public CPosTag()
{
tagger = new StanfordPosTagger(@"E:\v-wacui\QuestionAnswering\data\ResourceData\POSTAG\left3words-distsim-wsj-0-18.tagger");
tagger.InitTagger();
}
public static void AddPosTagCore(string input, string output, int iColSen, int iColApp)
{
//StanfordPosTagger tagger = new StanfordPosTagger(@"D:\hongsli\SVN\CodeBase\trunk\NifClickParser\Ref\C#\models\bidirectional-distsim-wsj-0-18.tagger");
//StanfordPosTagger tagger = new StanfordPosTagger(GlobalParamters.con_strStandfordPostagFile);//@"D:\hongsli\SVN\CodeBase\trunk\NifClickParser\Ref\C#\models\left3words-distsim-wsj-0-18.tagger");
StanfordPosTagger tagger = new StanfordPosTagger(@"E:\v-wacui\QuestionAnswering\data\ResourceData\POSTAG\left3words-distsim-wsj-0-18.tagger");
DateTime t1 = DateTime.Now;
int iCount = 0;
Console.WriteLine("start time: " +t1.ToString());
using (TextReader tr = new StreamReader(input))
{
using (TextWriter tw = new StreamWriter(output))
{
string text="";
string outtk="";
string outtg="";
while ((text = tr.ReadLine()) != null)
{
string[] cols = text.Split('\t');
if (cols.Length > iColSen)
{
tagger.AddTag(cols[iColSen], out outtk, out outtg);
for (int i = 0; i < iColApp; i++)
{
tw.Write(cols[i] + "\t");
}
//tw.Write(text + "\t");
tw.Write(outtk.Replace("\t", "#TAB#") + "\t");
tw.WriteLine(outtg.Replace("\t", "#TAB#") + "\t");
iCount++;
if (iCount % 1000 == 0)
{
Console.Write(string.Format("\rcompleted tagging {1} sentences in {0} ", (DateTime.Now - t1).ToString(), iCount));
}
}
}
}
}
Console.WriteLine(string.Format("completed tagging {1} sentences in {0} ", (DateTime.Now-t1).ToString(), iCount));
// StanfordCoreNLP pipeline = new StanfordCoreNLP();
// java.util.Properties props = new java.util.Properties();
// props.put("annotators", "tokenize, ssplit, pos");
// string text = "";
// using (TextReader tr = new StreamReader(input))
// {
// using (TextWriter tw = new StreamWriter(output))
// {
// while ((text = tr.ReadLine()) != null)
// {
// Annotation document = new Annotation(text);
// // run all Annotators on this text
// pipeline.annotate(document);
// // these are all the sentences in this document
// // a CoreMap is essentially a Map that uses class objects as keys and has values with custom types
// List<CoreMap> sentences = (List<CoreMap>)document.get((java.lang.Class)typeof(edu.stanford.nlp.ling.CoreAnnotations.SentencesAnnotation));
// StringBuilder sb = new StringBuilder();
// foreach (CoreMap sentence in sentences)
// {
// List<edu.stanford.nlp.ling.CoreLabel> tokens = (List<edu.stanford.nlp.ling.CoreLabel>)sentence.get((java.lang.Class)typeof(edu.stanford.nlp.ling.CoreAnnotations.TokensAnnotation));
// foreach (edu.stanford.nlp.ling.CoreLabel token in tokens)
// {
// String word = (String)token.get((java.lang.Class)typeof(edu.stanford.nlp.ling.CoreAnnotations.TextAnnotation));
// String pos = (String)token.get((java.lang.Class)typeof(edu.stanford.nlp.ling.CoreAnnotations.PartOfSpeechAnnotation));
// //String ne = (String) token.get((java.lang.Class) typeof(edu.stanford.nlp.ling.CoreAnnotations.NamedEntityTagAnnotation));
// sb.Append(word+"/"+pos+" ");
// }
// tw.WriteLine(sb.ToString().TrimEnd());
// //Tree tree = sentence.get(TreeAnnotation.class);
// //SemanticGraph dependencies = sentence.get(CollapsedCCProcessedDependenciesAnnotation.class);
// }
// // this is the coreference link graph
// // each link stores an arc in the graph; the first element in the Pair is the source, the second is the target
// // each node is stored as <sentence id, token id>. Both offsets start at 1!
// //List<Pair<IntTuple, IntTuple>> graph = document.get(CorefGraphAnnotation.class);
// }
// tw.Close();
// }
// tr.Close();
// }
//
}
