bool Get_XML_colour (CXMLelement & node,
const char * sName,
COLORREF & cValue,
const bool bUseDefault)
{
CString strValue;
if (!Get_XML_string (node, sName, strValue, true, true))
if (bUseDefault)
return false;
else
{
cValue = 0; // default is black
return false;
}
if (strValue.IsEmpty ())
ThrowErrorException ("No colour supplied for attribute named '%s'" ,
sName);
if (SetColour (strValue, cValue))
ThrowErrorException ("Invalid colour code \"%s\" for attribute named '%s'" ,
(LPCTSTR) strValue,
sName);
return true;
} // end of Get_XML_colour
void CXMLparser::ProcessEntity (void)
{
CString strName;
CString strValue;
CString strTemp;
strName = GetName ("entity");
strTemp = strName.Left (3);
strTemp.MakeUpper ();
if (strTemp == "XML")
ThrowErrorException ("Names starting with \"XML\" are reserved");
strValue = GetValue ("entity", strName);
// entities might have entities in them, so replace them
m_CustomEntityMap.SetAt (strName, ReplaceEntities (strValue));
if (m_xmlBuff [m_xmlPos] != '>')
ThrowErrorException ("Expected '>', got %c",
m_xmlBuff [m_xmlPos]);
m_xmlPos++; // skip >
// bypass spaces and comments
SkipComments (true);
} // end of CXMLparser::ProcessEntity
t_regexp * regcomp(const char *exp, const int options)
{
const char *error;
int erroroffset;
t_regexp * re;
pcre * program;
pcre_extra * extra;
program = pcre_compile(exp, options, &error, &erroroffset, NULL);
if (!program)
ThrowErrorException("Failed: %s at offset %d", Translate (error), erroroffset);
// study it for speed purposes
extra = pcre_study(program, 0, &error);
if (error)
{
pcre_free (program);
ThrowErrorException("Regexp study failed: %s", error);
}
// we need to allocate memory for the substring offsets
re = new t_regexp;
// remember program and extra stuff
re->m_program = program;
re->m_extra = extra;
re->m_iExecutionError = 0; // no error now
return re;
}
void CXMLparser::SkipComments (const bool bAndSpaces)
{
while (true) // loop in case consecutive comments, or comment - space - comment
{
if (bAndSpaces)
SkipSpaces ();
// comment?
if (m_xmlBuff [m_xmlPos] == '<' &&
m_xmlBuff [m_xmlPos + 1] == '!' &&
m_xmlBuff [m_xmlPos + 2] == '-' &&
m_xmlBuff [m_xmlPos + 3] == '-')
{
// skip comment
m_xmlPos += 4; // skip <!--
while (m_xmlBuff [m_xmlPos] != '-' ||
m_xmlBuff [m_xmlPos + 1] != '-' ||
m_xmlBuff [m_xmlPos + 2] != '>')
{
if (m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Unterminated comment at end of file");
if (m_xmlBuff [m_xmlPos] == '-' &&
m_xmlBuff [m_xmlPos + 1] == '-')
ThrowErrorException ("Invalid '--' inside comment");
if (m_xmlBuff [m_xmlPos] == '\n')
m_xmlLine++; // count lines
m_xmlPos++;
}
// here for end of comment
m_xmlPos += 3; // skip -->
} // end of comment
else
return; // no comment, all done
};
} // end of SkipComments
void CXMLparser::ProcessDoctype (void)
{
m_strDocumentName = GetName ("Root document");
// document name may be followed by SYSTEM or PUBLIC
if (isalpha (m_xmlBuff [m_xmlPos]))
{
CString strName = GetName ("External ID");
// we'll get and ignore these for future compatibility
if (strName == "SYSTEM")
GetValue ("SYSTEM", "DTD name"); // eg. SYSTEM "c:\blah\my.dtd"
else
if (strName == "PUBLIC")
{
GetValue ("PUBLIC", "Public ID"); // eg. PUBLIC "blah" "c:\blah\my.dtd"
GetValue ("PUBLIC", "DTD name");
}
else
ThrowErrorException ("Unsupported declaration \"%s\" inside <!DOCTYPE ...>",
(LPCTSTR) strName);
} // end of SYSTEM/PUBLIC
if (m_xmlBuff [m_xmlPos] == '[')
ProcessDTD ();
if (m_xmlBuff [m_xmlPos] != '>')
{
// shouldn't have end-of-file before the >
if (m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Unexpected end-of-file inside <!DOCTYPE> declaration");
ThrowErrorException ("Unsupported declaration inside <!DOCTYPE ...>");
} // end of looking for attributes
m_xmlPos++; // skip '>' symbol
} // end of CXMLparser::ProcessDoctype
void CXMLparser::ProcessDeclaration (const CString strDeclaration)
{
if (strDeclaration == "DOCTYPE")
ProcessDoctype ();
else
ThrowErrorException ("Unrecognised declaration: <!%s ...>",
(LPCTSTR) strDeclaration);
} // end of CXMLparser::ProcessDeclaration
CString CXMLparser::ReplaceEntities (const CString strSource)
{
// look for entities imbedded in the string
const char * p = strSource;
const char * pStart = strSource; // where buffer starts
const char * pEntity;
CString strFixedValue;
CString strEntity;
long length;
// quick check to eliminate ones that don't have imbedded entities
if (strSource.Find ('&') == -1)
return strSource;
strFixedValue.Empty ();
for ( ; *p; p++)
{
if (*p == '&')
{
// copy up to ampersand
length = p - pStart;
if (length > 0)
strFixedValue += CString (pStart, length);
p++; // skip ampersand
pEntity = p; // where entity starts
if (*p == '#')
p++;
while (isalnum (*p) ||
*p == '-' ||
*p == '.' ||
*p == '_')
p++;
if (*p != ';')
ThrowErrorException ("No closing \";\" in XML entity argument \"&%s\"",
(LPCTSTR) CString (pEntity, p - pEntity));
strFixedValue += Get_XML_Entity (CString (pEntity, p - pEntity)); // add to list
pStart = p + 1; // move on past the entity
} // end of having an ampersand
} // end of processing the value
// copy last bit
strFixedValue += pStart;
return strFixedValue;
} // end of CXMLparser::ReplaceEntities
void CXMLparser::ProcessDTD (void)
{
CString strName;
m_xmlPos++; // skip the [ symbol
SkipComments (true);
while (m_xmlBuff [m_xmlPos] != ']')
{
// shouldn't have end-of-file before the >
if (m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Unexpected end-of-file inside DTD declaration");
// look for declaration start
if (m_xmlBuff [m_xmlPos] != '<' || m_xmlBuff [m_xmlPos + 1] != '!')
ThrowErrorException ("Expected '<!', got %c",
m_xmlBuff [m_xmlPos]);
m_xmlPos += 2; // skip <!
strName = GetName ("DTD Declaration");
// we only support <!ENTITY ... > right now
if (strName == "ENTITY")
ProcessEntity ();
else
ThrowErrorException ("Unsupported declaration <!%s ...>",
(LPCTSTR) strName);
} // end of looking for declarations
m_xmlPos++; // skip ']' symbol
SkipComments (true);
} // end of CXMLparser::ProcessDTD
bool Get_XML_boolean (CXMLelement & node,
const char * sName,
bool & bValue,
const bool bUseDefault)
{
CString strValue;
if (!Get_XML_string (node, sName, strValue, bUseDefault))
if (bUseDefault)
return false;
// be nice and let them slip in Y or y
strValue.MakeLower ();
if (strValue == "0" ||
strValue == "" || // empty can be considered false
strValue == "no" ||
strValue == "n" ||
strValue == "false" ||
strValue == "f"
)
{
bValue = false;
return true;
}
if (strValue == "1" ||
strValue == "-1" ||
strValue == "yes" ||
strValue == "y" ||
strValue == "true" ||
strValue == "t"
)
{
bValue = true;
return true;
}
ThrowErrorException ("Boolean attribute named \"%s\" has unrecognised value \"%s\"",
sName,
(LPCTSTR) strValue);
bValue = false;
return true;
} // end of Get_XML_boolean
///////////////////////////////////////////////////////////////////////////////
// Write to a file. An assertion will occur if file was opened for reading.
// Inputs:
// unsigned int size Number of bytes to write to the file.
// const char* bufferToWrite Buffer containing data to write to the file.
// Exceptions:
// ErrorException thrown on failure.
///////////////////////////////////////////////////////////////////////////////
void RawFileLinux::Write(
unsigned int size,
const char* bufferToWrite
) {
//Write everything in one pass
assert( IsOpen() );
filestream.exceptions( std::fstream::badbit | std::fstream::failbit | std::fstream::goodbit );
try {
filestream.write( bufferToWrite, size );
lastPosition = position;
position = filestream.tellp();
filestream.clear();
} catch( std::fstream::failure e ) { //throw exception
ThrowErrorException( e );
}
}
bool Get_XML_number (CXMLelement & node,
const char * sName,
long & iValue,
const bool bUseDefault,
long iMinimum,
long iMaximum)
{
CString strValue;
if (!Get_XML_string (node, sName, strValue, bUseDefault))
if (bUseDefault)
return false;
else
strValue = "0"; // default is zero
__int64 iResult = 0;
const char * p = strValue;
bool bNegative = false;
// see if sign given
if (*p == '-')
{
bNegative = true;
p++;
}
else
if (*p == '+')
p++;
if (*p == 0)
ThrowErrorException ("No number suppied for numeric attribute named '%s'" ,
sName);
if (!isdigit (*p))
ThrowErrorException ("Invalid number \"%s\" for numeric attribute named '%s'" ,
(LPCTSTR) strValue,
sName);
for (iResult = 0; *p; p++)
{
if (!isdigit (*p))
ThrowErrorException ("Invalid character '%c' in numeric attribute named '%s'" ,
*p,
sName);
iResult *= 10;
iResult += *p - '0';
if (iResult > ((__int64) LONG_MAX + 1)) // to allow for LONG_MIN which will be one more
ThrowErrorException ("Value '%s' too large in numeric attribute named '%s'" ,
(LPCTSTR) strValue,
sName);
}
// make negative if necessary
if (bNegative)
iResult = 0 - iResult;
// fix up default 0, 0 to be 0, 1
if (iMinimum == 0 && iMaximum == 0)
iMaximum = 1;
if (iResult > iMaximum)
ThrowErrorException ("Value '%s' too large in numeric attribute named '%s'. "
"Range is %ld to %ld.",
(LPCTSTR) strValue,
sName,
iMinimum,
iMaximum);
if (iResult < iMinimum)
ThrowErrorException ("Value '%s' too small in numeric attribute named '%s'. "
"Range is %ld to %ld.",
(LPCTSTR) strValue,
sName,
iMinimum,
iMaximum);
iValue = (long) iResult;
return true;
} // end of Get_XML_number
CString CXMLparser::GetValue (const char * sType, const char * sName)
{
char cQuote;
UINT iStart;
CString strValue;
int iContentLen;
// in case spaces before the value
SkipSpaces ();
// find what sort of quote they are using
cQuote = m_xmlBuff [m_xmlPos];
if (cQuote == 0)
ThrowErrorException ("Unexpected end-of-file looking for value of %s named %s",
sType,
sName);
if (cQuote != '\'' && cQuote != '\"')
// value starting with > is really an omitted value
if (cQuote == '>')
ThrowErrorException ("Value for %s not supplied",
sType);
// detect a comment here as a special case
else
if (m_xmlBuff [m_xmlPos] == '<' &&
m_xmlBuff [m_xmlPos + 1] == '!' &&
m_xmlBuff [m_xmlPos + 2] == '-' &&
m_xmlBuff [m_xmlPos + 3] == '-')
ThrowErrorException ("Comment not valid here");
else
ThrowErrorException ("Value for %s named '%s' not supplied (in quotes)",
sType,
sName);
m_xmlPos++; // get past quote
iStart = m_xmlPos; // start of value
while (m_xmlBuff [m_xmlPos] != cQuote)
{
if (m_xmlBuff [m_xmlPos] == '\n' ||
m_xmlBuff [m_xmlPos] == '\r' ||
m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Value for %s named '%s' not terminated (no closing quote)",
sType,
sName);
// <![CDATA[ terminator invalid here (ie. ]]> )
if (m_xmlBuff [m_xmlPos] == ']' &&
m_xmlBuff [m_xmlPos + 1] == ']' &&
m_xmlBuff [m_xmlPos + 2] == '>')
ThrowErrorException ("Value for %s named '%s' may not contain CDATA terminator: ']]>'",
sType,
sName);
// don't let them slip in non-printables
if ((unsigned char) m_xmlBuff [m_xmlPos] < ' ')
ThrowErrorException ("Non-printable character, code %i, not permitted for "
"value for %s named '%s'",
(int) m_xmlBuff [m_xmlPos],
sType,
sName);
m_xmlPos++;
}
iContentLen = m_xmlPos - iStart;
m_xmlPos++; // get past quote
SkipSpaces ();
// construct value from buffer
return CString (&m_xmlBuff [iStart], iContentLen);
} // end of CXMLparser::GetValue
bool Get_XML_date (CXMLelement & node,
const char * sName,
CTime & tValue,
const bool bUseDefault)
{
CString strValue;
if (!Get_XML_string (node, sName, strValue, true, true))
if (bUseDefault)
return false;
else
{
tValue = 0; // default is no time
return false;
}
if (strValue.IsEmpty ())
ThrowErrorException ("No time supplied for attribute named '%s'" ,
sName);
// break up date into date portion / time portion
CStringList strDateTime;
StringToList (strValue, " ", strDateTime);
if (strDateTime.GetCount () < 1 || strDateTime.GetCount () > 2)
ThrowErrorException ("Date/time must consist of YYYY-MM-DD [ HH:MM:SS ]");
CString strDate = strDateTime.RemoveHead ();
CString strTime;
if (!strDateTime.IsEmpty ())
strTime = strDateTime.RemoveHead ();
CString strYear, strMonth, strDay;
CStringList strDateList;
StringToList (strDate, "-", strDateList);
if (strDateList.GetCount () != 3)
ThrowErrorException ("Date must consist of YYYY-MM-DD");
strYear = strDateList.RemoveHead ();
strMonth = strDateList.RemoveHead ();
strDay = strDateList.RemoveHead ();
if (!IsNumber (strYear))
ThrowErrorException ("Year is not numeric");
if (!IsNumber (strMonth))
ThrowErrorException ("Month is not numeric");
if (!IsNumber (strDay))
ThrowErrorException ("Day is not numeric");
int iYear, iMonth, iDay;
iYear = atoi (strYear);
iMonth = atoi (strMonth);
iDay = atoi (strDay);
if (iYear < 1900 || iYear > 2100)
ThrowErrorException ("Year must be in range 1900 to 2100");
if (iMonth < 1 || iMonth > 12)
ThrowErrorException ("Month must be in range 1 to 12");
if (iDay < 1 || iDay > 31)
ThrowErrorException ("Month must be in range 1 to 31");
int iHour = 0, iMinute = 0, iSecond = 0;
if (!strTime.IsEmpty ())
{
CString strHour, strMinute, strSecond;
CStringList strTimeList;
StringToList (strTime, ":", strTimeList);
if (strTimeList.GetCount () > 3)
ThrowErrorException ("Time must consist of HH, HH:MM, or HH:MM:SS");
strHour = strTimeList.RemoveHead ();
strMinute = "0";
strSecond = "0";
if (!strTimeList.IsEmpty ())
{
strMinute = strTimeList.RemoveHead ();
if (!strTimeList.IsEmpty ())
strSecond = strTimeList.RemoveHead ();
}
if (!IsNumber (strHour))
ThrowErrorException ("Hour is not numeric");
if (!IsNumber (strMinute))
ThrowErrorException ("Minute is not numeric");
if (!IsNumber (strSecond))
ThrowErrorException ("Second is not numeric");
iHour = atoi (strHour);
iMinute = atoi (strMinute);
iSecond = atoi (strSecond);
if (iHour < 0 || iHour > 23)
ThrowErrorException ("Hour must be in range 0 to 23");
if (iMinute < 0 || iMinute > 59)
ThrowErrorException ("Minute must be in range 0 to 59");
if (iSecond < 0 || iSecond > 59)
ThrowErrorException ("Minute must be in range 0 to 59");
} // end of having a time
tValue = CTime (iYear, iMonth, iDay, iHour, iMinute, iSecond, 0);
return true;
} // end of Get_XML_date
bool Get_XML_double (CXMLelement & node,
const char * sName,
double & dValue,
const bool bUseDefault,
const double dMinimum,
const double dMaximum)
{
CString strValue;
if (!Get_XML_string (node, sName, strValue, bUseDefault))
if (bUseDefault)
return false;
else
strValue = "0"; // default is zero
double dResult = 0;
const char * p = strValue;
int iDots = 0;
int iExponents = 0;
// see if sign given
if (*p == '-' || *p == '+')
p++;
if (*p == 0)
ThrowErrorException ("No number suppied for numeric attribute named '%s'" ,
sName);
if (!isdigit (*p))
ThrowErrorException ("Invalid number \"%s\" for numeric attribute named '%s'" ,
(LPCTSTR) strValue,
sName);
for ( ; *p; p++)
{
if (*p == '.')
{
if (++iDots > 1)
ThrowErrorException ("Too many decimal places for numeric attribute named '%s'" ,
sName);
} // end of decimal place
else if (toupper (*p) == 'E')
{
if (++iExponents > 1)
ThrowErrorException ("Too many 'E' characters for numeric attribute named '%s'" ,
sName);
// exponent may have sign
if (*p == '-' || *p == '+')
p++;
} // end of exponent
else if(!isdigit (*p))
ThrowErrorException ("Invalid character '%c' in numeric attribute named '%s'" ,
*p,
sName);
} // end of checking each character
myAtoF (strValue, &dResult);
/* // myAtoF does not return myAtoF
if (dResult == c || dResult == -HUGE_VAL)
ThrowErrorException ("Value '%s' out of range in numeric attribute named '%s'" ,
(LPCTSTR) strValue,
sName);
*/
if (dResult > dMaximum)
ThrowErrorException ("Value '%s' too large in numeric attribute named '%s'. "
"Range is %g to %g.",
(LPCTSTR) strValue,
sName,
dMinimum,
dMaximum);
if (dResult < dMinimum)
ThrowErrorException ("Value '%s' too small in numeric attribute named '%s'. "
"Range is %g to %g.",
(LPCTSTR) strValue,
sName,
dMinimum,
dMaximum);
dValue = dResult;
return true;
} // end of Get_XML_double
CString CXMLparser::Get_XML_Entity (CString & strName)
{
CString strEntityContents;
// look for &#nnn;
if (strName [0] == '#')
{
int iResult = 0;
// validate and work out number
if (strName [1] == 'x')
{
for (int i = 2; i < strName.GetLength (); i++)
{
if (!isxdigit (strName [i]))
{
ThrowErrorException ("Invalid hex number in XML entity: &%s;" ,
(LPCTSTR) strName);
return "";
}
int iNewDigit = toupper (strName [i]);
if (iNewDigit >= 'A')
iNewDigit -= 7;
if (iResult & 0xF0)
ThrowErrorException ("Invalid hex number in XML entity: &%s; "
"- maximum of 2 hex digits",
(LPCTSTR) strName);
iResult = (iResult << 4) + iNewDigit - '0';
}
} // end of hex entity
else
{
for (int i = 1; i < strName.GetLength (); i++)
{
if (!isdigit (strName [i]))
{
ThrowErrorException ("Invalid number in XML entity: &%s;" ,
(LPCTSTR) strName);
return "";
}
iResult *= 10;
iResult += strName [i] - '0';
}
} // end of decimal entity
if (iResult != 9) // we will accept tabs ;)
if (iResult < 32 || // don't allow nonprintable characters
iResult > 255) // don't allow characters more than 1 byte
{
ThrowErrorException ("Disallowed number in XML entity: &%s;" ,
(LPCTSTR) strName);
return "";
}
unsigned char cOneCharacterLine [2] = { (unsigned char) iResult, 0};
return (char *) cOneCharacterLine;
} // end of entity starting with #
// look up global entities first
if (App.m_EntityMap.Lookup (strName, strEntityContents))
return strEntityContents;
// then try ones for this document
else if (m_CustomEntityMap.Lookup (strName, strEntityContents))
return strEntityContents;
ThrowErrorException ("Unknown XML entity: &%s;" ,
(LPCTSTR) strName);
return ""; // to avoid warning
} // end of CXMLparser::Get_XML_Entity
void CXMLparser::BuildStructure (CFile * file)
{
m_xmlLength = file->GetLength ();
// can hardly be XML if empty file
if (m_xmlLength == 0)
ThrowErrorException ("Zero length XML file \"%s\"",
(LPCTSTR) file->GetFilePath ());
// sanity check - don't read 20 Mb files into memory
if (m_xmlLength > MAX_XML_DOCUMENT_SIZE)
ThrowErrorException ("XML file \"%s\"too large - maximum permitted is %i bytes",
(LPCTSTR) file->GetFilePath (),
MAX_XML_DOCUMENT_SIZE);
char * p = m_strxmlBuffer.GetBuffer (m_xmlLength + 1);
iLineLastItemFound = 0;
if (file->Read (p, m_xmlLength) != m_xmlLength)
{
m_strxmlBuffer.ReleaseBuffer (0);
ThrowErrorException ("Could not read XML file");
}
m_strxmlBuffer.ReleaseBuffer (m_xmlLength);
// look for Unicode (FF FE)
if ((unsigned char) m_strxmlBuffer [0] == 0xFF &&
(unsigned char) m_strxmlBuffer [1] == 0xFE)
{
CString buf2;
const char * q = m_strxmlBuffer;
q += 2; // skip indicator bytes
// find required buffer length
int length = WideCharToMultiByte (CP_UTF8, 0, (LPCWSTR) q, (m_xmlLength - 2) / 2,
NULL, 0, NULL, NULL);
// make a new string with enough length to hold it
char * p = buf2.GetBuffer (length);
// convert it
WideCharToMultiByte (CP_UTF8, 0, (LPCWSTR) q, (m_xmlLength - 2) / 2,
p, length, NULL, NULL);
buf2.ReleaseBuffer (length);
// copy to our buffer
m_strxmlBuffer = buf2;
}
else
// look for UTF-8 indicator bytes (EF BB BF)
if ((unsigned char) m_strxmlBuffer [0] == 0xEF &&
(unsigned char) m_strxmlBuffer [1] == 0xBB &&
(unsigned char) m_strxmlBuffer [2] == 0xBF)
m_strxmlBuffer = m_strxmlBuffer.Mid (3); // skip them
// convert tabs to spaces, we don't want tabs in our data
m_strxmlBuffer.Replace ('\t', ' ');
SeeIfBase64 (m_strxmlBuffer);
m_xmlBuff = m_strxmlBuffer; // get const char * pointer to buffer
ProcessNode (m_xmlRoot); // process root node
} // end of CXMLparser::BuildStructure
CString CXMLparser::GetName (const char * sType)
{
int iStart;
int iNameLen;
iStart = m_xmlPos; // start of name
if (m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Unexpected end-of-file looking for %s name",
sType);
// name must start with letter or underscore
if (!isalpha (m_xmlBuff [m_xmlPos]) && m_xmlBuff [m_xmlPos] != '_')
ThrowErrorException ("%s name must start with letter or underscore, "
"but starts with \"%c\"",
sType,
m_xmlBuff [m_xmlPos]);
// build up name
while (isalnum (m_xmlBuff [m_xmlPos]) ||
m_xmlBuff [m_xmlPos] == '-' ||
m_xmlBuff [m_xmlPos] == '.' ||
m_xmlBuff [m_xmlPos] == '_')
m_xmlPos++;
// OK - we now have a name, unless it is empty
iNameLen = m_xmlPos - iStart;
if (iNameLen == 0)
// name starting with > is really an omitted name
if (m_xmlBuff [m_xmlPos] == '>')
ThrowErrorException ("%s name not supplied",
sType);
// detect a comment inside an element as a special case
else
if (m_xmlBuff [m_xmlPos] == '<' &&
m_xmlBuff [m_xmlPos + 1] == '!' &&
m_xmlBuff [m_xmlPos + 2] == '-' &&
m_xmlBuff [m_xmlPos + 3] == '-')
ThrowErrorException ("Comment not valid here");
else
ThrowErrorException ("%s name cannot start with '%c'",
sType,
m_xmlBuff [m_xmlPos]);
// sanity check - make sure name not ridiculous
if (iNameLen > MAX_XML_NAME_LENGTH)
ThrowErrorException ("%s name is too long (%i characters) "
" maximum permitted is %i characters",
sType,
iNameLen,
MAX_XML_NAME_LENGTH);
// in case spaces before next thing
SkipSpaces ();
// construct name from buffer
return CString (&m_xmlBuff [iStart], iNameLen);
} // end of CXMLparser::GetName
void CXMLparser::ProcessAttributes (CXMLelement & node)
{
CAttribute * pAttribute;
CString strName;
CString strValue;
while (m_xmlBuff [m_xmlPos] != '>')
{
// look for "empty element" terminator: />
if (m_xmlBuff [m_xmlPos] == '/' && m_xmlBuff [m_xmlPos + 1] == '>')
{
node.bEmpty = true;
m_xmlPos++; // skip '/' symbol
break;
}
// look for terminator ?> at end of <?xml tag
if (m_xmlBuff [m_xmlPos] == '?' &&
m_xmlBuff [m_xmlPos + 1] == '>' &&
node.strName [0] == '?'
)
{
node.bEmpty = true;
m_xmlPos++; // skip '?' symbol
break;
}
// shouldn't have end-of-file before the >
if (m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Unexpected end-of-file inside element definition for %s",
(LPCTSTR) node.strName);
// we really should have a name now ...
strName = GetName ("Attribute");
// already there? error
if (node.AttributeMap.Lookup (strName, pAttribute))
ThrowErrorException ("Duplicate attribute name '%s'",
(LPCTSTR) strName);
// in case spaces before next thing
SkipSpaces ();
if (m_xmlBuff [m_xmlPos] != '=')
ThrowErrorException ("Attribute name '%s' not followed by '='",
(LPCTSTR) strName);
m_xmlPos++; // skip the = sign
int iLine = m_xmlLine; // line that = sign is on
strValue = GetValue ("attribute", strName);
pAttribute = new CAttribute;
if (!pAttribute)
ThrowErrorException ("Could not allocate memory for XML parsing of attribute %s",
(LPCTSTR) strName);
pAttribute->strName = strName;
pAttribute->strValue = ReplaceEntities (strValue);
pAttribute->iLine = iLine;
// add to map
node.AttributeMap.SetAt (strName, pAttribute);
} // end of looking for attributes
m_xmlPos++; // skip '>' symbol
} // end of CXMLparser::ProcessAttributes
void CXMLparser::AssembleContent (CXMLelement & node,
const UINT iFirstLine,
const UINT iNodeStart)
{
/*
work out content - it will be non-nested text belonging to this element
eg. <a>xxx<b>yyy</b>zzz</a>
"a" will have content xxxzzz
"b" will have content yyy
whitespace is preserved
You can reassemble an entire element (including nested ones) by
concatenating the contents of its children.
For cross-platform consistency, carriage returns are dropped,
and linefeeds become carriage-return/linefeed
*/
int iContentLen = m_xmlPos - iNodeStart;
int iLines = m_xmlLine - iFirstLine; // lines of content
const char * pi = &m_xmlBuff [iNodeStart]; // copy from
const char * pl = &m_xmlBuff [m_xmlPos]; // limit of copy
char * po = node.strContent.GetBuffer (iContentLen + 1 + iLines); // copy to
int iDepth = 0;
bool bInside = false;
char cQuote = ' ';
while (pi < pl) // copy, dropping nested elements
{
// they might have <blah value=">">
// so we have to skip special characters inside quotes
if (bInside && (*pi == '\'' || *pi == '\"'))
{
cQuote = *pi; // for terminator
pi++;
while (pi < pl && *pi != cQuote)
pi++;
pi++; // skip terminating quote
continue; // back to start of loop
} // end of quotes inside a definition
// <![CDATA[ is special - just copy contents
if (pi [0] == '<' &&
pi [1] == '!' &&
pi [2] == '[' &&
pi [3] == 'C' &&
pi [4] == 'D' &&
pi [5] == 'A' &&
pi [6] == 'T' &&
pi [7] == 'A' &&
pi [8] == '[')
{
pi += 9; // skip <![CDATA[
// this shouldn't happen
if (bInside)
ThrowErrorException ("Unexpected '<![CDATA[' inside element definition");
// should be terminated as we checked earlier, but may as well check again
while (pi < pl &&
(pi [0] != ']' ||
pi [1] != ']' ||
pi [2] != '>'))
{
// copy if not nested, and not inside an element definition
// -- omit carriage returns
if (iDepth == 0 &&
*pi != '\r')
{
// make linefeeds into carriage return/linefeed
if (*pi == '\n')
*po++ = '\r';
// special treatment for ampersands, so they don't get converted
// into entities ...
if (*pi == '&')
*po++ = '\x01';
else
*po++ = *pi; // copy it
}
pi++;
} // end of inside <![CDATA[
pi += 3; // skip ]]>
continue; // back to start of loop
} // end of <![CDATA[ section
if (*pi == '<') // going into an element
{
bInside = true;
if (pi [1] == '/') // </tag> drops a level
{
iDepth--;
if (iDepth < 0)
iDepth = 0; // cannot have negative depth
}
else
if (pi [1] != '!' && pi [1] != '?') // excepting special ones
iDepth++; // <tag> goes up a level
}
// copy if not nested, and not inside an element definition
// -- omit carriage returns
if (iDepth == 0 &&
!bInside &&
*pi != '\r')
{
// make linefeeds into carriage return/linefeed
if (*pi == '\n')
*po++ = '\r';
*po++ = *pi; // copy if not inside an element
}
// leaving an element (only occurs 'inside' a <... sequence )
if (bInside && pi [0] == '/' && pi [1] == '>')
{
iDepth--; // cancel level gain - this tag is empty
if (iDepth < 0)
iDepth = 0; // cannot have negative depth
}
else
if (*pi == '>') // leaving an element
bInside = false;
pi++; // onto next
}
*po = 0; // terminating null
node.strContent.ReleaseBuffer (-1);
// get rid of entities (such as <) in the content
node.strContent = ReplaceEntities (node.strContent);
// now, in case we had <![CDATA[ sections, replace 0x01 by &
node.strContent.Replace ('\x01', '&');
} // end of CXMLparser::AssembleContent
CString CMUSHclientDoc::FixSendText (const CString strSource,
const int iSendTo,
const t_regexp * regexp, // regular expression (for triggers, aliases)
const char * sLanguage, // language for send-to-script
const bool bMakeWildcardsLower,
const bool bExpandVariables, // convert @foo
const bool bExpandWildcards, // convert %x
const bool bFixRegexps, // convert \ to \\ for instance
const bool bIsRegexp, // true = regexp trigger
const bool bThrowExceptions)
{
CString strOutput; // result of expansion
const char * pText,
* pStartOfGroup;
pText = pStartOfGroup = strSource;
while (*pText)
{
switch (*pText)
{
/* -------------------------------------------------------------------- *
* Variable expansion - @foo becomes <contents of foo> *
* -------------------------------------------------------------------- */
case '@':
{
if (!bExpandVariables)
{
pText++; // just copy the @
break;
}
// copy up to the @ sign
strOutput += CString (pStartOfGroup, pText - pStartOfGroup);
pText++; // skip the @
// @@ becomes @
if (*pText == '@')
{
pStartOfGroup = ++pText;
strOutput += "@";
continue;
}
const char * pName;
bool bEscape = bFixRegexps;
// syntax @!variable defeats the escaping
if (*pText == '!')
{
pText++;
bEscape = false;
}
pName = pText;
// find end of variable name
while (*pText)
if (*pText == '_' || isalnum ((unsigned char) *pText))
pText++;
else
break;
/* -------------------------------------------------------------------- *
* We have a variable - look it up and do internal replacements *
* -------------------------------------------------------------------- */
CString strVariableName (pName, pText - pName);
if (strVariableName.IsEmpty ())
{
if (bThrowExceptions)
ThrowErrorException ("@ must be followed by a variable name");
} // end of no variable name
else
{
CVariable * variable_item;
strVariableName.MakeLower ();
if (GetVariableMap ().Lookup (strVariableName, variable_item))
{
// fix up so regexps don't get confused with [ etc. inside variable
CString strVariableContents;
if (bEscape)
{
const char * pi;
// allow for doubling in size, plus terminating null
char * po = strVariableContents.GetBuffer ((variable_item->strContents.GetLength () * 2) + 1);
for (pi = variable_item->strContents;
*pi;
pi++)
{
if (((unsigned char) *pi) < ' ')
continue; // drop non-printables
if (bIsRegexp)
if (!isalnum ((unsigned char) *pi) && *pi != ' ')
{
*po++ = '\\'; // escape it
*po++ = *pi;
}
else
*po++ = *pi; // just copy it
else // not regexp
if (*pi != '*')
*po++ = *pi; // copy all except asterisks
}
*po = 0; // terminating null
strVariableContents.ReleaseBuffer (-1);
} // end of escaping wanted
else
strVariableContents = variable_item->strContents;
// in the "send" box may need to convert script expansions etc.
if (!bFixRegexps)
strVariableContents = FixWildcard ((const char *) strVariableContents,
false, // not force variables to lowercase
iSendTo,
sLanguage).c_str ();
// fix up HTML sequences if we are sending it to a log file
if (m_bLogHTML && iSendTo == eSendToLogFile)
strVariableContents = FixHTMLString (strVariableContents);
strOutput += strVariableContents;
} // end of name existing in variable map
else
{
if (bThrowExceptions)
ThrowErrorException ("Variable '%s' is not defined.",
(LPCTSTR) strVariableName);
} // end of variable does not exist
} // end of not empty name
// get ready for next batch from beyond the variable
pStartOfGroup = pText;
}
break; // end of '@'
/* -------------------------------------------------------------------- *
* Wildcard substitution - %1 becomes <contents of wildcard 1> *
* - %<foo> becomes <contents of wildcard "foo" *
* - %% becomes % *
* -------------------------------------------------------------------- */
case '%':
if (!bExpandWildcards)
{
pText++; // just copy the %
break;
}
// see what comes after the % symbol
switch (pText [1])
{
/* -------------------------------------------------------------------- *
* %% *
* -------------------------------------------------------------------- */
case '%':
// copy up to - and including - the percent sign
strOutput += CString (pStartOfGroup, pText - pStartOfGroup + 1);
// get ready for next batch from beyond the %%
pText += 2; // don't reprocess the %%
pStartOfGroup = pText;
break; // end of %%
/* -------------------------------------------------------------------- *
* %0 to %9 *
* -------------------------------------------------------------------- */
case '0': // a digit?
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
{
// copy up to the percent sign
strOutput += CString (pStartOfGroup, pText - pStartOfGroup);
// output the appropriate replacement text
if (regexp)
{
CString strWildcard = FixWildcard (regexp->GetWildcard (string (1, *++pText)),
bMakeWildcardsLower,
iSendTo,
sLanguage).c_str ();
// fix up HTML sequences if we are sending it to a log file
if (m_bLogHTML && iSendTo == eSendToLogFile)
strWildcard = FixHTMLString (strWildcard);
strOutput += strWildcard;
}
// get ready for next batch from beyond the digit
pStartOfGroup = ++pText;
}
break; // end of %(digit) (eg. %2)
/* -------------------------------------------------------------------- *
* %<name> *
* -------------------------------------------------------------------- */
case '<':
{
// copy up to the % sign
strOutput += CString (pStartOfGroup, pText - pStartOfGroup);
pText +=2; // skip the %<
const char * pName = pText;
// find end of wildcard name
while (*pText)
if (*pText != '>')
pText++;
else
break;
string sWildcardName (pName, pText - pName);
if (!sWildcardName.empty () && regexp)
{
CString strWildcard = FixWildcard (regexp->GetWildcard (sWildcardName),
bMakeWildcardsLower,
iSendTo,
sLanguage).c_str ();
// fix up HTML sequences if we are sending it to a log file
if (m_bLogHTML && iSendTo == eSendToLogFile)
strWildcard = FixHTMLString (strWildcard);
strOutput += strWildcard;
}
// get ready for next batch from beyond the name
if (*pText == '>')
pText++;
pStartOfGroup = pText;
}
break; // end of %<foo>
/* -------------------------------------------------------------------- *
* %C - clipboard contents *
* -------------------------------------------------------------------- */
case 'C':
case 'c':
{
// copy up to the percent sign
strOutput += CString (pStartOfGroup, pText - pStartOfGroup);
CString strClipboard;
if (GetClipboardContents (strClipboard, m_bUTF_8))
strOutput += strClipboard;
else
if (bThrowExceptions)
ThrowErrorException ("No text on the Clipboard");
// get ready for next batch from beyond the 'c'
pText += 2;
pStartOfGroup = pText;
}
break; // end of %c
/* -------------------------------------------------------------------- *
* %(something else) *
* -------------------------------------------------------------------- */
default:
pText++;
break;
} // end of switch on character after '%'
break; // end of '%(something)'
/* -------------------------------------------------------------------- *
* All other characters - just increment pointer so we can copy later *
* -------------------------------------------------------------------- */
default:
pText++;
break;
} // end of switch on *pText
} // end of not end of string yet
// copy last group
strOutput += pStartOfGroup;
return strOutput;
} // end of CMUSHclientDoc::FixSendText
int regexec(register t_regexp *prog,
register const char *string,
const int start_offset)
{
int options = App.m_bRegexpMatchEmpty ? 0 : PCRE_NOTEMPTY; // don't match on an empty string
int count;
// exit if no regexp program to process (possibly because of previous error)
if (prog->m_program == NULL)
return false;
// inspired by a suggestion by Twisol (to remove a hard-coded limit on the number of wildcards)
int capturecount = 0;
// how many captures did we get?
pcre_fullinfo(prog->m_program, NULL, PCRE_INFO_CAPTURECOUNT, &capturecount);
// allocate enough memory
vector<int> offsets ((capturecount + 1) * 3); // we always get offset 0 - the whole match
LARGE_INTEGER start,
finish;
if (App.m_iCounterFrequency)
QueryPerformanceCounter (&start);
else
{
start.QuadPart = 0;
finish.QuadPart = 0;
}
pcre_callout = NULL;
count = pcre_exec(prog->m_program, prog->m_extra, string, strlen (string),
start_offset, options, &offsets [0], offsets.size ());
if (App.m_iCounterFrequency)
{
QueryPerformanceCounter (&finish);
prog->iTimeTaken += finish.QuadPart - start.QuadPart;
}
prog->m_iMatchAttempts++; // how many times did we try to match?
if (count == PCRE_ERROR_NOMATCH)
return false; // no match - don't save matching string etc.
// otherwise free program as an indicator that we can't keep trying to do this one
if (count <= 0)
{
pcre_free (prog->m_program);
prog->m_program = NULL;
pcre_free (prog->m_extra);
prog->m_extra = NULL;
prog->m_iExecutionError = count; // remember reason
ThrowErrorException (TFormat ("Error executing regular expression: %s",
(LPCTSTR) Convert_PCRE_Runtime_Error (count)));
}
// if, and only if, we match, we will save the matching string, the count
// and offsets, so we can extract the wildcards later on
prog->m_sTarget = string; // for extracting wildcards
prog->m_iCount = count; // ditto
prog->m_vOffsets.clear ();
copy (offsets.begin (), offsets.end (), back_inserter (prog->m_vOffsets));
return true; // match
}
void CXMLparser::ProcessNode (CXMLelement & parent)
{
CString strName;
CString strValue;
CString strPrefix;
UINT iNodeStart = m_xmlPos;
UINT iFirstLine = m_xmlLine;
while (m_xmlBuff [m_xmlPos])
{
// bypass spaces and comments
SkipComments (true);
// nothing left? all done
if (m_xmlBuff [m_xmlPos] == 0)
break;
// look for element start
if (m_xmlBuff [m_xmlPos] != '<')
{
// not element - must be content
// can't have content at root level
if (parent.strName.IsEmpty ())
ThrowErrorException ("Expected '<', got \"%c\" (content not permitted here)",
m_xmlBuff [m_xmlPos]);
if (m_xmlBuff [m_xmlPos] == '\n')
m_xmlLine++; // count lines
// <![CDATA[ terminator invalid here (ie. ]]> )
if (m_xmlBuff [m_xmlPos] == ']' &&
m_xmlBuff [m_xmlPos + 1] == ']' &&
m_xmlBuff [m_xmlPos + 2] == '>')
ThrowErrorException ("CDATA terminator ']]>' invalid");
// don't let them slip in non-printables
if ((unsigned char) m_xmlBuff [m_xmlPos] < ' ' &&
(unsigned char) m_xmlBuff [m_xmlPos] != '\n' &&
(unsigned char) m_xmlBuff [m_xmlPos] != '\r')
ThrowErrorException ("Non-printable character, code %i, not permitted",
(int) m_xmlBuff [m_xmlPos]);
m_xmlPos++; // skip character
continue; // keep processing content
}
m_xmlPos++; // skip <
// special cases: <?xml blah?> <!doctype blah> </end-of-something>
if (m_xmlBuff [m_xmlPos] == '!' ||
m_xmlBuff [m_xmlPos] == '?' ||
m_xmlBuff [m_xmlPos] == '/')
{
strPrefix = m_xmlBuff [m_xmlPos];
m_xmlPos++; // bypass special types
}
else
strPrefix.Empty ();
// special and bizarre case of <![CDATA[ ... ]]>
if (strPrefix == '!' &&
m_xmlBuff [m_xmlPos] == '[' &&
m_xmlBuff [m_xmlPos + 1] == 'C' &&
m_xmlBuff [m_xmlPos + 2] == 'D' &&
m_xmlBuff [m_xmlPos + 3] == 'A' &&
m_xmlBuff [m_xmlPos + 4] == 'T' &&
m_xmlBuff [m_xmlPos + 5] == 'A' &&
m_xmlBuff [m_xmlPos + 6] == '[')
{
// can't have content at root level
if (parent.strName.IsEmpty ())
ThrowErrorException ("<![CDATA[ invalid at root level");
m_xmlPos += 7; // skip [CDATA[
while (m_xmlBuff [m_xmlPos] != ']' ||
m_xmlBuff [m_xmlPos + 1] != ']' ||
m_xmlBuff [m_xmlPos + 2] != '>')
{
// shouldn't have end-of-file before the ]]>
if (m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Unexpected end-of-file inside [CDATA[ definition");
// don't let them slip in non-printables
if ((unsigned char) m_xmlBuff [m_xmlPos] < ' ' &&
(unsigned char) m_xmlBuff [m_xmlPos] != '\n' &&
(unsigned char) m_xmlBuff [m_xmlPos] != '\r')
ThrowErrorException ("Non-printable character, code %i, not permitted",
(int) m_xmlBuff [m_xmlPos]);
if (m_xmlBuff [m_xmlPos] == '\n')
m_xmlLine++; // count lines
m_xmlPos++; // try next character
}
m_xmlPos += 3; // skip ]]>
continue; // back to start of loop
} // end of <![CDATA[ section
int iLine = m_xmlLine; // line where we expect to see name
strName = GetName ("Element");
// check for sequence </blah xxx>
if (strPrefix == '/' && m_xmlBuff [m_xmlPos] != '>')
if (m_xmlBuff [m_xmlPos] == 0)
ThrowErrorException ("Unexpected end-of-file after '</%s'",
(LPCTSTR) strName);
else
ThrowErrorException ("Surplus characters after '</%s'",
(LPCTSTR) strName);
// check for sequence </blah>
if (strPrefix == '/')
{
if (strName != parent.strName)
if (parent.strName.IsEmpty ()) // can't go up a level here :)
ThrowErrorException ("Unexpected closing tag </%s> - "
"no corresponding opening tag",
(LPCTSTR) strName);
else
ThrowErrorException ("Elements terminated out of sequence, "
"expected </%s>, got </%s>",
(LPCTSTR) parent.strName,
(LPCTSTR) strName);
AssembleContent (parent, iFirstLine, iNodeStart);
m_xmlPos++; // skip terminating >
return; // terminated this node
}
if (strPrefix == '!')
ProcessDeclaration (strName); // process declaration, eg. <!DOCTYPE
else
{ // not a declaration, must be an ordinary element
// time to create a new element
CXMLelement * pElement = new CXMLelement;
if (!pElement)
ThrowErrorException ("Could not allocate memory for XML parsing of element %s",
(LPCTSTR) strName);
// remember its name
pElement->strName = strPrefix + strName;
pElement->iLine = iLine;
parent.ChildrenList.AddTail (pElement); // add to parent's list of children
// OK, we have a new sibling for our parent, let's see if it has any attributes ...
ProcessAttributes (*pElement); // process its attributes
// hmm - we have now got something like <blah> ...
// drop down to find children for it - unless 'empty' element
if (!pElement->bEmpty)
ProcessNode (*pElement);
} // end of non-declaration
} // end of processing buffer
if (!parent.strName.IsEmpty ())
ThrowErrorException ("Unexpected end-of-file while looking for </%s>",
(LPCTSTR) parent.strName);
} // end of CXMLparser::ProcessNode
static void readfont(CString strName)
{
#define MAXFIRSTLINELEN 1000
int i,row,numsread;
inchr theord;
int maxlen,cmtlines,ffright2left;
int smush,smush2;
char *fileline,magicnum[5];
CStdioFile fontfile (strName,
CFile::modeRead|CFile::shareDenyNone|CFile::typeText);
readmagic(fontfile,magicnum);
fileline = (char*)myalloc(sizeof(char)*(MAXFIRSTLINELEN+1));
if (fontfile.ReadString(fileline,MAXFIRSTLINELEN+1)==NULL) {
fileline[0] = '\0';
}
if (MYSTRLEN(fileline)>0 ? fileline[MYSTRLEN(fileline)-1]!='\n' : 0) {
skiptoeol(fontfile);
}
numsread = sscanf(fileline,"%*c%c %d %*d %d %d %d %d %d",
&hardblank,&charheight,&maxlen,&smush,&cmtlines,
&ffright2left,&smush2);
myfree(fileline);
if (strcmp(magicnum,FONTFILEMAGICNUMBER) || numsread<5)
ThrowErrorException ("Not a FIGlet 2 font file");
for (i=1;i<=cmtlines;i++) {
skiptoeol(fontfile);
}
if (numsread<6) {
ffright2left = 0;
}
if (numsread<7) { /* if no smush2, decode smush into smush2 */
if (smush == 0) smush2 = SM_KERN;
else if (smush < 0) smush2 = 0;
else smush2 = (smush & 31) | SM_SMUSH;
}
if (charheight<1) {
charheight = 1;
}
if (maxlen<1) {
maxlen = 1;
}
maxlen += 100; /* Give ourselves some extra room */
if (smushoverride == SMO_NO)
smushmode = smush2;
else if (smushoverride == SMO_FORCE)
smushmode |= smush2;
if (right2left<0) {
right2left = ffright2left;
}
if (justification<0) {
justification = 2*right2left;
}
fileline = (char*)myalloc(sizeof(char)*(maxlen+1));
/* Allocate "missing" character */
fcharlist = (fcharnode*)myalloc(sizeof(fcharnode));
fcharlist->ord = 0;
fcharlist->thechar = (char**)myalloc(sizeof(char*)*charheight);
fcharlist->next = NULL;
for (row=0;row<charheight;row++) {
fcharlist->thechar[row] = (char*)myalloc(sizeof(char));
fcharlist->thechar[row][0] = '\0';
}
for (theord=' ';theord<='~';theord++) {
readfontchar(fontfile,theord,fileline,maxlen);
}
for (theord=0;theord<=6;theord++) {
readfontchar(fontfile,deutsch[theord],fileline,maxlen);
}
while (fontfile.ReadString(fileline,maxlen+1)==NULL?0:
sscanf(fileline,"%li",&theord)==1) {
readfontchar(fontfile,theord,fileline,maxlen);
}
myfree(fileline);
}
