/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8.1)
- * Copyright (C) 2014 The Jalview Authors
+ * 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.HashSet;
+import java.util.List;
+import java.util.Set;
/**
* Routines for approximate Sequence Id resolution by name using string
*/
public class SequenceIdMatcher
{
- private Hashtable names;
+ /**
+ * weak hash for each sequence
+ */
+ private HashMap<SeqIdName, Set<SequenceI>> names;
+
+ // /**
+ // * cache of values removed for each query string.
+ // */
+ // private HashMap<String, List<SequenceI>> resolved;
+
+ /**
+ * do we index sequences on all 'words' in ID string ?
+ */
+ private boolean wordBased = 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)
+ */
+ static String WORD_SEP = "~. |#\\/<>!\"" + ((char) 0x00A4)
+ + "$%^*)}[@',?_";
+
+ /**
+ * @return true if matcher is word-based (ie string key matches one of the
+ * words within the body of one or more sequence IDs)
+ */
+ public boolean isWordBased()
+ {
+ return wordBased;
+ }
+
+ /**
+ * Construct a standard (non-word based) matcher. To configure word based
+ * matching, use the fully qualified constructor
+ *
+ * @param seqs
+ */
+ public SequenceIdMatcher(List<SequenceI> seqs)
+ {
+ this(false, seqs);
+ }
+
+ /**
+ * construct a new matcher for a set of sequences, configured as required.
+ * Note: enabling word based matching
+ *
+ * @param wordBasedMatch
+ * - when true, "myseq" matches "X|myseq" and "myseq"
+ * @param seqs
+ */
+ public SequenceIdMatcher(boolean wordBasedMatch, List<SequenceI> seqs)
+ {
+ wordBased = wordBasedMatch;
+ names = new HashMap<SeqIdName, Set<SequenceI>>();
+ addAll(seqs);
+ }
- public SequenceIdMatcher(SequenceI[] seqs)
+ /**
+ * add more sequences to this matcher - also used by the constructor
+ *
+ * @param seqs
+ */
+ public void addAll(List<SequenceI> seqs)
{
- names = new Hashtable();
- for (int i = 0; i < seqs.length; i++)
+ for (SequenceI seq : seqs)
{
- // 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)
+ addSeq(seq);
+ }
+ }
+
+ private void addSeqIdName(SeqIdName idname, SequenceI seq)
+ {
+ Set<SequenceI> seqset = names.get(idname);
+ if (seqset == null)
+ {
+ seqset = new HashSet<SequenceI>();
+ names.put(idname, seqset);
+ }
+ seqset.add(seq);
+ }
+
+ public void addSeq(SequenceI seq)
+ {
+ // TODO: deal with ID collisions - SequenceI should be appended to list
+ // associated with this key.
+ addSeqIdName(new SeqIdName(seq.getDisplayId(true)), seq);
+ if (wordBased)
+ {
+ for (SeqIdName key : getWordsFor(seq))
+ {
+ addSeqIdName(key, seq);
+ }
+ }
+ SequenceI dbseq = seq;
+ // TODO add test for database xref resolution
+ while (dbseq.getDatasetSequence() != null)
+ {
+ dbseq = dbseq.getDatasetSequence();
+ }
+ // add in any interesting identifiers
+ if (dbseq.getDBRefs() != null)
+ {
+ DBRefEntry dbr[] = dbseq.getDBRefs();
+ SeqIdName sid = null;
+ for (int r = 0; r < dbr.length; r++)
{
- DBRefEntry dbr[] = seqs[i].getDBRef();
- SeqIdName sid = null;
- for (int r = 0; r < dbr.length; r++)
+ sid = new SeqIdName(dbr[r].getAccessionId());
+ if (!names.containsKey(sid))
{
- sid = new SeqIdName(dbr[r].getAccessionId());
- if (!names.contains(sid))
+ addSeqIdName(sid, seq);
+ }
+ }
+ }
+ }
+
+ /**
+ * generate word based keys for the given sequence
+ *
+ * @param seq
+ * @return list of split keys
+ */
+ public static List<SeqIdName> getWordsFor(SequenceI seq)
+ {
+ ArrayList<SeqIdName> keys = new ArrayList<SeqIdName>();
+ String name = seq.getName(), limits = "/" + seq.getStart() + "-"
+ + seq.getEnd();
+ int namel = name.length();
+ char[] sep = new char[WORD_SEP.length()];
+ // find only the separators present in the ID.
+ for (int i = 0; i < sep.length; i++)
+ {
+ sep[i] = WORD_SEP.charAt(i);
+ if (seq.getName().indexOf("" + sep[i]) == -1)
+ {
+ sep[i] = 0;
+ }
+ }
+ ;
+ // make words
+ for (int i = 0; i < sep.length; i++)
+ {
+ if (sep[i] > 0)
+ {
+ int p = 0, m = -1;
+ while ((m = name.indexOf(sep[i], p)) > p)
+ {
+
+ if (m > 0 && m - p > 5)
{
- names.put(sid, seqs[i]);
+ // split to end of word m with this delimiter
+ keys.add(new SeqIdName(name.substring(p, m - 1) + limits));
}
+ if (namel - m > 5)
+ {
+ // index word after this delimiter m
+ keys.add(new SeqIdName(name.substring(m + 1) + limits));
+ }
+ p = m + 1;
+ }
+ if (namel - p > 4)
+ {
+ // index word after this delimiter m
+ keys.add(new SeqIdName(name.substring(p) + limits));
}
}
}
+ return keys;
+ }
+
+ /**
+ * convenience method to make a matcher from concrete array Note: in order to
+ * support word based matching, use the fully qualified constructor
+ *
+ * @param sequences
+ */
+ public SequenceIdMatcher(SequenceI[] sequences)
+ {
+ this(Arrays.asList(sequences));
}
/**
* @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;
+ ArrayList<SequenceI> best = new ArrayList<SequenceI>();
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);
+ addSeq(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);
+ addSeq(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)
+ boolean is_closer = ((q = Math.abs(matchlen - namlen)) > (w = Math
+ .abs(candlen - namlen)) && candlen > matchlen);
+ // if not closer, then check if current best is actually identical in case
+ // as
+ // well
+ if (is_closer
+ || (candName.equalsCase(cand.getName()) && !candName
+ .equalsCase(best.get(0).getName())))
{
best.clear();
match = cand;
matchlen = candlen;
best.add(match);
}
- if (q == w && candlen == matchlen)
+ else
{
- // record any ties
- best.add(cand);
+ if (q == w && candlen == matchlen)
+ {
+ // equivalently good, and matches with case as well. so
+ // record any ties
+ best.add(cand);
+ }
}
}
if (best.size() == 0)
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;
}
/**
* @return SequenceI
*/
private SequenceI findIdMatch(
- jalview.analysis.SequenceIdMatcher.SeqIdName nam)
+ jalview.analysis.SeqIdName nam)
{
- Vector matches = new Vector();
+ ArrayList<SequenceI> matches = new ArrayList<SequenceI>();
while (names.containsKey(nam))
{
- matches.addElement(names.remove(nam));
+ matches.addAll(names.remove(nam));
}
return pickbestMatch(nam, matches);
}
* SeqIdName
* @return SequenceI[]
*/
- private SequenceI[] findAllIdMatches(
- jalview.analysis.SequenceIdMatcher.SeqIdName nam)
+ private List<SequenceI> findAllIdMatches(
+ jalview.analysis.SeqIdName nam)
{
- Vector matches = new Vector();
+ ArrayList<SequenceI> matches = new ArrayList<SequenceI>();
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
- {
- 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 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
- * SeqIdName
- * @return boolean
- */
- public boolean equals(SeqIdName 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())
- {
- 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;
- }
- }
- }
}
git://source.jalview.org / jalview.git / blobdiff