void WordGenerator (CXMLElement *pCmdLine)
{
int i;
// Load input file
CString sFilespec = pCmdLine->GetAttribute(CONSTLIT("input"));
if (sFilespec.IsBlank())
{
printf("ERROR: input filename expected.\n");
return;
}
CFileReadBlock InputFile(sFilespec);
if (InputFile.Open() != NOERROR)
{
printf("ERROR: Unable to open file: %s\n", sFilespec.GetASCIIZPointer());
return;
}
// "Novel" means that we only generate words that are not
// in the input file.
bool bNovelWordsOnly = pCmdLine->GetAttributeBool(NOVEL_ATTRIB);
// Build up a word generator
CMarkovWordGenerator Generator;
TMap<CString, DWORD> InputWords;
// Read each line of the file
char *pPos = InputFile.GetPointer(0);
char *pEndPos = pPos + InputFile.GetLength();
while (pPos < pEndPos)
{
// Skip whitespace
while (pPos < pEndPos && (strIsWhitespace(pPos) || *pPos < ' '))
pPos++;
// Parse the line
char *pStart = pPos;
while (pPos < pEndPos && *pPos != '\r' && *pPos != '\n' && *pPos >= ' ')
pPos++;
CString sWord(pStart, pPos - pStart);
// Add the word to the generator
if (!sWord.IsBlank())
{
Generator.AddSample(strTrimWhitespace(sWord));
// If we are looking for novel words we need to keep a map
// of all words in the input file.
if (bNovelWordsOnly)
InputWords.Insert(sWord);
}
}
// If we have a count, then output a list of random words
int iCount;
if (pCmdLine->FindAttributeInteger(COUNT_ATTRIB, &iCount))
{
if (iCount > 0)
{
TArray<CString> Result;
Generator.GenerateUnique(iCount, &Result);
for (i = 0; i < Result.GetCount(); i++)
if (InputWords.Find(Result[i]))
{
Result.Delete(i);
i--;
}
Result.Sort();
for (i = 0; i < Result.GetCount(); i++)
printf("%s\n", Result[i].GetASCIIZPointer());
}
}
// Otherwise, output the generator as XML
else
{
CMemoryWriteStream Output;
if (Output.Create() != NOERROR)
{
printf("ERROR: Out of memory.\n");
return;
}
if (Generator.WriteAsXML(&Output) != NOERROR)
{
printf("ERROR: Unable to output generator as XML.\n");
return;
}
Output.Write("\0", 1);
printf(Output.GetPointer());
}
}
bool CHexeMarkupEvaluator::ProcessHeader (SHTTPRequestCtx &Ctx, CDatum dResult)
// ProcessHeader
//
// Outputs the given header
{
// Check for error
if (dResult.IsError())
{
m_Output.Write(strPattern(ERR_PROCESSING_HEADER, dResult.AsString()));
return true;
}
// Processing depends on result type
switch (dResult.GetBasicType())
{
case CDatum::typeNil:
return true;
// If this is a string or anything else, we expect both field and value
// are in the same string and we need to parse it.
default:
{
CString sData = dResult.AsString();
// Parse into field and value
char *pPos = sData.GetParsePointer();
while (strIsWhitespace(pPos))
pPos++;
// Look for the field name
char *pStart = pPos;
while (*pPos != ':' && *pPos != '\0')
pPos++;
CString sField(pStart, pPos - pStart);
if (sField.IsEmpty())
{
m_Output.Write(strPattern(ERR_NO_HEADER_FIELD, sData));
return true;
}
// Look for the value
CString sValue;
if (*pPos == ':')
{
pPos++;
while (strIsWhitespace(pPos))
pPos++;
sValue = CString(pPos);
}
// Done
CHTTPMessage::SHeader *pNewHeader = m_Headers.Insert();
pNewHeader->sField = sField;
pNewHeader->sValue = sValue;
}
}
return true;
}
bool CTimeDate::Parse (const CString &sFormat, const CString &sValue, CString *retsError)
// Parse
//
// Parse a date of the given format.
{
// Internet format:
//
// Sun, 06 Nov 1994 08:49:37 GMT
if (strEquals(sFormat, FORMAT_INTERNET))
{
char *pPos = sValue.GetASCIIZPointer();
char *pPosEnd = pPos + sValue.GetLength();
// Skip leading whitespace
while (strIsWhitespace(pPos))
pPos++;
// Skip day of week
while (pPos < pPosEnd
&& (*pPos < '0' || *pPos > '9'))
pPos++;
if (pPos >= pPosEnd)
goto InvalidValue;
// Day
bool bFail;
m_Time.wDay = strParseInt(pPos, 0, &pPos, &bFail);
if (bFail || m_Time.wDay < 1 || m_Time.wDay > 31)
goto InvalidValue;
pPos++;
if (pPos >= pPosEnd)
goto InvalidValue;
// Month
if (*pPos == 'A')
{
pPos++;
if (*pPos == 'p')
{
m_Time.wMonth = 4;
pPos += 3;
}
else if (*pPos == 'u')
{
m_Time.wMonth = 8;
pPos += 3;
}
else
goto InvalidValue;
}
else if (*pPos == 'D')
{
m_Time.wMonth = 12;
pPos += 4;
}
else if (*pPos == 'F')
{
m_Time.wMonth = 2;
pPos += 4;
}
else if (*pPos == 'J')
{
pPos++;
if (*pPos == 'a')
{
m_Time.wMonth = 1;
pPos += 3;
}
else if (*pPos == 'u')
{
pPos++;
if (*pPos == 'l')
{
m_Time.wMonth = 7;
pPos += 2;
}
else if (*pPos == 'n')
{
m_Time.wMonth = 6;
pPos += 2;
}
else
goto InvalidValue;
}
else
goto InvalidValue;
}
else if (*pPos == 'M')
{
pPos++;
if (*pPos == 'a')
{
pPos++;
if (*pPos == 'r')
{
m_Time.wMonth = 3;
pPos += 2;
}
else if (*pPos == 'y')
{
m_Time.wMonth = 5;
pPos += 2;
}
else
goto InvalidValue;
}
else
goto InvalidValue;
}
else if (*pPos == 'N')
{
m_Time.wMonth = 11;
pPos += 4;
}
else if (*pPos == 'O')
{
m_Time.wMonth = 10;
pPos += 4;
}
else if (*pPos == 'S')
{
m_Time.wMonth = 9;
pPos += 4;
}
else
goto InvalidValue;
if (pPos >= pPosEnd)
goto InvalidValue;
// Year
m_Time.wYear = strParseInt(pPos, 0, &pPos, &bFail);
if (bFail || m_Time.wYear < 1)
goto InvalidValue;
pPos++;
if (pPos >= pPosEnd)
goto InvalidValue;
// Hour
m_Time.wHour = strParseInt(pPos, 0, &pPos, &bFail);
if (bFail || m_Time.wHour > 23)
goto InvalidValue;
pPos++;
if (pPos >= pPosEnd)
goto InvalidValue;
// Minute
m_Time.wMinute = strParseInt(pPos, 0, &pPos, &bFail);
if (bFail || m_Time.wMinute > 59)
goto InvalidValue;
pPos++;
if (pPos >= pPosEnd)
goto InvalidValue;
// Second
m_Time.wSecond = strParseInt(pPos, 0, &pPos, &bFail);
if (bFail || m_Time.wSecond > 59)
goto InvalidValue;
// Millisecond is not stored
m_Time.wMilliseconds = 0;
// We don't store day of week
m_Time.wDayOfWeek = 0xffff;
// Done
return true;
InvalidValue:
if (retsError) *retsError = strPatternSubst(CONSTLIT("Invalid date value: %s"), sValue);
return false;
}
else
{
if (retsError) *retsError = strPatternSubst(CONSTLIT("Unsupported date format: %s."), sFormat);
return false;
}
}
inline bool IsDelimiterChar (char *pPos, bool bIsSpecialAttrib = false) { return (*pPos == '\0' || *pPos == ',' || *pPos == ';' || (!bIsSpecialAttrib && strIsWhitespace(pPos))); }
