/*!
Detect encoding of text buffer
\param pData Buffer
\param lenData Length of buffer
Supports UTF-8 and UCS-2 big and little endian
CJK encodings could be added
*/
UT_Error IE_Imp_Text::_recognizeEncoding(const char *szBuf, UT_uint32 iNumbytes)
{
if (IE_Imp_Text_Sniffer::_recognizeUTF8(szBuf, iNumbytes))
_setEncoding("UTF-8");
else
{
IE_Imp_Text_Sniffer::UCS2_Endian eUcs2 = IE_Imp_Text_Sniffer::UE_NotUCS;
eUcs2 = IE_Imp_Text_Sniffer::_recognizeUCS2(szBuf, iNumbytes, false);
if (eUcs2 == IE_Imp_Text_Sniffer::UE_BigEnd)
_setEncoding(XAP_EncodingManager::get_instance()->getUCS2BEName());
else if (eUcs2 == IE_Imp_Text_Sniffer::UE_LittleEnd)
_setEncoding(XAP_EncodingManager::get_instance()->getUCS2LEName());
else
_setEncoding(
#ifdef TOOLKIT_WIN
XAP_EncodingManager::get_instance()->getNative8BitEncodingName()
#else
"ISO-8859-1"
#endif
);
}
return UT_OK;
}
void XAP_UnixDialog_Encoding::event_Ok(void)
{
GtkTreeSelection * selection;
GtkTreeIter iter;
GtkTreeModel * model;
gint row = 0;
selection = gtk_tree_view_get_selection( GTK_TREE_VIEW(m_listEncodings) );
// if there is no selection, or the selection's data (GtkListItem widget)
// is empty, return cancel. GTK can make this happen.
if ( !selection ||
!gtk_tree_selection_get_selected (selection, &model, &iter)
)
{
_setAnswer (XAP_Dialog_Encoding::a_CANCEL);
return;
}
// get the ID of the selected Type
gtk_tree_model_get (model, &iter, 1, &row, -1);
if (row >= 0) {
_setSelectionIndex(static_cast<UT_uint32>(row));
_setEncoding (_getAllEncodings()[row]);
_setAnswer (XAP_Dialog_Encoding::a_OK);
} else {
UT_ASSERT_NOT_REACHED();
_setAnswer (XAP_Dialog_Encoding::a_CANCEL);
}
}
IE_Exp_Text::IE_Exp_Text(PD_Document * pDocument, bool bEncoded)
: IE_Exp(pDocument),
m_pListener(NULL),
m_bIsEncoded(false),
m_szEncoding(0),
m_bExplicitlySetEncoding(false),
m_bIs16Bit(false),
m_bUnicode(false),
m_bBigEndian(false),
m_bUseBOM(false)
{
m_error = UT_OK;
// Get encoding dialog prefs setting
bool bAlwaysPrompt = false;
XAP_App::getApp()->getPrefsValueBool(AP_PREF_KEY_AlwaysPromptEncoding, &bAlwaysPrompt);
m_bIsEncoded = bAlwaysPrompt | bEncoded;
const char *szEncodingName = pDocument->getEncodingName();
if (!szEncodingName || !*szEncodingName)
szEncodingName = XAP_EncodingManager::get_instance()->getNativeEncodingName();
_setEncoding(szEncodingName);
}
UT_Error IE_Exp_Text::_writeDocument(void)
{
// Don't call base method if user cancels encoding dialog
if (!(!m_bIsEncoded || m_bExplicitlySetEncoding || _doEncodingDialog(m_szEncoding)))
return UT_SAVE_CANCELLED;
// TODO If we're going to the clipboard and the OS supports unicode, set encoding.
// TODO Only supports Windows so far.
// TODO Should use a finer-grain technique than IsWinNT() since Win98 supports unicode clipboard.
if (getDocRange())
{
#ifdef WIN32
if (UT_IsWinNT())
_setEncoding(XAP_EncodingManager::get_instance()->getNativeUnicodeEncodingName());
#endif
}
m_pListener = _constructListener();
if (!m_pListener)
return UT_IE_NOMEMORY;
if (getDocRange())
getDoc()->tellListenerSubset(static_cast<PL_Listener *>(m_pListener),getDocRange());
else
getDoc()->tellListener(static_cast<PL_Listener *>(m_pListener));
DELETEP(m_pListener);
return ((m_error) ? UT_IE_COULDNOTWRITE : UT_OK);
}
PL_Listener * IE_Exp_Text::_constructListener(void)
{
if (!m_bExplicitlySetEncoding) {
const std::string & prop = getProperty ("encoding");
if (!prop.empty()) {
_setEncoding (prop.c_str());
}
}
return new Text_Listener(getDoc(),this,(getDocRange()!=NULL),m_szEncoding,
m_bIs16Bit,m_bUnicode,m_bUseBOM,m_bBigEndian);
}
bool IE_Imp_Text::pasteFromBuffer(PD_DocumentRange * pDocRange,
const unsigned char * pData, UT_uint32 lenData,
const char *szEncoding)
{
UT_return_val_if_fail(getDoc() == pDocRange->m_pDoc,false);
UT_return_val_if_fail(pDocRange->m_pos1 == pDocRange->m_pos2,false);
// Attempt to guess whether we're pasting 8 bit or unicode text
if (szEncoding)
_setEncoding(szEncoding);
else
_recognizeEncoding(reinterpret_cast<const char *>(pData), lenData);
ImportStreamClipboard stream(pData, lenData);
setClipboard (pDocRange->m_pos1);
_parseStream(&stream);
return true;
}
IE_Imp_Text::IE_Imp_Text(PD_Document * pDocument, const char * encoding)
: IE_Imp(pDocument),
m_szEncoding(0),
m_bExplicitlySetEncoding(false),
m_bIsEncoded(false),
m_bIs16Bit(false),
m_bUseBOM(false),
m_bBigEndian(false),
m_bBlockDirectionPending(true),
m_bFirstBlockData(true),
m_pBlock(0)
{
m_bIsEncoded = ((encoding != NULL) && (strlen(encoding) > 0));
if ( m_bIsEncoded )
{
m_bExplicitlySetEncoding = true ;
_setEncoding(encoding);
}
}
IE_Exp_Text::IE_Exp_Text(PD_Document * pDocument, const char * encoding)
: IE_Exp(pDocument),
m_pListener(NULL),
m_bIsEncoded(false),
m_szEncoding(0),
m_bExplicitlySetEncoding(false),
m_bIs16Bit(false),
m_bUnicode(false),
m_bBigEndian(false),
m_bUseBOM(false)
{
m_error = UT_OK;
m_bIsEncoded = ((encoding != NULL) && (strlen(encoding) > 0));
if ( m_bIsEncoded )
{
m_bExplicitlySetEncoding = true;
_setEncoding(encoding);
}
}
/*!
Request file encoding from user
This function should be identical to the one in ie_Exp_Text
*/
bool IE_Imp_Text::_doEncodingDialog(const char *szEncoding)
{
XAP_Dialog_Id id = XAP_DIALOG_ID_ENCODING;
XAP_DialogFactory * pDialogFactory
= static_cast<XAP_DialogFactory *>(XAP_App::getApp()->getDialogFactory());
XAP_Dialog_Encoding * pDialog
= static_cast<XAP_Dialog_Encoding *>(pDialogFactory->requestDialog(id));
UT_return_val_if_fail(pDialog, false);
pDialog->setEncoding(szEncoding);
// run the dialog
XAP_Frame * pFrame = XAP_App::getApp()->getLastFocussedFrame();
UT_return_val_if_fail(pFrame, false);
pDialog->runModal(pFrame);
// extract what they did
bool bOK = (pDialog->getAnswer() == XAP_Dialog_Encoding::a_OK);
if (bOK)
{
const gchar * s;
static gchar szEnc[16];
s = pDialog->getEncoding();
UT_return_val_if_fail (s, false);
strcpy(szEnc,s);
_setEncoding(static_cast<const char *>(szEnc));
getDoc()->setEncodingName(szEnc);
}
pDialogFactory->releaseDialog(pDialog);
return bOK;
}
/*
Construct text importer
\param pDocument Document to import text into
\param bEncoded True if we should show encoding dialog
Uses current document's encoding if it is set
*/
IE_Imp_Text::IE_Imp_Text(PD_Document * pDocument, bool bEncoded)
: IE_Imp(pDocument),
m_szEncoding(0),
m_bExplicitlySetEncoding(false),
m_bIsEncoded(false),
m_bIs16Bit(false),
m_bUseBOM(false),
m_bBigEndian(false),
m_bBlockDirectionPending(true),
m_bFirstBlockData(true),
m_pBlock(NULL)
{
// Get encoding dialog prefs setting
bool bAlwaysPrompt;
XAP_App::getApp()->getPrefsValueBool(AP_PREF_KEY_AlwaysPromptEncoding, &bAlwaysPrompt);
m_bIsEncoded = bAlwaysPrompt | bEncoded;
const char *szEncodingName = pDocument->getEncodingName();
if (!szEncodingName || !*szEncodingName)
szEncodingName = XAP_EncodingManager::get_instance()->getNativeEncodingName();
_setEncoding(szEncodingName);
}
/*!
Parse stream contents into the document
\param stream Stream to import from
This code is used for both files and the clipboard
*/
UT_Error IE_Imp_Text::_parseStream(ImportStream * pStream)
{
UT_return_val_if_fail(pStream, UT_ERROR);
bool bFirstChar = true;
UT_GrowBuf gbBlock(1024);
UT_UCSChar c;
if (!m_bExplicitlySetEncoding) {
std::string prop;
prop = getProperty ("encoding");
if (!prop.empty()) {
_setEncoding (prop.c_str());
}
}
pStream->init(m_szEncoding);
while (pStream->getChar(c))
{
// TODO We should switch fonts when we encounter
// TODO characters from different scripts
switch (c)
{
case UCS_CR:
case UCS_LF:
case UCS_LINESEP:
case UCS_PARASEP:
// we interpret either CRLF, CR, or LF as a paragraph break.
// we also accept U+2028 (line separator) and U+2029 (para separator)
// especially since these are recommended by Mac OS X.
// flush out what we have
if (gbBlock.getLength() > 0)
X_ReturnNoMemIfError(_insertSpan(gbBlock));
X_ReturnNoMemIfError(_insertBlock());
break;
case UCS_BOM:
// This is Byte Order Mark at the start of file, Zero Width Non Joiner elsewhere
if (bFirstChar)
break;
// if we encounter any of the following characters we will
// substitute a '?' as they correspond to control characters,
// though some text files use them for their character representations
// We do this instead of of immediately returning an error
// (and assuming they have no business in a text file) so we can
// still show usable text to a user who has one of these files.
case 0x0000:
case 0x0001:
case 0x0002:
case 0x0003:
case 0x0004:
case 0x0005:
case 0x0006:
case 0x0007:
case 0x0008:
case 0x000e:
case 0x000f:
case 0x0010:
case 0x0011:
case 0x0012:
case 0x0013:
case 0x0014:
case 0x0015:
case 0x0016:
case 0x0017:
case 0x0018:
case 0x0019:
case 0x001a:
case 0x001b:
case 0x001c:
case 0x001d:
case 0x001e:
case 0x001f:
// UT_ASSERT(!(c <= 0x001f));
c = '?';
/* return UT_ERROR; // fall through with modified character */
default:
X_ReturnNoMemIfError(gbBlock.append(reinterpret_cast<UT_GrowBufElement*>(&c),1));
break;
}
bFirstChar = false;
}
if (gbBlock.getLength() > 0)
X_ReturnNoMemIfError(_insertSpan(gbBlock));
return UT_OK;
}
