public static TaggedDocument readTaggedDocument(String id, String dir, boolean printData) {
ObjectInput inputObject;
TaggedDocument td = null;
try {
inputObject =
new ObjectInputStream(new BufferedInputStream(new FileInputStream(dir + id + ".ser")));
try {
td = (TaggedDocument) inputObject.readObject();
} catch (ClassNotFoundException e) {
td = null;
Logger.logln("Couldn't load TaggedDocument: " + id + ", from: " + dir);
e.printStackTrace();
} finally {
inputObject.close();
}
} catch (FileNotFoundException e) {
e.printStackTrace();
} catch (IOException e) {
e.printStackTrace();
}
if (printData == true && td != null) {
System.out.println(td.toString());
}
return td;
}
/**
* Reads a saved serialized HashMap of TreeData objects in 'dir' named 'id' and returns the
* TreeContainer object.
*
* @param id name of HashMap
* @param dir location of saved .ser file
* @param printData if true, will print HashMap to string (via toString method)
* @return HashMap of TreeData objects specified by 'id' and 'dir', or null if no TreeContainer
* found
*/
@SuppressWarnings("unchecked")
public static HashMap<String, TreeData> readTreeDataMap(
String id, String dir, boolean printData) {
ObjectInput inputObject;
HashMap<String, TreeData> tdHash = null;
try {
inputObject =
new ObjectInputStream(new BufferedInputStream(new FileInputStream(dir + id + ".ser")));
try {
tdHash = (HashMap<String, TreeData>) inputObject.readObject();
} catch (ClassNotFoundException e) {
tdHash = null;
Logger.logln("Couldn't load ArrayList<TreeData>: " + id + ", from: " + dir);
e.printStackTrace();
} finally {
inputObject.close();
}
} catch (FileNotFoundException e) {
e.printStackTrace();
} catch (IOException e) {
e.printStackTrace();
}
if (printData == true && tdHash != null) {
System.out.println(tdHash.toString());
}
return tdHash;
}
/**
* Generic object writer
*
* @param o the Object to write
* @param id name of object
* @param dir directory to write the object to
* @return true if no errors, false otherwise
*/
public static boolean writeObject(Object o, String id, String dir) {
ObjectOutputStream outObject = null;
System.out.println("Place to write: " + dir + id + ".ser");
try {
outObject =
new ObjectOutputStream(new BufferedOutputStream(new FileOutputStream(dir + id + ".ser")));
try {
outObject.writeObject(o);
} finally {
outObject.close();
}
} catch (FileNotFoundException e) {
Logger.logln("ERROR saving object: " + o.toString());
e.printStackTrace();
return false;
} catch (IOException e) {
Logger.logln("ERROR saving object: " + o.toString());
e.printStackTrace();
return false;
}
return true;
}
public static void fatalError() {
int ans =
JOptionPane.showConfirmDialog(
null,
"Anonymouth has encountered a fatal error.\n"
+ "The best thing to do is to examine the stack trace, and terminate the program.\n"
+ "If you would like to save your work/problem set prior to closing Anonymouth,\n"
+ "click \"no\" below. Otherwise, click \"yes\" (this will close Anonymouth)",
"Fatal error encountered. Terminate?",
JOptionPane.YES_NO_OPTION);
if (ans == 0) {
Logger.logln("Fatal error encountered, termination requested.", LogOut.STDERR);
System.exit(FATAL_ERROR);
}
}
public static Document removeUnicodeControlChars(Document dirtyDoc) {
String newFile = "./temp/" + dirtyDoc.getTitle();
Document cleanDoc = new Document();
try {
dirtyDoc.load();
cleanDoc.setText(
(dirtyDoc.stringify()).replaceAll("\\p{C}&&[^\\t\\n\\r]", " ").toCharArray());
cleanDoc.setAuthor(dirtyDoc.getAuthor());
cleanDoc.setTitle(dirtyDoc.getTitle());
FileWriter fw = new FileWriter(new File("./temp/" + dirtyDoc.getTitle()));
return cleanDoc;
} catch (Exception e) {
// TODO Auto-generated catch block
e.printStackTrace();
Logger.logln(
"ERROR! Could not load document: "
+ dirtyDoc.getTitle()
+ " (SentenceTools.removeUnicodeControlChars)");
return dirtyDoc;
}
}
/**
* Takes a text (one String representing an entire document), and breaks it up into sentences.
* Tries to find true ends of sentences: shouldn't break up sentences containing quoted sentences,
* checks for sentences ending in a quoted sentence (e.x. He said, "Hello." ), will not break
* sentences containing common abbreviations (such as Dr., Mr. U.S., etc.,e.x., i.e., and others),
* and checks for ellipses points. However, It is probably not perfect.
*
* @param text
* @return
*/
public ArrayList<String> makeSentenceTokens(String text) {
ArrayList<String> sents = new ArrayList<String>(MAX_SENTENCES);
boolean merge1 = false, mergeFinal = false;
int currentStart = 1;
int currentStop = 0;
int lenText = text.length();
String temp;
int openingQuoteIndex = 0;
int closingQuoteIndex = 0;
text = text.replaceAll("\u201C", "\"");
text = text.replaceAll("\u201D", "\"");
text = text.replaceAll("\\p{C}&&[^\\t\\n\\r]", " ");
int notEOSNumber = 0;
int numNotEOS = notEndsOfSentence.length;
String replacementString = "";
String safeString = "";
for (notEOSNumber = 0; notEOSNumber < numNotEOS; notEOSNumber++) {
replacementString = notEndsOfSentence[notEOSNumber].replaceAll("\\.", PERIOD_REPLACEMENT);
// System.out.println("REPLACEMENT: "+replacementString);
safeString = notEndsOfSentence[notEOSNumber].replaceAll("\\.", "\\\\.");
// System.out.println(safeString);
text = text.replaceAll("\\b(?i)" + safeString, replacementString);
}
Matcher sent = EOS_chars.matcher(text);
boolean foundEOS = sent.find(currentStart);
Matcher sentEnd;
int charNum = 0;
int lenString = 0;
int lastQuoteAt = 0;
boolean foundQuote = false;
boolean isSentence;
while (foundEOS == true) {
currentStop = sent.end();
// System.out.println("Start: "+currentStart+" and Stop: "+currentStop);
temp = text.substring(currentStart - 1, currentStop);
// System.out.println(temp);
lenString = temp.length();
lastQuoteAt = 0;
foundQuote = false;
for (charNum = 0; charNum < lenString; charNum++) {
if (temp.charAt(charNum) == '\"') {
lastQuoteAt = charNum;
if (foundQuote == true) {
foundQuote = false;
} else {
foundQuote = true;
}
// System.out.println("Found quote!!! here it is: "+temp.charAt(charNum)+" ... in
// position: "+lastQuoteAt+" ... foundQuote is: "+foundQuote);
}
}
if (foundQuote == true && ((closingQuoteIndex = temp.indexOf("\"", lastQuoteAt + 1)) == -1)) {
if ((currentStop = text.indexOf("\"", currentStart + lastQuoteAt + 1)) == -1) {
currentStop = text.length();
} else {
currentStop += 1;
merge1 = true;
}
}
safeString = text.substring(currentStart - 1, currentStop);
sentEnd = sentence_quote.matcher(safeString);
isSentence = sentEnd.find();
// System.out.println("RESULT OF sentence_quote matching: "+isSentence);
if (isSentence
== true) { // If it seems that the text looks like this: He said, "Hello." Then she said,
// "Hi."
// Then we want to split this up into two sentences (it's possible to have a sentence like
// this: He said, "Hello.")
// System.out.println("start: "+sentEnd.start()+" ... end: "+sentEnd.end());
currentStop = text.indexOf("\"", sentEnd.start() + currentStart) + 1;
safeString = text.substring(currentStart - 1, currentStop);
}
safeString = safeString.replaceAll(PERIOD_REPLACEMENT, ".");
// System.out.println(safeString);
if (mergeFinal) {
mergeFinal = false;
String prev = sents.remove(sents.size() - 1);
safeString = prev + safeString;
}
if (merge1) { // makes so that the merge happens on the next pass through
merge1 = false;
mergeFinal = true;
}
sents.add(safeString);
// System.out.println("start minus one: "+(currentStart-1)+" stop: "+currentStop);
if (currentStart < 0 || currentStop < 0) {
Logger.logln("Something went really wrong making sentence tokens.");
System.exit(0);
}
// System.out.println("The rest of the text: "+text.substring(currentStart));
currentStart = currentStop + 1;
if (currentStart >= lenText) {
foundEOS = false;
continue;
}
foundEOS = sent.find(currentStart);
}
return sents;
}