conflict support 7 - handle conflicts for special operations#8
Conversation
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: fc664b24fd
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".
| if entry.path ~= base_path | ||
| and (entry.stage == 2 or entry.stage == 3) | ||
| and by_path[entry.path] | ||
| and not by_path[entry.path][1] | ||
| then |
There was a problem hiding this comment.
Pair each stage-1 conflict with its own rename candidates
The grouping loop currently adds all stage-2/3-only paths for every base_path, so multiple structural conflicts in the same index get cross-wired into the same group. In that case classify_conflict_entries ends up using whichever stage-2/3 entry was seen last, and grouped_paths can hide or duplicate unrelated conflicts, which breaks filetree conflict entries and special-resolution targeting. This should only pair candidates that belong to the current base conflict, not every stage-2/3 orphan path in the repo.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
Resolved in polyphilz/3-pane-7
|
|
||
| callback(build_status_snapshot(output, head_oid, { | ||
| root = root, | ||
| enrich_conflicts = opts.schedule_callback ~= false, |
There was a problem hiding this comment.
Preserve conflict enrichment on async poll snapshots
This ties conflict enrichment to callback scheduling, so when schedule_callback=false the snapshot skips enrich_conflict_files. The poll fast-path (lua/glance/repo_sync.lua sets schedule_callback=false) will therefore publish raw conflict entries without conflict_class/conflict_paths; for rename/delete and rename/rename states that means get_conflict_info cannot reconstruct a logical conflict from a single filetree entry, and the new special conflict flow becomes unavailable in poll-driven refreshes.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
Resolved in polyphilz/3-pane-7
ffb3e09 to
ea69fc9
Compare
fc664b2 to
bcdaff0
Compare
No description provided.