public virtual void PrintSentences(IEnumerable <IList <IHasWord> > sentences, string filename)
{
try
{
PrintWriter pw = IOUtils.GetPrintWriter(filename);
foreach (IList <IHasWord> sentence in sentences)
{
pw.Print("<s> ");
// Note: Use <s sentence-id > to identify sentences
string sentString = SentenceUtils.ListToString(sentence);
if (sentence.Count > maxSentenceLength)
{
logger.Warning("Sentence length=" + sentence.Count + " is longer than maximum set length " + maxSentenceLength);
logger.Warning("Long Sentence: " + sentString);
}
pw.Print(sentString);
pw.Println(" </s>");
}
pw.Close();
}
catch (IOException ex)
{
throw new Exception(ex);
}
}
/*for(int i = 0; i < data.length; i++) {
* for(int j = 0; j < data[i].length; j++) {
* System.out.println(data[i][j]);
* }
* }*/
/// <summary>prints the full feature matrix in tab-delimited form.</summary>
/// <remarks>
/// prints the full feature matrix in tab-delimited form. These can be BIG
/// matrices, so be careful! [Can also use printFullFeatureMatrixWithValues]
/// </remarks>
public virtual void PrintFullFeatureMatrix(PrintWriter pw)
{
string sep = "\t";
for (int i = 0; i < featureIndex.Size(); i++)
{
pw.Print(sep + featureIndex.Get(i));
}
pw.Println();
for (int i_1 = 0; i_1 < labels.Length; i_1++)
{
pw.Print(labelIndex.Get(i_1));
ICollection <int> feats = Generics.NewHashSet();
for (int j = 0; j < data[i_1].Length; j++)
{
int feature = data[i_1][j];
feats.Add(int.Parse(feature));
}
for (int j_1 = 0; j_1 < featureIndex.Size(); j_1++)
{
if (feats.Contains(int.Parse(j_1)))
{
pw.Print(sep + "1");
}
else
{
pw.Print(sep + "0");
}
}
pw.Println();
}
}
/// <exception cref="System.IO.IOException"/>
private void AddResults(IHttpServletRequest request, IHttpServletResponse response)
{
string input = request.GetParameter("input");
if (input == null)
{
return;
}
input = input.Trim();
if (input.IsEmpty())
{
return;
}
PrintWriter @out = response.GetWriter();
if (input.Length > MaximumQueryLength)
{
@out.Print("This query is too long. If you want to run very long queries, please download and use our <a href=\"http://nlp.stanford.edu/software/CRF-NER.html\">publicly released distribution</a>.");
return;
}
string outputFormat = request.GetParameter("outputFormat");
if (outputFormat == null || outputFormat.Trim().IsEmpty())
{
outputFormat = this.format;
}
bool preserveSpacing;
string preserveSpacingStr = request.GetParameter("preserveSpacing");
if (preserveSpacingStr == null || preserveSpacingStr.Trim().IsEmpty())
{
preserveSpacing = this.spacing;
}
else
{
preserveSpacingStr = preserveSpacingStr.Trim();
preserveSpacing = bool.ValueOf(preserveSpacingStr);
}
string classifier = request.GetParameter("classifier");
if (classifier == null || classifier.Trim().IsEmpty())
{
classifier = this.defaultClassifier;
}
response.AddHeader("classifier", classifier);
response.AddHeader("outputFormat", outputFormat);
response.AddHeader("preserveSpacing", preserveSpacing.ToString());
if (outputFormat.Equals("highlighted"))
{
OutputHighlighting(@out, ners[classifier], input);
}
else
{
@out.Print(StringEscapeUtils.EscapeHtml4(ners[classifier].ClassifyToString(input, outputFormat, preserveSpacing)));
}
}
private static void PrintAnswersTokenizedTextTsv <In>(IList <In> l, PrintWriter @out)
where In : ICoreMap
{
foreach (IN wi in l)
{
@out.Print(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.TextAnnotation))));
@out.Print('\t');
@out.Println(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.AnswerAnnotation))));
}
@out.Println();
}
public virtual void PrintAnswers(IList <CoreLabel> doc, PrintWriter pw)
{
string prevAnswer = "O";
string prevClass = string.Empty;
string afterLast = string.Empty;
foreach (CoreLabel word in doc)
{
if (!prevAnswer.Equals("O") && !prevAnswer.Equals(word.Get(typeof(CoreAnnotations.AnswerAnnotation))))
{
pw.Print("</" + prevClass + ">");
prevClass = string.Empty;
}
pw.Print(word.Get(typeof(CoreAnnotations.BeforeAnnotation)));
if (!word.Get(typeof(CoreAnnotations.AnswerAnnotation)).Equals("O") && !word.Get(typeof(CoreAnnotations.AnswerAnnotation)).Equals(prevAnswer))
{
if (Sharpen.Runtime.EqualsIgnoreCase(word.Get(typeof(CoreAnnotations.AnswerAnnotation)), "PERSON") || Sharpen.Runtime.EqualsIgnoreCase(word.Get(typeof(CoreAnnotations.AnswerAnnotation)), "ORGANIZATION") || Sharpen.Runtime.EqualsIgnoreCase(word
.Get(typeof(CoreAnnotations.AnswerAnnotation)), "LOCATION"))
{
prevClass = "ENAMEX";
}
else
{
if (Sharpen.Runtime.EqualsIgnoreCase(word.Get(typeof(CoreAnnotations.AnswerAnnotation)), "DATE") || Sharpen.Runtime.EqualsIgnoreCase(word.Get(typeof(CoreAnnotations.AnswerAnnotation)), "TIME"))
{
prevClass = "TIMEX";
}
else
{
if (Sharpen.Runtime.EqualsIgnoreCase(word.Get(typeof(CoreAnnotations.AnswerAnnotation)), "PERCENT") || Sharpen.Runtime.EqualsIgnoreCase(word.Get(typeof(CoreAnnotations.AnswerAnnotation)), "MONEY"))
{
prevClass = "NUMEX";
}
else
{
log.Info("unknown type: " + word.Get(typeof(CoreAnnotations.AnswerAnnotation)));
System.Environment.Exit(0);
}
}
}
pw.Print("<" + prevClass + " TYPE=\"" + word.Get(typeof(CoreAnnotations.AnswerAnnotation)) + "\">");
}
pw.Print(word.Get(typeof(CoreAnnotations.OriginalTextAnnotation)));
afterLast = word.Get(typeof(CoreAnnotations.AfterAnnotation));
prevAnswer = word.Get(typeof(CoreAnnotations.AnswerAnnotation));
}
if (!prevAnswer.Equals("O"))
{
pw.Print("</" + prevClass + ">");
prevClass = string.Empty;
}
pw.Println(afterLast);
}
// put a single newline at the end [added 20091024].
private static void PrintAnswersAsIsText <In>(IList <In> l, PrintWriter @out)
where In : ICoreMap
{
foreach (IN wi in l)
{
@out.Print(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.BeforeAnnotation))));
@out.Print(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.TextAnnotation))));
@out.Print('/');
@out.Print(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.AnswerAnnotation))));
@out.Print(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.AfterAnnotation))));
}
}
/// <summary>
/// <inheritDoc/>
///
/// </summary>
public override void PrintSparseFeatureMatrix(PrintWriter pw)
{
string sep = "\t";
for (int i = 0; i < size; i++)
{
pw.Print(labelIndex.Get(labels[i]));
int[] datum = data[i];
foreach (int j in datum)
{
pw.Print(sep + featureIndex.Get(j));
}
pw.Println();
}
}
//logger.info(lwi.get(i));
public virtual void PrintAnswers(IList <CoreLabel> doc, PrintWriter pw)
{
string ansStr = ChineseStringUtils.CombineSegmentedSentence(doc, flags);
pw.Print(ansStr);
pw.Println();
}
public void SendToPrinter(string val)
{
try
{
Socket socket = new Socket("192.168.8.105", 1100);
// Exmp : Socket socket = new Socket("192.168.0.101", 80);
PrintWriter oStream = new PrintWriter(socket.OutputStream);
oStream.Print("\t\t Text to The Printer");
oStream.Print("\n\n\n");
oStream.Close();
socket.Close();
}
catch (Exception e)
{
}
}
/// <summary><inheritDoc/></summary>
/// <exception cref="Javax.Servlet.ServletException"/>
/// <exception cref="System.IO.IOException"/>
protected override void DoGet(IHttpServletRequest request, IHttpServletResponse response)
{
logger.Info("GET SPIED query from " + request.GetRemoteAddr());
if (request.GetCharacterEncoding() == null)
{
request.SetCharacterEncoding("utf-8");
}
response.SetContentType("text/json; charset=UTF-8");
PrintWriter @out = response.GetWriter();
try
{
string raw = request.GetParameter("q");
string seedwords = request.GetParameter("seedwords");
string model = request.GetParameter("model");
if (raw == null || string.Empty.Equals(raw))
{
@out.Print("{\"okay\":false,\"reason\":\"No data provided\"}");
}
else
{
Run(@out, raw, seedwords, model);
}
}
catch (Exception t)
{
WriteError(t, @out, request.ToString());
}
@out.Close();
}
/// <summary>Prints a sparse feature-value output of the Dataset.</summary>
/// <remarks>
/// Prints a sparse feature-value output of the Dataset. Prints the actual
/// <see cref="object.ToString()"/>
/// representations of features. This is probably
/// what you want for RVFDataset since the above two methods seem useless and
/// unused.
/// </remarks>
public virtual void PrintSparseFeatureValues(int datumNo, PrintWriter pw)
{
pw.Print(labelIndex.Get(labels[datumNo]));
pw.Print('\t');
pw.Println("LABEL");
int[] datum = data[datumNo];
double[] vals = values[datumNo];
System.Diagnostics.Debug.Assert(datum.Length == vals.Length);
for (int i = 0; i < datum.Length; i++)
{
pw.Print(featureIndex.Get(datum[i]));
pw.Print('\t');
pw.Println(vals[i]);
}
pw.Println();
}
public virtual void RecordScore(IKBestViterbiParser parser, PrintWriter pw)
{
double score = parser.GetBestScore();
totScore += score;
n++;
if (pw != null)
{
pw.Print(str + " score: " + nf.Format(score));
if (runningAverages)
{
pw.Print(" average score: " + nf.Format(totScore / n));
}
pw.Println();
}
}
/// <exception cref="System.IO.IOException"/>
private static void Print(Annotation annotation, PrintWriter pw, AnnotationOutputter.Options options)
{
IList <ICoreMap> sentences = annotation.Get(typeof(CoreAnnotations.SentencesAnnotation));
if (sentences != null)
{
for (int i = 0; i < sentences.Count; i++)
{
ICoreMap sentence = sentences[i];
StringBuilder sentenceToWrite = new StringBuilder();
foreach (CoreLabel token in sentence.Get(typeof(CoreAnnotations.TokensAnnotation)))
{
sentenceToWrite.Append(" ");
sentenceToWrite.Append(token.Lemma().ToLower());
if (token.Get(typeof(CoreAnnotations.PartOfSpeechAnnotation)).StartsWith("V"))
{
//verb
sentenceToWrite.Append("_V");
}
else
{
if (token.Get(typeof(CoreAnnotations.PartOfSpeechAnnotation)).StartsWith("N"))
{
//noun
sentenceToWrite.Append("_N");
}
}
}
pw.Print(sentenceToWrite.ToString());
}
}
}
private static void CountTaggings(Treebank tb, PrintWriter pw)
{
TwoDimensionalCounter <string, string> wtc = new TwoDimensionalCounter <string, string>();
tb.Apply(null);
foreach (string key in wtc.FirstKeySet())
{
pw.Print(key);
pw.Print('\t');
ICounter <string> ctr = wtc.GetCounter(key);
foreach (string k2 in ctr.KeySet())
{
pw.Print(k2 + '\t' + ctr.GetCount(k2) + '\t');
}
pw.Println();
}
}
private void PrintAnswersTokenizedTextTabbed(IList <In> l, PrintWriter @out)
{
string background = flags.backgroundSymbol;
string lastEntityType = null;
foreach (IN wi in l)
{
string entityType = wi.Get(typeof(CoreAnnotations.AnswerAnnotation));
string token = StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.TextAnnotation)));
if (entityType.Equals(lastEntityType))
{
// continue the same entity in column 1 or 3
@out.Print(' ');
@out.Print(token);
}
else
{
if (lastEntityType != null && !background.Equals(lastEntityType))
{
// different entity type. If previous not background/start, write in column 2
@out.Print('\t');
@out.Print(lastEntityType);
}
if (background.Equals(entityType))
{
// we'll print it in column 3. Normally, we're in column 2, unless we were at the start of doc
if (lastEntityType == null)
{
@out.Print('\t');
}
@out.Print('\t');
}
else
{
// otherwise we're printing in column 1 again
@out.Println();
}
@out.Print(token);
lastEntityType = entityType;
}
}
// if we're in the middle of printing an entity, then we should print its type
if (lastEntityType != null && !background.Equals(lastEntityType))
{
@out.Print('\t');
@out.Print(lastEntityType);
}
// finish line then add blank line
@out.Println();
@out.Println();
}
/// <exception cref="System.IO.IOException"/>
public override void Print(Annotation doc, OutputStream target, AnnotationOutputter.Options options)
{
PrintWriter writer = new PrintWriter(IOUtils.EncodedOutputStreamWriter(target, options.encoding));
IList <ICoreMap> sentences = doc.Get(typeof(CoreAnnotations.SentencesAnnotation));
foreach (ICoreMap sentence in sentences)
{
SemanticGraph sg = sentence.Get(typeof(SemanticGraphCoreAnnotations.BasicDependenciesAnnotation));
if (sg != null)
{
writer.Print(conllUWriter.PrintSemanticGraph(sg));
}
else
{
writer.Print(conllUWriter.PrintPOSAnnotations(sentence));
}
}
writer.Flush();
}
internal virtual void WriteError(Exception t, PrintWriter @out, string input)
{
StringWriter sw = new StringWriter();
PrintWriter pw = new PrintWriter(sw);
Sharpen.Runtime.PrintStackTrace(t, pw);
logger.Info("input is " + input);
logger.Info(sw.ToString());
@out.Print("{\"okay\":false, \"reason\":\"Something bad happened. Contact the author.\"}");
}
public virtual void Evaluate(Tree guess, Tree gold, PrintWriter pw, double weight)
{
ICollection <object> dep1 = MakeObjects(guess);
ICollection <object> dep2 = MakeObjects(gold);
double curPrecision = Precision(dep1, dep2);
double curRecall = Precision(dep2, dep1);
curF1 = (curPrecision > 0.0 && curRecall > 0.0 ? 2.0 / (1.0 / curPrecision + 1.0 / curRecall) : 0.0);
precision += curPrecision * weight;
recall += curRecall * weight;
f1 += curF1 * weight;
num += weight;
precision2 += dep1.Count * curPrecision * weight;
pnum2 += dep1.Count * weight;
recall2 += dep2.Count * curRecall * weight;
rnum2 += dep2.Count * weight;
if (curF1 > 0.9999)
{
exact += 1.0;
}
if (pw != null)
{
pw.Print(" P: " + ((int)(curPrecision * 10000)) / 100.0);
if (runningAverages)
{
pw.Println(" (sent ave " + ((int)(precision * 10000 / num)) / 100.0 + ") (evalb " + ((int)(precision2 * 10000 / pnum2)) / 100.0 + ")");
}
pw.Print(" R: " + ((int)(curRecall * 10000)) / 100.0);
if (runningAverages)
{
pw.Print(" (sent ave " + ((int)(recall * 10000 / num)) / 100.0 + ") (evalb " + ((int)(recall2 * 10000 / rnum2)) / 100.0 + ")");
}
pw.Println();
double cF1 = 2.0 / (rnum2 / recall2 + pnum2 / precision2);
pw.Print(str + " F1: " + ((int)(curF1 * 10000)) / 100.0);
if (runningAverages)
{
pw.Print(" (sent ave " + ((int)(10000 * f1 / num)) / 100.0 + ", evalb " + ((int)(10000 * cF1)) / 100.0 + ") Exact: " + ((int)(10000 * exact / num)) / 100.0);
}
// pw.println(" N: " + getNum());
pw.Println(" N: " + num);
}
}
private static void EmitSortedTrees(PriorityQueue <Triple <double, Tree, Tree> > queue, int worstKTreesToEmit, string filePrefix)
{
if (queue == null)
{
log.Info("Queue was not initialized properly");
}
try
{
PrintWriter guessPw = new PrintWriter(new BufferedWriter(new OutputStreamWriter(new FileOutputStream(filePrefix + ".kworst.guess"), "UTF-8")));
PrintWriter goldPw = new PrintWriter(new BufferedWriter(new OutputStreamWriter(new FileOutputStream(filePrefix + ".kworst.gold"), "UTF-8")));
IConstituentFactory cFact = new LabeledScoredConstituentFactory();
PrintWriter guessDepPw = new PrintWriter(new BufferedWriter(new OutputStreamWriter(new FileOutputStream(filePrefix + ".kworst.guess.deps"), "UTF-8")));
PrintWriter goldDepPw = new PrintWriter(new BufferedWriter(new OutputStreamWriter(new FileOutputStream(filePrefix + ".kworst.gold.deps"), "UTF-8")));
System.Console.Out.Printf("F1s of %d worst trees:\n", worstKTreesToEmit);
for (int i = 0; queue.Peek() != null && i < worstKTreesToEmit; i++)
{
Triple <double, Tree, Tree> trees = queue.Poll();
System.Console.Out.WriteLine(trees.First());
//Output the trees
goldPw.Println(trees.Second().ToString());
guessPw.Println(trees.Third().ToString());
//Output the set differences
ICollection <Constituent> goldDeps = Generics.NewHashSet();
Sharpen.Collections.AddAll(goldDeps, trees.Second().Constituents(cFact));
goldDeps.RemoveAll(trees.Third().Constituents(cFact));
foreach (Constituent c in goldDeps)
{
goldDepPw.Print(c.ToString() + " ");
}
goldDepPw.Println();
ICollection <Constituent> guessDeps = Generics.NewHashSet();
Sharpen.Collections.AddAll(guessDeps, trees.Third().Constituents(cFact));
guessDeps.RemoveAll(trees.Second().Constituents(cFact));
foreach (Constituent c_1 in guessDeps)
{
guessDepPw.Print(c_1.ToString() + " ");
}
guessDepPw.Println();
}
guessPw.Close();
goldPw.Close();
goldDepPw.Close();
guessDepPw.Close();
}
catch (UnsupportedEncodingException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
catch (FileNotFoundException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
}
/// <summary>Print an Annotation to an output stream.</summary>
/// <remarks>
/// Print an Annotation to an output stream.
/// The target OutputStream is assumed to already by buffered.
/// </remarks>
/// <param name="doc"/>
/// <param name="target"/>
/// <param name="options"/>
/// <exception cref="System.IO.IOException"/>
public override void Print(Annotation doc, OutputStream target, AnnotationOutputter.Options options)
{
PrintWriter writer = new PrintWriter(IOUtils.EncodedOutputStreamWriter(target, options.encoding));
// vv A bunch of nonsense to get tokens vv
if (doc.Get(typeof(CoreAnnotations.SentencesAnnotation)) != null)
{
foreach (ICoreMap sentence in doc.Get(typeof(CoreAnnotations.SentencesAnnotation)))
{
if (sentence.Get(typeof(CoreAnnotations.TokensAnnotation)) != null)
{
IList <CoreLabel> tokens = sentence.Get(typeof(CoreAnnotations.TokensAnnotation));
SemanticGraph depTree = sentence.Get(typeof(SemanticGraphCoreAnnotations.BasicDependenciesAnnotation));
for (int i = 0; i < tokens.Count; ++i)
{
// ^^ end nonsense to get tokens ^^
// Try to get the incoming dependency edge
int head = -1;
string deprel = null;
if (depTree != null)
{
ICollection <int> rootSet = depTree.GetRoots().Stream().Map(null).Collect(Collectors.ToSet());
IndexedWord node = depTree.GetNodeByIndexSafe(i + 1);
if (node != null)
{
IList <SemanticGraphEdge> edgeList = depTree.GetIncomingEdgesSorted(node);
if (!edgeList.IsEmpty())
{
System.Diagnostics.Debug.Assert(edgeList.Count == 1);
head = edgeList[0].GetGovernor().Index();
deprel = edgeList[0].GetRelation().ToString();
}
else
{
if (rootSet.Contains(i + 1))
{
head = 0;
deprel = "ROOT";
}
}
}
}
// Write the token
writer.Print(Line(i + 1, tokens[i], head, deprel));
writer.Println();
}
}
writer.Println();
}
}
// extra blank line at end of sentence
writer.Flush();
}
private void PrettyPrint(PrintWriter pw, int indent)
{
for (int i = 0; i < indent; i++)
{
pw.Print(" ");
}
pw.Println(LocalString());
foreach (Edu.Stanford.Nlp.Semgraph.Semgrex.SemgrexPattern child in GetChildren())
{
child.PrettyPrint(pw, indent + 1);
}
}
public virtual void PrintAnswers(IList <CoreLabel> doc, PrintWriter @out)
{
IList <string> sentence = new List <string>();
int wrong = 0;
foreach (CoreLabel wi in doc)
{
StringBuilder sb = new StringBuilder();
if (!wi.Get(typeof(CoreAnnotations.AnswerAnnotation)).Equals(wi.Get(typeof(CoreAnnotations.GoldAnswerAnnotation))))
{
wrong++;
}
if (!ThreeClasses && wi.Get(typeof(CoreAnnotations.AnswerAnnotation)).Equals("UPPER"))
{
sb.Append(wi.Word().ToUpper());
}
else
{
if (wi.Get(typeof(CoreAnnotations.AnswerAnnotation)).Equals("LOWER"))
{
sb.Append(wi.Word().ToLower());
}
else
{
if (wi.Get(typeof(CoreAnnotations.AnswerAnnotation)).Equals("INIT_UPPER"))
{
sb.Append(Sharpen.Runtime.Substring(wi.Word(), 0, 1).ToUpper()).Append(Sharpen.Runtime.Substring(wi.Word(), 1));
}
else
{
if (wi.Get(typeof(CoreAnnotations.AnswerAnnotation)).Equals("O"))
{
// in this case, if it contains a-z at all, then append "MIX" at the end
sb.Append(wi.Word());
Matcher alphaMatcher = alphabet.Matcher(wi.Word());
if (alphaMatcher.Matches())
{
sb.Append("/MIX");
}
}
}
}
}
if (verboseForTrueCasing)
{
sb.Append("/GOLD-").Append(wi.Get(typeof(CoreAnnotations.GoldAnswerAnnotation))).Append("/GUESS-").Append(wi.Get(typeof(CoreAnnotations.AnswerAnnotation)));
}
sentence.Add(sb.ToString());
}
@out.Print(StringUtils.Join(sentence, " "));
System.Console.Error.Printf("> wrong = %d ; total = %d%n", wrong, doc.Count);
@out.Println();
}
private static void OutputHighlighting(PrintWriter @out, CRFClassifier <ICoreMap> classifier, string input)
{
ICollection <string> labels = classifier.Labels();
string background = classifier.BackgroundSymbol();
IList <IList <ICoreMap> > sentences = classifier.Classify(input);
IDictionary <string, Color> tagToColorMap = NERGUI.MakeTagToColorMap(labels, background);
StringBuilder result = new StringBuilder();
int lastEndOffset = 0;
foreach (IList <ICoreMap> sentence in sentences)
{
foreach (ICoreMap word in sentence)
{
int beginOffset = word.Get(typeof(CoreAnnotations.CharacterOffsetBeginAnnotation));
int endOffset = word.Get(typeof(CoreAnnotations.CharacterOffsetEndAnnotation));
string answer = word.Get(typeof(CoreAnnotations.AnswerAnnotation));
if (beginOffset > lastEndOffset)
{
result.Append(StringEscapeUtils.EscapeHtml4(Sharpen.Runtime.Substring(input, lastEndOffset, beginOffset)));
}
// Add a color bar for any tagged words
if (!background.Equals(answer))
{
Color color = tagToColorMap[answer];
result.Append("<span style=\"color:#ffffff;background:" + NERGUI.ColorToHTML(color) + "\">");
}
result.Append(StringEscapeUtils.EscapeHtml4(Sharpen.Runtime.Substring(input, beginOffset, endOffset)));
// Turn off the color bar
if (!background.Equals(answer))
{
result.Append("</span>");
}
lastEndOffset = endOffset;
}
}
if (lastEndOffset < input.Length)
{
result.Append(StringEscapeUtils.EscapeHtml4(Sharpen.Runtime.Substring(input, lastEndOffset)));
}
result.Append("<br><br>");
result.Append("Potential tags:");
foreach (KeyValuePair <string, Color> stringColorEntry in tagToColorMap)
{
result.Append("<br> ");
Color color = stringColorEntry.Value;
result.Append("<span style=\"color:#ffffff;background:" + NERGUI.ColorToHTML(color) + "\">");
result.Append(StringEscapeUtils.EscapeHtml4(stringColorEntry.Key));
result.Append("</span>");
}
@out.Print(result);
}
/// <summary>Writes out a lot of redundant data from this Object to the Writer w.</summary>
/// <param name="w">Data is written to this Writer</param>
public virtual void WriteAllData(TextWriter w)
{
int numStates = index.Size();
PrintWriter @out = new PrintWriter(w);
// all lines have one rule per line
@out.Println("Unary ruleIterator");
for (IEnumerator <UnaryRule> rI = RuleIterator(); rI.MoveNext();)
{
@out.Println(rI.Current.ToString(index));
}
@out.Println("Unary closedRuleIterator");
for (IEnumerator <UnaryRule> rI_1 = ClosedRuleIterator(); rI_1.MoveNext();)
{
@out.Println(rI_1.Current.ToString(index));
}
@out.Println("Unary rulesWithParentIterator");
for (int i = 0; i < numStates; i++)
{
@out.Println(index.Get(i));
for (IEnumerator <UnaryRule> rI_2 = RuleIteratorByParent(i); rI_2.MoveNext();)
{
@out.Print(" ");
@out.Println(rI_2.Current.ToString(index));
}
}
@out.Println("Unary closedRulesWithParentIterator");
for (int i_1 = 0; i_1 < numStates; i_1++)
{
@out.Println(index.Get(i_1));
for (IEnumerator <UnaryRule> rI_2 = ClosedRuleIteratorByParent(i_1); rI_2.MoveNext();)
{
@out.Print(" ");
@out.Println(rI_2.Current.ToString(index));
}
}
@out.Flush();
}
public virtual string ToSummaryString()
{
StringWriter sw = new StringWriter();
PrintWriter pw = new PrintWriter(sw);
pw.Println("Number of data points: " + Size());
pw.Print("Number of labels: " + labelIndex.Size() + " [");
IEnumerator <L> iter = labelIndex.GetEnumerator();
while (iter.MoveNext())
{
pw.Print(iter.Current);
if (iter.MoveNext())
{
pw.Print(", ");
}
}
pw.Println("]");
pw.Println("Number of features (Phi(X) types): " + featureIndex.Size());
pw.Println("Number of active feature types: " + NumFeatureTypes());
pw.Println("Number of active feature tokens: " + NumFeatureTokens());
return(sw.ToString());
}
/// <exception cref="System.IO.IOException"/>
private static void OutputByWriter(IConsumer <StringWriter> printer, PrintWriter @out)
{
StringWriter output = new StringWriter();
printer.Accept(output);
output.Flush();
string escapedXml = StringEscapeUtils.EscapeHtml4(output.ToString());
string[] lines = escapedXml.Split("\n");
@out.Print("<div><pre>");
foreach (string line in lines)
{
int numSpaces = 0;
while (numSpaces < line.Length && line[numSpaces] == ' ')
{
@out.Print(" ");
++numSpaces;
}
@out.Print(Sharpen.Runtime.Substring(line, numSpaces));
@out.Print("\n");
}
@out.Print("</pre></div>");
}
public virtual void WriteSVMLightFormat(PrintWriter writer)
{
foreach (RVFDatum <L, F> datum in this)
{
writer.Print(this.labelIndex.IndexOf(datum.Label()));
ICounter <F> features = datum.AsFeaturesCounter();
foreach (F feature in features.KeySet())
{
double count = features.GetCount(feature);
writer.Format(Locale.English, " %s:%f", this.featureIndex.IndexOf(feature), count);
}
writer.Println();
}
}
/// <summary>
/// Modification of printFullFeatureMatrix to correct bugs and print values
/// (Rajat).
/// </summary>
/// <remarks>
/// Modification of printFullFeatureMatrix to correct bugs and print values
/// (Rajat). Prints the full feature matrix in tab-delimited form. These can be
/// BIG matrices, so be careful!
/// </remarks>
public virtual void PrintFullFeatureMatrixWithValues(PrintWriter pw)
{
string sep = "\t";
for (int i = 0; i < featureIndex.Size(); i++)
{
pw.Print(sep + featureIndex.Get(i));
}
pw.Println();
for (int i_1 = 0; i_1 < size; i_1++)
{
// changed labels.length to size
pw.Print(labelIndex.Get(labels[i_1]));
// changed i to labels[i]
IDictionary <int, double> feats = Generics.NewHashMap();
for (int j = 0; j < data[i_1].Length; j++)
{
int feature = data[i_1][j];
double val = values[i_1][j];
feats[int.Parse(feature)] = val;
}
for (int j_1 = 0; j_1 < featureIndex.Size(); j_1++)
{
if (feats.Contains(int.Parse(j_1)))
{
pw.Print(sep + feats[int.Parse(j_1)]);
}
else
{
pw.Print(sep);
pw.Print(' ');
}
}
pw.Println();
}
pw.Flush();
}
private static void PrintAnswersXML <In>(IList <In> doc, PrintWriter @out)
where In : ICoreMap
{
int num = 0;
foreach (IN wi in doc)
{
string prev = StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.BeforeAnnotation)));
@out.Print(prev);
@out.Print("<wi num=\"");
// tag.append(wi.get("position"));
@out.Print(num++);
@out.Print("\" entity=\"");
@out.Print(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.AnswerAnnotation))));
@out.Print("\">");
@out.Print(XMLUtils.EscapeXML(StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.TextAnnotation)))));
@out.Print("</wi>");
string after = StringUtils.GetNotNullString(wi.Get(typeof(CoreAnnotations.AfterAnnotation)));
@out.Print(after);
}
}
public static bool WriteFile(TransducerGraph graph, string dir, string name)
{
try
{
File baseDir = new File(dir);
if (baseDir.Exists())
{
if (!baseDir.IsDirectory())
{
return(false);
}
}
else
{
if (!baseDir.Mkdirs())
{
return(false);
}
}
File file = new File(baseDir, name + ".dot");
PrintWriter w;
try
{
w = new PrintWriter(new FileWriter(file));
string dotString = graph.AsDOTString();
w.Print(dotString);
w.Flush();
w.Close();
}
catch (FileNotFoundException)
{
log.Info("Failed to open file in writeToDOTfile: " + file);
return(false);
}
catch (IOException)
{
log.Info("Failed to open file in writeToDOTfile: " + file);
return(false);
}
return(true);
}
catch (Exception e)
{
Sharpen.Runtime.PrintStackTrace(e);
return(false);
}
}
