X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/4413bbd3d1689a23f441cbdca898b99d802bb6d6..71b52b1342c65a5497f6b3780481ff2deb932a56:/src/scan-gram.l diff --git a/src/scan-gram.l b/src/scan-gram.l index bcb7209b..e2e08f9a 100644 --- a/src/scan-gram.l +++ b/src/scan-gram.l @@ -1,7 +1,6 @@ /* Bison Grammar Scanner -*- C -*- - Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 - Free Software Foundation, Inc. + Copyright (C) 2002-2012 Free Software Foundation, Inc. This file is part of Bison, the GNU Compiler Compiler. @@ -40,13 +39,14 @@ #include #include #include +#include #include #define YY_DECL GRAM_LEX_DECL -#define YY_USER_INIT \ - code_start = scanner_cursor = loc->start; \ +#define YY_USER_INIT \ + code_start = scanner_cursor = loc->start; \ /* Location of scanner cursor. */ static boundary scanner_cursor; @@ -56,15 +56,21 @@ static boundary scanner_cursor; static size_t no_cr_read (FILE *, char *, size_t); #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size)) -#define RETURN_PERCENT_FLAG(Value) \ +#define RETURN_PERCENT_PARAM(Value) \ + RETURN_VALUE(PERCENT_PARAM, param, param_ ## Value) + +#define RETURN_PERCENT_FLAG(Value) \ + RETURN_VALUE(PERCENT_FLAG, uniqstr, uniqstr_new (Value)) + +#define RETURN_VALUE(Token, Field, Value) \ do { \ - val->uniqstr = uniqstr_new (Value); \ - return PERCENT_FLAG; \ + val->Field = Value; \ + return Token; \ } while (0) #define ROLLBACK_CURRENT_TOKEN \ do { \ - scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0); \ + scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0); \ yyless (0); \ } while (0) @@ -95,16 +101,17 @@ static void unexpected_newline (boundary, char const *); /* Strings and characters in directives/rules. */ %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER /* A identifier was just read in directives/rules. Special state - to capture the sequence `identifier :'. */ + to capture the sequence 'identifier :'. */ %x SC_AFTER_IDENTIFIER /* A complex tag, with nested angles brackets. */ %x SC_TAG - /* Three types of user code: - - prologue (code between `%{' `%}' in the first section, before %%); + /* Four types of user code: + - prologue (code between '%{' '%}' in the first section, before %%); - actions, printers, union, etc, (between braced in the middle section); - - epilogue (everything after the second %%). */ -%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE + - epilogue (everything after the second %%). + - predicate (code between '%?{' and '{' in middle section); */ +%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE SC_PREDICATE /* C and C++ comments in code. */ %x SC_COMMENT SC_LINE_COMMENT /* Strings and characters in code. */ @@ -112,33 +119,33 @@ static void unexpected_newline (boundary, char const *); /* Bracketed identifiers support. */ %x SC_BRACKETED_ID SC_RETURN_BRACKETED_ID -letter [-.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] -id {letter}({letter}|[0-9])* +letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] +id {letter}({letter}|[-0-9])* directive %{id} -int [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, as this simplifies our implementation. We disallow angle bracket to match them in nested pairs: several languages use them for generics/template types. */ -tag [^\0<>]+ +tag [^\0<>]+ /* Zero or more instances of backslash-newline. Following GCC, allow white space between the backslash and the newline. */ -splice (\\[ \f\t\v]*\n)* +splice (\\[ \f\t\v]*\n)* %% %{ /* Nesting level. Either for nested braces, or nested angle brackets (but not mixed). */ - int nesting IF_LINT (= 0); + int nesting PACIFY_CC (= 0); /* Parent context state, when applicable. */ - int context_state IF_LINT (= 0); + int context_state PACIFY_CC (= 0); /* Location of most recent identifier, when applicable. */ - location id_loc IF_LINT (= empty_location); + location id_loc PACIFY_CC (= empty_location); /* Where containing code started, when applicable. Its initial value is relevant only when yylex is invoked in the SC_EPILOGUE @@ -147,7 +154,7 @@ splice (\\[ \f\t\v]*\n)* /* Where containing comment or string or character literal started, when applicable. */ - boundary token_start IF_LINT (= scanner_cursor); + boundary token_start PACIFY_CC (= scanner_cursor); %} @@ -158,7 +165,7 @@ splice (\\[ \f\t\v]*\n)* { /* Comments and white space. */ - "," warn_at (*loc, _("stray `,' treated as white space")); + "," warn_at (*loc, _("stray ',' treated as white space")); [ \f\n\t\v] | "//".* ; "/*" { @@ -180,7 +187,7 @@ splice (\\[ \f\t\v]*\n)* `----------------------------*/ /* For directives that are also command line options, the regex must be - "%..." + "%..." after "[-_]"s are removed, and the directive must match the --long option name, with a single string argument. Otherwise, add exceptions to ../build-aux/cross-options.pl. */ @@ -204,7 +211,7 @@ splice (\\[ \f\t\v]*\n)* "%glr-parser" return PERCENT_GLR_PARSER; "%language" return PERCENT_LANGUAGE; "%left" return PERCENT_LEFT; - "%lex-param" return PERCENT_LEX_PARAM; + "%lex-param" RETURN_PERCENT_PARAM(lex); "%locations" RETURN_PERCENT_FLAG("locations"); "%merge" return PERCENT_MERGE; "%name"[-_]"prefix" return PERCENT_NAME_PREFIX; @@ -214,7 +221,8 @@ splice (\\[ \f\t\v]*\n)* "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER; "%nterm" return PERCENT_NTERM; "%output" return PERCENT_OUTPUT; - "%parse-param" return PERCENT_PARSE_PARAM; + "%param" RETURN_PERCENT_PARAM(both); + "%parse-param" RETURN_PERCENT_PARAM(parse); "%prec" return PERCENT_PREC; "%precedence" return PERCENT_PRECEDENCE; "%printer" return PERCENT_PRINTER; @@ -262,10 +270,10 @@ splice (\\[ \f\t\v]*\n)* } /* Characters. */ - "'" token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER; + "'" token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER; /* Strings. */ - "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING; + "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING; /* Prologue. */ "%{" code_start = loc->start; BEGIN SC_PROLOGUE; @@ -278,6 +286,13 @@ splice (\\[ \f\t\v]*\n)* BEGIN SC_BRACED_CODE; } + /* Semantic predicate. */ + "%?"[ \f\n\t\v]*"{" { + nesting = 0; + code_start = loc->start; + BEGIN SC_PREDICATE; + } + /* A type. */ "<*>" return TAG_ANY; "<>" return TAG_NONE; @@ -326,7 +341,7 @@ splice (\\[ \f\t\v]*\n)* { - \0 complain_at (*loc, _("invalid null character")); + \0 complain_at (*loc, _("invalid null character")); } @@ -339,16 +354,16 @@ splice (\\[ \f\t\v]*\n)* "[" { if (bracketed_id_str) { - ROLLBACK_CURRENT_TOKEN; - BEGIN SC_RETURN_BRACKETED_ID; - *loc = id_loc; - return ID; + ROLLBACK_CURRENT_TOKEN; + BEGIN SC_RETURN_BRACKETED_ID; + *loc = id_loc; + return ID; } else { - bracketed_id_start = loc->start; - bracketed_id_context_state = YY_START; - BEGIN SC_BRACKETED_ID; + bracketed_id_start = loc->start; + bracketed_id_context_state = YY_START; + BEGIN SC_BRACKETED_ID; } } ":" { @@ -378,33 +393,33 @@ splice (\\[ \f\t\v]*\n)* {id} { if (bracketed_id_str) { - complain_at (*loc, _("unexpected identifier in bracketed name: %s"), - quote (yytext)); + complain_at (*loc, _("unexpected identifier in bracketed name: %s"), + quote (yytext)); } else { - bracketed_id_str = uniqstr_new (yytext); - bracketed_id_loc = *loc; + bracketed_id_str = uniqstr_new (yytext); + bracketed_id_loc = *loc; } } "]" { BEGIN bracketed_id_context_state; if (bracketed_id_str) { - if (INITIAL == bracketed_id_context_state) - { - val->uniqstr = bracketed_id_str; - bracketed_id_str = 0; - *loc = bracketed_id_loc; - return BRACKETED_ID; - } + if (INITIAL == bracketed_id_context_state) + { + val->uniqstr = bracketed_id_str; + bracketed_id_str = 0; + *loc = bracketed_id_loc; + return BRACKETED_ID; + } } else complain_at (*loc, _("an identifier expected")); } . { complain_at (*loc, _("invalid character in bracketed name: %s"), - quote (yytext)); + quote (yytext)); } <> { BEGIN bracketed_id_context_state; @@ -426,37 +441,37 @@ splice (\\[ \f\t\v]*\n)* /*---------------------------------------------------------------. - | Scanning a Yacc comment. The initial `/ *' is already eaten. | + | Scanning a Yacc comment. The initial '/ *' is already eaten. | `---------------------------------------------------------------*/ { "*/" BEGIN context_state; - .|\n ; + .|\n ; <> unexpected_eof (token_start, "*/"); BEGIN context_state; } /*------------------------------------------------------------. - | Scanning a C comment. The initial `/ *' is already eaten. | + | Scanning a C comment. The initial '/ *' is already eaten. | `------------------------------------------------------------*/ { "*"{splice}"/" STRING_GROW; BEGIN context_state; - <> unexpected_eof (token_start, "*/"); BEGIN context_state; + <> unexpected_eof (token_start, "*/"); BEGIN context_state; } /*--------------------------------------------------------------. - | Scanning a line comment. The initial `//' is already eaten. | + | Scanning a line comment. The initial '//' is already eaten. | `--------------------------------------------------------------*/ { - "\n" STRING_GROW; BEGIN context_state; - {splice} STRING_GROW; - <> BEGIN context_state; + "\n" STRING_GROW; BEGIN context_state; + {splice} STRING_GROW; + <> BEGIN context_state; } @@ -488,7 +503,7 @@ splice (\\[ \f\t\v]*\n)* /*----------------------------------------------------------. | Scanning a Bison character literal, decoding its escapes. | - | The initial quote is already eaten. | + | The initial quote is already eaten. | `----------------------------------------------------------*/ @@ -596,13 +611,13 @@ splice (\\[ \f\t\v]*\n)* obstack_1grow (&obstack_for_string, c); } - \\a obstack_1grow (&obstack_for_string, '\a'); - \\b obstack_1grow (&obstack_for_string, '\b'); - \\f obstack_1grow (&obstack_for_string, '\f'); - \\n obstack_1grow (&obstack_for_string, '\n'); - \\r obstack_1grow (&obstack_for_string, '\r'); - \\t obstack_1grow (&obstack_for_string, '\t'); - \\v obstack_1grow (&obstack_for_string, '\v'); + \\a obstack_1grow (&obstack_for_string, '\a'); + \\b obstack_1grow (&obstack_for_string, '\b'); + \\f obstack_1grow (&obstack_for_string, '\f'); + \\n obstack_1grow (&obstack_for_string, '\n'); + \\r obstack_1grow (&obstack_for_string, '\r'); + \\t obstack_1grow (&obstack_for_string, '\t'); + \\v obstack_1grow (&obstack_for_string, '\v'); /* \\[\"\'?\\] would be shorter, but it confuses xgettext. */ \\("\""|"'"|"?"|"\\") obstack_1grow (&obstack_for_string, yytext[1]); @@ -615,7 +630,7 @@ splice (\\[ \f\t\v]*\n)* else obstack_1grow (&obstack_for_string, c); } - \\(.|\n) { + \\(.|\n) { char const *p = yytext + 1; /* Quote only if escaping won't make the character visible. */ if (isspace ((unsigned char) *p) && isprint ((unsigned char) *p)) @@ -632,21 +647,21 @@ splice (\\[ \f\t\v]*\n)* { - {splice}|\\{splice}[^\n\[\]] STRING_GROW; + {splice}|\\{splice}[^\n\[\]] STRING_GROW; } { - "'" STRING_GROW; BEGIN context_state; - \n unexpected_newline (token_start, "'"); BEGIN context_state; - <> unexpected_eof (token_start, "'"); BEGIN context_state; + "'" STRING_GROW; BEGIN context_state; + \n unexpected_newline (token_start, "'"); BEGIN context_state; + <> unexpected_eof (token_start, "'"); BEGIN context_state; } { - "\"" STRING_GROW; BEGIN context_state; - \n unexpected_newline (token_start, "\""); BEGIN context_state; - <> unexpected_eof (token_start, "\""); BEGIN context_state; + "\"" STRING_GROW; BEGIN context_state; + \n unexpected_newline (token_start, "\""); BEGIN context_state; + <> unexpected_eof (token_start, "\""); BEGIN context_state; } @@ -654,7 +669,7 @@ splice (\\[ \f\t\v]*\n)* | Strings, comments etc. can be found in user code. | `---------------------------------------------------*/ - + { "'" { STRING_GROW; @@ -684,42 +699,63 @@ splice (\\[ \f\t\v]*\n)* /*-----------------------------------------------------------. - | Scanning some code in braces (actions). The initial "{" is | - | already eaten. | + | Scanning some code in braces (actions, predicates). The | + | initial "{" is already eaten. | `-----------------------------------------------------------*/ - + { "{"|"<"{splice}"%" STRING_GROW; nesting++; "%"{splice}">" STRING_GROW; nesting--; - "}" { - obstack_1grow (&obstack_for_string, '}'); - - --nesting; - if (nesting < 0) - { - STRING_FINISH; - loc->start = code_start; - val->code = last_string; - BEGIN INITIAL; - return BRACED_CODE; - } - } - /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly - (as `<' `<%'). */ + /* Tokenize '<<%' correctly (as '<<' '%') rather than incorrrectly + (as '<' '<%'). */ "<"{splice}"<" STRING_GROW; <> { + int token = (YY_START == SC_BRACED_CODE) ? BRACED_CODE : BRACED_PREDICATE; unexpected_eof (code_start, "}"); STRING_FINISH; loc->start = code_start; val->code = last_string; BEGIN INITIAL; - return BRACED_CODE; + return token; } } + +{ + "}" { + obstack_1grow (&obstack_for_string, '}'); + + --nesting; + if (nesting < 0) + { + STRING_FINISH; + loc->start = code_start; + val->code = last_string; + BEGIN INITIAL; + return BRACED_CODE; + } + } +} + + +{ + "}" { + --nesting; + if (nesting < 0) + { + STRING_FINISH; + loc->start = code_start; + val->code = last_string; + BEGIN INITIAL; + return BRACED_PREDICATE; + } + else + obstack_1grow (&obstack_for_string, '}'); + } +} /*--------------------------------------------------------------. | Scanning some prologue: from "%{" (already scanned) to "%}". | @@ -767,8 +803,8 @@ splice (\\[ \f\t\v]*\n)* | By default, grow the string obstack with the input. | `-----------------------------------------------------*/ -. | -\n STRING_GROW; +. | + \n STRING_GROW; %% @@ -784,35 +820,35 @@ no_cr_read (FILE *fp, char *buf, size_t size) { char *w = memchr (buf, '\r', bytes_read); if (w) - { - char const *r = ++w; - char const *lim = buf + bytes_read; - - 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; - } + { + char const *r = ++w; + char const *lim = buf + bytes_read; + + 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 bytes_read; @@ -843,7 +879,7 @@ scan_integer (char const *number, int base, location loc) /*------------------------------------------------------------------. | 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. | +| to a single-byte character. | `------------------------------------------------------------------*/ static int @@ -871,22 +907,22 @@ convert_ucn_to_byte (char const *ucn) 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, - ' ', '!', '"', '#', '$', '%', '&', '\'', - '(', ')', '*', '+', ',', '-', '.', '/', - '0', '1', '2', '3', '4', '5', '6', '7', - '8', '9', ':', ';', '<', '=', '>', '?', - '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G', - 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', - 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', - 'X', 'Y', 'Z', '[', '\\', ']', '^', '_', - '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g', - 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', - 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', - 'x', 'y', 'z', '{', '|', '}', '~' + '\0', -1, -1, -1, -1, -1, -1, '\a', + '\b', '\t', '\n', '\v', '\f', '\r', -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, + ' ', '!', '"', '#', '$', '%', '&', '\'', + '(', ')', '*', '+', ',', '-', '.', '/', + '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; @@ -898,7 +934,7 @@ convert_ucn_to_byte (char const *ucn) /*----------------------------------------------------------------. -| Handle `#line INT "FILE"'. ARGS has already skipped `#line '. | +| Handle '#line INT "FILE"'. ARGS has already skipped '#line '. | `----------------------------------------------------------------*/ static void @@ -906,8 +942,8 @@ handle_syncline (char *args, location loc) { char *after_num; unsigned long int lineno = strtoul (args, &after_num, 10); - char *file = strchr (after_num, '"') + 1; - *strchr (file, '"') = '\0'; + char *file = mbschr (after_num, '"') + 1; + *mbschr (file, '"') = '\0'; if (INT_MAX <= lineno) { warn_at (loc, _("line number overflow")); @@ -920,8 +956,8 @@ handle_syncline (char *args, location loc) /*----------------------------------------------------------------. | For a token or comment starting at START, report message MSGID, | -| which should say that an end marker was found before | -| the expected TOKEN_END. | +| which should say that an end marker was found before | +| the expected TOKEN_END. | `----------------------------------------------------------------*/ static void @@ -930,6 +966,10 @@ unexpected_end (boundary start, char const *msgid, char const *token_end) location loc; loc.start = start; loc.end = scanner_cursor; + token_end = quote (token_end); + // Instead of '\'', display "'". + if (STREQ (token_end, "'\\''", '\'', '\\', '\'', '\'', 0,0,0,0,0)) + token_end = "\"'\""; complain_at (loc, _(msgid), token_end); } @@ -942,7 +982,7 @@ unexpected_end (boundary start, char const *msgid, char const *token_end) static void unexpected_eof (boundary start, char const *token_end) { - unexpected_end (start, N_("missing `%s' at end of file"), token_end); + unexpected_end (start, N_("missing %s at end of file"), token_end); } @@ -953,7 +993,7 @@ unexpected_eof (boundary start, char const *token_end) static void unexpected_newline (boundary start, char const *token_end) { - unexpected_end (start, N_("missing `%s' at end of line"), token_end); + unexpected_end (start, N_("missing %s at end of line"), token_end); }