Repository navigation
refactor: Bump @humanfs/node from 0.16.6 to 0.16.8 - #955
Conversation
|
🚀 Thanks for opening this pull request! We appreciate your effort in improving the project. Please let us know once your pull request is ready for review. Tip
Note Please respond to review comments from AI agents just like you would to comments from a human reviewer. Let the reviewer resolve their own comments, unless they have reviewed and accepted your commit, or agreed with your explanation for why the feedback was incorrect. Caution Pull requests must be written using an AI agent with human supervision. Pull requests written entirely by a human will likely be rejected, because of lower code quality, higher review effort and the higher risk of introducing bugs. Please note that AI review comments on this pull request alone do not satisfy this requirement. Our CI and AI review are safeguards, not development tools. If many issues are flagged, rethink your development approach. Invest more effort in planning and design rather than using review cycles to fix low-quality code. |
Closes #937
Security update of a development-only dependency. It fixes Dependabot alert #432, GHSA-p498-v437-472g (medium): "humanfs: Recursive copy follows symlinked files and copies data from outside the source tree" (vulnerable
< 0.16.8, patched in 0.16.8).@humanfs/nodeis a transitive dependency ofeslint10.7.0, which requires^0.16.6, so onlypackage-lock.jsonchanges.npm ls @humanfs/node --omit=devis empty, so the package is not part of the production dependency tree; hence therefactor:prefix.Changes
@humanfs/node:directorytopackage.json(f691b60).@humanwhocodes/retrydependency moved from^0.3.0to^0.4.0. This only affectsbytes()(whichtext()andjson()use),write()andappend(), the only methods that go through the retrier. With retry 0.3.x, these calls started immediately. With 0.4.x, every call to them enters the retrier's queue, at most 1000 run at the same time perNodeHfsinstance (a call that is waiting for a retry still counts), and further calls wait until one finishes. As before, a call is only retried whenreadFile,writeFileorappendFilefails withENFILEorEMFILE.copy()andcopyAll()now copy symlinks as symlinks (22bbaa4). This is the fix for GHSA-p498-v437-472g. Before, a symlink to a file was copied as a regular file with the content of the file it points to, even when that file was outside the source tree. A symlink to a directory was passed tofs.copyFile(), which does not copy directories, so the copy most likely failed rather than copying the directory's content.@humanfs/typesmoved fromdevDependenciestodependencies(956ce7a, closes Bug:HfsWalkEntrytype missing since@humanfs/typesis a devDependency humanwhocodes/humanfs#145).@humanfs/corebumped from^0.19.1to^0.19.2.Dependencies that move with it (all are
@humanfs/node's own dependencies):@humanfs/core0.19.1 → 0.19.2: 0.19.2: Include type dependencies at runtime (956ce7a, closes Bug:HfsWalkEntrytype missing since@humanfs/typesis a devDependency humanwhocodes/humanfs#145). The published package only changes itspackage.json.@humanfs/types0.15.0 added. It only contains type definitions.@humanwhocodes/retry: the nested 0.3.1 copy under@humanfs/nodeis removed, and@humanfs/nodenow uses the top-level 0.4.2 thateslintalready uses. Changes between them:DEBUGvalue before outputting debug messages (fix: Look for specific DEBUG value before outputting debug messages humanwhocodes/retry#40)ESLint 10.7.0 only calls
hfs.isDirectory()andhfs.walk()from@humanfs/node(inlib/eslint/eslint-helpers.js). Neither of them changed in these versions, and neither ever goes through the retrier. ESLint does not callcopy()orcopyAll()either, so neither the retry change nor the symlink fix affects linting.The lock file was generated on the current
masterwith npm 10.9.3. Unlike the lock in #937, it leaves the optional peer dependenciessocks,smart-bufferandgcp-metadata(with its nestedgaxiosandnode-fetch) untouched.Breaking Changes
None
Code Changes Required
None — the upgrade is a drop-in replacement.
Summary by CodeRabbit