X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/1d6412adebec68e299086248a6a92a4191bbbec9..a5d5099417229a48aa18477073266e9aa3b77a8e:/src/scan-gram.l diff --git a/src/scan-gram.l b/src/scan-gram.l index 4cc1b060..7c4b3487 100644 --- a/src/scan-gram.l +++ b/src/scan-gram.l @@ -31,11 +31,17 @@ #include "reader.h" /* Each time we match a string, move the end cursor to its end. */ +#define YY_USER_INIT \ +do { \ + LOCATION_RESET (*yylloc); \ + /* 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_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 use it. @@ -69,15 +75,13 @@ scanner_last_string_free (void) } -/* This is only to avoid GCC warnings. */ -#define YY_USER_INIT if (yycontrol) {;}; - - static int braces_level = 0; static int percent_percent_count = 0; -static void handle_dollar PARAMS ((char *cp)); -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 @@ -96,7 +100,7 @@ blanks [ \t\f]+ start of the next token. */ #define TR_POS 0 #if TR_POS - fprintf (stderr, "FOO1: "); + fprintf (stderr, "FOO1: %p: ", yylloc); LOCATION_PRINT (stderr, *yylloc); fprintf (stderr, "\n"); #endif @@ -118,18 +122,23 @@ 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; @@ -150,7 +159,7 @@ blanks [ \t\f]+ {eols} YY_LINES; YY_STEP; {blanks} YY_STEP; {id} { - yylval->symbol = getsym (yytext); + yylval->symbol = symbol_get (yytext, *yylloc); return ID; } @@ -282,9 +291,10 @@ blanks [ \t\f]+ assert (yy_top_state () == INITIAL); { YY_OBS_FINISH; - yylval->symbol = getsym (last_string); - 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; @@ -318,7 +328,7 @@ 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 @@ -337,11 +347,13 @@ 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; } @@ -358,8 +370,10 @@ blanks [ \t\f]+ 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; @@ -386,7 +400,9 @@ blanks [ \t\f]+ } [^\[\]\"\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; @@ -414,6 +430,9 @@ blanks [ \t\f]+ /* Comments. */ "/*" YY_OBS_GROW; yy_push_state (SC_COMMENT); "//".* YY_OBS_GROW; + + /* Not comments. */ + "/" YY_OBS_GROW; } @@ -437,10 +456,12 @@ blanks [ \t\f]+ "{" YY_OBS_GROW; braces_level++; - "$"("<".*">")?(-?[0-9]+|"$") { handle_dollar (yytext); } - "@"(-?[0-9]+|"$") { handle_at (yytext); } + "$"("<"[^>]+">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code, + yytext, *yylloc); } + "@"(-?[0-9]+|"$") { handle_at (current_braced_code, + yytext, *yylloc); } - [^\[\]$/\'\"@\{\}\n\r]+ YY_OBS_GROW; + [^$@\[\]/\'\"\{\}\n\r]+ YY_OBS_GROW; {eols} YY_OBS_GROW; YY_LINES; /* A lose $, or /, or etc. */ @@ -471,9 +492,9 @@ 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); @@ -483,7 +504,6 @@ blanks [ \t\f]+ yylval->string = last_string; return PROLOGUE; } - } @@ -508,28 +528,18 @@ 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) +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 `$0' with respect to the top of the - stack. It is not the same as the rule->length in the case of mid - rule actions. */ - int rule_length = 0; - symbol_list *rhs; - for (rhs = current_rule->next; rhs; rhs = rhs->next) - ++rule_length; - - ++cp; + char *cp = text + 1; /* Get the type name if explicit. */ if (*cp == '<') @@ -544,10 +554,10 @@ handle_dollar (char *cp) if (*cp == '$') { if (!type_name) - type_name = get_type_name (0, current_rule); + type_name = symbol_list_n_type_name_get (current_rule, location, 0); if (!type_name && typed) - complain (_("$$ of `%s' has no declared type"), - current_rule->sym->tag); + complain_at (location, _("$$ of `%s' has no declared type"), + current_rule->sym->tag); if (!type_name) type_name = ""; obstack_fgrow1 (&string_obstack, @@ -555,16 +565,22 @@ handle_dollar (char *cp) } else if (isdigit (*cp) || *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"), "$", n); else { if (!type_name && n > 0) - type_name = get_type_name (n, current_rule); + type_name = symbol_list_n_type_name_get (current_rule, location, + n); if (!type_name && typed) - complain (_("$%d of `%s' has no declared type"), + complain_at (location, _("$%d of `%s' has no declared type"), n, current_rule->sym->tag); if (!type_name) type_name = ""; @@ -575,31 +591,60 @@ handle_dollar (char *cp) } else { - char buf[] = "$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 `$0' with respect to the top of the - stack. It is not the same as the rule->length in the case of mid - rule actions. */ - int rule_length = 0; - symbol_list *rhs; - for (rhs = current_rule->next; rhs; rhs = rhs->next) - ++rule_length; + switch (braced_code_kind) + { + case action_braced_code: + handle_action_dollar (text, location); + break; + + case destructor_braced_code: + case printer_braced_code: + handle_symbol_code_dollar (text, location); + break; + } +} + +/*------------------------------------------------------. +| TEXT is a location token (i.e., a `@...'). Output to | +| STRING_OBSTACK a reference to this location. | +`------------------------------------------------------*/ + +static inline void +handle_action_at (char *text, location_t location) +{ + char *cp = text + 1; locations_flag = 1; - ++cp; if (*cp == '$') { @@ -607,21 +652,69 @@ handle_at (char *cp) } else if (isdigit (*cp) || *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) { @@ -629,8 +722,14 @@ 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); }