals

Added

• Ability to add and view errors during evaluation with partiql-eval's EvalContext
• AST sub-trees representing literal values are lowered to Values during planning

v0.4.1

Changed

• partiql-extension-ion-functions : Made IonExtension pub

Added

Fixes

v0.4.0

Changed

• BREAKING: partiql-eval: modifies visibility of types implementing EvalExpr and Evaluable
• BREAKING: removed from_ion method on Value
• BREAKING: partiql-ast: visit fn returns a partiql-ast::Recurse type to indicate if visitation of children nodes should continue
• BREAKING: partiql-logical-planner: modifies lower(parsed: &Parsed) to return a Result type of Result<logical::LogicalPlan<logical::BindingsOp>, LoweringError> rather than a logical::LogicalPlan<logical::BindingsOp>
• BREAKING: partiql-eval: modifies compile(&mut self, plan: &LogicalPlan<BindingsOp>) to return a Result type of Result<EvalPlan, PlanErr> rather than an EvalPlan
  • This is part of an effort to replace panics with Results
• BREAKING: partiql-logical-planner: Adds a LogicalPlanner to encapsulate the lower method
• BREAKING: partiql-eval: Adds a EvaluatorPlanner now requires a Catalog to be supplied at initialization
• BREAKING: partiql-logical-planner: CallDef and related types moved to partiql-catalog

Added

• Implements built-in function EXTRACT
• Add partiql-extension-ion extension for encoding/decoding Value to/from Ion data
• Add partiql-extension-ion-functions extension which contains an extension function for reading from an Ion file
• Add partiql-catalog including an experimental Catalog interface and implementation
• Implements the COLL_* functions -- COLL_AVG, COLL_COUNT, COLL_MAX, COLL_MIN, COLL_SUM

Fixes

• Fix parsing of EXTRACT datetime parts YEAR, TIMEZONE_HOUR, and TIMEZONE_MINUTE
• Fix logical plan to eval plan conversion for EvalOrderBySortSpec with arguments DESC and NULLS LAST
• Fix parsing of EXTRACT to allow keywords after the FROM

v0.3.0

Changed

• EvalExpr.evaluate function now returns a Cow of Value
• Evaluable trait's get_vars function returns by ref
• Refactor of partiql-eval crate
  • Operators previously implementing Evaluable (e.g. EvalScan, EvalFilter) are under the eval::evaluable module
  • Expressions previously implementing EvalExpr (e.g. EvalBinOpExpr, EvalLitExpr) are under the eval::expr module
• Refactor CallAgg partiql-ast node

Added

• Adds some benchmarks for parsing, compiling, planning, & evaluation
• Implements more built-in functions -- POSITION, OCTET_LEN, BIT_LEN, ABS, MOD, CARDINALITY, OVERLAY
• Implements PIVOT operator in evaluator
• Implements LIKE for non-string, non-literals
• serde feature to partiql-value and partiql-logical with Serialize and Deserialize traits
• Adds Display for LogicalPlan
• Expose partiql_value::parse_ion as a public API
• Adds some convenience methods on Value
  • Add Extend implementations for List and Bag
  • Add methods to iterate a Tuple's values without zipping its names
  • Allow collect() into a Tuple with any Into<String>
• Parse OUTER UNION/INTERSECT/EXCEPT
• Parse WITH clause
• Implements LIMIT and OFFSET operators in evaluator
• DATE/TIME/TIMESTAMP values
• Parse TABLE <id> references
• Implements GROUP BY operator in evaluator
• Implements HAVING operator in evaluator
• Implements ORDER BY operator in evaluator
• Implements SQL aggregation functions (AVG, COUNT, MAX, MIN, SUM) in evaluator

Fixes

• Some performance improvements from removing extraneous clones and tweaking buffer sizes
• Fix off by one error when checking preconditions to lower join ON
• Recognize aggregate fn names in parser
• Pass-through comments when processing special forms
• Make BY <x> optional in GROUP clause
• Fix JOIN parsing by defaulting to INNER and allowing elision of keywords
• Allow un-parenthesized subquery as the only argument of a function in parser
• Fix handling of List/Bag/Tuple in keyword argument preprocessing in parser
• Fixes Tuple value duplicate equality and hashing
• Properly skip comments when parsing

v0.2.0

This release is highlighted by the addition of end-to-end PartiQL query evaluation. This release introduces a logical plan and evaluation API along with AST lowering to logical plan with name resolution. There are some breaking changes to the parsed AST.

Changed

• BREAKING: Refactors the AST
  • Removes Location from the AST, replacing with a 'node id' that gives the AST node identity; the id can be used to retrieve Location
  • Removes redundancies and extraneous nesting
  • Refactor some AST nodes (including FROM, WHERE, and HAVING clauses) for better visitation
  • Refactor FromSource to not wrap in AstNode

Added

• Adds end-to-end PartiQL query evaluation with the following supported features
  • SELECT-FROM-WHERE
  • LATERAL LEFT, INNER, CROSS JOINs
  • UNPIVOT
  • SELECT VALUE
  • Query expressions
  • List, Bag, Tuple constructors
  • Path expressions (wildcard & unpivot path are not yet supported)
  • Subquery (supported in logical and eval plan; not yet in AST to plan conversion)
  • DISTINCT
  • Variable references
  • Literals
  • Arithmetic operators (+, -, *, /, %)
  • Logical operators (AND, OR, NOT)
  • Equality operators (= , !=)
  • Comparison operators (<, >, <=, >=)
  • IS [NOT] MISSING, IS [NOT] NULL
  • IN
  • BETWEEN
  • LIKE
  • Searched and simple case expressions
  • COALESCE and NULLIF
  • CONCAT
  • And the following functions
    • LOWER
    • UPPER
    • CHARACTER_LENGTH
    • LTRIM
    • BTRIM
    • RTRIM
    • SUBSTRING
    • EXISTS
• Adds Visit and Visitor traits for visiting AST
• Add AST node Visit impls via proc_macros
• Adds PartiQL Value, an in-memory representation of PartiQL values
  • Supports PartiQL values other than DATE, TIME, s-expressions
  • Supports basic arithmetic, logical, equality, and comparison operators
  • Supports partiql parsing of Ion into Value
• Defines logical plan and evaluation DAG
• AST lowering to logical plan with name resolution
• partiql-conformance-tests support for parsing and running evaluation tests from partiql-tests

v0.1.0

This release highlights a PartiQL Parser API that has been tested using Syntax success and fail tests that are defined in PartiQL conformance test suite; other highlights are, PartiQL CLI REPL and query visualizer, and an experimental PartiQL web playground.

As this is the first minor version after v0.0.0, it does not include any breaking changes.

Added

• Lexer & Parser for the majority of PartiQL query capabilities—see syntax success
  and fail tests for more details.
• AST for the currently parsed subset of PartiQL
• Tracking of locations in source text for ASTs and Errors
• Parser fuzz tester
• Conformance tests via test generation from partiql-tests
• PartiQL Playground proof of concept (POC)
• PartiQL CLI with REPL and query visualization features

Full Changelog: https://github.com/partiql/partiql-lang-rust/commits/v0.1.0