// @pymethod |PyIFilter|GetText|Description of GetText.
PyObject *PyIFilter::GetText(PyObject *self, PyObject *args)
{
IFilter *pIF = GetI(self);
if ( pIF == NULL )
return NULL;
// @pyparm <int>|nBufSize|size of text buffer to create
ULONG nBufSize = 0;
if ( !PyArg_ParseTuple(args, "|i:GetText", &nBufSize) )
return NULL;
HRESULT hr;
if (nBufSize == 0)
nBufSize = 8192; // 8k default
WCHAR *wBuffer = (WCHAR *)PyMem_Malloc((nBufSize+1)*sizeof(WCHAR));
if (!wBuffer){
PyErr_SetString(PyExc_MemoryError, "getting text");
return NULL;
}
PY_INTERFACE_PRECALL;
hr = pIF->GetText( &nBufSize, wBuffer );
PY_INTERFACE_POSTCALL;
if ( FAILED(hr) ) {
PyMem_Free(wBuffer);
return PyCom_BuildPyException(hr, pIF, IID_IFilter );
}
PyObject *obRet = PyWinObject_FromWCHAR(wBuffer, nBufSize);
PyMem_Free(wBuffer);
return obRet;
}
HRESULT Analyze(wchar_t* szPath) {
HRESULT hr = S_OK;
// Load the IFilter associated with the specified file
IFilter* pFilter;
hr = LoadIFilter(szPath, NULL, (void**)&pFilter);
if (SUCCEEDED(hr)) {
// Initialize the IFilter
DWORD dwFlags = 0;
hr = pFilter->Init(FILTER_INIT_OPTIONS,0,NULL,&dwFlags);
if (SUCCEEDED(hr)) {
wchar_t szBuffer[BUFLEN];
ULONG ulSize;
STAT_CHUNK ps;
while (SUCCEEDED(hr))
{
// Retrieve the next chunk in the document
hr = pFilter->GetChunk(&ps);
if ( (FILTER_E_EMBEDDING_UNAVAILABLE == hr) || (FILTER_E_LINK_UNAVAILABLE == hr) ) {
hr = S_OK;
continue;
} else if (FILTER_E_END_OF_CHUNKS == hr) {
hr = S_OK;
break;
}
while(SUCCEEDED(hr)) {
// Retrieve the next block of text in the current chunk
ulSize = BUFLEN;
hr = pFilter->GetText(&ulSize, szBuffer);
if ( (FILTER_E_NO_TEXT == hr) || (FILTER_E_NO_MORE_TEXT == hr) ) {
hr = S_OK;
break;
}
if (SUCCEEDED(hr) && (0 < ulSize)) {
szBuffer[ulSize] = '\0';
// Convert to UTF8
unsigned int cbMultiByte = WideCharToMultiByte(CP_UTF8, NULL, szBuffer, -1, NULL, 0, NULL, NULL);
if (0 == cbMultiByte) {
hr = E_FAIL;
tcerr << "WideCharToMultiByte#1 invocation failed" << endl;
errorMessagePrinted = true;
} else {
char* pchMultiByte = new char[cbMultiByte];
if (NULL == pchMultiByte) {
hr = E_OUTOFMEMORY;
} else {
if (0 == WideCharToMultiByte(CP_UTF8, NULL, szBuffer, -1, pchMultiByte, cbMultiByte, NULL, NULL)) {
hr = E_FAIL;
tcerr << "WideCharToMultiByte#2 invocation failed" << endl;
errorMessagePrinted = true;
} else {
// Write the UTF8 text to stdout
if (cbMultiByte > fwrite(pchMultiByte, 1, cbMultiByte, stdout)) {
hr = E_FAIL;
tcerr << "Unable to write converted bytes to output" << endl;
errorMessagePrinted = true;
}
}
delete[] pchMultiByte;
}
}
}
}
}
} else {
tcerr << "IFilter initialization failed with HRESULT " << hr << endl;
errorMessagePrinted = true;
}
pFilter->Release();
} else {
tcerr << "IFilter loading failed with HRESULT " << hr << endl;
errorMessagePrinted = true;
}
return hr;
}
signed char
IFilterEndAnalyzer::analyze(AnalysisResult& idx, InputStream *in) {
const string& filename = idx.fileName();
int p = filename.find_last_of('.');
if (p < 0 || extensions.find(filename.substr(p)) == extensions.end()) {
return -1;
}
string filepath;
bool fileisondisk = checkForFile(idx.depth(), filename);
if (fileisondisk) {
filepath = filename;
} else {
int p = filename.find_last_of(".");
if ( p > 0 ){
string ext = filename.substr(p).c_str();
strlwr((char*)ext.c_str());
p = ext.find_first_not_of("._abcdefghijklmnopqrstuvwxyz0123456789");
if ( p >= 0 )
filepath = writeToTempFile(in, "");
else
filepath = writeToTempFile(in, ext.c_str());
}else
filepath = writeToTempFile(in, "");
}
if (filepath.length() > 0) {
IFilter* filter = NULL;
void* pvfilter=NULL;
wchar_t tmp[MAX_PATH];
_cpycharToWide(tmp,filepath.c_str(),MAX_PATH);
HRESULT hr = LoadIFilter(tmp,NULL,&pvfilter);
if (hr == S_OK) {
filter = (IFilter*)pvfilter;
ULONG __i=0;
hr = filter->Init(IFILTER_INIT_APPLY_INDEX_ATTRIBUTES,0,NULL,&__i);
if (FAILED( hr )) {
if (!fileisondisk)
unlink(filepath.c_str());
return -1;
}
const int sbBufferLen = 1024;
wchar_t sbBuffer[sbBufferLen];
STAT_CHUNK ps;
hr = filter->GetChunk(&ps);
while ( SUCCEEDED(hr) ) {
if (ps.flags == CHUNK_TEXT) {
int resultText = 0;
while ( resultText >= 0 ) {
ULONG sizeBuffer=sbBufferLen;
resultText = filter->GetText(&sizeBuffer, sbBuffer);
if (sizeBuffer > 0 ) {
string str = wchartoutf8(sbBuffer,sbBuffer+sizeBuffer);
idx.addText(str.c_str(),str.length());
}
}
} else if ( ps.flags == CHUNK_VALUE ) {
PROPVARIANT *pVar;
while ( SUCCEEDED( hr = filter->GetValue( &pVar ) ) ) {
//printf("propid: %d\nkind:%d\n",ps.attribute.psProperty.propid,ps.attribute.psProperty.ulKind);
if ( ps.attribute.psProperty.propid == 2 &&
ps.attribute.psProperty.ulKind == 1 &&
pVar->vt == VT_LPWSTR ) {
string str = wchartoutf8(pVar->pwszVal,pVar->pwszVal+wcslen(pVar->pwszVal));
idx.addValue("title", str );
}
PropVariantClear( pVar );
CoTaskMemFree( pVar );
}
} else {
printf("other flag %d\n",ps.flags);
}
hr = filter->GetChunk(&ps);
}
filter->Release();
if (!fileisondisk)
unlink(filepath.c_str());
return 0;
}
DWORD dw = GetLastError();
if ( dw != 0 ) {
LPVOID lpMsgBuf;
FormatMessage(
FORMAT_MESSAGE_ALLOCATE_BUFFER |
FORMAT_MESSAGE_FROM_SYSTEM,
NULL,
dw,
MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT),
(LPTSTR) &lpMsgBuf,
0, NULL );
wprintf(L"%s\n", lpMsgBuf);
LocalFree(lpMsgBuf);
}
}
if (!fileisondisk && filepath.length()>0) {
unlink(filepath.c_str());
}
return -1;
}