/*
- * 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.analysis;
-import java.util.*;
+import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.SequenceI;
-import jalview.datamodel.*;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
/**
* Routines for approximate Sequence Id resolution by name using string
*/
public class SequenceIdMatcher
{
- private Hashtable names;
+ private HashMap<SeqIdName, List<SequenceI>> names;
+
+ private Map<SeqIdName, List<SequenceI>> excludes;
+
+ public SequenceIdMatcher(List<SequenceI> seqs)
+ {
+ names = new HashMap<>();
+ excludes = new HashMap<>();
+ addAll(seqs);
+ }
+
+ /**
+ * Adds sequences to this matcher
+ *
+ * @param seqs
+ */
+ public void addAll(List<SequenceI> seqs)
+ {
+ for (SequenceI seq : seqs)
+ {
+ add(seq);
+ }
+ }
- public SequenceIdMatcher(SequenceI[] seqs)
+ /**
+ * Adds one sequence to this matcher
+ *
+ * @param seq
+ */
+ public void add(SequenceI seq)
{
- names = new Hashtable();
- for (int i = 0; i < seqs.length; i++)
+ SeqIdName key = new SeqIdName(seq.getDisplayId(true));
+ addMatchCandidate(key, seq);
+ SequenceI dbseq = seq;
+ while (dbseq.getDatasetSequence() != null)
{
- // TODO: deal with ID collisions - SequenceI should be appended to list
- // associated with this key.
- names.put(new SeqIdName(seqs[i].getDisplayId(true)), seqs[i]);
- // add in any interesting identifiers
- if (seqs[i].getDBRef() != null)
+ dbseq = dbseq.getDatasetSequence();
+ }
+ // add in any interesting identifiers
+ if (dbseq.getDBRefs() != null)
+ {
+ DBRefEntry dbr[] = dbseq.getDBRefs();
+ for (int r = 0; r < dbr.length; r++)
{
- DBRefEntry dbr[] = seqs[i].getDBRef();
- SeqIdName sid = null;
- for (int r = 0; r < dbr.length; r++)
+ DBRefEntry dbref = dbr[r];
+ SeqIdName sid = new SeqIdName(dbref.getAccessionId());
+ if (dbref.getMap() != null
+ && dbref.getMap().getMap().isTripletMap())
{
- sid = new SeqIdName(dbr[r].getAccessionId());
- if (!names.contains(sid))
+ /*
+ * dbref with 3:1 or 1:3 mapping (e.g. CDS/protein);
+ * mark as not a valid match for this id
+ */
+ List<SequenceI> excluded = excludes.get(sid);
+ if (excluded == null)
{
- names.put(sid, seqs[i]);
+ excludes.put(sid, excluded = new ArrayList<>());
}
+ excluded.add(seq);
+ System.out.println("Excluding " + sid + "->" + seq);
+ continue;
}
+ addMatchCandidate(sid, seq);
}
}
}
+ void addMatchCandidate(SeqIdName key, SequenceI seq)
+ {
+ List<SequenceI> namesList = names.get(key);
+ if (namesList == null)
+ {
+ names.put(key, namesList = new ArrayList<>());
+ }
+ if (!namesList.contains(seq))
+ {
+ namesList.add(seq);
+ System.out.println("Adding " + key + "->" + seq);
+ }
+ }
+
+ /**
+ * convenience method to make a matcher from concrete array
+ *
+ * @param sequences
+ */
+ public SequenceIdMatcher(SequenceI[] sequences)
+ {
+ this(Arrays.asList(sequences));
+ }
+
/**
* returns the closest SequenceI in matches to SeqIdName and returns all the
* matches to the names hash.
* @param candName
* SeqIdName
* @param matches
- * Vector of SequenceI objects
+ * List of SequenceI objects
* @return SequenceI closest SequenceI to SeqIdName
*/
- private SequenceI pickbestMatch(SeqIdName candName, Vector matches)
+ private SequenceI pickbestMatch(SeqIdName candName,
+ List<SequenceI> matches)
{
- SequenceI[] st = pickbestMatches(candName, matches);
- return st == null || st.length == 0 ? null : st[0];
+ List<SequenceI> st = pickbestMatches(candName, matches);
+ return st == null || st.size() == 0 ? null : st.get(0);
}
/**
* @return Object[] { SequenceI closest SequenceI to SeqIdName, SequenceI[]
* ties }
*/
- private SequenceI[] pickbestMatches(SeqIdName candName, Vector matches)
+ private List<SequenceI> pickbestMatches(SeqIdName candName,
+ List<SequenceI> matches)
{
- ArrayList best = new ArrayList();
- SequenceI match = null;
+ List<SequenceI> best = new ArrayList<>();
if (candName == null || matches == null || matches.size() == 0)
{
return null;
}
- match = (SequenceI) matches.elementAt(0);
- matches.removeElementAt(0);
+ SequenceI match = matches.remove(0);
best.add(match);
- names.put(new SeqIdName(match.getName()), match);
+ addMatchCandidate(new SeqIdName(match.getName()), match);
int matchlen = match.getName().length();
int namlen = candName.id.length();
while (matches.size() > 0)
{
// look through for a better one.
- SequenceI cand = (SequenceI) matches.elementAt(0);
- matches.remove(0);
- names.put(new SeqIdName(cand.getName()), cand);
+ SequenceI cand = matches.remove(0);
+ addMatchCandidate(new SeqIdName(cand.getName()), cand);
int q, w, candlen = cand.getName().length();
// keep the one with an id 'closer' to the given seqnam string
- if ((q = Math.abs(matchlen - namlen)) > (w = Math.abs(candlen
- - namlen))
- && candlen > matchlen)
+ if ((q = Math.abs(matchlen - namlen)) > (w = Math
+ .abs(candlen - namlen)) && candlen > matchlen)
{
best.clear();
match = cand;
return null;
}
;
- return (SequenceI[]) best.toArray(new SequenceI[0]);
+ return best;
}
/**
*
* @param seqnam
* string to query Matcher with.
+ * @return a new array or (possibly) null
*/
public SequenceI[] findAllIdMatches(String seqnam)
{
SeqIdName nam = new SeqIdName(seqnam);
- return findAllIdMatches(nam);
+ List<SequenceI> m = findAllIdMatches(nam);
+ if (m != null)
+ {
+ return m.toArray(new SequenceI[m.size()]);
+ }
+ return null;
}
/**
* SeqIdName
* @return SequenceI
*/
- private SequenceI findIdMatch(
- jalview.analysis.SequenceIdMatcher.SeqIdName nam)
+ private SequenceI findIdMatch(SeqIdName nam)
{
- Vector matches = new Vector();
+ List<SequenceI> matches = new ArrayList<>();
while (names.containsKey(nam))
{
- matches.addElement(names.remove(nam));
+ List<SequenceI> candidates = names.remove(nam);
+ List<SequenceI> except = excludes.get(nam);
+ int j = candidates.size();
+ for (int i = 0; i < j; i++)
+ {
+ SequenceI candidate = candidates.get(i);
+ if (!except.contains(candidate))
+ {
+ matches.add(candidate);
+ }
+ }
}
return pickbestMatch(nam, matches);
}
* SeqIdName
* @return SequenceI[]
*/
- private SequenceI[] findAllIdMatches(
+ private List<SequenceI> findAllIdMatches(
jalview.analysis.SequenceIdMatcher.SeqIdName nam)
{
- Vector matches = new Vector();
+ List<SequenceI> matches = new ArrayList<>();
while (names.containsKey(nam))
{
- matches.addElement(names.remove(nam));
+ matches.addAll(names.remove(nam));
}
- SequenceI[] r = pickbestMatches(nam, matches);
+ List<SequenceI> r = pickbestMatches(nam, matches);
return r;
}
- private class SeqIdName
+ class SeqIdName
{
String id;
{
if (s != null)
{
- id = new String(s);
+ id = s.toLowerCase();
}
else
{
}
}
+ @Override
public int hashCode()
{
- return ((id.length() >= 4) ? id.substring(0, 4).hashCode() : id
- .hashCode());
+ return ((id.length() >= 4) ? id.substring(0, 4).hashCode()
+ : id.hashCode());
}
+ @Override
public boolean equals(Object s)
{
+ if (s == null)
+ {
+ return false;
+ }
if (s instanceof SeqIdName)
{
- return this.equals((SeqIdName) s);
+ return this.stringequals(((SeqIdName) s).id);
}
else
{
if (s instanceof String)
{
- return this.equals((String) s);
+ return this.stringequals(((String) s).toLowerCase());
}
}
* todo: (JBPNote) Set separator characters appropriately
*
* @param s
- * SeqIdName
* @return boolean
*/
- public boolean equals(SeqIdName s)
+ private boolean stringequals(String s)
{
- // TODO: JAL-732 patch for cases when name includes a list of IDs, and the
- // match contains one ID flanked
- if (id.length() > s.id.length())
+ if (id.length() > s.length())
{
- return id.startsWith(s.id) ? (WORD_SEP.indexOf(id.charAt(s.id
- .length())) > -1) : false;
+ return id.startsWith(s)
+ ? (WORD_SEP.indexOf(id.charAt(s.length())) > -1)
+ : false;
}
else
{
- return s.id.startsWith(id) ? (s.id.equals(id) ? true : (WORD_SEP
- .indexOf(s.id.charAt(id.length())) > -1)) : false;
+ return s.startsWith(id)
+ ? (s.equals(id) ? true
+ : (WORD_SEP.indexOf(s.charAt(id.length())) > -1))
+ : false;
}
}
- public boolean equals(String s)
+ /**
+ * toString method returns the wrapped sequence id. For debugging purposes
+ * only, behaviour not guaranteed not to change.
+ */
+ @Override
+ public String toString()
{
- if (id.length() > s.length())
- {
- return id.startsWith(s) ? (WORD_SEP.indexOf(id.charAt(s.length())) > -1)
- : false;
- }
- else
- {
- return s.startsWith(id) ? (s.equals(id) ? true : (WORD_SEP
- .indexOf(s.charAt(id.length())) > -1)) : false;
- }
+ return id;
}
}
}
git://source.jalview.org / jalview.git / blobdiff