X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fdatamodel%2FSequenceFeature.java;h=f17bd3349604541fd923c9da101bf16e77abea80;hb=bab4b08733a9543bdce5c0b9d1c32ca7d18ad9e3;hp=681bedbca9f08ec16f2aa29b613fc18f88adffb4;hpb=d423f22792e47dbc800ae220a58677f988971d06;p=jalview.git

diff --git a/src/jalview/datamodel/SequenceFeature.java b/src/jalview/datamodel/SequenceFeature.java
index 681bedb..f17bd33 100755
--- a/src/jalview/datamodel/SequenceFeature.java
+++ b/src/jalview/datamodel/SequenceFeature.java
@@ -1,23 +1,32 @@
 /*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.5)
- * Copyright (C) 2010 J Procter, AM Waterhouse, G Barton, M Clamp, S Searle
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
  * 
  * This file is part of Jalview.
  * 
  * Jalview is free software: you can redistribute it and/or
  * modify it under the terms of the GNU General Public License 
- * as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
- * 
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *  
  * Jalview is distributed in the hope that it will be useful, but 
  * WITHOUT ANY WARRANTY; without even the implied warranty 
  * of MERCHANTABILITY or FITNESS FOR A PARTICULAR 
  * PURPOSE.  See the GNU General Public License for more details.
  * 
- * You should have received a copy of the GNU General Public License along with Jalview.  If not, see <http://www.gnu.org/licenses/>.
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview.  If not, see <http://www.gnu.org/licenses/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
  */
 package jalview.datamodel;
 
-import java.util.*;
+import jalview.datamodel.features.FeatureLocationI;
+
+import java.util.Comparator;
+import java.util.HashMap;
+import java.util.Map;
+import java.util.Map.Entry;
+import java.util.Vector;
 
 /**
  * DOCUMENT ME!
@@ -25,145 +34,271 @@ import java.util.*;
  * @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 float score;
+  public final int end;
 
-  public String type;
+  public final String featureGroup;
+
+  public final float score;
+
+  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 java.util.Vector links;
+  public Vector<String> links;
 
-  // Feature group can be set from a features file
-  // as a group of features between STARTGROUP and ENDGROUP markers
-  public String featureGroup;
+  /**
+   * 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());
+  }
 
-  public SequenceFeature()
+  /**
+   * 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);
   }
 
   /**
-   * 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.
+   * Constructor including a score value
    * 
-   * @param cpy
+   * @param theType
+   * @param theDesc
+   * @param theBegin
+   * @param theEnd
+   * @param theScore
+   * @param group
    */
-  public SequenceFeature(SequenceFeature cpy)
+  public SequenceFeature(String theType, String theDesc, int theBegin,
+          int theEnd, float theScore, String group)
   {
-    if (cpy != null)
+    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);
+  }
+
+  /**
+   * A copy constructor that allows the value of final fields to be 'modified'
+   * 
+   * @param sf
+   * @param newType
+   * @param newBegin
+   * @param newEnd
+   * @param newGroup
+   * @param newScore
+   */
+  public SequenceFeature(SequenceFeature sf, String newType, int newBegin,
+          int newEnd, String newGroup, float newScore)
+  {
+    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)
-      {
-        featureGroup = new String(cpy.featureGroup);
-      }
-      if (cpy.otherDetails != null)
+      otherDetails = new HashMap<>();
+      for (Entry<String, Object> entry : sf.otherDetails.entrySet())
       {
-        try
-        {
-          otherDetails = (Hashtable) cpy.otherDetails.clone();
-        } catch (Exception e)
-        {
-          // Uncloneable objects in the otherDetails - don't complain
-        }
+        otherDetails.put(entry.getKey(), entry.getValue());
       }
-      if (cpy.links != null && cpy.links.size() > 0)
+    }
+    if (sf.links != null && sf.links.size() > 0)
+    {
+      links = new Vector<>();
+      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;
     }
 
-    if (!(type + description + featureGroup).equals(sf.type
-            + sf.description + sf.featureGroup))
+    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 (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!
    * 
@@ -174,11 +309,6 @@ public class SequenceFeature
     return type;
   }
 
-  public void setType(String type)
-  {
-    this.type = type;
-  }
-
   /**
    * DOCUMENT ME!
    * 
@@ -199,19 +329,17 @@ public class SequenceFeature
     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<>();
     }
 
-    links.insertElementAt(labelLink, 0);
+    if (!links.contains(labelLink))
+    {
+      links.insertElementAt(labelLink, 0);
+    }
   }
 
   public float getScore()
@@ -219,13 +347,8 @@ public class SequenceFeature
     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
@@ -244,6 +367,20 @@ public class SequenceFeature
   }
 
   /**
+   * 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
    * 
@@ -258,7 +395,7 @@ public class SequenceFeature
     {
       if (otherDetails == null)
       {
-        otherDetails = new Hashtable();
+        otherDetails = new HashMap<>();
       }
 
       otherDetails.put(key, value);
@@ -271,29 +408,150 @@ public class SequenceFeature
    */
   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;
+  }
+
+  /**
+   * Set the value of strand
+   * 
+   * @param strand
+   *          should be "+" for forward, or "-" for reverse
+   */
+  public void setStrand(String strand)
+  {
+    setValue(STRAND, strand);
   }
 
-  public void setPosition(int pos)
+  public void setPhase(String phase)
   {
-    begin = pos;
-    end = pos;
+    setValue(PHASE, phase);
   }
 
-  public int getPosition()
+  public String getPhase()
   {
-    return begin;
+    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;
+  }
+}
+
+class SFSortByEnd implements Comparator<SequenceFeature>
+{
+  @Override
+  public int compare(SequenceFeature a, SequenceFeature b)
+  {
+    return a.getEnd() - b.getEnd();
+  }
+}
+
+class SFSortByBegin implements Comparator<SequenceFeature>
+{
+  @Override
+  public int compare(SequenceFeature a, SequenceFeature b)
+  {
+    return a.getBegin() - b.getBegin();
+  }
 }