private static void writeLine(Writer out, String line, TextIdentifier docQuerySnrId)
throws IOException {
String[] tokens = line.split(" ");
if (tokens.length > SENTENCE_MAXTOKENS) {
return;
}
out.write("<D=" + docQuerySnrId.toValidString() + ">\n");
for (String token : tokens) {
out.write(token + "\n");
}
out.write("</D>\n");
}
/**
* A stream with one sentence per line is read in, sentences containing a query token (or alias
* thereof) are retained and printed out, formatted.
*
* @param args
* @throws IOException
*/
public static void main(String[] args) throws IOException {
if (args.length < 1 || args.length > 2) {
System.err.println("Format <query_expanded.xml> [extraction_window:int]");
System.err.println("A stream with one sentece per line is read in, sentences containing");
System.err.println("a query token are retained and printed out, formatted.");
System.err.println(
"The optional extraction_window parameter must be an integer between -1"
+ " and inf. It specifies the number of sentences printed out before and "
+ " after a sentence containing a query. E.g., extraction_window=1 will"
+ " print sentences immediately preceding and following query sentences."
+ " If set to -1, all sentences will be printed. Default is 0.");
return;
}
BufferedWriter out = new BufferedWriter(new OutputStreamWriter(System.out, "UTF-8"));
BufferedReader br = new BufferedReader(new InputStreamReader(System.in, "UTF-8"));
QueryList ql = new QueryList(args[0]);
int extractionWin = 0;
if (args.length > 1) {
try {
extractionWin = Integer.parseInt(args[1]);
if (extractionWin < -1) {
throw new NumberFormatException();
}
} catch (NumberFormatException e) {
System.err.println("Extraction window must be an integer between -1" + " and inf.");
return;
}
}
// look behind
CircularFifoBuffer lastSentences = null;
if (extractionWin > 0) {
lastSentences = new CircularFifoBuffer(extractionWin);
}
// look forward
int currentExtractionWin = 0;
int sNr = 1;
TextIdentifier docQueryId = null;
for (String line; (line = br.readLine()) != null; ) {
if (line.startsWith("#")) {
docQueryId = TextIdentifier.fromDelimited(line.substring(1));
if (docQueryId.getQueryId() == null || docQueryId.getQueryId().isEmpty()) {
throw new IllegalArgumentException("TextIdentifier without query id:" + line);
}
sNr = 1;
if (lastSentences != null) {
lastSentences.clear();
}
currentExtractionWin = 0;
} else {
assert (docQueryId != null);
TextIdentifier docQuerySnrId = new TextIdentifier(docQueryId).setSentNr(sNr);
try {
if (extractionWin == -1
|| hasQueryMatch(ql.getQueryById(docQuerySnrId.getQueryId()), line)) {
// old sentences in buffer
if (lastSentences != null) {
int oldSNr = sNr - lastSentences.size();
while (lastSentences.size() > 0) {
Object lineStrObject = lastSentences.remove();
writeLine(
out,
(String) lineStrObject,
new TextIdentifier(docQueryId).setSentNr(oldSNr++));
}
}
// current sentence
writeLine(out, line, new TextIdentifier(docQueryId).setSentNr(sNr));
currentExtractionWin = extractionWin;
} else { // no match in sentence
// within look-forward window of last query-containing sentence
if (currentExtractionWin > 0) {
writeLine(out, line, new TextIdentifier(docQueryId).setSentNr(sNr));
--currentExtractionWin;
} else { // possibly in look-behind window
if (lastSentences != null) {
lastSentences.add(line);
}
}
}
} catch (IllegalArgumentException e) {
throw new IllegalArgumentException(
"Error processing sentence with id:"
+ " \n"
+ docQueryId
+ "\n"
+ docQuerySnrId
+ "\n"
+ e);
}
sNr += 1;
}
}
out.flush();
br.close();
}
