*/
#include <stdio.h>
+#include <ctype.h>
#include <unistd.h>
#include <string.h>
#include <stdlib.h>
#undef ENTRY
};
+/*
+ * Static node for out of memory errors. Only "type" is used. lineno is
+ * always left at 0. The rest of the node content can be overwritten,
+ * but is never used.
+ */
+static struct ctf_node error_node = {
+ .type = NODE_ERROR,
+};
+
BT_HIDDEN
const char *node_type(struct ctf_node *node)
{
return gstr;
}
-/*
- * note: never use gc_string_append on a string that has external references.
- * gsrc will be garbage collected immediately, and gstr might be.
- * Should only be used to append characters to a string literal or constant.
- */
-BT_HIDDEN
-struct gc_string *gc_string_append(struct ctf_scanner *scanner,
- struct gc_string *gstr,
- struct gc_string *gsrc)
+void setstring(struct ctf_scanner *scanner, YYSTYPE *lvalp, const char *src)
{
- size_t newlen = strlen(gsrc->s) + strlen(gstr->s) + 1;
- size_t alloclen;
+ lvalp->gs = gc_string_alloc(scanner, strlen(src) + 1);
+ strcpy(lvalp->gs->s, src);
+}
- /* TODO: could be faster with find first bit or glib Gstring */
- /* sizeof long to account for malloc header (int or long ?) */
- for (alloclen = 8; alloclen < sizeof(long) + sizeof(*gstr) + newlen;
- alloclen *= 2);
+static
+int str_check(size_t str_len, size_t offset, size_t len)
+{
+ /* check overflow */
+ if (offset + len < offset)
+ return -1;
+ if (offset + len > str_len)
+ return -1;
+ return 0;
+}
- if (alloclen > gstr->alloclen) {
- struct gc_string *newgstr;
+static
+int bt_isodigit(int c)
+{
+ switch (c) {
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ return 1;
+ default:
+ return 0;
+ }
+}
- newgstr = gc_string_alloc(scanner, newlen);
- strcpy(newgstr->s, gstr->s);
- strcat(newgstr->s, gsrc->s);
- bt_list_del(&gstr->gc);
- free(gstr);
- gstr = newgstr;
- } else {
- strcat(gstr->s, gsrc->s);
+static
+int parse_base_sequence(const char *src, size_t len, size_t pos,
+ char *buffer, size_t *buf_len, int base)
+{
+ const size_t max_char = 3;
+ int nr_char = 0;
+
+ while (!str_check(len, pos, 1) && nr_char < max_char) {
+ char c = src[pos++];
+
+ if (base == 8) {
+ if (bt_isodigit(c))
+ buffer[nr_char++] = c;
+ else
+ break;
+ } else if (base == 16) {
+ if (isxdigit(c))
+ buffer[nr_char++] = c;
+ else
+ break;
+
+ } else {
+ /* Unsupported base */
+ return -1;
+ }
}
- bt_list_del(&gsrc->gc);
- free(gsrc);
- return gstr;
+ assert(nr_char > 0);
+ buffer[nr_char] = '\0';
+ *buf_len = nr_char;
+ return 0;
}
-void setstring(struct ctf_scanner *scanner, YYSTYPE *lvalp, const char *src)
+static
+int import_basic_string(struct ctf_scanner *scanner, YYSTYPE *lvalp,
+ size_t len, const char *src, char delim)
{
- lvalp->gs = gc_string_alloc(scanner, strlen(src) + 1);
- strcpy(lvalp->gs->s, src);
+ size_t pos = 0, dpos = 0;
+
+ if (str_check(len, pos, 1))
+ return -1;
+ if (src[pos++] != delim)
+ return -1;
+
+ while (src[pos] != delim) {
+ char c;
+
+ if (str_check(len, pos, 1))
+ return -1;
+ c = src[pos++];
+ if (c == '\\') {
+ if (str_check(len, pos, 1))
+ return -1;
+ c = src[pos++];
+
+ switch (c) {
+ case 'a':
+ c = '\a';
+ break;
+ case 'b':
+ c = '\b';
+ break;
+ case 'f':
+ c = '\f';
+ break;
+ case 'n':
+ c = '\n';
+ break;
+ case 'r':
+ c = '\r';
+ break;
+ case 't':
+ c = '\t';
+ break;
+ case 'v':
+ c = '\v';
+ break;
+ case '\\':
+ c = '\\';
+ break;
+ case '\'':
+ c = '\'';
+ break;
+ case '\"':
+ c = '\"';
+ break;
+ case '?':
+ c = '?';
+ break;
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ {
+ char oct_buffer[4];
+ size_t oct_len;
+
+ if (parse_base_sequence(src, len, pos - 1,
+ oct_buffer, &oct_len, 8))
+ return -1;
+ c = strtoul(&oct_buffer[0], NULL, 8);
+ pos += oct_len - 1;
+ break;
+ }
+ case 'x':
+ {
+ char hex_buffer[4];
+ size_t hex_len;
+
+ if (parse_base_sequence(src, len, pos,
+ hex_buffer, &hex_len, 16))
+ return -1;
+ c = strtoul(&hex_buffer[0], NULL, 16);
+ pos += hex_len;
+ break;
+ }
+ default:
+ return -1;
+ }
+ }
+ if (str_check(len, dpos, 1))
+ return -1;
+ lvalp->gs->s[dpos++] = c;
+ }
+
+ if (str_check(len, dpos, 1))
+ return -1;
+ lvalp->gs->s[dpos++] = '\0';
+
+ if (str_check(len, pos, 1))
+ return -1;
+ if (src[pos++] != delim)
+ return -1;
+
+ if (str_check(len, pos, 1))
+ return -1;
+ if (src[pos] != '\0')
+ return -1;
+ return 0;
+}
+
+int import_string(struct ctf_scanner *scanner, YYSTYPE *lvalp,
+ const char *src, char delim)
+{
+ size_t len;
+
+ len = strlen(src) + 1;
+ lvalp->gs = gc_string_alloc(scanner, len);
+ if (src[0] == 'L') {
+ // TODO: import wide string
+ printfl_error(yyget_lineno(scanner),
+ "Wide string not supported yet.");
+ return -1;
+ } else {
+ return import_basic_string(scanner, lvalp, len, src, delim);
+ }
}
static void init_scope(struct ctf_scanner_scope *scope,
struct ctf_node *node;
node = malloc(sizeof(*node));
- if (!node)
- return NULL;
+ if (!node) {
+ printfl_fatal(yyget_lineno(scanner->scanner), "out of memory");
+ return &error_node;
+ }
memset(node, 0, sizeof(*node));
node->type = type;
node->lineno = yyget_lineno(scanner->scanner);
switch (type) {
case NODE_ROOT:
+ node->type = NODE_ERROR;
printfn_fatal(node, "trying to create root node");
break;
case NODE_UNKNOWN:
default:
+ node->type = NODE_ERROR;
printfn_fatal(node, "unknown node type '%d'", (int) type);
break;
}
*/
%expect 2
%start file
-%token CHARACTER_CONSTANT_START SQUOTE STRING_LITERAL_START DQUOTE ESCSEQ CHAR_STRING_TOKEN LSBRAC RSBRAC LPAREN RPAREN LBRAC RBRAC RARROW STAR PLUS MINUS LT GT TYPEASSIGN COLON SEMICOLON DOTDOTDOT DOT EQUAL COMMA CONST CHAR DOUBLE ENUM ENV EVENT FLOATING_POINT FLOAT INTEGER INT LONG SHORT SIGNED STREAM STRING STRUCT TRACE CALLSITE CLOCK TYPEALIAS TYPEDEF UNSIGNED VARIANT VOID _BOOL _COMPLEX _IMAGINARY DECIMAL_CONSTANT OCTAL_CONSTANT HEXADECIMAL_CONSTANT TOK_ALIGN
+%token INTEGER_LITERAL STRING_LITERAL CHARACTER_LITERAL LSBRAC RSBRAC LPAREN RPAREN LBRAC RBRAC RARROW STAR PLUS MINUS LT GT TYPEASSIGN COLON SEMICOLON DOTDOTDOT DOT EQUAL COMMA CONST CHAR DOUBLE ENUM ENV EVENT FLOATING_POINT FLOAT INTEGER INT LONG SHORT SIGNED STREAM STRING STRUCT TRACE CALLSITE CLOCK TYPEALIAS TYPEDEF UNSIGNED VARIANT VOID _BOOL _COMPLEX _IMAGINARY TOK_ALIGN
%token <gs> IDENTIFIER ID_TYPE
%token ERROR
%union
{
long long ll;
+ unsigned long long ull;
char c;
struct gc_string *gs;
struct ctf_node *n;
}
+%type <gs> STRING_LITERAL CHARACTER_LITERAL
+
%type <gs> keywords
-%type <gs> s_char s_char_sequence c_char c_char_sequence
+%type <ull> INTEGER_LITERAL
%type <n> postfix_expression unary_expression unary_expression_or_range
%type <n> declaration
{ $$ = yylval.gs; }
;
-/* 1.5 Constants */
-
-c_char_sequence:
- c_char
- { $$ = $1; }
- | c_char_sequence c_char
- { $$ = gc_string_append(scanner, $1, $2); }
- ;
-
-c_char:
- CHAR_STRING_TOKEN
- { $$ = yylval.gs; }
- | ESCSEQ
- {
- reparent_error(scanner, "escape sequences not supported yet");
- }
- ;
-
-/* 1.6 String literals */
-
-s_char_sequence:
- s_char
- { $$ = $1; }
- | s_char_sequence s_char
- { $$ = gc_string_append(scanner, $1, $2); }
- ;
-
-s_char:
- CHAR_STRING_TOKEN
- { $$ = yylval.gs; }
- | ESCSEQ
- {
- reparent_error(scanner, "escape sequences not supported yet");
- }
- ;
/* 2: Phrase structure grammar */
$$->u.unary_expression.type = UNARY_STRING;
$$->u.unary_expression.u.string = yylval.gs->s;
}
- | DECIMAL_CONSTANT
+ | INTEGER_LITERAL
{
$$ = make_node(scanner, NODE_UNARY_EXPRESSION);
$$->u.unary_expression.type = UNARY_UNSIGNED_CONSTANT;
- sscanf(yylval.gs->s, "%" PRIu64,
- &$$->u.unary_expression.u.unsigned_constant);
+ $$->u.unary_expression.u.unsigned_constant = $1;
}
- | OCTAL_CONSTANT
- {
- $$ = make_node(scanner, NODE_UNARY_EXPRESSION);
- $$->u.unary_expression.type = UNARY_UNSIGNED_CONSTANT;
- sscanf(yylval.gs->s, "0%" PRIo64,
- &$$->u.unary_expression.u.unsigned_constant);
- }
- | HEXADECIMAL_CONSTANT
- {
- $$ = make_node(scanner, NODE_UNARY_EXPRESSION);
- $$->u.unary_expression.type = UNARY_UNSIGNED_CONSTANT;
- sscanf(yylval.gs->s, "0x%" PRIx64,
- &$$->u.unary_expression.u.unsigned_constant);
- }
- | STRING_LITERAL_START DQUOTE
+ | STRING_LITERAL
{
$$ = make_node(scanner, NODE_UNARY_EXPRESSION);
$$->u.unary_expression.type = UNARY_STRING;
- $$->u.unary_expression.u.string = "";
+ $$->u.unary_expression.u.string = $1->s;
}
- | STRING_LITERAL_START s_char_sequence DQUOTE
+ | CHARACTER_LITERAL
{
$$ = make_node(scanner, NODE_UNARY_EXPRESSION);
$$->u.unary_expression.type = UNARY_STRING;
- $$->u.unary_expression.u.string = $2->s;
- }
- | CHARACTER_CONSTANT_START c_char_sequence SQUOTE
- {
- $$ = make_node(scanner, NODE_UNARY_EXPRESSION);
- $$->u.unary_expression.type = UNARY_STRING;
- $$->u.unary_expression.u.string = $2->s;
+ $$->u.unary_expression.u.string = $1->s;
}
| LPAREN unary_expression RPAREN
{
- $$ = make_node(scanner, NODE_UNARY_EXPRESSION);
- $$->u.unary_expression.type = UNARY_NESTED;
- $$->u.unary_expression.u.nested_exp = $2;
+ $$ = $2;
}
| postfix_expression LSBRAC unary_expression RSBRAC
{
$$ = make_node(scanner, NODE_ENUMERATOR);
$$->u.enumerator.id = $1->s;
}
- | STRING_LITERAL_START DQUOTE
+ | STRING_LITERAL
{
$$ = make_node(scanner, NODE_ENUMERATOR);
- $$->u.enumerator.id = "";
- }
- | STRING_LITERAL_START s_char_sequence DQUOTE
- {
- $$ = make_node(scanner, NODE_ENUMERATOR);
- $$->u.enumerator.id = $2->s;
+ $$->u.enumerator.id = $1->s;
}
| IDENTIFIER EQUAL unary_expression_or_range
{
$$->u.enumerator.id = $1->s;
bt_list_splice(&($3)->tmp_head, &($$)->u.enumerator.values);
}
- | STRING_LITERAL_START DQUOTE EQUAL unary_expression_or_range
+ | STRING_LITERAL EQUAL unary_expression_or_range
{
$$ = make_node(scanner, NODE_ENUMERATOR);
- $$->u.enumerator.id = "";
- bt_list_splice(&($4)->tmp_head, &($$)->u.enumerator.values);
- }
- | STRING_LITERAL_START s_char_sequence DQUOTE EQUAL unary_expression_or_range
- {
- $$ = make_node(scanner, NODE_ENUMERATOR);
- $$->u.enumerator.id = $2->s;
- bt_list_splice(&($5)->tmp_head, &($$)->u.enumerator.values);
+ $$->u.enumerator.id = $1->s;
+ bt_list_splice(&($3)->tmp_head, &($$)->u.enumerator.values);
}
;