summaryrefslogtreecommitdiffstats
path: root/include/tools/solar.h
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-27 16:51:28 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-27 16:51:28 +0000
commit940b4d1848e8c70ab7642901a68594e8016caffc (patch)
treeeb72f344ee6c3d9b80a7ecc079ea79e9fba8676d /include/tools/solar.h
parentInitial commit. (diff)
downloadlibreoffice-940b4d1848e8c70ab7642901a68594e8016caffc.tar.xz
libreoffice-940b4d1848e8c70ab7642901a68594e8016caffc.zip
Adding upstream version 1:7.0.4.upstream/1%7.0.4upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r--include/tools/solar.h121
1 files changed, 121 insertions, 0 deletions
diff --git a/include/tools/solar.h b/include/tools/solar.h
new file mode 100644
index 000000000..721573287
--- /dev/null
+++ b/include/tools/solar.h
@@ -0,0 +1,121 @@
+/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+/*
+ * This file is part of the LibreOffice project.
+ *
+ * This Source Code Form is subject to the terms of the Mozilla Public
+ * License, v. 2.0. If a copy of the MPL was not distributed with this
+ * file, You can obtain one at http://mozilla.org/MPL/2.0/.
+ *
+ * This file incorporates work covered by the following license notice:
+ *
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed
+ * with this work for additional information regarding copyright
+ * ownership. The ASF licenses this file to you under the Apache
+ * License, Version 2.0 (the "License"); you may not use this file
+ * except in compliance with the License. You may obtain a copy of
+ * the License at http://www.apache.org/licenses/LICENSE-2.0 .
+ */
+
+#ifndef INCLUDED_TOOLS_SOLAR_H
+#define INCLUDED_TOOLS_SOLAR_H
+
+#include <sal/types.h>
+#include <osl/endian.h>
+
+/** Intermediate type to solve type clash with Windows headers.
+ Should be removed as soon as all code parts have been reviewed
+ and the correct type is known. Most of the times ULONG is meant
+ to be a 32-Bit unsigned integer type as sal_uInt32 is often
+ used for data exchange or for similar method args. */
+typedef sal_uIntPtr sal_uLong; /* Replaces type ULONG */
+
+// misc. macros to leverage platform and compiler differences
+
+#define DELETEZ( p ) ( delete p,p = NULL )
+
+// solar binary types
+
+/* Solar (portable) Binary (exchange) Type; OSI 6 subset
+ always little endian;
+ not necessarily aligned */
+
+typedef sal_uInt8 SVBT16[2];
+typedef sal_uInt8 SVBT32[4];
+typedef sal_uInt8 SVBT64[8];
+
+#ifdef __cplusplus
+
+inline sal_uInt16 SVBT16ToUInt16( const SVBT16 p ) { return static_cast<sal_uInt16>
+ (static_cast<sal_uInt16>(p[0])
+ + (static_cast<sal_uInt16>(p[1]) << 8)); }
+inline sal_Int16 SVBT16ToInt16( const SVBT16 p ) { return sal_Int16(SVBT16ToUInt16(p)); }
+inline sal_uInt32 SVBT32ToUInt32 ( const SVBT32 p ) { return static_cast<sal_uInt32>
+ (static_cast<sal_uInt32>(p[0])
+ + (static_cast<sal_uInt32>(p[1]) << 8)
+ + (static_cast<sal_uInt32>(p[2]) << 16)
+ + (static_cast<sal_uInt32>(p[3]) << 24)); }
+#if defined OSL_LITENDIAN
+inline double SVBT64ToDouble( const SVBT64 p )
+{
+ double n;
+ reinterpret_cast<sal_uInt8*>(&n)[0] = p[0];
+ reinterpret_cast<sal_uInt8*>(&n)[1] = p[1];
+ reinterpret_cast<sal_uInt8*>(&n)[2] = p[2];
+ reinterpret_cast<sal_uInt8*>(&n)[3] = p[3];
+ reinterpret_cast<sal_uInt8*>(&n)[4] = p[4];
+ reinterpret_cast<sal_uInt8*>(&n)[5] = p[5];
+ reinterpret_cast<sal_uInt8*>(&n)[6] = p[6];
+ reinterpret_cast<sal_uInt8*>(&n)[7] = p[7];
+ return n;
+}
+#else
+inline double SVBT64ToDouble( const SVBT64 p ) { double n;
+ reinterpret_cast<sal_uInt8*>(&n)[0] = p[7];
+ reinterpret_cast<sal_uInt8*>(&n)[1] = p[6];
+ reinterpret_cast<sal_uInt8*>(&n)[2] = p[5];
+ reinterpret_cast<sal_uInt8*>(&n)[3] = p[4];
+ reinterpret_cast<sal_uInt8*>(&n)[4] = p[3];
+ reinterpret_cast<sal_uInt8*>(&n)[5] = p[2];
+ reinterpret_cast<sal_uInt8*>(&n)[6] = p[1];
+ reinterpret_cast<sal_uInt8*>(&n)[7] = p[0];
+ return n; }
+#endif
+
+inline void ShortToSVBT16( sal_uInt16 n, SVBT16 p )
+{
+ p[0] = static_cast<sal_uInt8>(n);
+ p[1] = static_cast<sal_uInt8>(n >> 8);
+}
+inline void UInt32ToSVBT32 ( sal_uInt32 n, SVBT32 p )
+{
+ p[0] = static_cast<sal_uInt8>(n);
+ p[1] = static_cast<sal_uInt8>(n >> 8);
+ p[2] = static_cast<sal_uInt8>(n >> 16);
+ p[3] = static_cast<sal_uInt8>(n >> 24);
+}
+inline void Int32ToSVBT32 ( sal_Int32 n, SVBT32 p ) { UInt32ToSVBT32(sal_uInt32(n), p); }
+#if defined OSL_LITENDIAN
+inline void DoubleToSVBT64( double n, SVBT64 p ) { p[0] = reinterpret_cast<sal_uInt8*>(&n)[0];
+ p[1] = reinterpret_cast<sal_uInt8*>(&n)[1];
+ p[2] = reinterpret_cast<sal_uInt8*>(&n)[2];
+ p[3] = reinterpret_cast<sal_uInt8*>(&n)[3];
+ p[4] = reinterpret_cast<sal_uInt8*>(&n)[4];
+ p[5] = reinterpret_cast<sal_uInt8*>(&n)[5];
+ p[6] = reinterpret_cast<sal_uInt8*>(&n)[6];
+ p[7] = reinterpret_cast<sal_uInt8*>(&n)[7]; }
+#else
+inline void DoubleToSVBT64( double n, SVBT64 p ) { p[0] = reinterpret_cast<sal_uInt8*>(&n)[7];
+ p[1] = reinterpret_cast<sal_uInt8*>(&n)[6];
+ p[2] = reinterpret_cast<sal_uInt8*>(&n)[5];
+ p[3] = reinterpret_cast<sal_uInt8*>(&n)[4];
+ p[4] = reinterpret_cast<sal_uInt8*>(&n)[3];
+ p[5] = reinterpret_cast<sal_uInt8*>(&n)[2];
+ p[6] = reinterpret_cast<sal_uInt8*>(&n)[1];
+ p[7] = reinterpret_cast<sal_uInt8*>(&n)[0]; }
+#endif
+#endif
+
+#endif
+
+/* vim:set shiftwidth=4 softtabstop=4 expandtab: */