/**
* Reads parse trees either from standard input or a specified file, converting them to sentences
* and printing those sentences on standard output.
*
* <pre>
* usage: [-v|-help|-usage] [-tags] [filename]
* -v|-help|-usage: prints out this message
* -tags: indicates to spit out one S-expression per word, of the form
* (word (tag))
* filename is the file to be processed (standard input is assumed if
* this argument is "-" or is not present)
* </pre>
*/
public static void main(String[] args) {
InputStream inStream = System.in;
boolean tags = false;
String inFile = null;
for (int i = 0; i < args.length; i++) {
if (args[i].equals("-help") || args[i].equals("-usage") || args[i].equals("-v")) {
usage();
return;
} else if (args[i].equals("-tags")) tags = true;
else if (!args[i].equals("-")) inFile = args[i];
}
if (inFile != null) {
try {
inStream = new FileInputStream(inFile);
} catch (FileNotFoundException fnfe) {
System.err.println(fnfe);
System.exit(-1);
}
}
try {
SexpTokenizer tok = new SexpTokenizer(inStream, Language.encoding(), bufSize);
OutputStream os = System.out;
Writer writer = new BufferedWriter(new OutputStreamWriter(os, Language.encoding()));
PrintWriter pw = new PrintWriter(writer);
Sexp curr = null;
while ((curr = Sexp.read(tok)) != null)
pw.println(tags ? Util.collectTaggedWords(curr) : Util.collectLeaves(curr));
pw.flush();
pw.close();
} catch (Exception e) {
System.out.println(e);
}
}
/**
* Returns <code>true</code> if <code>preterminal</code> represents a terminal with one of the
* following parts of speech: <tt>VB, VBD, VBG, VBN, VBP</tt> or <tt>VBZ</tt>. It is an error to
* call this method with a <code>Sexp</code> object for which {@link #isPreterminal(Sexp)} returns
* <code>false</code>.<br>
*
* @param preterminal the preterminal to test
* @return <code>true</code> if <code>preterminal</code> is a verb
*/
public boolean isVerb(Sexp preterminal) {
return isVerbTag(preterminal.list().get(0).symbol());
}
/**
* Returns <code>true</code> if the specified S-expression represents a preterminal that is the
* possessive part of speech. This method is intended to be used by implementations of {@link
* Training#addBaseNPs(Sexp)}.
*/
public boolean isPossessivePreterminal(Sexp tree) {
return (isPreterminal(tree) && tree.list().get(0).symbol() == possessivePos);
}
/**
* Returns <code>true</code> if the specified S-expression is a preterminal whose part of speech
* is <code>","</code> or <code>":"</code>.
*/
public boolean isPuncToRaise(Sexp preterm) {
return (isPreterminal(preterm) && puncToRaise.contains(preterm.list().first()));
}
/**
* Returns <code>true</code> if the specified S-expression represents a preterminal whose terminal
* element is the null element (<code>"-NONE-"</code>) for the Penn Treebank.
*
* @see Training#relabelSubjectlessSentences(Sexp)
*/
public boolean isNullElementPreterminal(Sexp tree) {
return (isPreterminal(tree) && tree.list().get(0).symbol() == nullElementPreterminal);
}
/**
* Returns <code>true</code> if <code>tree</code> represents a preterminal subtree (part-of-speech
* tag and word). Specifically, this method returns <code>true</code> if <code>tree</code> is an
* instance of <code>SexpList</code>, has a length of 2 and has a first list element of type
* <code>Symbol</code>.
*/
public final boolean isPreterminal(Sexp tree) {
return (tree.isList()
&& tree.list().length() == 2
&& tree.list().get(0).isSymbol()
&& tree.list().get(1).isSymbol());
}