From a100265576b58e6f80070a893f05c810731f69e1 Mon Sep 17 00:00:00 2001
From: kwaclaw <kwaclaw>
Date: Fri, 23 Dec 2005 14:45:27 +0000
Subject: [PATCH] Added support for 64-bit integers when reporting byte
 indexes, line and column numbers. Can be turned on/off using the
 XML_LARGE_SIZE switch.

---
 Changes              |  4 +++-
 doc/reference.html   | 16 +++++++++++++---
 lib/expat.h          |  6 +++---
 lib/expat_external.h |  8 ++++++++
 lib/xmlparse.c       |  8 ++++----
 lib/xmltok.h         |  4 ++--
 6 files changed, 33 insertions(+), 13 deletions(-)

diff --git a/Changes b/Changes
index 16ffadd..536a634 100644
--- a/Changes
+++ b/Changes
@@ -1,8 +1,10 @@
-Release 1.95.9 TBD
+Release 2.0 TBD
         - We no longer use the "check" library for C unit testing; we
           always use the (partial) internal implementation of the API.
         - Report XML_NS setting via XML_GetFeatureList().
         - Fixed headers for use from C++.
+        - Added XML_LARGE_SIZE switch to enable 64bit values for
+          byte indexes and line/column numbers.
         - Updated to use libtool 1.5.10 (the most recent).
         - Added support for AmigaOS.
         - Some mostly minor bug fixes. SF issues include: 1006708,
diff --git a/doc/reference.html b/doc/reference.html
index 574ae31..b91797b 100644
--- a/doc/reference.html
+++ b/doc/reference.html
@@ -344,6 +344,16 @@ defined using the <code>wchar_t</code> type; otherwise, <code>unsigned
 short</code> is used.  Defining this implies
 <code>XML_UNICODE</code>.</dd>
 
+<dt>XML_LARGE_SIZE</dt>
+<dd>If defined, causes the <code>XML_Size</code> and <code>XML_Index</code>
+integer types to be at least 64 bits in size. This is intended to support
+processing of very large input streams, where the return values of
+<code><a href="#XML_GetCurrentByteIndex" >XML_GetCurrentByteIndex</a></code>,
+<code><a href="#XML_GetCurrentLineNumber" >XML_GetCurrentLineNumber</a></code> and
+<code><a href="#XML_GetCurrentColumnNumber" >XML_GetCurrentColumnNumber</a></code>
+could overflow. It may not be supported by all compilers, and is turned
+off by default.</dd>
+
 <dt>XML_CONTEXT_BYTES</dt>
 <dd>The number of input bytes of markup context which the parser will
 ensure are available for reporting via <code><a href=
@@ -1905,7 +1915,7 @@ The code should be one of the enums that can be returned from
 </div>
 
 <pre class="fcndec" id="XML_GetCurrentByteIndex">
-long XMLCALL
+XML_Index XMLCALL
 XML_GetCurrentByteIndex(XML_Parser p);
 </pre>
 <div class="fcndef">
@@ -1916,7 +1926,7 @@ the values returned by <code><a href= "#XML_GetCurrentLineNumber"
 </div>
 
 <pre class="fcndec" id="XML_GetCurrentLineNumber">
-int XMLCALL
+XML_Size XMLCALL
 XML_GetCurrentLineNumber(XML_Parser p);
 </pre>
 <div class="fcndef">
@@ -1925,7 +1935,7 @@ Return the line number of the position.  The first line is reported as
 </div>
 
 <pre class="fcndec" id="XML_GetCurrentColumnNumber">
-int XMLCALL
+XML_Size XMLCALL
 XML_GetCurrentColumnNumber(XML_Parser p);
 </pre>
 <div class="fcndef">
diff --git a/lib/expat.h b/lib/expat.h
index ac1053f..09c5d14 100644
--- a/lib/expat.h
+++ b/lib/expat.h
@@ -905,9 +905,9 @@ XML_GetErrorCode(XML_Parser parser);
    was detected; otherwise the location is the location of the last
    parse event, as described above.
 */
-XMLPARSEAPI(int) XML_GetCurrentLineNumber(XML_Parser parser);
-XMLPARSEAPI(int) XML_GetCurrentColumnNumber(XML_Parser parser);
-XMLPARSEAPI(long) XML_GetCurrentByteIndex(XML_Parser parser);
+XMLPARSEAPI(XML_Size) XML_GetCurrentLineNumber(XML_Parser parser);
+XMLPARSEAPI(XML_Size) XML_GetCurrentColumnNumber(XML_Parser parser);
+XMLPARSEAPI(XML_Index) XML_GetCurrentByteIndex(XML_Parser parser);
 
 /* Return the number of bytes in the current event.
    Returns 0 if the event is in an internal entity.
diff --git a/lib/expat_external.h b/lib/expat_external.h
index 7202e11..ddb162d 100644
--- a/lib/expat_external.h
+++ b/lib/expat_external.h
@@ -94,6 +94,14 @@ typedef char XML_Char;
 typedef char XML_LChar;
 #endif /* XML_UNICODE */
 
+#ifdef XML_LARGE_SIZE  /* Use large integers for counts and positions. */
+typedef long long XML_Index;
+typedef unsigned long long XML_Size;
+#else
+typedef long XML_Index;
+typedef unsigned long XML_Size;
+#endif /* XML_LARGE_SIZE */
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/lib/xmlparse.c b/lib/xmlparse.c
index 453077f..04f8b91 100644
--- a/lib/xmlparse.c
+++ b/lib/xmlparse.c
@@ -458,7 +458,7 @@ struct XML_ParserStruct {
   char *m_bufferEnd;
   /* allocated end of buffer */
   const char *m_bufferLim;
-  long m_parseEndByteIndex;
+  XML_Index m_parseEndByteIndex;
   const char *m_parseEndPtr;
   XML_Char *m_dataBuf;
   XML_Char *m_dataBufEnd;
@@ -1751,7 +1751,7 @@ XML_GetErrorCode(XML_Parser parser)
   return errorCode;
 }
 
-long XMLCALL
+XML_Index XMLCALL
 XML_GetCurrentByteIndex(XML_Parser parser)
 {
   if (eventPtr)
@@ -1780,7 +1780,7 @@ XML_GetInputContext(XML_Parser parser, int *offset, int *size)
   return (char *) 0;
 }
 
-int XMLCALL
+XML_Size XMLCALL
 XML_GetCurrentLineNumber(XML_Parser parser)
 {
   if (eventPtr && eventPtr >= positionPtr) {
@@ -1790,7 +1790,7 @@ XML_GetCurrentLineNumber(XML_Parser parser)
   return position.lineNumber + 1;
 }
 
-int XMLCALL
+XML_Size XMLCALL
 XML_GetCurrentColumnNumber(XML_Parser parser)
 {
   if (eventPtr && eventPtr >= positionPtr) {
diff --git a/lib/xmltok.h b/lib/xmltok.h
index 1ecd05f..ca867aa 100644
--- a/lib/xmltok.h
+++ b/lib/xmltok.h
@@ -111,8 +111,8 @@ extern "C" {
 
 typedef struct position {
   /* first line and first column are 0 not 1 */
-  unsigned long lineNumber;
-  unsigned long columnNumber;
+  XML_Size lineNumber;
+  XML_Size columnNumber;
 } POSITION;
 
 typedef struct {
-- 
2.11.4.GIT