X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/56c47203421107ee019a20f3b56cfb967ad75bd2..c0cac722f7e5d7498528a67a37477cdf9ed1745c:/src/scan-gram.l diff --git a/src/scan-gram.l b/src/scan-gram.l index 15d75433..ef3976af 100644 --- a/src/scan-gram.l +++ b/src/scan-gram.l @@ -34,14 +34,14 @@ #define YY_USER_INIT \ do { \ LOCATION_RESET (*yylloc); \ + yylloc->file = infile; \ /* This is only to avoid GCC warnings. */ \ if (yycontrol) {;}; \ } while (0) -#define YY_USER_ACTION LOCATION_COLUMNS (*yylloc, yyleng) -#define YY_LINES LOCATION_LINES (*yylloc, yyleng); lineno += yyleng; -#define YY_STEP LOCATION_STEP (*yylloc) - +#define YY_USER_ACTION LOCATION_COLUMNS (*yylloc, yyleng); +#define YY_LINES LOCATION_LINES (*yylloc, yyleng); +#define YY_STEP LOCATION_STEP (*yylloc); /* STRING_OBSTACK -- Used to store all the characters that we need to keep (to construct ID, STRINGS etc.). Use the following macros to @@ -76,12 +76,13 @@ scanner_last_string_free (void) } - static int braces_level = 0; static int percent_percent_count = 0; -static void handle_dollar PARAMS ((char *cp, location_t location)); -static void handle_at PARAMS ((char *cp)); +static void handle_dollar PARAMS ((braced_code_t code_kind, + char *cp, location_t location)); +static void handle_at PARAMS ((braced_code_t code_kind, + char *cp, location_t location)); %} %x SC_COMMENT @@ -89,7 +90,7 @@ static void handle_at PARAMS ((char *cp)); %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER %x SC_BRACED_CODE SC_PROLOGUE SC_EPILOGUE -id [.a-zA-Z][.a-zA-Z_0-9]* +id [.a-zA-Z_][.a-zA-Z_0-9]* int [0-9]+ eols (\n|\r|\n\r|\r\n)+ blanks [ \t\f]+ @@ -122,18 +123,23 @@ blanks [ \t\f]+ "%debug" return PERCENT_DEBUG; "%define" return PERCENT_DEFINE; "%defines" return PERCENT_DEFINES; + "%destructor" return PERCENT_DESTRUCTOR; + "%dprec" return PERCENT_DPREC; "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE; "%expect" return PERCENT_EXPECT; "%file-prefix" return PERCENT_FILE_PREFIX; "%fixed"[-_]"output"[-_]"files" return PERCENT_YACC; + "%glr"[-_]"parser" return PERCENT_GLR_PARSER; "%left" return PERCENT_LEFT; "%locations" return PERCENT_LOCATIONS; + "%merge" return PERCENT_MERGE; "%name"[-_]"prefix" return PERCENT_NAME_PREFIX; "%no"[-_]"lines" return PERCENT_NO_LINES; "%nonassoc" return PERCENT_NONASSOC; "%nterm" return PERCENT_NTERM; "%output" return PERCENT_OUTPUT; "%prec" return PERCENT_PREC; + "%printer" return PERCENT_PRINTER; "%pure"[-_]"parser" return PERCENT_PURE_PARSER; "%right" return PERCENT_RIGHT; "%skeleton" return PERCENT_SKELETON; @@ -154,17 +160,17 @@ blanks [ \t\f]+ {eols} YY_LINES; YY_STEP; {blanks} YY_STEP; {id} { - yylval->symbol = getsym (yytext, *yylloc); + yylval->symbol = symbol_get (yytext, *yylloc); return ID; } {int} yylval->integer = strtol (yytext, 0, 10); return INT; /* Characters. We don't check there is only one. */ - \' YY_OBS_GROW; yy_push_state (SC_ESCAPED_CHARACTER); + "'" YY_OBS_GROW; yy_push_state (SC_ESCAPED_CHARACTER); /* Strings. */ - \" YY_OBS_GROW; yy_push_state (SC_ESCAPED_STRING); + "\"" YY_OBS_GROW; yy_push_state (SC_ESCAPED_STRING); /* Comments. */ "/*" yy_push_state (SC_COMMENT); @@ -193,7 +199,7 @@ blanks [ \t\f]+ . { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": invalid character: `%c'\n", *yytext); + fprintf (stderr, _(": invalid character: `%c'\n"), *yytext); YY_STEP; } } @@ -237,7 +243,7 @@ blanks [ \t\f]+ <> { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a comment\n"); + fprintf (stderr, _(": unexpected end of file in a comment\n")); yy_pop_state (); } } @@ -250,7 +256,7 @@ blanks [ \t\f]+ { - \" { + "\"" { assert (yy_top_state () == INITIAL); YY_OBS_GROW; YY_OBS_FINISH; @@ -265,7 +271,7 @@ blanks [ \t\f]+ <> { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a string\n"); + fprintf (stderr, _(": unexpected end of file in a string\n")); assert (yy_top_state () == INITIAL); YY_OBS_FINISH; yylval->string = last_string; @@ -281,27 +287,28 @@ blanks [ \t\f]+ { - \' { + "'" { YY_OBS_GROW; assert (yy_top_state () == INITIAL); { YY_OBS_FINISH; - yylval->symbol = getsym (last_string, *yylloc); - symbol_class_set (yylval->symbol, token_sym); - symbol_user_token_number_set (yylval->symbol, last_string[1]); + yylval->symbol = symbol_get (last_string, *yylloc); + symbol_class_set (yylval->symbol, token_sym, *yylloc); + symbol_user_token_number_set (yylval->symbol, + (unsigned char) last_string[1], *yylloc); YY_OBS_FREE; yy_pop_state (); return ID; } } - [^\'\n\r\\] YY_OBS_GROW; + [^\n\r\\] YY_OBS_GROW; {eols} obstack_1grow (&string_obstack, '\n'); YY_LINES; <> { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a character\n"); + fprintf (stderr, _(": unexpected end of file in a character\n")); assert (yy_top_state () == INITIAL); YY_OBS_FINISH; yylval->string = last_string; @@ -322,7 +329,7 @@ blanks [ \t\f]+ if (c > 255) { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": invalid escape: %s\n", yytext); + fprintf (stderr, _(": invalid escape: %s\n"), quote (yytext)); YY_STEP; } else @@ -341,11 +348,13 @@ blanks [ \t\f]+ \\t obstack_1grow (&string_obstack, '\t'); \\v obstack_1grow (&string_obstack, '\v'); \\[\\""] obstack_1grow (&string_obstack, yytext[1]); - \\. { + \\(.|\n) { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unrecognized escape: %s\n", yytext); + fprintf (stderr, _(": unrecognized escape: %s\n"), quote (yytext)); YY_OBS_GROW; } + /* FLex wants this rule, in case of a `\<>'. */ + \\ YY_OBS_GROW; } @@ -356,20 +365,22 @@ blanks [ \t\f]+ { - \' { + "'" { YY_OBS_GROW; assert (yy_top_state () != INITIAL); yy_pop_state (); } - [^\[\]\'\n\r\\] YY_OBS_GROW; - \\. YY_OBS_GROW; + [^\[\]\'\n\r\\]+ YY_OBS_GROW; + \\(.|\n) YY_OBS_GROW; + /* FLex wants this rule, in case of a `\<>'. */ + \\ YY_OBS_GROW; {eols} YY_OBS_GROW; YY_LINES; <> { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a character\n"); + fprintf (stderr, _(": unexpected end of file in a character\n")); assert (yy_top_state () != INITIAL); yy_pop_state (); } @@ -383,20 +394,22 @@ blanks [ \t\f]+ { - \" { + "\"" { assert (yy_top_state () != INITIAL); YY_OBS_GROW; yy_pop_state (); } [^\[\]\"\n\r\\]+ YY_OBS_GROW; - \\. YY_OBS_GROW; + \\(.|\n) YY_OBS_GROW; + /* FLex wants this rule, in case of a `\<>'. */ + \\ YY_OBS_GROW; {eols} YY_OBS_GROW; YY_LINES; <> { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a string\n"); + fprintf (stderr, _(": unexpected end of file in a string\n")); assert (yy_top_state () != INITIAL); yy_pop_state (); } @@ -410,14 +423,17 @@ blanks [ \t\f]+ { /* Characters. We don't check there is only one. */ - \' YY_OBS_GROW; yy_push_state (SC_CHARACTER); + "'" YY_OBS_GROW; yy_push_state (SC_CHARACTER); /* Strings. */ - \" YY_OBS_GROW; yy_push_state (SC_STRING); + "\"" YY_OBS_GROW; yy_push_state (SC_STRING); /* Comments. */ "/*" YY_OBS_GROW; yy_push_state (SC_COMMENT); "//".* YY_OBS_GROW; + + /* Not comments. */ + "/" YY_OBS_GROW; } @@ -441,8 +457,10 @@ blanks [ \t\f]+ "{" YY_OBS_GROW; braces_level++; - "$"("<"[^>]+">")?(-?[0-9]+|"$") { handle_dollar (yytext, *yylloc); } - "@"(-?[0-9]+|"$") { handle_at (yytext); } + "$"("<"[^>]+">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code, + yytext, *yylloc); } + "@"(-?[0-9]+|"$") { handle_at (current_braced_code, + yytext, *yylloc); } [^$@\[\]/\'\"\{\}\n\r]+ YY_OBS_GROW; {eols} YY_OBS_GROW; YY_LINES; @@ -452,7 +470,7 @@ blanks [ \t\f]+ <> { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a braced code\n"); + fprintf (stderr, _(": unexpected end of file in a braced code\n")); yy_pop_state (); YY_OBS_FINISH; yylval->string = last_string; @@ -475,19 +493,18 @@ blanks [ \t\f]+ return PROLOGUE; } - [^%\[\]/\'\"\n\r]+ YY_OBS_GROW; - "%"+[^%\}\n\r]+ YY_OBS_GROW; - {eols} YY_OBS_GROW; YY_LINES; + [^%\[\]/\'\"\n\r]+ YY_OBS_GROW; + "%" YY_OBS_GROW; + {eols} YY_OBS_GROW; YY_LINES; <> { LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a prologue\n"); + fprintf (stderr, _(": unexpected end of file in a prologue\n")); yy_pop_state (); YY_OBS_FINISH; yylval->string = last_string; return PROLOGUE; } - } @@ -512,28 +529,18 @@ blanks [ \t\f]+ %% /*------------------------------------------------------------------. -| CP is pointing to a wannabee semantic value (i.e., a `$'). | +| TEXT is pointing to a wannabee semantic value (i.e., a `$'). | | | | Possible inputs: $[]($|integer) | | | | Output to the STRING_OBSTACK a reference to this semantic value. | `------------------------------------------------------------------*/ -static void -handle_dollar (char *cp, location_t location) +static inline void +handle_action_dollar (char *text, location_t location) { const char *type_name = NULL; - - /* RULE_LENGTH is the number of values in the current rule so far, - which says where to find `$0' with respect to the top of the - stack. It is not the same as the rule->length in the case of mid - rule actions. */ - int rule_length = 0; - symbol_list_t *rhs; - for (rhs = current_rule->next; rhs; rhs = rhs->next) - ++rule_length; - - ++cp; + char *cp = text + 1; /* Get the type name if explicit. */ if (*cp == '<') @@ -557,8 +564,13 @@ handle_dollar (char *cp, location_t location) obstack_fgrow1 (&string_obstack, "]b4_lhs_value([%s])[", type_name); } - else if (isdigit (*cp) || *cp == '-') + else if (('0' <= *cp && *cp <= '9') || *cp == '-') { + /* RULE_LENGTH is the number of values in the current rule so + far, which says where to find `$0' with respect to the top of + the stack. It is not the same as the rule->length in the + case of mid rule actions. */ + int rule_length = symbol_list_length (current_rule->next); int n = strtol (cp, &cp, 10); if (n > rule_length) @@ -580,53 +592,130 @@ handle_dollar (char *cp, location_t location) } else { - char buf[] = "$c"; - buf[1] = *cp; - complain (_("%s is invalid"), quote (buf)); + complain_at (location, _("%s is invalid"), quote (text)); } } -/*-------------------------------------------------------. -| CP is pointing to a location (i.e., a `@'). Output to | -| STRING_OBSTACK a reference to this location. | -`-------------------------------------------------------*/ + +/*---------------------------------------------------------------. +| TEXT is expexted tp be $$ in some code associated to a symbol: | +| destructor or printer. | +`---------------------------------------------------------------*/ + +static inline void +handle_symbol_code_dollar (char *text, location_t location) +{ + char *cp = text + 1; + if (*cp == '$') + obstack_sgrow (&string_obstack, "]b4_dollar_dollar["); + else + complain_at (location, _("%s is invalid"), quote (text)); +} + + +/*-----------------------------------------------------------------. +| Dispatch onto handle_action_dollar, or handle_destructor_dollar, | +| depending upon CODE_KIND. | +`-----------------------------------------------------------------*/ static void -handle_at (char *cp) +handle_dollar (braced_code_t braced_code_kind, + char *text, location_t location) { - /* RULE_LENGTH is the number of values in the current rule so far, - which says where to find `$0' with respect to the top of the - stack. It is not the same as the rule->length in the case of mid - rule actions. */ - int rule_length = 0; - symbol_list_t *rhs; - for (rhs = current_rule->next; rhs; rhs = rhs->next) - ++rule_length; + switch (braced_code_kind) + { + case action_braced_code: + handle_action_dollar (text, location); + break; + + case destructor_braced_code: + case printer_braced_code: + handle_symbol_code_dollar (text, location); + break; + } +} + + +/*------------------------------------------------------. +| TEXT is a location token (i.e., a `@...'). Output to | +| STRING_OBSTACK a reference to this location. | +`------------------------------------------------------*/ +static inline void +handle_action_at (char *text, location_t location) +{ + char *cp = text + 1; locations_flag = 1; - ++cp; if (*cp == '$') { obstack_sgrow (&string_obstack, "]b4_lhs_location["); } - else if (isdigit (*cp) || *cp == '-') + else if (('0' <= *cp && *cp <= '9') || *cp == '-') { + /* RULE_LENGTH is the number of values in the current rule so + far, which says where to find `$0' with respect to the top of + the stack. It is not the same as the rule->length in the + case of mid rule actions. */ + int rule_length = symbol_list_length (current_rule->next); int n = strtol (cp, &cp, 10); + if (n > rule_length) - complain (_("invalid value: %s%d"), "@", n); + complain_at (location, _("invalid value: %s%d"), "@", n); else obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[", rule_length, n); } else { - char buf[] = "@c"; - buf[1] = *cp; - complain (_("%s is invalid"), quote (buf)); + complain_at (location, _("%s is invalid"), quote (text)); + } +} + + +/*---------------------------------------------------------------. +| TEXT is expexted tp be @$ in some code associated to a symbol: | +| destructor or printer. | +`---------------------------------------------------------------*/ + +static inline void +handle_symbol_code_at (char *text, location_t location) +{ + char *cp = text + 1; + if (*cp == '$') + obstack_sgrow (&string_obstack, "]b4_at_dollar["); + else + complain_at (location, _("%s is invalid"), quote (text)); +} + + +/*-------------------------------------------------------------------. +| Dispatch onto handle_action_at, or handle_destructor_at, depending | +| upon CODE_KIND. | +`-------------------------------------------------------------------*/ + +static void +handle_at (braced_code_t braced_code_kind, + char *text, location_t location) +{ + switch (braced_code_kind) + { + case action_braced_code: + handle_action_at (text, location); + break; + + case destructor_braced_code: + case printer_braced_code: + handle_symbol_code_at (text, location); + break; } } + +/*-------------------------. +| Initialize the scanner. | +`-------------------------*/ + void scanner_initialize (void) { @@ -634,8 +723,14 @@ scanner_initialize (void) } +/*-----------------------------------------------. +| Free all the memory allocated to the scanner. | +`-----------------------------------------------*/ + void scanner_free (void) { obstack_free (&string_obstack, 0); + /* Reclaim Flex's buffers. */ + yy_delete_buffer (YY_CURRENT_BUFFER); }