fix: resolve the object type behind modifiers in statement_type

[sileht]

Statement.statement_type took the first two keywords, so any keyword
sitting between the verb and the object displaced the object out of the
window. sqlparse 0.6.0 made that visible by lexing MATERIALIZED as a
keyword, but it predates the bump: TEMPORARY, TEMP and UNIQUE were
already doing it.

CREATE TEMPORARY TABLE "CREATE TEMPORARY" -> "CREATE TABLE"
CREATE TEMP TABLE "CREATE TEMP" -> "CREATE TABLE"
CREATE UNIQUE INDEX "CREATE UNIQUE" -> "CREATE INDEX"
CREATE MATERIALIZED VIEW "CREATE MATERIALIZED"-> "CREATE MATERIALIZED VIEW"
DROP MATERIALIZED VIEW "DROP MATERIALIZED" -> "DROP MATERIALIZED VIEW"
ALTER MATERIALIZED VIEW "ALTER MATERIALIZED" -> "ALTER MATERIALIZED VIEW"
REFRESH MATERIALIZED VIEW "MATERIALIZED" -> "REFRESH MATERIALIZED VIEW"
CREATE OR REPLACE VIEW "CREATE OR REPLACE" -> "CREATE OR REPLACE VIEW"

Three kinds of word are now told apart: modifiers that do not change what
the object is and are dropped (TEMPORARY, TEMP, GLOBAL, LOCAL, UNIQUE,
RECURSIVE); qualifiers that are part of the object's name and let it
continue (MATERIALIZED); and verbs sqlparse does not lex as keywords at
all, recovered from the statement's first token (REFRESH, which came
through with ttype None and left the type headless).

OBJECT_VERBS deliberately holds only verbs that are always followed by
the keyword naming their object. TRUNCATE acts on a name, so pairing it
with the next keyword would read a trailing option as the object type and
invent "TRUNCATE CASCADE", "TRUNCATE RESTART" and friends; it stays out,
and tests pin that.

This fixes one comparison, not just the labels. "CREATE TEMPORARY" was
not in UnorderedTokenList.STATEMENT_TYPES, so temporary tables never
had their column lists sorted and two identical ones written in a
different column order compared as drifted. They now behave like plain
CREATE TABLE. Temporary tables that genuinely differ still compare
unequal, statements with no keywords still resolve to UNKNOWN so
_parse_statements filters exactly what it filtered before, and a
statement behind a leading comment is still typed from its keywords.

Claude-Session: https://claude.ai/code/session_01XkWQRxh5pfi2Fbuv3Z9wen
Co-Authored-By: Claude Opus 5 (1M context) noreply@anthropic.com

[sileht]

This pull request is part of a Mergify stack:

#	Pull Request	Link
1	ci(mergify): don't check Python 3.9 tests	#299
2	chore: drop Python 3.9 support	#300
3	chore(deps): bump sqlparse to 0.6.0	#301	👈

[mergify]

Merge Protections

🟢 All 6 merge protections satisfied — ready to merge.

Show 6 satisfied protections

🟢 🤖 Continuous Integration

• all of:
  • check-success=Test with Python 3.10
  • check-success=Test with Python 3.11
  • check-success=Test with Python 3.12
  • check-success=semgrep

🟢 👀 Review Requirements

• any of:
  • #approved-reviews-by>=1
  • author = dependabot[bot]

🟢 Enforce conventional commit

Make sure that we follow https://www.conventionalcommits.org/en/v1.0.0/

• title ~= ^(fix|feat|internal|docs|style|refactor|perf|test|build|ci|chore|revert|ui)(?:\(.+\))?!?:

🟢 🔎 Reviews

• #changes-requested-reviews-by = 0
• #review-requested = 0
• #review-threads-unresolved = 0

🟢 📕 PR description

• body ~= (?ms:.{48,})

🟢 🚦 Auto-queue

When all merge protections are satisfied, this pull request will be queued automatically.

[mergify]

@sileht this pull request is now in conflict 😩

[sileht]

Revision history

#	Type	Changes	Reason	Date
1	initial	db8670e		2026-08-28 14:27 UTC
2	rebase	db8670e → 5a9c785 (rebase only)		2026-08-28 14:27 UTC
3	content	5a9c785 → bc14ac6		2026-08-31 07:26 UTC
4	content	bc14ac6 → 18e652b		2026-08-31 07:38 UTC
5	content	18e652b → 0f64cce		2026-08-31 08:30 UTC
6	content	0f64cce → 83fb20f (raw)		2026-08-31 08:52 UTC
7	content	83fb20f → 35642ab (raw)		2026-08-31 15:41 UTC

Pull request has been modified.

[Copilot]

🟢 Approval recommended

The dependency bump is well-scoped and the tests are updated to explicitly pin the relevant externally visible behavior changes.

Pull request overview

This PR upgrades the project’s sqlparse dependency to 0.6.0 to pick up security/DoS fixes, and updates tests to pin the observable Statement.statement_type behavior changes introduced by the new lexer keywording.

Changes:

• Bump sqlparse minimum version to >=0.6.0 and update the lockfile to 0.6.0.
• Extend test_statement_type coverage to pin MATERIALIZED/TEMPORARY statement-type outcomes under the updated lexer behavior.

File summaries

File	Description
tests/test_sql_compare.py	Adds/updates statement type expectations for MATERIALIZED VIEW and TEMPORARY TABLE cases under sqlparse 0.6.0.
pyproject.toml	Raises the minimum supported sqlparse version to >=0.6.0.
poetry.lock	Locks sqlparse to 0.6.0 and updates associated metadata/hashes.

Review details

• Files reviewed: 2/3 changed files
• Comments generated: 1
• Review effort level: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[mergify]

@sileht this pull request is now in conflict 😩

[mergify]

Merge Queue Status

• ✅ Entered queue — 2026-09-01 09:07 UTC · Rule: default · triggered by merge protections
• ✅ Checks skipped · PR is already up-to-date
• ✅ Merged — 2026-09-01 09:08 UTC · at 35642ab80ab45c616329dce14d88b2fa29069c62 · squash

This pull request spent 47 seconds in the queue, including 9 seconds running CI.

Required conditions to merge

• github-review-approved [🛡 GitHub branch protection]
  • fix: resolve the object type behind modifiers in statement_type #301
• github-review-approved [🛡 GitHub repository ruleset rule Require pull request for default branch]
  • fix: resolve the object type behind modifiers in statement_type #301
• all of [🛡 Merge Protections rule Enforce conventional commit]:
  • title ~= ^(fix|feat|internal|docs|style|refactor|perf|test|build|ci|chore|revert|ui)(?:\(.+\))?!?:
    • fix: resolve the object type behind modifiers in statement_type #301
• all of [🛡 Merge Protections rule 👀 Review Requirements]:
  • any of:
    • #approved-reviews-by>=1
      • fix: resolve the object type behind modifiers in statement_type #301
    • author = dependabot[bot]
      • fix: resolve the object type behind modifiers in statement_type #301
• all of [🛡 Merge Protections rule 📕 PR description]:
  • body ~= (?ms:.{48,})
    • fix: resolve the object type behind modifiers in statement_type #301
• all of [🛡 Merge Protections rule 🔎 Reviews]:
  • #changes-requested-reviews-by = 0
    • fix: resolve the object type behind modifiers in statement_type #301
  • #review-requested = 0
    • fix: resolve the object type behind modifiers in statement_type #301
  • #review-threads-unresolved = 0
    • fix: resolve the object type behind modifiers in statement_type #301
• all of [🛡 Merge Protections rule 🤖 Continuous Integration]:
  • all of:
    • check-success=Test with Python 3.10
    • check-success=Test with Python 3.11
    • check-success=Test with Python 3.12
    • check-success=semgrep