State holds the open project (VFS project, watcher, watched directories)
and versioned buffers. New commands: project.open/close/children,
file.open/save/saveAll, buffer.get/edit/close. buffer.edit takes UTF-16
change sets based on a version and fails with the new stale_version code;
closing dirty buffers or projects without force fails with
unsaved_changes. Commands queue events (projectOpened, treeChanged,
bufferChanged, dirtyChanged, fileChangedOnDisk, ...) that the engine
returns in Outcome.events; Engine::pump_vfs maps changes on disk to
events and reloads clean buffers. project.children watches listed
directories, capped at 256 native / 32 polling, least recently listed
dropped first.
Control channel: events.subscribe streams {"event": ...} lines, with
bounded per-subscriber queues (lutra_control::Subscribers) and method docs
in methods.rs. testkit's EngineBackend and the desktop Host implement the
new Backend::subscribe. Verified with 14 engine integration tests, a TCP
events test and the running app over the control channel.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6.1 KiB
Target-app parity
Status: M1 in progress (text buffer, VFS, tree UI, engine workspace landed), no PHP feature ships · Last updated: 2026-10-14 · Target: PhpStorm 2026.x
How this is measured
Each feature area gets a weight (how much a typical PHP developer uses it) and a score (how much of the documented behaviour Lutra has, judged by behaviour, not by menu items). The headline depth is the weighted sum. Every score is marked measured or estimated and must point to evidence (tests, commands, screenshots of Lutra). Behaviour references come from public PhpStorm documentation only, noted with URLs in plan/phpstorm/.
Hours are the estimated agent wall-clock hours for one agent to reach behaviour parity in the area, from zero. They are first estimates made at the end of M0 and are recalibrated after each milestone from merged-commit timestamps.
Feature areas
| Area | Weight | Score | Missing | Hours (one agent) | Evidence |
|---|---|---|---|---|---|
| Editor core | 12 | 0% | everything user-visible: no editor in the UI yet; engine buffers, file commands and events exist | ~120 (est.) | crates/text: rope buffer, positions, edits (M1.1, unit tests + proptests). M1.3: engine buffers with versions and dirty state, file.open, buffer.get, buffer.edit (UTF-16 change sets, stale_version), file.save, file.saveAll, buffer.close with unsaved-changes guard, line ending and BOM kept on save, clean buffers reload on external change; 14 integration tests on temp dirs (crates/engine/tests/workspace.rs) |
| Project view, VFS | 6 | 0% | everything reachable in the app: the tree UI is not wired to the engine yet (M1.4) | ~37 (est.) | lutra-vfs (M1.2): lazy sorted listing, ignore rules (.gitignore, .ignore, info/exclude, global excludes, Lutra defaults), native file watching via notify with a polling fallback, capped UTF-8 reads, atomic writes; 34 tests on temp dirs, watcher tests run on both backends (crates/vfs/tests). M1.5: project tree UI (lazy listing, keyboard navigation, dimmed ignored entries, event refresh) in ui/src/components/project/, Vitest against a mocked engine. M1.3: project.open, project.close, project.children (watches listed directories, capped at 256 native / 32 polling, least recently listed dropped), treeChanged events from disk, events.subscribe on the control channel with a TCP test (crates/control/tests/events.rs) |
| Navigation | 10 | 0% | everything | ~120 (est.) | — |
| Search and replace | 6 | 0% | everything | ~80 (est.) | — |
| Code completion | 10 | 0% | everything | ~250 (est.) | — |
| Inspections and quick-fixes | 8 | 0% | everything | ~300 (est.) | — |
| Refactoring | 7 | 0% | everything | ~250 (est.) | — |
| Code style and formatting | 4 | 0% | everything | ~120 (est.) | — |
| Git | 8 | 0% | everything | ~150 (est.) | — |
| Terminal | 3 | 0% | everything | ~30 (est.) | — |
| Run configurations | 2 | 0% | everything | ~50 (est.) | — |
| Debugger | 5 | 0% | everything | ~120 (est.) | — |
| Testing | 4 | 0% | everything | ~80 (est.) | — |
| Composer and interpreters | 2 | 0% | everything | ~50 (est.) | — |
| Frameworks | 5 | 0% | everything | ~300 (est.) | — |
| Database tools | 3 | 0% | everything | ~250 (est.) | — |
| HTTP client | 1 | 0% | everything | ~60 (est.) | — |
| Web languages | 2 | 0% | everything | ~200 (est.) | — |
| Docker and remote | 1 | 0% | everything | ~120 (est.) | — |
| Settings, keymaps, themes | 1 | 0% | settings UI, keymaps, theme switching (theme tokens exist in lutra_engine::theme) |
~60 (est.) | — |
| Plug-ins | 0.5 | 0% | everything | ~120 (est.) | — |
| AI features | 0.5 | 0% | everything | ~60 (est.) | — |
Sum: about 2,930 h. Adding UI fidelity (~150 h), localization (~40 h), performance (~100 h) and stability hardening (~100 h) gives the ROADMAP's ~3,300 h to full parity. The range is wide (2,000–5,000 h): completion, inspections, refactoring and frameworks depend on an own type-inference engine whose size is hard to judge before M6.
Weights were reviewed at the end of M0 and kept: nothing built so far gives a reason to change them.
What ships (M0, 2026-10-14)
No PHP feature. The foundation, verified on Windows 11:
cargo xtask cipasses all 13 steps (fmt, clippy with no-panic lints, tests, lints self-test, layers, assets, clean-room, cargo-deny licences, pnpm install, ESLint, vue-tsc, Vitest, frontend licences).- Commands
app.aboutandapp.quitwith unit tests; a panicking command is caught by the guard (crates/engine/src/engine.rstests). - Control channel:
app.ping,engine.execute,engine.listCommands,ui.screenshot, with proptests and a TCP integration test (crates/control), and checked against the running app. - The desktop window opens themed from Rust tokens and shows the product name and version that the frontend gets from
app.aboutover IPC.
What ships (M1, in progress)
Nothing user-visible yet, so every score stays 0%.
- M1.1 text buffer (
crates/text):Bufferon aropeyrope with byte, char, UTF-16 and line counts; byte offset to line and column in UTF-16 code units or chars and back, all fallible (out of range, inside a character, inside a CRLF, inside a surrogate pair); atomicEditbatches in LSPTextEdit[]semantics that reject reversed, out-of-range and overlapping ranges;//detection with a keep-the-file's-ending save policy. 23 unit tests and 6 proptests (round-trips on emoji, ZWJ sequences, CJK and combining marks; random edit batches checked against aStringmodel, never panicking).
Revision history
| Date | Change | Summary |
|---|---|---|
| 2026-10-14 | minor | Project view evidence: project tree UI (M1.5), score unchanged |
| 2026-10-14 | minor | M1.3 engine workspace evidence (Editor core, Project view; scores unchanged) |
| 2026-10-14 | minor | M1.1 text buffer evidence (Editor core, score unchanged) |
| 2026-10-14 | minor | Project view evidence: native file watching (M1.2b), score unchanged |
| 2026-10-14 | minor | Project view evidence: lutra-vfs (M1.2), score unchanged |
| 2026-10-14 | minor | Hours per area, weights reviewed, M0 evidence |
| 2026-10-13 | major | Baseline |