*/
package jalview.util;
-import java.util.*;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.List;
+
+import jalview.bin.Cache;
/**
- * MapList Simple way of bijectively mapping a non-contiguous linear range to
- * another non-contiguous linear range Use at your own risk! TODO: efficient
- * implementation of private posMap method TODO: test/ensure that sense of from
- * and to ratio start position is conserved (codon start position recovery)
- * TODO: optimize to use int[][] arrays rather than vectors.
+ * A simple way of bijectively mapping a non-contiguous linear range to another
+ * non-contiguous linear range.
+ *
+ * Use at your own risk!
+ *
+ * TODO: efficient implementation of private posMap method
+ *
+ * TODO: test/ensure that sense of from and to ratio start position is conserved
+ * (codon start position recovery)
*/
public class MapList
{
+
/*
- * (non-Javadoc)
- *
- * @see java.lang.Object#equals(java.lang.Object)
+ * Subregions (base 1) described as { [start1, end1], [start2, end2], ...}
+ */
+ private List<int[]> fromShifts;
+
+ /*
+ * Same format as fromShifts, for the 'mapped to' sequence
+ */
+ private List<int[]> toShifts;
+
+ /*
+ * number of steps in fromShifts to one toRatio unit
+ */
+ private int fromRatio;
+
+ /*
+ * number of steps in toShifts to one fromRatio
*/
- public boolean equals(MapList obj)
+ private int toRatio;
+
+ /*
+ * lowest and highest value in the from Map
+ */
+ private int fromLowest;
+
+ private int fromHighest;
+
+ /*
+ * lowest and highest value in the to Map
+ */
+ private int toLowest;
+
+ private int toHighest;
+
+ /**
+ * Constructor
+ */
+ public MapList()
+ {
+ fromShifts = new ArrayList<>();
+ toShifts = new ArrayList<>();
+ }
+
+ /**
+ * Two MapList objects are equal if they are the same object, or they both
+ * have populated shift ranges and all values are the same.
+ */
+ @Override
+ public boolean equals(Object o)
{
+ if (o == null || !(o instanceof MapList))
+ {
+ return false;
+ }
+
+ MapList obj = (MapList) o;
if (obj == this)
+ {
return true;
- if (obj != null && obj.fromRatio == fromRatio && obj.toRatio == toRatio
- && obj.fromShifts != null && obj.toShifts != null)
- {
- int i, iSize = fromShifts.size(), j, jSize = obj.fromShifts.size();
- if (iSize != jSize)
- return false;
- for (i = 0, iSize = fromShifts.size(), j = 0, jSize = obj.fromShifts
- .size(); i < iSize;)
- {
- int[] mi = (int[]) fromShifts.elementAt(i++);
- int[] mj = (int[]) obj.fromShifts.elementAt(j++);
- if (mi[0] != mj[0] || mi[1] != mj[1])
- return false;
- }
- iSize = toShifts.size();
- jSize = obj.toShifts.size();
- if (iSize != jSize)
- return false;
- for (i = 0, j = 0; i < iSize;)
- {
- int[] mi = (int[]) toShifts.elementAt(i++);
- int[] mj = (int[]) obj.toShifts.elementAt(j++);
- if (mi[0] != mj[0] || mi[1] != mj[1])
- return false;
- }
- return true;
}
- return false;
+ if (obj.fromRatio != fromRatio || obj.toRatio != toRatio
+ || obj.fromShifts == null || obj.toShifts == null)
+ {
+ return false;
+ }
+ return Arrays.deepEquals(fromShifts.toArray(), obj.fromShifts.toArray())
+ && Arrays.deepEquals(toShifts.toArray(),
+ obj.toShifts.toArray());
}
- public Vector fromShifts;
-
- public Vector toShifts;
-
- int fromRatio; // number of steps in fromShifts to one toRatio unit
+ /**
+ * Returns a hashcode made from the fromRatio, toRatio, and from/to ranges
+ */
+ @Override
+ public int hashCode()
+ {
+ int hashCode = 31 * fromRatio;
+ hashCode = 31 * hashCode + toRatio;
+ for (int[] shift : fromShifts)
+ {
+ hashCode = 31 * hashCode + shift[0];
+ hashCode = 31 * hashCode + shift[1];
+ }
+ for (int[] shift : toShifts)
+ {
+ hashCode = 31 * hashCode + shift[0];
+ hashCode = 31 * hashCode + shift[1];
+ }
- int toRatio; // number of steps in toShifts to one fromRatio
+ return hashCode;
+ }
/**
+ * Returns the 'from' ranges as {[start1, end1], [start2, end2], ...}
*
- * @return series of intervals mapped in from
+ * @return
*/
- public int[] getFromRanges()
+ public List<int[]> getFromRanges()
{
- return getRanges(fromShifts);
+ return fromShifts;
}
- public int[] getToRanges()
+ /**
+ * Returns the 'to' ranges as {[start1, end1], [start2, end2], ...}
+ *
+ * @return
+ */
+ public List<int[]> getToRanges()
{
- return getRanges(toShifts);
+ return toShifts;
}
- private int[] getRanges(Vector shifts)
+ /**
+ * Flattens a list of [start, end] into a single [start1, end1, start2,
+ * end2,...] array.
+ *
+ * @param shifts
+ * @return
+ */
+ protected static int[] getRanges(List<int[]> shifts)
{
int[] rnges = new int[2 * shifts.size()];
- Enumeration e = shifts.elements();
int i = 0;
- while (e.hasMoreElements())
+ for (int[] r : shifts)
{
- int r[] = (int[]) e.nextElement();
rnges[i++] = r[0];
rnges[i++] = r[1];
}
}
/**
- * lowest and highest value in the from Map
- */
- int[] fromRange = null;
-
- /**
- * lowest and highest value in the to Map
- */
- int[] toRange = null;
-
- /**
*
* @return length of mapped phrase in from
*/
public int getFromLowest()
{
- return fromRange[0];
+ return fromLowest;
}
public int getFromHighest()
{
- return fromRange[1];
+ return fromHighest;
}
public int getToLowest()
{
- return toRange[0];
+ return toLowest;
}
public int getToHighest()
{
- return toRange[1];
- }
-
- private void ensureRange(int[] limits, int pos)
- {
- if (limits[0] > pos)
- limits[0] = pos;
- if (limits[1] < pos)
- limits[1] = pos;
+ return toHighest;
}
+ /**
+ * Constructor given from and to ranges as [start1, end1, start2, end2,...].
+ * There is no validation check that the ranges do not overlap each other.
+ *
+ * @param from
+ * contiguous regions as [start1, end1, start2, end2, ...]
+ * @param to
+ * same format as 'from'
+ * @param fromRatio
+ * phrase length in 'from' (e.g. 3 for dna)
+ * @param toRatio
+ * phrase length in 'to' (e.g. 1 for protein)
+ */
public MapList(int from[], int to[], int fromRatio, int toRatio)
{
- fromRange = new int[]
- { from[0], from[1] };
- toRange = new int[]
- { to[0], to[1] };
+ this();
+ this.fromRatio = fromRatio;
+ this.toRatio = toRatio;
+ fromLowest = Integer.MAX_VALUE;
+ fromHighest = Integer.MIN_VALUE;
- fromShifts = new Vector();
for (int i = 0; i < from.length; i += 2)
{
- ensureRange(fromRange, from[i]);
- ensureRange(fromRange, from[i + 1]);
-
- fromShifts.addElement(new int[]
- { from[i], from[i + 1] });
+ /*
+ * note lowest and highest values - bearing in mind the
+ * direction may be reversed
+ */
+ fromLowest = Math.min(fromLowest, Math.min(from[i], from[i + 1]));
+ fromHighest = Math.max(fromHighest, Math.max(from[i], from[i + 1]));
+ fromShifts.add(new int[] { from[i], from[i + 1] });
}
- toShifts = new Vector();
+
+ toLowest = Integer.MAX_VALUE;
+ toHighest = Integer.MIN_VALUE;
for (int i = 0; i < to.length; i += 2)
{
- ensureRange(toRange, to[i]);
- ensureRange(toRange, to[i + 1]);
- toShifts.addElement(new int[]
- { to[i], to[i + 1] });
+ toLowest = Math.min(toLowest, Math.min(to[i], to[i + 1]));
+ toHighest = Math.max(toHighest, Math.max(to[i], to[i + 1]));
+ toShifts.add(new int[] { to[i], to[i + 1] });
}
- this.fromRatio = fromRatio;
- this.toRatio = toRatio;
}
+ /**
+ * Copy constructor. Creates an identical mapping.
+ *
+ * @param map
+ */
public MapList(MapList map)
{
- this.fromRange = new int[]
- { map.fromRange[0], map.fromRange[1] };
- this.toRange = new int[]
- { map.toRange[0], map.toRange[1] };
+ this();
+ // TODO not used - remove?
+ this.fromLowest = map.fromLowest;
+ this.fromHighest = map.fromHighest;
+ this.toLowest = map.toLowest;
+ this.toHighest = map.toHighest;
+
this.fromRatio = map.fromRatio;
this.toRatio = map.toRatio;
if (map.fromShifts != null)
{
- this.fromShifts = new Vector();
- Enumeration e = map.fromShifts.elements();
- while (e.hasMoreElements())
+ for (int[] r : map.fromShifts)
{
- int[] el = (int[]) e.nextElement();
- fromShifts.addElement(new int[]
- { el[0], el[1] });
+ fromShifts.add(new int[] { r[0], r[1] });
}
}
if (map.toShifts != null)
{
- this.toShifts = new Vector();
- Enumeration e = map.toShifts.elements();
- while (e.hasMoreElements())
+ for (int[] r : map.toShifts)
+ {
+ toShifts.add(new int[] { r[0], r[1] });
+ }
+ }
+ }
+
+ /**
+ * Constructor given ranges as lists of [start, end] positions. There is no
+ * validation check that the ranges do not overlap each other.
+ *
+ * @param fromRange
+ * @param toRange
+ * @param fromRatio
+ * @param toRatio
+ */
+ public MapList(List<int[]> fromRange, List<int[]> toRange, int fromRatio,
+ int toRatio)
+ {
+ this();
+ fromRange = coalesceRanges(fromRange);
+ toRange = coalesceRanges(toRange);
+ this.fromShifts = fromRange;
+ this.toShifts = toRange;
+ this.fromRatio = fromRatio;
+ this.toRatio = toRatio;
+
+ fromLowest = Integer.MAX_VALUE;
+ fromHighest = Integer.MIN_VALUE;
+ for (int[] range : fromRange)
+ {
+ if (range.length != 2)
+ {
+ // throw new IllegalArgumentException(range);
+ Cache.log.error("Invalid format for fromRange "
+ + Arrays.toString(range) + " may cause errors");
+ }
+ fromLowest = Math.min(fromLowest, Math.min(range[0], range[1]));
+ fromHighest = Math.max(fromHighest, Math.max(range[0], range[1]));
+ }
+
+ toLowest = Integer.MAX_VALUE;
+ toHighest = Integer.MIN_VALUE;
+ for (int[] range : toRange)
+ {
+ if (range.length != 2)
+ {
+ // throw new IllegalArgumentException(range);
+ Cache.log.error("Invalid format for toRange "
+ + Arrays.toString(range) + " may cause errors");
+ }
+ toLowest = Math.min(toLowest, Math.min(range[0], range[1]));
+ toHighest = Math.max(toHighest, Math.max(range[0], range[1]));
+ }
+ }
+
+ /**
+ * Consolidates a list of ranges so that any contiguous ranges are merged.
+ * This assumes the ranges are already in start order (does not sort them).
+ * <p>
+ * The main use case for this method is when mapping cDNA sequence to its
+ * protein product, based on CDS feature ranges which derive from spliced
+ * exons, but are contiguous on the cDNA sequence. For example
+ * <pre>
+ * CDS 1-20 // from exon1
+ * CDS 21-35 // from exon2
+ * CDS 36-71 // from exon3
+ * 'coalesce' to range 1-71
+ * </pre>
+ *
+ * @param ranges
+ * @return the same list (if unchanged), else a new merged list, leaving the
+ * input list unchanged
+ */
+ public static List<int[]> coalesceRanges(final List<int[]> ranges)
+ {
+ if (ranges == null || ranges.size() < 2)
+ {
+ return ranges;
+ }
+
+ boolean changed = false;
+ List<int[]> merged = new ArrayList<>();
+ int[] lastRange = ranges.get(0);
+ int lastDirection = lastRange[1] >= lastRange[0] ? 1 : -1;
+ lastRange = new int[] { lastRange[0], lastRange[1] };
+ merged.add(lastRange);
+ boolean first = true;
+
+ for (final int[] range : ranges)
+ {
+ if (first)
+ {
+ first = false;
+ continue;
+ }
+
+ int direction = range[1] >= range[0] ? 1 : -1;
+
+ /*
+ * if next range is in the same direction as last and contiguous,
+ * just update the end position of the last range
+ */
+ boolean sameDirection = range[1] == range[0]
+ || direction == lastDirection;
+ boolean extending = range[0] == lastRange[1] + lastDirection;
+ if (sameDirection && extending)
+ {
+ lastRange[1] = range[1];
+ changed = true;
+ }
+ else
{
- int[] el = (int[]) e.nextElement();
- toShifts.addElement(new int[]
- { el[0], el[1] });
+ lastRange = new int[] { range[0], range[1] };
+ merged.add(lastRange);
+ // careful: merging [5, 5] after [7, 6] should keep negative direction
+ lastDirection = (range[1] == range[0]) ? lastDirection : direction;
}
}
+
+ return changed ? merged : ranges;
}
/**
* @return int[][] { int[] { fromStart, fromFinish, toStart, toFinish }, int
* [fromFinish-fromStart+2] { toStart..toFinish mappings}}
*/
- public int[][] makeFromMap()
+ protected int[][] makeFromMap()
{
+ // TODO not used - remove??
return posMap(fromShifts, fromRatio, toShifts, toRatio);
}
*
* @return int[to position]=position mapped in from
*/
- public int[][] makeToMap()
+ protected int[][] makeToMap()
{
+ // TODO not used - remove??
return posMap(toShifts, toRatio, fromShifts, fromRatio);
}
/**
* construct an int map for intervals in intVals
*
- * @param intVals
+ * @param shiftTo
* @return int[] { from, to pos in range }, int[range.to-range.from+1]
* returning mapped position
*/
- private int[][] posMap(Vector intVals, int ratio, Vector toIntVals,
- int toRatio)
+ private int[][] posMap(List<int[]> shiftTo, int ratio,
+ List<int[]> shiftFrom, int toRatio)
{
- int iv = 0, ivSize = intVals.size();
+ // TODO not used - remove??
+ int iv = 0, ivSize = shiftTo.size();
if (iv >= ivSize)
{
return null;
}
- int[] intv = (int[]) intVals.elementAt(iv++);
+ int[] intv = shiftTo.get(iv++);
int from = intv[0], to = intv[1];
if (from > to)
{
}
while (iv < ivSize)
{
- intv = (int[]) intVals.elementAt(iv++);
+ intv = shiftTo.get(iv++);
if (intv[0] < from)
{
from = intv[0];
int mp[][] = new int[to - from + 2][];
for (int i = 0; i < mp.length; i++)
{
- int[] m = shift(i + from, intVals, ratio, toIntVals, toRatio);
+ int[] m = shift(i + from, shiftTo, ratio, shiftFrom, toRatio);
if (m != null)
{
if (i == 0)
}
mp[i] = m;
}
- int[][] map = new int[][]
- { new int[]
- { from, to, tF, tT }, new int[to - from + 2] };
+ int[][] map = new int[][] { new int[] { from, to, tF, tT },
+ new int[to - from + 2] };
map[0][2] = tF;
map[0][3] = tT;
* shifts.insertElementAt(new int[] { pos, shift}, sidx); else
* rshift[1]+=shift; }
*/
+
/**
* shift from pos to To(pos)
*
/**
*
- * @param fromShifts
+ * @param shiftTo
* @param fromRatio
- * @param toShifts
+ * @param shiftFrom
* @param toRatio
* @return
*/
- private int[] shift(int pos, Vector fromShifts, int fromRatio,
- Vector toShifts, int toRatio)
+ protected static int[] shift(int pos, List<int[]> shiftTo, int fromRatio,
+ List<int[]> shiftFrom, int toRatio)
{
- int[] fromCount = countPos(fromShifts, pos);
+ // TODO: javadoc; tests
+ int[] fromCount = countPos(shiftTo, pos);
if (fromCount == null)
{
return null;
}
int fromRemainder = (fromCount[0] - 1) % fromRatio;
int toCount = 1 + (((fromCount[0] - 1) / fromRatio) * toRatio);
- int[] toPos = countToPos(toShifts, toCount);
+ int[] toPos = countToPos(shiftFrom, toCount);
if (toPos == null)
{
return null; // throw new Error("Bad Mapping!");
}
// System.out.println(fromCount[0]+" "+fromCount[1]+" "+toCount);
- return new int[]
- { toPos[0], fromRemainder, toPos[1] };
+ return new int[] { toPos[0], fromRemainder, toPos[1] };
}
/**
* count how many positions pos is along the series of intervals.
*
- * @param intVals
+ * @param shiftTo
* @param pos
* @return number of positions or null if pos is not within intervals
*/
- private int[] countPos(Vector intVals, int pos)
+ protected static int[] countPos(List<int[]> shiftTo, int pos)
{
- int count = 0, intv[], iv = 0, ivSize = intVals.size();
+ int count = 0, intv[], iv = 0, ivSize = shiftTo.size();
while (iv < ivSize)
{
- intv = (int[]) intVals.elementAt(iv++);
+ intv = shiftTo.get(iv++);
if (intv[0] <= intv[1])
{
if (pos >= intv[0] && pos <= intv[1])
{
- return new int[]
- { count + pos - intv[0] + 1, +1 };
+ return new int[] { count + pos - intv[0] + 1, +1 };
}
else
{
{
if (pos >= intv[1] && pos <= intv[0])
{
- return new int[]
- { count + intv[0] - pos + 1, -1 };
+ return new int[] { count + intv[0] - pos + 1, -1 };
}
else
{
/**
* count out pos positions into a series of intervals and return the position
*
- * @param intVals
+ * @param shiftFrom
* @param pos
* @return position pos in interval set
*/
- private int[] countToPos(Vector intVals, int pos)
+ protected static int[] countToPos(List<int[]> shiftFrom, int pos)
{
- int count = 0, diff = 0, iv = 0, ivSize = intVals.size(), intv[] =
- { 0, 0 };
+ int count = 0, diff = 0, iv = 0, ivSize = shiftFrom.size();
+ int[] intv = { 0, 0 };
while (iv < ivSize)
{
- intv = (int[]) intVals.elementAt(iv++);
+ intv = shiftFrom.get(iv++);
diff = intv[1] - intv[0];
if (diff >= 0)
{
if (pos <= count + 1 + diff)
{
- return new int[]
- { pos - count - 1 + intv[0], +1 };
+ return new int[] { pos - count - 1 + intv[0], +1 };
}
else
{
{
if (pos <= count + 1 - diff)
{
- return new int[]
- { intv[0] - (pos - count - 1), -1 };
+ return new int[] { intv[0] - (pos - count - 1), -1 };
}
else
{
* find series of intervals mapping from start-end in the From map.
*
* @param start
- * position in to map
+ * position mapped 'to'
* @param end
- * position in to map
- * @return series of ranges in from map
+ * position mapped 'to'
+ * @return series of [start, end] ranges in sequence mapped 'from'
*/
public int[] locateInFrom(int start, int end)
{
// inefficient implementation
int fromStart[] = shiftTo(start);
- int fromEnd[] = shiftTo(end); // needs to be inclusive of end of symbol
- // position
- if (fromStart == null || fromEnd == null)
- return null;
- int iv[] = getIntervals(fromShifts, fromStart, fromEnd, fromRatio);
- return iv;
+ // needs to be inclusive of end of symbol position
+ int fromEnd[] = shiftTo(end);
+
+ return getIntervals(fromShifts, fromStart, fromEnd, fromRatio);
}
/**
* find series of intervals mapping from start-end in the to map.
*
* @param start
- * position in from map
+ * position mapped 'from'
* @param end
- * position in from map
- * @return series of ranges in to map
+ * position mapped 'from'
+ * @return series of [start, end] ranges in sequence mapped 'to'
*/
public int[] locateInTo(int start, int end)
{
- // inefficient implementation
int toStart[] = shiftFrom(start);
int toEnd[] = shiftFrom(end);
- if (toStart == null || toEnd == null)
- return null;
- int iv[] = getIntervals(toShifts, toStart, toEnd, toRatio);
- return iv;
+ return getIntervals(toShifts, toStart, toEnd, toRatio);
}
/**
* like shift - except returns the intervals in the given vector of shifts
* which were spanned in traversing fromStart to fromEnd
*
- * @param fromShifts2
+ * @param shiftFrom
* @param fromStart
* @param fromEnd
* @param fromRatio2
* @return series of from,to intervals from from first position of starting
* region to final position of ending region inclusive
*/
- private int[] getIntervals(Vector fromShifts2, int[] fromStart,
- int[] fromEnd, int fromRatio2)
+ protected static int[] getIntervals(List<int[]> shiftFrom,
+ int[] fromStart, int[] fromEnd, int fromRatio2)
{
+ if (fromStart == null || fromEnd == null)
+ {
+ return null;
+ }
int startpos, endpos;
startpos = fromStart[0]; // first position in fromStart
endpos = fromEnd[0]; // last position in fromEnd
int endindx = (fromRatio2 - 1); // additional positions to get to last
// position from endpos
- int intv = 0, intvSize = fromShifts2.size();
+ int intv = 0, intvSize = shiftFrom.size();
int iv[], i = 0, fs = -1, fe_s = -1, fe = -1; // containing intervals
// search intervals to locate ones containing startpos and count endindx
// positions on from endpos
while (intv < intvSize && (fs == -1 || fe == -1))
{
- iv = (int[]) fromShifts2.elementAt(intv++);
+ iv = shiftFrom.get(intv++);
if (fe_s > -1)
{
endpos = iv[0]; // start counting from beginning of interval
i++;
}
if (fs == fe && fe == -1)
+ {
return null;
- Vector ranges = new Vector();
+ }
+ List<int[]> ranges = new ArrayList<>();
if (fs <= fe)
{
intv = fs;
i = fs;
// truncate initial interval
- iv = (int[]) fromShifts2.elementAt(intv++);
- iv = new int[]
- { iv[0], iv[1] };// clone
+ iv = shiftFrom.get(intv++);
+ iv = new int[] { iv[0], iv[1] };// clone
if (i == fs)
+ {
iv[0] = startpos;
+ }
while (i != fe)
{
- ranges.addElement(iv); // add initial range
- iv = (int[]) fromShifts2.elementAt(intv++); // get next interval
- iv = new int[]
- { iv[0], iv[1] };// clone
+ ranges.add(iv); // add initial range
+ iv = shiftFrom.get(intv++); // get next interval
+ iv = new int[] { iv[0], iv[1] };// clone
i++;
}
if (i == fe)
+ {
iv[1] = endpos;
- ranges.addElement(iv); // add only - or final range
+ }
+ ranges.add(iv); // add only - or final range
}
else
{
// walk from end of interval.
- i = fromShifts2.size() - 1;
+ i = shiftFrom.size() - 1;
while (i > fs)
{
i--;
}
- iv = (int[]) fromShifts2.elementAt(i);
- iv = new int[]
- { iv[1], iv[0] };// reverse and clone
+ iv = shiftFrom.get(i);
+ iv = new int[] { iv[1], iv[0] };// reverse and clone
// truncate initial interval
if (i == fs)
{
}
while (--i != fe)
{ // fix apparent logic bug when fe==-1
- ranges.addElement(iv); // add (truncated) reversed interval
- iv = (int[]) fromShifts2.elementAt(i);
- iv = new int[]
- { iv[1], iv[0] }; // reverse and clone
+ ranges.add(iv); // add (truncated) reversed interval
+ iv = shiftFrom.get(i);
+ iv = new int[] { iv[1], iv[0] }; // reverse and clone
}
if (i == fe)
{
// interval is already reversed
iv[1] = endpos;
}
- ranges.addElement(iv); // add only - or final range
+ ranges.add(iv); // add only - or final range
}
// create array of start end intervals.
int[] range = null;
i = 0;
while (intv < intvSize)
{
- iv = (int[]) ranges.elementAt(intv);
+ iv = ranges.get(intv);
range[i++] = iv[0];
range[i++] = iv[1];
- ranges.setElementAt(null, intv++); // remove
+ ranges.set(intv++, null); // remove
}
}
return range;
*/
public int getToPosition(int mpos)
{
+ // TODO not used - remove??
int[] mp = shiftTo(mpos);
if (mp != null)
{
int[] mp = shiftTo(mpos);
if (mp != null)
{
- return new int[]
- { mp[0], mp[0] + mp[2] * (getFromRatio() - 1) };
+ return new int[] { mp[0], mp[0] + mp[2] * (getFromRatio() - 1) };
}
return null;
}
*/
public int getMappedPosition(int pos)
{
+ // TODO not used - remove??
int[] mp = shiftFrom(pos);
if (mp != null)
{
public int[] getMappedWord(int pos)
{
+ // TODO not used - remove??
int[] mp = shiftFrom(pos);
if (mp != null)
{
- return new int[]
- { mp[0], mp[0] + mp[2] * (getToRatio() - 1) };
+ return new int[] { mp[0], mp[0] + mp[2] * (getToRatio() - 1) };
}
return null;
}
/**
- * test routine. not incremental.
*
- * @param ml
- * @param fromS
- * @param fromE
+ * @return a MapList whose From range is this maplist's To Range, and vice
+ * versa
*/
- public static void testMap(MapList ml, int fromS, int fromE)
+ public MapList getInverse()
{
- for (int from = 1; from <= 25; from++)
+ return new MapList(getToRanges(), getFromRanges(), getToRatio(),
+ getFromRatio());
+ }
+
+ /**
+ * test for containment rather than equivalence to another mapping
+ *
+ * @param map
+ * to be tested for containment
+ * @return true if local or mapped range map contains or is contained by this
+ * mapping
+ */
+ public boolean containsEither(boolean local, MapList map)
+ {
+ // TODO not used - remove?
+ if (local)
{
- int[] too = ml.shiftFrom(from);
- System.out.print("ShiftFrom(" + from + ")==");
- if (too == null)
- {
- System.out.print("NaN\n");
- }
- else
- {
- System.out.print(too[0] + " % " + too[1] + " (" + too[2] + ")");
- System.out.print("\t+--+\t");
- int[] toofrom = ml.shiftTo(too[0]);
- if (toofrom != null)
- {
- if (toofrom[0] != from)
- {
- System.err.println("Mapping not reflexive:" + from + " "
- + too[0] + "->" + toofrom[0]);
- }
- System.out.println("ShiftTo(" + too[0] + ")==" + toofrom[0]
- + " % " + toofrom[1] + " (" + toofrom[2] + ")");
- }
- else
- {
- System.out.println("ShiftTo(" + too[0] + ")=="
- + "NaN! - not Bijective Mapping!");
- }
- }
+ return ((getFromLowest() >= map.getFromLowest()
+ && getFromHighest() <= map.getFromHighest())
+ || (getFromLowest() <= map.getFromLowest()
+ && getFromHighest() >= map.getFromHighest()));
}
- int mmap[][] = ml.makeFromMap();
- System.out.println("FromMap : (" + mmap[0][0] + " " + mmap[0][1] + " "
- + mmap[0][2] + " " + mmap[0][3] + " ");
- for (int i = 1; i <= mmap[1].length; i++)
+ else
{
- if (mmap[1][i - 1] == -1)
- {
- System.out.print(i + "=XXX");
+ return ((getToLowest() >= map.getToLowest()
+ && getToHighest() <= map.getToHighest())
+ || (getToLowest() <= map.getToLowest()
+ && getToHighest() >= map.getToHighest()));
+ }
+ }
- }
- else
- {
- System.out.print(i + "=" + (mmap[0][2] + mmap[1][i - 1]));
- }
- if (i % 20 == 0)
- {
- System.out.print("\n");
- }
- else
- {
- System.out.print(",");
- }
+ /**
+ * String representation - for debugging, not guaranteed not to change
+ */
+ @Override
+ public String toString()
+ {
+ StringBuilder sb = new StringBuilder(64);
+ sb.append("[");
+ for (int[] shift : fromShifts)
+ {
+ sb.append(" ").append(Arrays.toString(shift));
}
- // test range function
- System.out.print("\nTest locateInFrom\n");
+ sb.append(" ] ");
+ sb.append(fromRatio).append(":").append(toRatio);
+ sb.append(" to [");
+ for (int[] shift : toShifts)
{
- int f = mmap[0][2], t = mmap[0][3];
- while (f <= t)
- {
- System.out.println("Range " + f + " to " + t);
- int rng[] = ml.locateInFrom(f, t);
- if (rng != null)
- {
- for (int i = 0; i < rng.length; i++)
- {
- System.out.print(rng[i] + ((i % 2 == 0) ? "," : ";"));
- }
- }
- else
- {
- System.out.println("No range!");
- }
- System.out.print("\nReversed\n");
- rng = ml.locateInFrom(t, f);
- if (rng != null)
- {
- for (int i = 0; i < rng.length; i++)
- {
- System.out.print(rng[i] + ((i % 2 == 0) ? "," : ";"));
- }
- }
- else
- {
- System.out.println("No range!");
- }
- System.out.print("\n");
- f++;
- t--;
- }
+ sb.append(" ").append(Arrays.toString(shift));
}
- System.out.print("\n");
- mmap = ml.makeToMap();
- System.out.println("ToMap : (" + mmap[0][0] + " " + mmap[0][1] + " "
- + mmap[0][2] + " " + mmap[0][3] + " ");
- for (int i = 1; i <= mmap[1].length; i++)
+ sb.append(" ]");
+ return sb.toString();
+ }
+
+ /**
+ * Extend this map list by adding the given map's ranges. There is no
+ * validation check that the ranges do not overlap existing ranges (or each
+ * other), but contiguous ranges are merged.
+ *
+ * @param map
+ */
+ public void addMapList(MapList map)
+ {
+ if (this.equals(map))
{
- if (mmap[1][i - 1] == -1)
- {
- System.out.print(i + "=XXX");
+ return;
+ }
+ this.fromLowest = Math.min(fromLowest, map.fromLowest);
+ this.toLowest = Math.min(toLowest, map.toLowest);
+ this.fromHighest = Math.max(fromHighest, map.fromHighest);
+ this.toHighest = Math.max(toHighest, map.toHighest);
- }
- else
- {
- System.out.print(i + "=" + (mmap[0][2] + mmap[1][i - 1]));
- }
- if (i % 20 == 0)
- {
- System.out.print("\n");
- }
- else
- {
- System.out.print(",");
- }
+ for (int[] range : map.getFromRanges())
+ {
+ addRange(range, fromShifts);
}
- System.out.print("\n");
- // test range function
- System.out.print("\nTest locateInTo\n");
+ for (int[] range : map.getToRanges())
{
- int f = mmap[0][2], t = mmap[0][3];
- while (f <= t)
- {
- System.out.println("Range " + f + " to " + t);
- int rng[] = ml.locateInTo(f, t);
- if (rng != null)
- {
- for (int i = 0; i < rng.length; i++)
- {
- System.out.print(rng[i] + ((i % 2 == 0) ? "," : ";"));
- }
- }
- else
- {
- System.out.println("No range!");
- }
- System.out.print("\nReversed\n");
- rng = ml.locateInTo(t, f);
- if (rng != null)
- {
- for (int i = 0; i < rng.length; i++)
- {
- System.out.print(rng[i] + ((i % 2 == 0) ? "," : ";"));
- }
- }
- else
- {
- System.out.println("No range!");
- }
- f++;
- t--;
- System.out.print("\n");
- }
- }
-
- }
-
- public static void main(String argv[])
- {
- MapList ml = new MapList(new int[]
- { 1, 5, 10, 15, 25, 20 }, new int[]
- { 51, 1 }, 1, 3);
- MapList ml1 = new MapList(new int[]
- { 1, 3, 17, 4 }, new int[]
- { 51, 1 }, 1, 3);
- MapList ml2 = new MapList(new int[]
- { 1, 60 }, new int[]
- { 1, 20 }, 3, 1);
- // test internal consistency
- int to[] = new int[51];
- MapList.testMap(ml, 1, 60);
- MapList mldna = new MapList(new int[]
- { 2, 2, 6, 8, 12, 16 }, new int[]
- { 1, 3 }, 3, 1);
- int[] frm = mldna.locateInFrom(1, 1);
- testLocateFrom(mldna, 1, 1, new int[]
- { 2, 2, 6, 7 });
- MapList.testMap(mldna, 1, 3);
- /*
- * for (int from=1; from<=51; from++) { int[] too=ml.shiftTo(from); int[]
- * toofrom=ml.shiftFrom(too[0]);
- * System.out.println("ShiftFrom("+from+")=="+too[0]+" %
- * "+too[1]+"\t+-+\tShiftTo("+too[0]+")=="+toofrom[0]+" % "+toofrom[1]); }
- */
- System.out.print("Success?\n"); // if we get here - something must be
- // working!
+ addRange(range, toShifts);
+ }
}
- private static void testLocateFrom(MapList mldna, int i, int j, int[] ks)
+ /**
+ * Adds the given range to a list of ranges. If the new range just extends
+ * existing ranges, the current endpoint is updated instead.
+ *
+ * @param range
+ * @param addTo
+ */
+ static void addRange(int[] range, List<int[]> addTo)
{
- int[] frm = mldna.locateInFrom(i, j);
- if (frm == ks || java.util.Arrays.equals(frm, ks))
+ /*
+ * list is empty - add to it!
+ */
+ if (addTo.size() == 0)
{
- System.out.println("Success test locate from " + i + " to " + j);
+ addTo.add(range);
+ return;
}
- else
+
+ int[] last = addTo.get(addTo.size() - 1);
+ boolean lastForward = last[1] >= last[0];
+ boolean newForward = range[1] >= range[0];
+
+ /*
+ * contiguous range in the same direction - just update endpoint
+ */
+ if (lastForward == newForward && last[1] == range[0])
{
- System.err.println("Failed test locate from " + i + " to " + j);
- for (int c = 0; c < frm.length; c++)
+ last[1] = range[1];
+ return;
+ }
+
+ /*
+ * next range starts at +1 in forward sense - update endpoint
+ */
+ if (lastForward && newForward && range[0] == last[1] + 1)
+ {
+ last[1] = range[1];
+ return;
+ }
+
+ /*
+ * next range starts at -1 in reverse sense - update endpoint
+ */
+ if (!lastForward && !newForward && range[0] == last[1] - 1)
+ {
+ last[1] = range[1];
+ return;
+ }
+
+ /*
+ * just add the new range
+ */
+ addTo.add(range);
+ }
+
+ /**
+ * Returns true if mapping is from forward strand, false if from reverse
+ * strand. Result is just based on the first 'from' range that is not a single
+ * position. Default is true unless proven to be false. Behaviour is not well
+ * defined if the mapping has a mixture of forward and reverse ranges.
+ *
+ * @return
+ */
+ public boolean isFromForwardStrand()
+ {
+ return isForwardStrand(getFromRanges());
+ }
+
+ /**
+ * Returns true if mapping is to forward strand, false if to reverse strand.
+ * Result is just based on the first 'to' range that is not a single position.
+ * Default is true unless proven to be false. Behaviour is not well defined if
+ * the mapping has a mixture of forward and reverse ranges.
+ *
+ * @return
+ */
+ public boolean isToForwardStrand()
+ {
+ return isForwardStrand(getToRanges());
+ }
+
+ /**
+ * A helper method that returns true unless at least one range has start >
+ * end. Behaviour is undefined for a mixture of forward and reverse ranges.
+ *
+ * @param ranges
+ * @return
+ */
+ private boolean isForwardStrand(List<int[]> ranges)
+ {
+ boolean forwardStrand = true;
+ for (int[] range : ranges)
+ {
+ if (range[1] > range[0])
{
- System.err.print(frm[c] + ((c % 2 == 0) ? "," : ";"));
+ break; // forward strand confirmed
}
- System.err.println("Expected");
- for (int c = 0; c < ks.length; c++)
+ else if (range[1] < range[0])
{
- System.err.print(ks[c] + ((c % 2 == 0) ? "," : ";"));
+ forwardStrand = false;
+ break; // reverse strand confirmed
}
}
+ return forwardStrand;
}
/**
*
- * @return a MapList whose From range is this maplist's To Range, and vice
- * versa
+ * @return true if from, or to is a three to 1 mapping
*/
- public MapList getInverse()
+ public boolean isTripletMap()
{
- return new MapList(getToRanges(), getFromRanges(), getToRatio(),
- getFromRatio());
+ return (toRatio == 3 && fromRatio == 1)
+ || (fromRatio == 3 && toRatio == 1);
}
/**
- * test for containment rather than equivalence to another mapping
+ * Returns a map which is the composite of this one and the input map. That
+ * is, the output map has the fromRanges of this map, and its toRanges are the
+ * toRanges of this map as transformed by the input map.
+ * <p>
+ * Returns null if the mappings cannot be traversed (not all toRanges of this
+ * map correspond to fromRanges of the input), or if this.toRatio does not
+ * match map.fromRatio.
+ *
+ * <pre>
+ * Example 1:
+ * this: from [1-100] to [501-600]
+ * input: from [10-40] to [60-90]
+ * output: from [10-40] to [560-590]
+ * Example 2 ('reverse strand exons'):
+ * this: from [1-100] to [2000-1951], [1000-951] // transcript to loci
+ * input: from [1-50] to [41-90] // CDS to transcript
+ * output: from [10-40] to [1960-1951], [1000-971] // CDS to gene loci
+ * </pre>
*
* @param map
- * to be tested for containment
- * @return true if local or mapped range map contains or is contained by this
- * mapping
+ * @return
*/
- public boolean containsEither(boolean local, MapList map)
+ public MapList traverse(MapList map)
{
- if (local)
+ if (map == null)
{
- return ((getFromLowest() >= map.getFromLowest() && getFromHighest() <= map
- .getFromHighest()) || (getFromLowest() <= map.getFromLowest() && getFromHighest() >= map
- .getFromHighest()));
+ return null;
}
- else
+
+ /*
+ * compound the ratios by this rule:
+ * A:B with M:N gives A*M:B*N
+ * reduced by greatest common divisor
+ * so 1:3 with 3:3 is 3:9 or 1:3
+ * 1:3 with 3:1 is 3:3 or 1:1
+ * 1:3 with 1:3 is 1:9
+ * 2:5 with 3:7 is 6:35
+ */
+ int outFromRatio = getFromRatio() * map.getFromRatio();
+ int outToRatio = getToRatio() * map.getToRatio();
+ int gcd = MathUtils.gcd(outFromRatio, outToRatio);
+ outFromRatio /= gcd;
+ outToRatio /= gcd;
+
+ List<int[]> toRanges = new ArrayList<>();
+ for (int[] range : getToRanges())
{
- return ((getToLowest() >= map.getToLowest() && getToHighest() <= map
- .getToHighest()) || (getToLowest() <= map.getToLowest() && getToHighest() >= map
- .getToHighest()));
+ int[] transferred = map.locateInTo(range[0], range[1]);
+ if (transferred == null || transferred.length % 2 != 0)
+ {
+ return null;
+ }
+
+ /*
+ * convert [start1, end1, start2, end2, ...]
+ * to [[start1, end1], [start2, end2], ...]
+ */
+ for (int i = 0; i < transferred.length;)
+ {
+ toRanges.add(new int[] { transferred[i], transferred[i + 1] });
+ i += 2;
+ }
}
+
+ return new MapList(getFromRanges(), toRanges, outFromRatio, outToRatio);
+ }
+
+ /**
+ * Answers true if the mapping is from one contiguous range to another, else
+ * false
+ *
+ * @return
+ */
+ public boolean isContiguous()
+ {
+ return fromShifts.size() == 1 && toShifts.size() == 1;
+ }
+
+ /**
+ * Returns the [start, end...] positions in the range mapped from, that are
+ * mapped to by part or all of the given begin-end of the range mapped to.
+ * Returns null if begin-end does not overlap any position mapped to.
+ *
+ * @param begin
+ * @param end
+ * @return
+ */
+ public int[] getOverlapsInFrom(final int begin, final int end)
+ {
+ int[] overlaps = MappingUtils.findOverlap(toShifts, begin, end);
+
+ return overlaps == null ? null : locateInFrom(overlaps[0], overlaps[1]);
+ }
+
+ /**
+ * Returns the [start, end...] positions in the range mapped to, that are
+ * mapped to by part or all of the given begin-end of the range mapped from.
+ * Returns null if begin-end does not overlap any position mapped from.
+ *
+ * @param begin
+ * @param end
+ * @return
+ */
+ public int[] getOverlapsInTo(final int begin, final int end)
+ {
+ int[] overlaps = MappingUtils.findOverlap(fromShifts, begin, end);
+
+ return overlaps == null ? null : locateInTo(overlaps[0], overlaps[1]);
}
}
git://source.jalview.org / jalview.git / blobdiff