public static void main(String[] args) throws IOException, ClassNotFoundException {
Timing tim = new Timing();
AnnotationPipeline ap = new AnnotationPipeline();
boolean verbose = false;
ap.addAnnotator(new TokenizerAnnotator(verbose, "en"));
ap.addAnnotator(new WordsToSentencesAnnotator(verbose));
// ap.addAnnotator(new NERCombinerAnnotator(verbose));
// ap.addAnnotator(new OldNERAnnotator(verbose));
// ap.addAnnotator(new NERMergingAnnotator(verbose));
ap.addAnnotator(new ParserAnnotator(verbose, -1));
/**
* ap.addAnnotator(new UpdateSentenceFromParseAnnotator(verbose)); ap.addAnnotator(new
* NumberAnnotator(verbose)); ap.addAnnotator(new
* QuantifiableEntityNormalizingAnnotator(verbose)); ap.addAnnotator(new
* StemmerAnnotator(verbose)); ap.addAnnotator(new MorphaAnnotator(verbose));
*/
// ap.addAnnotator(new SRLAnnotator());
String text =
("USAir said in the filings that Mr. Icahn first contacted Mr. Colodny last September to discuss the benefits of combining TWA and USAir -- either by TWA's acquisition of USAir, or USAir's acquisition of TWA.");
Annotation a = new Annotation(text);
ap.annotate(a);
System.out.println(a.get(CoreAnnotations.TokensAnnotation.class));
for (CoreMap sentence : a.get(CoreAnnotations.SentencesAnnotation.class)) {
System.out.println(sentence.get(TreeCoreAnnotations.TreeAnnotation.class));
}
if (TIME) {
System.out.println(ap.timingInformation());
System.err.println("Total time for AnnotationPipeline: " + tim.toSecondsString() + " sec.");
}
}
public static final String doCorefResolution(Annotation annotation) {
Map<Integer, CorefChain> corefs = annotation.get(CorefChainAnnotation.class);
List<CoreMap> sentences = annotation.get(CoreAnnotations.SentencesAnnotation.class);
List<String> resolved = new ArrayList<String>();
for (CoreMap sentence : sentences) {
List<CoreLabel> tokens = sentence.get(CoreAnnotations.TokensAnnotation.class);
for (CoreLabel token : tokens) {
Integer corefClustId = token.get(CorefCoreAnnotations.CorefClusterIdAnnotation.class);
CorefChain chain = corefs.get(corefClustId);
if (chain == null) resolved.add(token.word());
else {
int sentINdx = chain.getRepresentativeMention().sentNum - 1;
CoreMap corefSentence = sentences.get(sentINdx);
List<CoreLabel> corefSentenceTokens = corefSentence.get(TokensAnnotation.class);
CorefMention reprMent = chain.getRepresentativeMention();
if (token.index() < reprMent.startIndex || token.index() > reprMent.endIndex) {
for (int i = reprMent.startIndex; i < reprMent.endIndex; i++) {
CoreLabel matchedLabel = corefSentenceTokens.get(i - 1);
resolved.add(matchedLabel.word());
}
} else resolved.add(token.word());
}
}
}
String resolvedStr = "";
System.out.println();
for (String str : resolved) {
resolvedStr += str + " ";
}
System.out.println(resolvedStr);
return resolvedStr;
}
@Override
public void annotate(Annotation annotation) {
if (verbose) {
System.err.print("Adding TokensRegexNER annotations ... ");
}
List<CoreMap> sentences = annotation.get(CoreAnnotations.SentencesAnnotation.class);
if (sentences != null) {
for (CoreMap sentence : sentences) {
List<CoreLabel> tokens = sentence.get(CoreAnnotations.TokensAnnotation.class);
annotateMatched(tokens);
}
} else {
List<CoreLabel> tokens = annotation.get(CoreAnnotations.TokensAnnotation.class);
if (tokens != null) {
annotateMatched(tokens);
} else {
throw new RuntimeException("Unable to find sentences or tokens in " + annotation);
}
}
if (verbose) System.err.println("done.");
}
@Override
public void handle(HttpExchange httpExchange) throws IOException {
// Set common response headers
httpExchange.getResponseHeaders().add("Access-Control-Allow-Origin", "*");
Future<String> json =
corenlpExecutor.submit(
() -> {
try {
// Get the document
Properties props =
new Properties() {
{
setProperty("annotators", "tokenize,ssplit,pos,lemma,ner,depparse");
}
};
Annotation doc = getDocument(props, httpExchange);
if (!doc.containsKey(CoreAnnotations.SentencesAnnotation.class)) {
StanfordCoreNLP pipeline = mkStanfordCoreNLP(props);
pipeline.annotate(doc);
}
// Construct the matcher
Map<String, String> params = getURLParams(httpExchange.getRequestURI());
// (get the pattern)
if (!params.containsKey("pattern")) {
respondError("Missing required parameter 'pattern'", httpExchange);
return "";
}
String pattern = params.get("pattern");
// (get whether to filter / find)
String filterStr = params.getOrDefault("filter", "false");
final boolean filter =
filterStr.trim().isEmpty()
|| "true".equalsIgnoreCase(filterStr.toLowerCase());
// (create the matcher)
final SemgrexPattern regex = SemgrexPattern.compile(pattern);
// Run TokensRegex
return JSONOutputter.JSONWriter.objectToJSON(
(docWriter) -> {
if (filter) {
// Case: just filter sentences
docWriter.set(
"sentences",
doc.get(CoreAnnotations.SentencesAnnotation.class)
.stream()
.map(
sentence ->
regex
.matcher(
sentence.get(
SemanticGraphCoreAnnotations
.CollapsedCCProcessedDependenciesAnnotation
.class))
.matches())
.collect(Collectors.toList()));
} else {
// Case: find matches
docWriter.set(
"sentences",
doc.get(CoreAnnotations.SentencesAnnotation.class)
.stream()
.map(
sentence ->
(Consumer<JSONOutputter.Writer>)
(JSONOutputter.Writer sentWriter) -> {
SemgrexMatcher matcher =
regex.matcher(
sentence.get(
SemanticGraphCoreAnnotations
.CollapsedCCProcessedDependenciesAnnotation
.class));
int i = 0;
while (matcher.find()) {
sentWriter.set(
Integer.toString(i),
(Consumer<JSONOutputter.Writer>)
(JSONOutputter.Writer matchWriter) -> {
IndexedWord match = matcher.getMatch();
matchWriter.set("text", match.word());
matchWriter.set(
"begin", match.index() - 1);
matchWriter.set("end", match.index());
for (String capture :
matcher.getNodeNames()) {
matchWriter.set(
"$" + capture,
(Consumer<JSONOutputter.Writer>)
groupWriter -> {
IndexedWord node =
matcher.getNode(
capture);
groupWriter.set(
"text", node.word());
groupWriter.set(
"begin",
node.index() - 1);
groupWriter.set(
"end", node.index());
});
}
});
i += 1;
}
sentWriter.set("length", i);
}));
}
});
} catch (Exception e) {
e.printStackTrace();
try {
respondError(e.getClass().getName() + ": " + e.getMessage(), httpExchange);
} catch (IOException ignored) {
}
}
return "";
});
// Send response
byte[] response = new byte[0];
try {
response = json.get(5, TimeUnit.SECONDS).getBytes();
} catch (InterruptedException | ExecutionException | TimeoutException e) {
respondError("Timeout when executing Semgrex query", httpExchange);
}
if (response.length > 0) {
httpExchange.getResponseHeaders().add("Content-Type", "text/json");
httpExchange.getResponseHeaders().add("Content-Length", Integer.toString(response.length));
httpExchange.sendResponseHeaders(HTTP_OK, response.length);
httpExchange.getResponseBody().write(response);
httpExchange.close();
}
}
@Override
public void handle(HttpExchange httpExchange) throws IOException {
// Set common response headers
httpExchange.getResponseHeaders().add("Access-Control-Allow-Origin", "*");
// Get sentence.
Properties props;
Annotation ann;
StanfordCoreNLP.OutputFormat of;
log("[" + httpExchange.getRemoteAddress() + "] Received message");
try {
props = getProperties(httpExchange);
ann = getDocument(props, httpExchange);
of =
StanfordCoreNLP.OutputFormat.valueOf(
props.getProperty("outputFormat", "json").toUpperCase());
// Handle direct browser connections (i.e., not a POST request).
if (ann.get(CoreAnnotations.TextAnnotation.class).length() == 0) {
log("[" + httpExchange.getRemoteAddress() + "] Interactive connection");
staticPageHandle.handle(httpExchange);
return;
}
log("[" + httpExchange.getRemoteAddress() + "] API call");
} catch (Exception e) {
// Return error message.
e.printStackTrace();
String response = e.getMessage();
httpExchange.getResponseHeaders().add("Content-Type", "text/plain");
httpExchange.sendResponseHeaders(HTTP_BAD_INPUT, response.length());
httpExchange.getResponseBody().write(response.getBytes());
httpExchange.close();
return;
}
try {
// Annotate
StanfordCoreNLP pipeline = mkStanfordCoreNLP(props);
Future<Annotation> completedAnnotationFuture =
corenlpExecutor.submit(
() -> {
pipeline.annotate(ann);
return ann;
});
Annotation completedAnnotation = completedAnnotationFuture.get(5, TimeUnit.SECONDS);
// Get output
ByteArrayOutputStream os = new ByteArrayOutputStream();
StanfordCoreNLP.createOutputter(props, AnnotationOutputter.getOptions(pipeline))
.accept(completedAnnotation, os);
os.close();
byte[] response = os.toByteArray();
httpExchange.getResponseHeaders().add("Content-Type", getContentType(props, of));
httpExchange.getResponseHeaders().add("Content-Length", Integer.toString(response.length));
httpExchange.sendResponseHeaders(HTTP_OK, response.length);
httpExchange.getResponseBody().write(response);
httpExchange.close();
} catch (TimeoutException e) {
respondError("CoreNLP request timed out", httpExchange);
} catch (Exception e) {
// Return error message.
respondError(e.getClass().getName() + ": " + e.getMessage(), httpExchange);
}
}