URL: <http://savannah.gnu.org/bugs/?52531>
Summary: Content-Disposition parameter continuations should use filename*0*= not filename**0= Project: GNU Wget Submitted by: None Submitted on: Mon 27 Nov 2017 01:44:19 PM UTC Category: Program Logic Severity: 3 - Normal Priority: 5 - Normal Status: None Privacy: Public Assigned to: None Originator Name: Originator Email: r...@robwu.nl Open/Closed: Open Discussion Lock: Any Release: None Operating System: None Reproducibility: Every Time Fixed Release: None Planned Release: None Regression: None Work Required: None Patch Included: None _______________________________________________________ Details: http://git.savannah.gnu.org/cgit/wget.git/commit/?id=d03b40e31e9aa06669c3ae0113ed24273e1b60e5 fixed #44628 by supporting `filename*`. It also introduced a new test: ``` { "filename**0=\"A\"; filename**1=\"A.ext\"; filename*0=\"B\";\ +filename*1=\"B\"", "AA.ext", true }, }; ``` This seems to be an attempt at implementing parameter continuations. Parameter continuations are specified by [RFC 2231, section 4.1](https://tools.ietf.org/html/rfc2231#section-4.1). The following example is given (edited with [errata](https://www.rfc-editor.org/errata/rfc2231)): ``` Content-Type: application/x-stuff; title*0*=us-ascii'en'This%20is%20even%20more%20 title*1*=%2A%2A%2Afun%2A%2A%2A%20 title*2="isn't it!" ``` This shows two things: 1. The test should be `filename*0*=` instead of `filename**0=`. 2. When `filename*=` is used, the value should start with `encoding'language'` (this is "ext-value" from RFC 5987). In practice some servers neglect to send this prefix, so to be fault-tolerant it is okay to treat a missing prefix as if the prefix was `''`. This bug report is opened to suggest that 1 should be fixed. _______________________________________________________ Reply to this item at: <http://savannah.gnu.org/bugs/?52531> _______________________________________________ Message sent via/by Savannah http://savannah.gnu.org/