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 import jalview.analysis.SequenceIdMatcher;
24 import jalview.datamodel.AlignedCodonFrame;
25 import jalview.datamodel.AlignmentI;
26 import jalview.datamodel.SequenceDummy;
27 import jalview.datamodel.SequenceFeature;
28 import jalview.datamodel.SequenceI;
29 import jalview.schemes.AnnotationColourGradient;
30 import jalview.schemes.GraduatedColor;
31 import jalview.schemes.UserColourScheme;
32 import jalview.util.Format;
33 import jalview.util.MapList;
35 import java.io.IOException;
36 import java.util.ArrayList;
37 import java.util.Arrays;
38 import java.util.HashMap;
39 import java.util.Hashtable;
40 import java.util.Iterator;
41 import java.util.List;
43 import java.util.StringTokenizer;
44 import java.util.Vector;
47 * Parse and create Jalview Features files Detects GFF format features files and
48 * parses. Does not implement standard print() - call specific printFeatures or
49 * printGFF. Uses AlignmentI.findSequence(String id) to find the sequence object
50 * for the features annotation - this normally works on an exact match.
55 public class FeaturesFile extends AlignFile
58 * work around for GFF interpretation bug where source string becomes
59 * description rather than a group
61 private boolean doGffSource = true;
64 * Creates a new FeaturesFile object.
71 * Creates a new FeaturesFile object.
81 public FeaturesFile(String inFile, String type) throws IOException
86 public FeaturesFile(FileParse source) throws IOException
92 * Parse GFF or sequence features file using case-independent matching,
96 * - alignment/dataset containing sequences that are to be annotated
98 * - hashtable to store feature colour definitions
100 * - process html strings into plain text
101 * @return true if features were added
103 public boolean parse(AlignmentI align, Hashtable colours,
106 return parse(align, colours, null, removeHTML, false);
110 * Parse GFF or sequence features file optionally using case-independent
111 * matching, discarding URLs
114 * - alignment/dataset containing sequences that are to be annotated
116 * - hashtable to store feature colour definitions
118 * - process html strings into plain text
119 * @param relaxedIdmatching
120 * - when true, ID matches to compound sequence IDs are allowed
121 * @return true if features were added
123 public boolean parse(AlignmentI align, Map colours, boolean removeHTML,
124 boolean relaxedIdMatching)
126 return parse(align, colours, null, removeHTML, relaxedIdMatching);
130 * Parse GFF or sequence features file optionally using case-independent
134 * - alignment/dataset containing sequences that are to be annotated
136 * - hashtable to store feature colour definitions
138 * - hashtable to store associated URLs
140 * - process html strings into plain text
141 * @return true if features were added
143 public boolean parse(AlignmentI align, Map colours, Map featureLink,
146 return parse(align, colours, featureLink, removeHTML, false);
150 * Parse GFF or sequence features file
153 * - alignment/dataset containing sequences that are to be annotated
155 * - hashtable to store feature colour definitions
157 * - hashtable to store associated URLs
159 * - process html strings into plain text
160 * @param relaxedIdmatching
161 * - when true, ID matches to compound sequence IDs are allowed
162 * @return true if features were added
164 public boolean parse(AlignmentI align, Map colours, Map featureLink,
165 boolean removeHTML, boolean relaxedIdmatching)
171 SequenceI seq = null;
173 * keep track of any sequences we try to create from the data if it is a GFF3 file
175 ArrayList<SequenceI> newseqs = new ArrayList<SequenceI>();
176 String type, desc, token = null;
178 int index, start, end;
182 String featureGroup = null, groupLink = null;
183 Map typeLink = new Hashtable();
185 * when true, assume GFF style features rather than Jalview style.
187 boolean GFFFile = true;
188 while ((line = nextLine()) != null)
190 if (line.startsWith("#"))
195 st = new StringTokenizer(line, "\t");
196 if (st.countTokens() == 1)
198 if (line.trim().equalsIgnoreCase("GFF"))
200 // Start parsing file as if it might be GFF again.
205 if (st.countTokens() > 1 && st.countTokens() < 4)
208 type = st.nextToken();
209 if (type.equalsIgnoreCase("startgroup"))
211 featureGroup = st.nextToken();
212 if (st.hasMoreElements())
214 groupLink = st.nextToken();
215 featureLink.put(featureGroup, groupLink);
218 else if (type.equalsIgnoreCase("endgroup"))
220 // We should check whether this is the current group,
221 // but at present theres no way of showing more than 1 group
228 Object colour = null;
229 String colscheme = st.nextToken();
230 if (colscheme.indexOf("|") > -1
231 || colscheme.trim().equalsIgnoreCase("label"))
233 // Parse '|' separated graduated colourscheme fields:
234 // [label|][mincolour|maxcolour|[absolute|]minvalue|maxvalue|thresholdtype|thresholdvalue]
235 // can either provide 'label' only, first is optional, next two
236 // colors are required (but may be
237 // left blank), next is optional, nxt two min/max are required.
238 // first is either 'label'
239 // first/second and third are both hexadecimal or word equivalent
241 // next two are values parsed as floats.
242 // fifth is either 'above','below', or 'none'.
243 // sixth is a float value and only required when fifth is either
244 // 'above' or 'below'.
245 StringTokenizer gcol = new StringTokenizer(colscheme, "|",
248 int threshtype = AnnotationColourGradient.NO_THRESHOLD;
249 float min = Float.MIN_VALUE, max = Float.MAX_VALUE, threshval = Float.NaN;
250 boolean labelCol = false;
252 String mincol = gcol.nextToken();
256 .println("Expected either 'label' or a colour specification in the line: "
260 String maxcol = null;
261 if (mincol.toLowerCase().indexOf("label") == 0)
264 mincol = (gcol.hasMoreTokens() ? gcol.nextToken() : null); // skip
266 mincol = (gcol.hasMoreTokens() ? gcol.nextToken() : null);
268 String abso = null, minval, maxval;
271 // at least four more tokens
272 if (mincol.equals("|"))
278 gcol.nextToken(); // skip next '|'
280 // continue parsing rest of line
281 maxcol = gcol.nextToken();
282 if (maxcol.equals("|"))
288 gcol.nextToken(); // skip next '|'
290 abso = gcol.nextToken();
291 gcol.nextToken(); // skip next '|'
292 if (abso.toLowerCase().indexOf("abso") != 0)
299 minval = gcol.nextToken();
300 gcol.nextToken(); // skip next '|'
302 maxval = gcol.nextToken();
303 if (gcol.hasMoreTokens())
305 gcol.nextToken(); // skip next '|'
309 if (minval.length() > 0)
311 min = new Float(minval).floatValue();
313 } catch (Exception e)
316 .println("Couldn't parse the minimum value for graduated colour for type ("
318 + ") - did you misspell 'auto' for the optional automatic colour switch ?");
323 if (maxval.length() > 0)
325 max = new Float(maxval).floatValue();
327 } catch (Exception e)
330 .println("Couldn't parse the maximum value for graduated colour for type ("
337 // add in some dummy min/max colours for the label-only
344 colour = new jalview.schemes.GraduatedColor(
345 new UserColourScheme(mincol).findColour('A'),
346 new UserColourScheme(maxcol).findColour('A'), min,
348 } catch (Exception e)
351 .println("Couldn't parse the graduated colour scheme ("
357 ((jalview.schemes.GraduatedColor) colour)
358 .setColourByLabel(labelCol);
359 ((jalview.schemes.GraduatedColor) colour)
360 .setAutoScaled(abso == null);
361 // add in any additional parameters
362 String ttype = null, tval = null;
363 if (gcol.hasMoreTokens())
365 // threshold type and possibly a threshold value
366 ttype = gcol.nextToken();
367 if (ttype.toLowerCase().startsWith("below"))
369 ((jalview.schemes.GraduatedColor) colour)
370 .setThreshType(AnnotationColourGradient.BELOW_THRESHOLD);
372 else if (ttype.toLowerCase().startsWith("above"))
374 ((jalview.schemes.GraduatedColor) colour)
375 .setThreshType(AnnotationColourGradient.ABOVE_THRESHOLD);
379 ((jalview.schemes.GraduatedColor) colour)
380 .setThreshType(AnnotationColourGradient.NO_THRESHOLD);
381 if (!ttype.toLowerCase().startsWith("no"))
384 .println("Ignoring unrecognised threshold type : "
389 if (((GraduatedColor) colour).getThreshType() != AnnotationColourGradient.NO_THRESHOLD)
394 tval = gcol.nextToken();
395 ((jalview.schemes.GraduatedColor) colour)
396 .setThresh(new Float(tval).floatValue());
397 } catch (Exception e)
400 .println("Couldn't parse threshold value as a float: ("
405 // parse the thresh-is-min token ?
406 if (gcol.hasMoreTokens())
409 .println("Ignoring additional tokens in parameters in graduated colour specification\n");
410 while (gcol.hasMoreTokens())
412 System.err.println("|" + gcol.nextToken());
414 System.err.println("\n");
420 UserColourScheme ucs = new UserColourScheme(colscheme);
421 colour = ucs.findColour('A');
425 colours.put(type, colour);
427 if (st.hasMoreElements())
429 String link = st.nextToken();
430 typeLink.put(type, link);
431 if (featureLink == null)
433 featureLink = new Hashtable();
435 featureLink.put(type, link);
441 while (st.hasMoreElements())
446 // Still possible this is an old Jalview file,
447 // which does not have type colours at the beginning
448 seqId = token = st.nextToken();
449 seq = findName(align, seqId, relaxedIdmatching, newseqs);
452 desc = st.nextToken();
454 if (doGffSource && desc.indexOf(' ') == -1)
456 // could also be a source term rather than description line
457 group = new String(desc);
459 type = st.nextToken();
462 String stt = st.nextToken();
463 if (stt.length() == 0 || stt.equals("-"))
469 start = Integer.parseInt(stt);
471 } catch (NumberFormatException ex)
477 String stt = st.nextToken();
478 if (stt.length() == 0 || stt.equals("-"))
484 end = Integer.parseInt(stt);
486 } catch (NumberFormatException ex)
490 // TODO: decide if non positional feature assertion for input data
491 // where end==0 is generally valid
494 // treat as non-positional feature, regardless.
499 score = new Float(st.nextToken()).floatValue();
500 } catch (NumberFormatException ex)
505 sf = new SequenceFeature(type, desc, start, end, score, group);
509 sf.setValue("STRAND", st.nextToken());
510 sf.setValue("FRAME", st.nextToken());
511 } catch (Exception ex)
515 if (st.hasMoreTokens())
517 StringBuffer attributes = new StringBuffer();
518 while (st.hasMoreTokens())
520 attributes.append("\t" + st.nextElement());
522 // TODO validate and split GFF2 attributes field ? parse out
523 // ([A-Za-z][A-Za-z0-9_]*) <value> ; and add as
524 // sf.setValue(attrib, val);
525 sf.setValue("ATTRIBUTES", attributes.toString());
528 seq.addSequenceFeature(sf);
529 while ((seq = align.findName(seq, seqId, true)) != null)
531 seq.addSequenceFeature(new SequenceFeature(sf));
537 if (GFFFile && seq == null)
543 desc = st.nextToken();
545 if (!st.hasMoreTokens())
548 .println("DEBUG: Run out of tokens when trying to identify the destination for the feature.. giving up.");
549 // in all probability, this isn't a file we understand, so bail
554 token = st.nextToken();
556 if (!token.equals("ID_NOT_SPECIFIED"))
558 seq = findName(align, seqId = token, relaxedIdmatching, null);
566 index = Integer.parseInt(st.nextToken());
567 seq = align.getSequenceAt(index);
568 } catch (NumberFormatException ex)
576 System.out.println("Sequence not found: " + line);
580 start = Integer.parseInt(st.nextToken());
581 end = Integer.parseInt(st.nextToken());
583 type = st.nextToken();
585 if (!colours.containsKey(type))
587 // Probably the old style groups file
588 UserColourScheme ucs = new UserColourScheme(type);
589 colours.put(type, ucs.findColour('A'));
591 sf = new SequenceFeature(type, desc, "", start, end, featureGroup);
592 if (st.hasMoreTokens())
596 score = new Float(st.nextToken()).floatValue();
597 // update colourgradient bounds if allowed to
598 } catch (NumberFormatException ex)
604 if (groupLink != null && removeHTML)
606 sf.addLink(groupLink);
607 sf.description += "%LINK%";
609 if (typeLink.containsKey(type) && removeHTML)
611 sf.addLink(typeLink.get(type).toString());
612 sf.description += "%LINK%";
615 parseDescriptionHTML(sf, removeHTML);
617 seq.addSequenceFeature(sf);
620 && (seq = align.findName(seq, seqId, false)) != null)
622 seq.addSequenceFeature(new SequenceFeature(sf));
624 // If we got here, its not a GFFFile
629 } catch (Exception ex)
631 System.out.println("Error parsing feature file: " + ex + "\n" + line);
632 ex.printStackTrace(System.err);
640 private AlignmentI lastmatchedAl = null;
642 private SequenceIdMatcher matcher = null;
645 * clear any temporary handles used to speed up ID matching
647 private void resetMatcher()
649 lastmatchedAl = null;
653 private SequenceI findName(AlignmentI align, String seqId,
654 boolean relaxedIdMatching, List<SequenceI> newseqs)
656 SequenceI match = null;
657 if (relaxedIdMatching)
659 if (lastmatchedAl != align)
661 matcher = new SequenceIdMatcher(
662 (lastmatchedAl = align).getSequencesArray());
665 matcher.addAll(newseqs);
668 match = matcher.findIdMatch(seqId);
672 match = align.findName(seqId, true);
675 if (match==null && newseqs!=null)
677 match = new SequenceDummy(seqId);
678 if (relaxedIdMatching)
680 matcher.addAll(Arrays.asList(new SequenceI[]
686 public void parseDescriptionHTML(SequenceFeature sf, boolean removeHTML)
688 if (sf.getDescription() == null)
692 jalview.util.ParseHtmlBodyAndLinks parsed = new jalview.util.ParseHtmlBodyAndLinks(
693 sf.getDescription(), removeHTML, newline);
695 sf.description = (removeHTML) ? parsed.getNonHtmlContent()
697 for (String link : parsed.getLinks())
705 * generate a features file for seqs includes non-pos features by default.
708 * source of sequence features
710 * hash of feature types and colours
711 * @return features file contents
713 public String printJalviewFormat(SequenceI[] seqs, Map<String,Object> visible)
715 return printJalviewFormat(seqs, visible, true, true);
719 * generate a features file for seqs with colours from visible (if any)
724 * hash of Colours for each feature type
726 * when true only feature types in 'visible' will be output
728 * indicates if non-positional features should be output (regardless
730 * @return features file contents
732 public String printJalviewFormat(SequenceI[] seqs, Map visible,
733 boolean visOnly, boolean nonpos)
735 StringBuffer out = new StringBuffer();
736 SequenceFeature[] next;
737 boolean featuresGen = false;
738 if (visOnly && !nonpos && (visible == null || visible.size() < 1))
740 // no point continuing.
741 return "No Features Visible";
744 if (visible != null && visOnly)
746 // write feature colours only if we're given them and we are generating
748 // TODO: decide if feature links should also be written here ?
749 Iterator en = visible.keySet().iterator();
753 type = en.next().toString();
755 if (visible.get(type) instanceof GraduatedColor)
757 GraduatedColor gc = (GraduatedColor) visible.get(type);
758 color = (gc.isColourByLabel() ? "label|" : "")
759 + Format.getHexString(gc.getMinColor()) + "|"
760 + Format.getHexString(gc.getMaxColor())
761 + (gc.isAutoScale() ? "|" : "|abso|") + gc.getMin() + "|"
763 if (gc.getThreshType() != AnnotationColourGradient.NO_THRESHOLD)
765 if (gc.getThreshType() == AnnotationColourGradient.BELOW_THRESHOLD)
771 if (gc.getThreshType() != AnnotationColourGradient.ABOVE_THRESHOLD)
773 System.err.println("WARNING: Unsupported threshold type ("
774 + gc.getThreshType() + ") : Assuming 'above'");
779 color += "|" + gc.getThresh();
786 else if (visible.get(type) instanceof java.awt.Color)
788 color = Format.getHexString((java.awt.Color) visible.get(type));
792 // legacy support for integer objects containing colour triplet values
793 color = Format.getHexString(new java.awt.Color(Integer
794 .parseInt(visible.get(type).toString())));
802 // Work out which groups are both present and visible
803 Vector groups = new Vector();
805 boolean isnonpos = false;
807 for (int i = 0; i < seqs.length; i++)
809 next = seqs[i].getSequenceFeatures();
812 for (int j = 0; j < next.length; j++)
814 isnonpos = next[j].begin == 0 && next[j].end == 0;
815 if ((!nonpos && isnonpos)
816 || (!isnonpos && visOnly && !visible
817 .containsKey(next[j].type)))
822 if (next[j].featureGroup != null
823 && !groups.contains(next[j].featureGroup))
825 groups.addElement(next[j].featureGroup);
835 if (groups.size() > 0 && groupIndex < groups.size())
837 group = groups.elementAt(groupIndex).toString();
839 out.append("STARTGROUP\t");
848 for (int i = 0; i < seqs.length; i++)
850 next = seqs[i].getSequenceFeatures();
853 for (int j = 0; j < next.length; j++)
855 isnonpos = next[j].begin == 0 && next[j].end == 0;
856 if ((!nonpos && isnonpos)
857 || (!isnonpos && visOnly && !visible
858 .containsKey(next[j].type)))
860 // skip if feature is nonpos and we ignore them or if we only
861 // output visible and it isn't non-pos and it's not visible
866 && (next[j].featureGroup == null || !next[j].featureGroup
872 if (group == null && next[j].featureGroup != null)
876 // we have features to output
878 if (next[j].description == null
879 || next[j].description.equals(""))
881 out.append(next[j].type + "\t");
885 if (next[j].links != null
886 && next[j].getDescription().indexOf("<html>") == -1)
888 out.append("<html>");
891 out.append(next[j].description + " ");
892 if (next[j].links != null)
894 for (int l = 0; l < next[j].links.size(); l++)
896 String label = next[j].links.elementAt(l).toString();
897 String href = label.substring(label.indexOf("|") + 1);
898 label = label.substring(0, label.indexOf("|"));
900 if (next[j].description.indexOf(href) == -1)
902 out.append("<a href=\"" + href + "\">" + label + "</a>");
906 if (next[j].getDescription().indexOf("</html>") == -1)
908 out.append("</html>");
914 out.append(seqs[i].getName());
915 out.append("\t-1\t");
916 out.append(next[j].begin);
918 out.append(next[j].end);
920 out.append(next[j].type);
921 if (next[j].score != Float.NaN)
924 out.append(next[j].score);
933 out.append("ENDGROUP\t");
943 } while (groupIndex < groups.size() + 1);
947 return "No Features Visible";
950 return out.toString();
954 * generate a gff file for sequence features includes non-pos features by
961 public String printGFFFormat(SequenceI[] seqs, Map<String,Object> visible)
963 return printGFFFormat(seqs, visible, true, true);
966 public String printGFFFormat(SequenceI[] seqs, Map<String,Object> visible,
967 boolean visOnly, boolean nonpos)
969 StringBuffer out = new StringBuffer();
970 SequenceFeature[] next;
973 for (int i = 0; i < seqs.length; i++)
975 if (seqs[i].getSequenceFeatures() != null)
977 next = seqs[i].getSequenceFeatures();
978 for (int j = 0; j < next.length; j++)
980 isnonpos = next[j].begin == 0 && next[j].end == 0;
981 if ((!nonpos && isnonpos)
982 || (!isnonpos && visOnly && !visible
983 .containsKey(next[j].type)))
988 source = next[j].featureGroup;
991 source = next[j].getDescription();
994 out.append(seqs[i].getName());
998 out.append(next[j].type);
1000 out.append(next[j].begin);
1002 out.append(next[j].end);
1004 out.append(next[j].score);
1007 if (next[j].getValue("STRAND") != null)
1009 out.append(next[j].getValue("STRAND"));
1017 if (next[j].getValue("FRAME") != null)
1019 out.append(next[j].getValue("FRAME"));
1025 // TODO: verify/check GFF - should there be a /t here before attribute
1028 if (next[j].getValue("ATTRIBUTES") != null)
1030 out.append(next[j].getValue("ATTRIBUTES"));
1033 out.append(newline);
1039 return out.toString();
1043 * this is only for the benefit of object polymorphism - method does nothing.
1051 * this is only for the benefit of object polymorphism - method does nothing.
1053 * @return error message
1055 public String print()
1057 return "USE printGFFFormat() or printJalviewFormat()";