refactor: 👽️ Update diagnostic source from 'laravel-lsp' to 'laravel'. by mikebronner · Pull Request #5 · mike-bronner/zed-laravel

mikebronner · 2026-05-12T15:31:02Z

No description provided.

Resolve the four blockers from Holmes's review of the scope-aware Blade variable / controller→view binding rename: - Correctness: collect_variable_name_spans no longer descends into a non-capturing anonymous closure, so renaming a compact() binding stops clobbering an unrelated `function () { $name ... }` body variable. Arrow functions and `use ($name)` closures still descend — they share the outer variable. - Security: view_binding_rename_edit validates the source view name with validate_view_name and constrains the resolved view path under the project root before any disk read, closing a path traversal where a hostile `view('/tmp/evil', ...)` could read an out-of-project file and leak its absolute path into the WorkspaceEdit. - AC #5: prepare_blade_var_rename now gates on is_template_variable, rejecting variables that never surface in template markup (undefined names, @php-block / function-locals) instead of accepting any $var. - AC #6: extract a pure binding_rename_spans cross-file core and drive view_binding_rename_edit through it, with tests exercising the real controller<->view orchestration (array-key + compact) and the multi-controller non-contamination case. Also tightens string_literal_text (matching quotes) and string_content_span_at (cursor strictly before the closing quote), and makes for_loop_scopes_rename assert the exact span set, per the review's non-blocking notes. Tests: 1631 lib tests pass (+8 new); cargo fmt + clippy -D warnings clean.

@foreach

…ontroller) (#89) * chore: start work on #55 * feat: ✨ scope-aware Blade variable rename + controller→view binding rename Add Blade template variable rename to the rename engine (issue #55): - Scope-aware in-template rename: a variable introduced by @foreach / @forelse / @for is block-scoped; a file-level variable (controller-passed or @php-assigned) is file-scoped but skips any nested loop that re-binds the same name. Blade comments and @verbatim regions are masked out. - Controller→view binding rename: F2 on a view-data key in view('v', ['key' => …]) or compact('key') rewrites the controller key, the in-view file-scoped $key usages, and (for compact) the enclosing method's local $key so the controller stays valid. New pure module src/blade_var_rename.rs (fully unit-tested), wired into the prepare_rename / rename handlers ahead of the literal-symbol classifier. Docs and README updated to move the feature from Planned to shipped. Refs #55 * fix: 🐛 address PR #89 review on Blade variable rename (#55) Resolve the four blockers from Holmes's review of the scope-aware Blade variable / controller→view binding rename: - Correctness: collect_variable_name_spans no longer descends into a non-capturing anonymous closure, so renaming a compact() binding stops clobbering an unrelated `function () { $name ... }` body variable. Arrow functions and `use ($name)` closures still descend — they share the outer variable. - Security: view_binding_rename_edit validates the source view name with validate_view_name and constrains the resolved view path under the project root before any disk read, closing a path traversal where a hostile `view('/tmp/evil', ...)` could read an out-of-project file and leak its absolute path into the WorkspaceEdit. - AC #5: prepare_blade_var_rename now gates on is_template_variable, rejecting variables that never surface in template markup (undefined names, @php-block / function-locals) instead of accepting any $var. - AC #6: extract a pure binding_rename_spans cross-file core and drive view_binding_rename_edit through it, with tests exercising the real controller<->view orchestration (array-key + compact) and the multi-controller non-contamination case. Also tightens string_literal_text (matching quotes) and string_content_span_at (cursor strictly before the closing quote), and makes for_loop_scopes_rename assert the exact span set, per the review's non-blocking notes. Tests: 1631 lib tests pass (+8 new); cargo fmt + clippy -D warnings clean. * fix: 🐛 Scope compact() rename to its enclosing closure. `enclosing_function_local_spans` matched only the legacy `anonymous_function_creation_expression` node kind, which does not exist in tree-sitter-php 0.24 (the closure kind is `anonymous_function`). A `compact('name')` inside a route closure therefore never elected the closure as its scope, fell back to the whole-file root, and renamed unrelated `$name` occurrences in sibling closures. Match both names, per repo convention. Also addresses two review notes flagged alongside: anchor `mask_php_blocks` on a word boundary so `@phpunit` / `@phpdoc` no longer trips a spurious `@php` mask region, and correct the `view_binding_key_at` docstring to drop the unimplemented `->with` chained forms. Refs: #55 * fix: 🐛 balance parens in Blade loop scope detection (#55) @foreach/@forelse loop arguments were captured with a `$[^)]*$` regex that stops at the first `)`, so an iterable containing a call — `@foreach($users->where('active', true) as $user)` — truncated to `($users->where('active', true)`, dropping the ` as $user` tail. The loop then registered no variables, its scope went invisible, and the file-scope arm admitted every `$var` — the cross-scope clobber AC #1 forbids. Capture the loop argument list by balancing parens (ignoring parens inside string literals) and recover the binding by splitting on the last ` as ` instead of a `[^)]` run, so parenthesized iterables keep their variable. Also addresses PR #89 round-3 non-blocking notes: - canonicalize both sides of the view-path root-containment guard before the textual `starts_with` check, so a symlink can't leak an out-of-project path into the WorkspaceEdit. - document the namespaced-view (`package::view`) limitation on `validate_view_name`. Regression tests: method-call and nested-paren iterables in blade_loops, plus in_scope_spans loop-scoping for a parenthesized iterable. * fix: 🐛 Balance Blade loop directive parens across wrapped lines A @foreach/@forelse/@for/@while header whose argument list wraps across physical lines was never registered as a loop block — the per-line paren scan gave up at the first line end — so a rename inside it silently clobbered the whole file. Balance the directive parens across continuation lines (one code path; single-line fast path unchanged). Refs #55 * fix: 🐛 Harden Blade variable rename against fail-open clobbers The scope engine renamed file-wide whenever it couldn't resolve a loop's binding — a silent clobber on valid templates. Root-cause fixes: - Resolve loop blocks over the comment/@verbatim-masked source (as variable_spans already does), so a {{-- --}} comment in or around a loop header no longer desyncs the parse or forms a phantom binding block. - Recover every binding shape — array/list() destructuring, by-reference (&$item), key + destructured value — by extracting every $ident from the foreach target instead of matching one fixed pattern. - Refuse $loop (reserved, never header-bound) at the prepare gate. - Fail-closed backstop: when loop scope can't be resolved — an opaque @foreach/@forelse (binding unparseable) or a broken (unbalanced-paren) header — refuse the rename and never clobber into the region, rather than admitting file-wide spans. 15 regression tests cover each shape. Refs #55

…root targets Add a containment backstop at the file-create seam (FileAction::build_code_action, issue #199 AC #5/#6): refuse to offer any create quick-fix (View, BladeComponent, Livewire, Inertia, …) whose target_path escapes the project root, returning None instead of constructing an out-of-root ResourceOp::Create. Completes the "fail-closed containment on every FS-touching path" sweep (#130 → #143 → #148 → #194) across the third surface — the write seam, alongside the read/resolve paths. The guard uses path_within_root_lexical, NOT the fail-closed path_within_root the sibling read paths use: a create target never exists yet, so path.canonicalize() always fails for it and the fail-closed guard would refuse *every* create, including legitimate in-root ones. The lexical guard refuses out-of-root and interior-`..` escapes while admitting a not-yet-created in-root target, and still canonicalizes to catch symlink escapes when the target exists. (AC #5 named path_within_root; flagged on the PR — same class of AC defect as the #3/#4 dispute resolved via Option 1.) Add code_action_create_containment.rs: out-of-root, interior-`..`, in-root positive control, and under-root symlink-escape cases. Reframe the resolve-seam negative tests per the #199 escalation (Option 1): they assert the *invariant* (an out-of-root / symlink-escaping component never resolves), with the resolve_component_file guard documented as the innermost backstop whose reject branch is unreachable via the public API today — the upstream path_within_root_lexical filter catches both negatives first. Drop the false "only path_within_root catches" claim and correct the misleading assert messages. Fixes: #199 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

…for containment-invariant uniformity (#203) * chore: start work on #199 * fix: 🔒️ Add fail-closed path_within_root guard to resolve_component_file. Mirror the sibling resolve_component_existing_file: after the file_exists_cached check, refuse any candidate whose real path escapes config.root before returning it. resolve_component_path already drops out-of-root candidates with the lexical path_within_root_lexical filter, so this is defense-in-depth — it makes the fail-closed containment invariant hold uniformly across every FS-touching component resolver (the #130 → #143 → #148 → #194 chain). Add component_file_navigation_containment.rs pinning the invariant at the resolve_component_file boundary: out-of-root and under-root-symlink escapes return None, in-root files still resolve. Fixes: #199 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: 🔒️ Guard the build_code_action write/create seam against out-of-root targets Add a containment backstop at the file-create seam (FileAction::build_code_action, issue #199 AC #5/#6): refuse to offer any create quick-fix (View, BladeComponent, Livewire, Inertia, …) whose target_path escapes the project root, returning None instead of constructing an out-of-root ResourceOp::Create. Completes the "fail-closed containment on every FS-touching path" sweep (#130 → #143 → #148 → #194) across the third surface — the write seam, alongside the read/resolve paths. The guard uses path_within_root_lexical, NOT the fail-closed path_within_root the sibling read paths use: a create target never exists yet, so path.canonicalize() always fails for it and the fail-closed guard would refuse *every* create, including legitimate in-root ones. The lexical guard refuses out-of-root and interior-`..` escapes while admitting a not-yet-created in-root target, and still canonicalizes to catch symlink escapes when the target exists. (AC #5 named path_within_root; flagged on the PR — same class of AC defect as the #3/#4 dispute resolved via Option 1.) Add code_action_create_containment.rs: out-of-root, interior-`..`, in-root positive control, and under-root symlink-escape cases. Reframe the resolve-seam negative tests per the #199 escalation (Option 1): they assert the *invariant* (an out-of-root / symlink-escaping component never resolves), with the resolve_component_file guard documented as the innermost backstop whose reject branch is unreachable via the public API today — the upstream path_within_root_lexical filter catches both negatives first. Drop the false "only path_within_root catches" claim and correct the misleading assert messages. Fixes: #199 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: 🔒️ Guard sibling create paths in build_code_action multi-file actions The write/create-seam backstop (issue #199) only validated `self.target_path`, but the multi-file action types each emit a SECOND `ResourceOp::Create`: - `Livewire` → a Blade view at `get_livewire_view_path` - `BladeComponentWithClass` → a PHP class at `get_component_class_path` Both paths are derived from `self.name` — an independent diagnostic field, not coupled to `target_path`. Because `PathBuf::join`/`push` of an absolute-looking segment *replaces* the base, a forged diagnostic with an in-root `target_path` and `name = "/etc/passwd"` slipped past the guard and materialised a file outside the project root via the second create — the exact containment escape this chain exists to close. Guard each sibling path with `path_within_root_lexical(&path, root)` and return `None` on escape, and correct the guard comment that wrongly asserted every create materialises only `target_path`. Adds non-vacuous multi-file coverage in `code_action_create_containment.rs`: a `Livewire` / `BladeComponentWithClass` action with an in-root `target_path` but an escaping `name`-derived view/class path returns `None` (each would be `Some` without the guard), plus in-root positive controls for both types. Refs #199 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

@include

… the emit-safe guard. The component-diagnostic loop (main.rs:16683) built its "Expected at:" hint from an unfiltered `possible_paths.first()` — the fourth and last surface in the `from_diagnostic` → `CreateFile` family, still echoing a raw candidate after the view()/@extends/@include loops were closed. A dangling under-root symlink (or out-of-root namespace path) planted at the conventional component path was echoed into the message, parsed back into a create target, passed the lexical-only #199 backstop, and a client `CreateFile` could follow it out of the project tree. Route it through `in_root_expected_path_hint` so all four "Expected at:" surfaces share the emit-safe guard. Also folds in the test hardenings from the same review round: - New `emit_safe_refuses_unverifiable_candidate_behind_a_symlink_loop` pins the `ELOOP` fail-closed case of the emit-safe `None` arm (deterministic on every unix uid), completing the EACCES/ENOTDIR/ELOOP trio named in its contract. - `emit_safe_refuses_out_of_root_candidate` now writes the escapee to disk and asserts the precondition, so it unambiguously proves an *existing* out-of-root file is refused, not merely an absent one. - `expected_path_hint_picks_first_in_root_candidate` gains the `canonicalize().is_err()` precondition matching its sibling tests. - New `component_expected_path_hint_is_unknown_when_all_candidates_out_of_root` regression test mirrors AC #5 for the component surface. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

@include

…stic surfaces (Livewire diagnostic fallback + "Expected at:" message hint) (#202) * chore: start work on #201 * fix: 🔒️ Extend path containment guard to remaining diagnostic surfaces. Route the Livewire "component not found" diagnostic fallback through any_in_root_candidate_exists so an out-of-root loadViewsFrom/ component_namespaces registration can no longer make diagnostics stat-probe files outside the project tree, nor let an out-of-root view that exists on disk silently satisfy the check. Source the "Expected at:" message hint from the first in-root candidate (new in_root_expected_path_hint helper, shared by the view() and @extends/@include loops) so a maliciously-registered out-of-root namespace path is never echoed back to the client in the message text. Add Livewire-fallback and message-hint cases to view_diagnostic_containment. Fixes: #201 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: 🔒️ Refuse dangling under-root symlinks in the "Expected at:" diagnostic hint The "Expected at:" create-this-file hint baked into a "view not found" diagnostic is parsed back into a CreateFile target by `from_diagnostic`, so it is an *emitted* path. The branch sourced it via `path_within_root_lexical`, whose docstring explicitly disclaims emitted paths — its `unwrap_or(true)` admits a dangling under-root symlink whose target is missing, which a client `CreateFile` could follow out of the project tree (issues #134/#155). Add a third containment entry point, `path_within_root_emit_safe`: it shares the lexical out-of-root gate and still admits a genuinely-absent in-root path (the hint is for a missing view, so fail-closing would break "create view"), but refuses a leaf dangling symlink via `symlink_metadata` — the residual the lexical guard leaves open. Point `in_root_expected_path_hint` at it (both "Expected at:" call sites share the helper, AC #3/#4). Extract the shared lexical gate into `lexically_in_root` to keep the module's single-source-of- truth invariant. Closes Holmes's escalated AC-vs-documented-contract dispute on PR #202, per the decision to close the residual in this PR rather than defer it. Refs #201 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: 🔒️ Refuse non-NotFound lstat errors in emit-safe containment guard `path_within_root_emit_safe`'s `None` arm tested `symlink_metadata().is_err()`, which is true for EVERY lstat error — not just `NotFound`. A lexically-in-root candidate that fails to canonicalize and whose lstat returns `EACCES` (a no-search-permission parent hiding a dangling under-root symlink), `ENOTDIR`, or `ELOOP` was therefore ADMITTED and echoed into the "Expected at:" diagnostic hint, where a client `CreateFile` could follow it out of the project tree — the exact escape this guard exists to refuse. The doc comment already specified the discriminating condition as `NotFound`; the code failed open against it. Discriminate on `NotFound` so the guard fails closed on anything it cannot prove absent, admitting only a genuinely-absent speculative create target. - `is_err()` → `is_err_and(|e| e.kind() == ErrorKind::NotFound)` - Tighten the doc comment to state non-`NotFound` errors fail closed - Add two `#[cfg(unix)]` tests pinning the corrected contract: an unverifiable candidate behind a file path component (`ENOTDIR`, deterministic) and a dangling under-root symlink behind a no-search-permission parent (`EACCES`, the case Holmes named) must both be refused Addresses Holmes's review blocker on PR #202. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> * fix: 🔒️ Route the "Blade component not found" diagnostic hint through the emit-safe guard. The component-diagnostic loop (main.rs:16683) built its "Expected at:" hint from an unfiltered `possible_paths.first()` — the fourth and last surface in the `from_diagnostic` → `CreateFile` family, still echoing a raw candidate after the view()/@extends/@include loops were closed. A dangling under-root symlink (or out-of-root namespace path) planted at the conventional component path was echoed into the message, parsed back into a create target, passed the lexical-only #199 backstop, and a client `CreateFile` could follow it out of the project tree. Route it through `in_root_expected_path_hint` so all four "Expected at:" surfaces share the emit-safe guard. Also folds in the test hardenings from the same review round: - New `emit_safe_refuses_unverifiable_candidate_behind_a_symlink_loop` pins the `ELOOP` fail-closed case of the emit-safe `None` arm (deterministic on every unix uid), completing the EACCES/ENOTDIR/ELOOP trio named in its contract. - `emit_safe_refuses_out_of_root_candidate` now writes the escapee to disk and asserts the precondition, so it unambiguously proves an *existing* out-of-root file is refused, not merely an absent one. - `expected_path_hint_picks_first_in_root_candidate` gains the `canonicalize().is_err()` precondition matching its sibling tests. - New `component_expected_path_hint_is_unknown_when_all_candidates_out_of_root` regression test mirrors AC #5 for the component surface. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

The #246 strict clear fired for ANY unresolvable pending hop once the chain reached EloquentCollection mode — but heuristic guesses (local scopes like User::forCurrentTenant(), unmodeled builder methods) share the pending-hops queue with genuine relation claims. A scope followed by ->get() flipped the mode, the scope name failed to resolve as a relation, and effective_model was cleared — silencing column-typo diagnostics that main correctly flagged on the root table. - chain: type the queue — RelationHop { name, kind } with RelationHopKind::{Claim, Heuristic} replacing the untyped Vec<String> - cursor: the RelationProperty receiver seed (property access or executed relation assignment) is a Claim; mid-chain unknowns are Heuristic - finalize: a miss clears effective_model only for Claim hops; Heuristic misses keep the skip-and-continue fallback in every mode - tests: regression coverage for the scope-then-terminator shape (diagnostics + unit), plus a dedicated end-to-end completion test for the executed-relation collection shape (review follow-up, AC #5)

* chore: start work on #246 * fix: 🐛 type executed relation-query collections against the related model $var = $user->competitions()->select(…)->get() previously typed $var as an EloquentBuilder on the ROOT model (User), so collection chains on it false-positived "Column not found" against the users table. - flow: detect the executed-relation assignment shape (relation call at the chain root, recognised builder methods between, COLLECTION_TERMINATOR on top; instance or static Eloquent root) via resolve_collection_relation - extractor: emit the RelationProperty receiver for it — collection mode at the base model with the relation queued as a pending hop, resolved by the existing async finalize step - finalize: a collection-mode hop that fails to resolve now clears effective_model (stay quiet) instead of falling back to the root model; builder-mode heuristic hops keep the skip fallback Fixes #246 * fix: 🐛 gate strict relation-hop clear on claim hops, not collection mode The #246 strict clear fired for ANY unresolvable pending hop once the chain reached EloquentCollection mode — but heuristic guesses (local scopes like User::forCurrentTenant(), unmodeled builder methods) share the pending-hops queue with genuine relation claims. A scope followed by ->get() flipped the mode, the scope name failed to resolve as a relation, and effective_model was cleared — silencing column-typo diagnostics that main correctly flagged on the root table. - chain: type the queue — RelationHop { name, kind } with RelationHopKind::{Claim, Heuristic} replacing the untyped Vec<String> - cursor: the RelationProperty receiver seed (property access or executed relation assignment) is a Claim; mid-chain unknowns are Heuristic - finalize: a miss clears effective_model only for Claim hops; Heuristic misses keep the skip-and-continue fallback in every mode - tests: regression coverage for the scope-then-terminator shape (diagnostics + unit), plus a dedicated end-to-end completion test for the executed-relation collection shape (review follow-up, AC #5) * fix: 🐛 discriminate scope calls and mode-flips in executed-relation collections * fix: 🐛 collect unrecognised middle calls as heuristic hops in executed-relation collections The resolve_collection_relation shape gate rejected any assignment chain with an unrecognised middle call, so idiomatic Laravel chains kept the original #246 false positive: pivot builder methods (->wherePivot(...), absent from the recognised-builder catalog), scopes on the related model (->competitions()->approved()->get()), and root scopes before the relation (->forCurrentTenant()->competitions()->get()). Mirror the inline cursor collector: unrecognised middles no longer reject — each becomes a RelationHopKind::Heuristic hop carried on the receiver (call_hops) and resolved in source order after the relation claim by the shared finalize step; a miss keeps the running model. The round-3 mode-flip gate is unchanged: a recognised middle whose ChainEffect isn't None (mid-chain get()/toBase()/first()) still rejects. Proactive round-4 hardening per human review direction on PR #266.

`get_phpunit_env_context` parses two PHPUnit spellings, `<env name="` and `<server name="`, through separate arms of one match — different literals, different offsets (`e + 11` against `s + 14`). The suite drove the first spelling only, so the second had never reached `completion()` anywhere in the crate, and the pull request re-quoted AC #5 without its `<server name="…">` clause and without noting the narrowing. Two tests close it: the typed prefix and the empty-prefix repro, mirroring the `<env name="` pair. Both pin the offset arithmetic — rewriting `s + 14` to `s + 11` fails these two tests and nothing else in the crate. The module doc comment claimed that restoring the deleted `std::env::vars()` loop turns every test in the file red. It does not. The shadowing test's fixture `.env` declares the process variable's own name, so the loop's own `!seen_names.contains(&name)` guard skipped that variable before this fix as well. The comment now says nine of ten, names the test that stays green, and names the second mutation that discriminates it. Two neighbouring claims in the same block are corrected with it: the context list now carries both PHPUnit spellings, and the "each test asserts a positive control" claim now records its one deliberate exception. Verified by running the mutations, not by inspection: restoring the loop fails 9 of 10, the `.env` echo-format mutation fails the tenth, the offset mutation fails exactly the two new tests. Watson-Branch: #342

…s inline (#343) * chore: start work on #342 Watson-Branch: #342 * fix: 🐛 Stop offering the LSP process environment in env completion. The env-completion handler merged `std::env::vars()` into its item list and rendered each value inline in `detail` and `documentation`. The language server inherits Zed's environment, which inherits the login shell's, so any credential whose name matched the typed prefix was offered with its value on screen — worst during screen-sharing and recording, when a completion popup is most visible. It was also wrong independent of secrets: `${...}` interpolation inside `.env` and `env()` at runtime both resolve against the dotenv file set, never against the editor's process environment, so those names would not exist wherever the application actually runs. Completion now offers only what the project declares in its `.env` files. The `seen_names` set went with the loop — its only read site was the merge's duplicate check. The `.env` echo path (value plus source file) is untouched. Eight tests drive the real `textDocument/completion` entry point across all three env contexts with a real process variable set, including the issue's empty-prefix repro and the `Ok(None)` response shape. Fixes: #342 * test: ✅ Cover <server name="…"> in the env-completion leak suite. `get_phpunit_env_context` parses two PHPUnit spellings, `<env name="` and `<server name="`, through separate arms of one match — different literals, different offsets (`e + 11` against `s + 14`). The suite drove the first spelling only, so the second had never reached `completion()` anywhere in the crate, and the pull request re-quoted AC #5 without its `<server name="…">` clause and without noting the narrowing. Two tests close it: the typed prefix and the empty-prefix repro, mirroring the `<env name="` pair. Both pin the offset arithmetic — rewriting `s + 14` to `s + 11` fails these two tests and nothing else in the crate. The module doc comment claimed that restoring the deleted `std::env::vars()` loop turns every test in the file red. It does not. The shadowing test's fixture `.env` declares the process variable's own name, so the loop's own `!seen_names.contains(&name)` guard skipped that variable before this fix as well. The comment now says nine of ten, names the test that stays green, and names the second mutation that discriminates it. Two neighbouring claims in the same block are corrected with it: the context list now carries both PHPUnit spellings, and the "each test asserts a positive control" claim now records its one deliberate exception. Verified by running the mutations, not by inspection: restoring the loop fails 9 of 10, the `.env` echo-format mutation fails the tenth, the offset mutation fails exactly the two new tests. Watson-Branch: #342 * test: ✅ Assert the .env documentation panel in the env-leak suite. AC #8 requires the shadowing regression to confirm the .env-declared variable still completes with its value AND `Source: <file>`. AC #4 fixes what that names: `detail` is `"<value> (from <file>)"`, `documentation` is `Source: <file>`. The suite asserted `label` and `detail` and never read `documentation`, so deleting `.section(format!("Source: {}", source_file))` from `completion()` left all 3185 tests green. Sweeps the field enumeration rather than the flagged member: the panel is asserted whole, so dropping any one of the three `CompletionDoc` builder calls — `.header`, `.summary`, `.section` — fails 9 of the module's 10 tests, verified live for each. Test-only change; `main.rs` is untouched. Watson-Branch: #342 * chore: 🔁 Re-trigger CI on an unchanged tree. GitHub accepted the push of dbd4792 — CodeQL ran on that SHA — but never dispatched the `pull_request`-triggered CI workflow for it. ci.yml has no paths filter and no concurrency block, and the workflow is active, so the event was dropped on GitHub's side. Closing and reopening the pull request did not dispatch it either. This commit is empty: the tree is byte-identical to dbd4792. It exists only to raise a fresh synchronize event so the required checks run. Watson-Branch: #342

refactor: 👽️ Update diagnostic source from 'laravel-lsp' to 'laravel'.

8b808fa

mikebronner self-assigned this May 12, 2026

mikebronner merged commit 9a9eb09 into main May 12, 2026

mikebronner deleted the update-lsp-reference branch May 12, 2026 15:31

mikebronner mentioned this pull request Jun 8, 2026

feat: goto-definition + hover for Artisan command strings #65

Merged

10 of 11 tasks

mr-sherlock-holmes Bot mentioned this pull request Jun 13, 2026

feat: rename — Blade variable rename (scope-aware + cross-file from controller) #89

Merged

9 of 10 tasks

mr-sherlock-holmes Bot mentioned this pull request Jun 14, 2026

feat: Folio page routing support (goto / completion / find-references) #91

Merged

9 tasks done

mikebronner mentioned this pull request Jun 14, 2026

Folio completion: add a handler-level test asserting Folio routes are offered #101

Closed

5 tasks

mr-sherlock-holmes Bot mentioned this pull request Jun 14, 2026

feat: Flux component support (goto-definition + hover + completion) #99

Merged

9 tasks done

dr-john-h-watson Bot mentioned this pull request Jun 15, 2026

Internalize containment check in locate_slot_in_view so the traversal invariant can't be re-opened by a new caller #146

Closed

7 tasks

mr-sherlock-holmes Bot mentioned this pull request Jun 15, 2026

Run slot-navigation containment guard before file_exists_cached to close out-of-root existence oracle #159

Closed

9 tasks done

mikebronner mentioned this pull request Jun 16, 2026

fix: make @php/@endphp block masking string-literal safe #164

Merged

12 tasks done

This was referenced Jun 16, 2026

Add handler-level tests for Blade variable rename wiring #165

Merged

discovery: map the full boundary of string-keyed & cross-scope PHP-variable rename references #167

Merged

dr-john-h-watson Bot mentioned this pull request Jun 17, 2026

feat: silence Intelephense false 'undefined property' (P1014) on Eloquent magic members #73

Open

8 tasks

This was referenced Jun 17, 2026

Reject out-of-root candidates lexically before canonicalize() in resolve_component_path #197

Merged

Extend path_within_root containment guard to the <livewire:>-tag goto-definition flow (create_livewire_location_from_salsa) #198

Merged

mikebronner mentioned this pull request Jun 18, 2026

Add the fail-closed path_within_root guard to resolve_component_file for containment-invariant uniformity #203

Merged

9 tasks done

mr-sherlock-holmes Bot mentioned this pull request Jun 18, 2026

Extend the path_within_root containment guard to the remaining diagnostic surfaces (Livewire diagnostic fallback + "Expected at:" message hint) #202

Merged

10 tasks done

mr-sherlock-holmes Bot mentioned this pull request Jun 18, 2026

Extend the path_within_root emit-safe guard to the translation/config (and middleware/feature/env) "Expected at:" diagnostic surfaces #214

Closed

8 tasks

dr-john-h-watson Bot mentioned this pull request Jul 14, 2026

fix: registry reverse-index lag on provider-body edits + nondeterministic equal-priority merge #264

Merged

12 tasks done

mr-sherlock-holmes Bot mentioned this pull request Jul 23, 2026

fix: honor custom $collectionClass in Blade variable query-builder inference #275

Merged

9 tasks done

mr-sherlock-holmes Bot mentioned this pull request Aug 28, 2026

fix: 🐛 Env completion offers the LSP process's own environment, values inline #343

Merged

15 tasks done

This was referenced Aug 28, 2026

fix: mask sensitive .env values in LSP output #348

Merged

feat: ✨ Hover and go-to-definition on keys in .env buffers #353

Merged

Cache-hit regression tests are blind to config_lookup::resolve_value's disk reads #357

Merged

Navigation Menu

Uh oh!

refactor: 👽️ Update diagnostic source from 'laravel-lsp' to 'laravel'. - #5

refactor: 👽️ Update diagnostic source from 'laravel-lsp' to 'laravel'.#5
mikebronner merged 1 commit into
mainfrom
update-lsp-reference

mikebronner commented May 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Conversation

mikebronner commented May 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant