Niels, my debugging attempt produced no useful information so I reverted it. Now I understand what you wrote: the failure occurs in freemarker and the response is empty.
Placing backslashes or quotes in the test file name gives different errors, earlier or later. Removing resourceDirectory.ftl gives me exactly the same error message on Linux, indicating a silent empty response: org.xml.sax.SAXParseException; lineNumber: 1; columnNumber: 1; Premature end of file Perhaps there is a problem locating this template on Windows? Case-insensitive filesystem? Kind regards, Ben. On 17/02/16 09:48, Ben Caradoc-Davies wrote: > Niels, > > I added two System.stderr.println statements to print the HTML responses > before parsing is attempted: > https://github.com/geoserver/geoserver/commit/e670b241f8322a755a0c893b2ecb65b48c33f432 > > The next Winbuild failure log should give us an idea what is going on. > > Kind regards, > Ben. > > On 17/02/16 09:14, Ben Caradoc-Davies wrote: >> Niels, >> >> the only difference I can see is the child link. Could Windows be >> returning a resource called '"mydir/myres"' (with the quotes in the >> string) or a backslash or something else which would break the HTML >> parser? Some problem with resource path canonicalisation of wWindows? I >> thought this situation would be covered in the Resource unit tests, >> which pass on Windows. Can you confirm? >> >> The encoding of the child href is very thin (only ampersand is >> replaced). Encoding " to " might give a better error message. And >> what about percent-encoding? What happens to spaces or # or other >> characters in resource names? >> >> I am guessing a quote or backslash in the child resource href on Windows. >> >> Kind regards, >> Ben. >> >> On 17/02/16 04:54, Niels Charlier wrote: >>> Hi Ben, >>> >>> I solved an issue in the test with Italian dates on the windows server, >>> but testDirectoryHTML is still failing on windows, and it appears to be >>> the same problem as when we tested with |ISO-8859-15 |before we fixed >>> the template charset issue. it could be anything though, because it >>> seems that exceptions thrown by the template process are being captured >>> and ignored, and an empty result is returned instead. >>> >>> What is incomprehensible to me is why testResourceMetadataHTML is >>> working. The tests and the templates are nearly identical. This is a >>> mystery to me. Can you see any significant difference? >>> >>> @geosolutions: Can I get access to the server again to see if I can >>> figure it out? >>> >>> Regards >>> Niels >>> >>> >>> On 16-02-16 16:16, winbu...@geo-solutions.it wrote: >>>> See >>>> <http://winbuild.geo-solutions.it/jenkins/job/GeoServer-Master/2635/changes> -- Ben Caradoc-Davies <b...@transient.nz> Director Transient Software Limited <http://transient.nz/> New Zealand ------------------------------------------------------------------------------ Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 _______________________________________________ Geoserver-devel mailing list Geoserver-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geoserver-devel
