[Libreoffice-commits] core.git: Branch 'distro/mimo/mimo-5-4-7-2' - 4 commits - configure.ac external/liborcus shell/source

Andras Timar (via logerrit) logerrit at kemper.freedesktop.org
Mon May 10 21:32:37 UTC 2021


Rebased ref, commits from common ancestor:
commit 12f2ddf58c4b4b7ce35814be09a8f0b99bde18fb
Author:     Andras Timar <andras.timar at collabora.com>
AuthorDate: Mon May 10 22:48:59 2021 +0200
Commit:     Andras Timar <andras.timar at collabora.com>
CommitDate: Mon May 10 23:31:56 2021 +0200

    Bump version to 5.4.7.2.M7
    
    Change-Id: I0e5314312bbefd3bd59cd39f0f8b876de8463e25

diff --git a/configure.ac b/configure.ac
index 9c25ac076c8c..bab32d23e6fa 100644
--- a/configure.ac
+++ b/configure.ac
@@ -9,7 +9,7 @@ dnl in order to create a configure script.
 # several non-alphanumeric characters, those are split off and used only for the
 # ABOUTBOXPRODUCTVERSIONSUFFIX in openoffice.lst. Why that is necessary, no idea.
 
-AC_INIT([LibreOffice],[5.4.7.2.M6],[],[],[http://documentfoundation.org/])
+AC_INIT([LibreOffice],[5.4.7.2.M7],[],[],[http://documentfoundation.org/])
 
 AC_PREREQ([2.59])
 
commit 036e62326d08cfdcda8b390720ab9c0cf9f8c3f7
Author:     Luboš Luňák <l.lunak at collabora.com>
AuthorDate: Thu Apr 29 20:10:34 2021 +0200
Commit:     Andras Timar <andras.timar at collabora.com>
CommitDate: Mon May 10 23:31:56 2021 +0200

    allow utf-8 in xml names (liborcus) (tdf#141672)
    
    Change-Id: Ib150d55b588a572e4352396f18de2331983b2aae
    Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114892
    Tested-by: Jenkins
    Reviewed-by: Luboš Luňák <l.lunak at collabora.com>

diff --git a/external/liborcus/UnpackedTarball_liborcus.mk b/external/liborcus/UnpackedTarball_liborcus.mk
index 791436e66016..6df45cb2fb39 100644
--- a/external/liborcus/UnpackedTarball_liborcus.mk
+++ b/external/liborcus/UnpackedTarball_liborcus.mk
@@ -21,6 +21,10 @@ $(eval $(call gb_UnpackedTarball_add_patches,liborcus,\
 	external/liborcus/0001-protect-the-self-closing-xml-element-code-against-se.patch \
 ))
 
+$(eval $(call gb_UnpackedTarball_add_patches,liborcus,\
+	external/liborcus/allow-utf-8-in-xml-names.patch \
+))
+
 ifeq ($(OS),WNT)
 $(eval $(call gb_UnpackedTarball_add_patches,liborcus,\
 	external/liborcus/windows-constants-hack.patch \
diff --git a/external/liborcus/allow-utf-8-in-xml-names.patch b/external/liborcus/allow-utf-8-in-xml-names.patch
new file mode 100644
index 000000000000..efef24b84053
--- /dev/null
+++ b/external/liborcus/allow-utf-8-in-xml-names.patch
@@ -0,0 +1,289 @@
+From 9889cb660372bc6c3da22fc274c73ea11040415f Mon Sep 17 00:00:00 2001
+From: =?UTF-8?q?Lubo=C5=A1=20Lu=C5=88=C3=A1k?= <l.lunak at centrum.cz>
+Date: Thu, 29 Apr 2021 19:12:20 +0200
+Subject: [PATCH] allow utf-8 in xml names (#137)
+
+https://www.w3.org/TR/2006/REC-xml11-20060816/#NT-NameStartChar
+has a list of all allowed characters.
+---
+ include/orcus/sax_parser_base.hpp |   3 +
+ src/orcus_test_xml.cpp            |   1 +
+ src/parser/sax_parser_base.cpp    | 186 ++++++++++++++++++++++++++++--
+ test/xml/non-ascii/check.txt      |   4 +
+ test/xml/non-ascii/input.xml      |   4 +
+ 5 files changed, 190 insertions(+), 8 deletions(-)
+ create mode 100644 test/xml/non-ascii/check.txt
+ create mode 100644 test/xml/non-ascii/input.xml
+
+diff --git a/include/orcus/sax_parser_base.hpp b/include/orcus/sax_parser_base.hpp
+index a117b3a2..1aeb8b38 100644
+--- a/include/orcus/sax_parser_base.hpp
++++ b/include/orcus/sax_parser_base.hpp
+@@ -217,6 +217,9 @@ protected:
+     void element_name(parser_element& elem, const char* begin_pos);
+     void attribute_name(pstring& attr_ns, pstring& attr_name);
+     void characters_with_encoded_char(cell_buffer& buf);
++
++    int is_name_char();
++    int is_name_start_char();
+ };
+ 
+ }}
+diff --git a/src/orcus_test_xml.cpp b/src/orcus_test_xml.cpp
+index 98e83297..89c8af72 100644
+--- a/src/orcus_test_xml.cpp
++++ b/src/orcus_test_xml.cpp
+@@ -73,6 +73,7 @@ const char* sax_parser_test_dirs[] = {
+     SRCDIR"/test/xml/bom/",
+     SRCDIR"/test/xml/custom-decl-1/",
+     SRCDIR"/test/xml/cdata-1/"
++    SRCDIR"/test/xml/non-ascii/",
+ };
+ 
+ const char* sax_parser_parse_only_test_dirs[] = {
+diff --git a/src/parser/sax_parser_base.cpp b/src/parser/sax_parser_base.cpp
+index 743130da..ecbd7f99 100644
+--- a/src/parser/sax_parser_base.cpp
++++ b/src/parser/sax_parser_base.cpp
+@@ -296,20 +296,22 @@ void parser_base::value_with_encoded_char(cell_buffer& buf, pstring& str)
+         str = pstring(buf.get(), buf.size());
+ 
+     // Skip the closing quote.
+-    assert(cur_char() == '"');
++    assert(!has_char() || cur_char() == '"');
+     next();
+ }
+ 
+ bool parser_base::value(pstring& str, bool decode)
+ {
+     char c = cur_char();
+-    if (c != '"')
++    if (c != '"' && c != '\'')
+         throw malformed_xml_error("value must be quoted", offset());
+ 
++    char quote_char = c;
++
+     c = next_char_checked();
+ 
+     const char* p0 = mp_char;
+-    for (; c != '"'; c = next_char_checked())
++    for (; c != quote_char; c = next_char_checked())
+     {
+         if (decode && c == '&')
+         {
+@@ -330,19 +332,187 @@ bool parser_base::value(pstring& str, bool decode)
+     return false;
+ }
+ 
++// https://www.w3.org/TR/2006/REC-xml11-20060816/#NT-NameStartChar
++// Return length of the character in bytes, otherwise 0.
++template< bool only_start_name >
++static
++int is_name_char_helper(const char* mp_char, const char* mp_end)
++{
++    const unsigned char first = mp_char[0];
++    // Note that ':' technically is an allowed name character, but it is handled separately
++    // e.g. in element_name(), so here pretend it isn't.
++    if (/*first == ':' ||*/ first == '_' || (first >= 'A' && first <= 'Z') || (first >= 'a' && first <= 'z'))
++        return 1;
++    if (!only_start_name && (first == '-' || first == '.' || (first >= '0' && first <= '9')))
++        return 1;
++
++    if (first < 0x7f) // other ascii characters are not allowed
++        return 0;
++    if (mp_end < mp_char + 1)
++        return 0;
++    const unsigned char second = mp_char[1];
++
++    // 0xb7 = 0xc2 0xb7 utf-8
++    if (!only_start_name && first == 0xc2 && second == 0xb7)
++        return 2;
++
++    // [#xC0-#xD6] | [#xD8-#xF6] | [#xF8-#x2FF]
++    // 0xc0 = 0xc3 0x80 utf-8
++    if (first < 0xc3)
++        return 0;
++    // xd7 = 0xc3 0x97 utf-8, 0xf7 = 0xc3 0xb7 utf-8
++    if (first == 0xc3)
++        return second >= 0x80 && second <= 0xff && second != 0x97 && second != 0xb7 ? 2 : 0;
++    // 0x2ff = 0xcb 0xbf utf-8, 0x300 = 0xcc 0x80 utf-8
++    if (first >= 0xc4 && first <= 0xcb)
++        return 2;
++
++    // [#x0300-#x036F]
++    // 0x0300 = 0xcc 0x80 utf-8, 0x36f = 0xcd 0xaf utf-8
++    if (!only_start_name && first == 0xcc)
++        return 2;
++    if (!only_start_name && first == 0xcd && second <= 0xaf)
++        return 2;
++
++    // [#x370-#x37D] | [#x37F-#x1FFF]
++    // 0x370 = 0xcd 0xb0 utf-8, 0x37e = 0xcd 0xbe
++    if (first < 0xcd)
++        return 0;
++    if (first == 0xcd)
++        return second >= 0xb0 && second != 0xbe ? 2 : 0;
++    // 0x07ff = 0xdf 0xbf utf-8 (the last 2-byte utf-8)
++    if (first <= 0xdf)
++        return 2;
++
++    if (first < 0xe0)
++        return 0;
++    if (mp_end < mp_char + 2)
++        return 0;
++    const unsigned char third = mp_char[2];
++
++    // 0x0800 = 0xe0 0xa0 0x80 utf-8, 0x1fff = 0xe1 0xbf 0xbf utf-8, 0x2000 = 0xe2 0x80 0x80
++    if (first == 0xe0 || first == 0xe1)
++        return 3;
++
++    // [#x200C-#x200D]
++    // 0x200c = 0xe2 0x80 0x8c utf-8, 0x200d = 0xe2 0x80 0x8d utf-8
++    if (first < 0xe2)
++        return 0;
++    if (first == 0xe2 && second == 0x80 && (third == 0x8c || third == 0x8d))
++        return 3;
++
++    // [#x203F-#x2040]
++    // 0x203f = 0xe2 0x80 0xbf utf-8, 0x2040 = 0xe2 0x81 0x80 utf-8
++    if (!only_start_name && first == 0xe2 && second == 0x80 && third == 0xbf)
++        return 3;
++    if (!only_start_name && first == 0xe2 && second == 0x81 && third == 0x80)
++        return 3;
++
++    // [#x2070-#x218F]
++    // 0x2070 = 0xe2 0x81 0xb0 utf-8, 0x218f = 0xe2 0x86 0x8f utf-8
++    if (first == 0xe2)
++    {
++        if (second < 0x81)
++            return 0;
++        if (second >= 0x81 && second < 0x86)
++            return 3;
++        if (second == 0x86 && third <= 0x8f)
++            return 3;
++    }
++
++    // [#x2C00-#x2FEF]
++    // 0x2c00 = 0xe2 0xb0 0x80 utf-8, 0x2fef = 0xe2 0xbf 0xaf utf-8
++    if (first == 0xe2)
++    {
++        if (second < 0xb0)
++            return 0;
++        if (second < 0xbf)
++            return 3;
++        if (second == 0xbf && third <= 0xaf)
++            return 3;
++    }
++
++    // [#x3001-#xD7FF]
++    // 0x3001 = 0xe3 0x80 0x81 utf-8, 0xd7ff = 0xed 0x9f 0xbf utf-8, 0xd800 = 0xed 0xa0 0x80 utf-8
++    if (first < 0xe3)
++        return 0;
++    if (first < 0xed)
++        return 3;
++    if (first == 0xed && second <= 0x9f)
++        return 3;
++
++    // [#xF900-#xFDCF]
++    // 0xf900 = 0xef 0xa4 0x80 utf-8, 0xfdcf = 0xef 0xb7 0x8f utf-8
++    if (first == 0xef)
++    {
++        if (second < 0xa4)
++            return 0;
++        if (second < 0xb7)
++            return 3;
++        if (second == 0xb7 && third <= 0x8f)
++            return 3;
++    }
++
++    // [#xFDF0-#xFFFD]
++    // 0xfdf0 = 0xef 0xb7 0xb0 utf-8, 0xfffd = 0xef 0xbf 0xbd utf-8
++    if (first == 0xef)
++    {
++        assert(second >= 0xb7);
++        if (second == 0xb7 && third < 0xb0)
++            return 0;
++        if (second < 0xbe)
++            return 3;
++        if (second == 0xbf && third <= 0xbd)
++            return 3;
++    }
++
++    if (first < 0xf0)
++        return 0;
++    if (mp_end < mp_char + 3)
++        return 0;
++    // const unsigned char fourth = mp_char[3];
++
++    // [#x10000-#xEFFFF]
++    // 0x10000 = 0xf0 0x90 0x80 0x80 utf-8, 0xeffff = 0xf3 0xaf 0xbf 0xbf utf-8,
++    // 0xf0000 = 0xf3 0xb0 0x80 0x80 utf-8
++    if (first >= 0xf0 && first < 0xf2)
++        return 4;
++    if (first == 0xf3 && second < 0xb0)
++        return 4;
++
++    return 0;
++}
++
++int parser_base::is_name_char()
++{
++    return is_name_char_helper<false>(mp_char, mp_end);
++}
++
++int parser_base::is_name_start_char()
++{
++    return is_name_char_helper<true>(mp_char, mp_end);
++}
++
+ void parser_base::name(pstring& str)
+ {
+     const char* p0 = mp_char;
+-    char c = cur_char();
+-    if (!is_alpha(c))
++    int skip = is_name_start_char();
++    if (skip == 0)
+     {
+         ::std::ostringstream os;
+-        os << "name must begin with an alphabet, but got this instead '" << c << "'";
++        os << "name must begin with an alphabet, but got this instead '" << cur_char() << "'";
+         throw malformed_xml_error(os.str(), offset());
+     }
++    next(skip);
+ 
+-    while (is_alpha(c) || is_numeric(c) || is_name_char(c))
+-        c = next_char_checked();
++    for(;;)
++    {
++        cur_char_checked(); // check end of xml stream
++        skip = is_name_char();
++        if(skip == 0)
++            break;
++        next(skip);
++    }
+ 
+     str = pstring(p0, mp_char-p0);
+ }
+diff --git a/test/xml/non-ascii/check.txt b/test/xml/non-ascii/check.txt
+new file mode 100644
+index 00000000..77b7c003
+--- /dev/null
++++ b/test/xml/non-ascii/check.txt
+@@ -0,0 +1,4 @@
++/Myšička
++/Myšička at jméno="Žužla"
++/Myšička/Nožičky
++/Myšička/Nožičky"4"
+diff --git a/test/xml/non-ascii/input.xml b/test/xml/non-ascii/input.xml
+new file mode 100644
+index 00000000..c516744b
+--- /dev/null
++++ b/test/xml/non-ascii/input.xml
+@@ -0,0 +1,4 @@
++<?xml version="1.0" encoding="UTF-8"?>
++<Myšička jméno="Žužla">
++   <Nožičky>4</Nožičky>
++</Myšička>
+-- 
+2.26.2
+
commit 70acef544db618ded97e9550a2519930d30e3fba
Author:     Stephan Bergmann <sbergman at redhat.com>
AuthorDate: Tue Feb 16 09:30:09 2021 +0100
Commit:     Andras Timar <andras.timar at collabora.com>
CommitDate: Mon May 10 23:31:54 2021 +0200

    Improve checkExtension
    
    Reviewed-on: https://gerrit.libreoffice.org/c/core/+/110970
    Reviewed-by: Mike Kaganski <mike.kaganski at collabora.com>
    Tested-by: Jenkins
    (cherry picked from commit f456c4dacf700e064e112ef068ff7edb04239754)
    Reviewed-on: https://gerrit.libreoffice.org/c/core/+/110922
    Reviewed-by: Michael Stahl <michael.stahl at allotropia.de>
    (cherry picked from commit f19d95986756412e5d72047656eec17a720c5e57)
    
    Change-Id: Iff416a9c5930ad5903f7ee51a2abbc94d5f40800

diff --git a/shell/source/win32/SysShExec.cxx b/shell/source/win32/SysShExec.cxx
index 1d0932e5db8d..19e0223f37d0 100644
--- a/shell/source/win32/SysShExec.cxx
+++ b/shell/source/win32/SysShExec.cxx
@@ -389,21 +389,28 @@ void SAL_CALL CSysShExec::execute( const OUString& aCommand, const OUString& aPa
                 }
             }
             pathname = OUString(SAL_U(path));
+            // ShellExecuteExW appears to ignore trailing dots, so remove them:
+            while (pathname.endsWith(".", &pathname)) {}
             auto const n = pathname.lastIndexOf('.');
             if (n > pathname.lastIndexOf('\\')) {
                 auto const ext = pathname.copy(n + 1);
-                OUString env;
-                if (osl_getEnvironment(OUString("PATHEXT").pData, &env.pData) != osl_Process_E_None)
-                {
-                    SAL_INFO("shell", "osl_getEnvironment(PATHEXT) failed");
-                }
-                if (!(checkExtension(ext, env)
-                      && checkExtension(
-                          ext,
-                          ".COM;.EXE;.BAT;.CMD;.VBS;.VBE;.JS;.JSE;.WSF;.WSH;.MSC;.PY;.CLASS;.JAR")))
-                {
-                    throw css::lang::IllegalArgumentException(
-                        "XSystemShellExecute.execute, cannot process <" + aCommand + ">", {}, 0);
+                if (!ext.isEmpty()) {
+                    OUString env;
+                    if (osl_getEnvironment(OUString("PATHEXT").pData, &env.pData)
+                        != osl_Process_E_None)
+                    {
+                        SAL_INFO("shell", "osl_getEnvironment(PATHEXT) failed");
+                    }
+                    if (!(checkExtension(ext, env)
+                          && checkExtension(
+                              ext,
+                              ".COM;.EXE;.BAT;.CMD;.VBS;.VBE;.JS;.JSE;.WSF;.WSH;.MSC;.PY;.CLASS;"
+                                  ".JAR;.APPLICATION;.LNK;.SCR")))
+                    {
+                        throw css::lang::IllegalArgumentException(
+                            "XSystemShellExecute.execute, cannot process <" + aCommand + ">", {},
+                            0);
+                    }
                 }
             }
         }
commit af34e2c563a1535e25c4ca9a933cb90299b8eef2
Author:     Mike Kaganski <mike.kaganski at collabora.com>
AuthorDate: Tue Jan 28 01:28:24 2020 +0300
Commit:     Andras Timar <andras.timar at collabora.com>
CommitDate: Mon May 10 23:31:24 2021 +0200

    tdf#130216: normalize paths with .. segments
    
    ... which obviously are rejected by SHGetFileInfoW and SHParseDisplayName
    that it calls internally.
    
    Reviewed-on: https://gerrit.libreoffice.org/c/core/+/87565
    Tested-by: Jenkins
    Reviewed-by: Mike Kaganski <mike.kaganski at collabora.com>
    Signed-off-by: Xisco Fauli <xiscofauli at libreoffice.org>
    Reviewed-on: https://gerrit.libreoffice.org/c/core/+/87737
    Reviewed-by: Miklos Vajna <vmiklos at collabora.com>
    (cherry picked from commit fc043d38c256243fb782cc48e7708feaeabba4ae)
    
    Change-Id: I2f5f3c675ea6aa1c2d92eef30be4399a8d600255

diff --git a/shell/source/win32/SysShExec.cxx b/shell/source/win32/SysShExec.cxx
index a9e5a2c4ec7d..1d0932e5db8d 100644
--- a/shell/source/win32/SysShExec.cxx
+++ b/shell/source/win32/SysShExec.cxx
@@ -39,6 +39,7 @@
 #endif
 #include <windows.h>
 #include <shellapi.h>
+#include <Shlobj.h>
 #include <Shobjidl.h>
 #include <objbase.h>
 #if defined _MSC_VER
@@ -307,21 +308,33 @@ void SAL_CALL CSysShExec::execute( const OUString& aCommand, const OUString& aPa
                      + "> failed with " + OUString::number(e1)),
                     {}, 0);
             }
+            const int MAX_LONG_PATH = 32767; // max longpath on WinNT
+            if (pathname.getLength() >= MAX_LONG_PATH)
+            {
+                throw css::lang::IllegalArgumentException(
+                    "XSystemShellExecute.execute, path <" + pathname + "> too long", {}, 0);
+            }
+            wchar_t path[MAX_LONG_PATH];
+            wcscpy_s(path, SAL_W(pathname.getStr()));
             for (int i = 0;; ++i) {
+                // tdf#130216: normalize c:\path\to\something\..\else into c:\path\to\else
+                if (PathResolve(path, nullptr, PRF_VERIFYEXISTS | PRF_REQUIREABSOLUTE) == 0)
+                {
+                    throw css::lang::IllegalArgumentException(
+                        "XSystemShellExecute.execute, PathResolve(" + OUString(SAL_U(path))
+                            + ") failed",
+                        {}, 0);
+                }
                 SHFILEINFOW info;
-                if (SHGetFileInfoW(
-                        SAL_W(pathname.getStr()), 0, &info, sizeof info, SHGFI_EXETYPE)
-                    != 0)
+                if (SHGetFileInfoW(path, 0, &info, sizeof info, SHGFI_EXETYPE) != 0)
                 {
                     throw css::lang::IllegalArgumentException(
                         "XSystemShellExecute.execute, cannot process <" + aCommand + ">", {}, 0);
                 }
-                if (SHGetFileInfoW(
-                        SAL_W(pathname.getStr()), 0, &info, sizeof info, SHGFI_ATTRIBUTES)
-                    == 0)
+                if (SHGetFileInfoW(path, 0, &info, sizeof info, SHGFI_ATTRIBUTES) == 0)
                 {
                     throw css::lang::IllegalArgumentException(
-                        "XSystemShellExecute.execute, SHGetFileInfoW(" + pathname + ") failed", {},
+                        "XSystemShellExecute.execute, SHGetFileInfoW(" + OUString(SAL_U(path)) + ") failed", {},
                         0);
                 }
                 if ((info.dwAttributes & SFGAO_LINK) == 0) {
@@ -346,7 +359,7 @@ void SAL_CALL CSysShExec::execute( const OUString& aCommand, const OUString& aPa
                          + o3tl::runtimeToOUString(e3.what())),
                         {}, 0);
                 }
-                e2 = file->Load(SAL_W(pathname.getStr()), STGM_READ);
+                e2 = file->Load(path, STGM_READ);
                 if (FAILED(e2)) {
                     throw css::lang::IllegalArgumentException(
                         ("XSystemShellExecute.execute, IPersistFile.Load failed with "
@@ -360,16 +373,14 @@ void SAL_CALL CSysShExec::execute( const OUString& aCommand, const OUString& aPa
                          + OUString::number(e2)),
                         {}, 0);
                 }
-                wchar_t path[MAX_PATH];
                 WIN32_FIND_DATAW wfd;
-                e2 = link->GetPath(path, MAX_PATH, &wfd, SLGP_RAWPATH);
+                e2 = link->GetPath(path, SAL_N_ELEMENTS(path), &wfd, SLGP_RAWPATH);
                 if (FAILED(e2)) {
                     throw css::lang::IllegalArgumentException(
                         ("XSystemShellExecute.execute, IShellLink.GetPath failed with "
                          + OUString::number(e2)),
                         {}, 0);
                 }
-                pathname = SAL_U(path);
                 // Fail at some arbitrary nesting depth, to avoid an infinite loop:
                 if (i == 30) {
                     throw css::lang::IllegalArgumentException(
@@ -377,6 +388,7 @@ void SAL_CALL CSysShExec::execute( const OUString& aCommand, const OUString& aPa
                         {}, 0);
                 }
             }
+            pathname = OUString(SAL_U(path));
             auto const n = pathname.lastIndexOf('.');
             if (n > pathname.lastIndexOf('\\')) {
                 auto const ext = pathname.copy(n + 1);


More information about the Libreoffice-commits mailing list