peiyongz 2003/01/28 10:31:47
Modified: c/src/xercesc/dom/impl DOMWriterImpl.hpp DOMWriterImpl.cpp
Log:
Bug#13694: Allow Xerces to write the BOM to XML files
Revision Changes Path
1.12 +9 -1 xml-xerces/c/src/xercesc/dom/impl/DOMWriterImpl.hpp
Index: DOMWriterImpl.hpp
===================================================================
RCS file: /home/cvs/xml-xerces/c/src/xercesc/dom/impl/DOMWriterImpl.hpp,v
retrieving revision 1.11
retrieving revision 1.12
diff -u -r1.11 -r1.12
--- DOMWriterImpl.hpp 20 Jan 2003 16:50:13 -0000 1.11
+++ DOMWriterImpl.hpp 28 Jan 2003 18:31:47 -0000 1.12
@@ -57,6 +57,9 @@
/*
* $Id$
* $Log$
+ * Revision 1.12 2003/01/28 18:31:47 peiyongz
+ * Bug#13694: Allow Xerces to write the BOM to XML files
+ *
* Revision 1.11 2003/01/20 16:50:13 tng
* DOMWriter fix:
* 1. wrong wrong nested cdata message
@@ -415,6 +418,8 @@
void printIndent(int level) const;
//does the actual work for processNode while keeping track of the level
void processNode(const DOMNode* const nodeToWrite, int level);
+
+ void processBOM();
// -----------------------------------------------------------------------
// Private data members
1.29 +74 -4 xml-xerces/c/src/xercesc/dom/impl/DOMWriterImpl.cpp
Index: DOMWriterImpl.cpp
===================================================================
RCS file: /home/cvs/xml-xerces/c/src/xercesc/dom/impl/DOMWriterImpl.cpp,v
retrieving revision 1.28
retrieving revision 1.29
diff -u -r1.28 -r1.29
--- DOMWriterImpl.cpp 24 Jan 2003 20:21:46 -0000 1.28
+++ DOMWriterImpl.cpp 28 Jan 2003 18:31:47 -0000 1.29
@@ -57,6 +57,9 @@
/*
* $Id$
* $Log$
+ * Revision 1.29 2003/01/28 18:31:47 peiyongz
+ * Bug#13694: Allow Xerces to write the BOM to XML files
+ *
* Revision 1.28 2003/01/24 20:21:46 tng
* DOMWriter: Call XMLFormatTarget::flush when done.
*
@@ -181,6 +184,7 @@
static const int SPLIT_CDATA_SECTIONS_ID = 0x5;
static const int VALIDATION_ID = 0x6;
static const int WHITESPACE_IN_ELEMENT_CONTENT_ID = 0x7;
+static const int BYTE_ORDER_MARK_ID = 0x8;
// feature true false
// ================================================================================
@@ -207,7 +211,8 @@
false, true, // normalize-characters
true, true, // split-cdata-sections
false, true, // validation
- true, false // whitespace-in-element-content
+ true, false, // whitespace-in-element-content
+ true, true // byte-order-mark
};
// default end-of-line sequence
@@ -370,6 +375,11 @@
chLatin_e, chSingleQuote, chLF, chNull
};
+static const XMLByte BOM_utf16be[] = {(XMLByte)0xFE, (XMLByte)0xFF, (XMLByte) 0};
+static const XMLByte BOM_utf16le[] = {(XMLByte)0xFF, (XMLByte)0xFE, (XMLByte) 0};
+static const XMLByte BOM_ucs4be[] = {(XMLByte)0x00, (XMLByte)0x00, (XMLByte)0xFE,
(XMLByte)0xFF, (XMLByte) 0};
+static const XMLByte BOM_ucs4le[] = {(XMLByte)0xFF, (XMLByte)0xFE, (XMLByte)0x00,
(XMLByte)0x00, (XMLByte) 0};
+
//
// Notification of the error though error handler
//
@@ -426,11 +436,12 @@
setFeature(CANONICAL_FORM_ID, false);
setFeature(DISCARD_DEFAULT_CONTENT_ID, true );
setFeature(ENTITIES_ID, true );
- setFeature(FORMAT_PRETTY_PRINT_ID, false);
+ setFeature(FORMAT_PRETTY_PRINT_ID, false);
setFeature(NORMALIZE_CHARACTERS_ID, false);
setFeature(SPLIT_CDATA_SECTIONS_ID, true );
setFeature(VALIDATION_ID, false);
setFeature(WHITESPACE_IN_ELEMENT_CONTENT_ID, true );
+ setFeature(BYTE_ORDER_MARK_ID, false);
}
bool DOMWriterImpl::canSetFeature(const XMLCh* const featName
@@ -778,9 +789,13 @@
case DOMNode::DOCUMENT_NODE: // Not to be shown to Filter
{
+
+ // output BOM if needed
+ processBOM();
+
setURCharRef();
const DOMDocument *docu = (const DOMDocument*)nodeToWrite;
-
+
//[23] XMLDecl ::= '<?xml' VersionInfo EncodingDecl? SDDecl? S?
'?>'
//[24] VersionInfo ::= S 'version' Eq ("'" VersionNum "'" | '"'
VersionNum '"')
//[80] EncodingDecl ::= S 'encoding' Eq ('"' EncName '"' | "'" EncName
@@ -1220,6 +1235,8 @@
featureId = VALIDATION_ID;
else if (XMLString::equals(featName,
XMLUni::fgDOMWRTWhitespaceInElementContent))
featureId = WHITESPACE_IN_ELEMENT_CONTENT_ID;
+ else if (XMLString::equals(featName, XMLUni::fgDOMWRTBOM))
+ featureId = BYTE_ORDER_MARK_ID;
//feature name not resolvable
if (featureId == INVALID_FEATURE_ID)
@@ -1478,6 +1495,59 @@
{
DOMWriterImpl* writer = (DOMWriterImpl*) this;
delete writer;
+}
+
+void DOMWriterImpl::processBOM()
+{
+ // if the feature is not set, don't output bom
+ if (!getFeature(BYTE_ORDER_MARK_ID))
+ return;
+
+ if ((XMLString::compareIString(fEncoding, XMLUni::fgUTF16LEncodingString) ==
0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUTF16LEncodingString2) ==
0) )
+ {
+ fFormatter->writeBOM(BOM_utf16le, 2);
+ }
+ else if ((XMLString::compareIString(fEncoding, XMLUni::fgUTF16BEncodingString)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUTF16BEncodingString2)
== 0) )
+ {
+ fFormatter->writeBOM(BOM_utf16be, 2);
+ }
+ else if ((XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString2)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString3)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString4)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString5)
== 0) )
+ {
+#if defined(ENDIANMODE_LITTLE)
+ fFormatter->writeBOM(BOM_utf16le, 2);
+#elif defined(ENDIANMODE_BIG)
+ fFormatter->writeBOM(BOM_utf16be, 2);
+#endif
+ }
+ else if ((XMLString::compareIString(fEncoding, XMLUni::fgUCS4LEncodingString)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUCS4LEncodingString2)
== 0) )
+ {
+ fFormatter->writeBOM(BOM_ucs4le, 4);
+ }
+ else if ((XMLString::compareIString(fEncoding, XMLUni::fgUCS4BEncodingString)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUCS4BEncodingString2)
== 0) )
+ {
+ fFormatter->writeBOM(BOM_ucs4be, 4);
+ }
+ else if ((XMLString::compareIString(fEncoding, XMLUni::fgUCS4EncodingString)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUCS4EncodingString2)
== 0) ||
+ (XMLString::compareIString(fEncoding, XMLUni::fgUCS4EncodingString3)
== 0) )
+ {
+#if defined(ENDIANMODE_LITTLE)
+ fFormatter->writeBOM(BOM_ucs4le, 4);
+#elif defined(ENDIANMODE_BIG)
+ fFormatter->writeBOM(BOM_ucs4be, 4);
+#endif
+ }
+
+ return;
+
}
XERCES_CPP_NAMESPACE_END
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
