%{
#include "system.h"
+#include "mbswidth.h"
#include "complain.h"
#include "quote.h"
#include "getargs.h"
if (yycontrol) {;}; \
} while (0)
-#define YY_USER_ACTION LOCATION_COLUMNS (*yylloc, yyleng);
-#define YY_LINES LOCATION_LINES (*yylloc, yyleng);
-#define YY_STEP LOCATION_STEP (*yylloc);
+#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
char *cp, location_t location);
static void handle_at (braced_code_t code_kind,
char *cp, location_t location);
+static int convert_ucn_to_byte (char const *hex_text);
%}
-%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]*
+letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
+id {letter}({letter}|[0-9])*
int [0-9]+
-eols (\n|\r|\n\r|\r\n)+
-blanks [ \t\f]+
+
+/* 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)*
%%
%{
"%nterm" return PERCENT_NTERM;
"%output" return PERCENT_OUTPUT;
"%parse-param" return PERCENT_PARSE_PARAM;
- "%prec" { rule_length--; return PERCENT_PREC; }
+ "%prec" rule_length--; return PERCENT_PREC;
"%printer" return PERCENT_PRINTER;
"%pure"[-_]"parser" return PERCENT_PURE_PARSER;
"%right" return PERCENT_RIGHT;
"%yacc" return PERCENT_YACC;
"=" return EQUAL;
- ":" { rule_length = 0; return COLON; }
- "|" { rule_length = 0; 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 = 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, _("%s is invalid"), 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; yy_push_state (SC_ESCAPED_STRING);
/* Comments. */
- "/*" yy_push_state (SC_COMMENT);
+ "/*" BEGIN SC_YACC_COMMENT;
"//".* YY_STEP;
/* Prologue. */
"{" YY_OBS_GROW; ++braces_level; yy_push_state (SC_BRACED_CODE);
/* A type. */
- "<"[^>]+">" {
+ "<"{tag}">" {
obstack_grow (&string_obstack, yytext + 1, yyleng - 2);
YY_OBS_FINISH;
yylval->string = last_string;
}
- /*------------------------------------------------------------.
- | 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. |
- `------------------------------------------------------------*/
+ /*-------------------------------------------------------------------.
+ | Whatever the start condition (but those which correspond to |
+ | entities `swallowed' by Bison: SC_YACC_COMMENT, SC_ESCAPED_STRING, |
+ | and SC_ESCAPED_CHARACTER), no M4 character must escape as is. |
+ `-------------------------------------------------------------------*/
-<SC_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
+<SC_COMMENT,SC_LINE_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
{
- \[ if (YY_START != SC_COMMENT) obstack_sgrow (&string_obstack, "@<:@");
- \] if (YY_START != SC_COMMENT) obstack_sgrow (&string_obstack, "@:>@");
+ \[ obstack_sgrow (&string_obstack, "@<:@");
+ \] obstack_sgrow (&string_obstack, "@:>@");
}
+ /*---------------------------------------------------------------.
+ | Scanning a Yacc comment. The initial `/ *' is already eaten. |
+ `---------------------------------------------------------------*/
- /*-----------------------------------------------------------.
- | Scanning a C comment. The initial `/ *' is already eaten. |
- `-----------------------------------------------------------*/
-
-<SC_COMMENT>
+<SC_YACC_COMMENT>
{
- "*/" { /* End of the comment. */
- if (yy_top_state () == INITIAL)
- {
- YY_STEP;
- }
- else
- {
- YY_OBS_GROW;
- }
- yy_pop_state ();
+ "*/" {
+ YY_STEP;
+ BEGIN INITIAL;
}
- [^\[\]*\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;
+ [^*]+|"*" ;
+
+ <<EOF>> {
+ LOCATION_PRINT (stderr, *yylloc);
+ fprintf (stderr, _(": unexpected end of file in a comment\n"));
+ BEGIN INITIAL;
+ }
+}
+
+
+ /*------------------------------------------------------------.
+ | Scanning a C comment. The initial `/ *' is already eaten. |
+ `------------------------------------------------------------*/
+
+<SC_COMMENT>
+{
+ "*"{splice}"/" YY_OBS_GROW; yy_pop_state ();
+ [^*\[\]]+|"*" YY_OBS_GROW;
<<EOF>> {
LOCATION_PRINT (stderr, *yylloc);
}
+ /*--------------------------------------------------------------.
+ | Scanning a line comment. The initial `//' is already eaten. |
+ `--------------------------------------------------------------*/
+
+<SC_LINE_COMMENT>
+{
+ "\n" YY_OBS_GROW; yy_pop_state ();
+ ([^\n\[\]]|{splice})+ YY_OBS_GROW;
+ <<EOF>> yy_pop_state ();
+}
+
+
/*----------------------------------------------------------------.
| Scanning a C string, including its escapes. The initial `"' is |
| already eaten. |
return STRING;
}
- [^\"\n\r\\]+ YY_OBS_GROW;
-
- {eols} obstack_1grow (&string_obstack, '\n'); YY_LINES;
+ [^\"\\]+ YY_OBS_GROW;
<<EOF>> {
LOCATION_PRINT (stderr, *yylloc);
}
}
- [^\n\r\\] YY_OBS_GROW;
-
- {eols} obstack_1grow (&string_obstack, '\n'); YY_LINES;
+ [^'\\]+ YY_OBS_GROW;
<<EOF>> {
LOCATION_PRINT (stderr, *yylloc);
<SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>
{
- \\[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"), quote (yytext));
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)
+ {
+ LOCATION_PRINT (stderr, *yylloc);
+ fprintf (stderr, _(": invalid escape: %s\n"), quote (yytext));
+ YY_STEP;
+ }
+ else
+ obstack_1grow (&string_obstack, c);
}
\\a obstack_1grow (&string_obstack, '\a');
\\r obstack_1grow (&string_obstack, '\r');
\\t obstack_1grow (&string_obstack, '\t');
\\v obstack_1grow (&string_obstack, '\v');
- \\[\\""''] obstack_1grow (&string_obstack, yytext[1]);
+ \\[\"'?\\] 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)
+ {
+ LOCATION_PRINT (stderr, *yylloc);
+ fprintf (stderr, _(": invalid escape: %s\n"), quote (yytext));
+ YY_STEP;
+ }
+ else
+ obstack_1grow (&string_obstack, c);
+ }
\\(.|\n) {
LOCATION_PRINT (stderr, *yylloc);
fprintf (stderr, _(": unrecognized escape: %s\n"), quote (yytext));
yy_pop_state ();
}
- [^\[\]\'\n\r\\]+ YY_OBS_GROW;
- \\(.|\n) YY_OBS_GROW;
- /* FLex wants this rule, in case of a `\<<EOF>>'. */
+ [^'\[\]\\]+ YY_OBS_GROW;
+ \\{splice}[^\[\]] YY_OBS_GROW;
+ {splice} YY_OBS_GROW;
+ /* Needed for `\<<EOF>>', `\\<<newline>>[', and `\\<<newline>>]'. */
\\ YY_OBS_GROW;
- {eols} YY_OBS_GROW; YY_LINES;
-
<<EOF>> {
LOCATION_PRINT (stderr, *yylloc);
fprintf (stderr, _(": unexpected end of file in a character\n"));
yy_pop_state ();
}
- [^\[\]\"\n\r\\]+ YY_OBS_GROW;
- \\(.|\n) YY_OBS_GROW;
- /* FLex wants this rule, in case of a `\<<EOF>>'. */
+ [^\"\[\]\\]+ YY_OBS_GROW;
+ \\{splice}[^\[\]] YY_OBS_GROW;
+ {splice} YY_OBS_GROW;
+ /* Needed for `\<<EOF>>', `\\<<newline>>[', and `\\<<newline>>]'. */
\\ YY_OBS_GROW;
- {eols} YY_OBS_GROW; YY_LINES;
-
<<EOF>> {
LOCATION_PRINT (stderr, *yylloc);
fprintf (stderr, _(": unexpected end of file in a string\n"));
"\"" YY_OBS_GROW; yy_push_state (SC_STRING);
/* Comments. */
- "/*" YY_OBS_GROW; yy_push_state (SC_COMMENT);
- "//".* YY_OBS_GROW;
+ "/"{splice}"*" YY_OBS_GROW; yy_push_state (SC_COMMENT);
+ "/"{splice}"/" YY_OBS_GROW; yy_push_state (SC_LINE_COMMENT);
/* Not comments. */
"/" YY_OBS_GROW;
"{" YY_OBS_GROW; braces_level++;
- "$"("<"[^>]+">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code,
+ "$"("<"{tag}">")?(-?[0-9]+|"$") { 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;
+ [^$@\[\]/'\"\{\}]+ YY_OBS_GROW;
- /* A lose $, or /, or etc. */
+ /* A stray $, or /, or etc. */
. YY_OBS_GROW;
<<EOF>> {
return PROLOGUE;
}
- [^%\[\]/\'\"\n\r]+ YY_OBS_GROW;
+ [^%\[\]/'\"]+ YY_OBS_GROW;
"%" YY_OBS_GROW;
- {eols} YY_OBS_GROW; YY_LINES;
<<EOF>> {
LOCATION_PRINT (stderr, *yylloc);
/*---------------------------------------------------------------.
| Scanning the epilogue (everything after the second "%%", which |
- | has already been eaten. |
+ | has already been eaten). |
`---------------------------------------------------------------*/
<SC_EPILOGUE>
{
- ([^\[\]]|{eols})+ YY_OBS_GROW;
+ [^\[\]]+ YY_OBS_GROW;
<<EOF>> {
yy_pop_state ();
obstack_fgrow1 (&string_obstack,
"]b4_lhs_value([%s])[", type_name);
}
- else if (('0' <= *cp && *cp <= '9') || *cp == '-')
+ else
{
- int n = strtol (cp, &cp, 10);
+ long num;
+ errno = 0;
+ num = strtol (cp, 0, 10);
- if (n > rule_length)
- complain_at (location, _("invalid value: %s%d"), "$", n);
- else
+ if (INT_MIN <= num && num <= rule_length && ! errno)
{
+ int n = num;
if (!type_name && n > 0)
type_name = symbol_list_n_type_name_get (current_rule, location,
n);
"]b4_rhs_value([%d], [%d], [%s])[",
rule_length, n, type_name);
}
- }
- else
- {
- complain_at (location, _("%s is invalid"), quote (text));
+ else
+ complain_at (location, _("invalid value: %s"), text);
}
}
/*---------------------------------------------------------------.
-| TEXT is expexted tp be $$ in some code associated to a symbol: |
+| TEXT is expected to be $$ in some code associated to a symbol: |
| destructor or printer. |
`---------------------------------------------------------------*/
if (*cp == '$')
obstack_sgrow (&string_obstack, "]b4_dollar_dollar[");
else
- complain_at (location, _("%s is invalid"), quote (text));
+ complain_at (location, _("%s is invalid"), quote_n (1, text));
}
{
obstack_sgrow (&string_obstack, "]b4_lhs_location[");
}
- else if (('0' <= *cp && *cp <= '9') || *cp == '-')
+ else
{
- int n = strtol (cp, &cp, 10);
+ long num;
+ errno = 0;
+ num = strtol (cp, 0, 10);
- if (n > rule_length)
- complain_at (location, _("invalid value: %s%d"), "@", n);
+ if (INT_MIN <= num && num <= rule_length && ! errno)
+ {
+ int n = num;
+ obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[",
+ rule_length, n);
+ }
else
- obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[",
- rule_length, n);
- }
- else
- {
- complain_at (location, _("%s is invalid"), quote (text));
+ complain_at (location, _("invalid value: %s"), text);
}
}
/*---------------------------------------------------------------.
-| TEXT is expexted tp be @$ in some code associated to a symbol: |
+| TEXT is expected to be @$ in some code associated to a symbol: |
| destructor or printer. |
`---------------------------------------------------------------*/
if (*cp == '$')
obstack_sgrow (&string_obstack, "]b4_at_dollar[");
else
- complain_at (location, _("%s is invalid"), quote (text));
+ complain_at (location, _("%s is invalid"), quote_n (1, text));
}
}
+/*------------------------------------------------------------------.
+| Convert universal character name UCN to a single-byte character, |
+| and return that character. Return -1 if UCN does not correspond |
+| to a single-byte character. |
+`------------------------------------------------------------------*/
+
+static int
+convert_ucn_to_byte (char const *ucn)
+{
+ unsigned long code = strtoul (ucn + 2, 0, 16);
+
+ /* FIXME: Currently we assume Unicode-compatible unibyte characters
+ on ASCII hosts (i.e., Latin-1 on hosts with 8-bit bytes). On
+ non-ASCII hosts we support only the portable C character set.
+ These limitations should be removed once we add support for
+ multibyte characters. */
+
+ if (UCHAR_MAX < code)
+ return -1;
+
+#if ! ('$' == 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 '$', '@',
+ and '`', which not in the basic execution character set but
+ which are unibyte characters on all the platforms that we know
+ about. */
+ static signed char const table[] =
+ {
+ '\0', -1, -1, -1, -1, -1, -1, '\a',
+ '\b', '\t', '\n', '\v', '\f', '\r', -1, -1,
+ -1, -1, -1, -1, -1, -1, -1, -1,
+ -1, -1, -1, -1, -1, -1, -1, -1,
+ ' ', '!', '"', '#', '$', '%', '&', '\'',
+ '(', ')', '*', '+', ',', '-', '.', '/',
+ '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;
+}
+
+
/*-------------------------.
| Initialize the scanner. |
`-------------------------*/