On 2/3/21 1:23 AM, Yann Ylavic wrote:
> On Tue, Feb 2, 2021 at 8:50 PM <rpl...@apache.org> wrote:
>>
>> Author: rpluem
>> Date: Tue Feb 2 19:50:14 2021
>> New Revision: 1886141
>>
> []
>> if (!scheme || u[0] != '/' || u[1] != '/' || u[2] == '\0') {
>> - if (!scheme && (u = strchr(url, ':')) && (u - url) > 14) {
>> - ap_log_rerror(APLOG_MARK, APLOG_ERR, 0, r, APLOGNO(10262)
>> - "overlong proxy URL scheme in %s", url);
>> - return HTTP_BAD_REQUEST;
>> - }
>> ap_log_rerror(APLOG_MARK, APLOG_DEBUG, 0, r, APLOGNO(01113)
>> "HTTP: declining URL %s", url);
>> return DECLINED; /* only interested in HTTP, WS or FTP via proxy */
>> }
>> + if (!scheme && (u = strchr(url, ':')) && (u - url) > 14) {
>> + ap_log_rerror(APLOG_MARK, APLOG_ERR, 0, r, APLOGNO(10262)
>> + "overlong proxy URL scheme in %s", url);
>> + return HTTP_BAD_REQUEST;
>> + }
>
> Hmm, actually this !scheme here is unreachable now.
Good catch.
> I think we can remove this test/block completely now that
Agreed. I think we should just decline and leave it to further possible
protocol handlers to deal
with these URL's. And if all decline we just hit the catch all that there is no
handler for this URL.
So +1 to just remove the block. Will you do?
Regards
Rüdiger
