/*
- * 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 <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.util;
-import java.util.*;
+import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.DBRefSource;
+import jalview.datamodel.PDBEntry;
+import jalview.datamodel.SequenceI;
-import jalview.datamodel.*;
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.Hashtable;
+import java.util.List;
+import java.util.Map;
+import com.stevesoft.pat.Regex;
+
+/**
+ * Utilities for handling DBRef objects and their collections.
+ */
public class DBRefUtils
{
- /**
- * Utilities for handling DBRef objects and their collections.
+ /*
+ * lookup from lower-case form of a name to its canonical (standardised) form
*/
+ private static Map<String, String> canonicalSourceNameLookup = new HashMap<String, String>();
+
+ private static Map<String, String> dasCoordinateSystemsLookup = new HashMap<String, String>();
+
+ static
+ {
+ // TODO load these from a resource file?
+ canonicalSourceNameLookup.put("uniprotkb/swiss-prot",
+ DBRefSource.UNIPROT);
+ canonicalSourceNameLookup.put("uniprotkb/trembl", DBRefSource.UNIPROT);
+
+ // Ensembl values for dbname in xref REST service:
+ canonicalSourceNameLookup.put("uniprot/sptrembl", DBRefSource.UNIPROT);
+ canonicalSourceNameLookup.put("uniprot/swissprot", DBRefSource.UNIPROT);
+
+ canonicalSourceNameLookup.put("pdb", DBRefSource.PDB);
+ canonicalSourceNameLookup.put("ensembl", DBRefSource.ENSEMBL);
+
+ dasCoordinateSystemsLookup.put("pdbresnum", DBRefSource.PDB);
+ dasCoordinateSystemsLookup.put("uniprot", DBRefSource.UNIPROT);
+ dasCoordinateSystemsLookup.put("embl", DBRefSource.EMBL);
+ // dasCoordinateSystemsLookup.put("embl", DBRefSource.EMBLCDS);
+ }
+
/**
*
* @param dbrefs
- * Vector of DBRef objects to search
+ * array of DBRef objects to search
* @param sources
* String[] array of source DBRef IDs to retrieve
- * @return Vector
+ * @return
*/
public static DBRefEntry[] selectRefs(DBRefEntry[] dbrefs,
String[] sources)
{
- if (dbrefs == null)
- {
- return null;
- }
- if (sources == null)
+ if (dbrefs == null || sources == null)
{
return dbrefs;
}
- Hashtable srcs = new Hashtable();
- Vector res = new Vector();
-
- for (int i = 0; i < sources.length; i++)
+ HashSet<String> srcs = new HashSet<String>();
+ for (String src : sources)
{
- srcs.put(new String(sources[i]), new Integer(i));
+ srcs.add(src);
}
- for (int i = 0, j = dbrefs.length; i < j; i++)
+
+ List<DBRefEntry> res = new ArrayList<DBRefEntry>();
+ for (DBRefEntry dbr : dbrefs)
{
- if (srcs.containsKey(dbrefs[i].getSource()))
+ String source = getCanonicalName(dbr.getSource());
+ if (srcs.contains(source))
{
- res.addElement(dbrefs[i]);
+ res.add(dbr);
}
}
if (res.size() > 0)
{
DBRefEntry[] reply = new DBRefEntry[res.size()];
- for (int i = 0; i < res.size(); i++)
- {
- reply[i] = (DBRefEntry) res.elementAt(i);
- }
- return reply;
+ return res.toArray(reply);
}
- res = null;
- // there are probable memory leaks in the hashtable!
return null;
}
* @return boolean true if Source DBRefEntry is compatible with DAS
* CoordinateSystem name
*/
- public static Hashtable DasCoordinateSystemsLookup = null;
public static boolean isDasCoordinateSystem(String string,
DBRefEntry dBRefEntry)
{
- if (DasCoordinateSystemsLookup == null)
- {
- // TODO: Make a DasCoordinateSystemsLookup properties resource
- // Initialise
- DasCoordinateSystemsLookup = new Hashtable();
- DasCoordinateSystemsLookup.put("pdbresnum",
- jalview.datamodel.DBRefSource.PDB);
- DasCoordinateSystemsLookup.put("uniprot",
- jalview.datamodel.DBRefSource.UNIPROT);
- DasCoordinateSystemsLookup.put("EMBL",
- jalview.datamodel.DBRefSource.EMBL);
- // DasCoordinateSystemsLookup.put("EMBL",
- // jalview.datamodel.DBRefSource.EMBLCDS);
- }
-
- String coordsys = (String) DasCoordinateSystemsLookup.get(string
- .toLowerCase());
- if (coordsys != null)
+ if (string == null || dBRefEntry == null)
{
- return coordsys.equals(dBRefEntry.getSource());
+ return false;
}
- return false;
+ String coordsys = dasCoordinateSystemsLookup.get(string.toLowerCase());
+ return coordsys == null ? false : coordsys.equals(dBRefEntry
+ .getSource());
}
- public static Hashtable CanonicalSourceNameLookup = null;
-
/**
* look up source in an internal list of database reference sources and return
* the canonical jalview name for the source, or the original string if it has
*/
public static String getCanonicalName(String source)
{
- if (CanonicalSourceNameLookup == null)
+ if (source == null)
{
- CanonicalSourceNameLookup = new Hashtable();
- CanonicalSourceNameLookup.put("uniprotkb/swiss-prot",
- jalview.datamodel.DBRefSource.UNIPROT);
- CanonicalSourceNameLookup.put("uniprotkb/trembl",
- jalview.datamodel.DBRefSource.UNIPROT);
- CanonicalSourceNameLookup.put("pdb",
- jalview.datamodel.DBRefSource.PDB);
- }
- String canonical = (String) CanonicalSourceNameLookup.get(source
- .toLowerCase());
- if (canonical == null)
- {
- return source;
+ return null;
}
- return canonical;
+ String canonical = canonicalSourceNameLookup.get(source.toLowerCase());
+ return canonical == null ? source : canonical;
}
/**
- * find RefEntry corresponding to a particular pattern the equals method of
- * each entry is used, from String attributes right down to Mapping
- * attributes.
+ * Returns an array of those references that match the given entry, or null if
+ * no matches. Currently uses a comparator which matches if
+ * <ul>
+ * <li>database sources are the same</li>
+ * <li>accession ids are the same</li>
+ * <li>both have no mapping, or the mappings are the same</li>
+ * </ul>
*
* @param ref
* Set of references to search
* @param entry
- * pattern to collect - null any entry for wildcard match
+ * pattern to match
* @return
*/
public static DBRefEntry[] searchRefs(DBRefEntry[] ref, DBRefEntry entry)
matchDbAndIdAndEitherMapOrEquivalentMapList);
}
- public static DBRefEntry[] searchRefs(DBRefEntry[] ref, DBRefEntry entry,
+ /**
+ * Returns an array of those references that match the given accession id
+ * <ul>
+ * <li>database sources are the same</li>
+ * <li>accession ids are the same</li>
+ * <li>both have no mapping, or the mappings are the same</li>
+ * </ul>
+ *
+ * @param ref
+ * Set of references to search
+ * @param entry
+ * pattern to match
+ * @return
+ */
+ public static DBRefEntry[] searchRefs(DBRefEntry[] ref, String accId)
+ {
+ return searchRefs(ref, new DBRefEntry("", "", accId), matchId);
+ }
+
+ /**
+ * Returns an array of those references that match the given entry, according
+ * to the given comparator. Returns null if no matches.
+ *
+ * @param refs
+ * an array of database references to search
+ * @param entry
+ * an entry to compare against
+ * @param comparator
+ * @return
+ */
+ static DBRefEntry[] searchRefs(DBRefEntry[] refs, DBRefEntry entry,
DbRefComp comparator)
{
- if (ref == null || entry == null)
+ if (refs == null || entry == null)
+ {
return null;
- Vector rfs = new Vector();
- for (int i = 0; i < ref.length; i++)
+ }
+ List<DBRefEntry> rfs = new ArrayList<DBRefEntry>();
+ for (int i = 0; i < refs.length; i++)
{
- if (comparator.matches(entry, ref[i]))
+ if (comparator.matches(entry, refs[i]))
{
- rfs.addElement(ref[i]);
+ rfs.add(refs[i]);
}
}
- // TODO Auto-generated method stub
- if (rfs.size() > 0)
- {
- DBRefEntry[] rf = new DBRefEntry[rfs.size()];
- rfs.copyInto(rf);
- return rf;
- }
- return null;
+ return rfs.size() == 0 ? null : rfs.toArray(new DBRefEntry[rfs.size()]);
}
- public interface DbRefComp
+ interface DbRefComp
{
public boolean matches(DBRefEntry refa, DBRefEntry refb);
}
/**
* match on all non-null fields in refa
*/
+ // TODO unused - remove?
public static DbRefComp matchNonNullonA = new DbRefComp()
{
+ @Override
public boolean matches(DBRefEntry refa, DBRefEntry refb)
{
if (refa.getSource() == null
* either field is null or field matches for all of source, version, accession
* id and map.
*/
+ // TODO unused - remove?
public static DbRefComp matchEitherNonNull = new DbRefComp()
{
+ @Override
public boolean matches(DBRefEntry refa, DBRefEntry refb)
{
- if ((refa.getSource() == null || refb.getSource() == null)
- || refb.getSource().equals(refa.getSource()))
+ if (nullOrEqual(refa.getSource(), refb.getSource())
+ && nullOrEqual(refa.getVersion(), refb.getVersion())
+ && nullOrEqual(refa.getAccessionId(), refb.getAccessionId())
+ && nullOrEqual(refa.getMap(), refb.getMap()))
{
- if ((refa.getVersion() == null || refb.getVersion() == null)
- || refb.getVersion().equals(refa.getVersion()))
- {
- if ((refa.getAccessionId() == null || refb.getAccessionId() == null)
- || refb.getAccessionId().equals(refa.getAccessionId()))
- {
- if ((refa.getMap() == null || refb.getMap() == null)
- || (refb.getMap() != null && refb.getMap().equals(
- refa.getMap())))
- {
- return true;
- }
- }
- }
+ return true;
}
return false;
}
* accession ID and DB must be identical. Version is ignored. Map is either
* not defined or is a match (or is compatible?)
*/
+ // TODO unused - remove?
public static DbRefComp matchDbAndIdAndEitherMap = new DbRefComp()
{
+ @Override
public boolean matches(DBRefEntry refa, DBRefEntry refb)
{
if (refa.getSource() != null && refb.getSource() != null
&& refb.getSource().equals(refa.getSource()))
{
// We dont care about version
- // if ((refa.getVersion()==null || refb.getVersion()==null)
- // || refb.getVersion().equals(refa.getVersion()))
- // {
if (refa.getAccessionId() != null && refb.getAccessionId() != null
+ // FIXME should be && not || here?
|| refb.getAccessionId().equals(refa.getAccessionId()))
{
if ((refa.getMap() == null || refb.getMap() == null)
* or map but no maplist on either or maplist of map on a is the complement of
* maplist of map on b.
*/
+ // TODO unused - remove?
public static DbRefComp matchDbAndIdAndComplementaryMapList = new DbRefComp()
{
+ @Override
public boolean matches(DBRefEntry refa, DBRefEntry refb)
{
if (refa.getSource() != null && refb.getSource() != null
&& refb.getSource().equals(refa.getSource()))
{
// We dont care about version
- // if ((refa.getVersion()==null || refb.getVersion()==null)
- // || refb.getVersion().equals(refa.getVersion()))
- // {
if (refa.getAccessionId() != null && refb.getAccessionId() != null
|| refb.getAccessionId().equals(refa.getAccessionId()))
{
if ((refa.getMap() == null && refb.getMap() == null)
|| (refa.getMap() != null && refb.getMap() != null))
+ {
if ((refb.getMap().getMap() == null && refa.getMap().getMap() == null)
|| (refb.getMap().getMap() != null
&& refa.getMap().getMap() != null && refb
{
return true;
}
+ }
}
}
return false;
* or or map but no maplist on either or maplist of map on a is equivalent to
* the maplist of map on b.
*/
+ // TODO unused - remove?
public static DbRefComp matchDbAndIdAndEquivalentMapList = new DbRefComp()
{
+ @Override
public boolean matches(DBRefEntry refa, DBRefEntry refb)
{
if (refa.getSource() != null && refb.getSource() != null
*/
public static DbRefComp matchDbAndIdAndEitherMapOrEquivalentMapList = new DbRefComp()
{
+ @Override
public boolean matches(DBRefEntry refa, DBRefEntry refb)
{
- // System.err.println("Comparing A: "+refa.getSrcAccString()+(refa.hasMap()?" has map.":"."));
- // System.err.println("Comparing B: "+refb.getSrcAccString()+(refb.hasMap()?" has map.":"."));
if (refa.getSource() != null && refb.getSource() != null
&& refb.getSource().equals(refa.getSource()))
{
// We dont care about version
- // if ((refa.getVersion()==null || refb.getVersion()==null)
- // || refb.getVersion().equals(refa.getVersion()))
- // {
if (refa.getAccessionId() != null && refb.getAccessionId() != null
&& refb.getAccessionId().equals(refa.getAccessionId()))
{
};
/**
- * used by file parsers to generate DBRefs from annotation within file (eg
- * stockholm)
+ * accession ID only must be identical.
+ */
+ public static DbRefComp matchId = new DbRefComp()
+ {
+ @Override
+ public boolean matches(DBRefEntry refa, DBRefEntry refb)
+ {
+ if (refa.getAccessionId() != null && refb.getAccessionId() != null
+ && refb.getAccessionId().equals(refa.getAccessionId()))
+ {
+ return true;
+ }
+ return false;
+ }
+ };
+
+ /**
+ * Parses a DBRefEntry and adds it to the sequence, also a PDBEntry if the
+ * database is PDB.
+ * <p>
+ * Used by file parsers to generate DBRefs from annotation within file (eg
+ * Stockholm)
*
* @param dbname
* @param version
* @param acn
* @param seq
- * where to anotate with reference
+ * where to annotate with reference
* @return parsed version of entry that was added to seq (if any)
*/
public static DBRefEntry parseToDbRef(SequenceI seq, String dbname,
DBRefEntry ref = null;
if (dbname != null)
{
- String locsrc = jalview.util.DBRefUtils.getCanonicalName(dbname);
- if (locsrc.equals(jalview.datamodel.DBRefSource.PDB))
+ String locsrc = DBRefUtils.getCanonicalName(dbname);
+ if (locsrc.equals(DBRefSource.PDB))
{
- // check for chaincode and mapping
- // PFAM style stockhom PDB citation
- com.stevesoft.pat.Regex r = new com.stevesoft.pat.Regex(
- "([0-9][0-9A-Za-z]{3})\\s*(.?)\\s*;([0-9]+)-([0-9]+)");
+ /*
+ * Check for PFAM style stockhom PDB accession id citation e.g.
+ * "1WRI A; 7-80;"
+ */
+ Regex r = new com.stevesoft.pat.Regex(
+ "([0-9][0-9A-Za-z]{3})\\s*(.?)\\s*;\\s*([0-9]+)-([0-9]+)");
if (r.search(acn.trim()))
{
String pdbid = r.stringMatched(1);
String chaincode = r.stringMatched(2);
- String mapstart = r.stringMatched(3);
- String mapend = r.stringMatched(4);
+ if (chaincode == null)
+ {
+ chaincode = " ";
+ }
+ // String mapstart = r.stringMatched(3);
+ // String mapend = r.stringMatched(4);
if (chaincode.equals(" "))
{
chaincode = "_";
ref = new DBRefEntry(locsrc, version, pdbid + chaincode);
PDBEntry pdbr = new PDBEntry();
pdbr.setId(pdbid);
- pdbr.getProperty().put("CHAIN",chaincode);
+ pdbr.setType(PDBEntry.Type.PDB);
+ pdbr.setProperty(new Hashtable());
+ pdbr.setChainCode(chaincode);
+ // pdbr.getProperty().put("CHAIN", chaincode);
seq.addPDBId(pdbr);
}
+ else
+ {
+ System.err.println("Malformed PDB DR line:" + acn);
+ }
}
else
{
return ref;
}
+ /**
+ * Returns true if either object is null, or they are equal
+ *
+ * @param o1
+ * @param o2
+ * @return
+ */
+ public static boolean nullOrEqual(Object o1, Object o2)
+ {
+ if (o1 == null || o2 == null)
+ {
+ return true;
+ }
+ return (o1 == null ? o2.equals(o1) : o1.equals(o2));
+ }
+
}