Issue #SCB2303
Ronny Voelker <[email protected]>
| Newsgroups | gmane.comp.java.scarab.issues |
|---|---|
| Message-ID | <[email protected]> |
Activity report on *Defect SCB2303 - quick search can not work for Chinese * Scarab Link: http://www.solitone.org/scarab/issues/id/SCB2303 Module: Scarab Activity generated by Ronny Voelker ([email protected]) at 12/11/2007 11:32 *Comments* - By Ronny Voelker - 12/11/2007 11:32 --- "Yes, changing the method to POST would fix this specific bug, but we would still have to worry about every non-ascii character in a GET request. Changing the coyote parameters also didn't work for me. I have upgraded tomcat from 4.1.29 to the recent 4.1.37. Now the coyote parameters are working as expected (in fact the default settings are now sufficient). Btw. in the trunk also the complex query does not handle non-ascii character properly. I have introduce this when I unified some things in my refactorings. But the base cause is a bug in fulcrum-parser. The version included in Scarab calls a deprecated method URLDecoder.decode() which uses an arbitrary character encoding depending on the settings of the underlying OS. To fix this, I have added a patched version of fulcrum-parser. In the current trunk of fulcrum-parser it's already fixed, but the version is incompatible with the rest of the Turbine-version used in Scarab. "