Current line: 0.4.x series (SQL Spec v1.0 — feature freeze) SQL support reference:
docs/SQL_SPEC.md
- PEP 249-compliant
ExcelConnectionandExcelCursor - Parameter binding with qmark placeholders (
?) - Openpyxl engine (default) and optional pandas engine
- Autocommit and snapshot-based rollback semantics
- Formula injection defense for write operations
SELECTwith projection, expressions, aliases,WHERE,ORDER BY,LIMIT,OFFSETDISTINCT(single-table queries)- JOIN support:
INNER,LEFT,RIGHT,FULL OUTER,CROSS(including chained JOINs) - Aggregation:
COUNT,SUM,AVG,MIN,MAX,COUNT(DISTINCT ...) GROUP BYandHAVING- Subqueries in
WHERE ... [NOT] IN (SELECT ...)(single-column, non-correlated) - Compound queries:
UNION,UNION ALL,INTERSECT,EXCEPT - Expressions: arithmetic (
+ - * /) andCASE ... WHEN ... THEN ... ELSE ... END - DML:
INSERT(single, multi-row,INSERT ... SELECT),UPDATE,DELETE - UPSERT:
INSERT ... ON CONFLICT ... DO NOTHING / DO UPDATE - DDL:
CREATE TABLE,DROP TABLE,ALTER TABLE ADD/DROP/RENAME COLUMN
- Strict mypy configuration and CI validation
- Broad parser/executor regression coverage
- Trusted Publisher (OIDC) release flow for PyPI
- SQL syntax freeze: no new SQL syntax additions until v0.5.0
- Internal architecture improvements (exception hierarchy, parser/validator separation)
- Test consolidation and coverage hardening
- Documentation reconciliation with implemented behavior
- Async-friendly API surface (design investigation)
- Additional backend options (e.g., Polars-based engine)
- Performance tuning for large-sheet scans and join-heavy queries
- Documentation and example expansion around advanced SQL patterns
- Full ACID transaction guarantees on
.xlsx - Multi-writer concurrency on the same workbook path
- Stored procedures, triggers, and foreign key enforcement
- Full SQL dialect parity with SQLite/PostgreSQL
- Authoritative SQL behavior:
docs/SQL_SPEC.md(SQL Spec v1.0) - Authoritative feature matrix:
docs/SQL_SPEC.md#2-authoritative-feature-matrix