/// <summary>
/// Test the ToLower function.
/// Enhance JohnT: should ideally test the case where output is > 10 characters longer than
/// input. However, I have not yet been able to find a Unicode character that is IN FACT
/// longer when converted to lower case.
/// </summary>
public void TestToLower()
{
Assert.AreEqual("abc", Icu.ToLower("ABC", "en"));
Assert.AreEqual("abc", Icu.ToLower("abc", "en"));
Assert.AreEqual("abc", Icu.ToLower("Abc", "en"));
Assert.AreEqual(";,.", Icu.ToLower(";,.", "en"));
}
/// <summary>
/// This guess factors in the placement of an occurrence in its segment for making other
/// decisions like matching lowercase alternatives for sentence initial occurrences.
/// </summary>
public IAnalysis GetBestGuess(AnalysisOccurrence occurrence)
{
// first see if we can make a guess based on the lowercase form of a sentence initial (non-lowercase) wordform
// TODO: make it look for the first word in the sentence...may not be at Index 0!
if (occurrence.Analysis is IWfiWordform && occurrence.Index == 0)
{
ITsString tssWfBaseline = occurrence.BaselineText;
CoreWritingSystemDefinition ws = Cache.ServiceLocator.WritingSystemManager.Get(tssWfBaseline.get_WritingSystemAt(0));
string sLower = Icu.ToLower(tssWfBaseline.Text, ws.IcuLocale);
// don't bother looking up the lowercased wordform if the instanceOf is already in lowercase form.
if (sLower != tssWfBaseline.Text)
{
ITsString tssLower = TsStringUtils.MakeString(sLower, TsStringUtils.GetWsAtOffset(tssWfBaseline, 0));
IWfiWordform lowercaseWf;
if (Cache.ServiceLocator.GetInstance <IWfiWordformRepository>().TryGetObject(tssLower, out lowercaseWf))
{
IAnalysis bestGuess;
if (TryGetBestGuess(lowercaseWf, occurrence.BaselineWs, out bestGuess))
{
return(bestGuess);
}
}
}
}
if (occurrence.BaselineWs == -1)
{
return(null); // happens with empty translation lines
}
return(GetBestGuess(occurrence.Analysis, occurrence.BaselineWs));
}
internal static void DisplayEntries(FdoCache cache, IWin32Window owner, Mediator mediatorIn,
IHelpTopicProvider helpProvider, string helpFileKey, ITsString tssWfIn)
{
ITsString tssWf = tssWfIn;
List <int> rghvo = LexEntryUi.FindEntriesForWordform(cache, tssWf);
// if we do not find a match for the word then try converting it to lowercase and see if there
// is an entry in the lexicon for the Wordform in lowercase. This is needed for occurences of
// words which are capitalized at the beginning of sentences. LT-7444 RickM
if (rghvo == null || rghvo.Count == 0)
{
//We need to be careful when converting to lowercase therefore use Icu.ToLower()
//get the WS of the tsString
int wsWf = StringUtils.GetWsAtOffset(tssWf, 0);
//use that to get the locale for the WS, which is used for
string wsLocale = cache.LanguageWritingSystemFactoryAccessor.get_EngineOrNull(wsWf).IcuLocale;
string sLower = Icu.ToLower(tssWf.Text, wsLocale);
ITsTextProps ttp = tssWf.get_PropertiesAt(0);
ITsStrFactory tsf = TsStrFactoryClass.Create();
tssWf = tsf.MakeStringWithPropsRgch(sLower, sLower.Length, ttp);
rghvo = LexEntryUi.FindEntriesForWordform(cache, tssWf);
}
StringTable stOrig;
Mediator mediator;
IVwStylesheet styleSheet;
bool fRestore = EnsureFlexTypeSetup(cache, mediatorIn, out stOrig, out mediator, out styleSheet);
if (rghvo == null || rghvo.Count == 0)
{
int hvoLe = ShowFindEntryDialog(cache, mediator, tssWf, owner);
if (hvoLe == 0)
{
// Restore the original string table in the mediator if needed.
if (fRestore)
{
mediator.StringTbl = stOrig;
}
return;
}
rghvo = new List <int>(1);
rghvo.Add(hvoLe);
}
using (SummaryDialogForm form =
new SummaryDialogForm(rghvo, tssWf, helpProvider, helpFileKey, styleSheet, cache, mediator))
{
form.ShowDialog(owner);
if (form.ShouldLink)
{
form.LinkToLexicon();
}
}
// Restore the original string table in the mediator if needed.
if (fRestore)
{
mediator.StringTbl = stOrig;
}
}
/// <summary>
/// Get the lead character, either a single character or a composite matching something
/// in the sort rules. (We need to support multi-graph letters. See LT-9244.)
/// </summary>
private string GetLeadChar(string sEntryNFD, string sWs)
{
if (string.IsNullOrEmpty(sEntryNFD))
{
return("");
}
string sEntry = Icu.ToLower(sEntryNFD, sWs);
Dictionary <string, string> mapChars;
Set <string> sortChars = GetDigraphs(sWs, out mapChars);
string sEntryT = sEntry;
bool fChanged = false;
do
{
foreach (string key in mapChars.Keys)
{
sEntry = sEntry.Replace(key, mapChars[key]);
}
fChanged = sEntryT != sEntry;
sEntryT = sEntry;
} while (fChanged);
string sFirst = sEntry.Substring(0, 1);
foreach (string sChar in sortChars)
{
if (sEntry.StartsWith(sChar))
{
if (sFirst.Length < sChar.Length)
{
sFirst = sChar;
}
}
}
// We don't want sFirst for an ignored first character or digraph.
ILgCollatingEngine lce = LgIcuCollatorClass.Create();
lce.Open(sWs);
byte[] ka = (byte[])lce.get_SortKeyVariant(sFirst, LgCollatingOptions.fcoDefault);
if (ka[0] == 1)
{
string sT = sEntry.Substring(sFirst.Length);
return(GetLeadChar(sT, sWs));
}
return(sFirst);
}
/// <summary>
/// Convert string to lower case equivalent.
/// </summary>
/// <param name="input"></param>
/// <returns></returns>
public string ToLower(string input)
{
return(Icu.ToLower(input, m_icuLocale));
}
/// <summary>
/// Get the set of significant digraphs (multigraphs) for the writing system. At the
/// moment, these are derived from ICU sorting rules associated with the writing system.
/// </summary>
private Set <string> GetDigraphs(string sWs, out Dictionary <string, string> mapChars)
{
Set <string> digraphs = null;
if (m_mapWsDigraphs.TryGetValue(sWs, out digraphs))
{
mapChars = m_mapWsMapChars[sWs];
return(digraphs);
}
digraphs = new Set <string>();
mapChars = new Dictionary <string, string>();
int ws = m_cache.LanguageWritingSystemFactoryAccessor.GetWsFromStr(sWs);
IWritingSystem wsX = null;
ICollation coll = null;
string sIcuRules = null;
if (ws > 0)
{
wsX = m_cache.LanguageWritingSystemFactoryAccessor.get_EngineOrNull(ws);
if (wsX.CollationCount > 0)
{
coll = wsX.get_Collation(0);
sIcuRules = coll.IcuRules;
if (String.IsNullOrEmpty(sIcuRules))
{
// The ICU rules may not be loaded for built-in languages, but are
// still helpful for our purposes here.
string sIcuOrig = sIcuRules;
coll.LoadIcuRules(sWs);
sIcuRules = coll.IcuRules;
coll.IcuRules = sIcuOrig; // but we don't want to actually change anything!
}
}
}
if (!String.IsNullOrEmpty(sIcuRules) && sIcuRules.Contains("&"))
{
string[] rgsRules = sIcuRules.Split(new char[] { '&' }, StringSplitOptions.RemoveEmptyEntries);
for (int i = 0; i < rgsRules.Length; ++i)
{
string sRule = rgsRules[i];
// This is a valid rule that specifies that the digraph aa should be ignored
// [last tertiary ignorable] = \u02bc = aa
// but the code here will ignore this. YAGNI the chances of a user specifying a digraph
// as ignorable may never happen.
if (sRule.Contains("["))
{
sRule = sRule.Substring(0, sRule.IndexOf("["));
}
if (String.IsNullOrEmpty(sRule.Trim()))
{
continue;
}
sRule = sRule.Replace("<<<", "=");
sRule = sRule.Replace("<<", "=");
if (sRule.Contains("<"))
{
// "&N<ng<<<Ng<ny<<<Ny" => "&N<ng=Ng<ny=Ny"
// "&N<ñ<<<Ñ" => "&N<ñ=Ñ"
// There are other issues we are not handling proplerly such as the next line
// &N<\u006e\u0067
string[] rgsPieces = sRule.Split(new char[] { '<', '=' }, StringSplitOptions.RemoveEmptyEntries);
for (int j = 0; j < rgsPieces.Length; ++j)
{
string sGraph = rgsPieces[j];
sGraph = sGraph.Trim();
if (String.IsNullOrEmpty(sGraph))
{
continue;
}
sGraph = Icu.Normalize(sGraph, Icu.UNormalizationMode.UNORM_NFD);
if (sGraph.Length > 1)
{
sGraph = Icu.ToLower(sGraph, sWs);
if (!digraphs.Contains(sGraph))
{
digraphs.Add(sGraph);
}
}
}
}
else if (sRule.Contains("="))
{
// "&ae<<æ<<<Æ" => "&ae=æ=Æ"
string[] rgsPieces = sRule.Split(new char[] { '=' }, StringSplitOptions.RemoveEmptyEntries);
string sGraphPrimary = rgsPieces[0].Trim();
Debug.Assert(!String.IsNullOrEmpty(sGraphPrimary));
sGraphPrimary = Icu.ToLower(sGraphPrimary, sWs);
for (int j = 1; j < rgsPieces.Length; ++j)
{
string sGraph = rgsPieces[j];
sGraph = sGraph.Trim();
if (String.IsNullOrEmpty(sGraph))
{
continue;
}
sGraph = Icu.Normalize(sGraph, Icu.UNormalizationMode.UNORM_NFD);
sGraph = Icu.ToLower(sGraph, sWs);
if (sGraph != sGraphPrimary)
{
if (!mapChars.ContainsKey(sGraph))
{
mapChars.Add(sGraph, sGraphPrimary);
}
}
}
}
}
}
m_mapWsDigraphs.Add(sWs, digraphs);
m_mapWsMapChars.Add(sWs, mapChars);
return(digraphs);
}
internal static void DisplayEntry(FdoCache cache, IWin32Window owner, Mediator mediatorIn,
IHelpTopicProvider helpProvider, string helpFileKey, ITsString tssWfIn)
{
ITsString tssWf = tssWfIn;
LexEntryUi leui = null;
Mediator mediator = null;
try
{
leui = FindEntryForWordform(cache, tssWf);
// if we do not find a match for the word then try converting it to lowercase and see if there
// is an entry in the lexicon for the Wordform in lowercase. This is needed for occurences of
// words which are capitalized at the beginning of sentences. LT-7444 RickM
if (leui == null)
{
//We need to be careful when converting to lowercase therefore use Icu.ToLower()
//get the WS of the tsString
int wsWf = TsStringUtils.GetWsAtOffset(tssWf, 0);
//use that to get the locale for the WS, which is used for
string wsLocale = cache.ServiceLocator.WritingSystemManager.Get(wsWf).IcuLocale;
string sLower = Icu.ToLower(tssWf.Text, wsLocale);
ITsTextProps ttp = tssWf.get_PropertiesAt(0);
tssWf = cache.TsStrFactory.MakeStringWithPropsRgch(sLower, sLower.Length, ttp);
leui = FindEntryForWordform(cache, tssWf);
}
// Ensure that we have a valid mediator with the proper string table.
bool fRestore;
StringTable stOrig;
mediator = EnsureValidMediator(mediatorIn, out fRestore, out stOrig);
FdoCache cache2 = (FdoCache)mediator.PropertyTable.GetValue("cache");
if (cache2 != cache)
{
mediator.PropertyTable.SetProperty("cache", cache);
}
EnsureWindowConfiguration(mediator);
IVwStylesheet styleSheet = GetStyleSheet(cache, mediator);
if (leui == null)
{
ILexEntry entry = ShowFindEntryDialog(cache, mediator, tssWf, owner);
if (entry == null)
{
// Restore the original string table in the mediator if needed.
if (fRestore)
{
mediator.StringTbl = stOrig;
}
return;
}
leui = new LexEntryUi(entry);
}
if (mediator != null)
{
leui.Mediator = mediator;
}
leui.ShowSummaryDialog(owner, tssWf, helpProvider, helpFileKey, styleSheet);
// Restore the original string table in the mediator if needed.
if (fRestore)
{
mediator.StringTbl = stOrig;
}
}
finally
{
if (leui != null)
{
leui.Dispose();
}
if (mediator != mediatorIn)
{
mediator.Dispose();
}
}
}