in_http: replace WEBrick::HTTPUtils.parse_query with URI #4900
Conversation
WEBrick is no longer recommended for production use. WEBrick::HTTPUtils is just a util, so it might be harmless. However, it would be happy to reduce dependence if possible. Signed-off-by: Daijiro Fukuda <[email protected]>
| @@ -643,6 +643,10 @@ def include_cors_allow_origin | |||
|
|
|||
| !r.nil? | |||
| end | |||
|
|
|||
| def parse_query(query) | |||
| query.nil? ? {} : Hash[URI.decode_www_form(query, Encoding::ASCII_8BIT)] | |||
There was a problem hiding this comment.
';' separator to be ignored. Is it ok here?
(I'm not sure difference WEBrick::HTTPUtils.parse_query and URI.decode_www_form)
There was a problem hiding this comment.
There was a problem hiding this comment.
Oh, I didn't notice that.
If we want to keep that specification, it would be better not to replace it by force.
WEBrick::HTTPUtilsis just a util, so it might be harmless.
There was a problem hiding this comment.
It is a breaking change, but it might be acceptable for v1.19.
Should we move forward?
There was a problem hiding this comment.
Maybe we should give it some time.
We want to remove runtime dependence on WEBrick, but I don't see using WEBrick::HTTPUtils as a problem because it is just utils.
If more people are not happy with this change of specifications, then it might be better not to include this modification.
Which issue(s) this PR fixes:
What this PR does / why we need it:
Replace
WEBrick::HTTPUtils.parse_querywithURI.decode_www_forminin_http.WEBrick is no longer recommended for production use.
WEBrick::HTTPUtilsis just a util, so it might be harmless.However, it would be happy to reduce dependence if possible.
This changes some specifications unintentionally.
At least, this makes us unable to use
;delimitter.We need to consider whether this breaking change is acceptable or not.
Docs Changes:
Not needed. (There is no impact on the specification.)
Release Note:
The same as the title.