Add corenlp.sh for preprocessing

corenlp.sh

@@ -0,0 +1,20 @@
+#!/usr/bin/env bash
+#
+# Runs Stanford CoreNLP.
+# Simple uses for xml and plain text output to files are:
+#    ./corenlp.sh -file filename
+#    ./corenlp.sh -file filename -outputFormat text 
+
+scriptdir=`dirname $0`
+
+# echo java -mx3g -cp \"$scriptdir/*\" edu.stanford.nlp.pipeline.StanfordCoreNLP $*
+
+# $1 - path
+
+PATH=$1
+for FNAME in $PATH/*
+do
+    /usr/bin/java -mx2g -cp "$scriptdir/*" edu.stanford.nlp.pipeline.StanfordCoreNLP -annotators tokenize,ssplit,pos,lemma,ner,parse -ssplit.eolonly -tokenize.whitespace true -file $FNAME
+    # /usr/bin/java -mx2g -cp "$scriptdir/*" edu.stanford.nlp.pipeline.StanfordCoreNLP -annotators tokenize,ssplit,pos,lemma,ner,parse -file $FNAME
+    /bin/mv $(/usr/bin/basename $FNAME.xml) $PATH/
+done