/// <exception cref="System.IO.IOException"/>
public virtual void Annotate(ICoreMap document)
{
// write input file in GUTime format
IElement inputXML = ToInputXML(document);
File inputFile = File.CreateTempFile("gutime", ".input");
//Document doc = new Document(inputXML);
PrintWriter inputWriter = new PrintWriter(inputFile);
inputWriter.Println(XMLUtils.NodeToString(inputXML, false));
// new XMLOutputter().output(inputXML, inputWriter);
inputWriter.Close();
bool useFirstDate = (!document.ContainsKey(typeof(CoreAnnotations.CalendarAnnotation)) && !document.ContainsKey(typeof(CoreAnnotations.DocDateAnnotation)));
List <string> args = new List <string>();
args.Add("perl");
args.Add("-I" + this.gutimePath.GetPath());
args.Add(new File(this.gutimePath, "TimeTag.pl").GetPath());
if (useFirstDate)
{
args.Add("-FDNW");
}
args.Add(inputFile.GetPath());
// run GUTime on the input file
ProcessBuilder process = new ProcessBuilder(args);
StringWriter outputWriter = new StringWriter();
SystemUtils.Run(process, outputWriter, null);
string output = outputWriter.GetBuffer().ToString();
Pattern docClose = Pattern.Compile("</DOC>.*", Pattern.Dotall);
output = docClose.Matcher(output).ReplaceAll("</DOC>");
//The TimeTag.pl result file contains next tags which must be removed
output = output.ReplaceAll("<lex.*?>", string.Empty);
output = output.Replace("</lex>", string.Empty);
output = output.Replace("<NG>", string.Empty);
output = output.Replace("</NG>", string.Empty);
output = output.Replace("<VG>", string.Empty);
output = output.Replace("</VG>", string.Empty);
output = output.Replace("<s>", string.Empty);
output = output.Replace("</s>", string.Empty);
// parse the GUTime output
IElement outputXML;
try
{
outputXML = XMLUtils.ParseElement(output);
}
catch (Exception ex)
{
throw new Exception(string.Format("error:\n%s\ninput:\n%s\noutput:\n%s", ex, IOUtils.SlurpFile(inputFile), output), ex);
}
/*
* try {
* outputXML = new SAXBuilder().build(new StringReader(output)).getRootElement();
* } catch (JDOMException e) {
* throw new RuntimeException(String.format("error:\n%s\ninput:\n%s\noutput:\n%s",
* e, IOUtils.slurpFile(inputFile), output));
* } */
inputFile.Delete();
// get Timex annotations
IList <ICoreMap> timexAnns = ToTimexCoreMaps(outputXML, document);
document.Set(typeof(TimeAnnotations.TimexAnnotations), timexAnns);
if (outputResults)
{
System.Console.Out.WriteLine(timexAnns);
}
// align Timex annotations to sentences
int timexIndex = 0;
foreach (ICoreMap sentence in document.Get(typeof(CoreAnnotations.SentencesAnnotation)))
{
int sentBegin = BeginOffset(sentence);
int sentEnd = EndOffset(sentence);
// skip times before the sentence
while (timexIndex < timexAnns.Count && BeginOffset(timexAnns[timexIndex]) < sentBegin)
{
++timexIndex;
}
// determine times within the sentence
int sublistBegin = timexIndex;
int sublistEnd = timexIndex;
while (timexIndex < timexAnns.Count && sentBegin <= BeginOffset(timexAnns[timexIndex]) && EndOffset(timexAnns[timexIndex]) <= sentEnd)
{
++sublistEnd;
++timexIndex;
}
// set the sentence timexes
sentence.Set(typeof(TimeAnnotations.TimexAnnotations), timexAnns.SubList(sublistBegin, sublistEnd));
}
}
private void Init(IElement element)
{
Init(XMLUtils.NodeToString(element, false), element);
}