/**
* @param key
* @param options
*/
private void addSuffixPluralPattern(String key, HashMap<String, String> options) {
String option = options.get(key);
if (option != null) {
patternMap.put(
key, Pattern.compile("\\b(\\w{" + minPrefixLength + ",})(" + option + ")s?\\b"));
}
}
/**
* @param key
* @param options
*/
private void addWordPattern(String key, HashMap<String, String> options) {
String option = options.get(key);
if (option != null) {
patternMap.put(key, Pattern.compile("\\b(" + option + ")\\b"));
}
}
@Override
public void execute() throws ExecutionException {
interrupted = false;
// quit if setup failed
if (gracefulExit) {
gracefulExit("Plugin was not initialised correctly. Exiting gracefully ... ");
return;
}
AnnotationSet inputAS =
(inputASName == null || inputASName.trim().length() == 0)
? document.getAnnotations()
: document.getAnnotations(inputASName);
AnnotationSet outputAS =
(outputASName == null || outputASName.trim().length() == 0)
? document.getAnnotations()
: document.getAnnotations(outputASName);
AnnotationSet sentenceAS = null;
if (sentenceType != null && !sentenceType.isEmpty()) {
sentenceAS = inputAS.get(sentenceType);
}
// Document content
String docContent = document.getContent().toString();
int docLen = docContent.length();
// For matching purposes replace all whitespace characters with a single space
docContent = docContent.replaceAll("[\\s\\xA0\\u2007\\u202F]", " ");
fireStatusChanged("Locating anatomy, disease and procedure mentions in " + document.getName());
fireProgressChanged(0);
if (sentenceAS != null) {
for (Annotation sentence : sentenceAS) {
Long sentStartOffset = sentence.getStartNode().getOffset();
Long sentEndOffset = sentence.getEndNode().getOffset();
// Converting the sentence to lower case prevents the need to use case-insenstive regex
// matching, which should give a small performance boost
String sentenceContent =
docContent
.substring(sentStartOffset.intValue(), sentEndOffset.intValue())
.toLowerCase(Locale.ENGLISH);
if (diseaseType != null && !diseaseType.isEmpty()) {
doMatch(
patternMap.get("disease_suffix"),
sentenceContent,
inputAS,
outputAS,
"suffDisease",
sentStartOffset,
docLen);
doMatch(
patternMap.get("disease_abbrevs"),
sentenceContent,
inputAS,
outputAS,
"preDisease",
sentStartOffset,
docLen);
doMatch(
patternMap.get("disease_named_syndrome"),
sentenceContent,
inputAS,
outputAS,
"namedDisease",
sentStartOffset,
docLen);
doMatch(
patternMap.get("disease_sense"),
sentenceContent,
inputAS,
outputAS,
"tmpDiseaseSense",
sentStartOffset,
docLen);
doMatch(
patternMap.get("disease_sense_context"),
sentenceContent,
inputAS,
outputAS,
"tmpDiseaseSenseContext",
sentStartOffset,
docLen);
doMatch(
patternMap.get("disease_generic_context"),
sentenceContent,
inputAS,
outputAS,
"poDisease",
sentStartOffset,
docLen);
doMatch(
patternMap.get("disease_anatomy_context"),
sentenceContent,
inputAS,
outputAS,
"tmpDisease",
sentStartOffset,
docLen);
}
if (procedureType != null && !procedureType.isEmpty()) {
doMatch(
patternMap.get("procedure_suffix"),
sentenceContent,
inputAS,
outputAS,
"poProcedure",
sentStartOffset,
docLen);
doMatch(
patternMap.get("procedure_key"),
sentenceContent,
inputAS,
outputAS,
"poProcedure",
sentStartOffset,
docLen);
doMatch(
patternMap.get("procedure_anatomy_context"),
sentenceContent,
inputAS,
outputAS,
"tmpProcedure",
sentStartOffset,
docLen);
}
if (symptomType != null && !symptomType.isEmpty()) {
doMatch(
patternMap.get("symptom_key"),
sentenceContent,
inputAS,
outputAS,
"poSymptom",
sentStartOffset,
docLen);
}
if (testType != null && !testType.isEmpty()) {
doMatch(
patternMap.get("test_key"),
sentenceContent,
inputAS,
outputAS,
"poTest",
sentStartOffset,
docLen);
}
if (anatomyType != null && !anatomyType.isEmpty()) {
doMatch(
patternMap.get("anatomy_suffix_adjective"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatSuffAdj",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_suffix"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatSuff",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_prefix"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatPre",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_position"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatPos",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_space_region_junction"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatSpace",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_part_adjective"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatAdj",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_latin_noun"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatLatin",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_muscle"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatMuscle",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_part"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatPart",
sentStartOffset,
docLen);
doMatch(
patternMap.get("anatomy_fluid"),
sentenceContent,
inputAS,
outputAS,
"tmpAnatFluid",
sentStartOffset,
docLen);
}
}
// Run JAPE transducer to clean up the output
fireStatusChanged(
"Processing anatomical, disease and procedure mentions in " + document.getName());
try {
japeTransducer.setDocument(document);
japeTransducer.setInputASName(inputASName);
japeTransducer.setOutputASName(outputASName);
japeTransducer.addProgressListener(this);
japeTransducer.execute();
} catch (ExecutionException re) {
gate.util.Err.println("Unable to run " + japeURL);
gracefulExit = true;
} finally {
japeTransducer.setDocument(null);
}
// rename temporary annotations
if (!debug) {
renameAnnotations(outputAS, "tmpAnatomicalTerm", anatomyType);
renameAnnotations(outputAS, "suffDisease", diseaseType);
renameAnnotations(outputAS, "poDisease", diseaseType);
renameAnnotations(outputAS, "preDisease", diseaseType);
renameAnnotations(outputAS, "poProcedure", procedureType);
renameAnnotations(outputAS, "poSymptom", symptomType);
renameAnnotations(outputAS, "poTest", testType);
}
} else {
gracefulExit("No sentences to process!");
}
// want list of disease key words plus symptoms such as oedema? or just diseases
fireProcessFinished();
} // end execute()