/**
* Find some protected parts defined in Tag Validation Options dialog: printf variables, java
* MessageFormat patterns, user defined cusom tags.
*
* <p>These protected parts shouldn't affect statistic but just be displayed in gray in editor and
* take part in tag validation.
*/
public static List<ProtectedPart> applyCustomProtectedParts(
String source, Pattern protectedPartsPatterns, List<ProtectedPart> protectedParts) {
List<ProtectedPart> result;
if (protectedParts != null) {
// Remove already define protected parts first for prevent intersection
for (ProtectedPart pp : protectedParts) {
source = source.replace(pp.getTextInSourceSegment(), StaticUtils.TAG_REPLACEMENT);
}
result = protectedParts;
} else {
result = new ArrayList<ProtectedPart>();
}
Matcher placeholderMatcher = protectedPartsPatterns.matcher(source);
while (placeholderMatcher.find()) {
ProtectedPart pp = new ProtectedPart();
pp.setTextInSourceSegment(placeholderMatcher.group());
pp.setDetailsFromSourceFile(placeholderMatcher.group());
pp.setReplacementWordsCountCalculation(placeholderMatcher.group());
pp.setReplacementUniquenessCalculation(placeholderMatcher.group());
pp.setReplacementMatchCalculation(placeholderMatcher.group());
result.add(pp);
}
return result;
}
@Override
public String constructShortcuts(List<Element> elements, List<ProtectedPart> protectedParts) {
protectedParts.clear();
// create shortcuts
InlineTagHandler tagHandler = new InlineTagHandler();
StringBuilder r = new StringBuilder();
for (Element el : elements) {
if (el instanceof XMLContentBasedTag) {
XMLContentBasedTag tag = (XMLContentBasedTag) el;
String shortcut = null;
int shortcutLetter;
int tagIndex;
boolean tagProtected;
if ("bpt".equals(tag.getTag())) {
// XLIFF specification requires 'rid' and 'id' attributes,
// but some tools uses 'i' attribute like for TMX
tagHandler.startBPT(
tag.getAttribute("rid"), tag.getAttribute("id"), tag.getAttribute("i"));
shortcutLetter = calcTagShortcutLetter(tag, ignoreTypeForBptTags);
tagHandler.setTagShortcutLetter(shortcutLetter);
tagIndex = tagHandler.endBPT();
shortcut =
"<"
+ (shortcutLetter != 0 ? String.valueOf(Character.toChars(shortcutLetter)) : 'f')
+ tagIndex
+ '>';
tagProtected = false;
} else if ("ept".equals(tag.getTag())) {
tagHandler.startEPT(
tag.getAttribute("rid"), tag.getAttribute("id"), tag.getAttribute("i"));
tagIndex = tagHandler.endEPT();
shortcutLetter = tagHandler.getTagShortcutLetter();
shortcut =
"</"
+ (shortcutLetter != 0 ? String.valueOf(Character.toChars(shortcutLetter)) : 'f')
+ tagIndex
+ '>';
tagProtected = false;
} else if ("it".equals(tag.getTag())) {
tagHandler.startOTHER();
tagHandler.setCurrentPos(tag.getAttribute("pos"));
tagIndex = tagHandler.endOTHER();
// XLIFF specification requires 'open/close' values,
// but some tools may use 'begin/end' values like for TMX
shortcutLetter = calcTagShortcutLetter(tag);
if ("close".equals(tagHandler.getCurrentPos())
|| "end".equals(tagHandler.getCurrentPos())) {
// In some cases, even if we're able to compute a shortcut, it's better to force to "f"
// for better compatibility with corresponding TMX files
if (forceShortCutToF) {
shortcutLetter = 'f';
}
shortcut =
"</"
+ (shortcutLetter != 0
? String.valueOf(Character.toChars(shortcutLetter))
: 'f')
+ tagIndex
+ '>';
} else {
shortcut =
"<"
+ (shortcutLetter != 0
? String.valueOf(Character.toChars(shortcutLetter))
: 'f')
+ tagIndex
+ '>';
}
tagProtected = false;
} else if ("ph".equals(tag.getTag())) {
tagHandler.startOTHER();
tagIndex = tagHandler.endOTHER();
shortcutLetter = calcTagShortcutLetter(tag, ignoreTypeForPhTags);
shortcut =
"<"
+ (shortcutLetter != 0 ? String.valueOf(Character.toChars(shortcutLetter)) : 'f')
+ tagIndex
+ "/>";
tagProtected = false;
} else if ("mrk".equals(tag.getTag())) {
tagHandler.startOTHER();
tagIndex = tagHandler.endOTHER();
shortcutLetter = 'm';
shortcut =
"<m"
+ tagIndex
+ ">"
+ tag.getIntactContents().sourceToOriginal()
+ "</m"
+ tagIndex
+ ">";
tagProtected = true;
} else {
shortcutLetter = 'f';
tagIndex = -1;
tagProtected = false;
}
tag.setShortcutLetter(shortcutLetter);
tag.setShortcutIndex(tagIndex);
tag.setShortcut(shortcut);
r.append(shortcut);
ProtectedPart pp = new ProtectedPart();
pp.setTextInSourceSegment(shortcut);
pp.setDetailsFromSourceFile(tag.toOriginal());
if (tagProtected) {
// protected text with related tags, like <m0>Acme</m0>
if (StatisticsSettings.isCountingProtectedText()) {
// Protected texts are counted, but related tags are not counted in the word count
pp.setReplacementWordsCountCalculation(
StaticUtils.TAG_REPLACEMENT
+ tag.getIntactContents().sourceToOriginal()
+ StaticUtils.TAG_REPLACEMENT);
} else {
// All protected parts are not counted in the word count(default)
pp.setReplacementWordsCountCalculation(StaticUtils.TAG_REPLACEMENT);
}
pp.setReplacementUniquenessCalculation(StaticUtils.TAG_REPLACEMENT);
pp.setReplacementMatchCalculation(tag.getIntactContents().sourceToOriginal());
} else {
// simple tag, like <i0>
if (StatisticsSettings.isCountingStandardTags()) {
pp.setReplacementWordsCountCalculation(tag.toSafeCalcShortcut());
} else {
pp.setReplacementWordsCountCalculation(StaticUtils.TAG_REPLACEMENT);
}
pp.setReplacementUniquenessCalculation(StaticUtils.TAG_REPLACEMENT);
pp.setReplacementMatchCalculation(StaticUtils.TAG_REPLACEMENT);
}
protectedParts.add(pp);
} else if (el instanceof Tag) {
Tag tag = (Tag) el;
int tagIndex = tagHandler.paired(tag.getTag(), tag.getType());
tag.setIndex(tagIndex);
String shortcut = tag.toShortcut();
r.append(shortcut);
ProtectedPart pp = new ProtectedPart();
pp.setTextInSourceSegment(shortcut);
pp.setDetailsFromSourceFile(tag.toOriginal());
if (StatisticsSettings.isCountingStandardTags()) {
pp.setReplacementWordsCountCalculation(tag.toSafeCalcShortcut());
} else {
pp.setReplacementWordsCountCalculation(StaticUtils.TAG_REPLACEMENT);
}
pp.setReplacementUniquenessCalculation(StaticUtils.TAG_REPLACEMENT);
pp.setReplacementMatchCalculation(StaticUtils.TAG_REPLACEMENT);
protectedParts.add(pp);
} else {
r.append(el.toShortcut());
}
}
return r.toString();
}