Re: [xwiki-dev] Problems reading some files on (Kubutu) Linux inside eclipse

Dan, While treat source files as UTF-8 encoding,these invalid chars will be reported as warnings in my IntelliJ IDEA 5.0 development env(WIN32). I change the java option from -Dfile.encoding=utf-8 to -Dfile.encoding=gbk( my local charset),it seems rights, and it doesn't limit using utf-8 in xwiki links and document contents. By apply following changs, utf-8 seems work fine * install MySQL 5.0.27, select utf8 as default charset,and do not select STRICT mode on config instance. *found xwiki's config file hibernate.cfg.xml,change connect uri to <property name="connection.url"> jdbc:mysql://localhost/xwiki?useUnicode=true&amp;characterEncoding=utf-8</property> *found tomcat's config file server.xml,add URIEncoding option to Connector's parameter: <Connector port="8080" maxHttpHeaderSize="8192" URIEncoding="UTF-8" maxThreads="150" minSpareThreads="25" maxSpareThreads="75" enableLookups="false" redirectPort="8443" acceptCount="100" connectionTimeout="20000" disableUploadTimeout="true" /> *found xwiki's config file web.xml,change encoding charset from iso-8859-1 to UTF-8.appling this change to xwiki.cfg,velocity.properties too. *and source codes: D:\dev\xwiki\trunk\core\src\main\java\com\xpn\xwiki\web>svn diff Utils.java Index: Utils.java =================================================================== --- Utils.java （修订版 1748） +++ Utils.java （工作拷贝） @@ -393,18 +393,28 @@ } } + public static String FixUtf8(String str) + { + try { + byte b[]=str.getBytes("ISO-8859-1"); + int lp=b.length; + while(--lp>=0 && b[lp]>=0) {}; + if(lp>=0) return new String(b,"UTF-8"); + } + catch(Exception e) + { + } + return str; + } + public static String decode(String name, XWikiContext context) { try { // Make sure + is considered as a space String result = name.replace('+', ' '); - + String ret=URLDecoder.decode(result, context.getWiki().getEncoding( )); // It seems Internet Explorer can send us back UTF-8 // instead of ISO-8859-1 for URLs - if (Base64.isValidUTF8(result.getBytes(), false)) - result = new String(result.getBytes(), "UTF-8"); - - // Still need to decode URLs - return URLDecoder.decode(result, context.getWiki().getEncoding()); + return FixUtf8(ret); } catch (Exception e) { return name; } D:\dev\xwiki\trunk\core\src\main\java\com\xpn\xwiki\web>svn diff XWikiServletRequest.java Index: XWikiServletRequest.java =================================================================== --- XWikiServletRequest.java （修订版 1748） +++ XWikiServletRequest.java （工作拷贝） @@ -88,7 +88,7 @@ } public String getPathInfo() { - return request.getPathInfo(); + return Utils.FixUtf8(request.getPathInfo()); } public String getPathTranslated() { * Thanks Michael for give us detailed method to export utf-8 document as pdf(http://mail-archive.objectweb.org/xwiki-dev/2006-03/msg00030.html). I use this command to get info of charset: java -cp fop-0.20.5.jar org.apache.fop.fonts.apps.TTFReader -ttcname "SimSun" D:\MYFONTS\simsun.ttc simsun.xml Best regards, Jackson Wang 2006/12/14, Dan Murphy &lt;dm.subs(a)googlemail.com&gt;om>: