/**
* Extract a sub-sequence from the sequence given a Range object
*
* @param range
* @param sequence
* @return the sub-sequence, null if the range or sequence is null
*/
public static String extractRangeSequence(Range range, String sequence) {
if (range == null || sequence == null || (sequence != null && sequence.length() == 0)) {
return null;
}
// when the range is valid
if (validateRange(range, sequence).isEmpty()) {
Position pos1 = range.getStart();
Position pos2 = range.getEnd();
// both the start position and the end position have a status
// if both positions are undetermined, or of type 'n-?','n-n', 'c-c' or '?-c', no feature
// sequence can be extracted
// if the start position is negative, it means we cannot extract a sequence for this range
if ((pos1.isPositionUndetermined() && pos2.isPositionUndetermined()) || pos1.getStart() < 0) {
return null;
}
// a feature sequence can be extracted
else {
// the start position is the start position of the start interval
long startSequence = pos1.getStart();
// the end position is the end position of the end interval
long endSequence = pos2.getEnd();
// in case of undetermined, the start position is starting from 1
if (pos1.isPositionUndetermined() || PositionUtils.isLessThan(pos1)) {
startSequence = 1;
}
// in case of greater than, the start position is starting from fromIntervalStart + 1
else if (PositionUtils.isGreaterThan(pos1)) {
startSequence++;
}
// in case of undetermined, the end position is at the end of the sequence
if (pos2.isPositionUndetermined() || PositionUtils.isGreaterThan(pos2)) {
endSequence = sequence.length();
}
// in case of less than, the end position is at 'toIntervalEnd' - 1
else if (PositionUtils.isLessThan(pos2)) {
endSequence--;
}
return sequence.substring(
(int) Math.max(0, startSequence - 1),
(int) endSequence); // we make sure that we don't request index < 0.
}
}
return null;
}
/**
* Checks if the interval positions of the range are overlapping
*
* @param range
* @return true if the range intervals are overlapping
*/
public static boolean areRangePositionsOverlapping(Range range) {
// get the range status
Position start = range.getStart();
Position end = range.getEnd();
long fromStart = start.getStart();
long fromEnd = start.getEnd();
long toStart = end.getStart();
long toEnd = end.getEnd();
// both the end and the start have a specific status
// in the specific case where the start is superior to a position and the end is inferior to
// another position, we need to check that the
// range is not invalid because 'greater than' and 'less than' are both exclusive
if (PositionUtils.isGreaterThan(start)
&& PositionUtils.isLessThan(end)
&& toStart - fromEnd < 2) {
return true;
}
// we have a greater than start position and the end position is equal to the start position
else if (PositionUtils.isGreaterThan(start)
&& !PositionUtils.isGreaterThan(end)
&& fromStart == toEnd) {
return true;
}
// we have a less than end position and the start position is equal to the start position
else if (!PositionUtils.isLessThan(start)
&& PositionUtils.isLessThan(end)
&& fromStart == toEnd) {
return true;
}
// As the range positions are 0 when the status is undetermined, we can only check if the ranges
// are not overlapping when both start and end are not undetermined
else if (!start.isPositionUndetermined() && !end.isPositionUndetermined()) {
return PositionUtils.arePositionsOverlapping(fromStart, fromEnd, toStart, toEnd);
}
return false;
}
/**
* Method to check if the range is valid or not. If the range is valid, the method returns null
* otherwise it returns a message.
*
* @param range : the range to check
* @param sequence : the sequence of the polymer
* @return empty list if the range is within the sequence, coherent with its fuzzy type and not
* overlapping. If the range is not valid, it will return a list of error messages describing
* why the range is invalid
*/
public static List<String> validateRange(Range range, String sequence) {
if (range != null) {
Position start = range.getStart();
Position end = range.getEnd();
List<String> messages = PositionUtils.validateRangePosition(start, sequence);
messages.addAll(PositionUtils.validateRangePosition(end, sequence));
if (areRangeStatusInconsistent(range)) {
messages.add(
"The start status "
+ start.getStatus().getShortName()
+ " and end status "
+ end.getStatus().getShortName()
+ " are inconsistent");
}
if (areRangePositionsOverlapping(range)) {
messages.add(
"The range positions overlap : ("
+ start.getStart()
+ "-"
+ start.getEnd()
+ ") - ("
+ end.getStart()
+ "-"
+ end.getEnd()
+ ")");
}
return messages;
}
return Collections.EMPTY_LIST;
}
