fix: file history panel traversal across entry boundaries when wrap_entries is false

[dlyongemallo]

No description provided.

[Copilot]

Pull request overview

Adds coverage and fixes for file-history panel navigation so traversal across commit-entry boundaries behaves correctly when wrap_entries is disabled.

Changes:

• Refactors FileHistoryPanel:_get_entry_by_file_offset() to use non-wrapping traversal logic when wrap_entries = false (no modulo wrapping).
• Updates FileHistoryPanel:set_file_by_offset() to no-op when the computed next target is nil.
• Introduces a new functional test suite covering wrap_entries behavior for both FilePanel and FileHistoryPanel.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
lua/diffview/scene/views/file_history/file_history_panel.lua	Adjusts cross-entry traversal logic to avoid wrapping when wrap_entries is false; adds an early-return guard when no next target exists.
lua/diffview/tests/functional/wrap_entries_spec.lua	Adds functional tests validating wrapping vs non-wrapping traversal across entries and within a flat file list.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

When wrap is false and offset overshoots past the start/end, this now returns nil (via falling out of the loop). That makes count-based navigation (vim.v.count1 > remaining items) a no-op even when there is still valid movement possible (e.g. from a middle file with a large count), which is inconsistent with other navigation codepaths that clamp to the nearest boundary (see FilePanel:next_file/prev_file). Consider clamping here too: after exhausting entries, return the first/last file overall unless the cursor is already at that boundary, in which case return nil to signal no movement.

Suggested change

	-- Reached the boundary: return nil to signal no movement.
	-- Reached the boundary without wrapping: clamp to first/last file,
	-- unless already at that boundary (then signal no movement).
	local boundary_entry_idx = sign > 0 and #self.entries or 1
	local boundary_entry = self.entries[boundary_entry_idx]
	local boundary_files = boundary_entry.files
	local boundary_file = sign > 0 and boundary_files[#boundary_files] or boundary_files[1]
	local at_forward_boundary = sign > 0
	and entry_idx == #self.entries
	and file_idx == #cur_entry.files
	local at_backward_boundary = sign < 0
	and entry_idx == 1
	and file_idx == 1
	if at_forward_boundary or at_backward_boundary then
	return
	end
	return boundary_entry, boundary_file

[Copilot]

This comment says the +6 offset should land on entry3 file1, but the assertion below expects entries[3].files[2] ("g"), which is correct for a flat traversal starting at entry1/file1. Please update the comment to match the actual expected target so the test remains self-explanatory.

[Copilot]

The test name says it "clamps" a large forward offset at the boundary, but it asserts nil (no movement). If the intended behavior is to clamp to the last available file, the expectation should assert the last file; if the intended behavior is no-op on overshoot, the test name should be updated to reflect that.