Fix processing report files with Unicode paths (#58)

Jenkins replaces some ASCII symbols like colons with Unicode
storing builds data. Parsing reports letter when users open
'SLOCCount Results' window the plugin unable to retrieve
SLOC information - file was stored by path with Unicode.

It relates to JDK 8.
Using unmarshal(File f) indeed processes not the File but a URL
based on its absolute path. When JDK tries to open connection it
calles ParseUtil.decode(url.getFile()) against our path containing
Unicode. This method converts Unicode back to ASCII. So we get
java.io.FileNotFoundException wrapped in javax.xml.bind.JAXBException,
which is ignored by the plugin. Using unmarshal(java.io.InputStream is)
solves the issue due to JDK uses passed InputStream and all
described here above do not affect us.

src/main/java/hudson/plugins/sloccount/model/cloc/ClocReport.java

@@ -10,7 +10,11 @@
 import javax.xml.bind.annotation.XmlElement;
 import javax.xml.bind.annotation.XmlRootElement;
 import java.io.File;
+import java.io.FileInputStream;
+import java.io.IOException;
+import java.io.InputStream;
 import java.io.Serializable;
+import java.io.UncheckedIOException;
 
 /**
  * Cloc report parser and the parsed file.
@@ -72,7 +76,14 @@ public static ClocReport parse(File file) throws JAXBException {
             currentThread.setContextClassLoader(initialClassLoader);
         }
         Unmarshaller unmarshaller = context.createUnmarshaller();
-        return (ClocReport) unmarshaller.unmarshal(file);
+        // Jenkins replaces ASCII symbols like colons found in filenames to Unicode.
+        // Using File() here causes converting Unicode symbols back to ASCII.
+        // Such files could not be found of course.
+        try (InputStream is = new FileInputStream(file)) {
+            return (ClocReport) unmarshaller.unmarshal(is);
+        } catch (IOException e) {
+            throw new UncheckedIOException(e);
+        }
     }
 
     /**