-/*\r
- * Jalview - A Sequence Alignment Editor and Viewer\r
- * Copyright (C) 2006 AM Waterhouse, J Procter, G Barton, M Clamp, S Searle\r
- *\r
- * This program is free software; you can redistribute it and/or\r
- * modify it under the terms of the GNU General Public License\r
- * as published by the Free Software Foundation; either version 2\r
- * of the License, or (at your option) any later version.\r
- *\r
- * This program is distributed in the hope that it will be useful,\r
- * but WITHOUT ANY WARRANTY; without even the implied warranty of\r
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\r
- * GNU General Public License for more details.\r
- *\r
- * You should have received a copy of the GNU General Public License\r
- * along with this program; if not, write to the Free Software\r
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA\r
- */\r
-package jalview.analysis;\r
-\r
-import java.util.*;\r
-\r
-import jalview.datamodel.*;\r
-\r
-/**\r
- * <p>Title: </p>\r
- * SequenceIdMatcher\r
- * <p>Description: </p>\r
- * Routine which does approximate Sequence Id resolution by name using\r
- * string containment (on word boundaries) rather than equivalence. It also\r
- * attempts to resolve ties where no exact match is available by picking the\r
- * the id closest to the query.\r
- * <p>Copyright: Copyright (c) 2004</p>\r
- *\r
- * <p>Company: Dundee University</p>\r
- *\r
- * @author not attributable\r
- * @version 1.0\r
- */\r
-public class SequenceIdMatcher\r
-{\r
- private Hashtable names;\r
-\r
- public SequenceIdMatcher(SequenceI[] seqs)\r
- {\r
- names = new Hashtable();\r
- for (int i = 0; i < seqs.length; i++)\r
- {\r
- names.put(new SeqIdName(seqs[i].getName()), seqs[i]);\r
- }\r
- }\r
- /**\r
- * returns the closest SequenceI in matches to SeqIdName and returns all the matches\r
- * to the names hash.\r
- * @param candName SeqIdName\r
- * @param matches Vector of SequenceI objects\r
- * @return SequenceI closest SequenceI to SeqIdName\r
- */\r
- private SequenceI pickbestMatch(SeqIdName candName, Vector matches) {\r
- SequenceI match=null;\r
- if (candName==null || matches==null || matches.size()==0)\r
- return null;\r
- match=(SequenceI) matches.elementAt(0);\r
- matches.removeElementAt(0);\r
- names.put(new SeqIdName(match.getName()), match);\r
- int matchlen=match.getName().length();\r
- int namlen=candName.id.length();\r
- while (matches.size()>0) {\r
- // look through for a better one.\r
- SequenceI cand=(SequenceI) matches.elementAt(0);\r
- names.put(new SeqIdName(cand.getName()), cand);\r
- int candlen = cand.getName().length();\r
- // keep the one with an id 'closer' to the given seqnam string\r
- if (Math.abs(matchlen-namlen)>Math.abs(candlen-namlen) && candlen>matchlen) {\r
- match = cand;\r
- matchlen = candlen;\r
- }\r
- }\r
- return match;\r
- }\r
-\r
- /**\r
- * get SequenceI with closest SequenceI.getName() to seq.getName()\r
- * @param seq SequenceI\r
- * @return SequenceI\r
- */\r
- SequenceI findIdMatch(SequenceI seq)\r
- {\r
- SeqIdName nam = new SeqIdName(seq.getName());\r
- return findIdMatch(nam);\r
- }\r
-\r
- SequenceI findIdMatch(String seqnam)\r
- {\r
- SeqIdName nam = new SeqIdName(seqnam);\r
- return findIdMatch(nam);\r
- }\r
- /**\r
- * findIdMatch\r
- *\r
- * Return pointers to sequences (or sequence object containers)\r
- * which have same Id as a given set of different sequence objects\r
- *\r
- * @param seqs SequenceI[]\r
- * @return SequenceI[]\r
- */\r
- SequenceI[] findIdMatch(SequenceI[] seqs)\r
- {\r
- SequenceI[] namedseqs = null;\r
- int i = 0;\r
- SeqIdName nam;\r
-\r
- if (seqs.length > 0)\r
- {\r
- namedseqs = new SequenceI[seqs.length];\r
- do\r
- {\r
- nam = new SeqIdName(seqs[i].getName());\r
-\r
- if (names.containsKey(nam))\r
- {\r
- namedseqs[i] = findIdMatch(nam);\r
- }\r
- else\r
- {\r
- namedseqs[i] = null;\r
- }\r
- }\r
- while (++i < seqs.length);\r
- }\r
-\r
- return namedseqs;\r
- }\r
-\r
- /**\r
- * core findIdMatch search method\r
- * @param nam SeqIdName\r
- * @return SequenceI\r
- */\r
- private SequenceI findIdMatch(jalview.analysis.SequenceIdMatcher.SeqIdName\r
- nam)\r
- {\r
- Vector matches=new Vector();\r
- while (names.containsKey(nam))\r
- {\r
- matches.addElement(names.remove(nam));\r
- }\r
- return pickbestMatch(nam, matches);\r
- }\r
-\r
- private class SeqIdName\r
- {\r
- String id;\r
-\r
- SeqIdName(String s)\r
- {\r
- if (s!=null)\r
- id = new String(s);\r
- else\r
- id = "";\r
- }\r
-\r
- public int hashCode()\r
- {\r
- return ((id.length()>=4) ? id.substring(0, 4).hashCode() : id.hashCode());\r
- }\r
-\r
- public boolean equals(Object s)\r
- {\r
- if (s instanceof SeqIdName)\r
- {\r
- return this.equals( (SeqIdName) s);\r
- }\r
- else\r
- {\r
- if (s instanceof String)\r
- {\r
- return this.equals( (String) s);\r
- }\r
- }\r
-\r
- return false;\r
- }\r
-\r
- /**\r
- * Characters that define the end of a unique sequence ID at\r
- * the beginning of an arbitrary ID string\r
- * JBPNote: This is a heuristic that will fail for arbritrarily extended sequence id's\r
- * (like portions of an aligned set of repeats from one sequence)\r
- */\r
- private String WORD_SEP="~. |#\\/<>!\"£$%^*)}[@',?_";\r
-\r
- /**\r
- * matches if one ID properly contains another at a whitespace boundary.\r
- * TODO: (JBPNote) These are not efficient. should use char[] for speed\r
- * todo: (JBPNote) Set separator characters appropriately\r
- * @param s SeqIdName\r
- * @return boolean\r
- */\r
- public boolean equals(SeqIdName s)\r
- {\r
- if (id.length()>s.id.length()) {\r
- return id.startsWith(s.id) ?\r
- (WORD_SEP.indexOf(id.charAt(s.id.length()))>-1)\r
- : false;\r
- } else\r
- return s.id.startsWith(id) ?\r
- (s.id.equals(id) ? true :\r
- (WORD_SEP.indexOf(s.id.charAt(id.length()))>-1))\r
- : false;\r
- }\r
-\r
- public boolean equals(String s)\r
- {\r
- if (id.length()>s.length()) {\r
- return id.startsWith(s) ?\r
- (WORD_SEP.indexOf(id.charAt(s.length()))>-1)\r
- : false;\r
- } else\r
- return s.startsWith(id) ?\r
- (s.equals(id) ? true :\r
- (WORD_SEP.indexOf(s.charAt(id.length()))>-1))\r
- : false;\r
- }\r
- }\r
-}\r
+/*
+ * 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
+ *
+ * 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/>.
+ */
+package jalview.analysis;
+
+import java.util.*;
+
+import jalview.datamodel.*;
+
+/**
+ * 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 Hashtable names;
+
+ public SequenceIdMatcher(SequenceI[] seqs)
+ {
+ names = new Hashtable();
+ for (int i = 0; i < seqs.length; i++)
+ {
+ // 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)
+ {
+ DBRefEntry dbr[] = seqs[i].getDBRef();
+ SeqIdName sid = null;
+ for (int r = 0; r < dbr.length; r++)
+ {
+ sid = new SeqIdName(dbr[r].getAccessionId());
+ if (!names.contains(sid))
+ {
+ names.put(sid, seqs[i]);
+ }
+ }
+ }
+ }
+ }
+
+ /**
+ * 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 SequenceI closest SequenceI to SeqIdName
+ */
+ private SequenceI pickbestMatch(SeqIdName candName, Vector matches)
+ {
+ SequenceI[] st= pickbestMatches(candName, matches);
+ return st==null || st.length==0 ? null : st[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 SequenceI[] pickbestMatches(SeqIdName candName, Vector matches)
+ {
+ ArrayList best=new ArrayList();
+ SequenceI match = null;
+ if (candName == null || matches == null || matches.size() == 0)
+ {
+ return null;
+ }
+ match = (SequenceI) matches.elementAt(0);
+ matches.removeElementAt(0);
+ best.add(match);
+ names.put(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);
+ 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 (SequenceI[]) best.toArray(new SequenceI[0]);
+ }
+
+ /**
+ * 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.
+ */
+ public SequenceI[] findAllIdMatches(String seqnam)
+ {
+
+ SeqIdName nam = new SeqIdName(seqnam);
+ return findAllIdMatches(nam);
+ }
+
+ /**
+ * 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(
+ jalview.analysis.SequenceIdMatcher.SeqIdName nam)
+ {
+ Vector matches = new Vector();
+ while (names.containsKey(nam))
+ {
+ matches.addElement(names.remove(nam));
+ }
+ return pickbestMatch(nam, matches);
+ }
+ /**
+ * core findIdMatch search method for finding all equivalent matches
+ *
+ * @param nam
+ * SeqIdName
+ * @return SequenceI[]
+ */
+ private SequenceI[] findAllIdMatches(
+ jalview.analysis.SequenceIdMatcher.SeqIdName nam)
+ {
+ Vector matches = new Vector();
+ while (names.containsKey(nam))
+ {
+ matches.addElement(names.remove(nam));
+ }
+ 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