diff options
Diffstat (limited to 'ruby.c')
| -rw-r--r-- | ruby.c | 1716 |
1 files changed, 1131 insertions, 585 deletions
@@ -15,8 +15,10 @@ #include <windows.h> #include <sys/cygwin.h> #endif -#include "ruby/ruby.h" #include "ruby/encoding.h" +#include "ruby/thread.h" +#include "ruby/version.h" +#include "internal.h" #include "eval_intern.h" #include "dln.h" #include <stdio.h> @@ -44,49 +46,90 @@ #ifndef MAXPATHLEN # define MAXPATHLEN 1024 #endif +#ifndef O_ACCMODE +# define O_ACCMODE (O_RDONLY | O_WRONLY | O_RDWR) +#endif #include "ruby/util.h" +#include "mjit.h" + +void Init_ruby_description(void); + #ifndef HAVE_STDLIB_H char *getenv(); #endif -VALUE rb_parser_get_yydebug(VALUE); -VALUE rb_parser_set_yydebug(VALUE, VALUE); - -const char *ruby_get_inplace_mode(void); -void ruby_set_inplace_mode(const char *); +#ifndef DISABLE_RUBYGEMS +# define DISABLE_RUBYGEMS 0 +#endif +#if DISABLE_RUBYGEMS +#define DEFAULT_RUBYGEMS_ENABLED "disabled" +#else +#define DEFAULT_RUBYGEMS_ENABLED "enabled" +#endif -#define DISABLE_BIT(bit) (1U << disable_##bit) -enum disable_flag_bits { - disable_gems, - disable_rubyopt, - disable_flag_count +#define COMMA , +#define FEATURE_BIT(bit) (1U << feature_##bit) +#define EACH_FEATURES(X, SEP) \ + X(gems) \ + SEP \ + X(did_you_mean) \ + SEP \ + X(rubyopt) \ + SEP \ + X(frozen_string_literal) \ + SEP \ + X(jit) \ + /* END OF FEATURES */ +#define EACH_DEBUG_FEATURES(X, SEP) \ + X(frozen_string_literal) \ + /* END OF DEBUG FEATURES */ +#define AMBIGUOUS_FEATURE_NAMES 0 /* no ambiguous feature names now */ +#define DEFINE_FEATURE(bit) feature_##bit +#define DEFINE_DEBUG_FEATURE(bit) feature_debug_##bit +enum feature_flag_bits { + EACH_FEATURES(DEFINE_FEATURE, COMMA), + feature_debug_flag_first, + feature_debug_flag_begin = feature_debug_flag_first - 1, + EACH_DEBUG_FEATURES(DEFINE_DEBUG_FEATURE, COMMA), + feature_flag_count }; +#define DEBUG_BIT(bit) (1U << feature_debug_##bit) + #define DUMP_BIT(bit) (1U << dump_##bit) +#define DEFINE_DUMP(bit) dump_##bit +#define EACH_DUMPS(X, SEP) \ + X(version) \ + SEP \ + X(copyright) \ + SEP \ + X(usage) \ + SEP \ + X(help) \ + SEP \ + X(yydebug) \ + SEP \ + X(syntax) \ + SEP \ + X(parsetree) \ + SEP \ + X(parsetree_with_comment) \ + SEP \ + X(insns) \ + /* END OF DUMPS */ enum dump_flag_bits { - dump_version, - dump_copyright, - dump_usage, - dump_yydebug, - dump_syntax, - dump_parsetree, - dump_parsetree_with_comment, - dump_insns, - dump_flag_count + dump_version_v, + EACH_DUMPS(DEFINE_DUMP, COMMA), + dump_exit_bits = (DUMP_BIT(yydebug) | DUMP_BIT(syntax) | + DUMP_BIT(parsetree) | DUMP_BIT(parsetree_with_comment) | + DUMP_BIT(insns)) }; -struct cmdline_options { - int sflag, xflag; - int do_loop, do_print; - int do_line, do_split; - int do_search; - unsigned int disable; - int verbose; - int safe_level; - unsigned int setids; - unsigned int dump; +typedef struct ruby_cmdline_options ruby_cmdline_options_t; + +struct ruby_cmdline_options { const char *script; VALUE script_name; VALUE e_script; @@ -97,119 +140,180 @@ struct cmdline_options { } enc; } src, ext, intern; VALUE req_list; + unsigned int features; + unsigned int dump; +#if USE_MJIT + struct mjit_options mjit; +#endif + int safe_level; + int sflag, xflag; + unsigned int warning: 1; + unsigned int verbose: 1; + unsigned int do_loop: 1; + unsigned int do_print: 1; + unsigned int do_line: 1; + unsigned int do_split: 1; + unsigned int do_search: 1; + unsigned int setids: 2; }; -static void init_ids(struct cmdline_options *); +static void init_ids(ruby_cmdline_options_t *); #define src_encoding_index GET_VM()->src_encoding_index -static struct cmdline_options * -cmdline_options_init(struct cmdline_options *opt) +enum { + COMPILATION_FEATURES = ( + 0 + | FEATURE_BIT(frozen_string_literal) + | FEATURE_BIT(debug_frozen_string_literal) + ), + DEFAULT_FEATURES = ( + (FEATURE_BIT(debug_flag_first)-1) +#if DISABLE_RUBYGEMS + & ~FEATURE_BIT(gems) +#endif + & ~FEATURE_BIT(frozen_string_literal) + & ~FEATURE_BIT(jit) + ) +}; + +static ruby_cmdline_options_t * +cmdline_options_init(ruby_cmdline_options_t *opt) { MEMZERO(opt, *opt, 1); init_ids(opt); opt->src.enc.index = src_encoding_index; opt->ext.enc.index = -1; opt->intern.enc.index = -1; + opt->features = DEFAULT_FEATURES; +#ifdef MJIT_FORCE_ENABLE /* to use with: ./configure cppflags="-DMJIT_FORCE_ENABLE" */ + opt->features |= FEATURE_BIT(jit); +#endif return opt; } -static NODE *load_file(VALUE, const char *, int, struct cmdline_options *); -static void forbid_setid(const char *, struct cmdline_options *); -#define forbid_setid(s) forbid_setid(s, opt) +static rb_ast_t *load_file(VALUE parser, VALUE fname, VALUE f, int script, + ruby_cmdline_options_t *opt); +static VALUE open_load_file(VALUE fname_v, int *xflag); +static void forbid_setid(const char *, const ruby_cmdline_options_t *); +#define forbid_setid(s) forbid_setid((s), opt) static struct { int argc; char **argv; -#if !defined(PSTAT_SETCMD) && !defined(HAVE_SETPROCTITLE) - size_t len; -#endif } origarg; static void -usage(const char *name) +show_usage_line(const char *str, unsigned int namelen, unsigned int secondlen, int help) +{ + const unsigned int w = 16; + const int wrap = help && namelen + secondlen - 1 > w; + printf(" %.*s%-*.*s%-*s%s\n", namelen-1, str, + (wrap ? 0 : w - namelen + 1), + (help ? secondlen-1 : 0), str + namelen, + (wrap ? w + 3 : 0), (wrap ? "\n" : ""), + str + namelen + secondlen); +} + +static void +usage(const char *name, int help) { /* This message really ought to be max 23 lines. * Removed -h because the user already knows that option. Others? */ - static const char *const usage_msg[] = { - "-0[octal] specify record separator (\\0, if no argument)", - "-a autosplit mode with -n or -p (splits $_ into $F)", - "-c check syntax only", - "-Cdirectory cd to directory, before executing your script", - "-d set debugging flags (set $DEBUG to true)", - "-e 'command' one line of script. Several -e's allowed. Omit [programfile]", - "-Eex[:in] specify the default external and internal character encodings", - "-Fpattern split() pattern for autosplit (-a)", - "-i[extension] edit ARGV files in place (make backup if extension supplied)", - "-Idirectory specify $LOAD_PATH directory (may be used more than once)", - "-l enable line ending processing", - "-n assume 'while gets(); ... end' loop around your script", - "-p assume loop like -n but print line also like sed", - "-rlibrary require the library, before executing your script", - "-s enable some switch parsing for switches after script name", - "-S look for the script using PATH environment variable", - "-T[level=1] turn on tainting checks", - "-v print version number, then turn on verbose mode", - "-w turn warnings on for your script", - "-W[level=2] set warning level; 0=silence, 1=medium, 2=verbose", - "-x[directory] strip off text before #!ruby line and perhaps cd to directory", - "--copyright print the copyright", - "--version print the version", - NULL + struct message { + const char *str; + unsigned short namelen, secondlen; + }; +#define M(shortopt, longopt, desc) { \ + shortopt " " longopt " " desc, \ + (unsigned short)sizeof(shortopt), \ + (unsigned short)sizeof(longopt), \ +} + static const struct message usage_msg[] = { + M("-0[octal]", "", "specify record separator (\\0, if no argument)"), + M("-a", "", "autosplit mode with -n or -p (splits $_ into $F)"), + M("-c", "", "check syntax only"), + M("-Cdirectory", "", "cd to directory before executing your script"), + M("-d", ", --debug", "set debugging flags (set $DEBUG to true)"), + M("-e 'command'", "", "one line of script. Several -e's allowed. Omit [programfile]"), + M("-Eex[:in]", ", --encoding=ex[:in]", "specify the default external and internal character encodings"), + M("-Fpattern", "", "split() pattern for autosplit (-a)"), + M("-i[extension]", "", "edit ARGV files in place (make backup if extension supplied)"), + M("-Idirectory", "", "specify $LOAD_PATH directory (may be used more than once)"), + M("-l", "", "enable line ending processing"), + M("-n", "", "assume 'while gets(); ... end' loop around your script"), + M("-p", "", "assume loop like -n but print line also like sed"), + M("-rlibrary", "", "require the library before executing your script"), + M("-s", "", "enable some switch parsing for switches after script name"), + M("-S", "", "look for the script using PATH environment variable"), + M("-T[level=1]", "", "turn on tainting checks"), + M("-v", "", "print the version number, then turn on verbose mode"), + M("-w", "", "turn warnings on for your script"), + M("-W[level=2]", "", "set warning level; 0=silence, 1=medium, 2=verbose"), + M("-x[directory]", "", "strip off text before #!ruby line and perhaps cd to directory"), + M("--jit", "", "enable JIT with default options (experimental)"), + M("--jit-[option]","", "enable JIT with an option (experimental)"), + M("-h", "", "show this message, --help for more info"), + }; + static const struct message help_msg[] = { + M("--copyright", "", "print the copyright"), + M("--dump={insns|parsetree|...}[,...]", "", + "dump debug information. see below for available dump list"), + M("--enable={gems|rubyopt|...}[,...]", ", --disable={gems|rubyopt|...}[,...]", + "enable or disable features. see below for available features"), + M("--external-encoding=encoding", ", --internal-encoding=encoding", + "specify the default external or internal character encoding"), + M("--verbose", "", "turn on verbose mode and disable script from stdin"), + M("--version", "", "print the version number, then exit"), + M("--help", "", "show this message, -h for short message"), + }; + static const struct message dumps[] = { + M("insns", "", "instruction sequences"), + M("yydebug", "", "yydebug of yacc parser generator"), + M("parsetree", "", "AST"), + M("parsetree_with_comment", "", "AST with comments"), + }; + static const struct message features[] = { + M("gems", "", "rubygems (default: "DEFAULT_RUBYGEMS_ENABLED")"), + M("did_you_mean", "", "did_you_mean (default: "DEFAULT_RUBYGEMS_ENABLED")"), + M("rubyopt", "", "RUBYOPT environment variable (default: enabled)"), + M("frozen-string-literal", "", "freeze all string literals (default: disabled)"), + M("jit", "", "JIT compiler (default: disabled)"), }; - const char *const *p = usage_msg; + static const struct message mjit_options[] = { + M("--jit-warnings", "", "Enable printing JIT warnings"), + M("--jit-debug", "", "Enable JIT debugging (very slow)"), + M("--jit-wait", "", "Wait until JIT compilation is finished everytime (for testing)"), + M("--jit-save-temps", "", "Save JIT temporary files in $TMP or /tmp (for testing)"), + M("--jit-verbose=num", "", "Print JIT logs of level num or less to stderr (default: 0)"), + M("--jit-max-cache=num", "", "Max number of methods to be JIT-ed in a cache (default: 1000)"), + M("--jit-min-calls=num", "", "Number of calls to trigger JIT (for testing, default: 5)"), + }; + int i; + const int num = numberof(usage_msg) - (help ? 1 : 0); +#define SHOW(m) show_usage_line((m).str, (m).namelen, (m).secondlen, help) printf("Usage: %s [switches] [--] [programfile] [arguments]\n", name); - while (*p) - printf(" %s\n", *p++); + for (i = 0; i < num; ++i) + SHOW(usage_msg[i]); + + if (!help) return; + + for (i = 0; i < numberof(help_msg); ++i) + SHOW(help_msg[i]); + puts("Dump List:"); + for (i = 0; i < numberof(dumps); ++i) + SHOW(dumps[i]); + puts("Features:"); + for (i = 0; i < numberof(features); ++i) + SHOW(features[i]); + puts("JIT options (experimental):"); + for (i = 0; i < numberof(mjit_options); ++i) + SHOW(mjit_options[i]); } -VALUE rb_get_load_path(void); - -#ifdef MANGLED_PATH -static VALUE -rubylib_mangled_path(const char *s, unsigned int l) -{ - static char *newp, *oldp; - static int newl, oldl, notfound; - char *ptr; - VALUE ret; - - if (!newp && !notfound) { - newp = getenv("RUBYLIB_PREFIX"); - if (newp) { - oldp = newp = strdup(newp); - while (*newp && !ISSPACE(*newp) && *newp != ';') { - newp = CharNext(newp); /* Skip digits. */ - } - oldl = newp - oldp; - while (*newp && (ISSPACE(*newp) || *newp == ';')) { - newp = CharNext(newp); /* Skip whitespace. */ - } - newl = strlen(newp); - if (newl == 0 || oldl == 0) { - rb_fatal("malformed RUBYLIB_PREFIX"); - } - translit_char(newp, '\\', '/'); - } - else { - notfound = 1; - } - } - if (!newp || l < oldl || STRNCASECMP(oldp, s, oldl) != 0) { - return rb_str_new(s, l); - } - ret = rb_str_new(0, l + newl - oldl); - ptr = RSTRING_PTR(ret); - memcpy(ptr, newp, newl); - memcpy(ptr + newl, s + oldl, l - oldl); - ptr[l + newl - oldl] = 0; - return ret; -} -#else -#define rubylib_mangled_path rb_str_new -#endif +#define rubylib_path_new rb_str_new static void push_include(const char *path, VALUE (*filter)(VALUE)) @@ -224,7 +328,7 @@ push_include(const char *path, VALUE (*filter)(VALUE)) p++; if (!*p) break; for (s = p; *s && *s != sep; s = CharNext(s)); - rb_ary_push(load_path, (*filter)(rubylib_mangled_path(p, s - p))); + rb_ary_push(load_path, (*filter)(rubylib_path_new(p, s - p))); p = s; } } @@ -255,7 +359,13 @@ push_include_cygwin(const char *path, VALUE (*filter)(VALUE)) p = strncpy(RSTRING_PTR(buf), p, len); } } - if (cygwin_conv_to_posix_path(p, rubylib) == 0) +#ifdef HAVE_CYGWIN_CONV_PATH +#define CONV_TO_POSIX_PATH(p, lib) \ + cygwin_conv_path(CCP_WIN_A_TO_POSIX|CCP_RELATIVE, (p), (lib), sizeof(lib)) +#else +# error no cygwin_conv_path +#endif + if (CONV_TO_POSIX_PATH(p, rubylib) == 0) p = rubylib; push_include(p, filter); if (!*s) break; @@ -309,6 +419,7 @@ ruby_incpush_expand(const char *path) ruby_push_include(path, expand_include_path); } +#undef UTF8_PATH #if defined _WIN32 || defined __CYGWIN__ static HMODULE libruby; @@ -325,6 +436,41 @@ rb_libruby_handle(void) { return libruby; } + +static inline void +translit_char_bin(char *p, int from, int to) +{ + while (*p) { + if ((unsigned char)*p == from) + *p = to; + p++; + } +} +#endif + +#ifdef _WIN32 +# define UTF8_PATH 1 +#endif + +#ifndef UTF8_PATH +# define UTF8_PATH 0 +#endif +#if UTF8_PATH +# define IF_UTF8_PATH(t, f) t +#else +# define IF_UTF8_PATH(t, f) f +#endif + +#if UTF8_PATH +static VALUE +str_conv_enc(VALUE str, rb_encoding *from, rb_encoding *to) +{ + return rb_str_conv_enc_opts(str, from, to, + ECONV_UNDEF_REPLACE|ECONV_INVALID_REPLACE, + Qnil); +} +#else +# define str_conv_enc(str, from, to) (str) #endif void ruby_init_loadpath_safe(int safe_level); @@ -335,115 +481,172 @@ ruby_init_loadpath(void) ruby_init_loadpath_safe(0); } +#if defined(LOAD_RELATIVE) +static VALUE +runtime_libruby_path(void) +{ +#if defined _WIN32 || defined __CYGWIN__ + DWORD len = RSTRING_EMBED_LEN_MAX, ret; + VALUE path; + VALUE wsopath = rb_str_new(0, len*sizeof(WCHAR)); + WCHAR *wlibpath; + char *libpath; + + while (wlibpath = (WCHAR *)RSTRING_PTR(wsopath), + ret = GetModuleFileNameW(libruby, wlibpath, len), + (ret == len)) + { + rb_str_modify_expand(wsopath, len*sizeof(WCHAR)); + rb_str_set_len(wsopath, (len += len)*sizeof(WCHAR)); + } + if (!ret || ret > len) rb_fatal("failed to get module file name"); +#if defined __CYGWIN__ + { + const int win_to_posix = CCP_WIN_W_TO_POSIX | CCP_RELATIVE; + size_t newsize = cygwin_conv_path(win_to_posix, wlibpath, 0, 0); + if (!newsize) rb_fatal("failed to convert module path to cygwin"); + path = rb_str_new(0, newsize); + libpath = RSTRING_PTR(path); + if (cygwin_conv_path(win_to_posix, wlibpath, libpath, newsize)) { + rb_str_resize(path, 0); + } + } +#else + { + DWORD i; + for (len = ret, i = 0; i < len; ++i) { + if (wlibpath[i] == L'\\') { + wlibpath[i] = L'/'; + ret = i+1; /* chop after the last separator */ + } + } + } + len = WideCharToMultiByte(CP_UTF8, 0, wlibpath, ret, NULL, 0, NULL, NULL); + path = rb_utf8_str_new(0, len); + libpath = RSTRING_PTR(path); + WideCharToMultiByte(CP_UTF8, 0, wlibpath, ret, libpath, len, NULL, NULL); +#endif + rb_str_resize(wsopath, 0); + return path; +#elif defined(HAVE_DLADDR) + Dl_info dli; + VALUE fname, path; + const void* addr = (void *)(VALUE)expand_include_path; + + if (!dladdr((void *)addr, &dli)) { + return rb_str_new(0, 0); + } +#ifdef __linux__ + else if (origarg.argc > 0 && origarg.argv && dli.dli_fname == origarg.argv[0]) { + fname = rb_str_new_cstr("/proc/self/exe"); + path = rb_readlink(fname, NULL); + } +#endif + else { + fname = rb_str_new_cstr(dli.dli_fname); + path = rb_realpath_internal(Qnil, fname, 1); + } + rb_str_resize(fname, 0); + return path; +#else +# error relative load path is not supported on this platform. +#endif +} +#endif + +#define INITIAL_LOAD_PATH_MARK rb_intern_const("@gem_prelude_index") + +VALUE ruby_archlibdir_path, ruby_prefix_path; + void ruby_init_loadpath_safe(int safe_level) { - VALUE load_path; + VALUE load_path, archlibdir = 0; ID id_initial_load_path_mark; - extern const char ruby_initial_load_paths[]; const char *paths = ruby_initial_load_paths; #if defined LOAD_RELATIVE -# if defined HAVE_DLADDR || (defined __CYGWIN__ && defined CCP_WIN_A_TO_POSIX) -# define VARIABLE_LIBPATH 1 -# else -# define VARIABLE_LIBPATH 0 -# endif -# if VARIABLE_LIBPATH +#if !defined ENABLE_MULTIARCH +# define RUBY_ARCH_PATH "" +#elif defined RUBY_ARCH +# define RUBY_ARCH_PATH "/"RUBY_ARCH +#else +# define RUBY_ARCH_PATH "/"RUBY_PLATFORM +#endif char *libpath; VALUE sopath; -# else - char libpath[MAXPATHLEN + 1]; -# endif size_t baselen; - char *p; + const char *p; -#if defined _WIN32 || defined __CYGWIN__ -# if VARIABLE_LIBPATH - sopath = rb_str_new(0, MAXPATHLEN); + sopath = runtime_libruby_path(); libpath = RSTRING_PTR(sopath); - GetModuleFileName(libruby, libpath, MAXPATHLEN); -# else - GetModuleFileName(libruby, libpath, sizeof libpath); -# endif -#elif defined(__EMX__) - _execname(libpath, sizeof(libpath) - 1); -#elif defined(HAVE_DLADDR) - Dl_info dli; - if (dladdr((void *)(VALUE)expand_include_path, &dli)) { - VALUE rb_realpath_internal(VALUE basedir, VALUE path, int strict); - char fbuf[MAXPATHLEN]; - char *f = dln_find_file_r(dli.dli_fname, getenv(PATH_ENV), fbuf, sizeof(fbuf)); - VALUE fname = rb_str_new_cstr(f); - rb_str_freeze(fname); - sopath = rb_realpath_internal(Qnil, fname, 1); - } - else { - sopath = rb_str_new(0, 0); - } - libpath = RSTRING_PTR(sopath); -#endif -#if !VARIABLE_LIBPATH - libpath[sizeof(libpath) - 1] = '\0'; -#endif -#if defined DOSISH - translit_char(libpath, '\\', '/'); -#elif defined __CYGWIN__ - { -# if VARIABLE_LIBPATH - const int win_to_posix = CCP_WIN_A_TO_POSIX | CCP_RELATIVE; - size_t newsize = cygwin_conv_path(win_to_posix, libpath, 0, 0); - if (newsize > 0) { - VALUE rubylib = rb_str_new(0, newsize); - p = RSTRING_PTR(rubylib); - if (cygwin_conv_path(win_to_posix, libpath, p, newsize) == 0) { - rb_str_resize(sopath, 0); - sopath = rubylib; - libpath = p; - } - } -# else - char rubylib[FILENAME_MAX]; - cygwin_conv_to_posix_path(libpath, rubylib); - strncpy(libpath, rubylib, sizeof(libpath)); -# endif - } -#endif p = strrchr(libpath, '/'); if (p) { - *p = 0; - if (p - libpath > 3 && !(STRCASECMP(p - 4, "/bin") && strcmp(p - 4, "/lib"))) { - p -= 4; - *p = 0; - } - } -#if !VARIABLE_LIBPATH - else { - strlcpy(libpath, ".", sizeof(libpath)); - p = libpath + 1; - } -#define PREFIX_PATH() rb_str_new(libpath, baselen) + static const char libdir[] = "/" +#ifdef LIBDIR_BASENAME + LIBDIR_BASENAME #else - rb_str_set_len(sopath, p - libpath); -#define PREFIX_PATH() sopath + "lib" #endif + RUBY_ARCH_PATH; + const ptrdiff_t libdir_len = (ptrdiff_t)sizeof(libdir) + - rb_strlen_lit(RUBY_ARCH_PATH) - 1; + static const char bindir[] = "/bin"; + const ptrdiff_t bindir_len = (ptrdiff_t)sizeof(bindir) - 1; + + const char *p2 = NULL; +#ifdef ENABLE_MULTIARCH + multiarch: +#endif + if (p - libpath >= bindir_len && !STRNCASECMP(p - bindir_len, bindir, bindir_len)) { + p -= bindir_len; + archlibdir = rb_str_subseq(sopath, 0, p - libpath); + rb_str_cat_cstr(archlibdir, libdir); + OBJ_FREEZE_RAW(archlibdir); + } + else if (p - libpath >= libdir_len && !strncmp(p - libdir_len, libdir, libdir_len)) { + archlibdir = rb_str_subseq(sopath, 0, (p2 ? p2 : p) - libpath); + OBJ_FREEZE_RAW(archlibdir); + p -= libdir_len; + } +#ifdef ENABLE_MULTIARCH + else if (p2) { + p = p2; + } + else { + p2 = p; + p = rb_enc_path_last_separator(libpath, p, rb_ascii8bit_encoding()); + if (p) goto multiarch; + p = p2; + } +#endif + } baselen = p - libpath; + rb_str_resize(sopath, baselen); + libpath = RSTRING_PTR(sopath); +#define PREFIX_PATH() sopath #define BASEPATH() rb_str_buf_cat(rb_str_buf_new(baselen+len), libpath, baselen) - -#define RUBY_RELATIVE(path, len) rb_str_buf_cat(BASEPATH(), path, len) +#define RUBY_RELATIVE(path, len) rb_str_buf_cat(BASEPATH(), (path), (len)) #else - static const char exec_prefix[] = RUBY_EXEC_PREFIX; -#define RUBY_RELATIVE(path, len) rubylib_mangled_path(path, len) -#define PREFIX_PATH() RUBY_RELATIVE(exec_prefix, sizeof(exec_prefix)-1) + const size_t exec_prefix_len = strlen(ruby_exec_prefix); +#define RUBY_RELATIVE(path, len) rubylib_path_new((path), (len)) +#define PREFIX_PATH() RUBY_RELATIVE(ruby_exec_prefix, exec_prefix_len) #endif + rb_gc_register_address(&ruby_prefix_path); + ruby_prefix_path = PREFIX_PATH(); + OBJ_FREEZE_RAW(ruby_prefix_path); + if (!archlibdir) archlibdir = ruby_prefix_path; + rb_gc_register_address(&ruby_archlibdir_path); + ruby_archlibdir_path = archlibdir; + load_path = GET_VM()->load_path; if (safe_level == 0) { ruby_push_include(getenv("RUBYLIB"), identical_path); } - id_initial_load_path_mark = rb_intern_const("@gem_prelude_index"); + id_initial_load_path_mark = INITIAL_LOAD_PATH_MARK; while (*paths) { size_t len = strlen(paths); VALUE path = RUBY_RELATIVE(paths, len); @@ -452,7 +655,7 @@ ruby_init_loadpath_safe(int safe_level) paths += len + 1; } - rb_const_set(rb_cObject, rb_intern_const("TMP_RUBY_PREFIX"), rb_obj_freeze(PREFIX_PATH())); + rb_const_set(rb_cObject, rb_intern_const("TMP_RUBY_PREFIX"), ruby_prefix_path); } @@ -460,38 +663,38 @@ static void add_modules(VALUE *req_list, const char *mod) { VALUE list = *req_list; + VALUE feature; if (!list) { - *req_list = list = rb_ary_new(); - RBASIC(list)->klass = 0; + *req_list = list = rb_ary_tmp_new(0); } - rb_ary_push(list, rb_obj_freeze(rb_str_new2(mod))); + feature = rb_str_cat_cstr(rb_str_tmp_new(0), mod); + rb_ary_push(list, feature); } -extern void Init_ext(void); -extern VALUE rb_vm_top_self(void); - static void require_libraries(VALUE *req_list) { VALUE list = *req_list; + VALUE self = rb_vm_top_self(); ID require; - rb_thread_t *th = GET_THREAD(); - rb_block_t *prev_base_block = th->base_block; - int prev_parse_in_eval = th->parse_in_eval; - th->base_block = 0; - th->parse_in_eval = 0; + rb_encoding *extenc = rb_default_external_encoding(); - Init_ext(); /* should be called here for some reason :-( */ CONST_ID(require, "require"); while (list && RARRAY_LEN(list) > 0) { VALUE feature = rb_ary_shift(list); - rb_funcall2(rb_vm_top_self(), require, 1, &feature); + rb_enc_associate(feature, extenc); + RBASIC_SET_CLASS_RAW(feature, rb_cString); + OBJ_FREEZE(feature); + rb_funcallv(self, require, 1, &feature); } *req_list = 0; +} - th->parse_in_eval = prev_parse_in_eval; - th->base_block = prev_base_block; +static const struct rb_block* +toplevel_context(rb_binding_t *bind) +{ + return &bind->block; } static void @@ -499,11 +702,11 @@ process_sflag(int *sflag) { if (*sflag > 0) { long n; - VALUE *args; + const VALUE *args; VALUE argv = rb_argv; n = RARRAY_LEN(argv); - args = RARRAY_PTR(argv); + args = RARRAY_CONST_PTR(argv); while (n > 0) { VALUE v = *args++; char *s = StringValuePtr(v); @@ -558,25 +761,24 @@ process_sflag(int *sflag) } } -NODE *rb_parser_append_print(VALUE, NODE *); -NODE *rb_parser_while_loop(VALUE, NODE *, int, int); -static long proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt); +static long proc_options(long argc, char **argv, ruby_cmdline_options_t *opt, int envopt); static void -moreswitches(const char *s, struct cmdline_options *opt, int envopt) +moreswitches(const char *s, ruby_cmdline_options_t *opt, int envopt) { long argc, i, len; char **argv, *p; const char *ap = 0; VALUE argstr, argary; + void *ptr; while (ISSPACE(*s)) s++; if (!*s) return; - argstr = rb_str_tmp_new((len = strlen(s)) + 2); + argstr = rb_str_tmp_new((len = strlen(s)) + (envopt!=0)); argary = rb_str_tmp_new(0); p = RSTRING_PTR(argstr); - *p++ = ' '; + if (envopt) *p++ = ' '; memcpy(p, s, len + 1); ap = 0; rb_str_cat(argary, (char *)&ap, sizeof(ap)); @@ -591,9 +793,10 @@ moreswitches(const char *s, struct cmdline_options *opt, int envopt) argc = RSTRING_LEN(argary) / sizeof(ap); ap = 0; rb_str_cat(argary, (char *)&ap, sizeof(ap)); - argv = (char **)RSTRING_PTR(argary); + argv = ptr = ALLOC_N(char *, argc); + MEMMOVE(argv, RSTRING_PTR(argary), char *, argc); - while ((i = proc_options(argc, argv, opt, envopt)) > 1 && (argc -= i) > 0) { + while ((i = proc_options(argc, argv, opt, envopt)) > 1 && envopt && (argc -= i) > 0) { argv += i; if (**argv != '-') { *--*argv = '-'; @@ -604,66 +807,124 @@ moreswitches(const char *s, struct cmdline_options *opt, int envopt) } } + ruby_xfree(ptr); /* get rid of GC */ rb_str_resize(argary, 0); rb_str_resize(argstr, 0); } +static int +name_match_p(const char *name, const char *str, size_t len) +{ + if (len == 0) return 0; + do { + while (TOLOWER(*str) == *name) { + if (!--len || !*++str) return 1; + ++name; + } + if (*str != '-' && *str != '_') return 0; + while (ISALNUM(*name)) name++; + if (*name != '-' && *name != '_') return 0; + ++name; + ++str; + } while (len > 0); + return !*name; +} + #define NAME_MATCH_P(name, str, len) \ - ((len) < (int)sizeof(name) && strncmp((str), name, (len)) == 0) + ((len) < (int)sizeof(name) && name_match_p((name), (str), (len))) #define UNSET_WHEN(name, bit, str, len) \ - if (NAME_MATCH_P(name, str, len)) { \ + if (NAME_MATCH_P((name), (str), (len))) { \ *(unsigned int *)arg &= ~(bit); \ return; \ } #define SET_WHEN(name, bit, str, len) \ - if (NAME_MATCH_P(name, str, len)) { \ + if (NAME_MATCH_P((name), (str), (len))) { \ *(unsigned int *)arg |= (bit); \ return; \ } +#define LITERAL_NAME_ELEMENT(name) #name + static void -enable_option(const char *str, int len, void *arg) +feature_option(const char *str, int len, void *arg, const unsigned int enable) { -#define UNSET_WHEN_DISABLE(bit) UNSET_WHEN(#bit, DISABLE_BIT(bit), str, len) - UNSET_WHEN_DISABLE(gems); - UNSET_WHEN_DISABLE(rubyopt); + static const char list[] = EACH_FEATURES(LITERAL_NAME_ELEMENT, ", "); + unsigned int *argp = arg; + unsigned int mask = ~0U; +#if AMBIGUOUS_FEATURE_NAMES + unsigned int set = 0U; + int matched = 0; +#define SET_FEATURE(bit) \ + if (NAME_MATCH_P(#bit, str, len)) {set |= mask = FEATURE_BIT(bit); ++matched;} +#else +#define SET_FEATURE(bit) \ + if (NAME_MATCH_P(#bit, str, len)) {mask = FEATURE_BIT(bit); goto found;} +#endif + EACH_FEATURES(SET_FEATURE, ;); if (NAME_MATCH_P("all", str, len)) { - *(unsigned int *)arg = 0U; + found: + *argp = (*argp & ~mask) | (mask & enable); return; } - rb_warn("unknown argument for --enable: `%.*s'", len, str); +#if AMBIGUOUS_FEATURE_NAMES + if (matched == 1) goto found; + if (matched > 1) { + VALUE mesg = rb_sprintf("ambiguous feature: `%.*s' (", len, str); +#define ADD_FEATURE_NAME(bit) \ + if (FEATURE_BIT(bit) & set) { \ + rb_str_cat_cstr(mesg, #bit); \ + if (--matched) rb_str_cat_cstr(mesg, ", "); \ + } + EACH_FEATURES(ADD_FEATURE_NAME, ;); + rb_str_cat_cstr(mesg, ")"); + rb_exc_raise(rb_exc_new_str(rb_eRuntimeError, mesg)); +#undef ADD_FEATURE_NAME + } +#endif + rb_warn("unknown argument for --%s: `%.*s'", + enable ? "enable" : "disable", len, str); + rb_warn("features are [%.*s].", (int)strlen(list), list); +} + +static void +enable_option(const char *str, int len, void *arg) +{ + feature_option(str, len, arg, ~0U); } static void disable_option(const char *str, int len, void *arg) { -#define SET_WHEN_DISABLE(bit) SET_WHEN(#bit, DISABLE_BIT(bit), str, len) - SET_WHEN_DISABLE(gems); - SET_WHEN_DISABLE(rubyopt); - if (NAME_MATCH_P("all", str, len)) { - *(unsigned int *)arg = ~0U; - return; - } - rb_warn("unknown argument for --disable: `%.*s'", len, str); + feature_option(str, len, arg, 0U); +} + +RUBY_EXTERN const int ruby_patchlevel; +int ruby_env_debug_option(const char *str, int len, void *arg); + +static void +debug_option(const char *str, int len, void *arg) +{ + static const char list[] = EACH_DEBUG_FEATURES(LITERAL_NAME_ELEMENT, ", "); +#define SET_WHEN_DEBUG(bit) SET_WHEN(#bit, DEBUG_BIT(bit), str, len) + EACH_DEBUG_FEATURES(SET_WHEN_DEBUG, ;); +#ifdef RUBY_DEVEL + if (ruby_patchlevel < 0 && ruby_env_debug_option(str, len, 0)) return; +#endif + rb_warn("unknown argument for --debug: `%.*s'", len, str); + rb_warn("debug features are [%.*s].", (int)strlen(list), list); } static void dump_option(const char *str, int len, void *arg) { + static const char list[] = EACH_DUMPS(LITERAL_NAME_ELEMENT, ", "); #define SET_WHEN_DUMP(bit) SET_WHEN(#bit, DUMP_BIT(bit), str, len) - SET_WHEN_DUMP(version); - SET_WHEN_DUMP(copyright); - SET_WHEN_DUMP(usage); - SET_WHEN_DUMP(yydebug); - SET_WHEN_DUMP(syntax); - SET_WHEN_DUMP(parsetree); - SET_WHEN_DUMP(parsetree_with_comment); - SET_WHEN_DUMP(insns); + EACH_DUMPS(SET_WHEN_DUMP, ;); rb_warn("don't know how to dump `%.*s',", len, str); - rb_warn("but only [version, copyright, usage, yydebug, syntax, parsetree, parsetree_with_comment, insns]."); + rb_warn("but only [%.*s].", (int)strlen(list), list); } static void @@ -677,30 +938,64 @@ set_option_encoding_once(const char *type, VALUE *name, const char *e, long elen if (*name && rb_funcall(ename, rb_intern("casecmp"), 1, *name) != INT2FIX(0)) { rb_raise(rb_eRuntimeError, - "%s already set to %s", type, RSTRING_PTR(*name)); + "%s already set to %"PRIsVALUE, type, *name); } *name = ename; } #define set_internal_encoding_once(opt, e, elen) \ - set_option_encoding_once("default_internal", &opt->intern.enc.name, e, elen) + set_option_encoding_once("default_internal", &(opt)->intern.enc.name, (e), (elen)) #define set_external_encoding_once(opt, e, elen) \ - set_option_encoding_once("default_external", &opt->ext.enc.name, e, elen) + set_option_encoding_once("default_external", &(opt)->ext.enc.name, (e), (elen)) #define set_source_encoding_once(opt, e, elen) \ - set_option_encoding_once("source", &opt->src.enc.name, e, elen) + set_option_encoding_once("source", &(opt)->src.enc.name, (e), (elen)) + +#if USE_MJIT +static void +setup_mjit_options(const char *s, struct mjit_options *mjit_opt) +{ + if (*s == 0) return; + else if (strcmp(s, "-warnings") == 0) { + mjit_opt->warnings = 1; + } + else if (strcmp(s, "-debug") == 0) { + mjit_opt->debug = 1; + } + else if (strcmp(s, "-wait") == 0) { + mjit_opt->wait = 1; + } + else if (strcmp(s, "-save-temps") == 0) { + mjit_opt->save_temps = 1; + } + else if (strncmp(s, "-verbose=", 9) == 0) { + mjit_opt->verbose = atoi(s + 9); + } + else if (strncmp(s, "-max-cache=", 11) == 0) { + mjit_opt->max_cache_size = atoi(s + 11); + } + else if (strncmp(s, "-min-calls=", 11) == 0) { + mjit_opt->min_calls = atoi(s + 11); + } + else { + rb_raise(rb_eRuntimeError, + "invalid MJIT option `%s' (--help will show valid MJIT options)", s + 1); + } +} +#endif static long -proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) +proc_options(long argc, char **argv, ruby_cmdline_options_t *opt, int envopt) { long n, argc0 = argc; const char *s; + int warning = opt->warning; - if (argc == 0) + if (argc <= 0 || !argv) return 0; for (argc--, argv++; argc > 0; argc--, argv++) { const char *const arg = argv[0]; - if (arg[0] != '-' || !arg[1]) + if (!arg || arg[0] != '-' || !arg[1]) break; s = arg + 1; @@ -739,10 +1034,13 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) s++; goto reswitch; } - ruby_show_version(); + opt->dump |= DUMP_BIT(version_v); opt->verbose = 1; case 'w': - ruby_verbose = Qtrue; + if (!opt->warning) { + warning = 1; + ruby_verbose = Qtrue; + } s++; goto reswitch; @@ -757,17 +1055,20 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) v = 1; s += numlen; } - switch (v) { - case 0: - ruby_verbose = Qnil; - break; - case 1: - ruby_verbose = Qfalse; - break; - default: - ruby_verbose = Qtrue; - break; + if (!opt->warning) { + switch (v) { + case 0: + ruby_verbose = Qnil; + break; + case 1: + ruby_verbose = Qfalse; + break; + default: + ruby_verbose = Qtrue; + break; + } } + warning = 1; } goto reswitch; @@ -807,11 +1108,9 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) if (envopt) goto noenvopt; forbid_setid("-e"); if (!*++s) { - s = argv[1]; - argc--, argv++; - } - if (!s) { - rb_raise(rb_eRuntimeError, "no code specified for -e"); + if (!--argc) + rb_raise(rb_eRuntimeError, "no code specified for -e"); + s = *++argv; } if (!opt->e_script) { opt->e_script = rb_str_new(0, 0); @@ -827,7 +1126,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) if (*++s) { add_modules(&opt->req_list, s); } - else if (argv[1]) { + else if (argc > 1) { add_modules(&opt->req_list, argv[1]); argc--, argv++; } @@ -841,6 +1140,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) case 'x': if (envopt) goto noenvopt; + forbid_setid("-x"); opt->xflag = TRUE; s++; if (*s && chdir(s) < 0) { @@ -851,12 +1151,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) case 'C': case 'X': if (envopt) goto noenvopt; - s++; - if (!*s) { - s = argv[1]; - argc--, argv++; - } - if (!s || !*s) { + if (!*++s && (!--argc || !(s = *++argv) || !*s)) { rb_fatal("Can't chdir"); } if (chdir(s) < 0) { @@ -927,7 +1222,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) forbid_setid("-I"); if (*++s) ruby_incpush_expand(s); - else if (argv[1]) { + else if (argc > 1) { ruby_incpush_expand(argv[1]); argc--, argv++; } @@ -945,7 +1240,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) if (v > 0377) rb_rs = Qnil; else if (v == 0 && numlen >= 2) { - rb_rs = rb_str_new2("\n\n"); + rb_rs = rb_str_new2(""); } else { c = v & 0xff; @@ -962,31 +1257,39 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) s++; # define is_option_end(c, allow_hyphen) \ - (!(c) || (allow_hyphen && (c) == '-') || (c) == '=') + (!(c) || ((allow_hyphen) && (c) == '-') || (c) == '=') # define check_envopt(name, allow_envopt) \ - ((allow_envopt || !envopt) ? (void)0 : \ + (((allow_envopt) || !envopt) ? (void)0 : \ rb_raise(rb_eRuntimeError, "invalid switch in RUBYOPT: --" name)) -# define need_argument(name, s) \ - ((*s++ ? !*s : (!--argc || !(s = *++argv))) ? \ +# define need_argument(name, s, needs_arg, next_arg) \ + ((*(s) ? !*++(s) : (next_arg) && (!argc || !((s) = argv[1]) || (--argc, ++argv, 0))) && (needs_arg) ? \ rb_raise(rb_eRuntimeError, "missing argument for --" name) \ : (void)0) -# define is_option_with_arg(name, allow_hyphen, allow_envopt) \ - (strncmp(name, s, n = sizeof(name) - 1) == 0 && is_option_end(s[n], allow_hyphen) ? \ - (check_envopt(name, allow_envopt), s += n, need_argument(name, s), 1) : 0) +# define is_option_with_arg(name, allow_hyphen, allow_envopt) \ + is_option_with_optarg(name, allow_hyphen, allow_envopt, Qtrue, Qtrue) +# define is_option_with_optarg(name, allow_hyphen, allow_envopt, needs_arg, next_arg) \ + (strncmp((name), s, n = sizeof(name) - 1) == 0 && is_option_end(s[n], (allow_hyphen)) ? \ + (check_envopt(name, (allow_envopt)), s += n, \ + need_argument(name, s, needs_arg, next_arg), 1) : 0) if (strcmp("copyright", s) == 0) { if (envopt) goto noenvopt_long; opt->dump |= DUMP_BIT(copyright); } - else if (strcmp("debug", s) == 0) { - ruby_debug = Qtrue; - ruby_verbose = Qtrue; + else if (is_option_with_optarg("debug", Qtrue, Qtrue, Qfalse, Qfalse)) { + if (s && *s) { + ruby_each_words(s, debug_option, &opt->features); + } + else { + ruby_debug = Qtrue; + ruby_verbose = Qtrue; + } } else if (is_option_with_arg("enable", Qtrue, Qtrue)) { - ruby_each_words(s, enable_option, &opt->disable); + ruby_each_words(s, enable_option, &opt->features); } else if (is_option_with_arg("disable", Qtrue, Qtrue)) { - ruby_each_words(s, disable_option, &opt->disable); + ruby_each_words(s, disable_option, &opt->features); } else if (is_option_with_arg("encoding", Qfalse, Qtrue)) { char *p; @@ -1004,7 +1307,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) if (!*(s = ++p)) break; set_encoding_part(internal); if (!*(s = ++p)) break; -#if ALLOW_DEFAULT_SOURCE_ENCODING +#if defined ALLOW_DEFAULT_SOURCE_ENCODING && ALLOW_DEFAULT_SOURCE_ENCODING set_encoding_part(source); if (!*(s = ++p)) break; #endif @@ -1019,7 +1322,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) else if (is_option_with_arg("external-encoding", Qfalse, Qtrue)) { set_external_encoding_once(opt, s, 0); } -#if ALLOW_DEFAULT_SOURCE_ENCODING +#if defined ALLOW_DEFAULT_SOURCE_ENCODING && ALLOW_DEFAULT_SOURCE_ENCODING else if (is_option_with_arg("source-encoding", Qfalse, Qtrue)) { set_source_encoding_once(opt, s, 0); } @@ -1032,6 +1335,14 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) opt->verbose = 1; ruby_verbose = Qtrue; } + else if (strncmp("jit", s, 3) == 0) { +#if USE_MJIT + opt->features |= FEATURE_BIT(jit); + setup_mjit_options(s + 3, &opt->mjit); +#else + rb_warn("MJIT support is disabled."); +#endif + } else if (strcmp("yydebug", s) == 0) { if (envopt) goto noenvopt_long; opt->dump |= DUMP_BIT(yydebug); @@ -1041,7 +1352,7 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) } else if (strcmp("help", s) == 0) { if (envopt) goto noenvopt_long; - opt->dump |= DUMP_BIT(usage); + opt->dump |= DUMP_BIT(help); goto switch_end; } else { @@ -1084,19 +1395,18 @@ proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) # undef check_envopt # undef need_argument # undef is_option_with_arg +# undef is_option_with_optarg } } switch_end: + if (warning) opt->warning = warning; return argc0 - argc; } -void Init_prelude(void); - static void -ruby_init_gems(int enable) +ruby_init_prelude(void) { - if (enable) rb_define_module("Gem"); Init_prelude(); rb_const_remove(rb_cObject, rb_intern_const("TMP_RUBY_PREFIX")); } @@ -1116,8 +1426,10 @@ opt_enc_index(VALUE enc_name) return i; } -#define rb_progname (GET_VM()->progname) +#define rb_progname (GET_VM()->progname) +#define rb_orig_progname (GET_VM()->orig_progname) VALUE rb_argv0; +VALUE rb_e_script; static VALUE false_value(void) @@ -1140,7 +1452,7 @@ uscore_get(void) VALUE line; line = rb_lastline_get(); - if (TYPE(line) != T_STRING) { + if (!RB_TYPE_P(line, T_STRING)) { rb_raise(rb_eTypeError, "$_ value need to be String (%s given)", NIL_P(line) ? "nil" : rb_obj_classname(line)); } @@ -1150,7 +1462,7 @@ uscore_get(void) /* * call-seq: * sub(pattern, replacement) -> $_ - * sub(pattern) { block } -> $_ + * sub(pattern) {|...| block } -> $_ * * Equivalent to <code>$_.sub(<i>args</i>)</code>, except that * <code>$_</code> will be updated if substitution occurs. @@ -1158,39 +1470,35 @@ uscore_get(void) */ static VALUE -rb_f_sub(argc, argv) - int argc; - VALUE *argv; +rb_f_sub(int argc, VALUE *argv) { - VALUE str = rb_funcall3(uscore_get(), rb_intern("sub"), argc, argv); + VALUE str = rb_funcall_passing_block(uscore_get(), rb_intern("sub"), argc, argv); rb_lastline_set(str); return str; } /* * call-seq: - * gsub(pattern, replacement) -> string - * gsub(pattern) {|...| block } -> string + * gsub(pattern, replacement) -> $_ + * gsub(pattern) {|...| block } -> $_ * * Equivalent to <code>$_.gsub...</code>, except that <code>$_</code> - * receives the modified result. + * will be updated if substitution occurs. * Available only when -p/-n command line option specified. * */ static VALUE -rb_f_gsub(argc, argv) - int argc; - VALUE *argv; +rb_f_gsub(int argc, VALUE *argv) { - VALUE str = rb_funcall3(uscore_get(), rb_intern("gsub"), argc, argv); + VALUE str = rb_funcall_passing_block(uscore_get(), rb_intern("gsub"), argc, argv); rb_lastline_set(str); return str; } /* * call-seq: - * chop -> string + * chop -> $_ * * Equivalent to <code>($_.dup).chop!</code>, except <code>nil</code> * is never returned. See <code>String#chop!</code>. @@ -1201,7 +1509,7 @@ rb_f_gsub(argc, argv) static VALUE rb_f_chop(void) { - VALUE str = rb_funcall3(uscore_get(), rb_intern("chop"), 0, 0); + VALUE str = rb_funcall_passing_block(uscore_get(), rb_intern("chop"), 0, 0); rb_lastline_set(str); return str; } @@ -1219,41 +1527,44 @@ rb_f_chop(void) */ static VALUE -rb_f_chomp(argc, argv) - int argc; - VALUE *argv; +rb_f_chomp(int argc, VALUE *argv) { - VALUE str = rb_funcall3(uscore_get(), rb_intern("chomp"), argc, argv); + VALUE str = rb_funcall_passing_block(uscore_get(), rb_intern("chomp"), argc, argv); rb_lastline_set(str); return str; } -void rb_stdio_set_default_encoding(void); -VALUE rb_parser_dump_tree(NODE *node, int comment); -VALUE rb_realpath_internal(VALUE basedir, VALUE path, int strict); - static VALUE -process_options(int argc, char **argv, struct cmdline_options *opt) +process_options(int argc, char **argv, ruby_cmdline_options_t *opt) { - NODE *tree = 0; + rb_ast_t *ast = 0; VALUE parser; - VALUE iseq; + VALUE script_name; + const rb_iseq_t *iseq; rb_encoding *enc, *lenc; +#if UTF8_PATH + rb_encoding *uenc, *ienc = 0; +#endif const char *s; char fbuf[MAXPATHLEN]; int i = (int)proc_options(argc, argv, opt, 0); - rb_thread_t *th = GET_THREAD(); - rb_env_t *env = 0; + rb_binding_t *toplevel_binding; + const struct rb_block *base_block; + unsigned int dump = opt->dump & dump_exit_bits; + + if (opt->dump & (DUMP_BIT(usage)|DUMP_BIT(help))) { + const char *const progname = + (argc > 0 && argv && argv[0] ? argv[0] : + origarg.argc > 0 && origarg.argv && origarg.argv[0] ? origarg.argv[0] : + ruby_engine); + usage(progname, (opt->dump & DUMP_BIT(help))); + return Qtrue; + } argc -= i; argv += i; - if (opt->dump & DUMP_BIT(usage)) { - usage(origarg.argv[0]); - return Qtrue; - } - - if (!(opt->disable & DISABLE_BIT(rubyopt)) && + if ((opt->features & FEATURE_BIT(rubyopt)) && opt->safe_level == 0 && (s = getenv("RUBYOPT"))) { VALUE src_enc_name = opt->src.enc.name; VALUE ext_enc_name = opt->ext.enc.name; @@ -1269,28 +1580,35 @@ process_options(int argc, char **argv, struct cmdline_options *opt) opt->intern.enc.name = int_enc_name; } - if (opt->dump & DUMP_BIT(version)) { + if (opt->src.enc.name) + rb_warning("-K is specified; it is for 1.8 compatibility and may cause odd behavior"); + +#if USE_MJIT + if (opt->features & FEATURE_BIT(jit)) { + opt->mjit.on = TRUE; /* set mjit.on for ruby_show_version() API and check to call mjit_init() */ + } +#endif + if (opt->dump & (DUMP_BIT(version) | DUMP_BIT(version_v))) { +#if USE_MJIT + mjit_opts.on = opt->mjit.on; /* used by ruby_show_version(). mjit_init() still can't be called here. */ +#endif ruby_show_version(); - return Qtrue; + if (opt->dump & DUMP_BIT(version)) return Qtrue; } if (opt->dump & DUMP_BIT(copyright)) { ruby_show_copyright(); - } - - if (opt->safe_level >= 4) { - OBJ_TAINT(rb_argv); - OBJ_TAINT(GET_VM()->load_path); + return Qtrue; } if (!opt->e_script) { - if (argc == 0) { /* no more args */ + if (argc <= 0) { /* no more args */ if (opt->verbose) return Qtrue; opt->script = "-"; } else { opt->script = argv[0]; - if (opt->script[0] == '\0') { + if (!opt->script || opt->script[0] == '\0') { opt->script = "-"; } else if (opt->do_search) { @@ -1309,19 +1627,34 @@ process_options(int argc, char **argv, struct cmdline_options *opt) argc--; argv++; } + if (opt->script[0] == '-' && !opt->script[1]) { + forbid_setid("program input from stdin"); + } } opt->script_name = rb_str_new_cstr(opt->script); opt->script = RSTRING_PTR(opt->script_name); -#if defined DOSISH || defined __CYGWIN__ + +#if _WIN32 + translit_char_bin(RSTRING_PTR(opt->script_name), '\\', '/'); +#elif defined DOSISH translit_char(RSTRING_PTR(opt->script_name), '\\', '/'); #endif - rb_obj_freeze(opt->script_name); + ruby_gc_set_params(opt->safe_level); ruby_init_loadpath_safe(opt->safe_level); - rb_enc_find_index("encdb"); + +#if USE_MJIT + if (opt->mjit.on) + /* Using TMP_RUBY_PREFIX created by ruby_init_loadpath_safe(). */ + mjit_init(&opt->mjit); +#endif + + Init_ruby_description(); + Init_enc(); lenc = rb_locale_encoding(); rb_enc_associate(rb_progname, lenc); + rb_obj_freeze(rb_progname); parser = rb_parser_new(); if (opt->dump & DUMP_BIT(yydebug)) { rb_parser_set_yydebug(parser, Qtrue); @@ -1347,35 +1680,77 @@ process_options(int argc, char **argv, struct cmdline_options *opt) enc = rb_enc_from_index(opt->intern.enc.index); rb_enc_set_default_internal(rb_enc_from_encoding(enc)); opt->intern.enc.index = -1; +#if UTF8_PATH + ienc = enc; +#endif } - rb_enc_associate(opt->script_name, lenc); - { + script_name = opt->script_name; + rb_enc_associate(opt->script_name, + IF_UTF8_PATH(uenc = rb_utf8_encoding(), lenc)); +#if UTF8_PATH + if (uenc != lenc) { + opt->script_name = str_conv_enc(opt->script_name, uenc, lenc); + opt->script = RSTRING_PTR(opt->script_name); + } +#endif + rb_obj_freeze(opt->script_name); + if (IF_UTF8_PATH(uenc != lenc, 1)) { long i; - VALUE load_path = GET_VM()->load_path; + rb_vm_t *vm = GET_VM(); + VALUE load_path = vm->load_path; + const ID id_initial_load_path_mark = INITIAL_LOAD_PATH_MARK; + int modifiable = FALSE; + + rb_get_expanded_load_path(); for (i = 0; i < RARRAY_LEN(load_path); ++i) { - rb_enc_associate(RARRAY_PTR(load_path)[i], lenc); + VALUE path = RARRAY_AREF(load_path, i); + int mark = rb_attr_get(path, id_initial_load_path_mark) == path; +#if UTF8_PATH + VALUE newpath = rb_str_conv_enc(path, uenc, lenc); + if (newpath == path) continue; + path = newpath; +#else + path = rb_enc_associate(rb_str_dup(path), lenc); +#endif + if (mark) rb_ivar_set(path, id_initial_load_path_mark, path); + if (!modifiable) { + rb_ary_modify(load_path); + modifiable = TRUE; + } + RARRAY_ASET(load_path, i, path); + } + if (modifiable) { + rb_ary_replace(vm->load_path_snapshot, load_path); + } + } + Init_ext(); /* load statically linked extensions before rubygems */ + if (opt->features & FEATURE_BIT(gems)) { + rb_define_module("Gem"); + if (opt->features & FEATURE_BIT(did_you_mean)) { + rb_define_module("DidYouMean"); } } - ruby_init_gems(!(opt->disable & DISABLE_BIT(gems))); + ruby_init_prelude(); + if ((opt->features ^ DEFAULT_FEATURES) & COMPILATION_FEATURES) { + VALUE option = rb_hash_new(); +#define SET_COMPILE_OPTION(h, o, name) \ + rb_hash_aset((h), ID2SYM(rb_intern_const(#name)), \ + ((o)->features & FEATURE_BIT(name) ? Qtrue : Qfalse)); + SET_COMPILE_OPTION(option, opt, frozen_string_literal); + SET_COMPILE_OPTION(option, opt, debug_frozen_string_literal); + rb_funcallv(rb_cISeq, rb_intern_const("compile_option="), 1, &option); +#undef SET_COMPILE_OPTION + } ruby_set_argv(argc, argv); process_sflag(&opt->sflag); - { - /* set eval context */ - VALUE toplevel_binding = rb_const_get(rb_cObject, rb_intern("TOPLEVEL_BINDING")); - rb_binding_t *bind; + GetBindingPtr(rb_const_get(rb_cObject, rb_intern("TOPLEVEL_BINDING")), + toplevel_binding); + /* need to acquire env from toplevel_binding each time, since it + * may update after eval() */ - GetBindingPtr(toplevel_binding, bind); - GetEnvPtr(bind->env, env); - } - -#define PREPARE_PARSE_MAIN(expr) do { \ - th->parse_in_eval--; \ - th->base_block = &env->block; \ - expr; \ - th->parse_in_eval++; \ - th->base_block = 0; \ -} while (0) + base_block = toplevel_context(toplevel_binding); + rb_parser_set_context(parser, base_block, TRUE); if (opt->e_script) { VALUE progname = rb_progname; @@ -1385,28 +1760,35 @@ process_options(int argc, char **argv, struct cmdline_options *opt) } else { eenc = lenc; +#if UTF8_PATH + if (ienc) eenc = ienc; +#endif } +#if UTF8_PATH + if (eenc != uenc) { + opt->e_script = str_conv_enc(opt->e_script, uenc, eenc); + } +#endif rb_enc_associate(opt->e_script, eenc); - rb_vm_set_progname(rb_progname = opt->script_name); - require_libraries(&opt->req_list); - rb_vm_set_progname(rb_progname = progname); - - PREPARE_PARSE_MAIN({ - tree = rb_parser_compile_string(parser, opt->script, opt->e_script, 1); - }); + if (!(opt->dump & ~DUMP_BIT(version_v))) { + ruby_set_script_name(opt->script_name); + require_libraries(&opt->req_list); + } + ruby_set_script_name(progname); + rb_parser_set_options(parser, opt->do_print, opt->do_loop, + opt->do_line, opt->do_split); + ast = rb_parser_compile_string(parser, opt->script, opt->e_script, 1); } else { - if (opt->script[0] == '-' && !opt->script[1]) { - forbid_setid("program input from stdin"); - } - - PREPARE_PARSE_MAIN({ - tree = load_file(parser, opt->script, 1, opt); - }); + VALUE f; + f = open_load_file(script_name, &opt->xflag); + ast = load_file(parser, opt->script_name, f, 1, opt); + } + ruby_set_script_name(opt->script_name); + if (dump & DUMP_BIT(yydebug)) { + dump &= ~DUMP_BIT(yydebug); + if (!dump) return Qtrue; } - rb_progname = opt->script_name; - rb_vm_set_progname(rb_progname); - if (opt->dump & DUMP_BIT(yydebug)) return Qtrue; if (opt->ext.enc.index >= 0) { enc = rb_enc_from_index(opt->ext.enc.index); @@ -1425,112 +1807,128 @@ process_options(int argc, char **argv, struct cmdline_options *opt) rb_enc_set_default_internal(Qnil); rb_stdio_set_default_encoding(); - if (!tree) return Qfalse; + if (!ast->body.root) { + rb_ast_dispose(ast); + return Qfalse; + } process_sflag(&opt->sflag); opt->xflag = 0; - if (opt->safe_level >= 4) { - FL_UNSET(rb_argv, FL_TAINT); - FL_UNSET(GET_VM()->load_path, FL_TAINT); - } - - if (opt->dump & DUMP_BIT(syntax)) { + if (dump & DUMP_BIT(syntax)) { printf("Syntax OK\n"); - return Qtrue; + dump &= ~DUMP_BIT(syntax); + if (!dump) return Qtrue; } - if (opt->do_print) { - PREPARE_PARSE_MAIN({ - tree = rb_parser_append_print(parser, tree); - }); - } if (opt->do_loop) { - PREPARE_PARSE_MAIN({ - tree = rb_parser_while_loop(parser, tree, opt->do_line, opt->do_split); - }); rb_define_global_function("sub", rb_f_sub, -1); rb_define_global_function("gsub", rb_f_gsub, -1); rb_define_global_function("chop", rb_f_chop, 0); rb_define_global_function("chomp", rb_f_chomp, -1); } - if (opt->dump & DUMP_BIT(parsetree) || opt->dump & DUMP_BIT(parsetree_with_comment)) { - rb_io_write(rb_stdout, rb_parser_dump_tree(tree, opt->dump & DUMP_BIT(parsetree_with_comment))); + if (dump & (DUMP_BIT(parsetree)|DUMP_BIT(parsetree_with_comment))) { + rb_io_write(rb_stdout, rb_parser_dump_tree(ast->body.root, dump & DUMP_BIT(parsetree_with_comment))); rb_io_flush(rb_stdout); - return Qtrue; + dump &= ~DUMP_BIT(parsetree)&~DUMP_BIT(parsetree_with_comment); + if (!dump) { + rb_ast_dispose(ast); + return Qtrue; + } } - PREPARE_PARSE_MAIN({ + { VALUE path = Qnil; - if (!opt->e_script && strcmp(opt->script, "-")) - path = rb_realpath_internal(Qnil, opt->script_name, 1); - iseq = rb_iseq_new_main(tree, opt->script_name, path); - }); + if (!opt->e_script && strcmp(opt->script, "-")) { + path = rb_realpath_internal(Qnil, script_name, 1); +#if UTF8_PATH + if (uenc != lenc) { + path = str_conv_enc(path, uenc, lenc); + } +#endif + if (!ENCODING_GET(path)) { /* ASCII-8BIT */ + rb_enc_copy(path, opt->script_name); + } + } + base_block = toplevel_context(toplevel_binding); + iseq = rb_iseq_new_main(&ast->body, opt->script_name, path, vm_block_iseq(base_block)); + rb_ast_dispose(ast); + } - if (opt->dump & DUMP_BIT(insns)) { - rb_io_write(rb_stdout, rb_iseq_disasm(iseq)); + if (dump & DUMP_BIT(insns)) { + rb_io_write(rb_stdout, rb_iseq_disasm((const rb_iseq_t *)iseq)); rb_io_flush(rb_stdout); - return Qtrue; + dump &= ~DUMP_BIT(insns); + if (!dump) return Qtrue; } + if (opt->dump & dump_exit_bits) return Qtrue; rb_define_readonly_boolean("$-p", opt->do_print); rb_define_readonly_boolean("$-l", opt->do_line); rb_define_readonly_boolean("$-a", opt->do_split); + if ((rb_e_script = opt->e_script) != 0) { + rb_gc_register_mark_object(opt->e_script); + } + rb_set_safe_level(opt->safe_level); - return iseq; + { + rb_execution_context_t *ec = GET_EC(); + + if (opt->e_script) { + /* -e */ + rb_exec_event_hook_script_compiled(ec, iseq, opt->e_script); + } + else { + /* file */ + rb_exec_event_hook_script_compiled(ec, iseq, Qnil); + } + } + return (VALUE)iseq; +} + +#ifndef DOSISH +static void +warn_cr_in_shebang(const char *str, long len) +{ + if (str[len-1] == '\n' && str[len-2] == '\r') { + rb_warn("shebang line ending with \\r may cause problems"); + } } +#else +#define warn_cr_in_shebang(str, len) (void)0 +#endif struct load_file_arg { VALUE parser; - const char *fname; + VALUE fname; int script; - struct cmdline_options *opt; + ruby_cmdline_options_t *opt; + VALUE f; }; static VALUE -load_file_internal(VALUE arg) +load_file_internal(VALUE argp_v) { - extern VALUE rb_stdin; - struct load_file_arg *argp = (struct load_file_arg *)arg; + struct load_file_arg *argp = (struct load_file_arg *)argp_v; VALUE parser = argp->parser; - const char *fname = argp->fname; + VALUE orig_fname = argp->fname; int script = argp->script; - struct cmdline_options *opt = argp->opt; - VALUE f; + ruby_cmdline_options_t *opt = argp->opt; + VALUE f = argp->f; int line_start = 1; - NODE *tree = 0; + rb_ast_t *ast = 0; rb_encoding *enc; ID set_encoding; - if (!fname) - rb_load_fail(fname); - if (strcmp(fname, "-") == 0) { - f = rb_stdin; - } - else { - int fd, mode = O_RDONLY; -#if defined DOSISH || defined __CYGWIN__ - { - const char *ext = strrchr(fname, '.'); - if (ext && STRCASECMP(ext, ".exe") == 0) - mode |= O_BINARY; - } -#endif - if ((fd = open(fname, mode)) < 0) { - rb_load_fail(fname); - } - - f = rb_io_fdopen(fd, mode, fname); - } - CONST_ID(set_encoding, "set_encoding"); if (script) { VALUE c = 1; /* something not nil */ VALUE line; - char *p; + char *p, *str; + long len; int no_src_enc = !opt->src.enc.name; int no_ext_enc = !opt->ext.enc.name; int no_int_enc = !opt->intern.enc.name; @@ -1539,41 +1937,39 @@ load_file_internal(VALUE arg) rb_funcall(f, set_encoding, 1, rb_enc_from_encoding(enc)); if (opt->xflag) { + line_start--; search_shebang: - forbid_setid("-x"); - opt->xflag = FALSE; while (!NIL_P(line = rb_io_gets(f))) { line_start++; - if (RSTRING_LEN(line) > 2 - && RSTRING_PTR(line)[0] == '#' - && RSTRING_PTR(line)[1] == '!') { - if ((p = strstr(RSTRING_PTR(line), "ruby")) != 0) { + RSTRING_GETMEM(line, str, len); + if (len > 2 && str[0] == '#' && str[1] == '!') { + if (line_start == 1) warn_cr_in_shebang(str, len); + if ((p = strstr(str+2, ruby_engine)) != 0) { goto start_read; } } } - rb_raise(rb_eLoadError, "no Ruby script found in input"); + rb_loaderror("no Ruby script found in input"); } c = rb_io_getbyte(f); if (c == INT2FIX('#')) { c = rb_io_getbyte(f); - if (c == INT2FIX('!')) { - line = rb_io_gets(f); - if (NIL_P(line)) - return 0; - - if ((p = strstr(RSTRING_PTR(line), "ruby")) == 0) { + if (c == INT2FIX('!') && !NIL_P(line = rb_io_gets(f))) { + RSTRING_GETMEM(line, str, len); + warn_cr_in_shebang(str, len); + if ((p = strstr(str, ruby_engine)) == 0) { /* not ruby script, assume -x flag */ goto search_shebang; } start_read: - p += 4; - RSTRING_PTR(line)[RSTRING_LEN(line) - 1] = '\0'; - if (RSTRING_PTR(line)[RSTRING_LEN(line) - 2] == '\r') - RSTRING_PTR(line)[RSTRING_LEN(line) - 2] = '\0'; + str += len - 1; + if (*str == '\n') *str-- = '\0'; + if (*str == '\r') *str-- = '\0'; + /* ruby_engine should not contain a space */ if ((p = strstr(p, " -")) != 0) { + opt->warning = 0; moreswitches(p + 1, opt, 0); } @@ -1599,11 +1995,12 @@ load_file_internal(VALUE arg) rb_io_ungetbyte(f, c); } else { - if (f != rb_stdin) rb_io_close(f); - f = Qnil; + argp->f = f = Qnil; + } + if (!(opt->dump & ~DUMP_BIT(version_v))) { + ruby_set_script_name(opt->script_name); + require_libraries(&opt->req_list); /* Why here? unnatural */ } - rb_vm_set_progname(rb_progname = opt->script_name); - require_libraries(&opt->req_list); /* Why here? unnatural */ } if (opt->src.enc.index >= 0) { enc = rb_enc_from_index(opt->src.enc.index); @@ -1612,167 +2009,259 @@ load_file_internal(VALUE arg) enc = rb_locale_encoding(); } else { - enc = rb_usascii_encoding(); + enc = rb_utf8_encoding(); } + rb_parser_set_options(parser, opt->do_print, opt->do_loop, + opt->do_line, opt->do_split); if (NIL_P(f)) { f = rb_str_new(0, 0); rb_enc_associate(f, enc); - return (VALUE)rb_parser_compile_string(parser, fname, f, line_start); + return (VALUE)rb_parser_compile_string_path(parser, orig_fname, f, line_start); } rb_funcall(f, set_encoding, 2, rb_enc_from_encoding(enc), rb_str_new_cstr("-")); - tree = rb_parser_compile_file(parser, fname, f, line_start); + ast = rb_parser_compile_file_path(parser, orig_fname, f, line_start); rb_funcall(f, set_encoding, 1, rb_parser_encoding(parser)); - if (script && tree && rb_parser_end_seen_p(parser)) { + if (script && rb_parser_end_seen_p(parser)) { + /* + * DATA is a File that contains the data section of the executed file. + * To create a data section use <tt>__END__</tt>: + * + * $ cat t.rb + * puts DATA.gets + * __END__ + * hello world! + * + * $ ruby t.rb + * hello world! + */ rb_define_global_const("DATA", f); + argp->f = Qnil; } - else if (f != rb_stdin) { - rb_io_close(f); + return (VALUE)ast; +} + +static VALUE +open_load_file(VALUE fname_v, int *xflag) +{ + const char *fname = (fname_v = rb_str_encode_ospath(fname_v), + StringValueCStr(fname_v)); + long flen = RSTRING_LEN(fname_v); + VALUE f; + int e; + + if (flen == 1 && fname[0] == '-') { + f = rb_stdin; + } + else { + int fd; + /* open(2) may block if fname is point to FIFO and it's empty. Let's + use O_NONBLOCK. */ +#if defined O_NONBLOCK && HAVE_FCNTL && !(O_NONBLOCK & O_ACCMODE) + /* TODO: fix conflicting O_NONBLOCK in ruby/win32.h */ +# define MODE_TO_LOAD (O_RDONLY | O_NONBLOCK) +#elif defined O_NDELAY && HAVE_FCNTL && !(O_NDELAY & O_ACCMODE) +# define MODE_TO_LOAD (O_RDONLY | O_NDELAY) +#else +# define MODE_TO_LOAD (O_RDONLY) +#endif + int mode = MODE_TO_LOAD; +#if defined DOSISH || defined __CYGWIN__ +# define isdirsep(x) ((x) == '/' || (x) == '\\') + { + static const char exeext[] = ".exe"; + enum {extlen = sizeof(exeext)-1}; + if (flen > extlen && !isdirsep(fname[flen-extlen-1]) && + STRNCASECMP(fname+flen-extlen, exeext, extlen) == 0) { + mode |= O_BINARY; + *xflag = 1; + } + } +#endif + + if ((fd = rb_cloexec_open(fname, mode, 0)) < 0) { + e = errno; + if (!rb_gc_for_fd(e)) { + rb_load_fail(fname_v, strerror(e)); + } + if ((fd = rb_cloexec_open(fname, mode, 0)) < 0) { + rb_load_fail(fname_v, strerror(errno)); + } + } + rb_update_max_fd(fd); + +#if defined HAVE_FCNTL && MODE_TO_LOAD != O_RDONLY + /* disabling O_NONBLOCK */ + if (fcntl(fd, F_SETFL, 0) < 0) { + e = errno; + (void)close(fd); + rb_load_fail(fname_v, strerror(e)); + } +#endif + + e = ruby_is_fd_loadable(fd); + if (!e) { + e = errno; + (void)close(fd); + rb_load_fail(fname_v, strerror(e)); + } + + f = rb_io_fdopen(fd, mode, fname); + if (e < 0) { + /* + We need to wait if FIFO is empty. It's FIFO's semantics. + rb_thread_wait_fd() release GVL. So, it's safe. + */ + rb_thread_wait_fd(fd); + } } - return (VALUE)tree; + return f; } static VALUE -restore_lineno(VALUE lineno) +restore_load_file(VALUE arg) { - return rb_gv_set("$.", lineno); + struct load_file_arg *argp = (struct load_file_arg *)arg; + VALUE f = argp->f; + + if (!NIL_P(f) && f != rb_stdin) { + rb_io_close(f); + } + return Qnil; } -static NODE * -load_file(VALUE parser, const char *fname, int script, struct cmdline_options *opt) +static rb_ast_t * +load_file(VALUE parser, VALUE fname, VALUE f, int script, ruby_cmdline_options_t *opt) { struct load_file_arg arg; arg.parser = parser; arg.fname = fname; arg.script = script; arg.opt = opt; - return (NODE *)rb_ensure(load_file_internal, (VALUE)&arg, restore_lineno, rb_gv_get("$.")); + arg.f = f; + return (rb_ast_t *)rb_ensure(load_file_internal, (VALUE)&arg, + restore_load_file, (VALUE)&arg); } void * rb_load_file(const char *fname) { - struct cmdline_options opt; + VALUE fname_v = rb_str_new_cstr(fname); + return rb_load_file_str(fname_v); +} - return load_file(rb_parser_new(), fname, 0, cmdline_options_init(&opt)); +void * +rb_load_file_str(VALUE fname_v) +{ + return rb_parser_load_file(rb_parser_new(), fname_v); } -#if !defined(PSTAT_SETCMD) && !defined(HAVE_SETPROCTITLE) -#if !defined(_WIN32) -#define USE_ENVSPACE_FOR_ARG0 -#endif +void * +rb_parser_load_file(VALUE parser, VALUE fname_v) +{ + ruby_cmdline_options_t opt; + VALUE f = open_load_file(fname_v, &cmdline_options_init(&opt)->xflag); + return load_file(parser, fname_v, f, 0, &opt); +} -#ifdef USE_ENVSPACE_FOR_ARG0 -extern char **environ; -#endif +/* + * call-seq: + * Process.argv0 -> frozen_string + * + * Returns the name of the script being executed. The value is not + * affected by assigning a new value to $0. + * + * This method first appeared in Ruby 2.1 to serve as a global + * variable free means to get the script name. + */ -static size_t -get_arglen(int argc, char **argv) +static VALUE +proc_argv0(VALUE process) { - char *s = argv[0]; - int i; + return rb_orig_progname; +} - if (!argc) return 0; - s += strlen(s); - /* See if all the arguments are contiguous in memory */ - for (i = 1; i < argc; i++) { - if (argv[i] == s + 1) { - s++; - s += strlen(s); /* this one is ok too */ - } - else { - break; - } - } -#if defined(USE_ENVSPACE_FOR_ARG0) - if (environ && (s+1 == environ[0])) { - s++; - s += strlen(s); - for (i = 1; environ[i]; i++) { - if (environ[i] == s + 1) { - s++; - s += strlen(s); /* this one is ok too */ - } - } -# if defined(HAVE_SETENV) && defined(HAVE_UNSETENV) - { - char *t = malloc(s - environ[0] + 1); - for (i = 0; environ[i]; i++) { - size_t len = strlen(environ[i]) + 1; - memcpy(t, environ[i], len); - environ[i] = t; - t += len; - } - } -# else - ruby_setenv("", NULL); /* duplicate environ vars */ -# endif - } -#endif - return s - argv[0]; +static VALUE ruby_setproctitle(VALUE title); + +/* + * call-seq: + * Process.setproctitle(string) -> string + * + * Sets the process title that appears on the ps(1) command. Not + * necessarily effective on all platforms. No exception will be + * raised regardless of the result, nor will NotImplementedError be + * raised even if the platform does not support the feature. + * + * Calling this method does not affect the value of $0. + * + * Process.setproctitle('myapp: worker #%d' % worker_id) + * + * This method first appeared in Ruby 2.1 to serve as a global + * variable free means to change the process title. + */ + +static VALUE +proc_setproctitle(VALUE process, VALUE title) +{ + return ruby_setproctitle(title); +} + +static VALUE +ruby_setproctitle(VALUE title) +{ + const char *ptr = StringValueCStr(title); + setproctitle("%.*s", RSTRING_LENINT(title), ptr); + return title; } -#endif static void set_arg0(VALUE val, ID id) { - char *s; - long i; - if (origarg.argv == 0) rb_raise(rb_eRuntimeError, "$0 not initialized"); - StringValue(val); - s = RSTRING_PTR(val); - i = RSTRING_LEN(val); -#if defined(PSTAT_SETCMD) - if (i > PST_CLEN) { - union pstun un; - char buf[PST_CLEN + 1]; /* PST_CLEN is 64 (HP-UX 11.23) */ - strlcpy(buf, s, sizeof(buf)); - un.pst_command = buf; - pstat(PSTAT_SETCMD, un, PST_CLEN, 0, 0); - } - else { - union pstun un; - un.pst_command = s; - pstat(PSTAT_SETCMD, un, i, 0, 0); - } -#elif defined(HAVE_SETPROCTITLE) - setproctitle("%.*s", (int)i, s); -#else - - if ((size_t)i > origarg.len - origarg.argc) { - i = (long)(origarg.len - origarg.argc); - } - memcpy(origarg.argv[0], s, i); - - { - int j; - char *t = origarg.argv[0] + i; - *t = '\0'; + rb_progname = rb_str_new_frozen(ruby_setproctitle(val)); +} - if ((size_t)(i + 1) < origarg.len) { - memset(t + 1, '\0', origarg.len - i - 1); - } - for (j = 1; j < origarg.argc; j++) { - origarg.argv[j] = t; - } - } +static inline VALUE +external_str_new_cstr(const char *p) +{ +#if UTF8_PATH + VALUE str = rb_utf8_str_new_cstr(p); + str = str_conv_enc(str, NULL, rb_default_external_encoding()); + OBJ_TAINT_RAW(str); + return str; +#else + return rb_external_str_new_cstr(p); #endif - rb_progname = rb_obj_freeze(rb_external_str_new(s, i)); } +/*! Sets the current script name to this value. + * + * This is similar to <code>$0 = name</code> in Ruby level but also affects + * <code>Method#location</code> and others. + */ void ruby_script(const char *name) { if (name) { - rb_progname = rb_obj_freeze(rb_external_str_new(name, strlen(name))); + rb_orig_progname = rb_progname = external_str_new_cstr(name); rb_vm_set_progname(rb_progname); } } +/*! Sets the current script name to this value. + * + * Same as ruby_script() but accepts a VALUE. + */ +void +ruby_set_script_name(VALUE name) +{ + rb_orig_progname = rb_progname = rb_str_dup(name); + rb_vm_set_progname(rb_progname); +} + static void -init_ids(struct cmdline_options *opt) +init_ids(ruby_cmdline_options_t *opt) { rb_uid_t uid = getuid(); rb_uid_t euid = geteuid(); @@ -1788,7 +2277,7 @@ init_ids(struct cmdline_options *opt) #undef forbid_setid static void -forbid_setid(const char *s, struct cmdline_options *opt) +forbid_setid(const char *s, const ruby_cmdline_options_t *opt) { if (opt->setids & 1) rb_raise(rb_eSecurityError, "no %s allowed while running setuid", s); @@ -1816,10 +2305,12 @@ opt_W_getter(ID id, void *data) return INT2FIX(1); case Qtrue: return INT2FIX(2); + default: + return Qnil; } - return Qnil; /* not reached */ } +/*! Defines built-in variables */ void ruby_prog_init(void) { @@ -1833,6 +2324,15 @@ ruby_prog_init(void) rb_define_hooked_variable("$0", &rb_progname, 0, set_arg0); rb_define_hooked_variable("$PROGRAM_NAME", &rb_progname, 0, set_arg0); + rb_define_module_function(rb_mProcess, "argv0", proc_argv0, 0); + rb_define_module_function(rb_mProcess, "setproctitle", proc_setproctitle, 1); + + /* + * ARGV contains the command line arguments used to run ruby. + * + * A library like OptionParser can be used to process command-line + * arguments. + */ rb_define_global_const("ARGV", rb_argv); } @@ -1843,14 +2343,14 @@ ruby_set_argv(int argc, char **argv) VALUE av = rb_argv; #if defined(USE_DLN_A_OUT) - if (origarg.argv) + if (origarg.argc > 0 && origarg.argv) dln_argv0 = origarg.argv[0]; - else + else if (argc > 0 && argv) dln_argv0 = argv[0]; #endif rb_ary_clear(av); for (i = 0; i < argc; i++) { - VALUE arg = rb_external_str_new(argv[i], strlen(argv[i])); + VALUE arg = external_str_new_cstr(argv[i]); OBJ_FREEZE(arg); rb_ary_push(av, arg); @@ -1860,29 +2360,75 @@ ruby_set_argv(int argc, char **argv) void * ruby_process_options(int argc, char **argv) { - struct cmdline_options opt; + ruby_cmdline_options_t opt; VALUE iseq; + const char *script_name = (argc > 0 && argv[0]) ? argv[0] : ruby_engine; - ruby_script(argv[0]); /* for the time being */ + if (!origarg.argv || origarg.argc <= 0) { + origarg.argc = argc; + origarg.argv = argv; + } + ruby_script(script_name); /* for the time being */ rb_argv0 = rb_str_new4(rb_progname); rb_gc_register_mark_object(rb_argv0); iseq = process_options(argc, argv, cmdline_options_init(&opt)); + +#ifndef HAVE_SETPROCTITLE + ruby_init_setproctitle(argc, argv); +#endif + return (void*)(struct RData*)iseq; } +static void +fill_standard_fds(void) +{ + int f0, f1, f2, fds[2]; + struct stat buf; + f0 = fstat(0, &buf) == -1 && errno == EBADF; + f1 = fstat(1, &buf) == -1 && errno == EBADF; + f2 = fstat(2, &buf) == -1 && errno == EBADF; + if (f0) { + if (pipe(fds) == 0) { + close(fds[1]); + if (fds[0] != 0) { + dup2(fds[0], 0); + close(fds[0]); + } + } + } + if (f1 || f2) { + if (pipe(fds) == 0) { + close(fds[0]); + if (f1 && fds[1] != 1) + dup2(fds[1], 1); + if (f2 && fds[1] != 2) + dup2(fds[1], 2); + if (fds[1] != 1 && fds[1] != 2) + close(fds[1]); + } + } +} + +/*! Initializes the process for libruby. + * + * This function assumes this process is ruby(1) and it has just started. + * Usually programs that embed CRuby interpreter may not call this function, + * and may do their own initialization. + * argc and argv cannot be NULL. + */ void ruby_sysinit(int *argc, char ***argv) { #if defined(_WIN32) - void rb_w32_sysinit(int *argc, char ***argv); rb_w32_sysinit(argc, argv); #endif - origarg.argc = *argc; - origarg.argv = *argv; -#if !defined(PSTAT_SETCMD) && !defined(HAVE_SETPROCTITLE) - origarg.len = get_arglen(origarg.argc, origarg.argv); -#endif + if (*argc >= 0 && *argv) { + origarg.argc = *argc; + origarg.argv = *argv; #if defined(USE_DLN_A_OUT) - dln_argv0 = origarg.argv[0]; + dln_argv0 = origarg.argv[0]; #endif + } + fill_standard_fds(); } |
