diff --git a/.gitignore b/.gitignore
index 7f4a73c..d9e8ed4 100644
--- a/.gitignore
+++ b/.gitignore
@@ -2,8 +2,14 @@
 r?*/
 V?*
 
+# Build files
+/wordfreq
+cmake*
+lib*.h
+
 # tmp files
 *.DMP
 
 # OS caching
 .DS_Store
+.idea
diff --git a/Dockerfile b/Dockerfile
index 1e724c6..87ea1cc 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,10 +1,10 @@
-FROM yottadb/yottadb:r1.20
+FROM yottadb/yottadb:r1.23
 LABEL maintainer="Bart Moelans <bart.moelans@uantwerpen.be>"
 
 ##################################
 # Define global environment vars #
 ##################################
-ENV SRC_DIR=/data \
+ENV SRC_DIR=/root \
     ydb_dist=/opt/yottadb/current
 
 
@@ -14,11 +14,14 @@ ENV SRC_DIR=/data \
 #################################################
 
 RUN apt-get -y update \
-	&& apt-get -y install vim \
+	&& apt-get -y install vim make gcc gdb\
     && ln -s ${ydb_dist}/ydb /usr/local/bin/M \
 	&& echo "source ${ydb_dist}/ydb_env_set" >> /root/.bashrc \
     && echo "export ydb_gbldir=/data/g/sandbox.gld" >> /root/.bashrc
 
+WORKDIR ${SRC_DIR}/cmumps
+
 # by default yottadb entrypoint is ydb itself
-ENTRYPOINT ["/bin/bash"]
-WORKDIR ${SRC_DIR}
+ENTRYPOINT []
+
+
diff --git a/Makefile b/Makefile
new file mode 100644
index 0000000..59d1a4c
--- /dev/null
+++ b/Makefile
@@ -0,0 +1,9 @@
+CC=gcc
+CFLAGS=-g -pedantic -Wall -Wno-unknown-pragmas -std=c99
+YOTTA_INCLUDES=-I $(ydb_dist) -L $(ydb_dist)
+
+wordfreq.so: wordfreq.c
+	$(CC) $(CFLAGS) $(YOTTA_INCLUDES) -shared -fPIC -o $@ $^ -lyottadb
+
+wordfreq: wordfreq.c
+	$(CC) $(CFLAGS) $(YOTTA_INCLUDES) -DTEST -o $@ $^ -lyottadb
\ No newline at end of file
diff --git a/data/r/testc.m b/data/r/testc.m
new file mode 100644
index 0000000..1443f6b
--- /dev/null
+++ b/data/r/testc.m
@@ -0,0 +1,6 @@
+; make wordfreq.so && mumps -run "^testc"
+WFREQ
+ n lenght
+ s word="Antwerpen"
+ d &wordfreq.len(.length,"Antwerpen")
+ w !,length
diff --git a/data/r/wfreq.m b/data/r/wfreq.m
new file mode 100644
index 0000000..847884f
--- /dev/null
+++ b/data/r/wfreq.m
@@ -0,0 +1,8 @@
+; make wordfreq.so && mumps -run "^wfreq"
+WFREQ
+ k ^Test,^index
+ s ^Test("ab")=1
+ s ^Test("ba")=2
+ w $&wordfreq.windex("^Test","^index")
+ w !
+ zwr ^index
diff --git a/docker-compose.yml b/docker-compose.yml
index e5cf669..c3036ad 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -2,7 +2,14 @@ version: "3"
 services: 
   mumps:
     build: .
+    environment:
+      - GTMXC_wordfreq=/root/cmumps/wordfreq.xc
     ports:
       - 8080:8080
     volumes:
+      - .:/root/cmumps
       - ./data/:/data
+    security_opt:
+      - seccomp:unconfined
+    cap_add:
+      - SYS_PTRACE
diff --git a/wordfreq.c b/wordfreq.c
new file mode 100644
index 0000000..4af22b0
--- /dev/null
+++ b/wordfreq.c
@@ -0,0 +1,173 @@
+/****************************************************************
+ *								                                *
+ * Copyright (c) 2017-2018 YottaDB LLC. and/or its subsidiaries.*
+ * All rights reserved.						                    *
+ *								                                *
+ *	This source code contains the intellectual property	        *
+ *	of its copyright holder(s), and is made available	        *
+ *	under a license.  If you do not know the terms of	        *
+ *	the license, please stop and do not read further.	        *
+ *								                                *
+ ****************************************************************/
+
+#include "libyottadb.h"    /* for ydb_* macros/prototypes/typedefs */
+
+#include <sys/types.h>    /* needed for "getpid" */
+#include <unistd.h>    /* needed for "getpid" */
+#include <stdio.h>    /* for "printf" */
+#include <string.h>    /* for "strtok" */
+#include <ctype.h>    /* for "toupper" */
+
+/**
+ * Macro to create/fill-in a ydb_buffer_t structure from ydb_string_t
+ */
+#define YDB_STRING_TO_BUFFER(STRING, BUFFERP)					\
+{										\
+	(BUFFERP)->buf_addr = (STRING)->address;						\
+	(BUFFERP)->len_used = (BUFFERP)->len_alloc = (STRING)->length;	\
+}
+
+/**
+ * Test function to test interaction Mumps to C
+ */
+ydb_status_t len(const int argc, ydb_long_t *length, const ydb_string_t *s) {
+    *length = s->length;
+    return YDB_OK;
+}
+
+void read_input(ydb_string_t *var_name) {
+    ydb_buffer_t tmp_b, index, words;
+    ydb_buffer_t value;
+    char valuebuff[64], linebuff[1024], *lineptr, *ptr, *tmp_ptr;
+    int status;
+
+    YDB_STRING_TO_BUFFER(var_name, &words);
+    if (var_name->address[0] == '^') {
+        ydb_delete_s(&words, 0, NULL, YDB_DEL_TREE);
+    }
+
+    ydb_delete_s(&index, 0, NULL, YDB_DEL_TREE);
+    value.buf_addr = &valuebuff[0];
+    value.len_used = 0;
+    value.len_alloc = sizeof(valuebuff);
+    do {
+        lineptr = fgets(linebuff, sizeof(linebuff), stdin);
+        if (NULL == lineptr)
+            break;
+        ptr = strtok(lineptr, " ");
+        while (NULL != ptr) {    /* convert word pointed to by "ptr" to lowercase */
+            tmp_ptr = ptr;
+            while (*tmp_ptr) {
+                *tmp_ptr = tolower(*tmp_ptr);
+                tmp_ptr++;
+            }
+            /* */
+            tmp_b.buf_addr = ptr;
+            tmp_b.len_used = tmp_ptr - ptr;
+            if (tmp_ptr[-1] == '\n')
+                tmp_b.len_used--;    /* - 1 to remove trailing newline */
+            if (tmp_b.len_used) {
+                tmp_b.len_alloc = tmp_b.len_used;
+                status = ydb_incr_s(&words, 1, &tmp_b, NULL, &value);    /* M line : set value=$incr(words(tmp1)) */
+                YDB_ASSERT(YDB_OK == status);
+            }
+            ptr = strtok(NULL, " ");
+        }
+    } while (1);
+}
+
+ydb_status_t create_word_index(const int argc, ydb_string_t *var_name, ydb_string_t *index_name) {
+    ydb_buffer_t tmp_b, index, words, words_tmp1, null;
+    ydb_buffer_t index_subscr[2];
+    char tmp1buff[64], words_tmp1buff[64];
+    int status;
+    YDB_STRING_TO_BUFFER(var_name, &words);
+    YDB_STRING_TO_BUFFER(index_name, &index);
+
+    tmp_b.buf_addr = tmp1buff;            /* M line : set tmp1="" */
+    tmp_b.len_used = 0;
+    tmp_b.len_alloc = sizeof(tmp1buff);
+    words_tmp1.buf_addr = words_tmp1buff;
+    words_tmp1.len_used = 0;
+    words_tmp1.len_alloc = sizeof(words_tmp1buff);
+    null.buf_addr = NULL;
+    null.len_used = 0;
+    null.len_alloc = 0;
+    do {
+        status = ydb_subscript_next_s(&words, 1, &tmp_b, &tmp_b);    /* M line : set tmp1=$order(words(tmp1)) */
+        if (YDB_ERR_NODEEND == status)
+            break;
+        YDB_ASSERT(YDB_OK == status);
+        YDB_ASSERT(0 != tmp_b.len_used);
+        status = ydb_get_s(&words, 1, &tmp_b, &words_tmp1);    /* M line : set words_tmp1=words(tmp1) */
+        YDB_ASSERT(YDB_OK == status);
+        index_subscr[0] = words_tmp1;
+        index_subscr[1] = tmp_b;
+        status = ydb_set_s(&index, 2, index_subscr, &null);    /* M line : set index(words_tmp1,tmp1)="" */
+        YDB_ASSERT(YDB_OK == status);
+    } while (1);
+    return YDB_OK;
+}
+
+
+/* count and report word frequencies for http://www.cs.duke.edu/csed/code/code2007/ */
+int main() {
+    ydb_string_t *var_name, *index_name;
+
+    var_name = malloc(sizeof(ydb_string_t));
+    var_name->address = malloc(10 * sizeof(char));
+    index_name = malloc(sizeof(ydb_string_t));
+    index_name->address = malloc(10 * sizeof(char));
+    if (1) {
+        strcpy(var_name->address, "words");
+        var_name->length = 5;
+    } else {
+        strcpy(var_name->address, "^words");
+        var_name->length = 6;
+    }
+    strcpy(index_name->address, "index");
+    index_name->length = 5;
+
+    ydb_buffer_t tmp1, tmp2, index, words;
+    ydb_buffer_t index_subscr[2];
+    char tmp1buff[64], tmp2buff[64];
+    int status;
+
+    read_input(var_name);
+    YDB_STRING_TO_BUFFER(var_name, &words);
+    YDB_STRING_TO_BUFFER(index_name, &index);
+
+    create_word_index(2, var_name, index_name);
+
+    tmp1.buf_addr = tmp1buff;            /* M line : set tmp1="" */
+    tmp1.len_used = 0;
+    tmp1.len_alloc = sizeof(tmp1buff);
+
+    do {
+        status = ydb_subscript_previous_s(&index, 1, &tmp1, &tmp1);    /* M line : set tmp1=$order(index(tmp1),-1) */
+        if (YDB_ERR_NODEEND == status)
+            break;
+        YDB_ASSERT(YDB_OK == status);
+        YDB_ASSERT(0 != tmp1.len_used);
+        tmp2.buf_addr = tmp2buff;            /* M line : set tmp2="" */
+        tmp2.len_used = 0;
+        tmp2.len_alloc = sizeof(tmp2buff);
+        index_subscr[0] = tmp1;
+        index_subscr[1] = tmp2;
+        do {
+            status = ydb_subscript_next_s(&index, 2, index_subscr,
+                                          &tmp2); /* M line : set tmp2=$order(index(tmp1,tmp2)) */
+            if (YDB_ERR_NODEEND == status)
+                break;
+            YDB_ASSERT(YDB_OK == status);
+            YDB_ASSERT(0 != tmp2.len_used);
+            tmp1.buf_addr[tmp1.len_used] = '\0';
+            tmp2.buf_addr[tmp2.len_used] = '\0';
+            printf("%s\t%s\n", tmp1.buf_addr, tmp2.buf_addr);
+            index_subscr[1] = tmp2;
+        } while (1);
+    } while (1);
+    free(index_name);
+    free(var_name);
+    return YDB_OK;
+}
\ No newline at end of file
diff --git a/wordfreq.so b/wordfreq.so
new file mode 100755
index 0000000..28b18d3
Binary files /dev/null and b/wordfreq.so differ
diff --git a/wordfreq.xc b/wordfreq.xc
new file mode 100644
index 0000000..4ed5663
--- /dev/null
+++ b/wordfreq.xc
@@ -0,0 +1,3 @@
+/root/cmumps/wordfreq.so
+len: ydb_status_t len(O:ydb_long_t *, I:ydb_string_t *)
+windex: ydb_status_t create_word_index(I:ydb_string_t *, I:ydb_string_t *)
diff --git a/wordfreq_input.txt b/wordfreq_input.txt
new file mode 100644
index 0000000..e5d3534
--- /dev/null
+++ b/wordfreq_input.txt
@@ -0,0 +1 @@
+Lorem ipsum
\ No newline at end of file