I agree with changing the test code, but I did look at the implementation of Header [1] and think that using a TreeMap to store the fields is a bit of overkill.
I assume this was chosen just to get the case insensitive key matching behavior, but I think it will be a memory hog. And then the whole other structure to track the field values ... urgh. [1] https://svn.apache.org/repos/asf/harmony/enhanced/java/trunk/classlib/modules/luni/src/main/java/org/apache/harmony/luni/internal/net/www/protocol/http/Header.java Regards, Tim On 11/Jun/2010 09:58, Ray Chen (JIRA) wrote: > [ > https://issues.apache.org/jira/browse/HARMONY-6516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel > ] > > Ray Chen updated HARMONY-6516: > ------------------------------ > > Attachment: Harmony6516_v2.diff > > Change the test code, to test whether the list contains the item instead of > compare the value which depends on the order. > >> [java6][classlib][luni]Items in map which reterned by >> HttpURLConnection.getHeaderFields() should be in reverse order >> --------------------------------------------------------------------------------------------------------------------- >> >> Key: HARMONY-6516 >> URL: https://issues.apache.org/jira/browse/HARMONY-6516 >> Project: Harmony >> Issue Type: Bug >> Components: Classlib >> Reporter: Ray Chen >> Attachments: Harmony6516_v2.diff, Harmony_6516.diff >> >> >> RI6 will retrun a map in which the items are in reverse order, however RI5 >> didn't require that. >> So only fix this in java6 branch >
