// static demo class
/// <exception cref="System.IO.IOException"/>
public static void Main(string[] args)
{
if (args.Length < 2)
{
System.Console.Error.WriteLine("TokensRegexMatcher rules file [outFile]");
return;
}
string rules = args[0];
PrintWriter @out;
if (args.Length > 2)
{
@out = new PrintWriter(args[2]);
}
else
{
@out = new PrintWriter(System.Console.Out);
}
StanfordCoreNLP pipeline = new StanfordCoreNLP(PropertiesUtils.AsProperties("annotators", "tokenize,ssplit,pos,lemma,ner"));
Annotation annotation = new Annotation(IOUtils.SlurpFileNoExceptions(args[1]));
pipeline.Annotate(annotation);
// Load lines of file as TokenSequencePatterns
IList <TokenSequencePattern> tokenSequencePatterns = new List <TokenSequencePattern>();
foreach (string line in ObjectBank.GetLineIterator(rules))
{
TokenSequencePattern pattern = TokenSequencePattern.Compile(line);
tokenSequencePatterns.Add(pattern);
}
IList <ICoreMap> sentences = annotation.Get(typeof(CoreAnnotations.SentencesAnnotation));
int i = 0;
foreach (ICoreMap sentence in sentences)
{
IList <CoreLabel> tokens = sentence.Get(typeof(CoreAnnotations.TokensAnnotation));
@out.Println("Sentence #" + ++i);
@out.Print(" Tokens:");
foreach (CoreLabel token in tokens)
{
@out.Print(' ');
@out.Print(token.ToShortString("Text", "PartOfSpeech", "NamedEntityTag"));
}
@out.Println();
MultiPatternMatcher <ICoreMap> multiMatcher = TokenSequencePattern.GetMultiPatternMatcher(tokenSequencePatterns);
IList <ISequenceMatchResult <ICoreMap> > answers = multiMatcher.FindNonOverlapping(tokens);
int j = 0;
foreach (ISequenceMatchResult <ICoreMap> matched in answers)
{
@out.Println(" Match #" + ++j);
for (int k = 0; k <= matched.GroupCount(); k++)
{
@out.Println(" group " + k + " = " + matched.Group(k));
}
}
}
@out.Flush();
}
// static main only
/// <exception cref="System.IO.IOException"/>
public static void Main(string[] args)
{
StanfordCoreNLP pipeline = new StanfordCoreNLP(PropertiesUtils.AsProperties("annotators", "tokenize,ssplit,pos,lemma,ner"));
Annotation annotation = new Annotation("Casey is 21. Sally Atkinson's age is 30.");
pipeline.Annotate(annotation);
IList <ICoreMap> sentences = annotation.Get(typeof(CoreAnnotations.SentencesAnnotation));
IList <TokenSequencePattern> tokenSequencePatterns = new List <TokenSequencePattern>();
string[] patterns = new string[] { "(?$who [ ner: PERSON]+ ) /is/ (?$age [ pos: CD ] )", "(?$who [ ner: PERSON]+ ) /'s/ /age/ /is/ (?$age [ pos: CD ] )" };
foreach (string line in patterns)
{
TokenSequencePattern pattern = TokenSequencePattern.Compile(line);
tokenSequencePatterns.Add(pattern);
}
MultiPatternMatcher <ICoreMap> multiMatcher = TokenSequencePattern.GetMultiPatternMatcher(tokenSequencePatterns);
int i = 0;
foreach (ICoreMap sentence in sentences)
{
IList <CoreLabel> tokens = sentence.Get(typeof(CoreAnnotations.TokensAnnotation));
System.Console.Out.WriteLine("Sentence #" + ++i);
System.Console.Out.Write(" Tokens:");
foreach (CoreLabel token in tokens)
{
System.Console.Out.Write(' ');
System.Console.Out.Write(token.ToShortString("Text", "PartOfSpeech", "NamedEntityTag"));
}
System.Console.Out.WriteLine();
IList <ISequenceMatchResult <ICoreMap> > answers = multiMatcher.FindNonOverlapping(tokens);
int j = 0;
foreach (ISequenceMatchResult <ICoreMap> matched in answers)
{
System.Console.Out.WriteLine(" Match #" + ++j);
System.Console.Out.WriteLine(" match: " + matched.Group(0));
System.Console.Out.WriteLine(" who: " + matched.Group("$who"));
System.Console.Out.WriteLine(" age: " + matched.Group("$age"));
}
}
}
private void RunParallelApplyPats(IDictionary <string, DataInstance> sents, string label, E pattern, TwoDimensionalCounter <CandidatePhrase, E> wordsandLemmaPatExtracted, CollectionValuedMap <E, Triple <string, int, int> > matchedTokensByPat, ICollection
<CandidatePhrase> alreadyLabeledWords)
{
Redwood.Log(Redwood.Dbg, "Applying pattern " + pattern + " to a total of " + sents.Count + " sentences ");
IList <string> notAllowedClasses = new List <string>();
IList <string> sentids = CollectionUtils.ToList(sents.Keys);
if (constVars.doNotExtractPhraseAnyWordLabeledOtherClass)
{
foreach (string l in constVars.GetAnswerClass().Keys)
{
if (!l.Equals(label))
{
notAllowedClasses.Add(l);
}
}
notAllowedClasses.Add("OTHERSEM");
}
IDictionary <TokenSequencePattern, E> surfacePatternsLearnedThisIterConverted = null;
IDictionary <SemgrexPattern, E> depPatternsLearnedThisIterConverted = null;
if (constVars.patternType.Equals(PatternFactory.PatternType.Surface))
{
surfacePatternsLearnedThisIterConverted = new Dictionary <TokenSequencePattern, E>();
string patternStr = null;
try
{
patternStr = pattern.ToString(notAllowedClasses);
TokenSequencePattern pat = ((TokenSequencePattern)TokenSequencePattern.Compile(constVars.env[label], patternStr));
surfacePatternsLearnedThisIterConverted[pat] = pattern;
}
catch (Exception e)
{
log.Info("Error applying pattern " + patternStr + ". Probably an ill formed pattern (can be because of special symbols in label names). Contact the software developer.");
throw;
}
}
else
{
if (constVars.patternType.Equals(PatternFactory.PatternType.Dep))
{
depPatternsLearnedThisIterConverted = new Dictionary <SemgrexPattern, E>();
SemgrexPattern pat = SemgrexPattern.Compile(pattern.ToString(notAllowedClasses), new Env(constVars.env[label].GetVariables()));
depPatternsLearnedThisIterConverted[pat] = pattern;
}
else
{
throw new NotSupportedException();
}
}
//Apply the patterns and extract candidate phrases
int num;
int numThreads = constVars.numThreads;
//If number of sentences is less, do not create so many threads
if (sents.Count < 50)
{
numThreads = 1;
}
if (numThreads == 1)
{
num = sents.Count;
}
else
{
num = sents.Count / (numThreads - 1);
}
IExecutorService executor = Executors.NewFixedThreadPool(constVars.numThreads);
IList <IFuture <Triple <TwoDimensionalCounter <CandidatePhrase, E>, CollectionValuedMap <E, Triple <string, int, int> >, ICollection <CandidatePhrase> > > > list = new List <IFuture <Triple <TwoDimensionalCounter <CandidatePhrase, E>, CollectionValuedMap <E
, Triple <string, int, int> >, ICollection <CandidatePhrase> > > >();
for (int i = 0; i < numThreads; i++)
{
ICallable <Triple <TwoDimensionalCounter <CandidatePhrase, E>, CollectionValuedMap <E, Triple <string, int, int> >, ICollection <CandidatePhrase> > > task = null;
if (pattern.type.Equals(PatternFactory.PatternType.Surface))
{
//Redwood.log(Redwood.DBG, "Applying pats: assigning sentences " + i*num + " to " +Math.min(sentids.size(), (i + 1) * num) + " to thread " + (i+1));
task = new ApplyPatterns(sents, num == sents.Count ? sentids : sentids.SubList(i * num, Math.Min(sentids.Count, (i + 1) * num)), surfacePatternsLearnedThisIterConverted, label, constVars.removeStopWordsFromSelectedPhrases, constVars.removePhrasesWithStopWords
, constVars);
}
else
{
task = new ApplyDepPatterns(sents, num == sents.Count ? sentids : sentids.SubList(i * num, Math.Min(sentids.Count, (i + 1) * num)), depPatternsLearnedThisIterConverted, label, constVars.removeStopWordsFromSelectedPhrases, constVars.removePhrasesWithStopWords
, constVars);
}
IFuture <Triple <TwoDimensionalCounter <CandidatePhrase, E>, CollectionValuedMap <E, Triple <string, int, int> >, ICollection <CandidatePhrase> > > submit = executor.Submit(task);
list.Add(submit);
}
// Now retrieve the result
foreach (IFuture <Triple <TwoDimensionalCounter <CandidatePhrase, E>, CollectionValuedMap <E, Triple <string, int, int> >, ICollection <CandidatePhrase> > > future in list)
{
try
{
Triple <TwoDimensionalCounter <CandidatePhrase, E>, CollectionValuedMap <E, Triple <string, int, int> >, ICollection <CandidatePhrase> > result = future.Get();
Redwood.Log(ConstantsAndVariables.extremedebug, "Pattern " + pattern + " extracted phrases " + result.First());
wordsandLemmaPatExtracted.AddAll(result.First());
matchedTokensByPat.AddAll(result.Second());
Sharpen.Collections.AddAll(alreadyLabeledWords, result.Third());
}
catch (Exception e)
{
executor.ShutdownNow();
throw new Exception(e);
}
}
executor.Shutdown();
}
public WordsToSentencesAnnotator(Properties properties)
{
bool nlSplitting = bool.ValueOf(properties.GetProperty(StanfordCoreNLP.NewlineSplitterProperty, "false"));
if (nlSplitting)
{
bool whitespaceTokenization = bool.ValueOf(properties.GetProperty("tokenize.whitespace", "false"));
if (whitespaceTokenization)
{
if (Runtime.LineSeparator().Equals("\n"))
{
// this constructor will keep empty lines as empty sentences
WordToSentenceProcessor <CoreLabel> wts1 = new WordToSentenceProcessor <CoreLabel>(ArrayUtils.AsImmutableSet(new string[] { "\n" }));
this.countLineNumbers = true;
this.wts = wts1;
}
else
{
// throw "\n" in just in case files use that instead of
// the system separator
// this constructor will keep empty lines as empty sentences
WordToSentenceProcessor <CoreLabel> wts1 = new WordToSentenceProcessor <CoreLabel>(ArrayUtils.AsImmutableSet(new string[] { Runtime.LineSeparator(), "\n" }));
this.countLineNumbers = true;
this.wts = wts1;
}
}
else
{
// this constructor will keep empty lines as empty sentences
WordToSentenceProcessor <CoreLabel> wts1 = new WordToSentenceProcessor <CoreLabel>(ArrayUtils.AsImmutableSet(new string[] { PTBTokenizer.GetNewlineToken() }));
this.countLineNumbers = true;
this.wts = wts1;
}
}
else
{
string isOneSentence = properties.GetProperty("ssplit.isOneSentence");
if (bool.Parse(isOneSentence))
{
// this method treats null as false
// Treat as one sentence: You get a no-op sentence splitter that always returns all tokens as one sentence.
WordToSentenceProcessor <CoreLabel> wts1 = new WordToSentenceProcessor <CoreLabel>(true);
this.countLineNumbers = false;
this.wts = wts1;
}
else
{
// multi token sentence boundaries
string boundaryMultiTokenRegex = properties.GetProperty("ssplit.boundaryMultiTokenRegex");
// Discard these tokens without marking them as sentence boundaries
string tokenPatternsToDiscardProp = properties.GetProperty("ssplit.tokenPatternsToDiscard");
ICollection <string> tokenRegexesToDiscard = null;
if (tokenPatternsToDiscardProp != null)
{
string[] toks = tokenPatternsToDiscardProp.Split(",");
tokenRegexesToDiscard = Generics.NewHashSet(Arrays.AsList(toks));
}
// regular boundaries
string boundaryTokenRegex = properties.GetProperty("ssplit.boundaryTokenRegex");
string boundaryFollowersRegex = properties.GetProperty("ssplit.boundaryFollowersRegex");
// newline boundaries which are discarded.
ICollection <string> boundariesToDiscard = null;
string bounds = properties.GetProperty("ssplit.boundariesToDiscard");
if (bounds != null)
{
string[] toks = bounds.Split(",");
boundariesToDiscard = Generics.NewHashSet(Arrays.AsList(toks));
}
ICollection <string> htmlElementsToDiscard = null;
// HTML boundaries which are discarded
bounds = properties.GetProperty("ssplit.htmlBoundariesToDiscard");
if (bounds != null)
{
string[] elements = bounds.Split(",");
htmlElementsToDiscard = Generics.NewHashSet(Arrays.AsList(elements));
}
string nlsb = properties.GetProperty(StanfordCoreNLP.NewlineIsSentenceBreakProperty, StanfordCoreNLP.DefaultNewlineIsSentenceBreak);
this.countLineNumbers = false;
this.wts = new WordToSentenceProcessor <CoreLabel>(boundaryTokenRegex, boundaryFollowersRegex, boundariesToDiscard, htmlElementsToDiscard, WordToSentenceProcessor.StringToNewlineIsSentenceBreak(nlsb), (boundaryMultiTokenRegex != null) ? TokenSequencePattern
.Compile(boundaryMultiTokenRegex) : null, tokenRegexesToDiscard);
}
}
Verbose = bool.ValueOf(properties.GetProperty("ssplit.verbose", "false"));
}
public WordsToSentencesAnnotator(bool verbose, string boundaryTokenRegex, ICollection <string> boundaryToDiscard, ICollection <string> htmlElementsToDiscard, string newlineIsSentenceBreak, string boundaryMultiTokenRegex, ICollection <string> tokenRegexesToDiscard
)
: this(verbose, false, new WordToSentenceProcessor <CoreLabel>(boundaryTokenRegex, null, boundaryToDiscard, htmlElementsToDiscard, WordToSentenceProcessor.StringToNewlineIsSentenceBreak(newlineIsSentenceBreak), (boundaryMultiTokenRegex != null
) ? TokenSequencePattern.Compile(boundaryMultiTokenRegex) : null, tokenRegexesToDiscard))
{
}