🐛 fix: keep full review thread context in filtered timelines

[SigureMo]

修复 #41 可能忽略 review thread 的更新

[Copilot]

Pull request overview

This PR fixes a correctness gap in the timeline --after/--before filtering for PRs: review events can now be retained when their review threads contain updates that fall within the selected time window (so thread context isn’t lost).

Changes:

• Extend timeline filtering to consider “related timestamps” (e.g., review thread comment timestamps) in addition to the event’s own timestamp.
• Adjust PR review thread grouping/rendering so a window-matched thread shows full context, with per-comment window markers.
• Add CLI test coverage to validate that an older review is kept when its thread receives an in-window reply.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
tests/test_cli.py	Adds a regression test ensuring filtered PR timelines keep full review thread context when a thread update matches the window.
src/gh_llm/pager.py	Updates filtered collection logic to match events via related timestamps and threads the window through timeline fetches.
src/gh_llm/models.py	Adds related_timestamps to TimelineEvent to support window matching beyond the primary event timestamp.
src/gh_llm/github_api.py	Propagates timeline_window into parsing/rendering, collects review-thread comment timestamps, and renders window markers/full-context thread headers.

Comments suppressed due to low confidence (2)

src/gh_llm/pager.py:414

• For PR timelines, this condition disables the early-stop optimization when --after is set, so _collect_filtered_from_end() will walk all the way to the first timeline page. On large PRs this can turn an incremental read into an O(total_pages) fetch, which is a noticeable performance/latency regression.

Consider keeping an early termination strategy for PRs (e.g., precomputing the set of review IDs/threads with any comment timestamps in-window from reviewThreads, then fetching only those missing review nodes by ID, or otherwise bounding how far back the backward walk must go).

            if (
                meta.kind != "pr"
                and current_page.items
                and timeline_window.after is not None
                and current_page.items[0].timestamp <= timeline_window.after
            ):
                break

src/gh_llm/pager.py:505

• Similarly, for PR timelines this disables the forward early-stop when --before is set, which can force fetching every page from the start even when the requested window is small. That undermines the intended windowed/incremental behavior and can be expensive for long PR timelines.

If correctness requires pulling in older review nodes whose threads were updated in-window, consider a targeted strategy (derive “relevant review IDs” from reviewThreads and fetch only those older reviews) instead of scanning the entire timeline connection.

            if (
                meta.kind != "pr"
                and current_page.items
                and timeline_window.before is not None
                and current_page.items[-1].timestamp >= timeline_window.before
            ):
                break

---

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

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: b8d7068826

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Restrict related timestamp matching for --before windows

This change makes _event_matches_window include a review when any related_timestamps entry is inside the window, but reviews now carry full thread history from github_api._get_review_threads_by_review; as a result, a review submitted after a --before cutoff can still be included because the same thread has an older comment timestamp. That breaks time-filter correctness by showing out-of-range events whenever a thread spans multiple reviews.

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

Preserve minimized-comment filtering in full-context threads

The new and not force_full_context condition bypasses minimized-comment hiding whenever a thread matches the selected window, so filtered timelines (--after/--before) can display minimized comments even when show_minimized_details is false. This regresses the existing CLI contract where minimized comments stay collapsed unless the user explicitly opts in via expand-minimized behavior.

Useful? React with 👍 / 👎.