-package jalview.io;\r
-\r
-import jalview.datamodel.*;\r
-import jalview.util.*;\r
-\r
-import java.io.*;\r
-import java.util.*;\r
-\r
-public class PfamFile extends AlignFile {\r
-\r
- Vector ids;\r
-\r
- public PfamFile()\r
- {}\r
-\r
- public PfamFile(String inStr) {\r
- super(inStr);\r
- }\r
-\r
- public void initData() {\r
- super.initData();\r
- ids = new Vector();\r
- }\r
-\r
- public PfamFile(String inFile, String type) throws IOException {\r
- super(inFile,type);\r
- }\r
-\r
- public void parse() throws IOException{\r
- int i = 0;\r
- String line;\r
-\r
-\r
- Hashtable seqhash = new Hashtable();\r
- Vector headers = new Vector();\r
-\r
-\r
- while ((line = nextLine()) != null)\r
- {\r
-\r
- if (line.indexOf(" ") != 0)\r
- {\r
- if (line.indexOf("#") != 0)\r
- {\r
-\r
- StringTokenizer str = new StringTokenizer(line," ");\r
- String id = "";\r
-\r
- if (str.hasMoreTokens())\r
- {\r
- id = str.nextToken();\r
-\r
- StringBuffer tempseq;\r
-\r
- if (seqhash.containsKey(id))\r
- tempseq = (StringBuffer)seqhash.get(id);\r
- else\r
- {\r
- tempseq = new StringBuffer();\r
- seqhash.put(id,tempseq);\r
- }\r
-\r
- if (!(headers.contains(id)))\r
- headers.addElement(id);\r
-\r
-\r
- tempseq.append(str.nextToken());\r
- }\r
- }\r
- }\r
- }\r
-\r
- this.noSeqs = headers.size();\r
- if(noSeqs<1)\r
- throw new IOException("No sequences found (PFAM input)");\r
-\r
- for (i = 0; i < headers.size(); i++ ) {\r
-\r
- if ( seqhash.get(headers.elementAt(i)) != null) {\r
- if (maxLength < seqhash.get(headers.elementAt(i)).toString().length() )\r
- maxLength = seqhash.get(headers.elementAt(i)).toString().length();\r
-\r
- String head = headers.elementAt(i).toString();\r
- int start = 1;\r
- int end = seqhash.get(headers.elementAt(i)).toString().length();\r
-\r
- if (head.indexOf("/") > 0 ) {\r
- StringTokenizer st = new StringTokenizer(head,"/");\r
- if (st.countTokens() == 2) {\r
- ids.addElement(st.nextToken());\r
- String tmp = st.nextToken();\r
- st = new StringTokenizer(tmp,"-");\r
- if (st.countTokens() == 2) {\r
- start = Integer.valueOf(st.nextToken()).intValue();\r
- end = Integer.valueOf(st.nextToken()).intValue();\r
- } else\r
- {\r
- start = -1;\r
- end = -1;\r
- }\r
- } else\r
- ids.addElement(headers.elementAt(i));\r
-\r
- }\r
- else\r
- ids.addElement(headers.elementAt(i));\r
-\r
-\r
- Sequence newSeq = null;\r
- if (start != -1 && end != -1)\r
- {\r
- newSeq = new Sequence(ids.elementAt(i).toString(),\r
- seqhash.get(headers.elementAt(i).toString()).toString(),start,end);\r
- seqs.addElement(newSeq);\r
- }\r
- else\r
- {\r
- newSeq = new Sequence(ids.elementAt(i).toString(),\r
- seqhash.get(headers.elementAt(i).toString()).toString(),1,\r
- seqhash.get(headers.elementAt(i).toString()).toString().length());\r
- seqs.addElement(newSeq);\r
- }\r
-\r
- if(!isValidProteinSequence(newSeq.getSequence()))\r
- throw new IOException("Not a valid protein sequence - (PFAM input)");\r
- }\r
- else\r
- System.err.println("PFAM File reader: Can't find sequence for " + headers.elementAt(i));\r
-\r
- }\r
-\r
- }\r
-\r
- public static String print(SequenceI[] s) {\r
- StringBuffer out = new StringBuffer("");\r
-\r
- int max = 0;\r
- int maxid = 0;\r
-\r
- int i = 0;\r
-\r
- while (i < s.length && s[i] != null) {\r
- String tmp = s[i].getName() + "/" + s[i].getStart()+ "-" + s[i].getEnd();\r
-\r
- if (s[i].getSequence().length() > max) {\r
- max = s[i].getSequence().length();\r
- }\r
- if (tmp.length() > maxid) {\r
- maxid = tmp.length();\r
- }\r
- i++;\r
- }\r
-\r
- if (maxid < 15) {\r
- maxid = 15;\r
- }\r
-\r
- int j = 0;\r
- while ( j < s.length && s[j] != null) {\r
- out.append( new Format("%-" + maxid + "s").form(s[j].getName() + "/" + s[j].getStart() + "-" + s[j].getEnd() ) + " ");\r
-\r
- out.append(s[j].getSequence() + "\n");\r
- j++;\r
- }\r
- out.append("\n");\r
-\r
- return out.toString();\r
- }\r
-\r
- public String print() {\r
- return print(getSeqsAsArray());\r
- }\r
-}\r
+/*
+ * 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.
+ *
+ * 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/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
+package jalview.io;
+
+import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceI;
+import jalview.util.Format;
+import jalview.util.MessageManager;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.HashMap;
+
+public class PfamFile extends AlignFile
+{
+
+ public PfamFile()
+ {
+ }
+
+ public PfamFile(String inFile, String type) throws IOException
+ {
+ super(inFile, type);
+ }
+
+ public PfamFile(FileParse source) throws IOException
+ {
+ super(source);
+ }
+
+ @Override
+ public void initData()
+ {
+ super.initData();
+ }
+
+ @Override
+ public void parse() throws IOException
+ {
+ int i = 0;
+ String line;
+
+ HashMap<String, StringBuffer> seqhash = new HashMap<String, StringBuffer>();
+ ArrayList<String> headers = new ArrayList<String>();
+ boolean useTabs = false;
+ int spces;
+ while ((line = nextLine()) != null)
+ {
+ if (line.indexOf("#") == 0)
+ {
+ // skip comment lines
+ continue;
+ }
+ // locate first space or (if already checked), tab
+ if (useTabs)
+ {
+ spces = line.indexOf("\t");
+ }
+ else
+ {
+ spces = line.indexOf(" ");
+ // check to see if we ought to split on tabs instead.
+ if (!useTabs && spces == -1)
+ {
+ useTabs = true;
+ spces = line.indexOf("\t");
+ }
+ }
+ if (spces <= 0)
+ {
+ // no sequence data to split on
+ continue;
+ }
+ String id = line.substring(0, spces);
+ StringBuffer tempseq;
+
+ if (seqhash.containsKey(id))
+ {
+ tempseq = seqhash.get(id);
+ }
+ else
+ {
+ tempseq = new StringBuffer();
+ seqhash.put(id, tempseq);
+ }
+
+ if (!(headers.contains(id)))
+ {
+ headers.add(id);
+ }
+ if (spces + 1 < line.length())
+ {
+ tempseq.append(line.substring(spces + 1).trim());
+ }
+ }
+
+ this.noSeqs = headers.size();
+
+ if (noSeqs < 1)
+ {
+ throw new IOException(
+ MessageManager.getString("exception.pfam_no_sequences_found"));
+ }
+
+ for (i = 0; i < headers.size(); i++)
+ {
+ if (seqhash.get(headers.get(i)) != null)
+ {
+ if (maxLength < seqhash.get(headers.get(i)).toString().length())
+ {
+ maxLength = seqhash.get(headers.get(i)).toString().length();
+ }
+
+ Sequence newSeq = parseId(headers.get(i).toString());
+ newSeq.setSequence(seqhash.get(headers.get(i).toString())
+ .toString());
+ seqs.addElement(newSeq);
+ }
+ else
+ {
+ System.err.println("PFAM File reader: Can't find sequence for "
+ + headers.get(i));
+ }
+ }
+ }
+
+ public String print(SequenceI[] s)
+ {
+ StringBuffer out = new StringBuffer("");
+
+ int max = 0;
+ int maxid = 0;
+
+ int i = 0;
+
+ while ((i < s.length) && (s[i] != null))
+ {
+ String tmp = printId(s[i]);
+
+ if (s[i].getSequence().length > max)
+ {
+ max = s[i].getSequence().length;
+ }
+
+ if (tmp.length() > maxid)
+ {
+ maxid = tmp.length();
+ }
+
+ i++;
+ }
+
+ if (maxid < 15)
+ {
+ maxid = 15;
+ }
+
+ int j = 0;
+
+ while ((j < s.length) && (s[j] != null))
+ {
+ out.append(new Format("%-" + maxid + "s").form(printId(s[j]) + " "));
+
+ out.append(s[j].getSequenceAsString());
+ out.append(newline);
+ j++;
+ }
+
+ out.append(newline);
+
+ return out.toString();
+ }
+
+ @Override
+ public String print()
+ {
+ return print(getSeqsAsArray());
+ }
+}
git://source.jalview.org / jalview.git / blobdiff