2 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3 * Copyright (C) $$Year-Rel$$ The Jalview Authors
5 * This file is part of Jalview.
7 * Jalview is free software: you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation, either version 3
10 * of the License, or (at your option) any later version.
12 * Jalview is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15 * PURPOSE. See the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19 * The Jalview Authors are detailed in the 'AUTHORS' file.
23 // http://evolution.genetics.washington.edu/phylip/newick_doc.html
24 // TODO: Implement Basic NHX tag parsing and preservation
25 // TODO: http://evolution.genetics.wustl.edu/eddy/forester/NHX.html
26 // TODO: Extended SequenceNodeI to hold parsed NHX strings
29 import jalview.datamodel.SequenceNode;
30 import jalview.util.MessageManager;
32 import java.io.BufferedReader;
34 import java.io.FileReader;
35 import java.io.IOException;
36 import java.util.StringTokenizer;
38 import com.stevesoft.pat.Regex;
41 * Parse a new hanpshire style tree Caveats: NHX files are NOT supported and the
42 * tree distances and topology are unreliable when they are parsed. TODO: on
43 * this: NHX codes are appended in comments beginning with &&NHX. The codes are
44 * given below (from http://www.phylosoft.org/forester/NHX.html): Element Type
45 * Description Corresponding phyloXML element (parent element in parentheses) no
46 * tag string name of this node/clade (MUST BE FIRST, IF ASSIGNED)
47 * <name>(<clade>) : decimal branch length to parent node (MUST BE SECOND, IF
48 * ASSIGNED) <branch_length>(<clade>) :GN= string gene name <name>(<sequence>)
49 * :AC= string sequence accession <accession>(<sequence>) :ND= string node
50 * identifier - if this is being used, it has to be unique within each phylogeny
51 * <node_id>(<clade>) :B= decimal confidence value for parent branch
52 * <confidence>(<clade>) :D= 'T', 'F', or '?' 'T' if this node represents a
53 * duplication event - 'F' if this node represents a speciation event, '?' if
54 * this node represents an unknown event (D= tag should be replaced by Ev= tag)
55 * n/a :Ev=duplications>speciations>gene losses>event type>duplication type int
56 * int int string string event (replaces the =D tag), number of duplication,
57 * speciation, and gene loss events, type of event (transfer, fusion, root,
58 * unknown, other, speciation_duplication_loss, unassigned) <events>(<clade>)
59 * :E= string EC number at this node <annotation>(<sequence>) :Fu= string
60 * function at this node <annotation>(<sequence>)
61 * :DS=protein-length>from>to>support>name>from>... int int int double string
62 * int ... domain structure at this node <domain_architecture>(<sequence>) :S=
63 * string species name of the species/phylum at this node <taxonomy>(<clade>)
64 * :T= integer taxonomy ID of the species/phylum at this node <id>(<taxonomy>)
65 * :W= integer width of parent branch <width>(<clade>) :C=rrr.ggg.bbb
66 * integer.integer.integer color of parent branch <color>(<clade>) :Co= 'Y' or
67 * 'N' collapse this node when drawing the tree (default is not to collapse) n/a
68 * :XB= string custom data associated with a branch <property>(<clade>) :XN=
69 * string custom data associated with a node <property>(<clade>) :O= integer
70 * orthologous to this external node n/a :SN= integer subtree neighbors n/a :SO=
71 * integer super orthologous (no duplications on paths) to this external node
77 public class NewickFile extends FileParse
81 private boolean HasBootstrap = false;
83 private boolean HasDistances = false;
85 private boolean RootHasDistance = false;
88 boolean ReplaceUnderscores = false;
90 boolean printRootInfo = true;
92 private Regex[] NodeSafeName = new Regex[] {
93 new Regex().perlCode("m/[\\[,:'()]/"), // test for
96 new Regex().perlCode("s/'/''/"), // escaping quote
98 new Regex().perlCode("s/\\/w/_/") // unqoted whitespace
102 char QuoteChar = '\'';
105 * Creates a new NewickFile object.
110 * @throws IOException
113 public NewickFile(String inStr) throws IOException
115 super(inStr, DataSourceType.PASTE);
119 * Creates a new NewickFile object.
126 * @throws IOException
129 public NewickFile(String inFile, DataSourceType protocol)
132 super(inFile, protocol);
135 public NewickFile(FileParse source) throws IOException
141 * Creates a new NewickFile object.
146 public NewickFile(SequenceNode newtree)
152 * Creates a new NewickFile object.
159 public NewickFile(SequenceNode newtree, boolean bootstrap)
161 HasBootstrap = bootstrap;
166 * Creates a new NewickFile object.
175 public NewickFile(SequenceNode newtree, boolean bootstrap,
179 HasBootstrap = bootstrap;
180 HasDistances = distances;
184 * Creates a new NewickFile object.
192 * @param rootdistance
195 public NewickFile(SequenceNode newtree, boolean bootstrap,
196 boolean distances, boolean rootdistance)
199 HasBootstrap = bootstrap;
200 HasDistances = distances;
201 RootHasDistance = rootdistance;
218 * @return DOCUMENT ME!
220 private String ErrorStringrange(String Error, String Er, int r, int p,
223 return ((Error == null) ? "" : Error) + Er + " at position " + p + " ( "
224 + s.substring(((p - r) < 0) ? 0 : (p - r),
225 ((p + r) > s.length()) ? s.length() : (p + r))
230 // These are set automatically by the reader
231 public boolean HasBootstrap()
239 * @return DOCUMENT ME!
241 public boolean HasDistances()
246 public boolean HasRootDistance()
248 return RootHasDistance;
252 * parse the filesource as a newick file (new hampshire and/or extended)
254 * @throws IOException
255 * with a line number and character position for badly formatted NH
258 public void parse() throws IOException
262 { // fill nf with complete tree file
264 StringBuffer file = new StringBuffer();
266 while ((nf = nextLine()) != null)
271 nf = file.toString();
274 root = new SequenceNode();
276 SequenceNode realroot = null;
277 SequenceNode c = root;
281 // int flen = nf.length();
284 String nodename = null;
285 String commentString2 = null; // comments after simple node props
287 float DefDistance = (float) 0.001; // @param Default distance for a node -
289 int DefBootstrap = -1; // @param Default bootstrap for a node
291 float distance = DefDistance;
292 int bootstrap = DefBootstrap;
294 boolean ascending = false; // flag indicating that we are leaving the
297 Regex majorsyms = new Regex(
302 boolean parsednodename = false;
303 while (majorsyms.searchFrom(nf, cp) && (Error == null))
305 int fcp = majorsyms.matchedFrom();
307 switch (schar = nf.charAt(fcp))
311 // ascending should not be set
315 Error = ErrorStringrange(Error, "Unexpected '('", 7, fcp, nf);
321 if (c.right() == null)
323 c.setRight(new SequenceNode(null, c, null, DefDistance,
324 DefBootstrap, false));
325 c = (SequenceNode) c.right();
329 if (c.left() != null)
331 // Dummy node for polytomy - keeps c.left free for new node
332 SequenceNode tmpn = new SequenceNode(null, c, null, 0, 0, true);
333 tmpn.SetChildren(c.left(), c.right());
337 c.setLeft(new SequenceNode(null, c, null, DefDistance,
338 DefBootstrap, false));
339 c = (SequenceNode) c.left();
342 if (realroot == null)
348 distance = DefDistance;
349 bootstrap = DefBootstrap;
354 // Deal with quoted fields
357 Regex qnodename = new Regex(
360 if (qnodename.searchFrom(nf, fcp))
362 int nl = qnodename.stringMatched().length();
363 nodename = new String(
364 qnodename.stringMatched().substring(1, nl - 1));
365 // unpack any escaped colons
366 Regex xpandquotes = Regex
367 .perlCode("s/''/'/");
368 String widernodename = xpandquotes.replaceAll(nodename);
369 nodename = widernodename;
370 // jump to after end of quoted nodename
371 nextcp = fcp + nl + 1;
372 parsednodename = true;
376 Error = ErrorStringrange(Error,
377 "Unterminated quotes for nodename", 7, fcp, nf);
387 Error = ErrorStringrange(Error,
388 "Wayward semicolon (depth=" + d + ")", 7, fcp, nf);
390 // cp advanced at the end of default
394 // node string contains Comment or structured/extended NH format info
396 * if ((fcp-cp>1 && nf.substring(cp,fcp).trim().length()>1)) { // will
397 * process in remains System.err.println("skipped text:
398 * '"+nf.substring(cp,fcp)+"'"); }
400 // verify termination.
401 Regex comment = new Regex(
403 if (comment.searchFrom(nf, fcp))
405 // Skip the comment field
406 nextcp = comment.matchedFrom() + 1;
407 warningMessage = "Tree file contained comments which may confuse input algorithm.";
410 // cp advanced at the end of default to nextcp, ncp is unchanged so
411 // any node info can be read.
415 Error = ErrorStringrange(Error, "Unterminated comment", 3, fcp,
419 // Parse simpler field strings
420 String fstring = nf.substring(ncp, fcp);
421 // remove any comments before we parse the node info
422 // TODO: test newick file with quoted square brackets in node name (is
424 while (fstring.indexOf(']') > -1)
426 int cstart = fstring.indexOf('[');
427 int cend = fstring.indexOf(']');
428 commentString2 = fstring.substring(cstart + 1, cend);
429 fstring = fstring.substring(0, cstart)
430 + fstring.substring(cend + 1);
433 Regex uqnodename = new Regex(
434 "\\b([^' :;\\](),]+)");
435 Regex nbootstrap = new Regex(
436 "\\s*([0-9+]+)\\s*:");
437 Regex ndist = new Regex(
440 if (!parsednodename && uqnodename.search(fstring)
441 && ((uqnodename.matchedFrom(1) == 0) || (fstring
442 .charAt(uqnodename.matchedFrom(1) - 1) != ':'))) // JBPNote
445 if (nodename == null)
447 if (ReplaceUnderscores)
449 nodename = uqnodename.stringMatched(1).replace('_', ' ');
453 nodename = uqnodename.stringMatched(1);
458 Error = ErrorStringrange(Error,
459 "File has broken algorithm - overwritten nodename", 10,
463 // get comment bootstraps
465 if (nbootstrap.search(fstring))
467 if (nbootstrap.stringMatched(1)
468 .equals(uqnodename.stringMatched(1)))
470 nodename = null; // no nodename here.
472 if (nodename == null || nodename.length() == 0
473 || nbootstrap.matchedFrom(1) > (uqnodename.matchedFrom(1)
474 + uqnodename.stringMatched().length()))
478 bootstrap = (Integer.valueOf(nbootstrap.stringMatched(1)))
481 } catch (Exception e)
483 Error = ErrorStringrange(Error, "Can't parse bootstrap value",
484 4, ncp + nbootstrap.matchedFrom(), nf);
489 boolean nodehasdistance = false;
491 if (ndist.search(fstring))
495 distance = (Float.valueOf(ndist.stringMatched(1))).floatValue();
497 nodehasdistance = true;
498 } catch (Exception e)
500 Error = ErrorStringrange(Error,
501 "Can't parse node distance value", 7,
502 ncp + ndist.matchedFrom(), nf);
508 // Write node info here
510 // Trees without distances still need a render distance
511 c.dist = (HasDistances) ? distance : DefDistance;
512 // be consistent for internal bootstrap defaults too
513 c.setBootstrap((HasBootstrap) ? bootstrap : DefBootstrap);
516 RootHasDistance = nodehasdistance; // JBPNote This is really
517 // UGLY!!! Ensure root node gets
518 // its given distance
520 parseNHXNodeProps(c, commentString2);
521 commentString2 = null;
525 // Find a place to put the leaf
526 SequenceNode newnode = new SequenceNode(null, c, nodename,
527 (HasDistances) ? distance : DefDistance,
528 (HasBootstrap) ? bootstrap : DefBootstrap, false);
529 parseNHXNodeProps(c, commentString2);
530 commentString2 = null;
532 if (c.right() == null)
538 if (c.left() == null)
544 // Insert a dummy node for polytomy
545 // dummy nodes have distances
546 SequenceNode newdummy = new SequenceNode(null, c, null,
547 (HasDistances ? 0 : DefDistance), 0, true);
548 newdummy.SetChildren(c.left(), newnode);
556 // move back up the tree from preceding closure
559 if ((d > -1) && (c == null))
561 Error = ErrorStringrange(Error,
562 "File broke algorithm: Lost place in tree (is there an extra ')' ?)",
567 if (nf.charAt(fcp) == ')')
574 if (nf.charAt(fcp) == ',')
582 // Just advance focus, if we need to
583 if ((c.left() != null) && (!c.left().isLeaf()))
585 c = (SequenceNode) c.left();
591 // Reset new node properties to obvious fakes
593 distance = DefDistance;
594 bootstrap = DefBootstrap;
595 commentString2 = null;
596 parsednodename = false;
611 throw (new IOException(
612 MessageManager.formatMessage("exception.newfile", new String[]
613 { Error.toString() })));
617 throw (new IOException(
618 MessageManager.formatMessage("exception.newfile", new String[]
619 { MessageManager.getString("label.no_tree_read_in") })));
621 // THe next line is failing for topali trees - not sure why yet. if
622 // (root.right()!=null && root.isDummy())
623 root = (SequenceNode) root.right().detach(); // remove the imaginary root.
625 if (!RootHasDistance)
627 root.dist = (HasDistances) ? 0 : DefDistance;
632 * parse NHX codes in comment strings and update NewickFile state flags for
633 * distances and bootstraps, and add any additional properties onto the node.
636 * @param commentString
637 * @param commentString2
639 private void parseNHXNodeProps(SequenceNode c, String commentString)
641 // TODO: store raw comment on the sequenceNode so it can be recovered when
643 if (commentString != null && commentString.startsWith("&&NHX"))
645 StringTokenizer st = new StringTokenizer(commentString.substring(5),
647 while (st.hasMoreTokens())
649 String tok = st.nextToken();
650 int colpos = tok.indexOf("=");
654 String code = tok.substring(0, colpos);
655 String value = tok.substring(colpos + 1);
658 // parse out code/value pairs
659 if (code.toLowerCase().equals("b"))
662 Float iv = Float.valueOf(value);
663 v = iv.intValue(); // jalview only does integer bootstraps
669 } catch (Exception e)
672 "Couldn't parse code '" + code + "' = '" + value + "'");
673 e.printStackTrace(System.err);
684 * @return DOCUMENT ME!
686 public SequenceNode getTree()
692 * Generate a newick format tree according to internal flags for bootstraps,
693 * distances and root distances.
695 * @return new hampshire tree in a single line
697 public String print()
701 StringBuffer tf = new StringBuffer();
704 return (tf.append(";").toString());
711 * Generate a newick format tree according to internal flags for distances and
712 * root distances and user specificied writing of bootstraps.
714 * @param withbootstraps
715 * controls if bootstrap values are explicitly written.
717 * @return new hampshire tree in a single line
719 public String print(boolean withbootstraps)
723 boolean boots = this.HasBootstrap;
724 this.HasBootstrap = withbootstraps;
727 this.HasBootstrap = boots;
735 * Generate newick format tree according to internal flags for writing root
738 * @param withbootstraps
739 * explicitly write bootstrap values
741 * explicitly write distances
743 * @return new hampshire tree in a single line
745 public String print(boolean withbootstraps, boolean withdists)
749 boolean dists = this.HasDistances;
750 this.HasDistances = withdists;
752 String rv = print(withbootstraps);
753 this.HasDistances = dists;
760 * Generate newick format tree according to user specified flags
762 * @param withbootstraps
763 * explicitly write bootstrap values
765 * explicitly write distances
766 * @param printRootInfo
767 * explicitly write root distance
769 * @return new hampshire tree in a single line
771 public String print(boolean withbootstraps, boolean withdists,
772 boolean printRootInfo)
776 boolean rootinfo = printRootInfo;
777 this.printRootInfo = printRootInfo;
779 String rv = print(withbootstraps, withdists);
780 this.printRootInfo = rootinfo;
789 * @return DOCUMENT ME!
802 * @return DOCUMENT ME!
804 char setQuoteChar(char c)
806 char old = QuoteChar;
818 * @return DOCUMENT ME!
820 private String nodeName(String name)
822 if (NodeSafeName[0].search(name))
824 return QuoteChar + NodeSafeName[1].replaceAll(name) + QuoteChar;
828 return NodeSafeName[2].replaceAll(name);
838 * @return DOCUMENT ME!
840 private String printNodeField(SequenceNode c)
842 return ((c.getName() == null) ? "" : nodeName(c.getName()))
843 + ((HasBootstrap) ? ((c.getBootstrap() > -1)
844 ? ((c.getName() != null ? " " : "") + c.getBootstrap())
846 + ((HasDistances) ? (":" + c.dist) : "");
855 * @return DOCUMENT ME!
857 private String printRootField(SequenceNode root)
859 return (printRootInfo)
860 ? (((root.getName() == null) ? "" : nodeName(root.getName()))
862 ? ((root.getBootstrap() > -1)
863 ? ((root.getName() != null ? " " : "")
864 + +root.getBootstrap())
867 + ((RootHasDistance) ? (":" + root.dist) : ""))
871 // Non recursive call deals with root node properties
872 public void print(StringBuffer tf, SequenceNode root)
876 if (root.isLeaf() && printRootInfo)
878 tf.append(printRootField(root));
884 _print(tf, (SequenceNode) root.right());
885 _print(tf, (SequenceNode) root.left());
890 _print(tf, (SequenceNode) root.right());
892 if (root.left() != null)
897 _print(tf, (SequenceNode) root.left());
898 tf.append(")" + printRootField(root));
904 // Recursive call for non-root nodes
905 public void _print(StringBuffer tf, SequenceNode c)
911 tf.append(printNodeField(c));
917 _print(tf, (SequenceNode) c.left());
918 if (c.left() != null)
922 _print(tf, (SequenceNode) c.right());
927 _print(tf, (SequenceNode) c.right());
929 if (c.left() != null)
934 _print(tf, (SequenceNode) c.left());
935 tf.append(")" + printNodeField(c));
946 public static void main(String[] args)
950 if (args == null || args.length != 1)
953 "Takes one argument - file name of a newick tree file.");
957 File fn = new File(args[0]);
959 StringBuffer newickfile = new StringBuffer();
960 BufferedReader treefile = new BufferedReader(new FileReader(fn));
963 while ((l = treefile.readLine()) != null)
965 newickfile.append(l);
969 System.out.println("Read file :\n");
971 NewickFile trf = new NewickFile(args[0], DataSourceType.FILE);
973 System.out.println("Original file :\n");
975 Regex nonl = new Regex("\n+", "");
976 System.out.println(nonl.replaceAll(newickfile.toString()) + "\n");
978 System.out.println("Parsed file.\n");
979 System.out.println("Default output type for original input.\n");
980 System.out.println(trf.print());
981 System.out.println("Without bootstraps.\n");
982 System.out.println(trf.print(false));
983 System.out.println("Without distances.\n");
984 System.out.println(trf.print(true, false));
985 System.out.println("Without bootstraps but with distanecs.\n");
986 System.out.println(trf.print(false, true));
987 System.out.println("Without bootstraps or distanecs.\n");
988 System.out.println(trf.print(false, false));
989 System.out.println("With bootstraps and with distances.\n");
990 System.out.println(trf.print(true, true));
991 } catch (java.io.IOException e)
993 System.err.println("Exception\n" + e);