Update Worktrees when the file system changes - #15
Merged
Merged
Conversation
nathansobo
changed the title
nathansobo
force-pushed
the
rescan
branch
2 times, most recently
from
April 14, 2021 01:41
8832fe2 to
be8237d
Compare
The easy-parallel crate spawned new threads on each call, which was resulting in way too many threads. Co-Authored-By: Brooks Swinnerton <934497+bswinnerton@users.noreply.github.com>
This will allow us to re-parent elements when re-scanning when the file system changes.
Co-Authored-By: Max Brunsfeld <maxbrunsfeld@gmail.com>
Co-Authored-By: Brooks Swinnerton <934497+bswinnerton@users.noreply.github.com>
Co-Authored-By: Brooks Swinnerton <934497+bswinnerton@users.noreply.github.com>
Co-Authored-By: Antonio Scandurra <me@as-cii.com>
Co-Authored-By: Nathan Sobo <nathan@zed.dev> Co-Authored-By: Max Brunsfeld <max@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Co-Authored-By: Antonio Scandurra <me@as-cii.com>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Take a callback instead of an mpsc Sender. The run method blocks and invokes the callback for each batch of events. The caller controls the threading. The callback can return false to terminate the event stream. Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
We're going to need something that can be moved to a background thread. Worktree used to be easy to clone, but that's no longer really true. Instead we can take a snapshot.
Co-Authored-By: Max Brunsfeld <max@zed.dev> Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
1 task
Anthony-Eid
pushed a commit
to Anthony-Eid/zed
that referenced
this pull request
Aug 7, 2024
1 task
rtfeldman
added a commit
that referenced
this pull request
Feb 5, 2026
- Fix #6: Shell parser now includes redirect targets in normalized command string. 'echo hello > /etc/passwd' now normalizes to include '/etc/passwd' so deny patterns can match writes via redirects. fd-only redirects (2>&1) handled gracefully. - Fix #15: Test harness PermTest::run now panics on invalid regex patterns instead of silently dropping them via filter_map. Uses unwrap_or_else with descriptive panic message for all three pattern lists.
m-bers
pushed a commit
to m-bers/zed
that referenced
this pull request
Apr 19, 2026
…elix-your-grep-tool Truncate long lines in grep tool output to prevent context window blowups
Sipaha
pushed a commit
to Sipaha/spk-editor
that referenced
this pull request
May 5, 2026
Adds a `[target.x86_64-unknown-linux-gnu]` block to `.cargo/config.toml` forcing `-fuse-ld=mold` for ~5-10x faster links and lower peak RAM vs system `ld`. Documents the requirement in README.md and records the rationale as FORK.md decision zed-industries#15.
Sipaha
added a commit
to Sipaha/spk-editor
that referenced
this pull request
May 6, 2026
Adds a `[target.x86_64-unknown-linux-gnu]` block to `.cargo/config.toml` forcing `-fuse-ld=mold` for ~5-10x faster links and lower peak RAM vs system `ld`. Documents the requirement in README.md and records the rationale as FORK.md decision zed-industries#15.
Sipaha
added a commit
to Sipaha/spk-editor
that referenced
this pull request
May 19, 2026
…ask zed-industries#18) Replaces the inline-base64 image flow with proper chunked upload. Wire is binary WebSocket frames (no JSON, no base64). 16-byte header (u64 upload_id + u64 offset, BE) prefixes each chunk's raw bytes. Server buffers chunks in tmp files keyed by server-generated u64 id; 1h TTL GC. send_message_blocks resolves spk-upload://<id> ResourceLinks into acp::ContentBlock::Image just before forwarding to ACP, so the agent sees the same Image block as today. Mobile starts upload the moment a file is picked; UI shows per-attachment progress; Send is gated until all uploads complete. Reconnect resume via upload_status; force-kill resume via per-server EncryptedSharedPreferences. WS cap dials back from 32 MiB (zed-industries#15 hotfix) to 1 MiB once chunked lands.
ShalashaskaOcelot
pushed a commit
to ShalashaskaOcelot/zed
that referenced
this pull request
Jul 10, 2026
…do shortcuts Addresses several notebook keyboard-shortcut issues: - a/b (and the + toolbar buttons) now insert a cell and stay in COMMAND mode (select it, focus the notebook) instead of jumping into edit mode, so follow-up single-key shortcuts keep working (bug zed-industries#13). - Sync mode to focus: when the notebook root gains focus, force command mode (cx.on_focus), reducing the focus/mode desync that left shortcuts stuck (bug zed-industries#15). - Bind Escape to EnterCommandMode in the base NotebookEditor context (not just edit mode) as a recovery path. - Add ctrl-z / ctrl-shift-z for cell undo/redo in addition to z / shift-z. Tracking: archive phase 10 (stop-on-error + interrupt/restart queue-clear confirmed) and bug 10 (kernel picker Enter confirmed); rewrite phases 7/8/9/10 files so implemented tasks are checked (were left unchecked); update bug 3 (interrupt delivered + cancels batch; time.sleep delay is a known Windows limitation), file bug 15 (focus/mode desync), backlog immediate-interrupt via GenerateConsoleCtrlEvent and focus-return-after-chrome. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014eseBDGa7au9r2VxBHeh9X
ShalashaskaOcelot
pushed a commit
to ShalashaskaOcelot/zed
that referenced
this pull request
Jul 10, 2026
…de on run Rebind notebook cell copy/cut/paste from the invented single-key c/x/v to the OS-standard clipboard combos (ctrl/cmd-c, ctrl/cmd-x, ctrl/cmd-v) in the command-mode context. Edit mode keeps text copy/cut/paste. Free the base ctrl-c/cmd-c binding (was InterruptKernel, shadowed everywhere) and move interrupt to the Jupyter-standard `i i` in command mode; the toolbar Stop button is unchanged. Make Run (ctrl/cmd-enter) always return to command mode after executing so the cursor leaves the editor and command-mode shortcuts keep working. Archive phase 8 (undo/redo confirmed working). Reopen phase 7 for the rebind and add phase 11 for the run-to-command-mode change; update bug zed-industries#15. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014eseBDGa7au9r2VxBHeh9X
ShalashaskaOcelot
pushed a commit
to ShalashaskaOcelot/zed
that referenced
this pull request
Jul 15, 2026
… 32) - Esc while already in command mode collapses a multi-cell selection back to the primary cell, VS Code style; Esc from edit mode still only switches modes. - The output "…" menu's Copy/Collapse/Clear entries now re-select their cell via PlainClick when done, returning keyboard focus to the notebook so command-mode shortcuts work immediately after menu interactions (bug zed-industries#15 follow-up). Cell-toolbar buttons already refocused via handle_cell_toolbar_action. - Markdown preview render-on-exit-edit was already wired (editor Blurred → FinishedEditing → reparse) — verified, listed for runtime confirmation. Phase 32 complete; phase 35 (prompt Windows interrupt of C-blocking calls via CTRL_BREAK process groups) promoted to keep 5 phases in rotation (now 29, 31, 33, 34, 35). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014eseBDGa7au9r2VxBHeh9X
ShalashaskaOcelot
pushed a commit
to ShalashaskaOcelot/zed
that referenced
this pull request
Aug 6, 2026
Audited every CHANGELOG claim against the code: all phase and bug entries are genuinely implemented (symbol checks for each), and every commit hash inside this shallow clone's history resolves to the work it claims. Nothing needed pulling out for review. Fixes found along the way: - Bugs zed-industries#47 and zed-industries#54 were archived as user-confirmed by c199237 but that commit never deleted their bugs.md entries, so both were double-tracked. Removed. - Bug zed-industries#63 was still "open — needs a decision" although phase 59 implemented the decided behaviour. Moved to fix attempted - untested with the fix recorded. - Bugs zed-industries#7, zed-industries#15 and zed-industries#20 sat at fix attempted - untested with no pointer in the testing ledger, so they were invisible there. Added with test recipes. - Bug zed-industries#56's changelog hash was a transposition (c199273 → c199237). Phases 59 and 60 are implementation-complete with only manual tests left, so their test items move to awaiting_testing.md and the phase files are recorded in the changelog and deleted. Six phases remain in rotation. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HUB73REGigaN5TCgvSkM2s
jolutz
pushed a commit
to jolutz/zed
that referenced
this pull request
Aug 8, 2026
Update Worktrees when the file system changes
audivir
added a commit
to audivir/zed
that referenced
this pull request
Aug 10, 2026
…font scans Closes out High zed-industries#9 and Medium zed-industries#15 from BUG_REPORT.md: - resolve_project (walks parent directories checking typst.toml/main.typ - real filesystem I/O) ran synchronously in schedule_typst_update's foreground view.update closure on every debounced edit. It's now a plain (project_resolution: bool, active_path: &Path) function with no view dependency, called from the existing cx.background_executor() block instead; needs_new_world/world_arc selection moved there too since it depends on the resolved root. Covered by four new tests in typst_preview_view.rs exercising resolve_project directly. - TypstSystemWorld::new re-scanned every system font and re-parsed every bundled asset font on every project-root switch, even though neither depends on the root. Bundled + system fonts are now scanned once per process into a cached static_fonts() OnceLock<(FontBook, Vec<FontSlot>)> (FontSlot is now Clone); new() clones that and layers a fresh, root-scoped load_fonts_dir(&root) scan on top. Covered by two new tests in typst_world.rs. Also investigated the "path-normalization mismatch on symlinked roots" Low finding: traced worktree::LocalFile::abs_path to confirm Zed deliberately never canonicalizes worktree roots anywhere, so the mismatch this described is only reachable across multiple worktrees with divergent symlink paths to the same file - narrower than originally flagged, and fixing it locally would cut against that codebase-wide convention. Left open, documented in BUG_REPORT.md with the concrete reproduction condition instead of a fix. BUG_REPORT.md and TEST_PLAN.md updated to reflect what's fixed/tested now. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
audivir
added a commit
to audivir/zed
that referenced
this pull request
Aug 10, 2026
…font scans Closes out High zed-industries#9 and Medium zed-industries#15 from BUG_REPORT.md: - resolve_project (walks parent directories checking typst.toml/main.typ - real filesystem I/O) ran synchronously in schedule_typst_update's foreground view.update closure on every debounced edit. It's now a plain (project_resolution: bool, active_path: &Path) function with no view dependency, called from the existing cx.background_executor() block instead; needs_new_world/world_arc selection moved there too since it depends on the resolved root. Covered by four new tests in typst_preview_view.rs exercising resolve_project directly. - TypstSystemWorld::new re-scanned every system font and re-parsed every bundled asset font on every project-root switch, even though neither depends on the root. Bundled + system fonts are now scanned once per process into a cached static_fonts() OnceLock<(FontBook, Vec<FontSlot>)> (FontSlot is now Clone); new() clones that and layers a fresh, root-scoped load_fonts_dir(&root) scan on top. Covered by two new tests in typst_world.rs. Also investigated the "path-normalization mismatch on symlinked roots" Low finding: traced worktree::LocalFile::abs_path to confirm Zed deliberately never canonicalizes worktree roots anywhere, so the mismatch this described is only reachable across multiple worktrees with divergent symlink paths to the same file - narrower than originally flagged, and fixing it locally would cut against that codebase-wide convention. Left open, documented in BUG_REPORT.md with the concrete reproduction condition instead of a fix. BUG_REPORT.md and TEST_PLAN.md updated to reflect what's fixed/tested now.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
.gitignorefiles.gitignorefiles.gitdirectory are ignored