Skip to content

Commit 19d4ffd

Browse files
TimelordUKclaude
TimelordUK
and
claude
committed
chore: Bump version to 1.64.0 and update CHANGELOG
Release 1.64.0 brings major execution mode unification, query transformation pipeline enhancements, and QUALIFY clause support. Major Features: - Execution Mode Unification (Phases 0-3 complete) - Both -f and -q modes use unified execution path - Full transformer support in both modes - Dependency-aware execution with preprocessing - Query Transformation Pipeline - WHERE clause alias expansion - GROUP BY alias expansion - HAVING auto-aliasing - ORDER BY expression support - QUALIFY Clause Support - Window function filtering (Snowflake/BigQuery syntax) - Works with all window functions - Debug Capabilities - --show-transformations flag - Nvim \st and \sz keymaps - Pipeline visualization Improvements: - Python-based examples test framework - Formal and smoke testing for examples - Enhanced Neovim plugin keymaps - Documentation updates Bug Fixes: - Temp table persistence in --execute-statement mode - Transformer pipeline in dependency-aware execution - Qualified name resolution regression - Test suite improvements 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
1 parent bdd1c10 commit 19d4ffd

2 files changed

Lines changed: 161 additions & 1 deletion

File tree

‎CHANGELOG.md‎

Lines changed: 160 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -5,6 +5,166 @@ All notable changes to SQL CLI will be documented in this file.
55
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
66
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
77

8+
## [1.64.0] - 2025-11-01
9+
10+
### ✨ Major Features
11+
12+
#### **Execution Mode Unification & Query Transformation Pipeline**
13+
Complete unification of script mode (`-f`) and query mode (`-q`) execution paths, bringing sophisticated query transformation capabilities to both modes.
14+
15+
**Unified Execution Architecture**:
16+
- **Single execution path** - Both `-f` scripts and `-q` queries now use the same underlying execution engine
17+
- **Consistent transformer support** - All query transformations (WHERE, GROUP BY, HAVING, ORDER BY alias expansion) work in both modes
18+
- **Dependency-aware execution** - `--execute-statement N` now applies full preprocessing pipeline
19+
- **Shared infrastructure** - Eliminates code duplication and ensures feature parity
20+
21+
**Query Transformation Pipeline** (now available in both modes):
22+
- **WHERE clause alias expansion** - Use SELECT aliases in WHERE: `SELECT value * 2 AS doubled FROM data WHERE doubled > 100`
23+
- **GROUP BY alias expansion** - Reference SELECT aliases in GROUP BY: `SELECT region, SUM(sales) AS total FROM data GROUP BY region HAVING total > 1000`
24+
- **HAVING auto-aliasing** - Automatic aliases for aggregate expressions in HAVING clause
25+
- **ORDER BY expression support** - Complex expressions in ORDER BY automatically moved to SELECT with hidden columns
26+
27+
**New Debug Capabilities**:
28+
- **`--show-transformations` flag** - See the complete transformation pipeline for any query
29+
- **Nvim `\st` keymap** - Visualize transformations for query at cursor in Neovim plugin
30+
- **`\sz` keymap** - Alternative transformations debug view
31+
- **Detailed pipeline output** - Shows original SQL, intermediate steps, and final transformed query
32+
33+
**ORDER BY Expression Support**:
34+
- **Complex expressions in ORDER BY** - Use any SQL expression in ORDER BY clause
35+
- **Automatic SELECT injection** - Expressions automatically added to SELECT with hidden columns
36+
- **Aggregate support** - ORDER BY can use aggregates: `ORDER BY SUM(value) DESC`
37+
- **Works with transformers** - Integrates seamlessly with WHERE/GROUP BY alias expansion
38+
- **Example**: `SELECT region FROM sales GROUP BY region ORDER BY SUM(amount) DESC`
39+
40+
**Technical Architecture**:
41+
- **Three-phase execution** (Phases 0-2 complete):
42+
- Phase 0: Unified execution module foundation
43+
- Phase 1: Refactored `-q` mode to use unified path
44+
- Phase 2: Refactored `-f` mode to use unified path
45+
- Phase 3: Enabled full preprocessing pipeline in both modes
46+
- **Transformer orchestration** - Coordinated pipeline of AST transformers
47+
- **Preserved semantics** - All transformations maintain original query intent
48+
49+
#### **QUALIFY Clause Support**
50+
Industry-standard window function filtering using QUALIFY clause (Snowflake, BigQuery, Teradata syntax).
51+
52+
**New Capability**:
53+
```sql
54+
-- Top 3 products per category by sales
55+
SELECT category, product, sales,
56+
ROW_NUMBER() OVER (PARTITION BY category ORDER BY sales DESC) as rank
57+
FROM products
58+
QUALIFY rank <= 3;
59+
```
60+
61+
**Benefits**:
62+
- **Cleaner syntax** - No need for CTE wrapper around window functions
63+
- **Better readability** - Filter intent clear and concise
64+
- **Standard SQL** - Matches Snowflake/BigQuery syntax
65+
- **Performance** - Efficient filtering after window function evaluation
66+
67+
**Full Window Function Support**:
68+
- Works with ROW_NUMBER(), RANK(), DENSE_RANK()
69+
- Supports LAG(), LEAD(), FIRST_VALUE(), LAST_VALUE()
70+
- Compatible with all window functions (SUM, AVG, COUNT, etc.)
71+
- Handles complex PARTITION BY and ORDER BY clauses
72+
73+
### 🔧 Improvements
74+
75+
**Examples Testing Framework**:
76+
- **Python-based test runner** - Replaced bash scripts with robust Python framework
77+
- **Formal testing** - JSON expectations for critical examples
78+
- **Smoke testing** - 117+ examples validated for basic execution
79+
- **Data file hint support** - Examples automatically find their data files
80+
- **Clear output** - JSON validation failures clearly reported
81+
82+
**Neovim Plugin**:
83+
- **Transformation debug keymaps** - `\st` and `\sz` for pipeline visualization
84+
- **Better keymap organization** - Moved transformations to `\sz` to free up `\st`
85+
86+
**Documentation**:
87+
- **UNION ALL examples** - Added comprehensive subquery examples
88+
- **ORDER BY examples** - New example file for expression patterns
89+
- **Roadmap updates** - Documented ORDER BY completion status
90+
91+
### 🐛 Bug Fixes
92+
93+
**Temp Table Persistence**:
94+
- **Fixed `--execute-statement` mode** - Temp tables now properly persist across statement execution
95+
- **Materialization** - Temp tables correctly materialized and registered
96+
- **Dependency chain** - Multi-statement scripts with temp table dependencies work correctly
97+
98+
**Transformer Pipeline**:
99+
- **Dependency-aware execution** - Transformers now enabled in `--execute-statement` mode
100+
- **Qualified name resolution** - Fixed regression in table.column resolution after transformer changes
101+
- **Correlated subquery detection** - Phase 1 analyzer for future optimization
102+
103+
**Testing**:
104+
- **History file tests** - Ignored tests requiring persistent history file in CI
105+
- **Test output** - Clear JSON validation messages in examples framework
106+
- **CI pipeline** - Examples test suite integrated into continuous integration
107+
108+
### 📚 Documentation & Examples
109+
110+
**New Examples**:
111+
- `examples/order_by_expressions.sql` - ORDER BY expression patterns
112+
- `examples/union_all_subquery.sql` - UNION ALL with subqueries
113+
- Time series generation examples - How to create temporal test data
114+
115+
**Updated Documentation**:
116+
- `CLAUDE.md` - Examples test framework commands and usage
117+
- Roadmap - ORDER BY expression support completion notes
118+
- Test framework - Formal vs smoke test distinctions
119+
120+
### 🎯 Use Cases Enabled
121+
122+
**Complex Analytical Queries**:
123+
```sql
124+
-- Top regions by total sales (ORDER BY with aggregate)
125+
SELECT region, SUM(amount) AS total
126+
FROM sales
127+
GROUP BY region
128+
ORDER BY SUM(amount) DESC
129+
LIMIT 5;
130+
131+
-- Filtered window functions with QUALIFY
132+
SELECT salesperson, month, sales,
133+
ROW_NUMBER() OVER (PARTITION BY salesperson ORDER BY sales DESC) as rank
134+
FROM monthly_sales
135+
QUALIFY rank <= 3;
136+
137+
-- Multi-stage transformation pipeline (visible with --show-transformations)
138+
SELECT region, value * 2 AS doubled
139+
FROM data
140+
WHERE doubled > 100
141+
GROUP BY region
142+
HAVING SUM(doubled) > 1000
143+
ORDER BY SUM(doubled) DESC;
144+
```
145+
146+
**Unified Workflow**:
147+
- Write query in Neovim with `\st` to see transformations
148+
- Test with `-q` mode: `sql-cli -q "SELECT ..." --show-transformations`
149+
- Save to script file and run with `-f` mode (same execution path!)
150+
- Use `--execute-statement N` with full transformer support
151+
152+
### 🔧 Technical Details
153+
154+
**Files Modified**:
155+
- `src/main.rs` - Unified execution path integration
156+
- `src/execution/mod.rs` - New unified execution module
157+
- `src/query_plan/` - Transformer orchestration
158+
- `tests/integration/test_examples.py` - Python examples framework
159+
- `nvim-plugin/lua/sql-cli/` - Transformation debug keymaps
160+
161+
**Test Results**:
162+
- ✅ 457 library tests passing
163+
- ✅ 397 integration tests passing
164+
- ✅ 119 example tests (2 formal, 117 smoke)
165+
166+
---
167+
8168
## [1.63.0] - 2025-10-25
9169

10170
### ✨ Major Features

‎Cargo.toml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
[package]
22
name = "sql-cli"
3-
version = "1.63.0"
3+
version = "1.64.0"
44
edition = "2021"
55
autobins = false
66
autoexamples = false

0 commit comments

Comments
 (0)