X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fio%2FMSFfile.java;h=db29848afd96fbe66210e8fbc21625e68ea9f9be;hb=2305d687bac5ed8a76ecb93c54c7b090a928e362;hp=7c70a0967fdfbbe7e8b531d36500ba8d95c003dc;hpb=1e5dc0e8a71d9eb0be4748f994072b11f25b04e0;p=jalview.git
diff --git a/src/jalview/io/MSFfile.java b/src/jalview/io/MSFfile.java
index 7c70a09..db29848 100755
--- a/src/jalview/io/MSFfile.java
+++ b/src/jalview/io/MSFfile.java
@@ -1,27 +1,36 @@
/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8)
- * Copyright (C) 2012 J Procter, AM Waterhouse, LM Lui, J Engelhardt, 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 .
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
*/
package jalview.io;
-import java.io.*;
-import java.util.*;
+import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceI;
+import jalview.util.Comparison;
+import jalview.util.Format;
-import jalview.datamodel.*;
-import jalview.util.*;
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Hashtable;
+import java.util.List;
+import java.util.Locale;
+import java.util.StringTokenizer;
/**
* DOCUMENT ME!
@@ -50,7 +59,7 @@ public class MSFfile extends AlignFile
* @throws IOException
* DOCUMENT ME!
*/
- public MSFfile(String inFile, String type) throws IOException
+ public MSFfile(String inFile, DataSourceType type) throws IOException
{
super(inFile, type);
}
@@ -60,28 +69,24 @@ public class MSFfile extends AlignFile
super(source);
}
- {
- // TODO Auto-generated constructor stub
- }
-
/**
- * DOCUMENT ME!
+ * Read and parse MSF sequence data
*/
+ @Override
public void parse() throws IOException
{
- int i = 0;
boolean seqFlag = false;
- String key = new String();
- Vector headers = new Vector();
- Hashtable seqhash = new Hashtable();
- String line;
+ List headers = new ArrayList();
+ Hashtable seqhash = new Hashtable();
try
{
+ String line;
while ((line = nextLine()) != null)
{
StringTokenizer str = new StringTokenizer(line);
+ String key = null;
while (str.hasMoreTokens())
{
String inStr = str.nextToken();
@@ -90,31 +95,31 @@ public class MSFfile extends AlignFile
if (inStr.indexOf("Name:") != -1)
{
key = str.nextToken();
- headers.addElement(key);
+ headers.add(key);
}
- // if line has // set SeqFlag to 1 so we know sequences are coming
+ // if line has // set SeqFlag so we know sequences are coming
if (inStr.indexOf("//") != -1)
{
seqFlag = true;
}
// Process lines as sequence lines if seqFlag is set
- if ((inStr.indexOf("//") == -1) && (seqFlag == true))
+ if ((inStr.indexOf("//") == -1) && seqFlag)
{
- // seqeunce id is the first field
+ // sequence id is the first field
key = inStr;
- StringBuffer tempseq;
+ StringBuilder tempseq;
// Get sequence from hash if it exists
if (seqhash.containsKey(key))
{
- tempseq = (StringBuffer) seqhash.get(key);
+ tempseq = seqhash.get(key);
}
else
{
- tempseq = new StringBuffer();
+ tempseq = new StringBuilder(64);
seqhash.put(key, tempseq);
}
@@ -122,7 +127,8 @@ public class MSFfile extends AlignFile
while (str.hasMoreTokens())
{
// append the word to the sequence
- tempseq.append(str.nextToken());
+ String sequenceBlock = str.nextToken();
+ tempseq.append(sequenceBlock);
}
}
}
@@ -136,11 +142,11 @@ public class MSFfile extends AlignFile
this.noSeqs = headers.size();
// Add sequences to the hash
- for (i = 0; i < headers.size(); i++)
+ for (int i = 0; i < headers.size(); i++)
{
- if (seqhash.get(headers.elementAt(i)) != null)
+ if (seqhash.get(headers.get(i)) != null)
{
- String head = headers.elementAt(i).toString();
+ String head = headers.get(i);
String seq = seqhash.get(head).toString();
if (maxLength < head.length())
@@ -148,8 +154,11 @@ public class MSFfile extends AlignFile
maxLength = head.length();
}
- // Replace ~ with a sensible gap character
- seq = seq.replace('~', '-');
+ /*
+ * replace ~ (leading/trailing positions) with the gap character;
+ * use '.' as this is the internal gap character required by MSF
+ */
+ seq = seq.replace('~', '.');
Sequence newSeq = parseId(head);
@@ -160,7 +169,7 @@ public class MSFfile extends AlignFile
else
{
System.err.println("MSFFile Parser: Can't find sequence for "
- + headers.elementAt(i));
+ + headers.get(i));
}
}
}
@@ -176,7 +185,7 @@ public class MSFfile extends AlignFile
public int checkSum(String seq)
{
int check = 0;
- String sequence = seq.toUpperCase();
+ String sequence = seq.toUpperCase(Locale.ROOT);
for (int i = 0; i < sequence.length(); i++)
{
@@ -208,15 +217,17 @@ public class MSFfile extends AlignFile
*
* @return DOCUMENT ME!
*/
- public String print(SequenceI[] seqs)
+ @Override
+ public String print(SequenceI[] sqs, boolean jvSuffix)
{
- boolean is_NA = jalview.util.Comparison.isNucleotide(seqs);
+ boolean is_NA = Comparison.isNucleotide(sqs);
- SequenceI[] s = new SequenceI[seqs.length];
+ SequenceI[] s = new SequenceI[sqs.length];
- StringBuffer out = new StringBuffer("!!" + (is_NA ? "NA" : "AA")
- + "_MULTIPLE_ALIGNMENT 1.0");
+ StringBuilder out = new StringBuilder(256);
+ out.append("!!").append(is_NA ? "NA" : "AA")
+ .append("_MULTIPLE_ALIGNMENT 1.0");
// TODO: JBPNote : Jalview doesn't remember NA or AA yet.
out.append(newline);
out.append(newline);
@@ -224,14 +235,15 @@ public class MSFfile extends AlignFile
int maxid = 0;
int i = 0;
- while ((i < seqs.length) && (seqs[i] != null))
+ while ((i < sqs.length) && (sqs[i] != null))
{
- // Replace all internal gaps with . and external spaces with ~
- s[i] = new Sequence(seqs[i].getName(), seqs[i].getSequenceAsString()
- .replace('-', '.'), seqs[i].getStart(), seqs[i].getEnd());
+ /*
+ * modify to MSF format: uses '.' for internal gaps,
+ * and '~' for leading or trailing gaps
+ */
+ String seqString = sqs[i].getSequenceAsString().replace('-', '.');
- StringBuffer sb = new StringBuffer();
- sb.append(s[i].getSequence());
+ StringBuilder sb = new StringBuilder(seqString);
for (int ii = 0; ii < sb.length(); ii++)
{
@@ -256,21 +268,21 @@ public class MSFfile extends AlignFile
break;
}
}
+ s[i] = new Sequence(sqs[i].getName(), sb.toString(),
+ sqs[i].getStart(), sqs[i].getEnd());
- s[i].setSequence(sb.toString());
-
- if (s[i].getSequence().length > max)
+ if (sb.length() > max)
{
- max = s[i].getSequence().length;
+ max = sb.length();
}
i++;
}
- Format maxLenpad = new Format("%" + (new String("" + max)).length()
- + "d");
- Format maxChkpad = new Format("%" + (new String("1" + max)).length()
- + "d");
+ Format maxLenpad = new Format(
+ "%" + (new String("" + max)).length() + "d");
+ Format maxChkpad = new Format(
+ "%" + (new String("1" + max)).length() + "d");
i = 0;
int bigChecksum = 0;
@@ -283,7 +295,7 @@ public class MSFfile extends AlignFile
}
long maxNB = 0;
- out.append(" MSF: " + s[0].getSequence().length + " Type: "
+ out.append(" MSF: " + s[0].getLength() + " Type: "
+ (is_NA ? "N" : "P") + " Check: " + (bigChecksum % 10000)
+ " ..");
out.append(newline);
@@ -297,11 +309,11 @@ public class MSFfile extends AlignFile
while ((i < s.length) && (s[i] != null))
{
- nameBlock[i] = new String(" Name: " + printId(s[i]) + " ");
+ nameBlock[i] = new String(" Name: " + printId(s[i], jvSuffix) + " ");
- idBlock[i] = new String("Len: "
- + maxLenpad.form(s[i].getSequence().length) + " Check: "
- + maxChkpad.form(checksums[i]) + " Weight: 1.00" + newline);
+ idBlock[i] = new String("Len: " + maxLenpad.form(s[i].getLength())
+ + " Check: " + maxChkpad.form(checksums[i])
+ + " Weight: 1.00" + newline);
if (s[i].getName().length() > maxid)
{
@@ -341,12 +353,7 @@ public class MSFfile extends AlignFile
out.append(newline);
int len = 50;
- int nochunks = (max / len) + 1;
-
- if ((max % len) == 0)
- {
- nochunks--;
- }
+ int nochunks = (max / len) + (max % len > 0 ? 1 : 0);
for (i = 0; i < nochunks; i++)
{
@@ -354,7 +361,7 @@ public class MSFfile extends AlignFile
while ((j < s.length) && (s[j] != null))
{
- String name = printId(s[j]);
+ String name = printId(s[j], jvSuffix);
out.append(new Format("%-" + maxid + "s").form(name + " "));
@@ -363,8 +370,9 @@ public class MSFfile extends AlignFile
int start = (i * 50) + (k * 10);
int end = start + 10;
- if ((end < s[j].getSequence().length)
- && (start < s[j].getSequence().length))
+ int length = s[j].getLength();
+ if ((end < length)
+ && (start < length))
{
out.append(s[j].getSequence(start, end));
@@ -379,7 +387,7 @@ public class MSFfile extends AlignFile
}
else
{
- if (start < s[j].getSequence().length)
+ if (start < length)
{
out.append(s[j].getSequenceAsString().substring(start));
out.append(newline);
@@ -402,14 +410,4 @@ public class MSFfile extends AlignFile
return out.toString();
}
-
- /**
- * DOCUMENT ME!
- *
- * @return DOCUMENT ME!
- */
- public String print()
- {
- return print(getSeqsAsArray());
- }
}