diff options
Diffstat (limited to 'ruby.c')
| -rw-r--r-- | ruby.c | 2336 |
1 files changed, 2038 insertions, 298 deletions
@@ -1,428 +1,2168 @@ -/************************************************ +/********************************************************************** ruby.c - - $Author: matz $ - $Date: 1995/01/10 10:42:51 $ + $Author$ created at: Tue Aug 10 12:47:31 JST 1993 - Copyright (C) 1994 Yukihiro Matsumoto + Copyright (C) 1993-2007 Yukihiro Matsumoto + Copyright (C) 2000 Network Applied Communication Laboratory, Inc. + Copyright (C) 2000 Information-technology Promotion Agency, Japan -************************************************/ +**********************************************************************/ -#include "ruby.h" -#include "re.h" +#ifdef __CYGWIN__ +#include <windows.h> +#include <sys/cygwin.h> +#endif +#include "internal.h" +#include "ruby/thread.h" +#include "eval_intern.h" +#include "dln.h" #include <stdio.h> -#include <fcntl.h> #include <sys/types.h> -#include <sys/stat.h> -#include <signal.h> +#include <ctype.h> + +#ifdef __hpux +#include <sys/pstat.h> +#endif +#if defined(LOAD_RELATIVE) && defined(HAVE_DLADDR) +#include <dlfcn.h> +#endif + +#ifdef HAVE_UNISTD_H +#include <unistd.h> +#endif +#if defined(HAVE_FCNTL_H) +#include <fcntl.h> +#elif defined(HAVE_SYS_FCNTL_H) +#include <sys/fcntl.h> +#endif +#ifdef HAVE_SYS_PARAM_H +# include <sys/param.h> +#endif +#ifndef MAXPATHLEN +# define MAXPATHLEN 1024 +#endif +#ifndef O_ACCMODE +# define O_ACCMODE (O_RDONLY | O_WRONLY | O_RDWR) +#endif -#include "getopt.h" +#include "ruby/util.h" -static int version, copyright; +#ifndef HAVE_STDLIB_H +char *getenv(); +#endif -static struct option long_options[] = -{ - {"debug", 0, 0, 'd'}, - {"yydebug", 0, 0, 'y'}, - {"verbose", 0, 0, 'v'}, - {"version", 0, &version, 1}, - {"copyright", 0, ©right, 1}, - {0, 0, 0, 0} +#ifndef DISABLE_RUBYGEMS +# define DISABLE_RUBYGEMS 0 +#endif +#if DISABLE_RUBYGEMS +#define DEFAULT_RUBYGEMS_ENABLED "disabled" +#else +#define DEFAULT_RUBYGEMS_ENABLED "enabled" +#endif + +#define FEATURE_BIT(bit) (1U << feature_##bit) +enum feature_flag_bits { + feature_gems, + feature_did_you_mean, + feature_rubyopt, + feature_frozen_string_literal, + feature_debug_frozen_string_literal, + feature_flag_count +}; + +#define DEBUG_BIT(bit) (1U << feature_debug_##bit) + +#define DUMP_BIT(bit) (1U << dump_##bit) +enum dump_flag_bits { + dump_version, + dump_version_v, + dump_copyright, + dump_usage, + dump_help, + dump_yydebug, + dump_syntax, + dump_parsetree, + dump_parsetree_with_comment, + dump_insns, + dump_flag_count +}; + +struct cmdline_options { + int sflag, xflag; + int do_loop, do_print; + int do_line, do_split; + int do_search; + unsigned int features; + int verbose; + int safe_level; + unsigned int setids; + unsigned int dump; + const char *script; + VALUE script_name; + VALUE e_script; + struct { + struct { + VALUE name; + int index; + } enc; + } src, ext, intern; + VALUE req_list; }; -int debug = 0; -int verbose = 0; -static int sflag = FALSE; +static void init_ids(struct cmdline_options *); + +#define src_encoding_index GET_VM()->src_encoding_index + +enum { + COMPILATION_FEATURES = ( + 0 + | FEATURE_BIT(frozen_string_literal) + | FEATURE_BIT(debug_frozen_string_literal) + ), + DEFAULT_FEATURES = ( + ~0U +#if DISABLE_RUBYGEMS + & ~FEATURE_BIT(gems) +#endif + & ~FEATURE_BIT(frozen_string_literal) + & ~FEATURE_BIT(debug_frozen_string_literal) + ) +}; + +static struct cmdline_options * +cmdline_options_init(struct cmdline_options *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; + return opt; +} + +static NODE *load_file(VALUE, VALUE, int, struct cmdline_options *); +static void forbid_setid(const char *, struct cmdline_options *); +#define forbid_setid(s) forbid_setid((s), opt) -char *inplace = Qnil; -char *strdup(); -char *strstr(); -char *strchr(); -char *dln_find_file(); +static struct { + int argc; + char **argv; +} origarg; -extern int yydebug; -extern int nerrs; +static void +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 - 2 > 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? */ + + 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", ", --verbose", "print 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("-h", "", "show this message, --help for more info"), + }; + static const struct message help_msg[] = { + M("--copyright", "", "print the copyright"), + M("--enable=feature[,...]", ", --disable=feature[,...]", + "enable or disable features"), + M("--external-encoding=encoding", ", --internal-encoding=encoding", + "specify the default external or internal character encoding"), + M("--version", "", "print the version"), + M("--help", "", "show this message, -h for short message"), + }; + 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)"), + }; + 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); + 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("Features:"); + for (i = 0; i < numberof(features); ++i) + SHOW(features[i]); +} + +#define rubylib_path_new rb_str_new + +static void +push_include(const char *path, VALUE (*filter)(VALUE)) +{ + const char sep = PATH_SEP_CHAR; + const char *p, *s; + VALUE load_path = GET_VM()->load_path; + + p = path; + while (*p) { + while (*p == sep) + p++; + if (!*p) break; + for (s = p; *s && *s != sep; s = CharNext(s)); + rb_ary_push(load_path, (*filter)(rubylib_path_new(p, s - p))); + p = s; + } +} + +#ifdef __CYGWIN__ +static void +push_include_cygwin(const char *path, VALUE (*filter)(VALUE)) +{ + const char *p, *s; + char rubylib[FILENAME_MAX]; + VALUE buf = 0; + + p = path; + while (*p) { + unsigned int len; + while (*p == ';') + p++; + if (!*p) break; + for (s = p; *s && *s != ';'; s = CharNext(s)); + len = s - p; + if (*s) { + if (!buf) { + buf = rb_str_new(p, len); + p = RSTRING_PTR(buf); + } + else { + rb_str_resize(buf, len); + p = strncpy(RSTRING_PTR(buf), p, len); + } + } +#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; + p = s + 1; + } +} + +#define push_include push_include_cygwin +#endif + +void +ruby_push_include(const char *path, VALUE (*filter)(VALUE)) +{ + if (path == 0) + return; + push_include(path, filter); +} + +static VALUE +identical_path(VALUE path) +{ + return path; +} +static VALUE +locale_path(VALUE path) +{ + rb_enc_associate(path, rb_locale_encoding()); + return path; +} + +void +ruby_incpush(const char *path) +{ + ruby_push_include(path, locale_path); +} + +static VALUE +expand_include_path(VALUE path) +{ + char *p = RSTRING_PTR(path); + if (!p) + return path; + if (*p == '.' && p[1] == '/') + return path; + return rb_file_expand_path(path, Qnil); +} + +void +ruby_incpush_expand(const char *path) +{ + ruby_push_include(path, expand_include_path); +} + +#undef UTF8_PATH +#if defined _WIN32 || defined __CYGWIN__ +static HMODULE libruby; + +BOOL WINAPI +DllMain(HINSTANCE dll, DWORD reason, LPVOID reserved) +{ + if (reason == DLL_PROCESS_ATTACH) + libruby = dll; + return TRUE; +} + +HANDLE +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 +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); +} +#endif + +void ruby_init_loadpath_safe(int safe_level); + +void +ruby_init_loadpath(void) +{ + ruby_init_loadpath_safe(0); +} + +#if defined(LOAD_RELATIVE) && defined(HAVE_DLADDR) +static VALUE +dladdr_path(const void* addr) +{ + Dl_info dli; + VALUE fname, path; + + if (!dladdr(addr, &dli)) { + return rb_str_new(0, 0); + } +#ifdef __linux__ + else if (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; +} +#endif + +#define INITIAL_LOAD_PATH_MARK rb_intern_const("@gem_prelude_index") + +void +ruby_init_loadpath_safe(int safe_level) +{ + VALUE load_path; + ID id_initial_load_path_mark; + const char *paths = ruby_initial_load_paths; +#if defined LOAD_RELATIVE +# if defined HAVE_DLADDR || defined __CYGWIN__ || defined _WIN32 +# define VARIABLE_LIBPATH 1 +# else +# define VARIABLE_LIBPATH 0 +# endif +# if VARIABLE_LIBPATH + char *libpath; + VALUE sopath; +# else + char libpath[MAXPATHLEN + 1]; +# endif + size_t baselen; + char *p; + +#if defined _WIN32 || defined __CYGWIN__ + sopath = rb_str_new(0, MAXPATHLEN); + libpath = RSTRING_PTR(sopath); + GetModuleFileName(libruby, libpath, MAXPATHLEN); +#elif defined(HAVE_DLADDR) + sopath = dladdr_path((void *)(VALUE)expand_include_path); + libpath = RSTRING_PTR(sopath); +#endif + +#if !VARIABLE_LIBPATH + libpath[sizeof(libpath) - 1] = '\0'; +#endif +#if defined DOSISH + translit_char(libpath, '\\', '/'); +#elif defined __CYGWIN__ + { + 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; + } + } + } +#endif + p = strrchr(libpath, '/'); + if (p) { + static const char bindir[] = "/bin"; +#ifdef LIBDIR_BASENAME + static const char libdir[] = "/"LIBDIR_BASENAME; +#else + static const char libdir[] = "/lib"; +#endif + const ptrdiff_t bindir_len = (ptrdiff_t)sizeof(bindir) - 1; + const ptrdiff_t libdir_len = (ptrdiff_t)sizeof(libdir) - 1; + +#ifdef ENABLE_MULTIARCH + char *p2 = NULL; + + multiarch: +#endif + if (p - libpath >= bindir_len && !STRNCASECMP(p - bindir_len, bindir, bindir_len)) { + p -= bindir_len; + } + else if (p - libpath >= libdir_len && !strncmp(p - libdir_len, libdir, libdir_len)) { + 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 +#if !VARIABLE_LIBPATH + *p = 0; +#endif + } +#if !VARIABLE_LIBPATH + else { + strlcpy(libpath, ".", sizeof(libpath)); + p = libpath + 1; + } + baselen = p - libpath; +#define PREFIX_PATH() rb_str_new(libpath, baselen) +#else + baselen = p - libpath; + rb_str_resize(sopath, baselen); + libpath = RSTRING_PTR(sopath); +#define PREFIX_PATH() sopath +#endif -int xflag = FALSE; +#define BASEPATH() rb_str_buf_cat(rb_str_buf_new(baselen+len), libpath, baselen) -#ifdef USE_DLN -char *rb_dln_argv0; +#define RUBY_RELATIVE(path, len) rb_str_buf_cat(BASEPATH(), (path), (len)) +#else + 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 + load_path = GET_VM()->load_path; + + if (safe_level == 0) { + ruby_push_include(getenv("RUBYLIB"), identical_path); + } + + id_initial_load_path_mark = INITIAL_LOAD_PATH_MARK; + while (*paths) { + size_t len = strlen(paths); + VALUE path = RUBY_RELATIVE(paths, len); + rb_ivar_set(path, id_initial_load_path_mark, path); + rb_ary_push(load_path, path); + paths += len + 1; + } + + rb_const_set(rb_cObject, rb_intern_const("TMP_RUBY_PREFIX"), rb_obj_freeze(PREFIX_PATH())); +} + + +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_CLEAR_CLASS(list); + } + feature = rb_str_new2(mod); + RBASIC_CLEAR_CLASS(feature); + rb_ary_push(list, feature); +} -static void load_stdin(); -void rb_load_file(); +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_encoding *extenc = rb_default_external_encoding(); + int prev_parse_in_eval = th->parse_in_eval; + th->parse_in_eval = 0; + + CONST_ID(require, "require"); + while (list && RARRAY_LEN(list) > 0) { + VALUE feature = rb_ary_shift(list); + rb_enc_associate(feature, extenc); + RBASIC_SET_CLASS_RAW(feature, rb_cString); + OBJ_FREEZE(feature); + rb_funcall2(self, require, 1, &feature); + } + *req_list = 0; -static int do_loop = FALSE, do_print = FALSE; -static int do_check = FALSE, do_line = FALSE; -static int do_split = FALSE; + th->parse_in_eval = prev_parse_in_eval; +} -static char *script; +static rb_env_t* +toplevel_context(VALUE toplevel_binding) +{ + rb_env_t *env; + rb_binding_t *bind; + + GetBindingPtr(toplevel_binding, bind); + GetEnvPtr(bind->env, env); + return env; +} static void -proc_options(argcp, argvp) - int *argcp; - char ***argvp; -{ - int argc = *argcp; - char **argv = *argvp; - extern VALUE rb_load_path; - extern char *optarg; - extern int optind; - int c, i, j, script_given, do_search, opt_index; - extern VALUE RS, ORS, FS; - char *src; - - if (argc == 0) return; - - version = FALSE; - script_given = FALSE; - do_search = FALSE; - - optind = 0; - while ((c = getopt_long(argc, argv, "+acC:de:F:i:I:lnpR:svxX:yS", - long_options, &opt_index)) != EOF) { - switch (c) { +process_sflag(int *sflag) +{ + if (*sflag > 0) { + long n; + const VALUE *args; + VALUE argv = rb_argv; + + n = RARRAY_LEN(argv); + args = RARRAY_CONST_PTR(argv); + while (n > 0) { + VALUE v = *args++; + char *s = StringValuePtr(v); + char *p; + int hyphen = FALSE; + + if (s[0] != '-') + break; + n--; + if (s[1] == '-' && s[2] == '\0') + break; + + v = Qtrue; + /* check if valid name before replacing - with _ */ + for (p = s + 1; *p; p++) { + if (*p == '=') { + *p++ = '\0'; + v = rb_str_new2(p); + break; + } + if (*p == '-') { + hyphen = TRUE; + } + else if (*p != '_' && !ISALNUM(*p)) { + VALUE name_error[2]; + name_error[0] = + rb_str_new2("invalid name for global variable - "); + if (!(p = strchr(p, '='))) { + rb_str_cat2(name_error[0], s); + } + else { + rb_str_cat(name_error[0], s, p - s); + } + name_error[1] = args[-1]; + rb_exc_raise(rb_class_new_instance(2, name_error, rb_eNameError)); + } + } + s[0] = '$'; + if (hyphen) { + for (p = s + 1; *p; ++p) { + if (*p == '-') + *p = '_'; + } + } + rb_gv_set(s, v); + } + n = RARRAY_LEN(argv) - n; + while (n--) { + rb_ary_shift(argv); + } + *sflag = -1; + } +} + +static long proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt); + +static void +moreswitches(const char *s, struct cmdline_options *opt, int envopt) +{ + long argc, i, len; + char **argv, *p; + const char *ap = 0; + VALUE argstr, argary; + + while (ISSPACE(*s)) s++; + if (!*s) return; + argstr = rb_str_tmp_new((len = strlen(s)) + 2); + argary = rb_str_tmp_new(0); + + p = RSTRING_PTR(argstr); + *p++ = ' '; + memcpy(p, s, len + 1); + ap = 0; + rb_str_cat(argary, (char *)&ap, sizeof(ap)); + while (*p) { + ap = p; + rb_str_cat(argary, (char *)&ap, sizeof(ap)); + while (*p && !ISSPACE(*p)) ++p; + if (!*p) break; + *p++ = '\0'; + while (ISSPACE(*p)) ++p; + } + argc = RSTRING_LEN(argary) / sizeof(ap); + ap = 0; + rb_str_cat(argary, (char *)&ap, sizeof(ap)); + argv = (char **)RSTRING_PTR(argary); + + while ((i = proc_options(argc, argv, opt, envopt)) > 1 && (argc -= i) > 0) { + argv += i; + if (**argv != '-') { + *--*argv = '-'; + } + if ((*argv)[1]) { + ++argc; + --argv; + } + } + + /* 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) && name_match_p((name), (str), (len))) + +#define UNSET_WHEN(name, bit, 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))) { \ + *(unsigned int *)arg |= (bit); \ + return; \ + } + +static void +feature_option(const char *str, int len, void *arg, const unsigned int enable) +{ + unsigned int *argp = arg; + unsigned int mask = ~0U; +#define SET_FEATURE(bit) \ + if (NAME_MATCH_P(#bit, str, len)) {mask = FEATURE_BIT(bit); goto found;} + SET_FEATURE(gems); + SET_FEATURE(did_you_mean); + SET_FEATURE(rubyopt); + SET_FEATURE(frozen_string_literal); + if (NAME_MATCH_P("all", str, len)) { + found: + *argp = (*argp & ~mask) | (mask & enable); + return; + } + rb_warn("unknown argument for --%s: `%.*s'", + enable ? "enable" : "disable", len, str); +} + +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) +{ + feature_option(str, len, arg, 0U); +} + +static void +debug_option(const char *str, int len, void *arg) +{ +#define SET_WHEN_DEBUG(t, bit) SET_WHEN(#bit, t##_BIT(bit), str, len) + SET_WHEN_DEBUG(DEBUG, frozen_string_literal); + rb_warn("unknown argument for --debug: `%.*s'", len, str); +} + +static void +dump_option(const char *str, int len, void *arg) +{ +#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(help); + SET_WHEN_DUMP(yydebug); + SET_WHEN_DUMP(syntax); + SET_WHEN_DUMP(parsetree); + SET_WHEN_DUMP(parsetree_with_comment); + SET_WHEN_DUMP(insns); + 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]."); +} + +static void +set_option_encoding_once(const char *type, VALUE *name, const char *e, long elen) +{ + VALUE ename; + + if (!elen) elen = strlen(e); + ename = rb_str_new(e, elen); + + if (*name && + rb_funcall(ename, rb_intern("casecmp"), 1, *name) != INT2FIX(0)) { + rb_raise(rb_eRuntimeError, + "%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)) +#define set_external_encoding_once(opt, 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)) + +static long +proc_options(long argc, char **argv, struct cmdline_options *opt, int envopt) +{ + long n, argc0 = argc; + const char *s; + + if (argc == 0) + return 0; + + for (argc--, argv++; argc > 0; argc--, argv++) { + const char *const arg = argv[0]; + if (!arg || arg[0] != '-' || !arg[1]) + break; + + s = arg + 1; + reswitch: + switch (*s) { + case 'a': + if (envopt) goto noenvopt; + opt->do_split = TRUE; + s++; + goto reswitch; + case 'p': - do_print = TRUE; + if (envopt) goto noenvopt; + opt->do_print = TRUE; /* through */ case 'n': - do_loop = TRUE; - break; + if (envopt) goto noenvopt; + opt->do_loop = TRUE; + s++; + goto reswitch; case 'd': - debug = TRUE; - break; + ruby_debug = Qtrue; + ruby_verbose = Qtrue; + s++; + goto reswitch; case 'y': - yydebug = 1; - break; + if (envopt) goto noenvopt; + opt->dump |= DUMP_BIT(yydebug); + s++; + goto reswitch; case 'v': - verbose = TRUE; - show_version(); - break; + if (opt->verbose) { + s++; + goto reswitch; + } + opt->dump |= DUMP_BIT(version_v); + opt->verbose = 1; + case 'w': + ruby_verbose = Qtrue; + s++; + goto reswitch; + + case 'W': + { + size_t numlen; + int v = 2; /* -W as -W2 */ + + if (*++s) { + v = scan_oct(s, 1, &numlen); + if (numlen == 0) + v = 1; + s += numlen; + } + switch (v) { + case 0: + ruby_verbose = Qnil; + break; + case 1: + ruby_verbose = Qfalse; + break; + default: + ruby_verbose = Qtrue; + break; + } + } + goto reswitch; + + case 'c': + if (envopt) goto noenvopt; + opt->dump |= DUMP_BIT(syntax); + s++; + goto reswitch; + + case 's': + if (envopt) goto noenvopt; + forbid_setid("-s"); + if (!opt->sflag) opt->sflag = 1; + s++; + goto reswitch; + + case 'h': + if (envopt) goto noenvopt; + opt->dump |= DUMP_BIT(usage); + goto switch_end; + + case 'l': + if (envopt) goto noenvopt; + opt->do_line = TRUE; + rb_output_rs = rb_rs; + s++; + goto reswitch; + + case 'S': + if (envopt) goto noenvopt; + forbid_setid("-S"); + opt->do_search = TRUE; + s++; + goto reswitch; case 'e': - script_given++; - if (script == 0) script = "-e"; - lex_setsrc("-e", optarg, strlen(optarg)); - yyparse(); + if (envopt) goto noenvopt; + forbid_setid("-e"); + if (!*++s) { + 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); + if (opt->script == 0) + opt->script = "-e"; + } + rb_str_cat2(opt->e_script, s); + rb_str_cat2(opt->e_script, "\n"); break; - case 'i': - inplace = strdup(optarg); + case 'r': + forbid_setid("-r"); + if (*++s) { + add_modules(&opt->req_list, s); + } + else if (argc > 1) { + add_modules(&opt->req_list, argv[1]); + argc--, argv++; + } break; - case 'c': - do_check = TRUE; + case 'i': + if (envopt) goto noenvopt; + forbid_setid("-i"); + ruby_set_inplace_mode(s + 1); break; case 'x': - xflag = TRUE; + if (envopt) goto noenvopt; + opt->xflag = TRUE; + s++; + if (*s && chdir(s) < 0) { + rb_fatal("Can't chdir to %s", s); + } break; + case 'C': case 'X': - if (chdir(optarg) < 0) - Fatal("Can't chdir to %s", optarg); + if (envopt) goto noenvopt; + if (!*++s && (!--argc || !(s = *++argv) || !*s)) { + rb_fatal("Can't chdir"); + } + if (chdir(s) < 0) { + rb_fatal("Can't chdir to %s", s); + } break; - case 's': - sflag = TRUE; + case 'F': + if (envopt) goto noenvopt; + if (*++s) { + rb_fs = rb_reg_new(s, strlen(s), 0); + } break; - case 'l': - do_line = TRUE; - ORS = RS; - break; + case 'E': + if (!*++s && (!--argc || !(s = *++argv))) { + rb_raise(rb_eRuntimeError, "missing argument for -E"); + } + goto encoding; + + case 'U': + set_internal_encoding_once(opt, "UTF-8", 0); + ++s; + goto reswitch; + + case 'K': + if (*++s) { + const char *enc_name = 0; + switch (*s) { + case 'E': case 'e': + enc_name = "EUC-JP"; + break; + case 'S': case 's': + enc_name = "Windows-31J"; + break; + case 'U': case 'u': + enc_name = "UTF-8"; + break; + case 'N': case 'n': case 'A': case 'a': + enc_name = "ASCII-8BIT"; + break; + } + if (enc_name) { + opt->src.enc.name = rb_str_new2(enc_name); + if (!opt->ext.enc.name) + opt->ext.enc.name = opt->src.enc.name; + } + s++; + } + goto reswitch; - case 'R': + case 'T': { - char *p = optarg; - - while (*p) { - if (*p < '0' || '7' < *p) { - break; - } - p++; + size_t numlen; + int v = 1; + + if (*++s) { + v = scan_oct(s, 2, &numlen); + if (numlen == 0) + v = 1; + s += numlen; } - if (*p) { - RS = str_new2(optarg); + if (v > opt->safe_level) opt->safe_level = v; + } + goto reswitch; + + case 'I': + forbid_setid("-I"); + if (*++s) + ruby_incpush_expand(s); + else if (argc > 1) { + ruby_incpush_expand(argv[1]); + argc--, argv++; + } + break; + + case '0': + if (envopt) goto noenvopt; + { + size_t numlen; + int v; + char c; + + v = scan_oct(s, 4, &numlen); + s += numlen; + if (v > 0377) + rb_rs = Qnil; + else if (v == 0 && numlen >= 2) { + rb_rs = rb_str_new2("\n\n"); } else { - int i = strtoul(optarg, Qnil, 8); - - if (i == 0) RS = str_new(0, 0); - else if (i > 0xff) RS = Qnil; - else { - char c = i; - RS = str_new(&c, 1); - } + c = v & 0xff; + rb_rs = rb_str_new(&c, 1); } } - break; + goto reswitch; - case 'F': - FS = str_new2(optarg); + case '-': + if (!s[1] || (s[1] == '\r' && !s[2])) { + argc--, argv++; + goto switch_end; + } + s++; + +# define is_option_end(c, allow_hyphen) \ + (!(c) || ((allow_hyphen) && (c) == '-') || (c) == '=') +# define check_envopt(name, allow_envopt) \ + (((allow_envopt) || !envopt) ? (void)0 : \ + rb_raise(rb_eRuntimeError, "invalid switch in RUBYOPT: --" name)) +# 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) \ + 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 (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->features); + } + else if (is_option_with_arg("disable", Qtrue, Qtrue)) { + ruby_each_words(s, disable_option, &opt->features); + } + else if (is_option_with_arg("encoding", Qfalse, Qtrue)) { + char *p; + encoding: + do { +# define set_encoding_part(type) \ + if (!(p = strchr(s, ':'))) { \ + set_##type##_encoding_once(opt, s, 0); \ + break; \ + } \ + else if (p > s) { \ + set_##type##_encoding_once(opt, s, p-s); \ + } + set_encoding_part(external); + if (!*(s = ++p)) break; + set_encoding_part(internal); + if (!*(s = ++p)) break; +#if defined ALLOW_DEFAULT_SOURCE_ENCODING && ALLOW_DEFAULT_SOURCE_ENCODING + set_encoding_part(source); + if (!*(s = ++p)) break; +#endif + rb_raise(rb_eRuntimeError, "extra argument for %s: %s", + (arg[1] == '-' ? "--encoding" : "-E"), s); +# undef set_encoding_part + } while (0); + } + else if (is_option_with_arg("internal-encoding", Qfalse, Qtrue)) { + set_internal_encoding_once(opt, s, 0); + } + else if (is_option_with_arg("external-encoding", Qfalse, Qtrue)) { + set_external_encoding_once(opt, s, 0); + } +#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); + } +#endif + else if (strcmp("version", s) == 0) { + if (envopt) goto noenvopt_long; + opt->dump |= DUMP_BIT(version); + } + else if (strcmp("verbose", s) == 0) { + opt->verbose = 1; + ruby_verbose = Qtrue; + } + else if (strcmp("yydebug", s) == 0) { + if (envopt) goto noenvopt_long; + opt->dump |= DUMP_BIT(yydebug); + } + else if (is_option_with_arg("dump", Qfalse, Qfalse)) { + ruby_each_words(s, dump_option, &opt->dump); + } + else if (strcmp("help", s) == 0) { + if (envopt) goto noenvopt_long; + opt->dump |= DUMP_BIT(help); + goto switch_end; + } + else { + rb_raise(rb_eRuntimeError, + "invalid option --%s (-h will show valid options)", s); + } break; - case 'a': - do_split = TRUE; - break; + case '\r': + if (!s[1]) + break; - case 'C': - rb_set_kanjicode(optarg); - break; + default: + { + if (ISPRINT(*s)) { + rb_raise(rb_eRuntimeError, + "invalid option -%c (-h will show valid options)", + (int)(unsigned char)*s); + } + else { + rb_raise(rb_eRuntimeError, + "invalid option -\\x%02X (-h will show valid options)", + (int)(unsigned char)*s); + } + } + goto switch_end; - case 'S': - do_search = TRUE; + noenvopt: + /* "EIdvwWrKU" only */ + rb_raise(rb_eRuntimeError, "invalid switch in RUBYOPT: -%c", *s); break; - case 'I': - ary_unshift(rb_load_path, str_new2(optarg)); + noenvopt_long: + rb_raise(rb_eRuntimeError, "invalid switch in RUBYOPT: --%s", s); break; - default: + case 0: break; +# undef is_option_end +# undef check_envopt +# undef need_argument +# undef is_option_with_arg +# undef is_option_with_optarg } } - if (version) { - show_version(); - exit(0); + switch_end: + return argc0 - argc; +} + +static void +ruby_init_prelude(void) +{ + Init_prelude(); + rb_const_remove(rb_cObject, rb_intern_const("TMP_RUBY_PREFIX")); +} + +static int +opt_enc_index(VALUE enc_name) +{ + const char *s = RSTRING_PTR(enc_name); + int i = rb_enc_find_index(s); + + if (i < 0) { + rb_raise(rb_eRuntimeError, "unknown encoding name - %s", s); } - if (copyright) { - show_copyright(); + else if (rb_enc_dummy_p(rb_enc_from_index(i))) { + rb_raise(rb_eRuntimeError, "dummy encoding is not acceptable - %s ", s); } + return i; +} + +#define rb_progname (GET_VM()->progname) +#define rb_orig_progname (GET_VM()->orig_progname) +VALUE rb_argv0; + +static VALUE +false_value(void) +{ + return Qfalse; +} + +static VALUE +true_value(void) +{ + return Qtrue; +} + +#define rb_define_readonly_boolean(name, val) \ + rb_define_virtual_variable((name), (val) ? true_value : false_value, 0) + +static VALUE +uscore_get(void) +{ + VALUE line; + + line = rb_lastline_get(); + 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)); + } + return line; +} - if (argv[0] == Qnil) return; +/* + * call-seq: + * sub(pattern, replacement) -> $_ + * sub(pattern) {|...| block } -> $_ + * + * Equivalent to <code>$_.sub(<i>args</i>)</code>, except that + * <code>$_</code> will be updated if substitution occurs. + * Available only when -p/-n command line option specified. + */ - if (script_given == 0) { - if (argc == optind) { /* no more args */ - if (verbose) exit(0); - script = "-"; - load_stdin(); +static VALUE +rb_f_sub(int argc, VALUE *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) -> $_ + * gsub(pattern) {|...| block } -> $_ + * + * Equivalent to <code>$_.gsub...</code>, except that <code>$_</code> + * will be updated if substitution occurs. + * Available only when -p/-n command line option specified. + * + */ + +static VALUE +rb_f_gsub(int argc, VALUE *argv) +{ + VALUE str = rb_funcall_passing_block(uscore_get(), rb_intern("gsub"), argc, argv); + rb_lastline_set(str); + return str; +} + +/* + * call-seq: + * chop -> $_ + * + * Equivalent to <code>($_.dup).chop!</code>, except <code>nil</code> + * is never returned. See <code>String#chop!</code>. + * Available only when -p/-n command line option specified. + * + */ + +static VALUE +rb_f_chop(void) +{ + VALUE str = rb_funcall_passing_block(uscore_get(), rb_intern("chop"), 0, 0); + rb_lastline_set(str); + return str; +} + + +/* + * call-seq: + * chomp -> $_ + * chomp(string) -> $_ + * + * Equivalent to <code>$_ = $_.chomp(<em>string</em>)</code>. See + * <code>String#chomp</code>. + * Available only when -p/-n command line option specified. + * + */ + +static VALUE +rb_f_chomp(int argc, VALUE *argv) +{ + VALUE str = rb_funcall_passing_block(uscore_get(), rb_intern("chomp"), argc, argv); + rb_lastline_set(str); + return str; +} + +static VALUE +process_options(int argc, char **argv, struct cmdline_options *opt) +{ + NODE *tree = 0; + VALUE parser; + 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(); + VALUE toplevel_binding = Qundef; + + argc -= i; + argv += i; + + if (opt->dump & (DUMP_BIT(usage)|DUMP_BIT(help))) { + usage(origarg.argv[0], (opt->dump & DUMP_BIT(help))); + return Qtrue; + } + + 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; + VALUE int_enc_name = opt->intern.enc.name; + + opt->src.enc.name = opt->ext.enc.name = opt->intern.enc.name = 0; + moreswitches(s, opt, 1); + if (src_enc_name) + opt->src.enc.name = src_enc_name; + if (ext_enc_name) + opt->ext.enc.name = ext_enc_name; + if (int_enc_name) + opt->intern.enc.name = int_enc_name; + } + + if (opt->src.enc.name) + rb_warning("-K is specified; it is for 1.8 compatibility and may cause odd behavior"); + + if (opt->dump & (DUMP_BIT(version) | DUMP_BIT(version_v))) { + ruby_show_version(); + if (opt->dump & DUMP_BIT(version)) return Qtrue; + } + if (opt->dump & DUMP_BIT(copyright)) { + ruby_show_copyright(); + } + + if (!opt->e_script) { + if (argc == 0) { /* no more args */ + if (opt->verbose) + return Qtrue; + opt->script = "-"; } else { - script = argv[optind]; - if (do_search) { - script = dln_find_file(script, getenv("PATH")); - if (!script) script = argv[optind]; + opt->script = argv[0]; + if (!opt->script || opt->script[0] == '\0') { + opt->script = "-"; + } + else if (opt->do_search) { + char *path = getenv("RUBYPATH"); + + opt->script = 0; + if (path) { + opt->script = dln_find_file_r(argv[0], path, fbuf, sizeof(fbuf)); + } + if (!opt->script) { + opt->script = dln_find_file_r(argv[0], getenv(PATH_ENV), fbuf, sizeof(fbuf)); + } + if (!opt->script) + opt->script = argv[0]; } - rb_load_file(script); - optind++; + argc--; + argv++; } } - xflag = FALSE; - *argvp += optind; - *argcp -= optind; + opt->script_name = rb_str_new_cstr(opt->script); + opt->script = RSTRING_PTR(opt->script_name); - if (sflag) { - char *s; +#if _WIN32 + translit_char_bin(RSTRING_PTR(opt->script_name), '\\', '/'); +#elif defined DOSISH + translit_char(RSTRING_PTR(opt->script_name), '\\', '/'); +#endif - argc = *argcp; argv = *argvp; - for (; argc > 0 && argv[0][0] == '-'; argc--,argv++) { - if (argv[0][1] == '-') { - argc--,argv++; - break; + ruby_gc_set_params(opt->safe_level); + ruby_init_loadpath_safe(opt->safe_level); + 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); + } + if (opt->ext.enc.name != 0) { + opt->ext.enc.index = opt_enc_index(opt->ext.enc.name); + } + if (opt->intern.enc.name != 0) { + opt->intern.enc.index = opt_enc_index(opt->intern.enc.name); + } + if (opt->src.enc.name != 0) { + opt->src.enc.index = opt_enc_index(opt->src.enc.name); + src_encoding_index = opt->src.enc.index; + } + if (opt->ext.enc.index >= 0) { + enc = rb_enc_from_index(opt->ext.enc.index); + } + else { + enc = lenc; + } + rb_enc_set_default_external(rb_enc_from_encoding(enc)); + if (opt->intern.enc.index >= 0) { + 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); + rb_obj_freeze(opt->script_name); + { + long i; + VALUE load_path = GET_VM()->load_path; + const ID id_initial_load_path_mark = INITIAL_LOAD_PATH_MARK; + for (i = 0; i < RARRAY_LEN(load_path); ++i) { + VALUE path = RARRAY_AREF(load_path, i); + int mark = rb_attr_get(path, id_initial_load_path_mark) == path; + path = rb_enc_associate(rb_str_dup(path), lenc); + if (mark) rb_ivar_set(path, id_initial_load_path_mark, path); + RARRAY_ASET(load_path, i, 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_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 + } +#if UTF8_PATH + uenc = rb_utf8_encoding(); + if (uenc != lenc) { + opt->script_name = str_conv_enc(opt->script_name, uenc, lenc); + opt->script = RSTRING_PTR(opt->script_name); + } +#endif + ruby_set_argv(argc, argv); + process_sflag(&opt->sflag); + + toplevel_binding = rb_const_get(rb_cObject, rb_intern("TOPLEVEL_BINDING")); + +#define PREPARE_PARSE_MAIN(expr) do { \ + rb_env_t *env = toplevel_context(toplevel_binding); \ + th->parse_in_eval--; \ + th->base_block = &env->block; \ + expr; \ + th->parse_in_eval++; \ + th->base_block = 0; \ +} while (0) + + if (opt->e_script) { + VALUE progname = rb_progname; + rb_encoding *eenc; + if (opt->src.enc.index >= 0) { + eenc = rb_enc_from_index(opt->src.enc.index); + } + 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); + if (!(opt->dump & ~DUMP_BIT(version_v))) { + ruby_set_script_name(opt->script_name); + require_libraries(&opt->req_list); + } + ruby_set_script_name(progname); + + PREPARE_PARSE_MAIN({ + tree = 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_name, 1, opt); + }); + } + ruby_set_script_name(opt->script_name); + if (opt->dump & DUMP_BIT(yydebug)) return Qtrue; + + if (opt->ext.enc.index >= 0) { + enc = rb_enc_from_index(opt->ext.enc.index); + } + else { + enc = lenc; + } + rb_enc_set_default_external(rb_enc_from_encoding(enc)); + if (opt->intern.enc.index >= 0) { + /* Set in the shebang line */ + enc = rb_enc_from_index(opt->intern.enc.index); + rb_enc_set_default_internal(rb_enc_from_encoding(enc)); + } + else if (!rb_default_internal_encoding()) + /* Freeze default_internal */ + rb_enc_set_default_internal(Qnil); + rb_stdio_set_default_encoding(); + + if (!tree) return Qfalse; + + process_sflag(&opt->sflag); + opt->xflag = 0; + + if (opt->dump & DUMP_BIT(syntax)) { + printf("Syntax OK\n"); + 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))); + rb_io_flush(rb_stdout); + 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->dump & DUMP_BIT(insns)) { + rb_io_write(rb_stdout, rb_iseq_disasm((const rb_iseq_t *)iseq)); + rb_io_flush(rb_stdout); + 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); + + rb_set_safe_level(opt->safe_level); + + return (VALUE)iseq; +} + +struct load_file_arg { + VALUE parser; + VALUE fname; + int script; + int xflag; + struct cmdline_options *opt; + VALUE f; + VALUE lineno; +}; + +static VALUE +load_file_internal(VALUE argp_v) +{ + struct load_file_arg *argp = (struct load_file_arg *)argp_v; + VALUE parser = argp->parser; + VALUE orig_fname = argp->fname; + int script = argp->script; + struct cmdline_options *opt = argp->opt; + VALUE f = argp->f; + int line_start = 1; + NODE *tree = 0; + rb_encoding *enc; + ID set_encoding; + int xflag = argp->xflag; + + argp->script = 0; + CONST_ID(set_encoding, "set_encoding"); + if (script) { + VALUE c = 1; /* something not nil */ + VALUE line; + char *p; + int no_src_enc = !opt->src.enc.name; + int no_ext_enc = !opt->ext.enc.name; + int no_int_enc = !opt->intern.enc.name; + + enc = rb_ascii8bit_encoding(); + rb_funcall(f, set_encoding, 1, rb_enc_from_encoding(enc)); + + if (xflag || 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_engine)) != 0) { + goto start_read; + } + } } - argv[0][0] = '$'; - if (s = strchr(argv[0], '=')) { - *s++ = '\0'; - rb_gvar_set2((*argvp)[0], str_new2(s)); + 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_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'; + if ((p = strstr(p, " -")) != 0) { + moreswitches(p + 1, opt, 0); + } + + /* push back shebang for pragma may exist in next line */ + rb_io_ungetbyte(f, rb_str_new2("!\n")); } - else { - rb_gvar_set2((*argvp)[0], TRUE); + else if (!NIL_P(c)) { + rb_io_ungetbyte(f, c); + } + rb_io_ungetbyte(f, INT2FIX('#')); + if (no_src_enc && opt->src.enc.name) { + opt->src.enc.index = opt_enc_index(opt->src.enc.name); + src_encoding_index = opt->src.enc.index; + } + if (no_ext_enc && opt->ext.enc.name) { + opt->ext.enc.index = opt_enc_index(opt->ext.enc.name); } + if (no_int_enc && opt->intern.enc.name) { + opt->intern.enc.index = opt_enc_index(opt->intern.enc.name); + } + } + else if (!NIL_P(c)) { + rb_io_ungetbyte(f, c); + } + else { + if (f != rb_stdin) rb_io_close(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 */ } - *argcp = argc; *argvp = argv; } + if (opt->src.enc.index >= 0) { + enc = rb_enc_from_index(opt->src.enc.index); + } + else if (f == rb_stdin) { + enc = rb_locale_encoding(); + } + else { + enc = rb_utf8_encoding(); + } + if (NIL_P(f)) { + f = rb_str_new(0, 0); + rb_enc_associate(f, enc); + 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_path(parser, orig_fname, f, line_start); + rb_funcall(f, set_encoding, 1, rb_parser_encoding(parser)); + if (script && rb_parser_end_seen_p(parser)) argp->script = script; + return (VALUE)tree; } -static void -readin(fd, fname) - int fd; - char *fname; -{ - struct stat st; - char *ptr, *p, *pend; - - if (fstat(fd, &st) < 0) rb_sys_fail(fname); - if (!S_ISREG(st.st_mode)) - Fail("script is not a regular file - %s", fname); - - p = ptr = ALLOC_N(char, st.st_size+1); - if (read(fd, ptr, st.st_size) != st.st_size) { - free(ptr); - rb_sys_fail(fname); - } - pend = p + st.st_size; - if (xflag) { - char *s = p; - - *pend = '\0'; - xflag = FALSE; - while (p < pend) { - while (s < pend && *s != '\n') s++; - if (*s != '\n') break; - *s = '\0'; - if (p[0] == '#' && p[1] == '!' && strstr(p, "ruby")) { - if (p = strstr(p, "ruby -")) { - int argc; char *argv[2]; char **argvp = argv; - argc = 2; argv[0] = Qnil; argv[1] = p + 5; - proc_options(&argc, &argvp); - } - xflag = TRUE; - p = s + 1; - goto start_read; +static VALUE +open_load_file(VALUE fname_v, int *xflag) +{ + const char *fname = StringValueCStr(fname_v); + VALUE f; + int e; + + if (RSTRING_LEN(fname_v) == 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__ + { + const char *ext = strrchr(fname, '.'); + if (ext && STRCASECMP(ext, ".exe") == 0) { + mode |= O_BINARY; + *xflag = 1; } - p = s + 1; } - free(ptr); - Fail("No Ruby script found in input"); +#endif + + 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 + +#ifdef S_ISFIFO + { + struct stat st; + if (fstat(fd, &st) != 0) { + e = errno; + (void)close(fd); + rb_load_fail(fname_v, strerror(e)); + } + if (S_ISFIFO(st.st_mode)) { + /* + 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); + } + } +#endif + if (!ruby_is_fd_loadable(fd)) { + e = errno; + (void)close(fd); + rb_load_fail(fname_v, strerror(e)); + } + + f = rb_io_fdopen(fd, mode, fname); } - start_read: - lex_setsrc(fname, p, pend - p); - yyparse(); - free(ptr); + return f; } -void -rb_load_file(fname) - char *fname; +static VALUE +restore_load_file(VALUE arg) { - int fd; - char *ptr; - - if (fname[0] == '\0') { - load_stdin(); - return; + struct load_file_arg *argp = (struct load_file_arg *)arg; + VALUE f = argp->f; + VALUE lineno = argp->lineno; + + if (argp->script) { + /* + * 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); + } + else if (f != rb_stdin) { + rb_io_close(f); } + return rb_gv_set("$.", lineno); +} + +static NODE * +load_file(VALUE parser, VALUE fname, int script, struct cmdline_options *opt) +{ + struct load_file_arg arg; + arg.parser = parser; + arg.fname = fname; + arg.script = script; + arg.opt = opt; + arg.xflag = 0; + arg.lineno = rb_gv_get("$."); + arg.f = open_load_file(rb_str_encode_ospath(fname), &arg.xflag); + return (NODE *)rb_ensure(load_file_internal, (VALUE)&arg, + restore_load_file, (VALUE)&arg); +} + +void * +rb_load_file(const char *fname) +{ + VALUE fname_v = rb_str_new_cstr(fname); + return rb_load_file_str(fname_v); +} + +void * +rb_load_file_str(VALUE fname_v) +{ + struct cmdline_options opt; - fd = open(fname, O_RDONLY, 0); - if (fd < 0) rb_sys_fail(fname); - readin(fd, fname); - close(fd); + return load_file(rb_parser_new(), fname_v, 0, cmdline_options_init(&opt)); +} + +/* + * 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 VALUE +proc_argv0(VALUE process) +{ + return rb_orig_progname; +} + +/* + * 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) +{ + StringValue(title); + + setproctitle("%.*s", RSTRING_LENINT(title), RSTRING_PTR(title)); + + return title; } static void -load_stdin() +set_arg0(VALUE val, ID id) { - char buf[32]; - FILE *f; - char c; - int fd; + if (origarg.argv == 0) + rb_raise(rb_eRuntimeError, "$0 not initialized"); + + rb_progname = rb_str_new_frozen(proc_setproctitle(rb_mProcess, val)); +} + +static inline VALUE +external_str_new_cstr(const char *p) +{ +#if UTF8_PATH + VALUE str = rb_utf8_str_new_cstr(p); + return str_conv_enc(str, NULL, rb_default_external_encoding()); +#else + return rb_external_str_new_cstr(p); +#endif +} - sprintf(buf, "/tmp/ruby-f%d", getpid()); - f = fopen(buf, "w"); - fd = open(buf, O_RDONLY, 0); - if (fd < 0) rb_sys_fail(buf); - unlink(buf); - while ((c = getchar()) != EOF) { - putc(c, f); +/*! 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_orig_progname = rb_progname = external_str_new_cstr(name); + rb_vm_set_progname(rb_progname); } - fclose(f); +} - if (fd < 0) rb_sys_fail(buf); +/*! 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); +} - readin(fd, "-"); +static void +init_ids(struct cmdline_options *opt) +{ + rb_uid_t uid = getuid(); + rb_uid_t euid = geteuid(); + rb_gid_t gid = getgid(); + rb_gid_t egid = getegid(); + + if (uid != euid) opt->setids |= 1; + if (egid != gid) opt->setids |= 2; + if (uid && opt->setids) { + if (opt->safe_level < 1) opt->safe_level = 1; + } } -static VALUE Progname; -VALUE Argv; +#undef forbid_setid +static void +forbid_setid(const char *s, struct cmdline_options *opt) +{ + if (opt->setids & 1) + rb_raise(rb_eSecurityError, "no %s allowed while running setuid", s); + if (opt->setids & 2) + rb_raise(rb_eSecurityError, "no %s allowed while running setgid", s); + if (opt->safe_level > 0) + rb_raise(rb_eSecurityError, "no %s allowed in tainted mode", s); +} -static int origargc; -static char **origargv, **origenvp; +static void +verbose_setter(VALUE val, ID id, void *data) +{ + VALUE *variable = data; + *variable = RTEST(val) ? Qtrue : val; +} static VALUE -set_arg0(val, id) - VALUE val; - ID id; +opt_W_getter(ID id, void *data) { - char *s; - int i; - static int len; - - Check_Type(val, T_STRING); - if (len == 0) { - s = origargv[0]; - s += strlen(s); - /* See if all the arguments are contiguous in memory */ - for (i = 1; i < origargc; i++) { - if (origargv[i] == s + 1) - s += strlen(++s); /* this one is ok too */ - } - len = s - origargv[0]; - } - s = RSTRING(val)->ptr; - i = RSTRING(val)->len; - if (i > len) { - memcpy(origargv[0], s, len); - origargv[0][len] = '\0'; + VALUE *variable = data; + switch (*variable) { + case Qnil: + return INT2FIX(0); + case Qfalse: + return INT2FIX(1); + case Qtrue: + return INT2FIX(2); + default: + return Qnil; } - else { - memcpy(origargv[0], s, i); - s = origargv[0]+i; - *s++ = '\0'; - while (++i < len) - *s++ = ' '; - } - Progname = str_new2(origargv[0]); - - return val; } +/*! Defines built-in variables */ void -ruby_script(name) - char *name; +ruby_prog_init(void) { - if (name) { - Progname = str_new2(name); - } + rb_define_hooked_variable("$VERBOSE", &ruby_verbose, 0, verbose_setter); + rb_define_hooked_variable("$-v", &ruby_verbose, 0, verbose_setter); + rb_define_hooked_variable("$-w", &ruby_verbose, 0, verbose_setter); + rb_define_hooked_variable("$-W", &ruby_verbose, opt_W_getter, rb_gvar_readonly_setter); + rb_define_variable("$DEBUG", &ruby_debug); + rb_define_variable("$-d", &ruby_debug); + + 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); } void -ruby_init0(argc, argv, envp) - int argc; - char **argv, **envp; +ruby_set_argv(int argc, char **argv) { - extern VALUE errat; int i; + VALUE av = rb_argv; - origargc = argc; origargv = argv; origenvp = envp; +#if defined(USE_DLN_A_OUT) + if (origarg.argv) + dln_argv0 = origarg.argv[0]; + else + dln_argv0 = argv[0]; +#endif + rb_ary_clear(av); + for (i = 0; i < argc; i++) { + VALUE arg = external_str_new_cstr(argv[i]); - rb_call_inits(); + OBJ_FREEZE(arg); + rb_ary_push(av, arg); + } +} - rb_define_variable("$@", &errat, Qnil, Qnil, 0); - errat = str_new2(argv[0]); - rb_define_variable("$VERBOSE", &verbose, Qnil, Qnil, 0); - rb_define_variable("$DEBUG", &debug, Qnil, Qnil, 0); +void * +ruby_process_options(int argc, char **argv) +{ + struct cmdline_options opt; + VALUE iseq; + const char *script_name = (argc > 0 && argv[0]) ? argv[0] : ruby_engine; -#ifdef USE_DLN - rb_dln_argv0 = argv[0]; + 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 - proc_options(&argc, &argv); - if (do_check && nerrs == 0) { - printf("Syntax OK\n"); - exit(0); - } - if (do_print) { - yyappend_print(); + 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 (do_loop) { - yywhole_loop(do_line, do_split); + 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]); + } } +} - rb_define_variable("$0", &Progname, Qnil, set_arg0, 0); - ruby_script(script); - - rb_define_variable("$ARGV", &Argv, Qnil, Qnil, 0); - rb_define_variable("$*", &Argv, Qnil, Qnil, 0); - Argv = ary_new2(argc); - for (i=0; i < argc; i++) { - ary_push(Argv, str_new2(argv[i])); - } +/*! Initializes the process for ruby(1). + * + * This function assumes this process is ruby(1) and it has just started. + * Usually programs that embeds CRuby interpreter should not call this function, + * and should do their own initialization. + */ +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(USE_DLN_A_OUT) + dln_argv0 = origarg.argv[0]; +#endif + fill_standard_fds(); } |
