/**
* Takes the raw page lines represented as one continuous line and sorts the text by the y access
* of the word bounds. The words are then sliced into separate lines base on y changes. And
* finally each newly sorted line is sorted once more by each words x coordinate.
*/
public void sortAndFormatText() {
ArrayList<LineText> visiblePageLines = new ArrayList<LineText>(pageLines);
// create new array for storing the sorted lines
ArrayList<LineText> sortedPageLines = sortLinesVertically(visiblePageLines);
// try and insert the option words on existing lines
if (sortedPageLines.size() == 0) {
sortedPageLines = getVisiblePageLines(true);
} else {
insertOptionalLines(sortedPageLines);
}
// sort again
sortedPageLines = sortLinesVertically(sortedPageLines);
// do a rough check for duplicate strings that are sometimes generated
// by Chrystal Reports. Enable with
// -Dorg.icepdf.core.views.page.text.trim.duplicates=true
if (checkForDuplicates) {
for (final LineText lineText : sortedPageLines) {
final List<WordText> words = lineText.getWords();
if (words.size() > 0) {
final List<WordText> trimmedWords = new ArrayList<WordText>();
final Set<String> refs = new HashSet<String>();
for (final WordText wordText : words) {
// use regular rectangle so get a little rounding.
final String key = wordText.getText() + wordText.getBounds().getBounds();
if (refs.add(key)) {
trimmedWords.add(wordText);
}
}
lineText.setWords(trimmedWords);
}
}
}
// sort each line by x coordinate.
if (sortedPageLines.size() > 0) {
for (LineText lineText : sortedPageLines) {
Collections.sort(lineText.getWords(), new WordPositionComparator());
}
}
// recalculate the line bounds.
if (sortedPageLines.size() > 0) {
for (LineText lineText : sortedPageLines) {
lineText.getBounds();
}
}
// sort the lines
if (sortedPageLines.size() > 0 && !preserveColumns) {
Collections.sort(sortedPageLines, new LinePositionComparator());
}
// assign back the sorted lines.
this.sortedPageLines = sortedPageLines;
}
