Fix a bug with UTF-16 byte-order-marks on big-endian hosts. (CVS 1522)

FossilOrigin-Name: c17b864103fe5e6def0f650eadb7b2cc6e87144f

diff --git a/manifest b/manifest
index 3524b75b5f9e005f6a6313e6f9eb58fe8477232f..78ccfbd8d470894658eeecef42236060376edf48 100644 (file)
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C Fix\sanother\swebsite\stypo.\s(CVS\s1521)
-D 2004-06-02T00:08:56
+C Fix\sa\sbug\swith\sUTF-16\sbyte-order-marks\son\sbig-endian\shosts.\s(CVS\s1522)
+D 2004-06-02T00:29:24
 F Makefile.in ab7b0d5118e2da97bac66be8684a1034e3500f5a
 F Makefile.linux-gcc b86a99c493a5bfb402d1d9178dcdc4bd4b32f906
 F README f1de682fbbd94899d50aca13d387d1b3fd3be2dd
@@ -67,7 +67,7 @@ F src/test5.c a894fbfa8a7dcdc2cbc0c38c11149cc5c8252cea
 F src/tokenize.c 183c5d7da11affab5d70d903d33409c8c0ce6c5b
 F src/trigger.c 04b2c310d0d056b213609cab6df5fff03d5eaf88
 F src/update.c 259f06e7b22c684b2d3dda54a18185892d6e9573
-F src/utf.c c18b4e66bd13872b7199444addaa04785ff1b80a
+F src/utf.c c8be20ecdcb10659e23c43e35d835460e964d248
 F src/util.c 3b647719c0bece41491300b605cff96a7a26f03a
 F src/vacuum.c c91acc316127411980982938d050b299d42b81ef
 F src/vdbe.c cc754eddc0d2e741e57c4557c77754d8bbd93eeb
@@ -214,7 +214,7 @@ F www/support.tcl 1801397edd271cc39a2aadd54e701184b5181248
 F www/tclsqlite.tcl 19191cf2a1010eaeff74c51d83fd5f5a4d899075
 F www/vdbe.tcl 59288db1ac5c0616296b26dce071c36cb611dfe9
 F www/whentouse.tcl a8335bce47cc2fddb07f19052cb0cb4d9129a8e4
-P 023d1760c1a720632b25c98d14bf521ab91861e2
-R 5a6a7e0dc4da60097c6d0b9613de9a3a
-U drh
-Z 18ef1953bc240ed2b92d7ff4b96c3997
+P 24e2bbd0f233eaec1789e866e9abb88db66e2278
+R 770b5bbc0f504b74630be07ad566c456
+U danielk1977
+Z d278c6c962aa9ab7dcd916a21504bfc5

diff --git a/manifest.uuid b/manifest.uuid
index 1df538c5b46e980f085ad04e8cba3332c5d3398d..aa327ecee1d6a1aebe578a0a0cec3990844ff0a3 100644 (file)
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-24e2bbd0f233eaec1789e866e9abb88db66e2278
\ No newline at end of file
+c17b864103fe5e6def0f650eadb7b2cc6e87144f
\ No newline at end of file

diff --git a/src/utf.c b/src/utf.c
index 5b5507823a33d6909a915191440199a18155c82f..65dd05e4a3981e5dba864cda0e2289e7326d9eba 100644 (file)
--- a/src/utf.c
+++ b/src/utf.c
@@ -12,7 +12,7 @@
 ** This file contains routines used to translate between UTF-8, 
 ** UTF-16, UTF-16BE, and UTF-16LE.
 **
-** $Id: utf.c,v 1.15 2004/05/31 18:51:58 drh Exp $
+** $Id: utf.c,v 1.16 2004/06/02 00:29:24 danielk1977 Exp $
 **
 ** Notes on UTF-8:
 **
@@ -59,19 +59,6 @@ struct UtfString {
   int c;                /* Number of pZ bytes already read or written */
 };
 
-/* TODO: Implement this macro in os.h. It should be 1 on big-endian
-** machines, and 0 on little-endian.
-*/
-#define SQLITE_NATIVE_BIGENDIAN 0
-
-#if SQLITE_NATIVE_BIGENDIAN == 1
-#define BOM_BIGENDIAN 0x0000FFFE
-#define BOM_LITTLEENDIAN 0x0000FEFF
-#else
-#define BOM_BIGENDIAN 0x0000FEFF
-#define BOM_LITTLEENDIAN 0x0000FFFE
-#endif
-
 /*
 ** These two macros are used to interpret the first two bytes of the 
 ** unsigned char array pZ as a 16-bit unsigned int. BE16() for a big-endian
@@ -108,14 +95,10 @@ static int readUtf16Bom(UtfString *pStr, int big_endian){
   ** present.
   */
   if( pStr->n>1 ){
-    u32 bom = BE16(pStr->pZ);
-    if( bom==BOM_BIGENDIAN ){
-      pStr->c = 2;
-      return 1;
-    }
-    if( bom==BOM_LITTLEENDIAN ){
-      pStr->c = 2;
-      return 0;
+    u8 bom = sqlite3UtfReadBom(pStr->pZ, 2);
+    if( bom ){
+      pStr->c += 2;
+      return (bom==TEXT_Utf16le)?0:1;
     }
   }