fix(pagination): return adjacent page for reverse cursor pagination

[abhay-codes07]

Checklist

• Read CONTRIBUTING.md.
• PR title is short and non-generic (see previously merged PRs for examples).
• PR only implements a single feature or fixes a single bug.
• Tests passing (uv run pytest tests)?
• Lint / pre-commits passing (pre-commit run --all-files)?

Description

Fixes reverse (reverse=true, i.e. "previous page") cursor pagination in three list endpoints. The reference implementations already in the codebase (TableEditorService.list_rows, WorkflowsManagementService.list_workflows) do it in four steps: flip the scan direction, fetch limit + 1, compute cursors from scan order, then reverse the items and swap cursors/flags. Each of the fixed endpoints was missing a different piece:

• CasesService.search_cases: flipped the cursor comparison but not the ORDER BY, so LIMIT kept the first rows of the whole backward prefix — "previous page" returned page 1 from any page 3 or deeper. It also never swapped has_more/has_previous and its reverse branch overwrote next_cursor while leaving prev_cursor as null. Now uses the canonical scan-flip + reverse + swap.
• TablesService.list_rows (the search-style method behind search_rows): same missing ORDER BY flip; the post-fetch reversal/swap was already there, so it reversed the wrong window. Now flips the scan direction to match.
• CaseTableRowsService.list_rows: had two problems. Its reverse branch appended ascending sort keys to a statement already built with .order_by(created_at.desc(), id.desc()) — SQLAlchemy keeps the first sort keys, so the intended flip never took effect and the same wrong window came back. It also computed cursors from the already-reversed items and then swapped them again, so "next" pointed backward and "prev" pointed forward. The reverse branch now replaces the ORDER BY (order_by(None) first), and cursors are computed from scan order before the reversal, like the siblings.

Forward pagination is untouched in all three (the reverse handling is a no-op when reverse is false).

Related Issues

Fixes #3022

Screenshots / Recordings

N/A (API behavior fix)

Steps to QA

uv run pytest tests/unit/test_cases_service.py tests/unit/test_tables_service.py tests/unit/test_case_table_rows_service.py

The new regression tests page three deep and then walk back — the existing reverse tests only went back from page 2, where "previous page" and "page 1" coincide, which is why the bug never surfaced:

• test_search_cases_reverse_pagination_returns_adjacent_page: back from page 3 returns page 2 (previously page 1), flags are oriented in display direction, and the swapped cursors round-trip forward to page 3 and back to page 1.
• test_list_rows_reverse_pagination_returns_adjacent_page (tables): same walk over 7 rows with limit=2.
• test_list_rows_reverse_pagination_cursors_point_forward_and_back (case table rows): after going back a page, next_cursor returns to page 3 and prev_cursor continues to page 1 (previously each pointed the other way).

All pre-existing pagination tests in the three files pass unchanged.

---

Summary by cubic

Fixes reverse cursor pagination so “previous page” returns the adjacent page with correct cursors and flags across CasesService.search_cases, TablesService.list_rows, and CaseTableRowsService.list_rows. Forward pagination is unchanged.

• Bug Fixes
  • Implemented the canonical reverse flow: flip scan direction, fetch limit + 1, compute cursors from scan order, then reverse items and swap next/prev and has_more/has_previous. Added regression tests that walk back from page 3 to page 2.
  • CasesService.search_cases and TablesService.list_rows: now flip ORDER BY for reverse scans; case search also fixes swapped/overwritten cursors and flags.
  • CaseTableRowsService.list_rows: replaces (not appends) the base ORDER BY when reversing; computes cursors before reversal and swaps correctly so next points forward and prev points back.

^{Written for commit 8613676. Summary will update on new commits.}

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[zeropath-ai]

✅ No security or compliance issues detected. Reviewed everything up to 8613676.

Security Overview

• 🔎 Scanned files: 6 changed file(s)
• 🔗 Scan Link: https://zeropath.com/app/repositories/00dffd6c-8834-4dc9-b6d8-b44cd1622986?scanId=a128bae7-0f1d-4ab8-8c82-426f7c80ba8a&codeScanTypes=PrScan&tab=issues

Detected Code Changes

Change Type	Relevant files
Enhancement	► tracecat/cases/rows/service.py     Modify list_rows to adjust reverse pagination handling and ordering to reflect display order correctly ► tracecat/cases/service.py     Adjust search_cases to swap scan direction for reverse pagination and restore display order
Enhancement	► tracecat/tables/service.py     Modify list_rows to adjust reverse pagination handling and ordering to reflect display order correctly

[topher-lo]

@cubic review

[cubic-dev-ai]

@cubic review

@topher-lo I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

No issues found across 6 files

Confidence score: 5/5

• Automated review surfaced no issues in the provided summaries.
• No files require special attention.

_{Re-trigger cubic}