Hello,
i am using the bridge:inputFile component from the 'http://liferay.com/faces/bridge' taglib in a pure UTF-8 environment.
The component seems to have a problem with special characters (im my case umlauts) that are submitted in normal text input fields along the uploaded files.
The problem is, that the
com.liferay.faces.bridge.context.map.RequestParameterMapMultiPartImpl.RequestParameterMapMultiPartImpl(BridgeContext, ClientDataRequest){
...
boolean foundAtLeastOneUploadedFile = false;
if (diskFileItems != null) {
for (DiskFileItem diskFileItem : diskFileItems) {
String fieldName = diskFileItem.getFieldName();
int pos = fieldName.indexOf(namespace);
if (pos >= 0) {
fieldName = fieldName.substring(pos + namespace.length());
}
if (diskFileItem.isFormField()) {
[b]String requestParameterValue = diskFileItem.getString();[/b]
String fixedRequestParameterValue = portletContainer.fixRequestParameterValue(
requestParameterValue);
requestParameterMap.put(fieldName, fixedRequestParameterValue);
logger.debug("{0}=[{1}]", fieldName, fixedRequestParameterValue);
}
else {
...
}
always uses the default charset 'ISO-8859-1' from the DiskFileItem class when getting the parameter value (diskFileItem.getString()). This breaks the contents of the UTF-8 encoded textfields.
A possible solution might be passing the correct request encoding to the overloaded method 'public String getString(final String charset);'.
Best regards,
Max
