fix(Content-Disposition): properly parse and decode filename from header #7159
Conversation
|
Hi maintainers 👋 This PR isolates the ✅ Handles both quoted and UTF-8 encoded ( Thanks for reviewing! 🙏 |
There was a problem hiding this comment.
It's unclear why you're regressing the code to use var, using arrays instead of a hash object, and matching for both regular expressions at once. The only purpose of parseHeaders is to parse headers into an object representation. There shouldn't be any high-level manipulations that could lose or corrupt the original headers. For high-level header manipulations, the AxiosHeaders class was introduced. Similar logic can be written in AxiosHeaders.getAttachment() method, but definitely not in the low-level parseHeaders utility. This way, the user will be able to operate with the original headers and obtain the decoded file name if needed.
|
Thanks for the review, @DigitalBrainJS 🙏 You're absolutely right — the Once I refactor it accordingly, I’ll update the PR for re-review. |
…Disposition header according to RFC5987
|
All Node.js tests pass — 224/224 passing. This PR fixes filename extraction from the Summary of changes
Results
**Ready for review ** |
2 participants
Summary
Fixes filename extraction from the
Content-Dispositionheader by supporting both quoted and UTF-8 encoded (filename*) formats according to RFC5987.Changes
lib/helpers/parseHeaders.jsto correctly parse and decode bothfilenameandfilename*values.test/unit/helpers/parseHeaders.spec.js.Testing
npm test).Content-Disposition: attachment; filename="test.txt"Content-Disposition: attachment; filename*=UTF-8''résumé.pdfNotes
should support max content lengthis unrelated and can be ignored.Fixes #7142