fix(windows-utf8-patch): fix os.getenv() not supporting UTF-8 output (#1397)

This commit is contained in:
Takase 2023-04-08 01:11:01 +08:00 committed by George Sokianos
parent 612ebebb1f
commit d4ff3cb094
1 changed files with 47 additions and 17 deletions

View File

@ -1,6 +1,6 @@
diff -ruN lua-5.4.4/meson.build lua-5.4.4-mod/meson.build diff -ruN lua-5.4.4\meson.build lua-5.4.4-patched\meson.build
--- lua-5.4.4/meson.build 2022-11-16 10:33:38.424383300 +0800 --- lua-5.4.4\meson.build Wed Feb 22 18:16:56 2023
+++ lua-5.4.4-mod/meson.build 2022-11-16 09:40:57.697918000 +0800 +++ lua-5.4.4-patched\meson.build Wed Feb 22 04:10:01 2023
@@ -85,6 +85,7 @@ @@ -85,6 +85,7 @@
'src/lutf8lib.c', 'src/lutf8lib.c',
'src/lvm.c', 'src/lvm.c',
@ -9,9 +9,9 @@ diff -ruN lua-5.4.4/meson.build lua-5.4.4-mod/meson.build
dependencies: lua_lib_deps, dependencies: lua_lib_deps,
version: meson.project_version(), version: meson.project_version(),
soversion: lua_versions[0] + '.' + lua_versions[1], soversion: lua_versions[0] + '.' + lua_versions[1],
diff -ruN lua-5.4.4/src/luaconf.h lua-5.4.4-mod/src/luaconf.h diff -ruN lua-5.4.4\src\luaconf.h lua-5.4.4-patched\src\luaconf.h
--- lua-5.4.4/src/luaconf.h 2022-01-13 19:24:43.000000000 +0800 --- lua-5.4.4\src\luaconf.h Thu Jan 13 19:24:43 2022
+++ lua-5.4.4-mod/src/luaconf.h 2022-11-16 09:40:57.703926000 +0800 +++ lua-5.4.4-patched\src\luaconf.h Wed Feb 22 04:10:02 2023
@@ -782,5 +782,15 @@ @@ -782,5 +782,15 @@
@ -28,9 +28,9 @@ diff -ruN lua-5.4.4/src/luaconf.h lua-5.4.4-mod/src/luaconf.h
+ +
#endif #endif
diff -ruN lua-5.4.4/src/Makefile lua-5.4.4-mod/src/Makefile diff -ruN lua-5.4.4\src\Makefile lua-5.4.4-patched\src\Makefile
--- lua-5.4.4/src/Makefile 2021-07-15 22:01:52.000000000 +0800 --- lua-5.4.4\src\Makefile Thu Jul 15 22:01:52 2021
+++ lua-5.4.4-mod/src/Makefile 2022-11-16 09:40:57.708921800 +0800 +++ lua-5.4.4-patched\src\Makefile Wed Feb 22 04:10:02 2023
@@ -33,7 +33,7 @@ @@ -33,7 +33,7 @@
PLATS= guess aix bsd c89 freebsd generic linux linux-readline macosx mingw posix solaris PLATS= guess aix bsd c89 freebsd generic linux linux-readline macosx mingw posix solaris
@ -40,10 +40,10 @@ diff -ruN lua-5.4.4/src/Makefile lua-5.4.4-mod/src/Makefile
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 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) BASE_O= $(CORE_O) $(LIB_O) $(MYOBJS)
diff -ruN lua-5.4.4/src/utf8_wrappers.c lua-5.4.4-mod/src/utf8_wrappers.c diff -ruN lua-5.4.4\src\utf8_wrappers.c lua-5.4.4-patched\src\utf8_wrappers.c
--- lua-5.4.4/src/utf8_wrappers.c 1970-01-01 07:30:00.000000000 +0730 --- lua-5.4.4\src\utf8_wrappers.c Thu Jan 01 08:00:00 1970
+++ lua-5.4.4-mod/src/utf8_wrappers.c 2022-11-16 10:09:04.583866600 +0800 +++ lua-5.4.4-patched\src\utf8_wrappers.c Wed Feb 22 18:13:45 2023
@@ -0,0 +1,101 @@ @@ -0,0 +1,129 @@
+/** +/**
+ * Wrappers to provide Unicode (UTF-8) support on Windows. + * Wrappers to provide Unicode (UTF-8) support on Windows.
+ * + *
@ -58,12 +58,17 @@ diff -ruN lua-5.4.4/src/utf8_wrappers.c lua-5.4.4-mod/src/utf8_wrappers.c
+#include <stdlib.h> +#include <stdlib.h>
+#include <errno.h> +#include <errno.h>
+ +
+// A environment variable has the maximum length of 32767 characters
+// including the terminator.
+#define MAX_ENV_SIZE 32767
+// Set a high limit in case long paths are enabled. +// Set a high limit in case long paths are enabled.
+#define MAX_PATH_SIZE 4096 +#define MAX_PATH_SIZE 4096
+#define MAX_MODE_SIZE 128 +#define MAX_MODE_SIZE 128
+// cmd.exe argument length is reportedly limited to 8192. +// cmd.exe argument length is reportedly limited to 8192.
+#define MAX_CMD_SIZE 8192 +#define MAX_CMD_SIZE 8192
+ +
+static char env_value[MAX_ENV_SIZE];
+
+FILE *fopen_utf8(const char *pathname, const char *mode) { +FILE *fopen_utf8(const char *pathname, const char *mode) {
+ wchar_t pathname_w[MAX_PATH_SIZE]; + wchar_t pathname_w[MAX_PATH_SIZE];
+ wchar_t mode_w[MAX_MODE_SIZE]; + wchar_t mode_w[MAX_MODE_SIZE];
@ -144,11 +149,34 @@ diff -ruN lua-5.4.4/src/utf8_wrappers.c lua-5.4.4-mod/src/utf8_wrappers.c
+ } + }
+ return LoadLibraryExW(pathname_w, hFile, dwFlags); + return LoadLibraryExW(pathname_w, hFile, dwFlags);
+} +}
+
+char* getenv_utf8(const char *varname) {
+ /** This implementation is not thread safe.
+ * The string is only valid until the next call to getenv.
+ * This behavior is allowed per POSIX.1-2017 where it was said that:
+ * > The returned string pointer might be invalidated or the string content might be overwritten by a subsequent call to getenv(), setenv(), unsetenv(), or (if supported) putenv() but they shall not be affected by a call to any other function in this volume of POSIX.1-2017.
+ * > The returned string pointer might also be invalidated if the calling thread is terminated.
+ * > The getenv() function need not be thread-safe.
+ */
+ wchar_t *value_w;
+ wchar_t varname_w[MAX_ENV_SIZE];
+
+ if (!MultiByteToWideChar(CP_UTF8, MB_ERR_INVALID_CHARS, varname, -1, varname_w, MAX_ENV_SIZE))
+ return NULL;
+ value_w = _wgetenv((const wchar_t *) varname_w);
+ if (!value_w)
+ return NULL;
+
+ if (!WideCharToMultiByte(CP_UTF8, WC_ERR_INVALID_CHARS, value_w, -1, env_value, MAX_ENV_SIZE, NULL, NULL))
+ return NULL;
+
+ return env_value;
+}
+#endif +#endif
diff -ruN lua-5.4.4/src/utf8_wrappers.h lua-5.4.4-mod/src/utf8_wrappers.h diff -ruN lua-5.4.4\src\utf8_wrappers.h lua-5.4.4-patched\src\utf8_wrappers.h
--- lua-5.4.4/src/utf8_wrappers.h 1970-01-01 07:30:00.000000000 +0730 --- lua-5.4.4\src\utf8_wrappers.h Thu Jan 01 08:00:00 1970
+++ lua-5.4.4-mod/src/utf8_wrappers.h 2022-11-16 10:29:46.044102000 +0800 +++ lua-5.4.4-patched\src\utf8_wrappers.h Wed Feb 22 18:09:48 2023
@@ -0,0 +1,44 @@ @@ -0,0 +1,46 @@
+/** +/**
+ * Wrappers to provide Unicode (UTF-8) support on Windows. + * Wrappers to provide Unicode (UTF-8) support on Windows.
+ * + *
@ -180,9 +208,11 @@ diff -ruN lua-5.4.4/src/utf8_wrappers.h lua-5.4.4-mod/src/utf8_wrappers.h
+int remove_utf8(const char *pathname); +int remove_utf8(const char *pathname);
+int rename_utf8(const char *oldpath, const char *newpath); +int rename_utf8(const char *oldpath, const char *newpath);
+int system_utf8(const char *command); +int system_utf8(const char *command);
+char *getenv_utf8(const char *varname);
+#define remove remove_utf8 +#define remove remove_utf8
+#define rename rename_utf8 +#define rename rename_utf8
+#define system system_utf8 +#define system system_utf8
+#define getenv getenv_utf8
+#endif +#endif
+ +
+#ifdef loadlib_c +#ifdef loadlib_c