X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/ee000ba4fcd5b051bef4476341a7fdffbddf8ad3..ccdb39e0a6f95cb4a3f19c08885a7c421cdebedb:/src/scan-gram.l diff --git a/src/scan-gram.l b/src/scan-gram.l index cd8ca516..8f18c355 100644 --- a/src/scan-gram.l +++ b/src/scan-gram.l @@ -19,11 +19,12 @@ 02111-1307 USA */ -%option debug nodefault noyywrap nounput never-interactive stack +%option debug nodefault noyywrap never-interactive %option prefix="gram_" outfile="lex.yy.c" %{ #include "system.h" +#include "mbswidth.h" #include "complain.h" #include "quote.h" #include "getargs.h" @@ -34,14 +35,100 @@ #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_USER_ACTION extend_location (yylloc, yytext, yyleng); #define YY_STEP LOCATION_STEP (*yylloc) +#define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size)) + + +/* Read bytes from FP into buffer BUF of size SIZE. Return the + number of bytes read. Remove '\r' from input, treating \r\n + and isolated \r as \n. */ + +static size_t +no_cr_read (FILE *fp, char *buf, size_t size) +{ + size_t s = fread (buf, 1, size, fp); + if (s) + { + char *w = memchr (buf, '\r', s); + if (w) + { + char const *r = ++w; + char const *lim = buf + s; + + for (;;) + { + /* Found an '\r'. Treat it like '\n', but ignore any + '\n' that immediately follows. */ + w[-1] = '\n'; + if (r == lim) + { + int ch = getc (fp); + if (ch != '\n' && ungetc (ch, fp) != ch) + break; + } + else if (*r == '\n') + r++; + + /* Copy until the next '\r'. */ + do + { + if (r == lim) + return w - buf; + } + while ((*w++ = *r++) != '\r'); + } + + return w - buf; + } + } + + return s; +} + + +/* Extend *LOC to account for token TOKEN of size SIZE. */ + +static void +extend_location (location_t *loc, char const *token, int size) +{ + int line = loc->last_line; + int column = loc->last_column; + char const *p0 = token; + char const *p = token; + char const *lim = token + size; + + for (p = token; p < lim; p++) + switch (*p) + { + case '\r': + /* \r shouldn't survive no_cr_read. */ + abort (); + + case '\n': + line++; + column = 1; + p0 = p + 1; + break; + + case '\t': + column += mbsnwidth (p0, p - p0, 0); + column += 8 - ((column - 1) & 7); + p0 = p + 1; + break; + } + + loc->last_line = line; + loc->last_column = column + mbsnwidth (p0, p - p0, 0); +} + + /* STRING_OBSTACK -- Used to store all the characters that we need to keep (to construct ID, STRINGS etc.). Use the following macros to @@ -53,7 +140,6 @@ do { \ used, and which is used by YY_OBS_FREE to free the last string. */ static struct obstack string_obstack; -char *last_string; #define YY_OBS_GROW \ obstack_grow (&string_obstack, yytext, yyleng) @@ -64,52 +150,60 @@ char *last_string; last_string = obstack_finish (&string_obstack); \ } while (0) -#define YY_OBS_FREE \ - do { \ - obstack_free (&string_obstack, last_string); \ - } while (0) - -void -scanner_last_string_free (void) -{ - YY_OBS_FREE; -} +#define YY_OBS_FREE \ + obstack_free (&string_obstack, last_string) +/* Within well-formed rules, 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. -static int braces_level = 0; -static int percent_percent_count = 0; + Outside of well-formed rules, RULE_LENGTH has an undefined value. */ +static int rule_length; -static void handle_dollar PARAMS ((char *cp)); -static void handle_at PARAMS ((char *cp)); +static void handle_dollar (braced_code_t code_kind, + char *cp, location_t location); +static void handle_at (braced_code_t code_kind, + char *cp, location_t location); +static void handle_syncline (char *args, location_t *location); +static int convert_ucn_to_byte (char const *hex_text); +static void unexpected_end_of_file (location_t *, char const *); %} -%x SC_COMMENT +%x SC_COMMENT SC_LINE_COMMENT SC_YACC_COMMENT %x SC_STRING SC_CHARACTER %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER %x SC_BRACED_CODE SC_PROLOGUE SC_EPILOGUE -id [.a-zA-Z][.a-zA-Z_0-9]* -int [0-9]+ -eols (\n|\r|\n\r|\r\n)+ -blanks [ \t\f]+ +letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] +id {letter}({letter}|[0-9])* +directive %{letter}({letter}|[0-9]|-)* +int [0-9]+ + +/* POSIX says that a tag must be both an id and a C union member, but + historically almost any character is allowed in a tag. We disallow + NUL and newline, as this simplifies our implementation. */ +tag [^\0\n>]+ + +/* Zero or more instances of backslash-newline. Following GCC, allow + white space between the backslash and the newline. */ +splice (\\[ \f\t\v]*\n)* %% %{ + /* Nesting level of the current code in braces. */ + int braces_level IF_LINT (= 0); + + /* Scanner context when scanning C code. */ + int c_context IF_LINT (= 0); + + /* A string representing the most recently saved token. */ + char *last_string; + /* At each yylex invocation, mark the current position as the start of the next token. */ -#define TR_POS 0 -#if TR_POS - fprintf (stderr, "FOO1: %p: ", yylloc); - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, "\n"); -#endif YY_STEP; -#if TR_POS - fprintf (stderr, "BAR1: "); - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, "\n"); -#endif %} @@ -122,20 +216,27 @@ 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; + "%parse-param" return PERCENT_PARSE_PARAM; + "%prec" rule_length--; return PERCENT_PREC; + "%printer" return PERCENT_PRINTER; "%pure"[-_]"parser" return PERCENT_PURE_PARSER; "%right" return PERCENT_RIGHT; + "%lex-param" return PERCENT_LEX_PARAM; "%skeleton" return PERCENT_SKELETON; "%start" return PERCENT_START; "%term" return PERCENT_TOKEN; @@ -146,100 +247,114 @@ blanks [ \t\f]+ "%verbose" return PERCENT_VERBOSE; "%yacc" return PERCENT_YACC; + {directive} { + complain_at (*yylloc, _("invalid directive: %s"), quote (yytext)); + YY_STEP; + } + + ^"#line "{int}" \""[^\"]*"\"\n" handle_syncline (yytext + strlen ("#line "), yylloc); YY_STEP; + "=" return EQUAL; - ":" return COLON; - "|" return PIPE; + ":" rule_length = 0; return COLON; + "|" rule_length = 0; return PIPE; + "," return COMMA; ";" return SEMICOLON; - {eols} YY_LINES; YY_STEP; - {blanks} YY_STEP; + [ \f\n\t\v] YY_STEP; + {id} { - yylval->symbol = getsym (yytext, *yylloc); + yylval->symbol = symbol_get (yytext, *yylloc); + rule_length++; return ID; } - {int} yylval->integer = strtol (yytext, 0, 10); return INT; + {int} { + unsigned long num; + errno = 0; + num = strtoul (yytext, 0, 10); + if (INT_MAX < num || errno) + { + complain_at (*yylloc, _("integer out of range: %s"), quote (yytext)); + num = INT_MAX; + } + yylval->integer = num; + return INT; + } /* Characters. We don't check there is only one. */ - \' YY_OBS_GROW; yy_push_state (SC_ESCAPED_CHARACTER); + "'" YY_OBS_GROW; BEGIN SC_ESCAPED_CHARACTER; /* Strings. */ - \" YY_OBS_GROW; yy_push_state (SC_ESCAPED_STRING); + "\"" YY_OBS_GROW; BEGIN SC_ESCAPED_STRING; /* Comments. */ - "/*" yy_push_state (SC_COMMENT); + "/*" BEGIN SC_YACC_COMMENT; "//".* YY_STEP; /* Prologue. */ - "%{" yy_push_state (SC_PROLOGUE); + "%{" BEGIN SC_PROLOGUE; /* Code in between braces. */ - "{" YY_OBS_GROW; ++braces_level; yy_push_state (SC_BRACED_CODE); + "{" YY_OBS_GROW; braces_level = 0; BEGIN SC_BRACED_CODE; /* A type. */ - "<"[^>]+">" { + "<"{tag}">" { obstack_grow (&string_obstack, yytext + 1, yyleng - 2); YY_OBS_FINISH; yylval->string = last_string; return TYPE; } - - "%%" { + "%%" { + static int percent_percent_count; if (++percent_percent_count == 2) - yy_push_state (SC_EPILOGUE); + BEGIN SC_EPILOGUE; return PERCENT_PERCENT; } - . { - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": invalid character: `%c'\n", *yytext); + . { + complain_at (*yylloc, _("invalid character: %s"), quote (yytext)); YY_STEP; } } - /*------------------------------------------------------------. - | Whatever the start condition (but those which correspond to | - | entity `swallowed' by Bison: SC_ESCAPED_STRING and | - | SC_ESCAPED_CHARACTER), no M4 character must escape as is. | - `------------------------------------------------------------*/ + /*---------------------------------------------------------------. + | Scanning a Yacc comment. The initial `/ *' is already eaten. | + `---------------------------------------------------------------*/ - + { - \[ if (YY_START != SC_COMMENT) obstack_sgrow (&string_obstack, "@<:@"); - \] if (YY_START != SC_COMMENT) obstack_sgrow (&string_obstack, "@:>@"); -} + "*/" { + YY_STEP; + BEGIN INITIAL; + } + .|\n ; + <> unexpected_end_of_file (yylloc, "*/"); +} - /*-----------------------------------------------------------. - | Scanning a C comment. The initial `/ *' is already eaten. | - `-----------------------------------------------------------*/ + /*------------------------------------------------------------. + | Scanning a C comment. The initial `/ *' is already eaten. | + `------------------------------------------------------------*/ { - "*/" { /* End of the comment. */ - if (yy_top_state () == INITIAL) - { - YY_STEP; - } - else - { - YY_OBS_GROW; - } - yy_pop_state (); - } + "*"{splice}"/" YY_OBS_GROW; BEGIN c_context; + <> unexpected_end_of_file (yylloc, "*/"); +} - [^\[\]*\n\r]+ if (yy_top_state () != INITIAL) YY_OBS_GROW; - {eols} if (yy_top_state () != INITIAL) YY_OBS_GROW; YY_LINES; - . /* Stray `*'. */if (yy_top_state () != INITIAL) YY_OBS_GROW; - <> { - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a comment\n"); - yy_pop_state (); - } + /*--------------------------------------------------------------. + | Scanning a line comment. The initial `//' is already eaten. | + `--------------------------------------------------------------*/ + + +{ + "\n" YY_OBS_GROW; BEGIN c_context; + {splice} YY_OBS_GROW; + <> BEGIN c_context; } @@ -250,28 +365,17 @@ blanks [ \t\f]+ { - \" { - assert (yy_top_state () == INITIAL); + "\"" { YY_OBS_GROW; YY_OBS_FINISH; yylval->string = last_string; - yy_pop_state (); + rule_length++; + BEGIN INITIAL; return STRING; } - [^\"\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 string\n"); - assert (yy_top_state () == INITIAL); - YY_OBS_FINISH; - yylval->string = last_string; - yy_pop_state (); - return STRING; - } + .|\n YY_OBS_GROW; + <> unexpected_end_of_file (yylloc, "\""); } /*---------------------------------------------------------------. @@ -281,33 +385,21 @@ 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]); - YY_OBS_FREE; - yy_pop_state (); - return ID; - } - } - - [^\'\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"); - assert (yy_top_state () == INITIAL); YY_OBS_FINISH; - yylval->string = last_string; - yy_pop_state (); - return CHARACTER; + 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; + rule_length++; + BEGIN INITIAL; + return ID; } + + .|\n YY_OBS_GROW; + <> unexpected_end_of_file (yylloc, "'"); } @@ -317,20 +409,30 @@ blanks [ \t\f]+ { - \\[0-7]{3} { - long c = strtol (yytext + 1, 0, 8); - if (c > 255) + \\[0-7]{1,3} { + unsigned long c = strtoul (yytext + 1, 0, 8); + if (UCHAR_MAX < c) { - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": invalid escape: %s\n", yytext); + complain_at (*yylloc, _("invalid escape sequence: %s"), + quote (yytext)); YY_STEP; } else obstack_1grow (&string_obstack, c); } - \\x[0-9a-fA-F]{2} { - obstack_1grow (&string_obstack, strtol (yytext + 2, 0, 16)); + \\x[0-9a-fA-F]+ { + unsigned long c; + errno = 0; + c = strtoul (yytext + 2, 0, 16); + if (UCHAR_MAX < c || errno) + { + complain_at (*yylloc, _("invalid escape sequence: %s"), + quote (yytext)); + YY_STEP; + } + else + obstack_1grow (&string_obstack, c); } \\a obstack_1grow (&string_obstack, '\a'); @@ -340,10 +442,21 @@ blanks [ \t\f]+ \\r obstack_1grow (&string_obstack, '\r'); \\t obstack_1grow (&string_obstack, '\t'); \\v obstack_1grow (&string_obstack, '\v'); - \\[\\""] obstack_1grow (&string_obstack, yytext[1]); - \\. { - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unrecognized escape: %s\n", yytext); + \\[\"\'?\\] obstack_1grow (&string_obstack, yytext[1]); + \\(u|U[0-9a-fA-F]{4})[0-9a-fA-F]{4} { + int c = convert_ucn_to_byte (yytext); + if (c < 0) + { + complain_at (*yylloc, _("invalid escape sequence: %s"), + quote (yytext)); + YY_STEP; + } + else + obstack_1grow (&string_obstack, c); + } + \\(.|\n) { + complain_at (*yylloc, _("unrecognized escape sequence: %s"), + quote (yytext)); YY_OBS_GROW; } } @@ -356,23 +469,9 @@ blanks [ \t\f]+ { - \' { - YY_OBS_GROW; - assert (yy_top_state () != INITIAL); - yy_pop_state (); - } - - [^\[\]\'\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 character\n"); - assert (yy_top_state () != INITIAL); - yy_pop_state (); - } + "'" YY_OBS_GROW; BEGIN c_context; + \\{splice}[\'\\] YY_OBS_GROW; + <> unexpected_end_of_file (yylloc, "'"); } @@ -383,23 +482,9 @@ blanks [ \t\f]+ { - \" { - assert (yy_top_state () != INITIAL); - YY_OBS_GROW; - yy_pop_state (); - } - - [^\[\]\"\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 string\n"); - assert (yy_top_state () != INITIAL); - yy_pop_state (); - } + "\"" YY_OBS_GROW; BEGIN c_context; + \\{splice}[\"\\] YY_OBS_GROW; + <> unexpected_end_of_file (yylloc, "\""); } @@ -409,15 +494,10 @@ blanks [ \t\f]+ { - /* Characters. We don't check there is only one. */ - \' YY_OBS_GROW; yy_push_state (SC_CHARACTER); - - /* Strings. */ - \" YY_OBS_GROW; yy_push_state (SC_STRING); - - /* Comments. */ - "/*" YY_OBS_GROW; yy_push_state (SC_COMMENT); - "//".* YY_OBS_GROW; + "'" YY_OBS_GROW; c_context = YY_START; BEGIN SC_CHARACTER; + "\"" YY_OBS_GROW; c_context = YY_START; BEGIN SC_STRING; + "/"{splice}"*" YY_OBS_GROW; c_context = YY_START; BEGIN SC_COMMENT; + "/"{splice}"/" YY_OBS_GROW; c_context = YY_START; BEGIN SC_LINE_COMMENT; } @@ -428,37 +508,31 @@ blanks [ \t\f]+ { + "{"|"<"{splice}"%" YY_OBS_GROW; braces_level++; + "%"{splice}">" YY_OBS_GROW; braces_level--; "}" { YY_OBS_GROW; - if (--braces_level == 0) + braces_level--; + if (braces_level < 0) { - yy_pop_state (); YY_OBS_FINISH; yylval->string = last_string; + rule_length++; + BEGIN INITIAL; return BRACED_CODE; } } - "{" YY_OBS_GROW; braces_level++; - - "$"("<"[^>]+">")?(-?[0-9]+|"$") { handle_dollar (yytext); } - "@"(-?[0-9]+|"$") { handle_at (yytext); } + /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly + (as `<' `<%'). */ + "<"{splice}"<" YY_OBS_GROW; - [^$@\[\]/\'\"\{\}\n\r]+ YY_OBS_GROW; - {eols} YY_OBS_GROW; YY_LINES; - - /* A lose $, or /, or etc. */ - . YY_OBS_GROW; - - <> { - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a braced code\n"); - yy_pop_state (); - YY_OBS_FINISH; - yylval->string = last_string; - return BRACED_CODE; - } + "$"("<"{tag}">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code, + yytext, *yylloc); } + "@"(-?[0-9]+|"$") { handle_at (current_braced_code, + yytext, *yylloc); } + <> unexpected_end_of_file (yylloc, "}"); } @@ -469,71 +543,62 @@ blanks [ \t\f]+ { "%}" { - yy_pop_state (); - YY_OBS_FINISH; - yylval->string = last_string; - return PROLOGUE; - } - - [^%\[\]/\'\"\n\r]+ YY_OBS_GROW; - "%"+[^%\}\n\r]+ YY_OBS_GROW; - {eols} YY_OBS_GROW; YY_LINES; - - <> { - LOCATION_PRINT (stderr, *yylloc); - fprintf (stderr, ": unexpected end of file in a prologue\n"); - yy_pop_state (); YY_OBS_FINISH; yylval->string = last_string; + BEGIN INITIAL; return PROLOGUE; } + <> unexpected_end_of_file (yylloc, "%}"); } /*---------------------------------------------------------------. | Scanning the epilogue (everything after the second "%%", which | - | has already been eaten. | + | has already been eaten). | `---------------------------------------------------------------*/ { - ([^\[\]]|{eols})+ YY_OBS_GROW; - <> { - yy_pop_state (); YY_OBS_FINISH; yylval->string = last_string; + BEGIN INITIAL; return EPILOGUE; } } + /*----------------------------------------------------------------. + | By default, grow the string obstack with the input, escaping M4 | + | quoting characters. | + `----------------------------------------------------------------*/ + + +{ + \$ obstack_sgrow (&string_obstack, "$]["); + \@ obstack_sgrow (&string_obstack, "@@"); + \[ obstack_sgrow (&string_obstack, "@{"); + \] obstack_sgrow (&string_obstack, "@}"); + .|\n YY_OBS_GROW; +} + + %% /*------------------------------------------------------------------. -| 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 `$0' with respect to the top of the - stack. It is not the same as the rule->length in the case of mid - rule actions. */ - int rule_length = 0; - symbol_list *rhs; - for (rhs = current_rule->next; rhs; rhs = rhs->next) - ++rule_length; - - ++cp; + char *cp = text + 1; /* Get the type name if explicit. */ if (*cp == '<') @@ -548,27 +613,29 @@ handle_dollar (char *cp) if (*cp == '$') { 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, "]b4_lhs_value([%s])[", type_name); } - else if (isdigit (*cp) || *cp == '-') + else { - int n = strtol (cp, &cp, 10); + long num; + errno = 0; + num = strtol (cp, 0, 10); - if (n > rule_length) - complain (_("invalid value: %s%d"), "$", n); - else + if (INT_MIN <= num && num <= rule_length && ! errno) { + int n = num; if (!type_name && n > 0) - type_name = get_type_name (n, current_rule); + type_name = symbol_list_n_type_name_get (current_rule, location, + n); if (!type_name && typed) - complain (_("$%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 = ""; @@ -576,56 +643,224 @@ handle_dollar (char *cp) "]b4_rhs_value([%d], [%d], [%s])[", rule_length, n, type_name); } + else + complain_at (location, _("integer out of range: %s"), quote (text)); } +} + + +/*---------------------------------------------------------------. +| TEXT is expected to 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 == '$') + obstack_sgrow (&string_obstack, "]b4_dollar_dollar["); else - { - char buf[] = "$c"; - buf[1] = *cp; - complain (_("%s is invalid"), quote (buf)); - } + complain_at (location, _("invalid value: %s"), quote (text)); } -/*-------------------------------------------------------. -| CP is pointing to a location (i.e., a `@'). Output to | -| STRING_OBSTACK a reference to this location. | -`-------------------------------------------------------*/ + +/*-----------------------------------------------------------------. +| 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 `$0' with respect to the top of the - stack. It is not the same as the rule->length in the case of mid - rule actions. */ - int rule_length = 0; - symbol_list *rhs; - for (rhs = current_rule->next; rhs; rhs = rhs->next) - ++rule_length; + switch (braced_code_kind) + { + case action_braced_code: + handle_action_dollar (text, location); + break; + + case destructor_braced_code: + case printer_braced_code: + handle_symbol_code_dollar (text, location); + break; + } +} + + +/*------------------------------------------------------. +| TEXT is a location token (i.e., a `@...'). Output to | +| STRING_OBSTACK a reference to this location. | +`------------------------------------------------------*/ +static inline void +handle_action_at (char *text, location_t location) +{ + char *cp = text + 1; locations_flag = 1; - ++cp; if (*cp == '$') { obstack_sgrow (&string_obstack, "]b4_lhs_location["); } - else if (isdigit (*cp) || *cp == '-') + else { - int n = strtol (cp, &cp, 10); - if (n > rule_length) - complain (_("invalid value: %s%d"), "@", n); + long num; + errno = 0; + num = strtol (cp, 0, 10); + + if (INT_MIN <= num && num <= rule_length && ! errno) + { + int n = num; + obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[", + rule_length, n); + } else - obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[", - rule_length, n); + complain_at (location, _("integer out of range: %s"), quote (text)); } +} + + +/*---------------------------------------------------------------. +| TEXT is expected to 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 == '$') + obstack_sgrow (&string_obstack, "]b4_at_dollar["); else + complain_at (location, _("invalid value: %s"), quote (text)); +} + + +/*-------------------------------------------------------------------. +| Dispatch onto handle_action_at, or handle_destructor_at, depending | +| upon CODE_KIND. | +`-------------------------------------------------------------------*/ + +static void +handle_at (braced_code_t braced_code_kind, + char *text, location_t location) +{ + switch (braced_code_kind) { - char buf[] = "@c"; - buf[1] = *cp; - complain (_("%s is invalid"), quote (buf)); + case action_braced_code: + handle_action_at (text, location); + break; + + case destructor_braced_code: + case printer_braced_code: + handle_symbol_code_at (text, location); + break; } } + +/*------------------------------------------------------------------. +| Convert universal character name UCN to a single-byte character, | +| and return that character. Return -1 if UCN does not correspond | +| to a single-byte character. | +`------------------------------------------------------------------*/ + +static int +convert_ucn_to_byte (char const *ucn) +{ + unsigned long code = strtoul (ucn + 2, 0, 16); + + /* FIXME: Currently we assume Unicode-compatible unibyte characters + on ASCII hosts (i.e., Latin-1 on hosts with 8-bit bytes). On + non-ASCII hosts we support only the portable C character set. + These limitations should be removed once we add support for + multibyte characters. */ + + if (UCHAR_MAX < code) + return -1; + +#if ! ('$' == 0x24 && '@' == 0x40 && '`' == 0x60 && '~' == 0x7e) + { + /* A non-ASCII host. Use CODE to index into a table of the C + basic execution character set, which is guaranteed to exist on + all Standard C platforms. This table also includes '$', '@', + and '`', which are not in the basic execution character set but + which are unibyte characters on all the platforms that we know + about. */ + static signed char const table[] = + { + '\0', -1, -1, -1, -1, -1, -1, '\a', + '\b', '\t', '\n', '\v', '\f', '\r', -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, + ' ', '!', '"', '#', '$', '%', '&', '\'', + '(', ')', '*', '+', ',', '-', '.', '/', + '0', '1', '2', '3', '4', '5', '6', '7', + '8', '9', ':', ';', '<', '=', '>', '?', + '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G', + 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', + 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', + 'X', 'Y', 'Z', '[', '\\', ']', '^', '_', + '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g', + 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', + 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', + 'x', 'y', 'z', '{', '|', '}', '~' + }; + + code = code < sizeof table ? table[code] : -1; + } +#endif + + return code; +} + + +/*----------------------------------------------------------------. +| Handle `#line INT "FILE"'. ARGS has already skipped `#line '. | +`----------------------------------------------------------------*/ + +static void +handle_syncline (char *args, location_t *location) +{ + int lineno = strtol (args, &args, 10); + const char *file = NULL; + file = strchr (args, '"') + 1; + *strchr (file, '"') = 0; + /* FIXME: Leaking... Can't free, as some locations are still + pointing to the old file name. */ + infile = xstrdup (file); + location->file = infile; + location->last_line = lineno; +} + + +/*-------------------------------------------------------------. +| Report an unexpected end of file at LOC. An end of file was | +| encountered and the expected TOKEN_END was missing. After | +| reporting the problem, pretend that TOKEN_END was found. | +`-------------------------------------------------------------*/ + +static void +unexpected_end_of_file (location_t *loc, char const *token_end) +{ + size_t i = strlen (token_end); + + complain_at (*loc, _("missing `%s' at end of file"), token_end); + + /* Adjust location's last column so that any later message does not + mention the characters just inserted. */ + loc->last_column -= i; + + while (i != 0) + unput (token_end[--i]); +} + + +/*-------------------------. +| Initialize the scanner. | +`-------------------------*/ + void scanner_initialize (void) { @@ -633,8 +868,14 @@ 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); }