// Generates non-fatal failures and returns false if regex is invalid;
// otherwise returns true.
bool ValidateRegex(const char* regex) {
if (regex == NULL) {
// TODO([email protected]): fix the source file location in the
// assertion failures to match where the regex is used in user
// code.
ADD_FAILURE() << "NULL is not a valid simple regular expression.";
return false;
}
bool is_valid = true;
// True iff ?, *, or + can follow the previous atom.
bool prev_repeatable = false;
for (int i = 0; regex[i]; i++) {
if (regex[i] == '\\') { // An escape sequence
i++;
if (regex[i] == '\0') {
ADD_FAILURE() << FormatRegexSyntaxError(regex, i - 1)
<< "'\\' cannot appear at the end.";
return false;
}
if (!IsValidEscape(regex[i])) {
ADD_FAILURE() << FormatRegexSyntaxError(regex, i - 1)
<< "invalid escape sequence \"\\" << regex[i] << "\".";
is_valid = false;
}
prev_repeatable = true;
} else { // Not an escape sequence.
const char ch = regex[i];
if (ch == '^' && i > 0) {
ADD_FAILURE() << FormatRegexSyntaxError(regex, i)
<< "'^' can only appear at the beginning.";
is_valid = false;
} else if (ch == '$' && regex[i + 1] != '\0') {
ADD_FAILURE() << FormatRegexSyntaxError(regex, i)
<< "'$' can only appear at the end.";
is_valid = false;
} else if (IsInSet(ch, "()[]{}|")) {
ADD_FAILURE() << FormatRegexSyntaxError(regex, i)
<< "'" << ch << "' is unsupported.";
is_valid = false;
} else if (IsRepeat(ch) && !prev_repeatable) {
ADD_FAILURE() << FormatRegexSyntaxError(regex, i)
<< "'" << ch << "' can only follow a repeatable token.";
is_valid = false;
}
prev_repeatable = !IsInSet(ch, "^$?*+");
}
}
return is_valid;
}
bool CISpellAffix::AddToCharset ( char * szRangeL, char * szRangeU )
{
if ( !szRangeL || !szRangeU )
return false;
int iLengthL = strlen ( szRangeL );
int iLengthU = strlen ( szRangeU );
bool bSetL = ( iLengthL>0 && szRangeL[0]=='[' && szRangeL[iLengthL-1]==']' );
bool bSetR = ( iLengthU>0 && szRangeU[0]=='[' && szRangeU[iLengthU-1]==']' );
if ( bSetL!=bSetR )
return false;
if ( bSetL )
{
szRangeL [iLengthL - 1] = '\0';
szRangeL = szRangeL + 1;
szRangeU [iLengthU - 1] = '\0';
szRangeU = szRangeU + 1;
BYTE uMinL, uMaxL;
if ( !GetSetMinMax ( szRangeL, uMinL, uMaxL ) )
return false;
BYTE uMinU, uMaxU;
if ( !GetSetMinMax ( szRangeU, uMinU, uMaxU ) )
return false;
if ( ( uMaxU - uMinU )!=( uMaxL - uMinL ) )
return false;
for ( BYTE i=0; i<=( uMaxL - uMinL ); ++i )
if ( IsInSet ( uMinL + i, szRangeL ) && IsInSet ( uMinU + i, szRangeU ) )
AddCharPair ( uMinL + i, uMinU + i );
} else
{
if ( iLengthL > 4 || iLengthU > 4 )
return false;
const char * szL = szRangeL;
const char * szU = szRangeU;
AddCharPair ( GetWordchar(szL), GetWordchar(szU) );
}
m_bUseDictConversion = true;
return true;
}
bool CISpellAffixRule::CheckPrefix ( const CSphString & sWord ) const
{
int iCondI = 0;
for ( int i = 0; iCondI < m_iCondLen && i < m_iWordLen; ++i )
{
if ( m_sCondition.cstr()[iCondI]!='[' )
{
if ( m_sCondition.cstr()[iCondI]!=sWord.cstr()[i] )
return false;
++iCondI;
} else
{
int iRangeEnd = -1;
for ( int j=iCondI; j<m_iCondLen && iRangeEnd==-1; ++j )
if ( m_sCondition.cstr()[j]==']' )
iRangeEnd = j;
if ( iRangeEnd==-1 )
return false;
else
{
if ( !IsInSet ( sWord.cstr () [i], m_sCondition.SubString ( iCondI + 1, iRangeEnd - iCondI - 1 ).cstr () ) )
return false;
}
}
}
return true;
}
bool CISpellAffixRule::CheckSuffix ( const CSphString & sWord ) const
{
int iCondI = m_iCondLen-1;
for ( int i=m_iWordLen-1; iCondI>=0 && i>=0; --i )
{
if ( m_sCondition.cstr()[iCondI]!=']' )
{
if ( m_sCondition.cstr()[iCondI]!=sWord.cstr()[i] )
return false;
--iCondI;
} else
{
int iRangeStart = -1;
for ( int j=iCondI; j>=0 && iRangeStart==-1; --j )
if ( m_sCondition.cstr()[j]=='[' )
iRangeStart = j;
if ( iRangeStart==-1 )
return false;
else
{
if ( !IsInSet ( sWord.cstr () [i], m_sCondition.SubString ( iRangeStart + 1, iCondI - iRangeStart - 1 ).cstr () ) )
return false;
iCondI = iRangeStart - 1;
}
}
}
return true;
}
// Search for the collection which contains a specific type
static
_cmsParametricCurvesCollection *GetParametricCurveByType(int Type, int* index)
{
_cmsParametricCurvesCollection* c;
int Position;
for (c = ParametricCurves; c != NULL; c = c ->Next) {
Position = IsInSet(Type, c);
if (Position != -1) {
if (index != NULL)
*index = Position;
return c;
}
}
return NULL;
}
// Returns true iff "\\c" is a supported escape sequence.
bool IsValidEscape(char c) {
return (IsAsciiPunct(c) || IsInSet(c, "dDfnrsStvwW"));
}
bool IsAsciiWhiteSpace(char ch) { return IsInSet(ch, " \f\n\r\t\v"); }
bool IsRepeat(char ch) { return IsInSet(ch, "?*+"); }
bool IsAsciiPunct(char ch) {
return IsInSet(ch, "^-!\"#$%&'()*+,./:;<=>?@[\\]_`{|}~");
}
