Fix errors in scripts for dataset creation

data/scripts/createIds.py

@@ -0,0 +1,21 @@
+# A python script that creates 'test.fileids' file
+# for a certain dataset. This file contains the id
+# of each file in the dataset in the form dataset_{id}.
+
+import sys
+
+f = open("test.fileids", 'w')
+
+name = raw_input("Give dataset: ")
+if (name != "radio" and name != "paramythi"):
+    sys.exit("Wrong dataset (should be either radio or paramythi)")
+name = name.strip('\n') + "_"
+
+n = int(input("Give number of examples: "))
+
+for i in range(n):
+    if name == "radio_":
+        f.write(name + format(i, '02d') + '\n')
+    if name == "paramythi_":
+        f.write("Paramythi_horis_onoma_" + format(i, '04d') + '\n')
+f.close()

data/scripts/createText.py

@@ -1,7 +1,6 @@
-#!/bin/bash
-
-# A script that reads test.transcriptions file and write each
-# sentence in 'train-text.txt' file line by line.
+# A python script that reads 'test.transcriptions' file and
+# write each sentence in 'train-text.txt' file line by line.
+# Necessary to build a language model from transciptions.
 
 # Open read file
 r = open("../../test.transciptions", 'r')

data/scripts/merge.sh

@@ -1,5 +1,10 @@
 #!/bin/bash
 
+# A bash script that takes two language models
+# as input and merges them usign SRILM.
+# Usage: ./merge <language model 1> <language model 2>
+
+# Check arguments
 if [ ! -f "$1" ]; then
     echo "First .lm model does not exist"
     exit 0