-package jalview.analysis;\r
-\r
-import com.stevesoft.pat.Regex;\r
-\r
-import jalview.datamodel.*;\r
-\r
-public class ParseProperties\r
-{\r
- /**\r
- * Methods for parsing free text properties on alignments and sequences.\r
- * There are a number of ways we might want to do this:\r
- * arbitrary regex. and an associated score name for the number that's extracted.\r
- * Regex that provides both score and name.\r
- * \r
- * We may also want to :\r
- * - modify description to remove parsed numbers (this behaviour is dangerous since exporting the alignment would lose the original form then)\r
- * - \r
- * \r
- */\r
- /**\r
- * The alignment being operated on\r
- */\r
- private AlignmentI al=null;\r
- \r
- /**\r
- * initialise a new property parser\r
- * @param al\r
- */\r
- public ParseProperties(AlignmentI al) {\r
- this.al = al;\r
- }\r
-\r
- public int getScoresFromDescription(String ScoreName, String ScoreDescriptions, String regex, boolean repeat)\r
- {\r
- return getScoresFromDescription(new String[] { ScoreName }, new String[] { ScoreDescriptions}, regex, repeat);\r
- }\r
-\r
- public int getScoresFromDescription(String[] ScoreNames, String[] ScoreDescriptions, String regex, boolean repeat) \r
- {\r
- return getScoresFromDescription(al.getSequencesArray(), ScoreNames, ScoreDescriptions, regex, repeat);\r
- }\r
- /**\r
- * Extract scores for sequences by applying regex to description string.\r
- * @param seqs seuqences to extract annotation from.\r
- * @param ScoreNames labels for each numeric field in regex match\r
- * @param ScoreDescriptions description for each numeric field in regex match\r
- * @param regex Regular Expression string for passing to <code>new com.stevesoft.patt.Regex(regex)</code>\r
- * @param repeat true means the regex will be applied multiple times along the description string of each sequence\r
- * @return total number of sequences that matched the regex\r
- */\r
- public int getScoresFromDescription(SequenceI[] seqs, String[] ScoreNames, String[] ScoreDescriptions, String regex, boolean repeat) \r
- {\r
- int count=0;\r
- Regex pattern = new Regex(regex);\r
- if (pattern.numSubs()>ScoreNames.length)\r
- {\r
- // Check that we have enough labels and descriptions for any parsed scores.\r
- int onamelen = ScoreNames.length;\r
- String[] tnames = new String[pattern.numSubs()+1];\r
- System.arraycopy(ScoreNames, 0, tnames, 0, ScoreNames.length);\r
- String base = tnames[ScoreNames.length-1];\r
- ScoreNames = tnames;\r
- String descrbase = ScoreDescriptions[onamelen-1];\r
- if (descrbase == null)\r
- descrbase = "Score parsed from ("+regex+")";\r
- tnames = new String[pattern.numSubs()+1];\r
- System.arraycopy(ScoreDescriptions, 0, tnames, 0, ScoreDescriptions.length);\r
- ScoreDescriptions = tnames;\r
- for (int i=onamelen; i<ScoreNames.length; i++)\r
- {\r
- ScoreNames[i] = base+"_"+i;\r
- ScoreDescriptions[i] = descrbase+" (column "+i+")";\r
- }\r
- }\r
- for (int i=0; i<seqs.length; i++)\r
- {\r
- String descr = seqs[i].getDescription();\r
- if (descr==null)\r
- continue;\r
- int pos=0;\r
- boolean added=false;\r
- while ((repeat || pos==0) && pattern.searchFrom(descr, pos))\r
- {\r
- pos = pattern.matchedTo();\r
- for (int cols=0; cols<pattern.numSubs(); cols++)\r
- {\r
- String sstring = pattern.stringMatched(cols+1);\r
- double score=Double.NaN;\r
- try {\r
- score = new Double(sstring).doubleValue();\r
- }\r
- catch (Exception e)\r
- {\r
- // don't try very hard to parse if regex was wrong.\r
- continue;\r
- }\r
- // add score to sequence annotation.\r
- AlignmentAnnotation an = new AlignmentAnnotation(ScoreNames[cols], ScoreDescriptions[cols], null);\r
- an.setScore(score);\r
- System.out.println("Score: "+ScoreNames[cols]+"="+score); // DEBUG\r
- an.setSequenceRef(seqs[i]);\r
- seqs[i].addAlignmentAnnotation(an);\r
- al.addAnnotation(an);\r
- added=true;\r
- }\r
- }\r
- if (added)\r
- {\r
- count++;\r
- }\r
- }\r
- return count; \r
- }\r
- public static void main(String argv[]) {\r
- SequenceI[] seqs = new SequenceI[] { new Sequence("sq1","THISISAPLACEHOLDER"),\r
- new Sequence("sq2","THISISAPLACEHOLDER"),\r
- new Sequence("sq3","THISISAPLACEHOLDER"),\r
- new Sequence("sq4","THISISAPLACEHOLDER")};\r
- seqs[0].setDescription("1 mydescription1");\r
- seqs[1].setDescription("mydescription2");\r
- seqs[2].setDescription("2. 0.1 mydescription3");\r
- seqs[3].setDescription("3 0.01 mydescription4");\r
- //seqs[4].setDescription("5 mydescription5");\r
- Alignment al = new Alignment(seqs);\r
- ParseProperties pp = new ParseProperties(al);\r
- String regex = ".*([-0-9.+]+)";\r
- System.out.println("Matched "+pp.getScoresFromDescription("my Score", "my Score Description",regex, true)+" for "+regex);\r
- regex = ".*([-0-9.+]+).+([-0-9.+]+).*";\r
- System.out.println("Matched "+pp.getScoresFromDescription("my Score", "my Score Description",regex, true)+" for "+regex);\r
- System.out.println("Finished.");\r
- }\r
-}\r
+package jalview.analysis;
+
+import com.stevesoft.pat.Regex;
+
+import jalview.datamodel.*;
+
+public class ParseProperties
+{
+ /**
+ * Methods for parsing free text properties on alignments and sequences.
+ * There are a number of ways we might want to do this:
+ * arbitrary regex. and an associated score name for the number that's extracted.
+ * Regex that provides both score and name.
+ *
+ * We may also want to :
+ * - modify description to remove parsed numbers (this behaviour is dangerous since exporting the alignment would lose the original form then)
+ * -
+ *
+ */
+ /**
+ * The alignment being operated on
+ */
+ private AlignmentI al=null;
+
+ /**
+ * initialise a new property parser
+ * @param al
+ */
+ public ParseProperties(AlignmentI al) {
+ this.al = al;
+ }
+
+ public int getScoresFromDescription(String ScoreName, String ScoreDescriptions, String regex, boolean repeat)
+ {
+ return getScoresFromDescription(new String[] { ScoreName }, new String[] { ScoreDescriptions}, regex, repeat);
+ }
+
+ public int getScoresFromDescription(String[] ScoreNames, String[] ScoreDescriptions, String regex, boolean repeat)
+ {
+ return getScoresFromDescription(al.getSequencesArray(), ScoreNames, ScoreDescriptions, regex, repeat);
+ }
+ /**
+ * Extract scores for sequences by applying regex to description string.
+ * @param seqs seuqences to extract annotation from.
+ * @param ScoreNames labels for each numeric field in regex match
+ * @param ScoreDescriptions description for each numeric field in regex match
+ * @param regex Regular Expression string for passing to <code>new com.stevesoft.patt.Regex(regex)</code>
+ * @param repeat true means the regex will be applied multiple times along the description string of each sequence
+ * @return total number of sequences that matched the regex
+ */
+ public int getScoresFromDescription(SequenceI[] seqs, String[] ScoreNames, String[] ScoreDescriptions, String regex, boolean repeat)
+ {
+ int count=0;
+ Regex pattern = new Regex(regex);
+ if (pattern.numSubs()>ScoreNames.length)
+ {
+ // Check that we have enough labels and descriptions for any parsed scores.
+ int onamelen = ScoreNames.length;
+ String[] tnames = new String[pattern.numSubs()+1];
+ System.arraycopy(ScoreNames, 0, tnames, 0, ScoreNames.length);
+ String base = tnames[ScoreNames.length-1];
+ ScoreNames = tnames;
+ String descrbase = ScoreDescriptions[onamelen-1];
+ if (descrbase == null)
+ descrbase = "Score parsed from ("+regex+")";
+ tnames = new String[pattern.numSubs()+1];
+ System.arraycopy(ScoreDescriptions, 0, tnames, 0, ScoreDescriptions.length);
+ ScoreDescriptions = tnames;
+ for (int i=onamelen; i<ScoreNames.length; i++)
+ {
+ ScoreNames[i] = base+"_"+i;
+ ScoreDescriptions[i] = descrbase+" (column "+i+")";
+ }
+ }
+ for (int i=0; i<seqs.length; i++)
+ {
+ String descr = seqs[i].getDescription();
+ if (descr==null)
+ continue;
+ int pos=0;
+ boolean added=false;
+ int reps=0;
+ while ((repeat || pos==0) && pattern.searchFrom(descr, pos))
+ {
+ pos = pattern.matchedTo();
+ for (int cols=0; cols<pattern.numSubs(); cols++)
+ {
+ String sstring = pattern.stringMatched(cols+1);
+ double score=Double.NaN;
+ try {
+ score = new Double(sstring).doubleValue();
+ }
+ catch (Exception e)
+ {
+ // don't try very hard to parse if regex was wrong.
+ continue;
+ }
+ // add score to sequence annotation.
+ AlignmentAnnotation an = new AlignmentAnnotation(ScoreNames[cols] +((reps>0) ? "_"+reps : ""), ScoreDescriptions[cols], null);
+ an.setScore(score);
+ System.out.println("Score: "+ScoreNames[cols]+"="+score); // DEBUG
+ an.setSequenceRef(seqs[i]);
+ seqs[i].addAlignmentAnnotation(an);
+ al.addAnnotation(an);
+ added=true;
+ }
+ reps++; // repeated matches
+ }
+ if (added)
+ {
+ count++;
+ }
+ }
+ return count;
+ }
+ public static void main(String argv[]) {
+ SequenceI[] seqs = new SequenceI[] { new Sequence("sq1","THISISAPLACEHOLDER"),
+ new Sequence("sq2","THISISAPLACEHOLDER"),
+ new Sequence("sq3","THISISAPLACEHOLDER"),
+ new Sequence("sq4","THISISAPLACEHOLDER")};
+ seqs[0].setDescription("1 mydescription1");
+ seqs[1].setDescription("mydescription2");
+ seqs[2].setDescription("2. 0.1 mydescription3");
+ seqs[3].setDescription("3 0.01 mydescription4");
+ //seqs[4].setDescription("5 mydescription5");
+ Alignment al = new Alignment(seqs);
+ ParseProperties pp = new ParseProperties(al);
+ String regex = ".*([-0-9.+]+)";
+ System.out.println("Matched "+pp.getScoresFromDescription("my Score", "my Score Description",regex, true)+" for "+regex);
+ regex = ".*([-0-9.+]+).+([-0-9.+]+).*";
+ System.out.println("Matched "+pp.getScoresFromDescription("my Score", "my Score Description",regex, true)+" for "+regex);
+ System.out.println("Finished.");
+ }
+}
git://source.jalview.org / jalview.git / commitdiff