Chris,

On Mon, Aug 5, 2024 at 5:10 PM Christopher Schultz <
ch...@christopherschultz.net> wrote:

> Coty,
>
> On 8/5/24 14:56, Coty Sutherland wrote:
> > On Mon, Aug 5, 2024 at 2:43 PM Rainer Jung <rainer.j...@kippdata.de>
> wrote:
> >
> >> Am 05.08.24 um 20:18 schrieb Coty Sutherland:
> >>> I'm testing and see an issue with o.a.c.http2.TestStreamProcessor
> getting
> >>> some unexpected result:
> >>>
> >>> Testcase: testPrepareHeaders[0: loop [0], useAsyncIO[false]] took 0.061
> >> sec
> >>>       FAILED
> >>> expected:<...-Header-[etag]-[W/"9[34]-1447269522000"]
> >>> 3-H...> but was:<...-Header-[etag]-[W/"9[57]-1447269522000"]
> >>> 3-H...>
> >>> junit.framework.AssertionFailedError:
> >>> expected:<...-Header-[etag]-[W/"9[34]-1447269522000"]
> >>> 3-H...> but was:<...-Header-[etag]-[W/"9[57]-1447269522000"]
> >>> 3-H...>
> >>>       at
> >>>
> >>
> org.apache.coyote.http2.TestStreamProcessor.testPrepareHeaders(TestStreamProcessor.java:167)
> >>>       at
> >>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
> >>> Method)
> >>>       at
> >>>
> >>
> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77)
> >>>       at
> >>>
> >>
> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> >>
> >> The test class contains:
> >>
> >>           // Different line-endings -> different files size -> different
> >> weak eTag
> >>           if (JrePlatform.IS_WINDOWS) {
> >>
> >>   expected.append("3-Header-[etag]-[W/\"957-1447269522000\"]\n");
> >>           } else {
> >>
> >>   expected.append("3-Header-[etag]-[W/\"934-1447269522000\"]\n");
> >>           }
> >>
> >> The number behind the "W" (weak" in the etag header ist the file size,
> >> the second (correct) number the timestamp. It seems you get the 957 size
> >> exoected for Windows line endings instead of the 934 for Unix line
> >> endings. You could check the file test/webapp/index.html for size/line
> >> endings. Maybe your git settings lead to a DOS checkout instead of a
> >> unix one?
> >>
> >
> > Bah, yeah. It's not a git setting since I'm running the test from the
> > src.zip but the line endings are Windows (vim flags [noeol][dos]) :) My
> > fault for trying to run the test suite from the zip on Linux I guess heh
>
> You could use this:
>
>
> https://github.com/ChristopherSchultz/apache-tomcat-stuff/blob/main/bin/test-tomcat-release.sh
>
>
Thanks, maybe I'll do that next time :) Seems the most recent Fedora
upgrade I did made things a bit screwy.

Reply via email to