Repository navigation
chore(deps): bump tree-sitter-go from 0.23.4 to 0.25.0 - #15
dependabot[bot] wants to merge 1 commit into
Conversation
8cb712f to
a38533c
Compare
Bumps [tree-sitter-go](https://github.com/tree-sitter/tree-sitter-go) from 0.23.4 to 0.25.0. - [Release notes](https://github.com/tree-sitter/tree-sitter-go/releases) - [Commits](tree-sitter/tree-sitter-go@v0.23.4...v0.25.0) --- updated-dependencies: - dependency-name: tree-sitter-go dependency-version: 0.25.0 dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com>
a38533c to
0a22ad2
Compare
🌱 graft blast radiusNothing outside this diff depends on it. 0 areas changed; no indexed dependents at depth 2.
Open the interactive graph → — click an area to see its dependent symbols at file:line. |
|
Closing this Dependabot bump: tree-sitter-go 0.25 needs the tree-sitter 0.25 runtime, and #40 upgrades both together. On its own this PR fails CI. |
|
OK, I won't notify you again about this release, but will get in touch when a new version is available. If you'd rather skip all updates until the next major or minor version, let me know by commenting If you change your mind, just re-open this PR and I'll resolve any conflicts on it. |
…lhq#15) * fix(init): keep machine-wide wiring when init omits a host init converges a repository by retracting every host not named in --agents, and that sweep also ran at machine-wide scope. So `graft init B --agents cursor` removed the user-level Claude hooks, shim and MCP entry and the Codex hooks that every other repository on the machine depends on. The init-time retract is now repository-only; `uninstall` still removes the machine-wide files. Ported from nishadherath/fork_Graft a241f6f (MIT). Upstream issue trailhq#482. Co-authored-by: Nishad Herath <82203546+nishadherath@users.noreply.github.com> * fix(cli-meta): start npm through the shell on Windows On Windows npm is the npm.cmd shim, which spawnSync cannot start without a shell, so `graft version` always read "unreachable", the update check stored nothing and `graft upgrade` failed with ENOENT. All npm calls now go through one helper that passes a joined command line to the shell on Windows (avoiding Node 24's DEP0190 warning) and spawns npm directly elsewhere. The tests put a fake npm first on PATH. Ported from ansidium/trail-code-se7en 4fa301e (MIT). Upstream issue trailhq#487. Co-authored-by: ansidium <ansikun@gmail.com> * fix: keep a --dir on another drive out of the repo's ignore files On Windows a --dir on another drive has no relative path, so relative() returns the absolute target. That passed the startsWith("..") outside-repo check and appended /C:/... to the repository's .gitignore and .ignore. Both writers now share repoRelativeDir, which rejects anything not strictly inside the root. Ported from nishadherath/fork_Graft 50122fb (MIT). Upstream issue trailhq#384. Co-authored-by: Nishad Herath <82203546+nishadherath@users.noreply.github.com> * fix(graph): re-parse a cached entry that holds a parse error A grammar abort is a process-global, run-dependent failure, but the extraction cache stores it as a per-file entry keyed on content hash and replays it on every later build. An untouched file then stays broken forever, and a whole language can drop out of a graph that still exits 0 with a success line. Let an entry carrying an error fall through to the parse path instead of being replayed. The entry is still persisted, so the freshness probe does not report the file as new on every query; it just never stands in for a result again. Re-parsing costs only files that produced no nodes, and a run that succeeds repairs the entry. Upstream issue trailhq#312, PR trailhq#317 (MIT). Cherry-picked from Ercaner1988/Graft 14c0bcd. * fix(ingest): stop spreading a child subtree into push in walkFilesystem `out.push(...walkFilesystem(full, includes))` turns the child's array into push()'s argument list, and V8 caps that at roughly 128k. A single unignored subtree with more paths than that throws "Maximum call stack size exceeded" at trivial directory depth — the message is about the argument list, not recursion. Only the git-less path reaches this: when listGitFiles returns null the walker collects every non-dot file first and filters by language later, so any large cache, browser profile or dataset directory in a repo that was never `git init`-ed is fatal. Appending in a loop removes the cap without changing what is collected or in what order. Two tests, deliberately split: - a cheap one that pins the accumulation itself (a wide, deep subtree outside Git comes back flat and complete). It runs everywhere and would catch a botched loop, but it cannot reach the argument cap. - one that actually reproduces the crash with 150k files. It fails with "Maximum call stack size exceeded" on the previous code and passes with this one. CI also runs windows-latest, where creating that many files is far slower, so it is skipped there. Measured while writing this: on Node 22 a spread of 125k elements still succeeds and 150k throws, and `--stack-size` does not move the boundary — it is an argument-count limit, not stack depth. Upstream issue trailhq#314, PR trailhq#315 (MIT). Cherry-picked from Ercaner1988/Graft 88a0442. * test(ingest): run the argument-cap walker test on Windows The 150,000-file regression test for the walkFilesystem stack overflow was POSIX-only because creating the files is slow on Windows. It takes about 35 seconds there, fails with "Maximum call stack size exceeded" without the fix, and covers the platform graft is used on daily. * test(generic): expect a cached grammar failure to be re-parsed The re-parse fix for cached parse errors changes one assertion in the throwing-grammar test: an unchanged file whose last parse failed is now parsed again instead of replayed. The test's point still holds: the failure is cached as a failure, not an empty success, and the rebuild reports it again. * style(cli-meta): spawn npm with the repo's platform shell idiom Use one spawnSync call with `shell: win`, as signup.ts, cli.ts and shim-template.ts already do, and mark the PATH lookup of npm as intended for SonarCloud. Behaviour is unchanged: a joined command line through the shell on Windows, a direct spawn elsewhere. * style(cli-meta): mark the Windows npm shell spawn as reviewed for Semgrep The shell is required to start npm.cmd on Windows and every argument is a fixed token, so suppress the spawn-shell-true finding inline with that reason. --------- Co-authored-by: Ryan Duguid <ryanduguid@users.noreply.github.com> Co-authored-by: Nishad Herath <82203546+nishadherath@users.noreply.github.com> Co-authored-by: ansidium <ansikun@gmail.com> Co-authored-by: datrixlab <325650023+datrixlab@users.noreply.github.com> (cherry picked from commit 8766308)
Bumps tree-sitter-go from 0.23.4 to 0.25.0.
Release notes
Sourced from tree-sitter-go's releases.
Commits
15476780.25.03f912e9chore: generate179ca03feat: expose statement liste25214efix: allow the terminator to be omitted for the last elementedea6bffix: give index expressions a dynamic precedence of 1e1076e5feat: support generic type aliases00a299eci: update test failures, use macos-1593c2bb6build: update bindings1496eb7feat: use the new reserved rules apic350fa5ci: bump actions/checkout from 4 to 5