1. Edit the sakai.properties file to enable unzip file in content.
2. Update the ChineseFile.zip to content tool. The zip file only contain a zero-size file with Chinese file name encoded with GBK (a Chinese charset).
3. Unzip the file in the content tool. A folder named 'ChineseFile' is created as expected, but there's no file extracted.
1. zip doesn't contain the information of how to encode the file name(not content). In my tests, Linux programs encode file names with UTF-8, and Windows encode file names depend on language version (in Chinese version Windows, most, programs use GBK, and 7-zip always using UTF-8)
2. JDK6 seems using ISO-8859-1 to unzip file, and JDK7 seems using UTF-8 to unzip file by default. Neither can handle file with GBK encoded.
3. Good news, JDK7 provides a new constructor ZipFile(File file, int mode, Charset charset) can specified which charset to be used. But this also limits the JDK version.
1. Add a new property content.zip.charset to allow user define which charsets would be tried to extract the zip file one by one. And the default setting is only try UTF-8
To handle the ChineseFile.zip, the following properties may be added.
Patch is attached via 2.9.x