X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fio%2FAlignFile.java;h=497f0a580836af85016458f6f32c3ecd97adc9a3;hb=6fd29e7e2753c9066fa31b76ab71827f96d471ff;hp=3ef721003a82c83babb5a951563de6ff79ff3c2a;hpb=a79f9e113c51c032070c670e45ce3eb464691166;p=jalview.git

diff --git a/src/jalview/io/AlignFile.java b/src/jalview/io/AlignFile.java
index 3ef7210..497f0a5 100755
--- a/src/jalview/io/AlignFile.java
+++ b/src/jalview/io/AlignFile.java
@@ -1,30 +1,37 @@
 /*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.7)
- * Copyright (C) 2011 J Procter, AM Waterhouse, J Engelhardt, LM Lui, 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.io;
 
-import jalview.datamodel.Alignment;
 import jalview.datamodel.AlignmentAnnotation;
+import jalview.datamodel.AlignmentI;
 import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceGroup;
 import jalview.datamodel.SequenceI;
+import jalview.util.MessageManager;
 
 import java.io.IOException;
+import java.util.ArrayList;
 import java.util.Enumeration;
 import java.util.Hashtable;
+import java.util.List;
 import java.util.Vector;
 
 /**
@@ -34,55 +41,103 @@ import java.util.Vector;
  * @version $Revision$
  */
 public abstract class AlignFile extends FileParse
+        implements AlignmentFileReaderI, AlignmentFileWriterI
 {
   int noSeqs = 0;
 
   int maxLength = 0;
 
   /**
-   * Sequences to be added to form a new alignment.
+   * Sequences to be added to form a new alignment. TODO: remove vector in this
+   * class
    */
   protected Vector<SequenceI> seqs;
 
   /**
    * annotation to be added to generated alignment object
    */
-  protected Vector annotations;
+  protected Vector<AlignmentAnnotation> annotations;
+
+  /**
+   * SequenceGroups to be added to the alignment object
+   */
+  protected List<SequenceGroup> seqGroups;
 
   /**
    * Properties to be added to generated alignment object
    */
-  protected Hashtable properties;
+  private Hashtable properties;
 
   long start;
 
   long end;
 
-  boolean jvSuffix = true;
+  /**
+   * true if parse() has been called
+   */
+  private boolean parseCalled = false;
+
+  private boolean parseImmediately = true;
+
+  /**
+   * @return if doParse() was called at construction time
+   */
+  protected boolean isParseImmediately()
+  {
+    return parseImmediately;
+  }
 
   /**
    * Creates a new AlignFile object.
    */
   public AlignFile()
   {
+    // Shouldn't we init data structures (JBPNote: not sure - initData is for
+    // initialising the structures used for reading from a datasource, and the
+    // bare constructor hasn't got any datasource)
+    initData();
+  }
+
+  public AlignFile(SequenceI[] seqs)
+  {
+    this();
+    setSeqs(seqs);
   }
 
   /**
    * Constructor which parses the data from a file of some specified type.
    * 
-   * @param inFile
-   *          Filename to read from.
-   * @param type
+   * @param dataObject
+   *          Filename, URL or Pasted String to read from.
+   * @param sourceType
+   *          What type of file to read from (File, URL, Pasted String)
+   */
+  public AlignFile(String dataObject, DataSourceType sourceType)
+          throws IOException
+  {
+    this(true, dataObject, sourceType);
+  }
+
+  /**
+   * Constructor which (optionally delays) parsing of data from a file of some
+   * specified type.
+   * 
+   * @param parseImmediately
+   *          if false, need to call 'doParse()' to begin parsing data
+   * @param dataObject
+   *          Filename, URL or Pasted String to read from.
+   * @param sourceType
    *          What type of file to read from (File, URL)
+   * @throws IOException
    */
-  public AlignFile(String inFile, String type) throws IOException
+  public AlignFile(boolean parseImmediately, String dataObject,
+          DataSourceType sourceType) throws IOException
   {
-    super(inFile, type);
+    super(dataObject, sourceType);
     initData();
-    parse();
-    // sets the index of each sequence in the alignment
-    for( int i=0,c=seqs.size(); i<c; i++ ) {  
-    	seqs.get(i).setIndex(i);  
+    if (parseImmediately)
+    {
+      doParse();
     }
   }
 
@@ -95,16 +150,51 @@ public abstract class AlignFile extends FileParse
    */
   public AlignFile(FileParse source) throws IOException
   {
+    this(true, source);
+  }
+
+  /**
+   * Construct a new parser to read from the position where some other parsing
+   * process left
+   * 
+   * @param parseImmediately
+   *          if false, need to call 'doParse()' to begin parsing data
+   * @param source
+   */
+  public AlignFile(boolean parseImmediately, FileParse source)
+          throws IOException
+  {
     super(source);
     initData();
-    parse();
-    // sets the index of each sequence in the alignment
-    for( int i=0,c=seqs.size(); i<c; i++ ) {  
-    	seqs.get(i).setIndex(i);  
+
+    // stash flag in case parse needs to know if it has to autoconfigure or was
+    // configured after construction
+    this.parseImmediately = parseImmediately;
+
+    if (parseImmediately)
+    {
+      doParse();
     }
   }
 
   /**
+   * called if parsing was delayed till after parser was constructed
+   * 
+   * @throws IOException
+   */
+  public void doParse() throws IOException
+  {
+    if (parseCalled)
+    {
+      throw new IOException(
+              "Implementation error: Parser called twice for same data.\n"
+                      + "Need to call initData() again before parsing can be reattempted.");
+    }
+    parseCalled = true;
+    parse();
+  }
+
+  /**
    * Return the seqs Vector
    */
   public Vector<SequenceI> getSeqs()
@@ -112,16 +202,22 @@ public abstract class AlignFile extends FileParse
     return seqs;
   }
 
+  public List<SequenceGroup> getSeqGroups()
+  {
+    return seqGroups;
+  }
+
   /**
    * Return the Sequences in the seqs Vector as an array of Sequences
    */
+  @Override
   public SequenceI[] getSeqsAsArray()
   {
     SequenceI[] s = new SequenceI[seqs.size()];
 
     for (int i = 0; i < seqs.size(); i++)
     {
-      s[i] = (SequenceI) seqs.elementAt(i);
+      s[i] = seqs.elementAt(i);
     }
 
     return s;
@@ -133,7 +229,8 @@ public abstract class AlignFile extends FileParse
    * 
    * @param al
    */
-  public void addAnnotations(Alignment al)
+  @Override
+  public void addAnnotations(AlignmentI al)
   {
     addProperties(al);
     for (int i = 0; i < annotations.size(); i++)
@@ -145,8 +242,7 @@ public abstract class AlignFile extends FileParse
        * Rna.GetBasePairsFromAlignmentAnnotation(annotations.elementAt(i));
        * Rna.HelixMap(pairArray);
        */
-      AlignmentAnnotation an = (AlignmentAnnotation) annotations
-              .elementAt(i);
+      AlignmentAnnotation an = annotations.elementAt(i);
       an.validateRangeAndDisplay();
       al.addAnnotation(an);
     }
@@ -154,13 +250,24 @@ public abstract class AlignFile extends FileParse
   }
 
   /**
+   * register sequence groups on the alignment for **output**
+   * 
+   * @param al
+   */
+  public void addSeqGroups(AlignmentI al)
+  {
+    this.seqGroups = al.getGroups();
+
+  }
+
+  /**
    * Add any additional information extracted from the file to the alignment
    * properties.
    * 
    * @note implicitly called by addAnnotations()
    * @param al
    */
-  public void addProperties(Alignment al)
+  public void addProperties(AlignmentI al)
   {
     if (properties != null && properties.size() > 0)
     {
@@ -187,8 +294,8 @@ public abstract class AlignFile extends FileParse
   {
     if (key == null)
     {
-      throw new Error(
-              "Implementation error: Cannot have null alignment property key.");
+      throw new Error(MessageManager.getString(
+              "error.implementation_error_cannot_have_null_alignment"));
     }
     if (value == null)
     {
@@ -215,8 +322,10 @@ public abstract class AlignFile extends FileParse
    */
   protected void initData()
   {
-    seqs = new Vector();
-    annotations = new Vector();
+    seqs = new Vector<SequenceI>();
+    annotations = new Vector<AlignmentAnnotation>();
+    seqGroups = new ArrayList<SequenceGroup>();
+    parseCalled = false;
   }
 
   /**
@@ -225,9 +334,10 @@ public abstract class AlignFile extends FileParse
    * @param s
    *          DOCUMENT ME!
    */
-  protected void setSeqs(SequenceI[] s)
+  @Override
+  public void setSeqs(SequenceI[] s)
   {
-    seqs = new Vector();
+    seqs = new Vector<SequenceI>();
 
     for (int i = 0; i < s.length; i++)
     {
@@ -241,16 +351,6 @@ public abstract class AlignFile extends FileParse
   public abstract void parse() throws IOException;
 
   /**
-   * Print out in alignment file format the Sequences in the seqs Vector.
-   */
-  public abstract String print();
-
-  public void addJVSuffix(boolean b)
-  {
-    jvSuffix = b;
-  }
-
-  /**
    * A general parser for ids.
    * 
    * @String id Id to be parsed
@@ -263,7 +363,15 @@ public abstract class AlignFile extends FileParse
     if (space > -1)
     {
       seq = new Sequence(id.substring(0, space), "");
-      seq.setDescription(id.substring(space + 1));
+      String desc = id.substring(space + 1);
+      seq.setDescription(desc);
+
+      /*
+       * it is tempting to parse Ensembl style gene description e.g.
+       * chromosome:GRCh38:7:140696688:140721955:1 and set the
+       * start position of the sequence, but this causes much confusion
+       * for reverse strand feature locations
+       */
     }
     else
     {
@@ -274,38 +382,54 @@ public abstract class AlignFile extends FileParse
   }
 
   /**
-   * Creates the output id. Adds prefix Uniprot format source|id And suffix
-   * Jalview /start-end
+   * Creates the output id. Adds prefix Uniprot format source|id and optionally
+   * suffix Jalview /start-end
+   * 
+   * @param jvsuffix
    * 
    * @String id Id to be parsed
    */
+  String printId(SequenceI seq, boolean jvsuffix)
+  {
+    return seq.getDisplayId(jvsuffix);
+  }
+
   String printId(SequenceI seq)
   {
-    return seq.getDisplayId(jvSuffix);
+    return printId(seq, true);
   }
 
   /**
    * vector of String[] treeName, newickString pairs
    */
-  Vector newickStrings = null;
+  Vector<String[]> newickStrings = null;
 
   protected void addNewickTree(String treeName, String newickString)
   {
     if (newickStrings == null)
     {
-      newickStrings = new Vector();
+      newickStrings = new Vector<String[]>();
     }
-    newickStrings.addElement(new String[]
-    { treeName, newickString });
+    newickStrings.addElement(new String[] { treeName, newickString });
   }
 
   protected int getTreeCount()
   {
-    if (newickStrings == null)
+    return newickStrings == null ? 0 : newickStrings.size();
+  }
+
+  @Override
+  public void addGroups(AlignmentI al)
+  {
+
+    for (SequenceGroup sg : getSeqGroups())
     {
-      return 0;
+      al.addGroup(sg);
     }
-    return newickStrings.size();
   }
 
+  protected void addSequence(SequenceI seq)
+  {
+    seqs.add(seq);
+  }
 }