[SYSTEMDS-3790] Fix transformencode robustness for non-existing columns

This patch fixes endless loops in transformencode, if the tfspec
references columns outside the column range.

Author: mboehm7

src/main/java/org/apache/sysds/runtime/transform/encode/EncoderFactory.java

@@ -127,15 +127,24 @@ public static MultiColumnEncoder createEncoder(String spec, String[] colnames, i
 			// Error out if the first level encoders have overlaps
 			if (intersect(rcIDs, binIDs, haIDs, weIDs, bowIDs))
 				throw new DMLRuntimeException("More than one encoders (recode, binning, hashing, word_embedding, bag_of_words) on one column is not allowed");
-
+			
 			List<Integer> ptIDs = except(UtilFunctions.getSeqList(1, clen, 1), naryUnionDistinct(rcIDs, haIDs, binIDs, weIDs, bowIDs));
-			List<Integer> oIDs = Arrays.asList(ArrayUtils
-				.toObject(TfMetaUtils.parseJsonIDList(jSpec, colnames, TfMethod.OMIT.toString(), minCol, maxCol)));
-			List<Integer> mvIDs = Arrays.asList(ArrayUtils.toObject(
-				TfMetaUtils.parseJsonObjectIDList(jSpec, colnames, TfMethod.IMPUTE.toString(), minCol, maxCol)));
+			List<Integer> oIDs = new ArrayList<>(Arrays.asList(ArrayUtils
+				.toObject(TfMetaUtils.parseJsonIDList(jSpec, colnames, TfMethod.OMIT.toString(), minCol, maxCol))));
+			List<Integer> mvIDs = new ArrayList<>(Arrays.asList(ArrayUtils.toObject(
+				TfMetaUtils.parseJsonObjectIDList(jSpec, colnames, TfMethod.IMPUTE.toString(), minCol, maxCol))));
 			List<Integer> udfIDs = TfMetaUtils.parseUDFColIDs(jSpec, colnames, minCol, maxCol);
 
-
+			// robustness for transformencode specs w/ non-existing columns (so far, endless loops)
+			rcIDs.removeIf(i -> i > clen);
+			ptIDs.removeIf(i -> i > clen);
+			oIDs.removeIf(i -> i > clen);
+			mvIDs.removeIf(i -> i > clen);
+			udfIDs.removeIf(i -> i > clen);
+			binIDs.removeIf(i -> i > clen);
+			weIDs.removeIf(i -> i > clen);
+			bowIDs.removeIf(i -> i > clen);
+			
 			// create individual encoders
 			if(!rcIDs.isEmpty())
 				for(Integer id : rcIDs)

src/test/scripts/functions/builtin/adasynRealData.dml

@@ -28,9 +28,7 @@ tfspec = read($2, data_type="scalar", value_type="string")
 upsample = as.logical($3)
 
 if( tfspec != " " ) {
-  F = M[, 1:ncol(M)] # FIXME
   [X,meta] = transformencode(target=F, spec=tfspec);
-  X = X[,1:ncol(X)-1];
   X = imputeByMode(X);
 }
 else {