diff options
| author | George Hazan <george.hazan@gmail.com> | 2015-05-23 14:55:03 +0000 |
|---|---|---|
| committer | George Hazan <george.hazan@gmail.com> | 2015-05-23 14:55:03 +0000 |
| commit | 7060a1f3dcde4355124316e6aab3a4e336812155 (patch) | |
| tree | c90151cb9386abce03de741f271c3751b3a7eb6f /protocols/IcqOscarJ/src/fam_04message.cpp | |
| parent | bcdf15018dd56e91974d3723d2a7013ff3dcd699 (diff) | |
more utf fixes
git-svn-id: http://svn.miranda-ng.org/main/trunk@13775 1316c22d-e87f-b044-9b9b-93d7a3e3ba9c
Diffstat (limited to 'protocols/IcqOscarJ/src/fam_04message.cpp')
| -rw-r--r-- | protocols/IcqOscarJ/src/fam_04message.cpp | 36 |
1 files changed, 20 insertions, 16 deletions
diff --git a/protocols/IcqOscarJ/src/fam_04message.cpp b/protocols/IcqOscarJ/src/fam_04message.cpp index a9ca209a1e..3dc6b83c19 100644 --- a/protocols/IcqOscarJ/src/fam_04message.cpp +++ b/protocols/IcqOscarJ/src/fam_04message.cpp @@ -241,6 +241,7 @@ void CIcqProto::handleRecvServMsgType1(BYTE *buf, size_t wLen, DWORD dwUin, char // different encodings (just like the new format of Offline Messages).
DWORD dwRecvTime;
char* szMsg = NULL;
+ bool bUtf8 = false;
PROTORECVEVENT pre = { 0 };
int bAdded;
@@ -249,27 +250,23 @@ void CIcqProto::handleRecvServMsgType1(BYTE *buf, size_t wLen, DWORD dwUin, char WORD wMsgPart = 1;
while (oscar_tlv *pMessageTLV = pChain->getTLV(0x0101, wMsgPart)) { // Loop thru all message parts
if (pMessageTLV->wLen > 4) {
- WORD wMsgLen;
- BYTE *pMsgBuf;
- WORD wEncoding;
- WORD wCodePage;
char *szMsgPart = NULL;
- int bMsgPartUnicode = FALSE;
+ bool bMsgPartUnicode = false;
// The message begins with a encoding specification
// The first WORD is believed to have the following meaning:
// 0x00: US-ASCII
// 0x02: Unicode UCS-2 Big Endian encoding
// 0x03: local 8bit encoding
- pMsgBuf = pMessageTLV->pData;
+ BYTE *pMsgBuf = pMessageTLV->pData;
+ WORD wEncoding, wCodePage;
unpackWord(&pMsgBuf, &wEncoding);
unpackWord(&pMsgBuf, &wCodePage);
- wMsgLen = pMessageTLV->wLen - 4;
+ WORD wMsgLen = pMessageTLV->wLen - 4;
debugLogA("Message (format 1) - Part %d: Encoding is 0x%X, page is 0x%X", wMsgPart, wEncoding, wCodePage);
switch (wEncoding) {
-
case 2: // UCS-2
{
WCHAR *usMsgPart = (WCHAR*)SAFE_MALLOC(wMsgLen + 2);
@@ -279,7 +276,7 @@ void CIcqProto::handleRecvServMsgType1(BYTE *buf, size_t wLen, DWORD dwUin, char szMsgPart = make_utf8_string(usMsgPart);
if (!IsUSASCII(szMsgPart, mir_strlen(szMsgPart)))
- bMsgPartUnicode = TRUE;
+ bMsgPartUnicode = true;
SAFE_FREE(&usMsgPart);
}
break;
@@ -291,21 +288,21 @@ void CIcqProto::handleRecvServMsgType1(BYTE *buf, size_t wLen, DWORD dwUin, char szMsgPart = (char*)SAFE_MALLOC(wMsgLen + 1);
memcpy(szMsgPart, pMsgBuf, wMsgLen);
szMsgPart[wMsgLen] = '\0';
-
break;
}
+
// Check if the new part is compatible with the message
- if (!pre.flags && bMsgPartUnicode) { // make the resulting message utf-8 encoded - need to append utf-8 encoded part
+ if (!bUtf8 && bMsgPartUnicode) { // make the resulting message utf-8 encoded - need to append utf-8 encoded part
if (szMsg) { // not necessary to convert - appending first part, only set flags
char *szUtfMsg = ansi_to_utf8_codepage(szMsg, getWord(hContact, "CodePage", m_wAnsiCodepage));
SAFE_FREE(&szMsg);
szMsg = szUtfMsg;
}
+ bUtf8 = true;
}
- if (!bMsgPartUnicode) { // convert message part to utf-8 and append
+ if (!bMsgPartUnicode && bUtf8) { // convert message part to utf-8 and append
char *szUtfPart = ansi_to_utf8_codepage((char*)szMsgPart, getWord(hContact, "CodePage", m_wAnsiCodepage));
-
SAFE_FREE(&szMsgPart);
szMsgPart = szUtfPart;
}
@@ -321,11 +318,12 @@ void CIcqProto::handleRecvServMsgType1(BYTE *buf, size_t wLen, DWORD dwUin, char if (_strnicmp(szMsg, "<html>", 6) == 0) // strip HTML formating from AIM message
szMsg = EliminateHtml(szMsg, mir_strlen(szMsg));
- if (!pre.flags && !IsUSASCII(szMsg, mir_strlen(szMsg))) { // message is Ansi and contains national characters, create Unicode part by codepage
+ if (!bUtf8 && !IsUSASCII(szMsg, mir_strlen(szMsg))) { // message is Ansi and contains national characters, create Unicode part by codepage
char *usMsg = convertMsgToUserSpecificUtf(hContact, szMsg);
if (usMsg) {
SAFE_FREE(&szMsg);
szMsg = usMsg;
+ bUtf8 = true;
}
}
@@ -1529,6 +1527,7 @@ void CIcqProto::handleMessageTypes(DWORD dwUin, char *szUID, DWORD dwTimestamp, case MTYPE_PLAIN: /* plain message */
{
PROTORECVEVENT pre = { 0 };
+ bool bUtf8 = false;
// Check if this message is marked as UTF8 encoded
if (dwDataLen > 12) {
@@ -1545,6 +1544,10 @@ void CIcqProto::handleMessageTypes(DWORD dwUin, char *szUID, DWORD dwTimestamp, usMsg[dwExtraLen] = '\0';
SAFE_FREE(&szMsg);
szMsg = (char*)make_utf8_string(usMsg);
+
+ if (!IsUnicodeAscii(usMsg, dwExtraLen))
+ bUtf8 = true; // only mark real non-ascii messages as unicode
+
bDoubleMsg = 1;
}
}
@@ -1557,6 +1560,7 @@ void CIcqProto::handleMessageTypes(DWORD dwUin, char *szUID, DWORD dwTimestamp, while ((dwGuidLen >= 38) && (dwDataLen >= dwGuidLen)) {
if (!strncmp(pMsg, CAP_UTF8MSGS, 38)) { // Found UTF8 cap, convert message to ansi
+ bUtf8 = true;
break;
}
else if (!strncmp(pMsg, CAP_RTFMSGS, 38)) { // Found RichText cap
@@ -1571,9 +1575,9 @@ void CIcqProto::handleMessageTypes(DWORD dwUin, char *szUID, DWORD dwTimestamp, }
hContact = HContactFromUIN(dwUin, &bAdded);
- sendMessageTypesAck(hContact, true, pAckParams);
+ sendMessageTypesAck(hContact, bUtf8, pAckParams);
- if (!pre.flags && !IsUSASCII(szMsg, mir_strlen(szMsg))) { // message is Ansi and contains national characters, create Unicode part by codepage
+ if (!bUtf8 && !IsUSASCII(szMsg, mir_strlen(szMsg))) { // message is Ansi and contains national characters, create Unicode part by codepage
char *usMsg = convertMsgToUserSpecificUtf(hContact, szMsg);
if (usMsg) {
SAFE_FREE(&szMsg);
|