From 75763bd57338b3efe8300ece094aacc8e1a7a467 Mon Sep 17 00:00:00 2001 From: Dan Gudmundsson Date: Mon, 25 Mar 2013 12:44:20 +0100 Subject: erts: Change erlang:open_port spawn to handle unicode Previously only 'spawn_executable' handled unicode input. Also change 'cd' option to always handle unicode. Update open_port documentation and tests --- erts/doc/src/erlang.xml | 66 +++--- erts/emulator/beam/erl_bif_port.c | 69 ++----- erts/emulator/beam/erl_unicode.c | 12 +- erts/emulator/beam/global.h | 6 + erts/emulator/sys/win32/sys.c | 412 ++++++++++++-------------------------- erts/emulator/test/port_SUITE.erl | 71 +++++-- 6 files changed, 242 insertions(+), 394 deletions(-) diff --git a/erts/doc/src/erlang.xml b/erts/doc/src/erlang.xml index 5befd51974..81e9084e07 100644 --- a/erts/doc/src/erlang.xml +++ b/erts/doc/src/erlang.xml @@ -2613,7 +2613,28 @@ os_prompt%

Returns a port identifier as the result of opening a new Erlang port. A port can be seen as an external Erlang - process. PortName is one of the following:

+ process. +

+

The name of the executable as well as the arguments + given in cd, env, args and arg0 is subject to + Unicode file name translation if the system is running + in Unicode file name mode. To avoid + translation or force i.e. UTF-8, supply the executable + and/or arguments as a binary in the correct + encoding. See the file module, the + + file:native_name_encoding/0 function and the + stdlib users guide + for details.

+ +

The characters in the name (if given as a list) + can only be > 255 if the Erlang VM is started in + Unicode file name translation mode, otherwise the name + of the executable is limited to the ISO-latin-1 + character set.

 + +

PortName is one of the following:

{spawn, Command} @@ -2668,25 +2689,6 @@ os_prompt% executed, the appropriate command interpreter will implicitly be invoked, but there will still be no command argument expansion or implicit PATH search.

- -

The name of the executable as well as the arguments - given in args and arg0 is subject to - Unicode file name translation if the system is running - in Unicode file name mode. To avoid - translation or force i.e. UTF-8, supply the executable - and/or arguments as a binary in the correct - encoding. See the file module, the - - file:native_name_encoding/0 function and the - stdlib users guide - for details.

- -

The characters in the name (if given as a list) - can only be > 255 if the Erlang VM is started in - Unicode file name translation mode, otherwise the name - of the executable is limited to the ISO-latin-1 - character set.

If the FileName cannot be run, an error exception, with the posix error code as the reason, is @@ -2762,11 +2764,7 @@ os_prompt% strings. The one exception is Val being the atom false (in analogy with os:getenv/1), which removes the environment variable. -

-

If Unicode filename encoding is in effect (see the erl manual - page), the strings (both Name and - Value) may contain characters with codepoints > 255.

+

{args, [ string() | binary() ]} @@ -2794,21 +2792,6 @@ os_prompt% should not be given in this list. The proper executable name will automatically be used as argv[0] where applicable.

-

When the Erlang VM is running in Unicode file name - mode, the arguments can contain any Unicode characters and - will be translated into whatever is appropriate on the - underlying OS, which means UTF-8 for all platforms except - Windows, which has other (more transparent) ways of - dealing with Unicode arguments to programs. To avoid - Unicode translation of arguments, they can be supplied as - binaries in whatever encoding is deemed appropriate.

- -

The characters in the arguments (if given as a - list of characters) can only be > 255 if the Erlang - VM is started in Unicode file name mode, - otherwise the arguments are limited to the - ISO-latin-1 character set.

 -

If one, for any reason, wants to explicitly set the program name in the argument vector, the arg0 option can be used.

@@ -2824,9 +2807,6 @@ os_prompt% responds to this is highly system dependent and no specific effect is guaranteed.

-

The unicode file name translation rules of the - args option apply to this option as well.

- exit_status diff --git a/erts/emulator/beam/erl_bif_port.c b/erts/emulator/beam/erl_bif_port.c index 109c54fd7f..5b26e53c9f 100644 --- a/erts/emulator/beam/erl_bif_port.c +++ b/erts/emulator/beam/erl_bif_port.c @@ -784,43 +784,29 @@ open_port(Process* p, Eterm name, Eterm settings, int *err_typep, int *err_nump) goto badarg; } - if (*tp == am_spawn || *tp == am_spawn_driver) { /* A process port */ + if (*tp == am_spawn || *tp == am_spawn_driver || *tp == am_spawn_executable) { /* A process port */ + int encoding; if (arity != make_arityval(2)) { goto badarg; } name = tp[1]; - if (is_atom(name)) { - name_buf = (char *) erts_alloc(ERTS_ALC_T_TMP, - atom_tab(atom_val(name))->len+1); - sys_memcpy((void *) name_buf, - (void *) atom_tab(atom_val(name))->name, - atom_tab(atom_val(name))->len); - name_buf[atom_tab(atom_val(name))->len] = '\0'; - } else if ((i = is_string(name))) { - name_buf = (char *) erts_alloc(ERTS_ALC_T_TMP, i + 1); - if (intlist_to_buf(name, name_buf, i) != i) - erl_exit(1, "%s:%d: Internal error\n", __FILE__, __LINE__); - name_buf[i] = '\0'; - } else { + encoding = erts_get_native_filename_encoding(); + /* Do not convert the command to utf-16le yet, do that in win32 specific code */ + /* since the cmd is used for comparsion with drivers names and copied to port info */ + if (encoding == ERL_FILENAME_WIN_WCHAR) { + encoding = ERL_FILENAME_UTF8; + } + if ((name_buf = erts_convert_filename_to_encoding(name, NULL, 0, ERTS_ALC_T_TMP,0,1, encoding, NULL)) + == NULL) { goto badarg; } + if (*tp == am_spawn_driver) { opts.spawn_type = ERTS_SPAWN_DRIVER; + } else if (*tp == am_spawn_executable) { + opts.spawn_type = ERTS_SPAWN_EXECUTABLE; } - driver = &spawn_driver; - } else if (*tp == am_spawn_executable) { /* A program */ - /* - * {spawn_executable,Progname} - */ - - if (arity != make_arityval(2)) { - goto badarg; - } - name = tp[1]; - if ((name_buf = erts_convert_filename_to_native(name, NULL, 0, ERTS_ALC_T_TMP,0,1, NULL)) == NULL) { - goto badarg; - } - opts.spawn_type = ERTS_SPAWN_EXECUTABLE; + driver = &spawn_driver; } else if (*tp == am_fd) { /* An fd port */ int n; @@ -861,29 +847,8 @@ open_port(Process* p, Eterm name, Eterm settings, int *err_typep, int *err_nump) } if (edir != NIL) { - /* A working directory is expressed differently if spawn_executable, i.e. Unicode is handles - for spawn_executable... */ - if (opts.spawn_type != ERTS_SPAWN_EXECUTABLE) { - Eterm iolist; - DeclareTmpHeap(heap,4,p); - int r; - - UseTmpHeap(4,p); - heap[0] = edir; - heap[1] = make_list(heap+2); - heap[2] = make_small(0); - heap[3] = NIL; - iolist = make_list(heap); - r = erts_iolist_to_buf(iolist, (char*) dir, MAXPATHLEN); - UnUseTmpHeap(4,p); - if (ERTS_IOLIST_TO_BUF_FAILED(r)) { - goto badarg; - } - opts.wd = (char *) dir; - } else { - if ((opts.wd = erts_convert_filename_to_native(edir, NULL, 0, ERTS_ALC_T_TMP,0,1,NULL)) == NULL) { - goto badarg; - } + if ((opts.wd = erts_convert_filename_to_native(edir, NULL, 0, ERTS_ALC_T_TMP,0,1,NULL)) == NULL) { + goto badarg; } } @@ -961,11 +926,11 @@ static char **convert_args(Eterm l) int n; int i = 0; Eterm str; - /* We require at least one element in list (argv[0]) */ if (is_not_list(l) && is_not_nil(l)) { return NULL; } n = list_length(l); + /* We require at least one element in argv[0] + NULL at end */ pp = erts_alloc(ERTS_ALC_T_TMP, (n + 2) * sizeof(char **)); pp[i++] = erts_default_arg0; while (is_list(l)) { diff --git a/erts/emulator/beam/erl_unicode.c b/erts/emulator/beam/erl_unicode.c index ad6f8b993a..71794e1101 100644 --- a/erts/emulator/beam/erl_unicode.c +++ b/erts/emulator/beam/erl_unicode.c @@ -1981,9 +1981,19 @@ BIF_RETTYPE binary_to_existing_atom_2(BIF_ALIST_2) * string routines, that will certainly fail on some OS. */ -char *erts_convert_filename_to_native(Eterm name, char *statbuf, size_t statbuf_size, ErtsAlcType_t alloc_type, int allow_empty, int allow_atom, Sint *used) +char *erts_convert_filename_to_native(Eterm name, char *statbuf, size_t statbuf_size, + ErtsAlcType_t alloc_type, int allow_empty, + int allow_atom, Sint *used) { int encoding = erts_get_native_filename_encoding(); + return erts_convert_filename_to_encoding(name, statbuf, statbuf_size, alloc_type, + allow_empty, allow_atom, encoding, used); +} + +char *erts_convert_filename_to_encoding(Eterm name, char *statbuf, size_t statbuf_size, + ErtsAlcType_t alloc_type, int allow_empty, + int allow_atom, int encoding, Sint *used) +{ char* name_buf = NULL; if ((allow_atom && is_atom(name)) || diff --git a/erts/emulator/beam/global.h b/erts/emulator/beam/global.h index 26ed5f82c1..f7e4c81181 100755 --- a/erts/emulator/beam/global.h +++ b/erts/emulator/beam/global.h @@ -788,6 +788,12 @@ char *erts_convert_filename_to_native(Eterm name, char *statbuf, ErtsAlcType_t alloc_type, int allow_empty, int allow_atom, Sint *used /* out */); +char *erts_convert_filename_to_encoding(Eterm name, char *statbuf, + size_t statbuf_size, + ErtsAlcType_t alloc_type, + int allow_empty, int allow_atom, + int encoding, + Sint *used /* out */); Eterm erts_convert_native_to_filename(Process *p, byte *bytes); Eterm erts_utf8_to_list(Process *p, Uint num, byte *bytes, Uint sz, Uint left, Uint *num_built, Uint *num_eaten, Eterm tail); diff --git a/erts/emulator/sys/win32/sys.c b/erts/emulator/sys/win32/sys.c index 922967c698..93bbae7dee 100755 --- a/erts/emulator/sys/win32/sys.c +++ b/erts/emulator/sys/win32/sys.c @@ -69,17 +69,14 @@ static void async_read_file(struct async_io* aio, LPVOID buf, DWORD numToRead); static int async_write_file(struct async_io* aio, LPVOID buf, DWORD numToWrite); static int get_overlapped_result(struct async_io* aio, LPDWORD pBytesRead, BOOL wait); -static BOOL create_child_process(char *, HANDLE, HANDLE, +static BOOL create_child_process(wchar_t *, HANDLE, HANDLE, HANDLE, LPHANDLE, LPDWORD, BOOL, - LPVOID, LPTSTR, unsigned, - char **, int *); + LPVOID, wchar_t*, unsigned, + wchar_t **, int *); static int create_pipe(LPHANDLE, LPHANDLE, BOOL, BOOL); -static int application_type(const char* originalName, char fullPath[MAX_PATH], +static int application_type(const wchar_t* originalName, wchar_t fullPath[MAX_PATH], BOOL search_in_path, BOOL handle_quotes, int *error_return); -static int application_type_w(const WCHAR *originalName, WCHAR fullPath[MAX_PATH], - BOOL search_in_path, BOOL handle_quotes, - int *error_return); HANDLE erts_service_event; @@ -1173,7 +1170,7 @@ spawn_init(void) } static ErlDrvData -spawn_start(ErlDrvPort port_num, char* name, SysDriverOpts* opts) +spawn_start(ErlDrvPort port_num, char* utf8_name, SysDriverOpts* opts) { HANDLE hToChild = INVALID_HANDLE_VALUE; /* Write handle to child. */ HANDLE hFromChild = INVALID_HANDLE_VALUE; /* Read handle from child. */ @@ -1189,7 +1186,9 @@ spawn_start(ErlDrvPort port_num, char* name, SysDriverOpts* opts) SECURITY_ATTRIBUTES sa = {sizeof(SECURITY_ATTRIBUTES), NULL, TRUE}; char* envir = opts->envir; int errno_return = -1; - + wchar_t *name; + int len; + if (opts->read_write & DO_READ) neededSelects++; if (opts->read_write & DO_WRITE) @@ -1247,26 +1246,40 @@ spawn_start(ErlDrvPort port_num, char* name, SysDriverOpts* opts) * Spawn the port program. */ - DEBUGF(("Spawning \"%s\"\n", name)); + if ((len = MultiByteToWideChar(CP_UTF8, 0, utf8_name, -1, NULL, 0)) > 0) { + name = erts_alloc(ERTS_ALC_T_TMP, len*sizeof(wchar_t)); + MultiByteToWideChar(CP_UTF8, 0, utf8_name, -1, name, len); + } else { /* Not valid utf-8, just convert byte to wchar */ + int i; + len = strlen(utf8_name); + name = erts_alloc(ERTS_ALC_T_TMP, (1+len)*sizeof(wchar_t)); + for(i=0; iport_pid, - &pid, - opts->hide_window, - (LPVOID) envir, - (LPTSTR) opts->wd, - opts->spawn_type, - opts->argv, - &errno_return); + ok = create_child_process(name, + hChildStdin, + hChildStdout, + hChildStderr, + &dp->port_pid, + &pid, + opts->hide_window, + (LPVOID) envir, + (wchar_t *) opts->wd, + opts->spawn_type, + (wchar_t **) opts->argv, + &errno_return); CloseHandle(hChildStdin); CloseHandle(hChildStdout); if (close_child_stderr && hChildStderr != INVALID_HANDLE_VALUE && hChildStderr != 0) { CloseHandle(hChildStderr); } + erts_free(ERTS_ALC_T_TMP, name); + if (envir != NULL) { erts_free(ERTS_ALC_T_ENVIRONMENT, envir); } @@ -1344,9 +1357,9 @@ create_file_thread(AsyncIo* aio, int mode) * Example: input = "\"Program Files\"\\erl arg1 arg2" * gives 19 as result. * The length returned is equivalent with length(argv[0]) if the - * comman line should have been prepared by _setargv for the main function + * command line should have been prepared by _setargv for the main function */ -int parse_command(char* cmd){ +int parse_command(wchar_t* cmd){ #define NORMAL 2 #define STRING 1 #define STOP 0 @@ -1354,17 +1367,17 @@ int parse_command(char* cmd){ int state = NORMAL; while (cmd[i]) { switch (cmd[i]) { - case '"': + case L'"': if (state == NORMAL) state = STRING; else state = NORMAL; break; - case '\\': - if ((state == STRING) && (cmd[i+1]=='"')) + case L'\\': + if ((state == STRING) && (cmd[i+1]==L'"')) i++; break; - case ' ': + case L' ': if (state == NORMAL) state = STOP; break; @@ -1379,7 +1392,7 @@ int parse_command(char* cmd){ return i; } -static BOOL need_quotes(WCHAR *str) +static BOOL need_quotes(wchar_t *str) { int in_quote = 0; int backslashed = 0; @@ -1443,7 +1456,7 @@ static BOOL need_quotes(WCHAR *str) static BOOL create_child_process ( - char *origcmd, /* Command line for child process (including + wchar_t *origcmd, /* Command line for child process (including * name of executable). Or whole executable if st is * ERTS_SPAWN_EXECUTABLE */ @@ -1454,57 +1467,53 @@ create_child_process LPDWORD pdwID, /* Pointer to variable to received Process ID */ BOOL hide, /* Hide the window unconditionally. */ LPVOID env, /* Environment for the child */ - LPTSTR wd, /* Working dir for the child */ + wchar_t *wd, /* Working dir for the child */ unsigned st, /* Flags for spawn, tells us how to interpret origcmd */ - char **argv, /* Argument vector if given. */ + wchar_t **argv, /* Argument vector if given. */ int *errno_return /* Place to put an errno in in case of failure */ ) -{ +{ PROCESS_INFORMATION piProcInfo = {0}; BOOL ok = FALSE; int applType; /* Not to be changed for different types of executables */ int staticCreateFlags = GetPriorityClass(GetCurrentProcess()); int createFlags = DETACHED_PROCESS; - char *newcmdline = NULL; + wchar_t *newcmdline = NULL; int cmdlength; - char* thecommand; - LPTSTR appname = NULL; + wchar_t* thecommand; + wchar_t* appname = NULL; HANDLE hProcess = GetCurrentProcess(); - - *errno_return = -1; + STARTUPINFOW siStartInfo = {0}; + wchar_t execPath[MAX_PATH]; + *errno_return = -1; + siStartInfo.cb = sizeof(STARTUPINFOW); + siStartInfo.dwFlags = STARTF_USESTDHANDLES; + siStartInfo.hStdInput = hStdin; + siStartInfo.hStdOutput = hStdout; + siStartInfo.hStdError = hStderr; if (st != ERTS_SPAWN_EXECUTABLE) { - STARTUPINFO siStartInfo = {0}; - char execPath[MAX_PATH]; - - siStartInfo.cb = sizeof(STARTUPINFO); - siStartInfo.dwFlags = STARTF_USESTDHANDLES; - siStartInfo.hStdInput = hStdin; - siStartInfo.hStdOutput = hStdout; - siStartInfo.hStdError = hStderr; - /* * Parse out the program name from the command line (it can be quoted and * contain spaces). */ - newcmdline = erts_alloc(ERTS_ALC_T_TMP, 2048); + newcmdline = (wchar_t *) erts_alloc(ERTS_ALC_T_TMP, 2048*sizeof(wchar_t)); cmdlength = parse_command(origcmd); - thecommand = (char *) erts_alloc(ERTS_ALC_T_TMP, cmdlength+1); - strncpy(thecommand, origcmd, cmdlength); - thecommand[cmdlength] = '\0'; - DEBUGF(("spawn command: %s\n", thecommand)); - - applType = application_type(thecommand, execPath, TRUE, - TRUE, errno_return); - DEBUGF(("application_type returned for (%s) is %d\n", thecommand, applType)); + thecommand = (wchar_t *) erts_alloc(ERTS_ALC_T_TMP, (cmdlength+1)*sizeof(wchar_t)); + wcsncpy(thecommand, origcmd, cmdlength); + thecommand[cmdlength] = L'\0'; + DEBUGF(("spawn command: %S\n", thecommand)); + + applType = application_type(thecommand, execPath, TRUE, TRUE, errno_return); + DEBUGF(("application_type returned for (%S) is %d\n", thecommand, applType)); erts_free(ERTS_ALC_T_TMP, (void *) thecommand); if (applType == APPL_NONE) { erts_free(ERTS_ALC_T_TMP,newcmdline); return FALSE; } - newcmdline[0] = '\0'; + newcmdline[0] = L'\0'; if (applType == APPL_DOS) { /* @@ -1513,11 +1522,11 @@ create_child_process * a normal process inside of a hidden console application, * and then run that hidden console as a detached process. */ - + siStartInfo.wShowWindow = SW_HIDE; siStartInfo.dwFlags |= STARTF_USESHOWWINDOW; createFlags = CREATE_NEW_CONSOLE; - strcat(newcmdline, "cmd.exe /c "); + wcscat(newcmdline, L"cmd.exe /c "); } else if (hide) { DEBUGF(("hiding window\n")); siStartInfo.wShowWindow = SW_HIDE; @@ -1525,35 +1534,25 @@ create_child_process createFlags = 0; } - strcat(newcmdline, execPath); - strcat(newcmdline, origcmd+cmdlength); - DEBUGF(("Creating child process: %s, createFlags = %d\n", newcmdline, createFlags)); - ok = CreateProcessA(appname, - newcmdline, - NULL, - NULL, - TRUE, - createFlags | staticCreateFlags | - CREATE_UNICODE_ENVIRONMENT, - env, - wd, - &siStartInfo, + wcscat(newcmdline, execPath); + wcscat(newcmdline, origcmd+cmdlength); + DEBUGF(("Creating child process: %S, createFlags = %d\n", newcmdline, createFlags)); + ok = CreateProcessW(appname, + newcmdline, + NULL, + NULL, + TRUE, + createFlags | staticCreateFlags | + CREATE_UNICODE_ENVIRONMENT, + env, + wd, + &siStartInfo, &piProcInfo); } else { /* ERTS_SPAWN_EXECUTABLE, filename and args are in unicode ({utf16,little}) */ int run_cmd = 0; - STARTUPINFOW siStartInfo = {0}; - WCHAR execPath[MAX_PATH]; - - siStartInfo.cb = sizeof(STARTUPINFOW); - siStartInfo.dwFlags = STARTF_USESTDHANDLES; - siStartInfo.hStdInput = hStdin; - siStartInfo.hStdOutput = hStdout; - siStartInfo.hStdError = hStderr; - - applType = application_type_w((WCHAR *) origcmd, execPath, FALSE, FALSE, - errno_return); + applType = application_type(origcmd, execPath, FALSE, FALSE, errno_return); if (applType == APPL_NONE) { return FALSE; } @@ -1573,37 +1572,37 @@ create_child_process createFlags = 0; } if (run_cmd) { - WCHAR cmdPath[MAX_PATH]; + wchar_t cmdPath[MAX_PATH]; int cmdType; - cmdType = application_type_w(L"cmd.exe", cmdPath, TRUE, FALSE, errno_return); + cmdType = application_type(L"cmd.exe", cmdPath, TRUE, FALSE, errno_return); if (cmdType == APPL_NONE || cmdType == APPL_DOS) { return FALSE; } - appname = (char *) erts_alloc(ERTS_ALC_T_TMP, (wcslen(cmdPath)+1)*sizeof(WCHAR)); - wcscpy((WCHAR *) appname,cmdPath); + appname = (wchar_t *) erts_alloc(ERTS_ALC_T_TMP, (wcslen(cmdPath)+1)*sizeof(wchar_t)); + wcscpy(appname,cmdPath); } else { - appname = (char *) erts_alloc(ERTS_ALC_T_TMP, (wcslen(execPath)+1)*sizeof(WCHAR)); - wcscpy((WCHAR *) appname, execPath); + appname = (wchar_t *) erts_alloc(ERTS_ALC_T_TMP, (wcslen(execPath)+1)*sizeof(wchar_t)); + wcscpy(appname, execPath); } if (argv == NULL) { BOOL orig_need_q = need_quotes(execPath); - WCHAR *ptr; + wchar_t *ptr; int ocl = wcslen(execPath); if (run_cmd) { - newcmdline = (char *) erts_alloc(ERTS_ALC_T_TMP, - (ocl + ((orig_need_q) ? 3 : 1) - + 11)*sizeof(WCHAR)); - memcpy(newcmdline,L"cmd.exe /c ",11*sizeof(WCHAR)); - ptr = (WCHAR *) (newcmdline + (11*sizeof(WCHAR))); + newcmdline = (wchar_t *) erts_alloc(ERTS_ALC_T_TMP, + (ocl + ((orig_need_q) ? 3 : 1) + + 11)*sizeof(wchar_t)); + memcpy(newcmdline,L"cmd.exe /c ",11*sizeof(wchar_t)); + ptr = newcmdline + 11; } else { - newcmdline = (char *) erts_alloc(ERTS_ALC_T_TMP, - (ocl + ((orig_need_q) ? 3 : 1))*sizeof(WCHAR)); - ptr = (WCHAR *) newcmdline; + newcmdline = (wchar_t *) erts_alloc(ERTS_ALC_T_TMP, + (ocl + ((orig_need_q) ? 3 : 1))*sizeof(wchar_t)); + ptr = (wchar_t *) newcmdline; } if (orig_need_q) { *ptr++ = L'"'; } - memcpy(ptr,execPath,ocl*sizeof(WCHAR)); + memcpy(ptr,execPath,ocl*sizeof(wchar_t)); ptr += ocl; if (orig_need_q) { *ptr++ = L'"'; @@ -1611,12 +1610,12 @@ create_child_process *ptr = L'\0'; } else { int sum = 1; /* '\0' */ - WCHAR **ar = (WCHAR **) argv; - WCHAR *n; - char *save_arg0 = NULL; - if (argv[0] == erts_default_arg0 || run_cmd) { + wchar_t **ar = argv; + wchar_t *n; + wchar_t *save_arg0 = NULL; + if (argv[0] == (wchar_t *) erts_default_arg0 || run_cmd) { save_arg0 = argv[0]; - argv[0] = (char *) execPath; + argv[0] = execPath; } if (run_cmd) { sum += 11; /* cmd.exe /c */ @@ -1629,11 +1628,11 @@ create_child_process sum++; /* space */ ++ar; } - ar = (WCHAR **) argv; - newcmdline = erts_alloc(ERTS_ALC_T_TMP, sum*sizeof(WCHAR)); - n = (WCHAR *) newcmdline; + ar = argv; + newcmdline = (wchar_t *) erts_alloc(ERTS_ALC_T_TMP, sum*sizeof(wchar_t)); + n = newcmdline; if (run_cmd) { - memcpy(n,L"cmd.exe /c ",11*sizeof(WCHAR)); + memcpy(n,L"cmd.exe /c ",11*sizeof(wchar_t)); n += 11; } while (*ar != NULL) { @@ -1642,7 +1641,7 @@ create_child_process if (q) { *n++ = L'"'; } - memcpy(n,*ar,sum*sizeof(WCHAR)); + memcpy(n,*ar,sum*sizeof(wchar_t)); n += sum; if (q) { *n++ = L'"'; @@ -1657,16 +1656,16 @@ create_child_process } DEBUGF(("Creating child process: %s, createFlags = %d\n", newcmdline, createFlags)); - ok = CreateProcessW((WCHAR *) appname, - (WCHAR *) newcmdline, - NULL, - NULL, - TRUE, - createFlags | staticCreateFlags | - CREATE_UNICODE_ENVIRONMENT, - env, - (WCHAR *) wd, - &siStartInfo, + ok = CreateProcessW((wchar_t *) appname, + (wchar_t *) newcmdline, + NULL, + NULL, + TRUE, + createFlags | staticCreateFlags | + CREATE_UNICODE_ENVIRONMENT, + env, + wd, + &siStartInfo, &piProcInfo); } /* end SPAWN_EXECUTABLE */ @@ -1775,180 +1774,23 @@ static int create_pipe(HANDLE *phRead, HANDLE *phWrite, BOOL inheritRead, BOOL o return TRUE; } - - - -static int application_type -( - const char *originalName, /* Name of the application to find. */ - char fullPath[MAX_PATH], /* Filled with complete path to - * application. */ - BOOL search_in_path, /* If we should search the system wide path */ - BOOL handle_quotes, /* If we should handle quotes around executable */ - int *error_return /* A place to put an error code */ - ) -{ - int applType, i; - HANDLE hFile; - char *ext, *rest; - char buf[2]; - DWORD read; - IMAGE_DOS_HEADER header; - static char extensions[][5] = {"", ".com", ".exe", ".bat"}; - int is_quoted; - int len; - - /* Look for the program as an external program. First try the name - * as it is, then try adding .com, .exe, and .bat, in that order, to - * the name, looking for an executable. - * NOTE! that we does not support execution of .com programs on Windows NT - * - * - * Using the raw SearchPath() procedure doesn't do quite what is - * necessary. If the name of the executable already contains a '.' - * character, it will not try appending the specified extension when - * searching (in other words, SearchPath will not find the program - * "a.b.exe" if the arguments specified "a.b" and ".exe"). - * So, first look for the file as it is named. Then manually append - * the extensions, looking for a match. (') - */ - - len = strlen(originalName); - is_quoted = handle_quotes && len > 0 && originalName[0] == '"' && - originalName[len-1] == '"'; - - applType = APPL_NONE; - *error_return = ENOENT; - for (i = 0; i < (int) (sizeof(extensions) / sizeof(extensions[0])); i++) { - if(is_quoted) { - lstrcpyn(fullPath, originalName+1, MAX_PATH - 7); - len = strlen(fullPath); - if(len > 0) { - fullPath[len-1] = '\0'; - } - } else { - lstrcpyn(fullPath, originalName, MAX_PATH - 5); - } - lstrcat(fullPath, extensions[i]); - SearchPath((search_in_path) ? NULL : ".", fullPath, NULL, MAX_PATH, fullPath, &rest); - - /* - * Ignore matches on directories or data files, return if identified - * a known type. - */ - - if (GetFileAttributes(fullPath) & FILE_ATTRIBUTE_DIRECTORY) { - continue; - } - - ext = strrchr(fullPath, '.'); - if ((ext != NULL) && (strcmpi(ext, ".bat") == 0)) { - *error_return = EACCES; - applType = APPL_DOS; - break; - } - - hFile = CreateFile(fullPath, GENERIC_READ, FILE_SHARE_READ, NULL, - OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL); - if (hFile == INVALID_HANDLE_VALUE) { - continue; - } - - *error_return = EACCES; /* If considered an error, - it's an access error */ - header.e_magic = 0; - ReadFile(hFile, (void *) &header, sizeof(header), &read, NULL); - if (header.e_magic != IMAGE_DOS_SIGNATURE) { - /* - * Doesn't have the magic number for relocatable executables. If - * filename ends with .com, assume it's a DOS application anyhow. - * Note that we didn't make this assumption at first, because some - * supposed .com files are really 32-bit executables with all the - * magic numbers and everything. - */ - - CloseHandle(hFile); - if ((ext != NULL) && (strcmpi(ext, ".com") == 0)) { - applType = APPL_DOS; - break; - } - continue; - } - if (header.e_lfarlc != sizeof(header)) { - /* - * All Windows 3.X and Win32 and some DOS programs have this value - * set here. If it doesn't, assume that since it already had the - * other magic number it was a DOS application. - */ - - CloseHandle(hFile); - applType = APPL_DOS; - break; - } - - /* - * The DWORD at header.e_lfanew points to yet another magic number. - */ - - buf[0] = '\0'; - SetFilePointer(hFile, header.e_lfanew, NULL, FILE_BEGIN); - ReadFile(hFile, (void *) buf, 2, &read, NULL); - CloseHandle(hFile); - - if ((buf[0] == 'L') && (buf[1] == 'E')) { - applType = APPL_DOS; - } else if ((buf[0] == 'N') && (buf[1] == 'E')) { - applType = APPL_WIN3X; - } else if ((buf[0] == 'P') && (buf[1] == 'E')) { - applType = APPL_WIN32; - } else { - continue; - } - break; - } - - if (applType == APPL_NONE) { - return APPL_NONE; - } - - if ((applType == APPL_DOS) || (applType == APPL_WIN3X)) { - /* - * Replace long path name of executable with short path name for - * 16-bit applications. Otherwise the application may not be able - * to correctly parse its own command line to separate off the - * application name from the arguments. - */ - - GetShortPathName(fullPath, fullPath, MAX_PATH); - } - if (is_quoted) { - /* restore quotes on quoted program name */ - len = strlen(fullPath); - memmove(fullPath+1,fullPath,len); - fullPath[0]='"'; - fullPath[len+1]='"'; - fullPath[len+2]='\0'; - } - return applType; -} - -static int application_type_w (const WCHAR *originalName, /* Name of the application to find. */ - WCHAR wfullpath[MAX_PATH],/* Filled with complete path to +static int application_type (const wchar_t *originalName, /* Name of the application to find. */ + wchar_t wfullpath[MAX_PATH],/* Filled with complete path to * application. */ - BOOL search_in_path, /* If we should search the system wide path */ - BOOL handle_quotes, /* If we should handle quotes around executable */ - int *error_return) /* A place to put an error code */ + BOOL search_in_path, /* If we should search the system wide path */ + BOOL handle_quotes, /* If we should handle quotes around executable */ + int *error_return) /* A place to put an error code */ { int applType, i; HANDLE hFile; - WCHAR *ext, *rest; + wchar_t *ext, *rest; char buf[2]; DWORD read; IMAGE_DOS_HEADER header; - static WCHAR extensions[][5] = {L"", L".com", L".exe", L".bat"}; + static wchar_t extensions[][5] = {L"", L".com", L".exe", L".bat"}; int is_quoted; int len; - WCHAR xfullpath[MAX_PATH]; + wchar_t xfullpath[MAX_PATH]; len = wcslen(originalName); is_quoted = handle_quotes && len > 0 && originalName[0] == L'"' && @@ -2063,7 +1905,7 @@ static int application_type_w (const WCHAR *originalName, /* Name of the applica if (is_quoted) { /* restore quotes on quoted program name */ len = wcslen(wfullpath); - memmove(wfullpath+1,wfullpath,len*sizeof(WCHAR)); + memmove(wfullpath+1,wfullpath,len*sizeof(wchar_t)); wfullpath[0]=L'"'; wfullpath[len+1]=L'"'; wfullpath[len+2]=L'\0'; diff --git a/erts/emulator/test/port_SUITE.erl b/erts/emulator/test/port_SUITE.erl index 13aa0f4c00..8f7f98f9e3 100644 --- a/erts/emulator/test/port_SUITE.erl +++ b/erts/emulator/test/port_SUITE.erl @@ -1,7 +1,7 @@ %% %% %CopyrightBegin% %% -%% Copyright Ericsson AB 1997-2012. All Rights Reserved. +%% Copyright Ericsson AB 1997-2013. All Rights Reserved. %% %% The contents of this file are subject to the Erlang Public License, %% Version 1.1, (the "License"); you may not use this file except in @@ -942,6 +942,20 @@ cd(Config) when is_list(Config) -> Other2 -> test_server:fail({env, Other2}) end, + _ = open_port({spawn, Cmd}, + [{cd, unicode:characters_to_binary(TestDir)}, + {line, 256}]), + receive + {_, {data, {eol, String2}}} -> + case filename_equal(String2, TestDir) of + true -> + ok; + false -> + test_server:fail({cd, String2}) + end; + Other3 -> + test_server:fail({env, Other3}) + end, test_server:timetrap_cancel(Dog), ok. @@ -1346,19 +1360,28 @@ spawn_executable(Config) when is_list(Config) -> EchoArgs1 = filename:join([DataDir,"echo_args"]), ExactFile1 = filename:nativename(os:find_executable(EchoArgs1)), [ExactFile1] = run_echo_args(DataDir,[]), + [ExactFile1] = run_echo_args(DataDir,[binary]), ["echo_args"] = run_echo_args(DataDir,["echo_args"]), + ["echo_args"] = run_echo_args(DataDir,[binary, "echo_args"]), ["echo_arguments"] = run_echo_args(DataDir,["echo_arguments"]), + ["echo_arguments"] = run_echo_args(DataDir,[binary, "echo_arguments"]), [ExactFile1,"hello world","dlrow olleh"] = run_echo_args(DataDir,[ExactFile1,"hello world","dlrow olleh"]), [ExactFile1] = run_echo_args(DataDir,[default]), + [ExactFile1] = run_echo_args(DataDir,[binary, default]), [ExactFile1,"hello world","dlrow olleh"] = run_echo_args(DataDir,[switch_order,ExactFile1,"hello world", "dlrow olleh"]), [ExactFile1,"hello world","dlrow olleh"] = + run_echo_args(DataDir,[binary,switch_order,ExactFile1,"hello world", + "dlrow olleh"]), + [ExactFile1,"hello world","dlrow olleh"] = run_echo_args(DataDir,[default,"hello world","dlrow olleh"]), [ExactFile1,"hello world","dlrow olleh"] = run_echo_args_2("\""++ExactFile1++"\" "++"\"hello world\" \"dlrow olleh\""), + [ExactFile1,"hello world","dlrow olleh"] = + run_echo_args_2(unicode:characters_to_binary("\""++ExactFile1++"\" "++"\"hello world\" \"dlrow olleh\"")), PrivDir = ?config(priv_dir, Config), SpaceDir =filename:join([PrivDir,"With Spaces"]), @@ -1373,6 +1396,8 @@ spawn_executable(Config) when is_list(Config) -> ["echo_arguments"] = run_echo_args(SpaceDir,["echo_arguments"]), [ExactFile2,"hello world","dlrow olleh"] = run_echo_args(SpaceDir,[ExactFile2,"hello world","dlrow olleh"]), + [ExactFile2,"hello world","dlrow olleh"] = + run_echo_args(SpaceDir,[binary, ExactFile2,"hello world","dlrow olleh"]), [ExactFile2] = run_echo_args(SpaceDir,[default]), [ExactFile2,"hello world","dlrow olleh"] = run_echo_args(SpaceDir,[switch_order,ExactFile2,"hello world", @@ -1381,6 +1406,8 @@ spawn_executable(Config) when is_list(Config) -> run_echo_args(SpaceDir,[default,"hello world","dlrow olleh"]), [ExactFile2,"hello world","dlrow olleh"] = run_echo_args_2("\""++ExactFile2++"\" "++"\"hello world\" \"dlrow olleh\""), + [ExactFile2,"hello world","dlrow olleh"] = + run_echo_args_2(unicode:characters_to_binary("\""++ExactFile2++"\" "++"\"hello world\" \"dlrow olleh\"")), ExeExt = case string:to_lower(lists:last(string:tokens(ExactFile2,"."))) of @@ -1408,9 +1435,12 @@ spawn_executable(Config) when is_list(Config) -> [default,"hello world","dlrow olleh"]), [ExactFile3,"hello world","dlrow olleh"] = run_echo_args_2("\""++ExactFile3++"\" "++"\"hello world\" \"dlrow olleh\""), + [ExactFile3,"hello world","dlrow olleh"] = + run_echo_args_2(unicode:characters_to_binary("\""++ExactFile3++"\" "++"\"hello world\" \"dlrow olleh\"")), {'EXIT',{enoent,_}} = (catch run_echo_args(SpaceDir,"fnurflmonfi", [default,"hello world", "dlrow olleh"])), + NonExec = "kronxfrt"++ExeExt, file:write_file(filename:join([SpaceDir,NonExec]), <<"Not an executable">>), @@ -1511,25 +1541,40 @@ run_echo_args_2(FullnameAndArgs) -> run_echo_args(Where,Args) -> - run_echo_args(Where,"echo_args",Args). + run_echo_args(Where,"echo_args",Args). run_echo_args(Where,Prog,Args) -> - ArgvArg = case Args of - [] -> - []; - [default|T] -> - [{args,T}]; - [switch_order,H|T] -> - [{args,T},{arg0,H}]; - [H|T] -> - [{arg0,H},{args,T}] + {Binary, ArgvArg} = pack_argv(Args), + Command0 = filename:join([Where,Prog]), + Command = case Binary of + true -> unicode:characters_to_binary(Command0); + false -> Command0 end, - Command = filename:join([Where,Prog]), Port = open_port({spawn_executable,Command},ArgvArg++[eof]), Data = collect_data(Port), Port ! {self(), close}, receive {Port, closed} -> ok end, parse_echo_args_output(Data). - + +pack_argv([binary|Args]) -> + {true, pack_argv(Args, true)}; +pack_argv(Args) -> + {false, pack_argv(Args, false)}. + +pack_argv(Args, Binary) -> + case Args of + [] -> + []; + [default|T] -> + [{args,[make_bin(Arg,Binary) || Arg <- T]}]; + [switch_order,H|T] -> + [{args,[make_bin(Arg,Binary) || Arg <- T]},{arg0,make_bin(H,Binary)}]; + [H|T] -> + [{arg0,make_bin(H,Binary)},{args,[make_bin(Arg,Binary) || Arg <- T]}] + end. + +make_bin(Str, false) -> Str; +make_bin(Str, true) -> unicode:characters_to_binary(Str). + collect_data(Port) -> receive {Port, {data, Data}} -> -- cgit v1.2.3 From 892fb14882ef356c90eda8aa02ca43c3db8b53da Mon Sep 17 00:00:00 2001 From: Dan Gudmundsson Date: Mon, 22 Apr 2013 13:18:29 +0200 Subject: erts: Change erlang:open_port/2 spec --- erts/preloaded/ebin/erlang.beam | Bin 93520 -> 93584 bytes erts/preloaded/src/erlang.erl | 6 +++--- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/erts/preloaded/ebin/erlang.beam b/erts/preloaded/ebin/erlang.beam index a5c43435df..226e4c8a1f 100644 Binary files a/erts/preloaded/ebin/erlang.beam and b/erts/preloaded/ebin/erlang.beam differ diff --git a/erts/preloaded/src/erlang.erl b/erts/preloaded/src/erlang.erl index 6929ca3fa5..0df0768365 100644 --- a/erts/preloaded/src/erlang.erl +++ b/erts/preloaded/src/erlang.erl @@ -1714,15 +1714,15 @@ nodes(_Arg) -> erlang:nif_error(undefined). -spec open_port(PortName, PortSettings) -> port() when - PortName :: {spawn, Command :: string()} | - {spawn_driver, Command :: [byte()]} | + PortName :: {spawn, Command :: string() | binary()} | + {spawn_driver, Command :: string() | binary()} | {spawn_executable, FileName :: file:name() } | {fd, In :: non_neg_integer(), Out :: non_neg_integer()}, PortSettings :: [Opt], Opt :: {packet, N :: 1 | 2 | 4} | stream | {line, L :: non_neg_integer()} - | {cd, Dir :: string()} + | {cd, Dir :: string() | binary()} | {env, Env :: [{Name :: string(), Val :: string() | false}]} | {args, [string() | binary()]} | {arg0, string() | binary()} -- cgit v1.2.3 From 0057676c3ebfd603658631088c977219ba2b0cc6 Mon Sep 17 00:00:00 2001 From: Dan Gudmundsson Date: Wed, 15 May 2013 15:16:50 +0200 Subject: kernel: Handle unicode in os:cmd Allow unicode strings in os:cmd and try to convert the result bytes to unicode list. Also test it. --- lib/kernel/src/os.erl | 37 +++++++++++++++++++-------------- lib/kernel/test/os_SUITE.erl | 26 +++++++++++++++++------ lib/kernel/test/os_SUITE_data/my_echo.c | 28 +++++++++++++++++++++++++ 3 files changed, 69 insertions(+), 22 deletions(-) diff --git a/lib/kernel/src/os.erl b/lib/kernel/src/os.erl index 86f2f94f1f..c92bb463c1 100644 --- a/lib/kernel/src/os.erl +++ b/lib/kernel/src/os.erl @@ -183,20 +183,25 @@ extensions() -> Command :: atom() | io_lib:chars(). cmd(Cmd) -> validate(Cmd), - case type() of - {unix, _} -> - unix_cmd(Cmd); - {win32, Wtype} -> - Command0 = case {os:getenv("COMSPEC"),Wtype} of - {false,windows} -> lists:concat(["command.com /c", Cmd]); - {false,_} -> lists:concat(["cmd /c", Cmd]); - {Cspec,_} -> lists:concat([Cspec," /c",Cmd]) - end, - %% open_port/2 awaits string() in Command, but io_lib:chars() can be - %% deep lists according to io_lib module description. - Command = lists:flatten(Command0), - Port = open_port({spawn, Command}, [stream, in, eof, hide]), - get_data(Port, []) + Bytes = case type() of + {unix, _} -> + unix_cmd(Cmd); + {win32, Wtype} -> + Command0 = case {os:getenv("COMSPEC"),Wtype} of + {false,windows} -> lists:concat(["command.com /c", Cmd]); + {false,_} -> lists:concat(["cmd /c", Cmd]); + {Cspec,_} -> lists:concat([Cspec," /c",Cmd]) + end, + %% open_port/2 awaits string() in Command, but io_lib:chars() can be + %% deep lists according to io_lib module description. + Command = lists:flatten(Command0), + Port = open_port({spawn, Command}, [stream, in, eof, hide]), + get_data(Port, []) + end, + String = unicode:characters_to_list(list_to_binary(Bytes)), + if %% Convert to unicode list if possible otherwise return bytes + is_list(String) -> String; + true -> Bytes end. unix_cmd(Cmd) -> @@ -337,7 +342,7 @@ mk_cmd(Cmd) when is_atom(Cmd) -> % backward comp. mk_cmd(Cmd) -> %% We insert a new line after the command, in case the command %% contains a comment character. - io_lib:format("(~ts\n) @@ -345,7 +350,7 @@ validate(Atom) when is_atom(Atom) -> validate(List) when is_list(List) -> validate1(List). -validate1([C|Rest]) when is_integer(C), 0 =< C, C < 256 -> +validate1([C|Rest]) when is_integer(C) -> validate1(Rest); validate1([List|Rest]) when is_list(List) -> validate1(List), diff --git a/lib/kernel/test/os_SUITE.erl b/lib/kernel/test/os_SUITE.erl index 73ed704ae3..9b474c4cdf 100644 --- a/lib/kernel/test/os_SUITE.erl +++ b/lib/kernel/test/os_SUITE.erl @@ -20,7 +20,7 @@ -export([all/0, suite/0,groups/0,init_per_suite/1, end_per_suite/1, init_per_group/2,end_per_group/2]). --export([space_in_cwd/1, quoting/1, space_in_name/1, bad_command/1, +-export([space_in_cwd/1, quoting/1, cmd_unicode/1, space_in_name/1, bad_command/1, find_executable/1, unix_comment_in_command/1, deep_list_command/1, evil/1]). -include_lib("test_server/include/test_server.hrl"). @@ -28,9 +28,8 @@ suite() -> [{ct_hooks,[ts_install_cth]}]. all() -> - [space_in_cwd, quoting, space_in_name, bad_command, - find_executable, unix_comment_in_command, deep_list_command, - evil]. + [space_in_cwd, quoting, cmd_unicode, space_in_name, bad_command, + find_executable, unix_comment_in_command, deep_list_command, evil]. groups() -> []. @@ -95,6 +94,21 @@ quoting(Config) when is_list(Config) -> ?line [] = receive_all(), ok. + +cmd_unicode(doc) -> "Test that unicode arguments work."; +cmd_unicode(suite) -> []; +cmd_unicode(Config) when is_list(Config) -> + ?line DataDir = ?config(data_dir, Config), + ?line Echo = filename:join(DataDir, "my_echo"), + + ?line comp("one", os:cmd(Echo ++ " one")), + ?line comp("one::two", os:cmd(Echo ++ " one two")), + ?line comp("åäö::ϼΩ", os:cmd(Echo ++ " åäö " ++ [1020, 937])), + ?t:sleep(5), + ?line [] = receive_all(), + ok. + + space_in_name(doc) -> "Test that program with a space in its name can be executed."; space_in_name(suite) -> []; @@ -302,8 +316,8 @@ comp(Expected, Got) -> Expected -> ok; Other -> - ok = io:format("Expected: ~s\n", [Expected]), - ok = io:format("Got: ~s\n", [Other]), + ok = io:format("Expected: ~ts\n", [Expected]), + ok = io:format("Got: ~ts\n", [Other]), test_server:fail() end. diff --git a/lib/kernel/test/os_SUITE_data/my_echo.c b/lib/kernel/test/os_SUITE_data/my_echo.c index 2127511dd1..712c828bb5 100644 --- a/lib/kernel/test/os_SUITE_data/my_echo.c +++ b/lib/kernel/test/os_SUITE_data/my_echo.c @@ -1,3 +1,30 @@ +#ifdef __WIN32__ +#include + +int wmain(int argc, wchar_t **argv) +{ + char* sep = ""; + int len; + + /* + * Echo all arguments separated with '::', so that we can check that + * quotes are interpreted correctly. + */ + + while (argc-- > 1) { + char *utf8; + len = WideCharToMultiByte(CP_UTF8, 0, argv[1], -1, NULL, 0, NULL, NULL); + utf8 = malloc(len*sizeof(char)); + WideCharToMultiByte(CP_UTF8, 0, argv++[1], -1, utf8, len, NULL, NULL); + printf("%s%s", sep, utf8); + free(utf8); + sep = "::"; + } + putchar('\n'); + return 0; +} +#else + #include int @@ -17,3 +44,4 @@ main(int argc, char** argv) putchar('\n'); return 0; } +#endif -- cgit v1.2.3