[PATCH 34/34] msxml3: skip a possible BOM in doparse

[Marcus Meissner]

Hi,

Issue encountered in Corel X4 installer. (see bug 11747),
it gets a UTF16 document with BOM in front.

trace:msxml:domdoc_load (0x131d10)->({VT_BSTR: L"C:\\users\\marcus\\Temp\\{180A87F9-82B9-42A6-8F0E-3AC30A58AC8C}\\{5A51C731-A877-4C6E-8D51-CD16CDC46C9B}\\CorelDRAW Graphics Suite X4 Setup Files\\Setup.xml"})
trace:msxml:create_moniker_from_url L"C:\\users\\marcus\\Temp\\{180A87F9-82B9-42A6-8F0E-3AC30A58AC8C}\\{5A51C731-A877-4C6E-8D51-CD16CDC46C9B}\\CorelDRAW Graphics Suite X4 Setup Files\\Setup.xml"
trace:msxml:bind_url 0x131570
trace:msxml:bsc_AddRef (0x138da0) ref=2
trace:msxml:bsc_QueryInterface interface {6d5140c1-7436-11ce-8034-00aa006009fa} not implemented
trace:msxml:bsc_QueryInterface interface {aaa74ef9-8ee7-4659-88d9-f8c504da73cc} not implemented
trace:msxml:bsc_OnStartBinding (0x138da0)->(ff 0x13a238)
trace:msxml:bsc_QueryInterface interface {79eac9e4-baf9-11ce-8c82-00aa004ba90b} not implemented
trace:msxml:bsc_OnDataAvailable (0x138da0)->(5 36572 0x33e908 0x33e8fc)
trace:msxml:bsc_OnStopBinding (0x138da0)->(00000000 (null))
fixme:msxml:doparse start of xml is ffffffff fffffffe 3c 00 49 00 43

Bug is only fixed partially by this.

Ciao, Marcus
---
 dlls/msxml3/domdoc.c |   10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/dlls/msxml3/domdoc.c b/dlls/msxml3/domdoc.c
index 49e6168..3504204 100644
--- a/dlls/msxml3/domdoc.c
+++ b/dlls/msxml3/domdoc.c
@@ -486,6 +486,16 @@ static xmlDocPtr doparse(domdoc* This, char const* ptr, int len, xmlCharEncoding
         sax_serror                      /* serror */
     };
 
+    /* UTF-16 BOM at start of data */
+    if ((len > 2) && (ptr[0] == (char)0xff)  && (ptr[1] == (char)0xfe)) {
+        ptr += 2;
+        len -= 2;
+        if (encoding == XML_CHAR_ENCODING_NONE)
+            encoding = XML_CHAR_ENCODING_UTF16LE;
+        else
+            FIXME("Not changing xml encoding type from %d to XML_CHAR_ENCODING_UTF16LE.\n", encoding);
+    }
+
     pctx = xmlCreateMemoryParserCtxt(ptr, len);
     if (!pctx)
     {
-- 
1.7.10.4