X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FSequenceIdMatcher.java;h=98886472f1de8e750b46438ddb1680a81fa206c8;hb=63cead20b09743e899a22cb89f1c09e4d41cc8c0;hp=f4f305b08da6a19e53c2bd61c0f1c53d4f1e787b;hpb=fbec1b33d0fc169d72be059a5d2cf12b248270e1;p=jalview.git
diff --git a/src/jalview/analysis/SequenceIdMatcher.java b/src/jalview/analysis/SequenceIdMatcher.java
index f4f305b..9888647 100755
--- a/src/jalview/analysis/SequenceIdMatcher.java
+++ b/src/jalview/analysis/SequenceIdMatcher.java
@@ -1,185 +1,411 @@
-/*
- * Jalview - A Sequence Alignment Editor and Viewer
- * Copyright (C) 2005 AM Waterhouse, J Procter, G Barton, M Clamp, S Searle
- *
- * This program 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 2
- * of the License, or (at your option) any later version.
- *
- * This program 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 this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
- */
-package jalview.analysis;
-
-import java.util.*;
-
-import jalview.datamodel.*;
-
-/**
- *
Title:
- * SequenceIdMatcher
- * Description:
- * Routine which does approximate Sequence Id resolution by name using
- * string containment (on word boundaries) rather than equivalence
- * Copyright: Copyright (c) 2004
- *
- * Company: Dundee University
- *
- * @author not attributable
- * @version 1.0
- */
-public class SequenceIdMatcher
-{
- private Hashtable names;
-
- public SequenceIdMatcher(SequenceI[] seqs)
- {
- names = new Hashtable();
- for (int i = 0; i < seqs.length; i++)
- {
- names.put(new SeqIdName(seqs[i].getName()), seqs[i]);
- }
- }
-
- SequenceI findIdMatch(SequenceI seq)
- {
- SeqIdName nam = new SeqIdName(seq.getName());
-
- if (names.containsKey(nam))
- {
- return (SequenceI) names.get(nam);
- }
-
- return null;
- }
-
- SequenceI findIdMatch(String seqnam)
- {
- SeqIdName nam = new SeqIdName(seqnam);
-
- if (names.containsKey(nam))
- {
- return (SequenceI) names.get(nam);
- }
-
- return null;
- }
-
- /**
- * findIdMatch
- *
- * Return pointers to sequences (or sequence object containers)
- * which have same Id as a given set of different sequence objects
- *
- * @param seqs SequenceI[]
- * @return SequenceI[]
- */
- SequenceI[] findIdMatch(SequenceI[] seqs)
- {
- SequenceI[] namedseqs = null;
- int i = 0;
- SeqIdName nam;
-
- if (seqs.length > 0)
- {
- namedseqs = new SequenceI[seqs.length];
- do
- {
- nam = new SeqIdName(seqs[i].getName());
-
- if (names.containsKey(nam))
- {
- namedseqs[i] = (SequenceI) names.get(nam);
- }
- else
- {
- namedseqs[i] = null;
- }
- }
- while (++i < seqs.length);
- }
-
- return namedseqs;
- }
-
- private class SeqIdName
- {
- String id;
-
- SeqIdName(String s)
- {
- if (s!=null)
- id = new String(s);
- else
- id = "";
- }
-
- public int hashCode()
- {
- return ((id.length()>=4) ? id.substring(0, 4).hashCode() : id.hashCode());
- }
-
- public boolean equals(Object s)
- {
- if (s instanceof SeqIdName)
- {
- return this.equals( (SeqIdName) s);
- }
- else
- {
- if (s instanceof String)
- {
- return this.equals( (String) s);
- }
- }
-
- return false;
- }
-
- /**
- * Characters that define the end of a unique sequence ID at
- * the beginning of an arbitrary ID string
- */
- private String WORD_SEP="~. |#\\/<>!\"£$%^*)}[@',?";
-
- /**
- * matches if one ID properly contains another at a whitespace boundary.
- * TODO: (JBPNote) These are not efficient. should use char[] for speed
- * todo: (JBPNote) Set separator characters appropriately
- * @param s SeqIdName
- * @return boolean
- */
- public boolean equals(SeqIdName s)
- {
- if (id.length()>s.id.length()) {
- return id.startsWith(s.id) ?
- (WORD_SEP.indexOf(id.charAt(s.id.length()))>-1)
- : false;
- } else
- return s.id.startsWith(id) ?
- (s.id.equals(id) ? true :
- (WORD_SEP.indexOf(s.id.charAt(id.length()))>-1))
- : false;
- }
-
- public boolean equals(String s)
- {
- 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;
- }
- }
-}
+/*
+ * 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 .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
+package jalview.analysis;
+
+import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.SequenceI;
+
+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
+ * containment (on word boundaries) rather than equivalence. It also attempts to
+ * resolve ties where no exact match is available by picking the the id closest
+ * to the query.
+ */
+public class SequenceIdMatcher
+{
+ private HashMap> names;
+
+ private Map> excludes;
+
+ public SequenceIdMatcher(List seqs)
+ {
+ names = new HashMap<>();
+ excludes = new HashMap<>();
+ addAll(seqs);
+ }
+
+ /**
+ * Adds sequences to this matcher
+ *
+ * @param seqs
+ */
+ public void addAll(List seqs)
+ {
+ for (SequenceI seq : seqs)
+ {
+ add(seq);
+ }
+ }
+
+ /**
+ * Adds one sequence to this matcher
+ *
+ * @param seq
+ */
+ public void add(SequenceI seq)
+ {
+ SeqIdName key = new SeqIdName(seq.getDisplayId(true));
+ addMatchCandidate(key, seq);
+ SequenceI dbseq = seq;
+ while (dbseq.getDatasetSequence() != 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 dbref = dbr[r];
+ SeqIdName sid = new SeqIdName(dbref.getAccessionId());
+ if (dbref.getMap() != null
+ && dbref.getMap().getMap().isTripletMap())
+ {
+ /*
+ * dbref with 3:1 or 1:3 mapping (e.g. CDS/protein);
+ * mark as not a valid match for this id
+ */
+ List excluded = excludes.get(sid);
+ if (excluded == null)
+ {
+ 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 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
+ * List of SequenceI objects
+ * @return SequenceI closest SequenceI to SeqIdName
+ */
+ private SequenceI pickbestMatch(SeqIdName candName,
+ List matches)
+ {
+ List st = pickbestMatches(candName, matches);
+ return st == null || st.size() == 0 ? null : st.get(0);
+ }
+
+ /**
+ * 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
+ * @return Object[] { SequenceI closest SequenceI to SeqIdName, SequenceI[]
+ * ties }
+ */
+ private List pickbestMatches(SeqIdName candName,
+ List matches)
+ {
+ List best = new ArrayList<>();
+ if (candName == null || matches == null || matches.size() == 0)
+ {
+ return null;
+ }
+ SequenceI match = matches.remove(0);
+ best.add(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 = 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)
+ {
+ best.clear();
+ match = cand;
+ matchlen = candlen;
+ best.add(match);
+ }
+ if (q == w && candlen == matchlen)
+ {
+ // record any ties
+ best.add(cand);
+ }
+ }
+ if (best.size() == 0)
+ {
+ return null;
+ }
+ ;
+ return best;
+ }
+
+ /**
+ * get SequenceI with closest SequenceI.getName() to seq.getName()
+ *
+ * @param seq
+ * SequenceI
+ * @return SequenceI
+ */
+ public SequenceI findIdMatch(SequenceI seq)
+ {
+ SeqIdName nam = new SeqIdName(seq.getName());
+ return findIdMatch(nam);
+ }
+
+ public SequenceI findIdMatch(String seqnam)
+ {
+ SeqIdName nam = new SeqIdName(seqnam);
+ return findIdMatch(nam);
+ }
+
+ /**
+ * Find all matches for a given sequence name.
+ *
+ * @param seqnam
+ * string to query Matcher with.
+ * @return a new array or (possibly) null
+ */
+ public SequenceI[] findAllIdMatches(String seqnam)
+ {
+
+ SeqIdName nam = new SeqIdName(seqnam);
+ List m = findAllIdMatches(nam);
+ if (m != null)
+ {
+ return m.toArray(new SequenceI[m.size()]);
+ }
+ return null;
+ }
+
+ /**
+ * findIdMatch
+ *
+ * Return pointers to sequences (or sequence object containers) which have
+ * same Id as a given set of different sequence objects
+ *
+ * @param seqs
+ * SequenceI[]
+ * @return SequenceI[]
+ */
+ public SequenceI[] findIdMatch(SequenceI[] seqs)
+ {
+ SequenceI[] namedseqs = null;
+ int i = 0;
+ SeqIdName nam;
+
+ if (seqs.length > 0)
+ {
+ namedseqs = new SequenceI[seqs.length];
+ do
+ {
+ nam = new SeqIdName(seqs[i].getName());
+
+ if (names.containsKey(nam))
+ {
+ namedseqs[i] = findIdMatch(nam);
+ }
+ else
+ {
+ namedseqs[i] = null;
+ }
+ } while (++i < seqs.length);
+ }
+
+ return namedseqs;
+ }
+
+ /**
+ * core findIdMatch search method
+ *
+ * @param nam
+ * SeqIdName
+ * @return SequenceI
+ */
+ private SequenceI findIdMatch(SeqIdName nam)
+ {
+ List matches = new ArrayList<>();
+ while (names.containsKey(nam))
+ {
+ List candidates = names.remove(nam);
+ List 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);
+ }
+
+ /**
+ * core findIdMatch search method for finding all equivalent matches
+ *
+ * @param nam
+ * SeqIdName
+ * @return SequenceI[]
+ */
+ private List findAllIdMatches(
+ jalview.analysis.SequenceIdMatcher.SeqIdName nam)
+ {
+ List matches = new ArrayList<>();
+ while (names.containsKey(nam))
+ {
+ matches.addAll(names.remove(nam));
+ }
+ List r = pickbestMatches(nam, matches);
+ return r;
+ }
+
+ class SeqIdName
+ {
+ String id;
+
+ SeqIdName(String s)
+ {
+ if (s != null)
+ {
+ id = s.toLowerCase();
+ }
+ else
+ {
+ id = "";
+ }
+ }
+
+ @Override
+ public int 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.stringequals(((SeqIdName) s).id);
+ }
+ else
+ {
+ if (s instanceof String)
+ {
+ return this.stringequals(((String) s).toLowerCase());
+ }
+ }
+
+ return false;
+ }
+
+ /**
+ * Characters that define the end of a unique sequence ID at the beginning
+ * of an arbitrary ID string JBPNote: This is a heuristic that will fail for
+ * arbritrarily extended sequence id's (like portions of an aligned set of
+ * repeats from one sequence)
+ */
+ private String WORD_SEP = "~. |#\\/<>!\"" + ((char) 0x00A4)
+ + "$%^*)}[@',?_";
+
+ /**
+ * matches if one ID properly contains another at a whitespace boundary.
+ * TODO: (JBPNote) These are not efficient. should use char[] for speed
+ * todo: (JBPNote) Set separator characters appropriately
+ *
+ * @param s
+ * @return boolean
+ */
+ private boolean stringequals(String s)
+ {
+ 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;
+ }
+ }
+
+ /**
+ * toString method returns the wrapped sequence id. For debugging purposes
+ * only, behaviour not guaranteed not to change.
+ */
+ @Override
+ public String toString()
+ {
+ return id;
+ }
+ }
+}