Merge pull request #1041 from takase1121/lua-utf8-windows

Support UTF-8 on Windows (Lua)
This commit is contained in:
Jefferson González 2022-06-17 09:47:02 -04:00 committed by GitHub
commit 64cb6a290d
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 210 additions and 0 deletions

2
.gitignore vendored
View File

@ -18,3 +18,5 @@ compile_commands.json
error.txt
lite-xl*
LiteXL*
!resources/windows/*.diff

View File

@ -0,0 +1,195 @@
diff -ruN lua-5.4.3/meson.build newlua/meson.build
--- lua-5.4.3/meson.build 2022-05-29 21:04:17.850449500 +0800
+++ newlua/meson.build 2022-06-10 19:23:55.685139800 +0800
@@ -82,6 +82,7 @@
'src/lutf8lib.c',
'src/lvm.c',
'src/lzio.c',
+ 'src/utf8_wrappers.c',
dependencies: lua_lib_deps,
override_options: project_options,
implicit_include_directories: false,
Binary files lua-5.4.3/src/lua54.dll and newlua/src/lua54.dll differ
diff -ruN lua-5.4.3/src/luaconf.h newlua/src/luaconf.h
--- lua-5.4.3/src/luaconf.h 2021-03-15 21:32:52.000000000 +0800
+++ newlua/src/luaconf.h 2022-06-10 19:15:03.014745300 +0800
@@ -786,5 +786,15 @@
+#if defined(lua_c) || defined(luac_c) || (defined(LUA_LIB) && \
+ (defined(lauxlib_c) || defined(liolib_c) || \
+ defined(loadlib_c) || defined(loslib_c)))
+#include "utf8_wrappers.h"
+#endif
+
+
+
+
+
#endif
diff -ruN lua-5.4.3/src/Makefile newlua/src/Makefile
--- lua-5.4.3/src/Makefile 2021-02-10 02:47:17.000000000 +0800
+++ newlua/src/Makefile 2022-06-10 19:22:45.267931400 +0800
@@ -33,7 +33,7 @@
PLATS= guess aix bsd c89 freebsd generic linux linux-readline macosx mingw posix solaris
LUA_A= liblua.a
-CORE_O= lapi.o lcode.o lctype.o ldebug.o ldo.o ldump.o lfunc.o lgc.o llex.o lmem.o lobject.o lopcodes.o lparser.o lstate.o lstring.o ltable.o ltm.o lundump.o lvm.o lzio.o
+CORE_O= lapi.o lcode.o lctype.o ldebug.o ldo.o ldump.o lfunc.o lgc.o llex.o lmem.o lobject.o lopcodes.o lparser.o lstate.o lstring.o ltable.o ltm.o lundump.o lvm.o lzio.o utf8_wrappers.o
LIB_O= lauxlib.o lbaselib.o lcorolib.o ldblib.o liolib.o lmathlib.o loadlib.o loslib.o lstrlib.o ltablib.o lutf8lib.o linit.o
BASE_O= $(CORE_O) $(LIB_O) $(MYOBJS)
diff -ruN lua-5.4.3/src/utf8_wrappers.c newlua/src/utf8_wrappers.c
--- lua-5.4.3/src/utf8_wrappers.c 1970-01-01 07:30:00.000000000 +0730
+++ newlua/src/utf8_wrappers.c 2022-06-10 19:13:11.904613300 +0800
@@ -0,0 +1,101 @@
+/**
+ * Wrappers to provide Unicode (UTF-8) support on Windows.
+ *
+ * Copyright (c) 2018 Peter Wu <peter@lekensteyn.nl>
+ * SPDX-License-Identifier: (GPL-2.0-or-later OR MIT)
+ */
+
+#ifdef _WIN32
+#include <windows.h> /* for MultiByteToWideChar */
+#include <wchar.h> /* for _wrename */
+#include <stdio.h>
+#include <stdlib.h>
+#include <errno.h>
+
+// Set a high limit in case long paths are enabled.
+#define MAX_PATH_SIZE 4096
+#define MAX_MODE_SIZE 128
+// cmd.exe argument length is reportedly limited to 8192.
+#define MAX_CMD_SIZE 8192
+
+FILE *fopen_utf8(const char *pathname, const char *mode) {
+ wchar_t pathname_w[MAX_PATH_SIZE];
+ wchar_t mode_w[MAX_MODE_SIZE];
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, pathname, -1, pathname_w, MAX_PATH_SIZE) ||
+ !MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, mode, -1, mode_w, MAX_MODE_SIZE)) {
+ errno = EINVAL;
+ return NULL;
+ }
+ return _wfopen(pathname_w, mode_w);
+}
+
+FILE *freopen_utf8(const char *pathname, const char *mode, FILE *stream) {
+ wchar_t pathname_w[MAX_PATH_SIZE];
+ wchar_t mode_w[MAX_MODE_SIZE];
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, pathname, -1, pathname_w, MAX_PATH_SIZE) ||
+ !MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, mode, -1, mode_w, MAX_MODE_SIZE)) {
+ // Close stream as documented for the error case.
+ fclose(stream);
+ errno = EINVAL;
+ return NULL;
+ }
+ return _wfreopen(pathname_w, mode_w, stream);
+}
+
+int remove_utf8(const char *pathname) {
+ wchar_t pathname_w[MAX_PATH_SIZE];
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, pathname, -1, pathname_w, MAX_PATH_SIZE)) {
+ errno = EINVAL;
+ return -1;
+ }
+ return _wremove(pathname_w);
+}
+
+int rename_utf8(const char *oldpath, const char *newpath) {
+ wchar_t oldpath_w[MAX_PATH_SIZE];
+ wchar_t newpath_w[MAX_PATH_SIZE];
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, oldpath, -1, oldpath_w, MAX_PATH_SIZE) ||
+ !MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, newpath, -1, newpath_w, MAX_PATH_SIZE)) {
+ errno = EINVAL;
+ return -1;
+ }
+ return _wrename(oldpath_w, newpath_w);
+}
+
+FILE *popen_utf8(const char *command, const char *mode) {
+ wchar_t command_w[MAX_CMD_SIZE];
+ wchar_t mode_w[MAX_MODE_SIZE];
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, command, -1, command_w, MAX_CMD_SIZE) ||
+ !MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, mode, -1, mode_w, MAX_MODE_SIZE)) {
+ errno = EINVAL;
+ return NULL;
+ }
+ return _wpopen(command_w, mode_w);
+}
+
+int system_utf8(const char *command) {
+ wchar_t command_w[MAX_CMD_SIZE];
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, command, -1, command_w, MAX_CMD_SIZE)) {
+ errno = EINVAL;
+ return -1;
+ }
+ return _wsystem(command_w);
+}
+
+DWORD GetModuleFileNameA_utf8(HMODULE hModule, LPSTR lpFilename, DWORD nSize) {
+ wchar_t filename_w[MAX_PATH + 1];
+ if (!GetModuleFileNameW(hModule, filename_w, MAX_PATH + 1)) {
+ return 0;
+ }
+ return WideCharToMultiByte(CP_UTF8, WC_ERR_INVALID_CHARS, filename_w, -1, lpFilename, nSize, NULL, NULL);
+}
+
+HMODULE LoadLibraryExA_utf8(LPCSTR lpLibFileName, HANDLE hFile, DWORD dwFlags) {
+ wchar_t pathname_w[MAX_PATH_SIZE];
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, lpLibFileName, -1, pathname_w, MAX_PATH_SIZE)) {
+ SetLastError(ERROR_INVALID_NAME);
+ return NULL;
+ }
+ return LoadLibraryExW(pathname_w, hFile, dwFlags);
+}
+#endif
diff -ruN lua-5.4.3/src/utf8_wrappers.h newlua/src/utf8_wrappers.h
--- lua-5.4.3/src/utf8_wrappers.h 1970-01-01 07:30:00.000000000 +0730
+++ newlua/src/utf8_wrappers.h 2022-06-10 19:22:53.554879400 +0800
@@ -0,0 +1,42 @@
+/**
+ * Wrappers to provide Unicode (UTF-8) support on Windows.
+ *
+ * Copyright (c) 2018 Peter Wu <peter@lekensteyn.nl>
+ * SPDX-License-Identifier: (GPL-2.0-or-later OR MIT)
+ */
+
+#ifdef _WIN32
+
+#if defined(loadlib_c) || defined(lauxlib_c) || defined(liolib_c) || defined(luac_c)
+#include <stdio.h> /* for loadlib_c */
+FILE *fopen_utf8(const char *pathname, const char *mode);
+#define fopen fopen_utf8
+#endif
+
+#ifdef lauxlib_c
+FILE *freopen_utf8(const char *pathname, const char *mode, FILE *stream);
+#define freopen freopen_utf8
+#endif
+
+#ifdef liolib_c
+FILE *popen_utf8(const char *command, const char *mode);
+#define _popen popen_utf8
+#endif
+
+#ifdef loslib_c
+int remove_utf8(const char *pathname);
+int rename_utf8(const char *oldpath, const char *newpath);
+int system_utf8(const char *command);
+#define remove remove_utf8
+#define rename rename_utf8
+#define system system_utf8
+#endif
+
+#ifdef loadlib_c
+#include <windows.h>
+DWORD GetModuleFileNameA_utf8(HMODULE hModule, LPSTR lpFilename, DWORD nSize);
+HMODULE LoadLibraryExA_utf8(LPCSTR lpLibFileName, HANDLE hFile, DWORD dwFlags);
+#define GetModuleFileNameA GetModuleFileNameA_utf8
+#define LoadLibraryExA LoadLibraryExA_utf8
+#endif
+#endif

View File

@ -22,6 +22,7 @@ show_help() {
echo "-B --bundle Create an App bundle (macOS only)"
echo "-P --portable Create a portable binary package."
echo "-O --pgo Use profile guided optimizations (pgo)."
echo "-U --windows-lua-utf Use the UTF8 patch for Lua."
echo " macOS: disabled when used with --bundle,"
echo " Windows: Implicit being the only option."
echo
@ -35,6 +36,9 @@ main() {
local bundle
local portable
local pgo
local patch_lua
local lua_subproject_path
for i in "$@"; do
case $i in
@ -76,6 +80,10 @@ main() {
pgo="-Db_pgo=generate"
shift
;;
-U|--windows-lua-utf)
patch_lua="true"
shift
;;
*)
# unknown option
;;
@ -103,6 +111,11 @@ main() {
$pgo \
"${build_dir}"
lua_subproject_path=subprojects/lua-*/
if [[ $patch_lua == "true" ]] && [[ ! -z $force_fallback ]] && [[ -d $lua_subproject_path ]]; then
patch -d $lua_subproject_path -p1 --forward < resources/windows/001-lua-unicode.diff
fi
meson compile -C "${build_dir}"
if [ ! -z ${pgo+x} ]; then