*/
package jalview.datamodel;
-import java.util.*;
+import jalview.datamodel.features.FeatureLocationI;
+
+import java.util.HashMap;
+import java.util.Map;
+import java.util.Map.Entry;
+import java.util.Vector;
/**
* DOCUMENT ME!
* @author $author$
* @version $Revision$
*/
-public class SequenceFeature
+public class SequenceFeature implements FeatureLocationI
{
- public int begin;
+ /*
+ * score value if none is set; preferably Float.Nan, but see
+ * JAL-2060 and JAL-2554 for a couple of blockers to that
+ */
+ private static final float NO_SCORE = 0f;
+
+ private static final String STATUS = "status";
+
+ private static final String STRAND = "STRAND";
+
+ // private key for Phase designed not to conflict with real GFF data
+ private static final String PHASE = "!Phase";
+
+ // private key for ENA location designed not to conflict with real GFF data
+ private static final String LOCATION = "!Location";
+
+ /*
+ * ATTRIBUTES is reserved for the GFF 'column 9' data, formatted as
+ * name1=value1;name2=value2,value3;...etc
+ */
+ private static final String ATTRIBUTES = "ATTRIBUTES";
- public int end;
+ /*
+ * type, begin, end, featureGroup, score and contactFeature are final
+ * to ensure that the integrity of SequenceFeatures data store
+ * can't be broken by direct update of these fields
+ */
+ public final String type;
+
+ public final int begin;
+
+ public final int end;
+
+ public final String featureGroup;
- public float score;
+ public final float score;
- public String type;
+ private final boolean contactFeature;
public String description;
- public Hashtable otherDetails;
+ /*
+ * a map of key-value pairs; may be populated from GFF 'column 9' data,
+ * other data sources (e.g. GenBank file), or programmatically
+ */
+ public Map<String, Object> otherDetails;
+
+ public Vector<String> links;
- public java.util.Vector links;
+ /**
+ * Constructs a duplicate feature. Note: Uses makes a shallow copy of the
+ * otherDetails map, so the new and original SequenceFeature may reference the
+ * same objects in the map.
+ *
+ * @param cpy
+ */
+ public SequenceFeature(SequenceFeature cpy)
+ {
+ this(cpy, cpy.getBegin(), cpy.getEnd(), cpy.getFeatureGroup(), cpy
+ .getScore());
+ }
- // Feature group can be set from a features file
- // as a group of features between STARTGROUP and ENDGROUP markers
- public String featureGroup;
+ /**
+ * Constructor
+ *
+ * @param theType
+ * @param theDesc
+ * @param theBegin
+ * @param theEnd
+ * @param group
+ */
+ public SequenceFeature(String theType, String theDesc, int theBegin,
+ int theEnd, String group)
+ {
+ this(theType, theDesc, theBegin, theEnd, NO_SCORE, group);
+ }
- public SequenceFeature()
+ /**
+ * Constructor including a score value
+ *
+ * @param theType
+ * @param theDesc
+ * @param theBegin
+ * @param theEnd
+ * @param theScore
+ * @param group
+ */
+ public SequenceFeature(String theType, String theDesc, int theBegin,
+ int theEnd, float theScore, String group)
{
+ this.type = theType;
+ this.description = theDesc;
+ this.begin = theBegin;
+ this.end = theEnd;
+ this.featureGroup = group;
+ this.score = theScore;
+
+ /*
+ * for now, only "Disulfide/disulphide bond" is treated as a contact feature
+ */
+ this.contactFeature = "disulfide bond".equalsIgnoreCase(type)
+ || "disulphide bond".equalsIgnoreCase(type);
}
/**
- * Constructs a duplicate feature. Note: Uses clone on the otherDetails so
- * only shallow copies are made of additional properties and method will
- * silently fail if unclonable objects are found in the hash.
+ * A copy constructor that allows the value of final fields to be 'modified'
*
- * @param cpy
+ * @param sf
+ * @param newType
+ * @param newBegin
+ * @param newEnd
+ * @param newGroup
+ * @param newScore
*/
- public SequenceFeature(SequenceFeature cpy)
+ public SequenceFeature(SequenceFeature sf, String newType, int newBegin,
+ int newEnd, String newGroup, float newScore)
{
- if (cpy != null)
+ this(newType, sf.getDescription(), newBegin, newEnd, newScore,
+ newGroup);
+
+ if (sf.otherDetails != null)
{
- begin = cpy.begin;
- end = cpy.end;
- score = cpy.score;
- if (cpy.type != null)
- {
- type = new String(cpy.type);
- }
- if (cpy.description != null)
- {
- description = new String(cpy.description);
- }
- if (cpy.featureGroup != null)
+ otherDetails = new HashMap<String, Object>();
+ for (Entry<String, Object> entry : sf.otherDetails.entrySet())
{
- featureGroup = new String(cpy.featureGroup);
+ otherDetails.put(entry.getKey(), entry.getValue());
}
- if (cpy.otherDetails != null)
- {
- try
- {
- otherDetails = (Hashtable) cpy.otherDetails.clone();
- } catch (Exception e)
- {
- // Uncloneable objects in the otherDetails - don't complain
- }
- }
- if (cpy.links != null && cpy.links.size() > 0)
+ }
+ if (sf.links != null && sf.links.size() > 0)
+ {
+ links = new Vector<String>();
+ for (int i = 0, iSize = sf.links.size(); i < iSize; i++)
{
- links = new Vector();
- for (int i = 0, iSize = cpy.links.size(); i < iSize; i++)
- {
- links.addElement(cpy.links.elementAt(i));
- }
+ links.addElement(sf.links.elementAt(i));
}
}
}
- public SequenceFeature(String type, String desc, String status,
- int begin, int end, String featureGroup)
+ /**
+ * A copy constructor that allows the value of final fields to be 'modified'
+ *
+ * @param sf
+ * @param newBegin
+ * @param newEnd
+ * @param newGroup
+ * @param newScore
+ */
+ public SequenceFeature(SequenceFeature sf, int newBegin, int newEnd,
+ String newGroup, float newScore)
{
- this.type = type;
- this.description = desc;
- setValue("status", status);
- this.begin = begin;
- this.end = end;
- this.featureGroup = featureGroup;
+ this(sf, sf.getType(), newBegin, newEnd, newGroup, newScore);
}
- public SequenceFeature(String type, String desc, int begin, int end,
- float score, String featureGroup)
+ /**
+ * Two features are considered equal if they have the same type, group,
+ * description, start, end, phase, strand, and (if present) 'Name', ID' and
+ * 'Parent' attributes.
+ *
+ * Note we need to check Parent to distinguish the same exon occurring in
+ * different transcripts (in Ensembl GFF). This allows assembly of transcript
+ * sequences from their component exon regions.
+ */
+ @Override
+ public boolean equals(Object o)
{
- this.type = type;
- this.description = desc;
- this.begin = begin;
- this.end = end;
- this.score = score;
- this.featureGroup = featureGroup;
+ return equals(o, false);
}
- public boolean equals(SequenceFeature sf)
+ /**
+ * Overloaded method allows the equality test to optionally ignore the
+ * 'Parent' attribute of a feature. This supports avoiding adding many
+ * superficially duplicate 'exon' or CDS features to genomic or protein
+ * sequence.
+ *
+ * @param o
+ * @param ignoreParent
+ * @return
+ */
+ public boolean equals(Object o, boolean ignoreParent)
{
- if (begin != sf.begin || end != sf.end || score != sf.score)
+ if (o == null || !(o instanceof SequenceFeature))
+ {
+ return false;
+ }
+
+ SequenceFeature sf = (SequenceFeature) o;
+ boolean sameScore = Float.isNaN(score) ? Float.isNaN(sf.score)
+ : score == sf.score;
+ if (begin != sf.begin || end != sf.end || !sameScore)
{
return false;
}
- if (!(type + description + featureGroup).equals(sf.type
- + sf.description + sf.featureGroup))
+ if (getStrand() != sf.getStrand())
{
return false;
}
+ if (!(type + description + featureGroup + getPhase()).equals(
+ sf.type + sf.description + sf.featureGroup + sf.getPhase()))
+ {
+ return false;
+ }
+ if (!equalAttribute(getValue("ID"), sf.getValue("ID")))
+ {
+ return false;
+ }
+ if (!equalAttribute(getValue("Name"), sf.getValue("Name")))
+ {
+ return false;
+ }
+ if (!ignoreParent)
+ {
+ if (!equalAttribute(getValue("Parent"), sf.getValue("Parent")))
+ {
+ return false;
+ }
+ }
return true;
}
/**
+ * Returns true if both values are null, are both non-null and equal
+ *
+ * @param att1
+ * @param att2
+ * @return
+ */
+ protected static boolean equalAttribute(Object att1, Object att2)
+ {
+ if (att1 == null && att2 == null)
+ {
+ return true;
+ }
+ if (att1 != null)
+ {
+ return att1.equals(att2);
+ }
+ return att2.equals(att1);
+ }
+
+ /**
* DOCUMENT ME!
*
* @return DOCUMENT ME!
*/
+ @Override
public int getBegin()
{
return begin;
}
- public void setBegin(int start)
- {
- this.begin = start;
- }
-
/**
* DOCUMENT ME!
*
* @return DOCUMENT ME!
*/
+ @Override
public int getEnd()
{
return end;
}
- public void setEnd(int end)
- {
- this.end = end;
- }
-
/**
* DOCUMENT ME!
*
return type;
}
- public void setType(String type)
- {
- this.type = type;
- }
-
/**
* DOCUMENT ME!
*
return featureGroup;
}
- public void setFeatureGroup(String featureGroup)
- {
- this.featureGroup = featureGroup;
- }
-
public void addLink(String labelLink)
{
if (links == null)
{
- links = new java.util.Vector();
+ links = new Vector<String>();
}
- links.insertElementAt(labelLink, 0);
+ if (!links.contains(labelLink))
+ {
+ links.insertElementAt(labelLink, 0);
+ }
}
public float getScore()
return score;
}
- public void setScore(float value)
- {
- score = value;
- }
-
/**
- * Used for getting values which are not in the basic set. eg STRAND, FRAME
+ * Used for getting values which are not in the basic set. eg STRAND, PHASE
* for GFF file
*
* @param key
}
/**
+ * Returns a property value for the given key if known, else the specified
+ * default value
+ *
+ * @param key
+ * @param defaultValue
+ * @return
+ */
+ public Object getValue(String key, Object defaultValue)
+ {
+ Object value = getValue(key);
+ return value == null ? defaultValue : value;
+ }
+
+ /**
* Used for setting values which are not in the basic set. eg STRAND, FRAME
* for GFF file
*
{
if (otherDetails == null)
{
- otherDetails = new Hashtable();
+ otherDetails = new HashMap<String, Object>();
}
otherDetails.put(key, value);
*/
public void setStatus(String status)
{
- setValue("status", status);
+ setValue(STATUS, status);
}
public String getStatus()
{
+ return (String) getValue(STATUS);
+ }
+
+ public void setAttributes(String attr)
+ {
+ setValue(ATTRIBUTES, attr);
+ }
+
+ public String getAttributes()
+ {
+ return (String) getValue(ATTRIBUTES);
+ }
+
+ /**
+ * Return 1 for forward strand ('+' in GFF), -1 for reverse strand ('-' in
+ * GFF), and 0 for unknown or not (validly) specified
+ *
+ * @return
+ */
+ public int getStrand()
+ {
+ int strand = 0;
if (otherDetails != null)
{
- String stat = (String) otherDetails.get("status");
- if (stat != null)
- return new String(stat);
+ Object str = otherDetails.get(STRAND);
+ if ("-".equals(str))
+ {
+ strand = -1;
+ }
+ else if ("+".equals(str))
+ {
+ strand = 1;
+ }
}
- return null;
+ return strand;
}
- public void setPosition(int pos)
+ /**
+ * Set the value of strand
+ *
+ * @param strand
+ * should be "+" for forward, or "-" for reverse
+ */
+ public void setStrand(String strand)
{
- begin = pos;
- end = pos;
+ setValue(STRAND, strand);
}
- public int getPosition()
+ public void setPhase(String phase)
{
- return begin;
+ setValue(PHASE, phase);
+ }
+
+ public String getPhase()
+ {
+ return (String) getValue(PHASE);
+ }
+
+ /**
+ * Sets the 'raw' ENA format location specifier e.g. join(12..45,89..121)
+ *
+ * @param loc
+ */
+ public void setEnaLocation(String loc)
+ {
+ setValue(LOCATION, loc);
+ }
+
+ /**
+ * Gets the 'raw' ENA format location specifier e.g. join(12..45,89..121)
+ *
+ * @param loc
+ */
+ public String getEnaLocation()
+ {
+ return (String) getValue(LOCATION);
}
+ /**
+ * Readable representation, for debug only, not guaranteed not to change
+ * between versions
+ */
+ @Override
+ public String toString()
+ {
+ return String.format("%d %d %s %s", getBegin(), getEnd(), getType(),
+ getDescription());
+ }
+
+ /**
+ * Overridden to ensure that whenever two objects are equal, they have the
+ * same hashCode
+ */
+ @Override
+ public int hashCode()
+ {
+ String s = getType() + getDescription() + getFeatureGroup()
+ + getValue("ID") + getValue("Name") + getValue("Parent")
+ + getPhase();
+ return s.hashCode() + getBegin() + getEnd() + (int) getScore()
+ + getStrand();
+ }
+
+ /**
+ * Answers true if the feature's start/end values represent two related
+ * positions, rather than ends of a range. Such features may be visualised or
+ * reported differently to features on a range.
+ */
+ @Override
+ public boolean isContactFeature()
+ {
+ return contactFeature;
+ }
+
+ /**
+ * Answers true if the sequence has zero start and end position
+ *
+ * @return
+ */
+ public boolean isNonPositional()
+ {
+ return begin == 0 && end == 0;
+ }
}