void JoinMatch(int start, int end,int index,bool inpattern)
{
// concatenate the match value
bool started = false;
for (int i = start; i <= end; ++i)
{
if (i < 1 || i > wordCount) continue; // ignore off end
if (unmarked[i]) continue; // ignore words masked
char* word = wordStarts[i];
if (started)
{
strcat(wildcardOriginalText[index], wildcardSeparator);
strcat(wildcardCanonicalText[index], wildcardSeparator);
}
else started = true;
strcat(wildcardOriginalText[index], word);
if (wordCanonical[i]) strcat(wildcardCanonicalText[index], wordCanonical[i]);
else strcat(wildcardCanonicalText[index], word);
}
if (strstr(wildcardCanonicalText[index], "unknown-word")) strcpy(wildcardCanonicalText[index], "unknown-word"); // if any are unknown, the composite is unknown
// force a particular case? do we know the word?
int lookup = STANDARD_LOOKUP;
if (uppercaseFind > 0)
lookup = UPPERCASE_LOOKUP;
else if (uppercaseFind == 0) lookup = LOWERCASE_LOOKUP;
WORDP D = FindWord(wildcardCanonicalText[index], 0, lookup);
if (D) strcpy(wildcardCanonicalText[index], D->word); // but may not be found if original has plural or such or if uses _
D = FindWord(wildcardOriginalText[index], 0, lookup);
if (D) strcpy(wildcardOriginalText[index], D->word); // but may not be found if original has plural or such or if uses _
if (trace & TRACE_OUTPUT && !inpattern && CheckTopicTrace()) Log(STDTRACELOG, (char*)"_%d=%s/%s ", index, wildcardOriginalText[index], wildcardCanonicalText[index]);
}
static char* SpellCheck( int i, int language)
{
// on entry we will have passed over words which are KnownWord (including bases) or isInitialWord (all initials)
// wordstarts from 1 ... wordCount is the incoming sentence words (original). We are processing the ith word here.
char* word = wordStarts[i];
if (!*word) return NULL;
if (!stricmp(word,loginID) || !stricmp(word,computerID)) return word; // dont change his/our name ever
size_t len = strlen(word);
if (len > 2 && word[len-2] == '\'') return word; // dont do anything with ' words
// test for run togetherness like "talkabout fingers"
int breakAt = SplitWord(word);
if (breakAt > 0)// we found a split, insert 2nd word into word stream
{
char* tokens[3];
WORDP D = FindWord(word,breakAt,PRIMARY_CASE_ALLOWED);
tokens[1] = D->word;
tokens[2] = word+breakAt;
ReplaceWords(i,1,2,tokens);
fixedSpell = true;
return NULL;
}
// now imagine partial runtogetherness, like "talkab out fingers"
if (i < wordCount)
{
char tmp[MAX_WORD_SIZE];
strcpy(tmp,word);
strcat(tmp,wordStarts[i+1]);
breakAt = SplitWord(tmp);
if (breakAt > 0) // replace words with the dual pair
{
char* tokens[3];
WORDP D = FindWord(tmp,breakAt,PRIMARY_CASE_ALLOWED);
tokens[1] = D->word;
tokens[2] = tmp+breakAt;
ReplaceWords(i,2,2,tokens);
fixedSpell = true;
return NULL;
}
}
// remove any nondigit characters repeated more than once. Dont do this earlier, we want substitutions to have a chance at it first. ammmmmmazing
static char word1[MAX_WORD_SIZE];
char* ptr = word-1;
char* ptr1 = word1;
while (*++ptr)
{
*ptr1 = *ptr;
while (ptr[1] == *ptr1 && ptr[2] == *ptr1 && (*ptr1 < '0' || *ptr1 > '9')) ++ptr; // skip double repeats
++ptr1;
}
*ptr1 = 0;
if (FindCanonical(word1,0,true) && !IsUpperCase(*word1)) return word1; // this is a different form of a canonical word so its ok
// now use word spell checker
char* d = SpellFix(word,i,PART_OF_SPEECH,language);
return (d) ? d : NULL;
}
static int SplitWord(char* word)
{
WORDP D2;
bool good;
int breakAt = 0;
if (IsDigit(*word))
{
while (IsDigit(word[++breakAt]) || word[breakAt] == '.'){;} // find end of number
if (word[breakAt]) // found end of number
{
D2 = FindWord(word+breakAt,0,PRIMARY_CASE_ALLOWED);
if (D2)
{
good = (D2->properties & (PART_OF_SPEECH|FOREIGN_WORD)) != 0 || (D2->internalBits & HAS_SUBSTITUTE) != 0;
if (good && (D2->systemFlags & AGE_LEARNED))// must be common words we find
{
char number[MAX_WORD_SIZE];
strncpy(number,word,breakAt);
number[breakAt] = 0;
StoreWord(number,ADJECTIVE|NOUN|ADJECTIVE_NUMBER|NOUN_NUMBER);
return breakAt; // split here
}
}
}
}
// try all combinations of breaking the word into two known words
breakAt = 0;
size_t len = strlen(word);
for (unsigned int k = 1; k < len-1; ++k)
{
if (!stricmp(language,"english") && k == 1 && *word != 'a' && *word != 'A' && *word != 'i' && *word != 'I') continue; // only a and i are allowed single-letter words
else if (!stricmp(language,"french") && k == 1 && *word != 'y' && *word != 'a' && *word != 'A' && !SameUTF(word,"à") && !SameUTF(word, "À") && !SameUTF(word, "ô") && !SameUTF(word,"Ô")) continue; // in french only y, a and ô are allowed single-letter words
WORDP D1 = FindWord(word,k,PRIMARY_CASE_ALLOWED);
if (!D1) continue;
good = (D1->properties & (PART_OF_SPEECH|FOREIGN_WORD)) != 0 || (D1->internalBits & HAS_SUBSTITUTE) != 0;
if (!good || !(D1->systemFlags & AGE_LEARNED)) continue; // must be normal common words we find
D2 = FindWord(word+k,len-k,PRIMARY_CASE_ALLOWED);
if (!D2) continue;
good = (D2->properties & (PART_OF_SPEECH|FOREIGN_WORD)) != 0 || (D2->internalBits & HAS_SUBSTITUTE) != 0;
if (!good || !(D2->systemFlags & AGE_LEARNED) ) continue; // must be normal common words we find
if (!breakAt) breakAt = k; // found a split
else // found multiple places to split... dont know what to do
{
breakAt = -1;
break;
}
}
return breakAt;
}
static char* SpellCheck(unsigned int i)
{
// on entry we will have passed over words which are KnownWord (including bases) or isInitialWord (all initials)
// wordstarts from 1 ... wordCount is the incoming sentence words (original). We are processing the ith word here.
char* word = wordStarts[i];
if (!*word) return NULL;
if (!stricmp(word,loginID) || !stricmp(word,computerID)) return word; // dont change his/our name ever
size_t len = strlen(word);
if (len > 2 && word[len-2] == '\'') return word; // dont do anything with ' words
// test for run togetherness like "talkabout fingers"
int breakAt = SplitWord(word);
if (breakAt > 0)// we found a split, insert 2nd word into word stream
{
++wordCount;
memmove(wordStarts+i+1,wordStarts+i,sizeof(char*) * (wordCount-i)); // open up a slot for a new word
wordStarts[i+1] = reuseAllocation(wordStarts[i+1],wordStarts[i]+breakAt); // set this to the second word (shared from within 1st word)
return FindWord(wordStarts[i],breakAt,PRIMARY_CASE_ALLOWED)->word; // 1st word gets replaced, we added valid word after
}
// now imagine partial runtogetherness, like "talkab out fingers"
if (i < wordCount)
{
char tmp[MAX_WORD_SIZE];
strcpy(tmp,word);
strcat(tmp,wordStarts[i+1]);
breakAt = SplitWord(tmp);
if (breakAt > 0) // replace words with the dual pair
{
wordStarts[i+1] = reuseAllocation(wordStarts[i+1],StoreWord(tmp+breakAt)->word); // set this to the second word (shared from within 1st word)
return FindWord(tmp,breakAt,PRIMARY_CASE_ALLOWED)->word; // 1st word gets replaced, we added valid word after
}
}
// remove any nondigit characters repeated more than once. Dont do this earlier, we want substitutions to have a chance at it first. ammmmmmazing
static char word1[MAX_WORD_SIZE];
char* ptr = word-1;
char* ptr1 = word1;
while (*++ptr)
{
*ptr1 = *ptr;
while (ptr[1] == *ptr1 && ptr[2] == *ptr1 && (*ptr1 < '0' || *ptr1 > '9')) ++ptr; // skip double repeats
++ptr1;
}
*ptr1 = 0;
if (FindCanonical(word1,0,true) && !IsUpperCase(*word1)) return word1; // this is a different form of a canonical word so its ok
// now use word spell checker
char* d = SpellFix(word,i,PART_OF_SPEECH);
return (d) ? d : NULL;
}
pDICT FindWord(pDICT root, char KeyWord[])
{
if(root == NULL)
return root;
else
{
if(strcmp(KeyWord, root->data.word)<0)
root ->leftlink = FindWord(root->leftlink, KeyWord);
else
{
if(strcmp(KeyWord,root->data.word)>0)
root->rightlink = FindWord(root->rightlink,KeyWord);
else
{
if(root ->leftlink ==NULL && root->rightlink ==NULL)
{
// pDICT temp= NULL;
//temp->leftlink= root;
free(root);
root = NULL;
}
else
{
if(root -> leftlink ==NULL)
{//has right link
pDICT temp = root;
root = root->rightlink;
free(temp);
}
else
{
if(root ->rightlink ==NULL)
{//has left link
pDICT temp = root;
root = root->leftlink;
free(temp);
}
else
{//have two links
pDICT temp = FirstWord(root->rightlink);
root->data = temp->data;
root ->rightlink = FindWord(root->rightlink,temp->data.word);
}
}
}
}
}
}
return root;
}
bool bgParserASE::ReadHelperObject()
{
bool hr = true;
TCHAR szWordArray[2][MAX_PATH * 4];
D3DXVECTOR3 v3Data;
FaceInfo i4Data;
int iData;
int iLoop, iLoopMax;
int iNumObj = m_pModel->m_ObjectList.size();
m_pModel->m_ObjectList.resize(iNumObj + 1);
m_pModel->m_ObjectList[iNumObj].vpObj = new HelperObject;
m_pModel->m_ObjectList[iNumObj].eNodeType = OBJECT_NODE_TYPE_HELPEROBJECT;
IF_FALSE_RETURN(ReadNodeInfo(iNumObj));
IF_FALSE_RETURN(FindWord(_T("*BOUNDINGBOX_MIN")));
_stscanf(m_szLine, _T("%s %f%f%f"), m_szWord, &v3Data.x, &v3Data.z, &v3Data.y);
static_cast<HelperObject*>(m_pModel->m_ObjectList[iNumObj].vpObj)->vBBoxMin = v3Data;
IF_FALSE_RETURN(FindWord(_T("*BOUNDINGBOX_MAX")));
_stscanf(m_szLine, _T("%s %f%f%f"), m_szWord, &v3Data.x, &v3Data.z, &v3Data.y);
static_cast<HelperObject*>(m_pModel->m_ObjectList[iNumObj].vpObj)->vBBoxMax = v3Data;
///// TM_ANIMATION - 애니메이션
_tcscpy(szWordArray[0], _T("*TM_ANIMATION"));
_tcscpy(szWordArray[1], _T("}"));
switch (FindWordArray(szWordArray, 2))
{
case 0: // *TM_ANIMATION 애니메이션이 있다면 ====================================
{
m_pModel->m_ObjectList[iNumObj].bAnim = true;
IF_FALSE_RETURN(ReadTMAnimation(iNumObj));
}
break;
case 1: // } 애니메이션 없이 끝난다면 ===============================
{
m_pModel->m_ObjectList[iNumObj].bAnim = false;
}
break;
case -1: // 찾는 단어 없음 =========================================================
default: // 나머지 (배열 요소가 2개이므로 나올 수 없음)
return false;
break;
}
return hr;
}
//This should find a word inside a phrase, not considering variants of my word (For ex 'tris' won't be found inside 'i love tetris')
int FindWord(char str[], char myWord[], int p)
{
if (strlen(str)-strlen(myWord)<p) return 0;
if (!(((p==0)||(str[p-1]==' '))&&((str[p+strlen(myWord)]==' ')||(str[p+strlen(myWord)]=='\0')))) return FindWord(str, myWord, p+1);
int i;
for (i=0; myWord[i]!='\0'; i++)
{
if (myWord[i]!=str[p+i]) return FindWord(str, myWord, p+1);
}
return 1 + FindWord(str, myWord, p+1);
}
int ProcessExpress(char *instring) /*对原表达式进行预处理*/
{
int len,pos;
inlength=-1;
parencount=0;
tokencount=LASTOPERAND;
len=strlen(instring);
instring[len]='\0';
for (pos=0;pos<len;)
{
if (instring[pos]==' ') /*忽略空格字符*/
pos++;
else if (isalpha(instring[pos])) /*处理字母*/
pos=FindWord(instring,pos);
else if (isdigit(instring[pos]) || instring[pos]=='.') /*处理数字*/
pos=FindNumber(instring,pos);
else /*处理符号*/
pos=FindSymbol(instring,pos);
if (pos==-1)
return 0;
}
if (parencount!=0)
printf("左右括号不匹配\n");
PutToken(0); /*在infix中添加ENDEXPR*/
return 1;
}
void DuminoSolver::Play(const std::vector<std::vector<unsigned char>> &grid,
const Dictionary &dict)
{
for (int i = 0; i < 5; ++i)
{
for (int j = 0; j < 5; ++j)
{
used[i][j] = false;
}
}
found_words = new Dictionary();
vhid = new VirtualHID();
for (int i = 0; i < 5; ++i)
{
for (int j = 0; j < 5; ++j)
{
FindWord(grid, dict.root, cv::Point(j, i), "");
}
}
delete found_words;
delete vhid;
}
static bool FindPhrase(char* word, int start,bool reverse, int & actualStart, int& actualEnd)
{ // Phrases are dynamic, might not be marked, so have to check each word separately. -- faulty in not respecting ignored(unmarked) words
if (start > wordCount) return false;
bool matched = false;
actualEnd = start;
int oldend;
oldend = start = 0; // allowed to match anywhere or only next
unsigned int n = BurstWord(word);
for (int i = 0; i < n; ++i) // use the set of burst words - but "Andy Warhol" might be a SINGLE word.
{
WORDP D = FindWord(GetBurstWord(i));
bool junk;
matched = MatchTest(reverse,D,actualEnd,NULL,NULL,0,junk,actualStart,actualEnd);
if (matched)
{
if (oldend > 0 && actualStart != (oldend + 1)) // do our words match in sequence NO. retry later in sentence
{
++start;
actualStart = actualEnd = start;
i = -1;
oldend = start = 0;
matched = false;
continue;
}
if (i == 0) start = actualStart; // where we matched INITIALLY
oldend = actualEnd;
}
else break;
}
if (matched) actualStart = start;
return matched;
}
int main( void )
{
char *text, *word, *ptr, *wordEnd;
size_t len;
char chr, tmpchr;
puts( "This program search character in a text and print all words, that contain given character." );
puts( "Enter character(a-z, A-Z, 0-9):" );
while( 1 )
{
chr = (char)getchar();
if( IsGraph( tmpchr = (char)getchar() ) ) {
ungetc( tmpchr, stdin );
}
if( IsAlnum( chr ) == false )
{
puts( "Wrong character, try again:" );
}
else
{
break;
}
}
text = Input();
if( text == NULL )
{
puts( "Not enough memory." );
return 1;
}
ptr = text;
while( wordEnd = FindWord( ptr, chr ) )
{
len = 1;
word = wordEnd;
while( IsAlnum( word[-1] ) && word >= text )
{
word--;
len++;
}
word = memcpy( malloc( len + 1 ), word, len );
if( word == NULL )
{
puts( "Memory is not enough" );
break;
}
else
{
word[len] = 0;
printf( "\"%s\" is the word, that contain '%c'\n", word, chr );
free( word );
}
ptr = wordEnd + 1;
}
free( text );
return 0;
}
int main(int argc, const char * argv[])
{
char str[]="Just testing this code out", myWord[]="test";
printf("FindWord:\t %i\n", FindWord(str, myWord, 0));
return 0;
}
bool bgParserASE::ReadScene()
{
bool hr = true;
IF_FALSE_RETURN(FindWord(_T("*SCENE_FIRSTFRAME")));
_stscanf(m_szLine, _T("%s %d"), m_szWord, &m_pModel->m_Scene.iFirstFrame);
IF_FALSE_RETURN(FindWord(_T("*SCENE_LASTFRAME")));
_stscanf(m_szLine, _T("%s %d"), m_szWord, &m_pModel->m_Scene.iLastFrame);
IF_FALSE_RETURN(FindWord(_T("*SCENE_FRAMESPEED")));
_stscanf(m_szLine, _T("%s %d"), m_szWord, &m_pModel->m_Scene.iFrameSpeed);
IF_FALSE_RETURN(FindWord(_T("*SCENE_TICKSPERFRAME")));
_stscanf(m_szLine, _T("%s %d"), m_szWord, &m_pModel->m_Scene.iTicksPerFrame);
IF_FALSE_RETURN(FindWord(_T("}"))); // SCENE 탈출
return hr;
}
void SpellCheck::NextWord(void)
{
if (FindWord(false) == false)
{
ShowWindow(SW_HIDE);
DoneMessage();
}
else
ShowWordFromSelection();
}
static void CompleteWildcard()
{
WORDP D = FindWord(wildcardCanonicalText[wildcardIndex]);
if (D && D->properties & D->internalBits & UPPERCASE_HASH) // but may not be found if original has plural or such or if uses _
{
strcpy(wildcardCanonicalText[wildcardIndex],D->word);
}
++wildcardIndex;
if (wildcardIndex > MAX_WILDCARDS) wildcardIndex = 0;
}
bool bgParserASE::Read()
{
bool hr = true;
TCHAR szWordArray[2][MAX_PATH * 4];
// m_ASE 데이터 초기화
ZeroMemory(&m_pModel->m_Scene, sizeof(SceneInfo));
m_pModel->m_MaterialList.clear();
m_pModel->m_ObjectList.clear();
// 파일 앞부분 1회만 등장하는 섹션 읽기
IF_FALSE_RETURN(FindWord(_T("*SCENE")));
IF_FALSE_RETURN(ReadScene());
IF_FALSE_RETURN(FindWord(_T("*MATERIAL_LIST")));
IF_FALSE_RETURN(ReadMaterial());
// 여러번 등장하는 섹션 반복해서 읽기
_tcscpy(szWordArray[0], _T("*GEOMOBJECT"));
_tcscpy(szWordArray[1], _T("*HELPEROBJECT"));
while (!feof(m_pFile))
{
switch (FindWordArray(szWordArray, 2))
{
case 0: IF_FALSE_RETURN(ReadGeomObject()); break;
case 1: IF_FALSE_RETURN(ReadHelperObject()); break;
case -1: // 찾는 단어 없음 (파일의 끝)
{
ConvertToModel(); // 읽은 데이터를 모델용 데이터로 컨버팅
LinkNode(); // 노드 관계 연결
OperationTM(); // 노드 관계에 따른 행렬 연산
}
break;
default: // 나머지 (배열 요소가 2개이므로 나올 수 없음)
return false;
break;
}
}
return hr;
}
//pDICT FindWord(pDICT root, char key[])
//{
// if (root == NULL)
// return root;
// if(strcmp(key,root->data.word)<0)
// root->leftlink = FindWord(root->leftlink,key);
//
// else if(strcmp(key,root->data.word)>0)
// root->rightlink =FindWord(root->rightlink,key);
// else
// {
// if(root->leftlink==NULL)
// {
// pDICT temp = root ->rightlink;
// free(root);
// return temp;
// }
// else if(root->rightlink ==NULL)
// {
// pDICT temp = root->leftlink;
// free(root);
// return temp;
// }
// pDICT temp = FirstWord(root->rightlink);
// root->data = temp ->data;
// root ->rightlink = FindWord(root->rightlink,temp->data.word);
// }
// return root;
//}
void DeleteWord(pDICT root)
{
pDICT temp;
char Key[1024];
printf("Enter Keyword to delete: " );
gets(Key);
flushall();
temp = FindWord(root,Key);
if(temp == root)
printf("The word has been deleted.\n");
else
printf("No word in dictionary.\n");
}
void ReadComputerID()
{
strcpy(computerID,"anonymous");
WORDP D = FindWord("defaultbot",0); // do we have a FACT with the default bot in it as verb
if (D)
{
FACT* F = GetVerbHead(D);
if (F)
{
D = Meaning2Word(F->subject);
strcpy(computerID,D->word);
}
}
}
/*
* Manual page at function.def
*/
INT16 CGEN_PUBLIC CFunction::OnGet()
{
// Delegate to running function // ------------------------------------
FNC_DELEGATE OnGet(); // Use a weird macro (see function.def)
// Initialize // ------------------------------------
CDlpObject* iCont = GetActiveInstance(); // Determine field container
const char* lpsId = GetNextToken(TRUE); // Determine field name
// Validate // ------------------------------------
DLPASSERT(iCont); // Check set target
if (!dlp_strlen(lpsId)) // If no field name committed
return IERROR(this,FNC_EXPECT,"field identifier after -get",0,0); // Error
SWord* lpWrd = iCont->FindWord(lpsId,WL_TYPE_FIELD); // Find field in container
if (!lpWrd) // If not found
{ // >>
iCont = this; // Use this instance as container
lpWrd = FindWord(lpsId,WL_TYPE_FIELD); // And seek again
} // <<
if (!lpWrd) return IERROR(this,ERR_NOTFIELD,lpsId,0,0); // If still not found --> Error
// Push field value // ------------------------------------
switch (lpWrd->ex.fld.nType) // Branch for field variable type
{ // >>
case T_BOOL : { PushLogic ( *( BOOL*)lpWrd->lpData); break; }// - Boolean
case T_UCHAR : { PushNumber (CMPLX(*( UINT8*)lpWrd->lpData)); break; }// - Unsigned character
case T_CHAR : { PushNumber (CMPLX(*( INT8*)lpWrd->lpData)); break; }// - Signed character
case T_USHORT : { PushNumber (CMPLX(*( UINT16*)lpWrd->lpData)); break; }// - Unsigned short integer
case T_SHORT : { PushNumber (CMPLX(*( INT16*)lpWrd->lpData)); break; }// - Signed short integer
case T_UINT : { PushNumber (CMPLX(*( UINT32*)lpWrd->lpData)); break; }// - Unsigned integer
case T_INT : { PushNumber (CMPLX(*( INT32*)lpWrd->lpData)); break; }// - Signed integer
case T_ULONG : { PushNumber (CMPLX(*( UINT64*)lpWrd->lpData)); break; }// - Unsigned long integer
case T_LONG : { PushNumber (CMPLX(*( INT64*)lpWrd->lpData)); break; }// - Signed long integer
case T_FLOAT : { PushNumber (CMPLX(*( FLOAT32*)lpWrd->lpData)); break; }// - Single precision floating point
case T_DOUBLE : { PushNumber (CMPLX(*( FLOAT64*)lpWrd->lpData)); break; }// - Double precision floating point
case T_COMPLEX : { PushNumber ( *(COMPLEX64*)lpWrd->lpData); break; }// - Double precision complex floating point
case T_INSTANCE: { PushInstance(*(CDlpObject**) lpWrd->lpData); break; }// - Instance
case T_TEXT : /* Fall through */ // - Text (deprecated type!)
case T_CSTRING : /* Fall through */ // - Constant string
case T_STRING : { PushString(*(char**) lpWrd->lpData); break; }// - String
default : { // - Other types
if (lpWrd->ex.fld.nType > 0 && lpWrd->ex.fld.nType <= 256) // Character array?
PushString((char*)lpWrd->lpData); // Push value
else // Type unknown!
DLPASSERT(FMSG("Unknown field type")); // Error
} // <<
} // <<
return O_K; // Done.
}
bool FindWord(char* word)
{
char arr[256];
for (int i = 0; i < 4; i++)
{
for (int j = 0; j < 4; j++)
{
ResetBitMap();
if (FindWord(arr, 256, word, strlen(word), i, j, 0))
{
return true;
}
}
}
}
// loading simple list of words from a file, lemmatizing it, and storing
void CMorphDictionary::LoadFrequentRoots(string path)
{
string load_path = path+"predictroots.txt";
FILE*fp = fopen (load_path.c_str(), "r");
if (!fp) return;
char buffer[1000];
while (fgets (buffer, 1000, fp))
{
string WordStr = buffer;
Trim(WordStr);
if (WordStr.empty()) continue;
RmlMakeUpper (WordStr, m_Language);
vector<CFindWordNode> FindResults;
bool retval = FindWord(WordStr, true, false, false, FindResults);
if (!retval) continue;
set<size_t> UsedFlexia;
for (size_t i=0; i<FindResults.size(); i++)
{
CFormInfo P;
P.Create( this,
FindResults[i].m_nBase,
FindResults[i].m_LemmaInfo,
FindResults[i].m_nFlex
);
for (size_t j=0; j < P.GetCount(); j++)
{
size_t FlexNo = P.GetFlexNoOfForm(j);
if (UsedFlexia.find(FlexNo) == UsedFlexia.end())
{
CPredictEndRoot R;
R.m_BaseNo = FindResults[i].m_nBase;
R.m_LemmaInfo = FindResults[i].m_LemmaInfo;
R.m_EndRoot = P.GetWordForm(j);
reverse(R.m_EndRoot.begin(),R.m_EndRoot.end());
R.m_FlexNo = FlexNo;
UsedFlexia.insert(FlexNo);
m_PredictEndRoots.push_back(R);
};
};
};
};
fclose(fp);
sort(m_PredictEndRoots.begin(), m_PredictEndRoots.end());
};
bool wordBreak(string strtomatch, unordered_set<string> &myset) {
unsigned strlength = strtomatch.length();
BuildMatrix(&pmatrix,strlength,strlength);
for(int i=0;i<strlength;++i){
for(int j=i;j<strlength;++j){
pmatrix[i][j] = false;
}
}
for(unsigned i=0;i<strlength;++i){
for(unsigned j=i;j<strlength;++j){
string strdata;
strdata.assign(strtomatch,i,j-i+1);
if(myset.find(strdata)!=myset.end()){
pmatrix[i][j] = true;
}
}
}
for(int i=strlength-1;i>=0;--i){
bool isallfalse = true;
for(int j = i;j<strlength;++j){
if(pmatrix[i][j]){
isallfalse = false;
break;
}
}
if(isallfalse&&(i-1>=0)){
for(int j = 0;j<strlength;++j){
pmatrix[j][i-1] = false;
}
}
}
bool ret = FindWord(strlength);
ReleaseMatrix(&pmatrix,strtomatch.length());
return ret;
}
void Engine::ExecuteWord(const std::string &word)
{
Word *w = FindWord(word);
if(w != NULL)
{
w->Execute(false);
return;
}
// integer?
std::istringstream is(word);
int number;
if(is >> number)
{
LiteralWord lit(number);
lit.Execute(false);
return;
}
throw std::string("unknown word");
}
//align 1居中 3左对齐
//currentView 10页脚 9页眉
void CmyWord::SetSeekView(const CString& text,long align, long currentView, const CString& findW)
{
Pane mPane;
m_activeWind = m_wdApp.GetActiveWindow();
mPane = m_activeWind.GetActivePane();
m_view = mPane.GetView();
m_view.SetSeekView(currentView);
if(!findW.IsEmpty())
{
FindWord(findW,text);
}
else
{
m_wdSel.TypeText(text);
}
_ParagraphFormat parafm = m_wdSel.GetParagraphFormat();
parafm.SetAlignment(align);
m_wdSel.SetParagraphFormat(parafm);
m_view.SetSeekView(wdSeekMainDocument);
mPane.ReleaseDispatch();
}
void ReadNewUser()
{
userFirstLine = 1;
inputCount = 0;
ClearUserVariables();
ResetTopicSystem();
// set his random seed
bool hasUpperCharacters;
unsigned int rand = (unsigned int) Hashit((unsigned char *) loginID,strlen(loginID),hasUpperCharacters);
char word[MAX_WORD_SIZE];
randIndex = rand & 4095;
sprintf(word,"%d",randIndex);
SetUserVariable("$randindex",word );
strcpy(word,computerID);
word[0] = toUppercaseData[(unsigned char)word[0]];
SetUserVariable("$bot",word );
SetUserVariable("$login",loginName);
sprintf(readBuffer,"^%s",computerID);
WORDP D = FindWord(readBuffer,0,LOWERCASE_LOOKUP);
if (!D)
{
ReportBug("Cannot find bot %s\r\n",computerID);
return;
}
int oldjump = jumpIndex;
char* macroArgumentList = AllocateBuffer();
*macroArgumentList = 0;
globalDepth = 2;
unsigned int result;
currentOutputBase = macroArgumentList;
DoFunction(D->word,macroArgumentList,macroArgumentList,result);
globalDepth = 0;
jumpIndex = oldjump; // return to old error handler
FreeBuffer();
}
/* Add one word to the dictionary.
*/
static int IndexWord(char *Word,int Follows)
{
WORD *thisWord,*lastWord;
int wordIndex;
if(Word != END_SENTENCE)
{
thisWord = FindWord(Word);
if(!thisWord)
{
thisWord = AddWord(Word);
if(!thisWord) Niall_Error("Out of memory.");
}
wordIndex = WordIndex(thisWord);
}
else wordIndex = -1;
lastWord = GetWord(Follows);
if(!lastWord) Niall_Error("Corrupted brain (Can't find last word).");
Associate(lastWord,wordIndex);
return(wordIndex);
}
void DuminoSolver::FindWord(const DuminoGrid &grid,
DictionaryItem* item,
cv::Point p,
std::string word)
{
for (int i = 0; i < 8; ++i)
{
cv::Point new_p = p + nbrs[i];
if(new_p.x < 0 || new_p.x >= 5 || new_p.y < 0 || new_p.y >= 5)
continue;
int index = grid[new_p.y][new_p.x];
if((item->children[index] != NULL) && !used[new_p.y][new_p.x])
{
used[new_p.y][new_p.x] = true;
word.push_back(item->children[index]->value);
if(item->children[index]->endOfWord && word.length() > 2)
{
if(!found_words->ContainsWord(word))
{
found_words->InsertWord(word);
Type(word);
usleep(1000 * 100);
}
}
FindWord(grid, item->children[index], new_p, word);
word.pop_back();
used[new_p.y][new_p.x] = false;
}
}
}
char* SpellFix(char* originalWord,int start,uint64 posflags,int language)
{
size_t len = strlen(originalWord);
if (len >= 100 || len == 0) return NULL;
if (IsDigit(*originalWord)) return NULL; // number-based words and numbers must be treated elsewhere
char letterLow = GetLowercaseData(*originalWord);
char letterHigh = GetUppercaseData(*originalWord);
bool hasUnderscore = (strchr(originalWord,'_')) ? true : false;
bool isUpper = IsUpperCase(originalWord[0]);
if (IsUpperCase(originalWord[1])) isUpper = false; // not if all caps
if (trace == TRACE_SPELLING) Log(STDTRACELOG,(char*)"Spell: %s\r\n",originalWord);
char word[MAX_WORD_SIZE];
MakeLowerCopy(word,originalWord);
// mark positions of the letters and make lower case
char base[257];
memset(base,0,257);
char* ptr = word - 1;
char c;
int position = 0;
while ((c = *++ptr) && position < 255)
{
base[position++ + 1] = GetLowercaseData(c);
}
// Priority is to a word that looks like what the user typed, because the user probably would have noticed if it didnt and changed it. So add/delete has priority over tranform
WORDP choices[4000];
WORDP bestGuess[4000];
unsigned int index = 0;
unsigned int bestGuessindex = 0;
int min = 30;
unsigned char realWordLetterCounts[LETTERMAX];
memset(realWordLetterCounts,0,LETTERMAX);
for (int i = 0; i < (int)len; ++i) ++realWordLetterCounts[(unsigned char)letterIndexData[(unsigned char)word[i]]]; // compute number of each kind of character
uint64 pos = PART_OF_SPEECH; // all pos allowed
WORDP D;
if (posflags == PART_OF_SPEECH && start < wordCount) // see if we can restrict word based on next word
{
D = FindWord(wordStarts[start+1],0,PRIMARY_CASE_ALLOWED);
uint64 flags = (D) ? D->properties : (-1); // if we dont know the word, it could be anything
if (flags & PREPOSITION) pos &= -1 ^ (PREPOSITION|NOUN); // prep cannot be preceeded by noun or prep
if (!(flags & (PREPOSITION|VERB|CONJUNCTION|ADVERB)) && flags & DETERMINER) pos &= -1 ^ (DETERMINER|ADJECTIVE|NOUN|ADJECTIVE_NUMBER|NOUN_NUMBER); // determiner cannot be preceeded by noun determiner adjective
if (!(flags & (PREPOSITION|VERB|CONJUNCTION|DETERMINER|ADVERB)) && flags & ADJECTIVE) pos &= -1 ^ (NOUN);
if (!(flags & (PREPOSITION|NOUN|CONJUNCTION|DETERMINER|ADVERB|ADJECTIVE)) && flags & VERB) pos &= -1 ^ (VERB); // we know all helper verbs we might be
if (D && *D->word == '\'' && D->word[1] == 's' ) pos &= NOUN; // we can only be a noun if possessive - contracted 's should already be removed by now
}
if (posflags == PART_OF_SPEECH && start > 1)
{
D = FindWord(wordStarts[start-1],0,PRIMARY_CASE_ALLOWED);
uint64 flags = (D) ? D->properties : (-1); // if we dont know the word, it could be anything
if (flags & DETERMINER) pos &= -1 ^ (VERB|CONJUNCTION|PREPOSITION|DETERMINER);
}
posflags &= pos; // if pos types are known and restricted and dont match
static int range[] = {0,-1,1,-2,2};
for (unsigned int i = 0; i < 5; ++i)
{
if (language == ENGLISH && i >= 3) break; // only allow +-2 for spanish
MEANING offset = lengthLists[len + range[i]];
if (trace == TRACE_SPELLING) Log(STDTRACELOG,(char*)"\r\n Begin offset %d\r\n",i);
while (offset)
{
D = Meaning2Word(offset);
offset = D->spellNode;
if (PART_OF_SPEECH == posflags && D->systemFlags & PATTERN_WORD){;} // legal generic match
else if (!(D->properties & posflags)) continue; // wrong kind of word
if (*D->word != letterLow && *D->word != letterHigh && language == ENGLISH) continue; // we assume no one misspells starting letter
char* under = strchr(D->word,'_');
// SPELLING lists have no underscore or space words in them
if (hasUnderscore && !under) continue; // require keep any underscore
if (!hasUnderscore && under) continue; // require not have any underscore
if (isUpper && !(D->internalBits & UPPERCASE_HASH) && start != 1) continue; // dont spell check to lower a word in upper
int val = EditDistance(D, D->length, len, base+1,min,realWordLetterCounts,language);
if (val <= min) // as good or better
{
if (val < min)
{
if (trace == TRACE_SPELLING) Log(STDTRACELOG,(char*)" Better: %s against %s value: %d\r\n",D->word,originalWord,val);
index = 0;
min = val;
}
else if ( val == min && trace == TRACE_SPELLING) Log(STDTRACELOG,(char*)" Equal: %s against %s value: %d\r\n",D->word,originalWord,val);
if (!(D->internalBits & BEEN_HERE))
{
choices[index++] = D;
if (index > 3998) break;
AddInternalFlag(D,BEEN_HERE);
}
}
}
}
// try endings ing, s, etc
if (start) // no stem spell if COMING from a stem spell attempt (start == 0)
{
char* stem = StemSpell(word,start);
if (stem)
{
WORDP D = FindWord(stem,0,(tokenControl & ONLY_LOWERCASE) ? PRIMARY_CASE_ALLOWED : STANDARD_LOOKUP);
if (D)
{
for (unsigned int j = 0; j < index; ++j)
{
if (choices[j] == D) // already in our list
{
D = NULL;
break;
}
}
}
if (D) choices[index++] = D;
}
}
if (!index) return NULL;
// take our guesses, and pick the most common (earliest learned or most frequently used) word
uint64 commonmin = 0;
bestGuess[0] = NULL;
for (unsigned int j = 0; j < index; ++j) RemoveInternalFlag(choices[j],BEEN_HERE);
if (index == 1)
{
if (trace == TRACE_SPELLING) Log(STDTRACELOG,(char*)" Single best spell: %s\r\n",choices[0]->word);
return choices[0]->word; // pick the one
}
for (unsigned int j = 0; j < index; ++j)
{
uint64 common = choices[j]->systemFlags & COMMONNESS;
if (common < commonmin) continue;
if (choices[j]->internalBits & UPPERCASE_HASH && index > 1) continue; // ignore proper names for spell better when some other choice exists
if (common > commonmin)
{
commonmin = common;
bestGuessindex = 0;
}
bestGuess[bestGuessindex++] = choices[j];
}
if (bestGuessindex)
{
if (trace == TRACE_SPELLING) Log(STDTRACELOG,(char*)" Pick spell: %s\r\n",bestGuess[0]->word);
return bestGuess[0]->word;
}
return NULL;
}
static char* StemSpell(char* word,unsigned int i)
{
static char word1[MAX_WORD_SIZE];
strcpy(word1,word);
size_t len = strlen(word);
char* ending = NULL;
char* best = NULL;
// suffixes
if (len < 5){;} // too small to have a suffix we care about (suffix == 2 at min)
else if (!strnicmp(word+len-3,(char*)"ing",3))
{
word1[len-3] = 0;
best = SpellFix(word1,0,VERB,ENGLISH);
if (best && FindWord(best,0,LOWERCASE_LOOKUP)) return GetPresentParticiple(best);
}
else if (!strnicmp(word+len-2,(char*)"ed",2))
{
word1[len-2] = 0;
best = SpellFix(word1,0,VERB,ENGLISH);
if (best)
{
char* past = GetPastTense(best);
if (!past) return NULL;
size_t pastlen = strlen(past);
if (past[pastlen-1] == 'd') return past;
ending = "ed";
}
}
else
{
unsigned int i = 0;
char* suffix;
while ((suffix = stems[i].word))
{
uint64 kind = stems[i++].flags;
size_t suffixlen = strlen(suffix);
if (!strnicmp(word+len-suffixlen,suffix,suffixlen))
{
word1[len-suffixlen] = 0;
best = SpellFix(word1,0,kind,ENGLISH);
if (best)
{
ending = suffix;
break;
}
}
}
}
if (!ending && word[len-1] == 's')
{
word1[len-1] = 0;
best = SpellFix(word1,0,VERB|NOUN,ENGLISH);
if (best)
{
WORDP F = FindWord(best,0,(tokenControl & ONLY_LOWERCASE) ? PRIMARY_CASE_ALLOWED : STANDARD_LOOKUP);
if (F && F->properties & NOUN) return GetPluralNoun(F);
ending = "s";
}
}
if (ending)
{
strcpy(word1,best);
strcat(word1,ending);
return word1;
}
return NULL;
}
bool SpellCheckSentence()
{
WORDP D,E;
fixedSpell = false;
bool lowercase = false;
int language = ENGLISH;
char* lang = GetUserVariable((char*)"$cs_language");
if (lang && !stricmp(lang,(char*)"spanish")) language = SPANISH;
// check for all uppercase
for (int i = FindOOBEnd(1) + 1; i <= wordCount; ++i) // skip start of sentence
{
char* word = wordStarts[i];
size_t len = strlen(word);
for (int j = 0; j < (int)len; ++j)
{
if (IsLowerCase(word[j]))
{
lowercase = true;
i = j = 1000;
}
}
}
if (!lowercase && wordCount > 2) // must have several words in uppercase
{
for (int i = FindOOBEnd(1); i <= wordCount; ++i)
{
char* word = wordStarts[i];
MakeLowerCase(word);
}
}
int startWord = FindOOBEnd(1);
for (int i = startWord; i <= wordCount; ++i)
{
char* word = wordStarts[i];
if (!word || !word[1] || *word == '"' ) continue; // illegal or single char or quoted thingy
size_t len = strlen(word);
// dont spell check uppercase not at start or joined word
if (IsUpperCase(word[0]) && (i != startWord || strchr(word,'_')) && tokenControl & NO_PROPER_SPELLCHECK) continue;
// dont spell check email or other things with @ or . in them
if (strchr(word,'@') || strchr(word,'.') || strchr(word,'$')) continue;
// dont spell check names of json objects or arrays
if (!strnicmp(word,"ja-",3) || !strnicmp(word,"jo-",3)) continue;
char* known = ProbableKnownWord(word);
if (known && !strcmp(known,word)) continue; // we know it
if (known && strcmp(known,word))
{
char* tokens[2];
if (!IsUpperCase(*known)) // revised the word to lower case (avoid to upper case like "fields" to "Fields"
{
WORDP D = FindWord(known,0,LOWERCASE_LOOKUP);
if (D)
{
tokens[1] = D->word;
ReplaceWords(i,1,1,tokens);
fixedSpell = true;
continue;
}
}
else // is uppercase a concept member? then revise upwards
{
WORDP D = FindWord(known,0,UPPERCASE_LOOKUP);
if (IsConceptMember(D))
{
tokens[1] = D->word;
ReplaceWords(i,1,1,tokens);
fixedSpell = true;
continue;
}
}
}
char* p = word -1;
unsigned char c;
char* hyphen = 0;
while ((c = *++p) != 0)
{
++len;
if (c == '-') hyphen = p; // note is hyphenated - use trailing
}
if (len == 0 || GetTemperatureLetter(word)) continue; // bad ignore utf word or llegal length - also no composite words
if (c && c != '@' && c != '.') // illegal word character
{
if (IsDigit(word[0]) || len == 1){;} // probable numeric?
// accidental junk on end of word we do know immedately?
else if (i > 1 && !IsAlphaUTF8OrDigit(wordStarts[i][len-1]) )
{
WORDP entry,canonical;
char word[MAX_WORD_SIZE];
strcpy(word,wordStarts[i]);
word[len-1] = 0;
uint64 sysflags = 0;
uint64 cansysflags = 0;
WORDP revise;
GetPosData(i,word,revise,entry,canonical,sysflags,cansysflags,true,true); // dont create a non-existent word
if (entry && entry->properties & PART_OF_SPEECH)
{
wordStarts[i] = reuseAllocation(wordStarts[i],entry->word);
fixedSpell = true;
continue; // not a legal word character, leave it alone
}
}
}
// see if we know the other case
if (!(tokenControl & (ONLY_LOWERCASE|STRICT_CASING)) || (i == startSentence && !(tokenControl & ONLY_LOWERCASE)))
{
WORDP E = FindWord(word,0,SECONDARY_CASE_ALLOWED);
bool useAlternateCase = false;
if (E && E->systemFlags & PATTERN_WORD) useAlternateCase = true;
if (E && E->properties & (PART_OF_SPEECH|FOREIGN_WORD))
{
// if the word we find is UPPER case, and this might be a lower case noun plural, don't change case.
size_t len = strlen(word);
if (word[len-1] == 's' )
{
WORDP F = FindWord(word,len-1);
if (!F || !(F->properties & (PART_OF_SPEECH|FOREIGN_WORD))) useAlternateCase = true;
else continue;
}
else useAlternateCase = true;
}
else if (E) // does it have a member concept fact
{
if (IsConceptMember(E))
{
useAlternateCase = true;
break;
}
}
if (useAlternateCase)
{
char* tokens[2];
tokens[1] = E->word;
ReplaceWords(i,1,1,tokens);
fixedSpell = true;
continue;
}
}
// merge with next token?
char join[MAX_WORD_SIZE * 3];
if (i != wordCount && *wordStarts[i+1] != '"' )
{
// direct merge as a single word
strcpy(join,word);
strcat(join,wordStarts[i+1]);
WORDP D = FindWord(join,0,(tokenControl & ONLY_LOWERCASE) ? PRIMARY_CASE_ALLOWED : STANDARD_LOOKUP);
strcpy(join,word);
if (!D || !(D->properties & PART_OF_SPEECH) ) // merge these two, except "going to" or wordnet composites of normal words // merge as a compound word
{
strcat(join,(char*)"_");
strcat(join,wordStarts[i+1]);
D = FindWord(join,0,(tokenControl & ONLY_LOWERCASE) ? PRIMARY_CASE_ALLOWED : STANDARD_LOOKUP);
}
if (D && D->properties & PART_OF_SPEECH && !(D->properties & AUX_VERB)) // merge these two, except "going to" or wordnet composites of normal words
{
WORDP P1 = FindWord(word,0,LOWERCASE_LOOKUP);
WORDP P2 = FindWord(wordStarts[i+1],0,LOWERCASE_LOOKUP);
if (!P1 || !P2 || !(P1->properties & PART_OF_SPEECH) || !(P2->properties & PART_OF_SPEECH))
{
char* tokens[2];
tokens[1] = D->word;
ReplaceWords(i,2,1,tokens);
fixedSpell = true;
continue;
}
}
}
// break apart slashed pair like eat/feed
char* slash = strchr(word,'/');
if (slash && slash != word && slash[1]) // break apart word/word
{
if ((wordCount + 2 ) >= REAL_SENTENCE_LIMIT) continue; // no room
*slash = 0;
D = StoreWord(word);
*slash = '/';
E = StoreWord(slash+1);
char* tokens[4];
tokens[1] = D->word;
tokens[2] = "/";
tokens[3] = E->word;
ReplaceWords(i,1,3,tokens);
fixedSpell = true;
--i;
continue;
}
// see if hypenated word should be separate or joined (ignore obvious adjective suffix)
if (hyphen && !stricmp(hyphen,(char*)"-like"))
{
StoreWord(word,ADJECTIVE_NORMAL|ADJECTIVE); // accept it as a word
continue;
}
else if (hyphen && (hyphen-word) > 1)
{
char test[MAX_WORD_SIZE];
char first[MAX_WORD_SIZE];
// test for split
*hyphen = 0;
strcpy(test,hyphen+1);
strcpy(first,word);
*hyphen = '-';
WORDP E = FindWord(test,0,LOWERCASE_LOOKUP);
WORDP D = FindWord(first,0,LOWERCASE_LOOKUP);
if (*first == 0)
{
wordStarts[i] = AllocateString(wordStarts[i] + 1); // -pieces want to lose the leading hypen (2-pieces)
fixedSpell = true;
}
else if (D && E) // 1st word gets replaced, we added another word after
{
if ((wordCount + 1 ) >= REAL_SENTENCE_LIMIT) continue; // no room
char* tokens[3];
tokens[1] = D->word;
tokens[2] = E->word;
ReplaceWords(i,1,2,tokens);
fixedSpell = true;
--i;
}
else if (!stricmp(test,(char*)"old") || !stricmp(test,(char*)"olds")) // break apart 5-year-old
{
if ((wordCount + 1 ) >= REAL_SENTENCE_LIMIT) continue; // no room
D = StoreWord(first);
E = StoreWord(test);
char* tokens[3];
tokens[1] = D->word;
tokens[2] = E->word;
ReplaceWords(i,1,2,tokens);
fixedSpell = true;
--i;
}
else // remove hyphen entirely?
{
strcpy(test,first);
strcat(test,hyphen+1);
D = FindWord(test,0,(tokenControl & ONLY_LOWERCASE) ? PRIMARY_CASE_ALLOWED : STANDARD_LOOKUP);
if (D)
{
wordStarts[i] = D->word;
fixedSpell = true;
--i;
}
}
continue; // ignore hypenated errors that we couldnt solve, because no one mistypes a hypen
}
// leave uppercase in first position if not adjusted yet... but check for lower case spell error
if (IsUpperCase(word[0]) && tokenControl & NO_PROPER_SPELLCHECK)
{
char lower[MAX_WORD_SIZE];
MakeLowerCopy(lower,word);
WORDP D = FindWord(lower,0,LOWERCASE_LOOKUP);
if (!D && i == startWord)
{
char* okword = SpellFix(lower,i,PART_OF_SPEECH,language);
if (okword)
{
char* tokens[2];
WORDP E = StoreWord(okword);
tokens[1] = E->word;
ReplaceWords(i,1,1,tokens);
fixedSpell = true;
}
}
continue;
}
if (*word != '\'' && (!FindCanonical(word, i,true) || IsUpperCase(word[0]))) // dont check quoted or findable words unless they are capitalized
{
word = SpellCheck(i,language);
// dont spell check proper names to improper, if word before or after is lower case originally
if (word && i != 1 && originalCapState[i] && !IsUpperCase(*word))
{
if (!originalCapState[i-1]) return false;
else if (i != wordCount && !originalCapState[i+1]) return false;
}
if (word && !*word) // performed substitution on prior word, restart this one
{
fixedSpell = true;
--i;
continue;
}
if (word)
{
char* tokens[2];
tokens[1] = word;
ReplaceWords(i,1,1,tokens);
fixedSpell = true;
continue;
}
}
}
return fixedSpell;
}
