private static void TestGetCorrectTerm(CSpellApi cSpellApi)
{
// init
// all lowerCase
string inText = "hotflashes";
// test process:
TokenObj inToken = new TokenObj(inText);
TokenObj outToken = NonWordCorrector.GetCorrectTerm(inToken, cSpellApi);
// result
string outText = outToken.GetTokenStr();
// print out
Console.WriteLine("--------- GetCorrectTerm( ) -----------");
Console.WriteLine("In: [" + inText + "]");
Console.WriteLine("Out: [" + outText + "]");
}
// public method
// Use: for loop, the latest and greatest implementation
// original implementation with for loop, To be deleted
// the core of spell-correction, include split
// inTokenList is the whole text
public static List <TokenObj> Process(List <TokenObj> inTokenList, CSpellApi cSpellApi, bool debugFlag)
{
DebugPrint.PrintProcess("3-4. NonWord-Split & 1To1", debugFlag);
DebugPrint.PrintInText(TextObj.TokenListToText(inTokenList), debugFlag);
// init the output TokenList
List <TokenObj> outTokenList = new List <TokenObj>();
// process: go through each token for detection and correction
// for the 1-to-1 and split correction
int tarPos = 0; // the position of the tokenObj in the inTokenList
// remove space token from the list
List <TokenObj> nonSpaceTokenList = TextObj.GetNonSpaceTokenObjList(inTokenList);
// use the inTokenList to keep the same spcae token
TokenObj outTokenObj = null;
int maxLegitTokenLength = cSpellApi.GetMaxLegitTokenLength();
foreach (TokenObj tokenObj in inTokenList)
{
/// <summary>
/// no context
/// TokenObj outTokenObj = SpellCorrector.GetCorrectTerm(tokenObj,
/// cSpellApi, debugFlag);
///
/// </summary>
// skip empty space tokens and long tokens
// SCR-3, use legit token
if (tokenObj.IsLegitToken(maxLegitTokenLength) == true)
{
// correct term is the highest ranked candidate
outTokenObj = NonWordCorrector.GetCorrectTerm(tokenObj, cSpellApi, debugFlag, tarPos, nonSpaceTokenList);
// used tarPos for context module
tarPos++;
}
else
{
outTokenObj = tokenObj;
}
// add the corrected tokenObj to the output token list
// use FlatMap because there might be a split
Split1To1Corrector.AddSplit1To1Correction(outTokenList, outTokenObj);
}
return(outTokenList);
}
