organicmaps/icu
13
Fork 0
mirror of https://github.com/unicode-org/icu.git synced 2025-04-17 02:37:25 +00:00
Code Issues 1 Projects Releases Wiki Activity Actions 8

ICU-965 in Any-Title make can't -> Can't, not Can'T

Browse source 
X-SVN-Rev: 5144
This commit is contained in:
Alan Liu 2001-06-29 21:19:49 +00:00
parent 3225d6b8fe
commit be55a412be
2 changed files with 20 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

21
icu4c/source/i18n/titletrn.cpp
View file

@ -18,7 +18,8 @@ const char* TitlecaseTransliterator::_ID = "Any-Title";
TitlecaseTransliterator::TitlecaseTransliterator(UnicodeFilter* adoptedFilter) :
Transliterator(_ID, adoptedFilter) {
setMaximumContextLength(1);
// Need to look back 2 characters in the case of "can't"
setMaximumContextLength(2);
}
/**
@ -63,12 +64,23 @@ void TitlecaseTransliterator::handleTransliterate(
// don't filter characters in the range contextStart..start-1
// (the left context).
// NOTE: This method contains some special case code to handle
// apostrophes between alpha characters. We want to have
// "can't" => "Can't" (not "Can'T"). This may be incorrect
// for some locales, e.g., "l'arbre" => "L'Arbre" (?).
// TODO: Revisit this.
// Determine if there is a preceding letter character in the
// left context (if there is any left context).
UBool wasLastCharALetter = FALSE;
if (offsets.start > offsets.contextStart) {
wasLastCharALetter =
u_isalpha(text.charAt(offsets.start - 1));
UChar c = text.charAt(offsets.start - 1);
// Handle the case "Can'|t", where the | marks the context
// boundary. We only handle a single apostrophe.
if (c == 0x0027 /*'*/ && (offsets.start-2) >= offsets.contextStart) {
c = text.charAt(offsets.start - 2);
}
wasLastCharALetter = u_isalpha(c);
}
// The buffer used to batch up changes to be made
@ -116,6 +128,9 @@ void TitlecaseTransliterator::handleTransliterate(
buffer.append(newChar);
}
wasLastCharALetter = TRUE;
} else if (c == 0x0027 /*'*/ && wasLastCharALetter) {
// Ignore a single embedded apostrophe, so that "can't" =>
// "Can't", not "Can'T".
} else {
wasLastCharALetter = FALSE;
}

4
icu4c/source/test/intltest/transtst.cpp
View file

@ -977,8 +977,8 @@ void TransliteratorTest::TestCaseMap(void) {
"THE QUICK BROWN FOx JUMPED OVER THE LAzy DOGS.");
expect(*toLower, "The quIck brown fOX jUMPED OVER THE LAzY dogs.",
"the quick brown foX jumped over the lazY dogs.");
expect(*toTitle, "the quick brown foX jumped over the laZy dogs.",
"The Quick Brown FoX Jumped Over The LaZy Dogs.");
expect(*toTitle, "the quick brown foX can't jump over the laZy dogs.",
"The Quick Brown FoX Can't Jump Over The LaZy Dogs.");
delete toUpper;
delete toLower;