END_TEST
START_TEST (test_FormulaTokenizer_empty)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula("");
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
fail_unless( t->value.ch == '\0' );
Token_free(t);
/**
* Subsequent calls to FormulaTokenizer_nextToken() should continue to
* return TT_END.
*/
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
fail_unless( t->value.ch == '\0' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
fail_unless( t->value.ch == '\0' );
Token_free(t);
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_FormulaTokenizer_numbers_nan_inf)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula("NaN Inf");
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL );
fail_unless( t->value.real != t->value.real );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL );
fail_unless( util_isInf(t->value.real) == 1 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
Token_free(t);
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_FormulaTokenizer_names)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula("foobar Foo2Bar _Foo_Bar");
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_NAME );
fail_unless( !strcmp(t->value.name, "foobar") );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_NAME );
fail_unless( !strcmp(t->value.name, "Foo2Bar") );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_NAME );
fail_unless( !strcmp(t->value.name, "_Foo_Bar") );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
Token_free(t);
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_FormulaTokenizer_numbers_exp_bug)
{
/* we were failing to correctly parse an exponentional followed
* by a sign with no space
*/
const char *formula = "12.3e1+.314E1";
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula(formula);
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == 12.3 );
fail_unless( t->exponent == 1 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_PLUS );
fail_unless( t->value.ch == '+' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == .314 );
fail_unless( t->exponent == 1 );
Token_free(t);
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_FormulaTokenizer_operators)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula("+-*/^(),");
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_PLUS );
fail_unless( t->value.ch == '+' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_MINUS );
fail_unless( t->value.ch == '-' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_TIMES );
fail_unless( t->value.ch == '*' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_DIVIDE );
fail_unless( t->value.ch == '/' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_POWER );
fail_unless( t->value.ch == '^' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_LPAREN );
fail_unless( t->value.ch == '(' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_RPAREN );
fail_unless( t->value.ch == ')' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_COMMA );
fail_unless( t->value.ch == ',' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless(t->type == TT_END );
fail_unless( t->value.ch == '\0' );
Token_free(t);
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_Token_free_NULL)
{
Token_free(NULL);
}
END_TEST
START_TEST (test_Token_free)
{
Token_t *t = Token_create();
t->type = TT_NAME;
t->value.name = safe_strdup("foo");
Token_free(t);
}
END_TEST
START_TEST (test_FormulaTokenizer_numbers_locale)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula("2.72");
Token_t *t;
setlocale(LC_NUMERIC, "de_DE");
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL );
fail_unless( t->value.real == 2.72 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
Token_free(t);
setlocale(LC_NUMERIC, "C");
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_Token_create)
{
Token_t *t = Token_create();
fail_unless( t->type == TT_UNKNOWN );
fail_unless( t->value.ch == '\0' );
fail_unless( t->value.name == NULL );
fail_unless( t->value.integer == 0 );
fail_unless( t->value.real == 0.0 );
fail_unless( t->exponent == 0 );
Token_free(t);
}
END_TEST
START_TEST (test_FormulaTokenizer_whitespace)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula(" \t \n ");
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
fail_unless( t->value.ch == '\0' );
Token_free(t);
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_FormulaTokenizer_unknown)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula("*****@*****.**");
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_NAME );
fail_unless( !strcmp(t->value.name, "bbornstein") );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_UNKNOWN );
fail_unless( t->value.ch == '@' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_NAME );
fail_unless( !strcmp(t->value.name, "acm") );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_UNKNOWN );
fail_unless( t->value.ch == '.' );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_NAME );
fail_unless( !strcmp(t->value.name, "org") );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
Token_free(t);
FormulaTokenizer_free(ft);
}
END_TEST
START_TEST (test_FormulaTokenizer_numbers)
{
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula("123 3.14 .007 6.7 5.");
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_INTEGER );
fail_unless( t->value.integer == 123 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL );
fail_unless( t->value.real == 3.14 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL );
fail_unless( t->value.real == .007 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL );
fail_unless( t->value.real == 6.7 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL );
fail_unless( t->value.real == 5.0 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
Token_free(t);
FormulaTokenizer_free(ft);
}
CK_CPPSTART
#endif
/**
* The following are private and used only within FormulaParser.c; however,
* I don't know how else to make them "public" only for testing than to put
* them here
*/
#define START_STATE 0
#define ACCEPT_STATE 0
#define ERROR_STATE 27
#define NUM_STATES 27
/**
* The Action[] table contains 144 entries. To test them all would be
* laborious and silly. Instead, for a few token type, test the first,
* last and middle entry in each token "run". Also, test some error
* states.
*/
START_TEST (test_FormulaParser_getAction)
{
int i;
Token_t *t = Token_create();
t->type = TT_NAME;
fail_unless( FormulaParser_getAction( 0, t) == 6, NULL );
fail_unless( FormulaParser_getAction(10, t) == 6, NULL );
fail_unless( FormulaParser_getAction(25, t) == 6, NULL );
fail_unless( FormulaParser_getAction( 1, t) == ERROR_STATE, NULL );
t->type = TT_INTEGER;
fail_unless( FormulaParser_getAction( 0, t) == 1, NULL );
fail_unless( FormulaParser_getAction(10, t) == 1, NULL );
fail_unless( FormulaParser_getAction(25, t) == 1, NULL );
fail_unless( FormulaParser_getAction( 1, t) == ERROR_STATE, NULL );
t->type = TT_PLUS;
fail_unless( FormulaParser_getAction( 1, t) == -9, NULL );
fail_unless( FormulaParser_getAction(16, t) == -2, NULL );
fail_unless( FormulaParser_getAction(24, t) == -11, NULL );
fail_unless( FormulaParser_getAction( 2, t) == ERROR_STATE, NULL );
t->type = TT_MINUS;
fail_unless( FormulaParser_getAction( 0, t) == 5, NULL );
fail_unless( FormulaParser_getAction(16, t) == -2, NULL );
fail_unless( FormulaParser_getAction(25, t) == 5, NULL );
fail_unless( FormulaParser_getAction( 2, t) == ERROR_STATE, NULL );
t->type = TT_END;
fail_unless( FormulaParser_getAction( 1, t) == -9, NULL );
fail_unless( FormulaParser_getAction(17, t) == -5, NULL );
fail_unless( FormulaParser_getAction(24, t) == -11, NULL );
fail_unless( FormulaParser_getAction( 3, t) == ERROR_STATE, NULL );
/**
* TT_UNKNOWN should always yield an error state.
*/
t->type = TT_UNKNOWN;
for (i = 0; i < NUM_STATES; i++)
{
fail_unless( FormulaParser_getAction(i, t) == ERROR_STATE, NULL );
}
Token_free(t);
}
END_TEST
START_TEST (test_FormulaTokenizer_numbers_exp)
{
const char *formula = "12.3e1 .314E1 7e-3 .067e2 5E0 2e+12 3e 4";
FormulaTokenizer_t *ft = FormulaTokenizer_createFromFormula(formula);
Token_t *t;
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == 12.3 );
fail_unless( t->exponent == 1 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == .314 );
fail_unless( t->exponent == 1 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == 7 );
fail_unless( t->exponent == -3 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == .067 );
fail_unless( t->exponent == 2 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == 5.0 );
fail_unless( t->exponent == 0 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == 2 );
fail_unless( t->exponent == 12 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_REAL_E );
fail_unless( t->value.real == 3.0 );
fail_unless( t->exponent == 0 );
Token_free(t);
/**
* Gobble-up the '4'. This last token is here as a test to ensure the
* previous token is not interpreted as 3e4.
*/
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_INTEGER );
fail_unless( t->value.integer == 4 );
Token_free(t);
t = FormulaTokenizer_nextToken(ft);
fail_unless( t->type == TT_END );
Token_free(t);
FormulaTokenizer_free(ft);
}
/**
* @if conly
* @memberof ASTNode_t
* @endif
*/
LIBSBML_EXTERN
ASTNode_t *
SBML_parseFormula (const char *formula)
{
long rule, state, action;
ASTNode_t *node = NULL;
FormulaTokenizer_t *tokenizer = NULL;
Stack_t *stack = NULL;
Token_t *token = NULL;
if (formula == NULL) return NULL;
tokenizer = FormulaTokenizer_createFromFormula(formula);
token = FormulaTokenizer_nextToken(tokenizer);
stack = Stack_create(20);
Stack_push(stack, (void *) START_STATE);
while (1)
{
state = (long) Stack_peek(stack);
action = FormulaParser_getAction(state, token);
if (action == ACCEPT_STATE)
{
node = Stack_peekAt(stack, 1);
break;
}
else if (action == ERROR_STATE)
{
/**
* Free ASTNodes on the Stack, skip the states.
*/
while (Stack_size(stack) > 1)
{
Stack_pop(stack);
ASTNode_free( Stack_pop(stack) );
}
node = NULL;
break;
}
/**
* Shift
*/
else if (action > 0)
{
Stack_push( stack, ASTNode_createFromToken(token) );
Stack_push( stack, (void *) action );
Token_free(token);
token = FormulaTokenizer_nextToken(tokenizer);
}
/**
* Reduce
*/
else if (action < 0)
{
rule = -action;
node = FormulaParser_reduceStackByRule(stack, rule);
state = (long) Stack_peek(stack);
Stack_push( stack, node );
Stack_push( stack, (void *) FormulaParser_getGoto(state, rule) );
}
}
FormulaTokenizer_free(tokenizer);
Stack_free(stack);
Token_free(token);
return node;
}