private bool TryGetTwficLowercaseSentenceInitialForm(int hvoAnnotation, out int hvoWfiLower)
{
hvoWfiLower = 0;
int ktagMatchingLowercaseForm = InterlinVc.MatchingLowercaseWordForm(Cache);
if (Cache.MainCacheAccessor.get_IsPropInCache(hvoAnnotation, ktagMatchingLowercaseForm, (int)CellarModuleDefns.kcptReferenceAtom, 0))
hvoWfiLower = Cache.GetObjProperty(hvoAnnotation, ktagMatchingLowercaseForm);
else
{
StTxtPara.TwficInfo twficInfo = new StTxtPara.TwficInfo(m_fdoCache, hvoAnnotation);
if (twficInfo.IsFirstTwficInSegment)
{
ITsString tssWfBaseline = StTxtPara.TssSubstring(twficInfo.Object);
CpeTracker tracker = new CpeTracker(Cache.LanguageWritingSystemFactoryAccessor, tssWfBaseline);
ILgCharacterPropertyEngine cpe = tracker.CharPropEngine(0);
string sLower = cpe.ToLower(tssWfBaseline.Text);
hvoWfiLower = Cache.LangProject.WordformInventoryOA.GetWordformId(sLower,
StringUtils.GetWsAtOffset(tssWfBaseline, 0));
}
}
return hvoWfiLower != 0;
}
public SegmentBreaker(ITsString text, ILgWritingSystemFactory wsf)
{
tssText = text;
m_cpeTracker = new CpeTracker(wsf, text);
// Make sure this is always something.
m_cpe = m_cpeTracker.CharPropEngine(0);
}
/// ------------------------------------------------------------------------------------
/// <summary>
/// Determine if the specified text exists in the string
/// </summary>
/// <param name="wordFormTss">text to look for</param>
/// <param name="sourceTss">text to search in</param>
/// <param name="wsf">source of char prop engines</param>
/// <param name="fMatchWholeWord">True to match a whole word, false otherwise</param>
/// <param name="ichMin">The start of the string where the text was found (undefined if
/// this method returns false)</param>
/// <param name="ichLim">The limit (one character position past the end) of the string
/// where the text was found (undefined if this method returns false)</param>
/// <returns><c>true</c> if the text is matched exactly (matching case)
/// </returns>
/// <remarks>
/// TODO: Add a parameter to make it possible to do a case-insensitive match?</remarks>
/// ------------------------------------------------------------------------------------
public static bool FindTextInString(ITsString wordFormTss, ITsString sourceTss,
ILgWritingSystemFactory wsf, bool fMatchWholeWord, out int ichMin, out int ichLim)
{
ichMin = ichLim = 0;
if (wordFormTss == null || sourceTss == null || wordFormTss.Length == 0 ||
sourceTss.Length == 0)
{
// Nothing we can really do
return false;
}
int ichWordForm = 0;
bool fMatchInProgress = false;
CpeTracker sourceTracker = new CpeTracker(wsf, sourceTss);
CpeTracker wordTracker = new CpeTracker(wsf, wordFormTss);
string wordForm = wordFormTss.Text;
string sourceText = sourceTss.Text;
bool fWordFormAndSrcAreBothNormalized = wordForm.IsNormalized() && sourceText.IsNormalized();
bool fPrevCharWasWordForming = false;
// Must use local temp variable because some callers pass same variable as both
// ichMin and ichLim.
int ichLimT;
for (int ichSrc = 0; ichSrc < sourceText.Length; ichSrc = ichLimT)
{
ichLimT = ichSrc + 1;
bool fWordForming = sourceTracker.CharPropEngine(ichSrc).get_IsWordForming(sourceText[ichSrc]);
if (!fMatchInProgress && fPrevCharWasWordForming && fWordForming && fMatchWholeWord)
continue;
fPrevCharWasWordForming = fWordForming;
if (!fMatchInProgress && !fWordForming)
continue;
bool fMatch = (wordForm[ichWordForm] == sourceText[ichSrc]);
if (fMatch)
{
ichWordForm++;
}
else
{
if (sourceText[ichSrc] == kchObject)
fMatch = true;
else if (!fWordFormAndSrcAreBothNormalized)
{
int ichWfCharLim = ichWordForm + 1;
while (ichWfCharLim < wordForm.Length &&
!wordTracker.CharPropEngine(ichWfCharLim).get_IsLetter(wordForm[ichWfCharLim]) &&
wordTracker.CharPropEngine(ichWfCharLim).get_IsWordForming(wordForm[ichWfCharLim]))
{
ichWfCharLim++;
}
while (ichLimT < sourceText.Length &&
!sourceTracker.CharPropEngine(ichLimT).get_IsLetter(sourceText[ichLimT]) &&
sourceTracker.CharPropEngine(ichLimT).get_IsWordForming(sourceText[ichLimT]))
{
ichLimT++;
}
if (wordForm.Substring(ichWordForm, ichWfCharLim - ichWordForm).Normalize() ==
sourceText.Substring(ichSrc, ichLimT - ichSrc).Normalize())
{
ichWordForm = ichWfCharLim;
fMatch = true;
}
}
}
if (fMatch)
{
// If this character is the start of a match but the previous character was
// word-forming, then this is a bogus match, so we keep looking.
if (!fMatchInProgress)
{
if (ichSrc == 0 || !sourceTracker.CharPropEngine(ichSrc - 1).get_IsWordForming(sourceText[ichSrc - 1]) || !fMatchWholeWord)
{
ichMin = ichSrc;
fMatchInProgress = true;
}
else
ichWordForm = 0;
}
if (fMatchInProgress && ichWordForm == wordForm.Length)
{
if (++ichSrc < sourceText.Length && sourceTracker.CharPropEngine(ichSrc).get_IsWordForming(sourceText[ichSrc]) &&
fMatchWholeWord)
{
ichWordForm = 0;
fMatchInProgress = false;
}
else
{
ichLim = ichLimT;
return true;
}
}
}
else
{
ichWordForm = 0;
fMatchInProgress = false;
}
}
return false;
}
/// <summary>
/// Start it off analyzing a string.
/// </summary>
/// <param name="tss"></param>
/// <param name="cpe">engine to use.</param>
public WordMaker(ITsString tss, ILgWritingSystemFactory wsf)
{
Init(tss);
m_tracker = new CpeTracker(wsf, tss);
m_cpe = m_tracker.CharPropEngine(0); // a default for functions that don't depend on wordforming.
}
