Skip to content

CLAUDE.md

Latest commit

 

History

History
38 lines (30 loc) · 1.77 KB

CLAUDE.md

File metadata and controls

38 lines (30 loc) · 1.77 KB

logq — PartiQL Query Engine for Server Logs

Project Overview

logq is a Rust CLI tool implementing PartiQL to query server log files (ELB, ALB, S3, Squid, JSONL). The goal is full PartiQL spec compliance (with pragmatic exclusions) using a bottom-up phased approach.

Architecture

  • Parser (src/syntax/): nom-based parser producing AST nodes
  • Logical Planner (src/logical/): AST → logical plan tree
  • Physical Executor (src/execution/): Stream-based pull execution with RecordStream trait
  • Common Types (src/common/): Value enum, type definitions shared across layers

Implementation Plan

See docs/plans/2026-04-04-partiql-completion-design-final.md for the full design.

Phases:

  • Phase 0: Code cleanup (case-sensitivity, failure→anyhow, dedup, bug fixes)
  • Phase 1: Foundation (NULL/MISSING propagation, float arithmetic, type coercion, IS NULL/MISSING, ordering, multi-branch CASE WHEN, parse_logic refactor)
  • Phase 2: Expressions (LIKE, BETWEEN, IN, CAST, ||, COALESCE/NULLIF, string functions, date_part)
  • Phase 3: Clauses (SELECT VALUE, DISTINCT, path wildcards, JOINs, LATERAL, subqueries)
  • Phase 4: Set operations (UNION/INTERSECT/EXCEPT)

Rules

  • Run cargo test before every commit
  • Never commit code that breaks existing passing tests
  • Write tests BEFORE implementation (test-first)
  • Commit after every meaningful unit of work
  • Update CHANGELOG.md after completing each sub-phase item

Out of Scope

PIVOT, Ion literals, bag literals (<<>>), correlated subqueries, window functions, schema-based type checking.

Test Oracle

  • Spec examples from PartiQL Specification PDF as ground truth
  • Hand-crafted edge cases for log-domain behavior
  • All existing ~58 tests must never regress

Progress

Track in CHANGELOG.md