@Test
public void splitWordAndLength() {
StringBuilder sb = new StringBuilder();
List<String> brokenStrings = new ArrayList<String>();
BreakIterator boundary = BreakIterator.getWordInstance();
boundary.setText(val);
int start = boundary.first();
for (int end = boundary.next(); end != BreakIterator.DONE; start = end, end = boundary.next()) {
int lengthOfNext = end - start;
if ((sb.length() + lengthOfNext) > 180) {
brokenStrings.add(sb.toString());
sb = new StringBuilder(); // or set to 0
}
sb.append(val.substring(start, end));
// if last element
if (end == val.length()) {
brokenStrings.add(sb.toString());
}
}
for (String x : brokenStrings) {
System.out.println(x);
}
}
/**
* 格式化字符串，如果字符串超过指定长度，则自动折行。
*
* @param text 要格式化的字符串
* @param maxLength 行的长度
* @param locale 国家地区
* @param prefix1 首行前缀
* @param prefix2 第二行及后面行的前缀
* @return 格式化后的字符串
*/
private String formatLines(
String text, int maxLength, Locale locale, String prefix1, String prefix) {
BreakIterator boundary = BreakIterator.getLineInstance(locale);
StringBuffer result = new StringBuffer(prefix1);
boundary.setText(text);
int start = boundary.first();
int end = boundary.next();
int lineLength = 0;
while (end != BreakIterator.DONE) {
String word = text.substring(start, end);
lineLength = lineLength + word.length();
if (lineLength >= maxLength) {
result.append("\n").append(prefix);
lineLength = word.length();
}
result.append(word);
start = end;
end = boundary.next();
}
return result.toString();
}
/**
* Converts a line of text into an array of lower case words using a BreakIterator.wordInstance().
*
* <p>This method is under the Jive Open Source Software License and was written by Mark Imbriaco.
*
* @param text a String of text to convert into an array of words
* @return text broken up into an array of words.
*/
public static String[] toLowerCaseWordArray(String text) {
if (text == null || text.length() == 0) {
return new String[0];
}
List<String> wordList = new ArrayList<String>();
BreakIterator boundary = BreakIterator.getWordInstance();
boundary.setText(text);
int start = 0;
for (int end = boundary.next(); end != BreakIterator.DONE; start = end, end = boundary.next()) {
String tmp = text.substring(start, end).trim();
// Remove characters that are not needed.
tmp = replace(tmp, "+", "");
tmp = replace(tmp, "/", "");
tmp = replace(tmp, "\\", "");
tmp = replace(tmp, "#", "");
tmp = replace(tmp, "*", "");
tmp = replace(tmp, ")", "");
tmp = replace(tmp, "(", "");
tmp = replace(tmp, "&", "");
if (tmp.length() > 0) {
wordList.add(tmp);
}
}
return wordList.toArray(new String[wordList.size()]);
}
/**
* LIU: Finds the longest substring that fits a given width composed of subunits returned by a
* BreakIterator. If the smallest subunit is too long, returns 0.
*
* @param fMtr metrics to use
* @param line the string to be fix into width
* @param width line.substring(0, result) must be <= width
* @param breaker the BreakIterator that will be used to find subunits
* @return maximum characters, at boundaries returned by breaker, that fit into width, or zero on
* failure
*/
private int findFittingBreak(FontMetrics fMtr, String line, int width, BreakIterator breaker) {
breaker.setText(line);
int last = breaker.first();
int end = breaker.next();
while (end != BreakIterator.DONE && visibleWidth(fMtr, line.substring(0, end)) <= width) {
last = end;
end = breaker.next();
}
return last;
}
private void prepTxt(String txtInit, float fontSizeInit, float maxWidth) {
if (txtInit == null) throw new NullPointerException();
setFontSize(fontSizeInit);
txt = txtInit;
fontSize = fontSizeInit;
areaWidth = maxWidth - pad;
lineHeight = getTextAsc() + getTextDesc();
if (lineList == null) lineList = new ArrayList<String>();
else lineList.clear();
BreakIterator boundary = BreakIterator.getWordInstance();
boundary.setText(txt);
int start = boundary.first();
int end = boundary.next();
int prevEnd = start;
while (end != BreakIterator.DONE) {
String line = txt.substring(start, end);
String prevLine = txt.substring(start, prevEnd);
float lineWidth = getTextWidth(line);
if (lineWidth > areaWidth) {
// If the first word is longer than lineWidth
// prevLine is empty and should be ignored
if (prevLine.length() > 0) lineList.add(prevLine);
start = prevEnd;
}
prevEnd = end;
end = boundary.next();
}
String line = txt.substring(start, prevEnd);
lineList.add(line);
if (lines == null || lines.length != lineList.size()) lines = new String[lineList.size()];
if (lineWidths == null || lineWidths.length != lineList.size())
lineWidths = new float[lineList.size()];
lineList.toArray(lines);
maxLineWidth = 0;
for (int i = 0; i < lines.length; i++) {
lineWidths[i] = getTextWidth(lines[i]);
if (maxLineWidth < lineWidths[i]) maxLineWidth = lineWidths[i];
}
areaWidth = maxLineWidth;
areaHeight = lineHeight * lines.length;
width = areaWidth + pad * 2;
height = areaHeight + pad * 2;
}
@Override
public Object evaluate(DeferredObject[] arguments) throws HiveException {
assert (arguments.length >= 1 && arguments.length <= 3);
if (arguments[0].get() == null) {
return null;
}
// if there is more than 1 argument specified, a different natural language
// locale is being specified
Locale locale = null;
if (arguments.length > 1 && arguments[1].get() != null) {
Text language = (Text) converters[1].convert(arguments[1].get());
Text country = null;
if (arguments.length > 2 && arguments[2].get() != null) {
country = (Text) converters[2].convert(arguments[2].get());
}
if (country != null) {
locale = new Locale(language.toString().toLowerCase(), country.toString().toUpperCase());
} else {
locale = new Locale(language.toString().toLowerCase());
}
} else {
locale = Locale.getDefault();
}
// get the input and prepare the output
Text chunk = (Text) converters[0].convert(arguments[0].get());
String text = chunk.toString();
ArrayList<ArrayList<Text>> result = new ArrayList<ArrayList<Text>>();
// Parse out sentences using Java's text-handling API
BreakIterator bi = BreakIterator.getSentenceInstance(locale);
bi.setText(text);
int idx = 0;
while (bi.next() != BreakIterator.DONE) {
String sentence = text.substring(idx, bi.current());
idx = bi.current();
result.add(new ArrayList<Text>());
// Parse out words in the sentence
BreakIterator wi = BreakIterator.getWordInstance(locale);
wi.setText(sentence);
int widx = 0;
ArrayList<Text> sent_array = result.get(result.size() - 1);
while (wi.next() != BreakIterator.DONE) {
String word = sentence.substring(widx, wi.current());
widx = wi.current();
if (Character.isLetterOrDigit(word.charAt(0))) {
sent_array.add(new Text(word));
}
}
}
return result;
}
static List<String> extractTokens(String document) {
BreakIterator iterator = BreakIterator.getWordInstance();
iterator.setText(document);
ArrayList<String> result = new ArrayList<String>();
int start = iterator.first();
for (int end = iterator.next(); end != BreakIterator.DONE; start = end, end = iterator.next()) {
String s = document.substring(start, end).toLowerCase().replaceAll("[^a-z]", "");
if (!ENGLISH_STOP_WORDS.contains(s)) {
result.add(s);
}
}
return result;
}
/**
* Extract out sentences from the reviews. to take into account the negative lists Later Check
* Stanford Document tokenizer
*/
private void BreakInLines() {
// this.Lines = review.split(". ");
BreakIterator border = BreakIterator.getSentenceInstance(Locale.US);
border.setText(review);
// System.out.println(review);
int start = border.first();
// iterate, creating sentences out of all the Strings between the given boundaries
for (int end = border.next(); end != BreakIterator.DONE; start = end, end = border.next()) {
// System.out.println(review.substring(start,end));
Lines.add(review.substring(start, end));
NumOfSentences++;
}
// System.out.println(NumOfSentences);
}
private void calcBoxDimensions(CharSequence txtInit, float fontSizeInit, float maxWidth) {
if (txtInit == null) throw new NullPointerException();
setFontSize(fontSizeInit);
txt = txtInit;
fontSize = fontSizeInit;
areaWidth = maxWidth - pad;
linesHeight = getTextAsc() + getTextDesc();
if (lineList == null) lineList = new ArrayList<CharSequence>();
else lineList.clear();
BreakIterator boundary = BreakIterator.getWordInstance();
boundary.setText(txt.toString());
int start = boundary.first();
int end = boundary.next();
int prevEnd = start;
while (end != BreakIterator.DONE) {
CharSequence line = txt.subSequence(start, end);
CharSequence prevLine = txt.subSequence(start, prevEnd);
float lineWidth = getTextWidth(line, 0, line.length());
if (lineWidth > areaWidth) {
// If the first word is longer than lineWidth
// prevLine is empty and should be ignored
if (prevLine.length() > 0) lineList.add(prevLine);
start = prevEnd;
}
prevEnd = end;
end = boundary.next();
}
CharSequence line = txt.subSequence(start, prevEnd);
lineList.add(line);
maxLinesWidth = 0;
for (CharSequence seq : lineList) {
float lineWidth = getTextWidth(seq, 0, seq.length());
if (maxLinesWidth < lineWidth) maxLinesWidth = lineWidth;
}
areaWidth = maxLinesWidth;
areaHeight = linesHeight * lineList.size();
width = areaWidth + pad * 2;
height = areaHeight + pad * 2;
}
public void check(String name, String in, String[] out, BreakIterator bi, TestHarness harness) {
harness.checkPoint(name);
bi.setText(in);
int index = 0;
int from = bi.current();
harness.check(from, 0);
while (true) {
int to = bi.next();
if (to == BreakIterator.DONE) break;
harness.check(in.substring(from, to), out[index]);
++index;
from = to;
}
harness.check(index, out.length);
harness.checkPoint("backwards " + name);
bi.last();
index = out.length - 1;
from = bi.current();
harness.check(from, in.length());
while (true) {
int to = bi.previous();
if (to == BreakIterator.DONE) break;
harness.check(in.substring(to, from), out[index]);
--index;
from = to;
}
harness.check(index, -1);
}
/**
* set / update the text of the displayLabels. these are the Week column headers above the days on
* the Calendar part of the <code>CDateTime</code>.
*/
private void updateDaysOfWeek() {
if (dayPanel != null) {
Calendar tmpcal = cdt.getCalendarInstance();
tmpcal.set(Calendar.DAY_OF_WEEK, tmpcal.getFirstDayOfWeek());
Locale locale = cdt.getLocale();
boolean ltr =
(ComponentOrientation.getOrientation(locale).isLeftToRight()
&& !locale.getLanguage().equals("zh")); // $NON-NLS-1$
BreakIterator iterator = BreakIterator.getCharacterInstance(locale);
for (int x = 0; x < dayLabels.length; x++) {
String str = getFormattedDate("E", tmpcal.getTime()); // $NON-NLS-1$
if (dayLabels[x].getData(CDT.Key.Compact, Boolean.class)) {
iterator.setText(str);
int start, end;
if (ltr) {
start = iterator.first();
end = iterator.next();
} else {
end = iterator.last();
start = iterator.previous();
}
dayLabels[x].setText(str.substring(start, end));
} else {
dayLabels[x].setText(str);
}
tmpcal.add(Calendar.DAY_OF_WEEK, 1);
}
}
}
static HashMap<String, WordTuple> findWordsInSentences(
String target, BreakIterator wordIterator, ArrayList<Integer> sentences) {
HashMap<String, WordTuple> wordMap = new HashMap<String, WordTuple>();
wordIterator.setText(target);
int start = wordIterator.first();
int end = wordIterator.next();
while (end != BreakIterator.DONE) {
String word = target.substring(start, end);
if (Character.isLetterOrDigit(word.charAt(0))) {
// System.out.println(word);
// System.out.println(start + "-" + end);
// check which sentence the word is in by comparing end with values in sentences
int sentenceNo = 0;
for (int i = 0; i < sentences.size(); i++) {
if (end <= sentences.get(i)) {
sentenceNo = i;
break;
}
}
// lowercase the word
String wordLc = word.toLowerCase();
// check if word exists in hashmap
if (wordMap.containsKey(wordLc)) {
// if exists, add sentence number to word's list in hashmap
WordTuple wordTuple = wordMap.get(wordLc);
ArrayList<Integer> sentenceList = wordTuple.getSentenceList();
sentenceList.add(sentenceNo);
wordMap.put(wordLc, wordTuple);
} else {
// if it does not exist, create list, add sentence number to list, and add list to hashmap
// with word as key
ArrayList<Integer> sentenceList = new ArrayList<Integer>();
sentenceList.add(sentenceNo);
WordTuple wordTuple = new WordTuple();
wordTuple.setSentenceList(sentenceList);
wordMap.put(wordLc, wordTuple);
}
}
start = end;
end = wordIterator.next();
}
return wordMap;
}
private static Tokens splitText(final String text) {
final List<Token> l = new LinkedList<>();
// use a BreakIterator to iterate our way through the words of the text
final BreakIterator wordIterator = BreakIterator.getWordInstance(new Locale("en", "US"));
wordIterator.setText(text);
// simply iterate through the text, keeping track of a start and end index of the current word
int startIdx = wordIterator.first();
for (int endIdx = wordIterator.next();
endIdx != DONE;
startIdx = endIdx, endIdx = wordIterator.next()) {
final String word = text.substring(startIdx, endIdx);
l.add(new Token(startIdx, word));
}
return new Tokens(l);
}
/**
* 返回一段文字中的单词的数组
*
* @param text
* @return
*/
public static final String[] toLowerCaseWordArray(String text) {
if (text == null || text.length() == 0) return new String[0];
ArrayList wordList = new ArrayList();
BreakIterator boundary = BreakIterator.getWordInstance();
boundary.setText(text);
int start = 0;
for (int end = boundary.next(); end != -1; end = boundary.next()) {
String tmp = text.substring(start, end).trim();
tmp = replace(tmp, "+", "");
tmp = replace(tmp, "/", "");
tmp = replace(tmp, "\\", "");
tmp = replace(tmp, "#", "");
tmp = replace(tmp, "*", "");
tmp = replace(tmp, ")", "");
tmp = replace(tmp, "(", "");
tmp = replace(tmp, "&", "");
if (tmp.length() > 0) wordList.add(tmp);
start = end;
}
return (String[]) wordList.toArray(new String[wordList.size()]);
}
public static String convertStringToTitleCase(String toConvert) {
BreakIterator wordBreaker = BreakIterator.getWordInstance();
wordBreaker.setText(toConvert);
int end;
String word = "";
for (int start = wordBreaker.first();
(end = wordBreaker.next()) != BreakIterator.DONE;
start = end) {
word += StringProcessing.wordToTitleCase(toConvert.substring(start, end));
}
return word;
}
static ArrayList<Integer> findSentenceBoundaries(String target, BreakIterator iterator) {
ArrayList<Integer> sentenceBoundaryList = new ArrayList<Integer>();
iterator.setText(target);
int boundary = iterator.first();
while (boundary != BreakIterator.DONE) {
boundary = iterator.next();
if (boundary != -1) {
sentenceBoundaryList.add(boundary);
}
}
return sentenceBoundaryList;
}
public static void javaBreakIterator() {
BreakIterator wordIterator = BreakIterator.getWordInstance();
String text = "Let's pause, and then reflect.";
wordIterator.setText(text);
int boundary = wordIterator.first();
while (boundary != BreakIterator.DONE) {
int begin = boundary;
System.out.print(boundary + "-");
boundary = wordIterator.next();
int end = boundary;
if (end == BreakIterator.DONE) break;
System.out.print(boundary + " [" + text.substring(begin, end) + "];");
}
System.out.println();
}
static String extractShortDescription(String description) {
if (description == null) {
return null;
}
int dot = description.indexOf(".");
if (dot != -1) {
BreakIterator breakIterator = BreakIterator.getSentenceInstance(Locale.US);
breakIterator.setText(description);
String text = description.substring(breakIterator.first(), breakIterator.next()).trim();
return removeSpaceBetweenLine(text);
} else {
String[] lines = description.split(NEW_LINE);
return lines[0].trim();
}
}
@Test
public void testSentenceDetection() {
BreakIterator sentenceIterator = BreakIterator.getSentenceInstance(Locale.US);
sentenceIterator.setText(TEST_STRING);
int start = sentenceIterator.first();
int end = -1;
List<String> sentenceList = new ArrayList<String>();
while ((end = sentenceIterator.next()) != BreakIterator.DONE) {
String sentence = TEST_STRING.substring(start, end);
start = end;
sentenceList.add(sentence);
System.out.println("Sentence: " + sentence);
}
}
/**
* Called to summarize a document when no hits were found. By default this just returns the first
* {@code maxPassages} sentences; subclasses can override to customize.
*/
protected Passage[] getEmptyHighlight(String fieldName, BreakIterator bi, int maxPassages) {
// BreakIterator should be un-next'd:
List<Passage> passages = new ArrayList<>();
int pos = bi.current();
assert pos == 0;
while (passages.size() < maxPassages) {
int next = bi.next();
if (next == BreakIterator.DONE) {
break;
}
Passage passage = new Passage();
passage.score = Float.NaN;
passage.startOffset = pos;
passage.endOffset = next;
passages.add(passage);
pos = next;
}
return passages.toArray(new Passage[passages.size()]);
}
/** Returns the next word in the text */
public String nextWord() {
if (!first) {
currentWordPos = nextWordPos;
currentWordEnd = getNextWordEnd(text, currentWordPos);
nextWordPos = getNextWordStart(text, currentWordEnd + 1);
int current = sentanceIterator.current();
if (current == currentWordPos) startsSentance = true;
else {
startsSentance = false;
if (currentWordEnd > current) sentanceIterator.next();
}
}
// The nextWordPos has already been populated
String word = null;
try {
word = document.getText(currentWordPos, currentWordEnd - currentWordPos);
} catch (BadLocationException ex) {
moreTokens = false;
}
wordCount++;
first = false;
if (nextWordPos == -1) moreTokens = false;
return word;
}
/**
* Initializes fields comment, inlineTags of the object
*
* @param commentText the processed comment text
*/
private void procComment(String commentText) {
// initialize inlineTags
ArrayList<Tag> result = new ArrayList<Tag>();
String noInlineTags = replaceAtSigns(commentText);
/*
* Pattern p = Pattern.compile("\\{\\s*@[^}]*\\}"); // matches inline
* tags // Pattern p =
* Pattern.compile("\\{\\s*@([^\\s\\}]+)\\s*([^\\}]*)\\s*}"); // matches
* inline tags Matcher m = p.matcher(commentText); int start = 0, end =
* 0; // create an array of tag objects of kind "Text" and "@link"; as
* explained in the // doclet API, for a comment // This is a {@link Doc
* commentlabel} example. // create an array of Tag objects: // *
* tags[0] is a Tag with name "Text" and text consisting of "This is a "
* // * tags[1] is a SeeTag with name "@link", and label "commentlabel"
* // * tags[2] is a Tag with name "Text" and text consisting of
* " example." while (m.find()) { end = m.start(); String linkText =
* m.group(); // System.out.print("String = \"" +
* commentText.substring(start, end)); //
* System.out.println("\"; linkText = \"" + linkText + "\""); //
* result.add(new X10Tag("Text", commentText.substring(start, end),
* this)); result.add(X10Tag.processInlineTag(linkText, this)); //int
* index = commentText.indexOf(linkText); //commentText =
* commentText.substring(0, index) + commentText.substring(index +
* linkText.length()); // result.add(new X10SeeTag(true, linkText,
* this)); // "true" signifies an @link tag, as opposed to an @see tag
* start = m.end(); }
*/
if (!commentText.startsWith("@")) { // make sure that there is a
// beginning paragraph
// initialize comment
int blockTagStart = noInlineTags.indexOf("@"); // start of block
// tags within
// comment
blockTagStart = (blockTagStart == -1) ? commentText.length() : blockTagStart;
this.comment = commentText.substring(0, blockTagStart).trim();
if (!comment.equals("")) {
result.addAll(createInlineTags(comment, this));
}
// }
// add constraints, if any
// String decl = declString();
// if (decl != null) {
// result.add(new X10Tag(decl, this));
// }
// initialize firstSentenceTags
BreakIterator b = BreakIterator.getSentenceInstance();
b.setText(comment);
int start = 0;
int end = 0;
start = b.first();
end = b.next();
String firstSentence = ((start <= end) ? comment.substring(start, end).trim() : "");
// System.out.println("X10Doc.initializeFields(): firstSentence = \""
// + firstSentence + "\"");
firstSentenceTags = createInlineTags(firstSentence, this).toArray(new X10Tag[0]);
} else {
firstSentenceTags = new X10Tag[0];
}
inlineTags = result.toArray(new X10Tag[0]);
// TODO: creating Tag objects for block tags and storing them in a field
// of this object
Pattern blockTagPattern = Pattern.compile("\\s*@[^@]*");
Matcher blockTagMatcher = blockTagPattern.matcher(noInlineTags);
while (blockTagMatcher.find()) {
String tagText = blockTagMatcher.group();
int start = blockTagMatcher.start();
processBlockTag(commentText.substring(start, start + tagText.length()));
}
}
// algorithm: treat sentence snippets as miniature documents
// we can intersect these with the postings lists via BreakIterator.preceding(offset),s
// score each sentence as norm(sentenceStartOffset) * sum(weight * tf(freq))
private Passage[] highlightDoc(
String field,
BytesRef terms[],
int contentLength,
BreakIterator bi,
int doc,
TermsEnum termsEnum,
DocsAndPositionsEnum[] postings,
int n)
throws IOException {
PassageScorer scorer = getScorer(field);
if (scorer == null) {
throw new NullPointerException("PassageScorer cannot be null");
}
PriorityQueue<OffsetsEnum> pq = new PriorityQueue<>();
float weights[] = new float[terms.length];
// initialize postings
for (int i = 0; i < terms.length; i++) {
DocsAndPositionsEnum de = postings[i];
int pDoc;
if (de == EMPTY) {
continue;
} else if (de == null) {
postings[i] = EMPTY; // initially
if (!termsEnum.seekExact(terms[i])) {
continue; // term not found
}
de =
postings[i] = termsEnum.docsAndPositions(null, null, DocsAndPositionsEnum.FLAG_OFFSETS);
if (de == null) {
// no positions available
throw new IllegalArgumentException(
"field '" + field + "' was indexed without offsets, cannot highlight");
}
pDoc = de.advance(doc);
} else {
pDoc = de.docID();
if (pDoc < doc) {
pDoc = de.advance(doc);
}
}
if (doc == pDoc) {
weights[i] = scorer.weight(contentLength, de.freq());
de.nextPosition();
pq.add(new OffsetsEnum(de, i));
}
}
pq.add(new OffsetsEnum(EMPTY, Integer.MAX_VALUE)); // a sentinel for termination
PriorityQueue<Passage> passageQueue =
new PriorityQueue<>(
n,
new Comparator<Passage>() {
@Override
public int compare(Passage left, Passage right) {
if (left.score < right.score) {
return -1;
} else if (left.score > right.score) {
return 1;
} else {
return left.startOffset - right.startOffset;
}
}
});
Passage current = new Passage();
OffsetsEnum off;
while ((off = pq.poll()) != null) {
final DocsAndPositionsEnum dp = off.dp;
int start = dp.startOffset();
if (start == -1) {
throw new IllegalArgumentException(
"field '" + field + "' was indexed without offsets, cannot highlight");
}
int end = dp.endOffset();
// LUCENE-5166: this hit would span the content limit... however more valid
// hits may exist (they are sorted by start). so we pretend like we never
// saw this term, it won't cause a passage to be added to passageQueue or anything.
assert EMPTY.startOffset() == Integer.MAX_VALUE;
if (start < contentLength && end > contentLength) {
continue;
}
if (start >= current.endOffset) {
if (current.startOffset >= 0) {
// finalize current
current.score *= scorer.norm(current.startOffset);
// new sentence: first add 'current' to queue
if (passageQueue.size() == n && current.score < passageQueue.peek().score) {
current.reset(); // can't compete, just reset it
} else {
passageQueue.offer(current);
if (passageQueue.size() > n) {
current = passageQueue.poll();
current.reset();
} else {
current = new Passage();
}
}
}
// if we exceed limit, we are done
if (start >= contentLength) {
Passage passages[] = new Passage[passageQueue.size()];
passageQueue.toArray(passages);
for (Passage p : passages) {
p.sort();
}
// sort in ascending order
Arrays.sort(
passages,
new Comparator<Passage>() {
@Override
public int compare(Passage left, Passage right) {
return left.startOffset - right.startOffset;
}
});
return passages;
}
// advance breakiterator
assert BreakIterator.DONE < 0;
current.startOffset = Math.max(bi.preceding(start + 1), 0);
current.endOffset = Math.min(bi.next(), contentLength);
}
int tf = 0;
while (true) {
tf++;
BytesRef term = terms[off.id];
if (term == null) {
// multitermquery match, pull from payload
term = off.dp.getPayload();
assert term != null;
}
current.addMatch(start, end, term);
if (off.pos == dp.freq()) {
break; // removed from pq
} else {
off.pos++;
dp.nextPosition();
start = dp.startOffset();
end = dp.endOffset();
}
if (start >= current.endOffset || end > contentLength) {
pq.offer(off);
break;
}
}
current.score += weights[off.id] * scorer.tf(tf, current.endOffset - current.startOffset);
}
// Dead code but compiler disagrees:
assert false;
return null;
}
