supported parsing multiple blobs

[niyaton]

Kenja java parser changes its behaviour depends on program arguments now.
If it gets only one arg (old style):

• read all lines from stdin and regards as source code.
• kenja-java-parser will output a syntax tree file which path was given as a argument.

If it gets two args (new added):

• read lines per line
• each line means hash of blob to parse.
• kenja-java-parser will read the contents of blob by using jGit API.
• the location of git repository is given as first argument.
• the location of syntax trees dir is given as second argument.
• kenja-java-parser automatically make syntax tree files in syntax trees dir and these file name is same as hash of blob.