int ft_atoi(const char *str)
{
int i;
int mult;
int ret;
if (!str)
return (0);
else if (skip_space(str) < 0)
return (0);
i = skip_space(str);
ret = 0;
mult = 1;
while (str[i] && is_char(str[i]) == 0 && i >= 0)
{
ret += ((str[i] - 48) * mult);
mult *= 10;
i--;
}
if (is_char(str[i]) >= 1 && is_char(str[i - 1]) == 1)
return (0);
else if (str[i] == '-' && ret > 0)
ret *= -1;
else if (str[i] == '\200')
return (0);
return (ret);
}
INTPTR do_phrases(utterance_struct &u) {
phrase_struct p;
CFSWString res;
p.phone_c = 0;
p.syl_c = 0;
p.word_c = 0;
for (INTPTR i = 0; i < u.s.GetLength(); i++) {
CFSWString c = u.s.GetAt(i);
CFSWString pc = res.GetAt(res.GetLength() - 1);
CFSWString nc = u.s.GetAt(i + 1);
CFSWString nnc = u.s.GetAt(i + 2);
if ((is_comma(c) || is_colon(c) || is_semicolon(c)) && is_space(nc) && is_char(nnc)) {
res.Trim();
if (res.GetLength() > 0) {
push_ph_res(u, p, res);
}
} else
if (is_bbracket(c)) {
res.Trim();
if (res.GetLength() > 0) {
push_ph_res(u, p, res);
}
p.s = L"sulgudes";
u.phr_vector.AddItem(p);
} else
if (is_ebracket(c)) {
res.Trim();
if (res.GetLength() > 0) {
push_ph_res(u, p, res);
}
} else
if (is_space(c)) { // komatud sidesõnad
CFSWString tempm = u.s.Mid(i + 1, -1);
res.Trim();
if (is_conju(tempm.Left(tempm.Find(sp))) && res.GetLength() > 0) {
push_ph_res(u, p, res);
} else
res += c;
} else
if (is_bhyphen(c)) {
res.Trim();
if (res.GetLength() > 0 && ((is_char(pc) && is_space(nc)) || (is_space(nc) && is_char(nnc)) || (is_space(pc) && is_char(nc)))) {
push_ph_res(u, p, res);
} else
res += c;
} else
res += c;
}
if (res.GetLength() > 0) {
// if (is_ending(res.GetAt(res.GetLength() - 1))) {
// res.Delete(res.GetLength() - 1, 1);
// }
push_ph_res(u, p, res);
}
return u.phr_vector.GetSize();
}
CFSWString DealWithText(CFSWString text) {
/* Proovin kogu sõnniku minema loopida */
CFSWString res;
text.Trim();
text.Replace(L"\n\n", L"\n", 1);
text.Replace(L"‘", L"'", 1);
text.Replace(L"`", L"'", 1);
text.Replace(L"´", L"'", 1);
text.Replace(L"’", L"'", 1);
for (INTPTR i = 0; i < text.GetLength(); i++) {
CFSWString c = text.GetAt(i);
CFSWString pc = res.GetAt(res.GetLength() - 1);
CFSWString nc = text.GetAt(i + 1);
if (c == L"'") {
if (is_vowel(pc))
res += L"q";
else
res += c;
}
else
if (is_char(c)) res += c;
else
if (is_digit(c)) res += c;
else
if (is_hyphen(c) && is_char(pc) && is_char(nc)) res += sp;
else
if (is_symbol(c)) res += c;
else
if (is_colon(c) && !is_colon(pc)) res += c;
else
if (is_bbracket(c) && !is_bbracket(pc)) res += c;
else
if (is_ebracket(c) && is_ending(nc)) res += L"";
else
if (is_ebracket(c) && !is_ebracket(pc)) res += c;
else
if (is_comma(c) && !is_comma(pc)) res += c;
else
if (is_fchar(c)) res += replace_fchar(c);
else
if (is_space(c) && !is_whitespace(pc)) res += c;
else
if (is_break(c) && !is_break(pc)) {
res += c;
} //kahtlane
else
if (is_tab(c) && !is_whitespace(pc)) res += c;
else
if (is_ending(c) && !is_ending(pc) && !is_whitespace(pc)) res += c;
}
res.Trim();
return res;
}
bool is_valid_term(const char* code_line) {
unsigned long find_index=0;
unsigned long find_length=strlen(code_line);
char find_char='\0';
char term_left[VARIABLES_NAME_LENGTH]={0};
char term_right[VARIABLES_NAME_LENGTH]={0};
for (;find_index<find_length;++find_index) {
find_char=*(char*)(code_line+find_index);
if (('+'==find_char) || ('-'==find_char) || ('*'==find_char) || ('/'==find_char) || ('='==find_char) || ('<'==find_char) || ('>'==find_char)) {
memcpy(term_left,code_line,find_index);
memcpy(term_right,code_line+find_index+1,find_length-find_index-1);
if (is_valid_variables_name(term_left)) {
if (is_valid_variables_name(term_right)) {
return true;
} else if (is_number(term_right)) {
return true;
} else if (is_float(term_right)) {
return true;
} else if (is_char(term_right)) {
return true;
} else if (is_string(term_right)) {
return true;
}
}
}
if (('<'==find_char && '='==*(char*)(code_line+find_index+1)) || ('>'==find_char && '='==*(char*)(code_line+find_index+1) || ('<'==find_char && '>'==*(char*)(code_line+find_index+1)))) {
memset(term_left,0,VARIABLES_NAME_LENGTH);
memset(term_right,0,VARIABLES_NAME_LENGTH);
memcpy(term_left,code_line,find_index);
memcpy(term_right,code_line+find_index+2,find_length-find_index-2);
if (is_valid_variables_name(term_left)) {
if (is_valid_variables_name(term_right)) {
return true;
} else if (is_number(term_right)) {
return true;
} else if (is_float(term_right)) {
return true;
} else if (is_char(term_right)) {
return true;
} else if (is_string(term_right)) {
return true;
}
}
return false;
}
}
return false;
}
int
parse_line(FILE *f, char *line,int *row, http_conf *g)
{
char *split;
char *name;
int err;
name = split = strchr(line, '=');
if(split == NULL) return 2;//配置文件错误
while(!is_char(*name)) name++;
if(strncmp(line, "port", 4) == 0){
split ++;
g->port = atoi(split);
if(g->port <= 0 || g->port > 65535)return 3;//port不正确
}
else if(strncmp(line, "web", 3) == 0) {
if((err = set_web(f, g, row))) return err;
g->web_count ++;
}
else if(strncmp(line, "mimetype", 8) == 0){
if((err = set_mimetype(f, g, row))) return err;;
}
else {
return 4;//不能识别的配置节点
}
return 0;
}
STATIC bool skip_whitespace(mp_lexer_t *lex, bool stop_at_newline) {
bool had_physical_newline = false;
while (!is_end(lex)) {
if (is_physical_newline(lex)) {
if (stop_at_newline && lex->nested_bracket_level == 0) {
break;
}
had_physical_newline = true;
next_char(lex);
} else if (is_whitespace(lex)) {
next_char(lex);
} else if (is_char(lex, '#')) {
next_char(lex);
while (!is_end(lex) && !is_physical_newline(lex)) {
next_char(lex);
}
// had_physical_newline will be set on next loop
} else if (is_char_and(lex, '\\', '\n')) {
// line-continuation, so don't set had_physical_newline
next_char(lex);
next_char(lex);
} else {
break;
}
}
return had_physical_newline;
}
int check_domain(char *domain,int check)
{
int len = 64;
int have_dot = 0;
int i;
for (i = 0; domain[i] != '\0';i++)
{
if(is_char(domain[i]))
continue;
if(is_digital(domain[i]))
continue;
if(is_hyphen(domain[i]))
continue;
if(is_dot(domain[i]))
{
have_dot = 1;
continue;
}
/*error!*/
return 0;
}
if (!have_dot)
return 0;
if (i>len || i<3)
return 0;
/*ok*/
return 1;
}
void Printer::print(LispObjRef obj) {
if (is_nil(obj))
output_ << "NIL";
else if (is_fixnum(obj))
output_ << get_ctype<FixnumType>(obj); // (CFixnum)(boost::get<FixnumType>(*obj));
else if (is_floatnum(obj))
output_ << get_ctype<FloatnumType>(obj); //(CFloatnum)(boost::get<FloatnumType>(*obj));
else if (is_string(obj))
output_ << "\"" << get_ctype<StringType>(obj) << "\""; // ""(CString)(boost::get<StringType>(*obj)) << "\"";
else if (is_symbol(obj))
output_ << get_ctype<SymbolType>(obj).name; // static_cast<LispSymbol>(boost::get<SymbolType>(*obj)).first;
else if (is_cons(obj)) {
output_ << "(";
print_cons(obj);
output_ << ")";
} else if (is_char(obj)) {
CChar c = get_ctype<CharType>(obj);
if (isprint(c)) {
output_ << c;
} else {
output_ << "#" << std::hex << (int) c << std::dec;
}
}
else
output_ << "#UNPRINTABLE#";
}
static obj_t make_string_fn(obj_t args, Reporter rep)
{
size_t nargs, i, k;
char ch;
string str;
nargs = list_length(args);
if (nargs == 0 || nargs > 2) {
reportf(rep, "make-string: "
"length and optional fill char expected");
return unspecific;
}
if (!is_num(list_ref(args, 0))) {
reportf(rep, "make-string: "
"first argument must be a non-negative integer");
return unspecific;
}
if (nargs == 2 && !is_char(list_ref(args, 1))) {
reportf(rep, "make-string: "
"second argument must be a character");
return unspecific;
}
k = fetch_num(list_ref(args, 0));
ch = (nargs == 1) ? 0 : fetch_char(list_ref(args, 1));
str = string_alloc(k);
for (i = 0; i < str->len; i++)
str->data[i] = ch;
return make_string(str);
}
void print(Value x)
{
if (is_nil(x))
prints("nil");
else if (is_eof(x))
printf("#eof");
else if (is_fixnum(x))
printf("%d", as_fixnum(x));
else if (is_bool(x))
printf("%s", as_bool(x) ? "true" : "false");
else if (is_char(x))
printf("'%c'", as_char(x));
else if (is_pair(x))
print_list(x);
else if (is_symbol(x))
prints(as_symbol(x)->value);
else if (is_string(x))
print_string(as_string(x));
else if (is_procedure(x))
printf("#<procedure %s>", as_procedure(x)->name->value);
else if (is_module(x))
printf("#<module>");
else if (is_type(x))
printf("#<type %s>", as_type(x)->name->value);
else if (is_ptr(x))
printf("#<object %p>", as_ptr(x));
else if (is_undefined(x))
printf("#undefined");
else
printf("#ufo");
}
calcu_type is_valid_calculate(const char* in_code_line) {
unsigned long find_flag_index=0;
unsigned long string_length=strlen(in_code_line);
char calculate_left[VARIABLES_NAME_LENGTH]={0};
char calculate_right[VARIABLES_NAME_LENGTH]={0};
calcu_type return_result=calcu_err;
for (;find_flag_index<string_length-1;++find_flag_index) {
if ('+'==*(char*)(in_code_line+find_flag_index))
return_result=calcu_add;
else if ('-'==*(char*)(in_code_line+find_flag_index))
return_result=calcu_dec;
else if ('*'==*(char*)(in_code_line+find_flag_index))
return_result=calcu_mul;
else if ('/'==*(char*)(in_code_line+find_flag_index))
return_result=calcu_div;
if (return_result) {
memcpy(calculate_left,in_code_line,find_flag_index-1);
memcpy(calculate_right,(const void*)(in_code_line+find_flag_index+1),string_length-find_flag_index-1);
if (is_valid_variables_name(calculate_left)) {
if (is_valid_variables_name(calculate_right))
return return_result;
else if (is_number(calculate_right))
return return_result;
else if (is_float(calculate_right))
return return_result;
else if (is_char(calculate_right))
return return_result;
else if (is_string(calculate_right))
return return_result;
return calcu_err;
} else if (is_number(calculate_left) && is_number(calculate_right))
return return_result;
else if (is_float(calculate_left) && is_float(calculate_right))
return return_result;
else if (is_char(calculate_left) && is_char(calculate_right) && (calcu_add==return_result || calcu_dec==return_result))
return return_result;
else if (is_string(calculate_left) && is_string(calculate_right) && calcu_add==return_result)
return return_result;
return calcu_err;
}
}
return calcu_err;
}
//////////////////////////////////////////////////////
//self_evaluating
//number, boolean, string, character and vector object is self-evaluating.
//////////////////////////////////////////////////////
static cellpoint is_self_evaluating(cellpoint exp)
{
if (is_true(is_number(exp)) || is_true(is_boolean(exp)) ||
is_true(is_null(exp)) || is_true(is_string(exp)) ||
is_true(is_char(exp)) || is_true(is_vector(exp))){
reg = a_true;
}else {
reg = a_false;
}
return reg;
}
char *
rtrim_char(char *s, int(*is_char)(int))
{
if(!s || strlen(s) == 0)
return s;
char* back = s + strlen(s);
while(back > s && is_char(*--back));
*(back+1) = '\0';
return s;
}
char *
ltrim_char(char *sx, int(*is_char)(int))
{
size_t len = strlen(sx);
char *s = sx;
if(!s)
return s;
while( (s - sx) <= len && is_char(*s))
s++;
return s;
}
static int skip_space(const char *str)
{
int i;
i = 0;
if (str[i] == '\0')
return (0);
while (str[i] && is_char(str[i]) >= 1)
i++;
if (i > 1 && is_char(str[i]) == 2)
i--;
if (is_char(str[i]) == 1 && ft_isdigit((int)str[i + 1]) == 1)
i++;
while (str[i] && ft_isdigit((int)str[i]) == 1)
i++;
if (i > 0)
i--;
if (ft_isalpha(str[i]) == 1)
i--;
return (i);
}
int check_name (char *name,int check)
{
int len = 32;
int i;
int ret = 1;
if (!is_char(name[0]))
return 0;
if (check)
{
i = strlen(name);
if (i>len)
return 0;
return 1;
}
for (i = 1; name[i] != '\0';i++)
{
if(is_char(name[i]))
continue;
if(is_digital(name[i]))
continue;
if(is_underline(name[i]))
continue;
if(is_dot(name[i]))
continue;
/*error*/
return 0;
}
if (i>len)
return 0;
/*ok */
return 1;
}
int main(int argc, char **argv)
{
unsigned int ch = 0x666;
assert(is_digit(ch) == true);
assert(is_base_char(ch) == false);
assert(is_char(ch) == true);
assert(is_extender(ch) == false);
assert(is_combining_char(ch) == false);
assert(is_ideographic(ch) == false);
ch = 0xf40;
assert(is_digit(ch) == false);
assert(is_base_char(ch) == true);
assert(is_char(ch) == true);
assert(is_extender(ch) == false);
assert(is_combining_char(ch) == false);
assert(is_ideographic(ch) == false);
printf("The test pass.\n");
return 0;
}
// quoted-pair = "\" CHAR
const char* parse_quoted_pair(unsigned char **p)
{
if (parse_char(p, '\\'))
return ERR;
if (is_char(**p))
(*p)++;
else {
(*p)--;
return ERR;
}
return NULL;
}
// token = 1*<any CHAR except CTLs or separators>
const char* parse_token(unsigned char **p)
{
int len = 0;
while (1) {
if (is_char(**p) && !(is_ctl(**p) || is_separator(**p))) {
(*p)++;
len++;
continue;
} else {
if (len > 0) {
return NULL;
}
return ERR;
}
}
}
void lexA(){
f.open(file, ifstream::in);
skip_blanks();
char peek = f.peek();
if( is_number(peek) ){
get_number(peek);
}else if( is_char(peek) ){
check_keyword(peek);
}else if( is_special(peek) ){
handle_special(peek);
}else{
get_identifier();
}
}
void print_ptr_rec(ptr x) {
/*printf("%u\n", x);*/
if (is_fixnum(x)) {
printf("%d", to_fixnum(x));
} else if (x == bool_f) {
printf("#f");
} else if (x == bool_t) {
printf("#t");
} else if (is_null(x)) {
print_null();
} else if (is_char(x)) {
printf("%s", beautify(to_char(x)));
} else if (is_pair(x)) {
printf("(");
print_pair(x);
printf(")");
} else {
printf("#<unknown 0x%08x>", x);
}
}
static void print_atom(FILE *out, obj_t obj)
{
if (eq(obj, unspecific)) {
fputs("#<unspecified>", out);
} else if (is_null(obj)) {
fputs("()", out);
} else if (is_bool(obj)) {
if (fetch_bool(obj))
fputs("#t", out);
else
fputs("#f", out);
} else if (is_symbol(obj)) {
print_string(out, fetch_symbol(obj));
} else if (is_num(obj)) {
fprintf(out, "%ld", (long)fetch_num(obj));
} else if (is_char(obj)) {
char ch = fetch_char(obj);
switch (ch) {
case ' ':
fputs("#\\space", out);
break;
case '\n':
fputs("#\\newline", out);
break;
default:
fprintf(out, "#\\%c", ch);
break;
}
} else if (is_string(obj)) {
putc('"', out);
print_string(out, fetch_string(obj));
putc('"', out);
} else if (is_pair(obj)) {
if (!print_quotation(out, obj))
print_list(out, obj);
} else {
fputs("#<unknown>", out); /* TODO: function, lambda */
}
}
int main(void)
{
unsigned char size = 0xff;
char *beg, *end, str[ size ];
//char *x = "MynameisAlexI'm29yearsoldIwasbornin1983";
char *x = "1913-isbestyearofRussiain20century";
strcpy ( str, x );
str[ strlen ( x ) ] = '\0';
beg = &str;end = &str[ strlen ( x ) - 1 ];
while ( beg != end )
{
if ( !is_digit( *beg ) )
{
beg++;
}
else if ( !is_char( *end ) )
{
end--;
}
else
{
swap (beg, end);
beg++;end--;
}
}
printf ( "%s\n", x );
printf ( "%s\n", str );
return 0;
}
/* [email protected] * buffer = "*****@*****.**" * at = "@akaedu.com" * we need get "akaedu.com" */ char * search_domain(char *at) { char *tmp; tmp = at; tmp ++; for (;*tmp != '\0';tmp++ ) { if(is_char(*tmp)) continue; if(is_digital(*tmp)) continue; if(is_hyphen(*tmp)) continue; if(is_dot(*tmp)) continue; return tmp; } return tmp; }
/* [email protected] * buffer = "*****@*****.**" * at = "@akaedu.com" * we need get "tom" */ char * search_name(char *at,char * buffer) { char *tmp; tmp = at; tmp --; for (;*tmp != '\0', tmp != buffer;tmp-- ) { if(is_char(*tmp)) continue; if(is_digital(*tmp)) continue; if(is_underline(*tmp)) continue; if(is_dot(*tmp)) continue; tmp ++; return tmp; } return tmp; }
char_event_t input_common_readch() {
if (auto mc = lookahead_pop_evt()) {
return *mc;
}
wchar_t res;
mbstate_t state = {};
while (1) {
auto evt = readb();
if (!evt.is_char()) {
return evt;
}
wint_t b = evt.get_char();
if (MB_CUR_MAX == 1) {
return b; // single-byte locale, all values are legal
}
char bb = b;
size_t sz = std::mbrtowc(&res, &bb, 1, &state);
switch (sz) {
case (size_t)(-1): {
std::memset(&state, '\0', sizeof(state));
debug(2, L"Illegal input");
return char_event_type_t::check_exit;
}
case (size_t)(-2): {
break;
}
case 0: {
return 0;
}
default: { return res; }
}
}
}
boost::tribool request_parser::consume(request& req, char input)
{
switch (state_)
{
case method_start:
if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
state_ = method;
req.method.push_back(input);
return boost::indeterminate;
}
case method:
if (input == ' ')
{
state_ = uri;
return boost::indeterminate;
}
else if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
req.method.push_back(input);
return boost::indeterminate;
}
case uri:
if (input == ' ')
{
state_ = http_version_h;
return boost::indeterminate;
}
else if (is_ctl(input))
{
return false;
}
else
{
req.uri.push_back(input);
return boost::indeterminate;
}
case http_version_h:
if (input == 'H')
{
state_ = http_version_t_1;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_t_1:
if (input == 'T')
{
state_ = http_version_t_2;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_t_2:
if (input == 'T')
{
state_ = http_version_p;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_p:
if (input == 'P')
{
state_ = http_version_slash;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_slash:
if (input == '/')
{
req.http_version_major = 0;
req.http_version_minor = 0;
state_ = http_version_major_start;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_major_start:
if (is_digit(input))
{
req.http_version_major = req.http_version_major * 10 + input - '0';
state_ = http_version_major;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_major:
if (input == '.')
{
state_ = http_version_minor_start;
return boost::indeterminate;
}
else if (is_digit(input))
{
req.http_version_major = req.http_version_major * 10 + input - '0';
return boost::indeterminate;
}
else
{
return false;
}
case http_version_minor_start:
if (is_digit(input))
{
req.http_version_minor = req.http_version_minor * 10 + input - '0';
state_ = http_version_minor;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_minor:
if (input == '\r')
{
state_ = expecting_newline_1;
return boost::indeterminate;
}
else if (is_digit(input))
{
req.http_version_minor = req.http_version_minor * 10 + input - '0';
return boost::indeterminate;
}
else
{
return false;
}
case expecting_newline_1:
if (input == '\n')
{
state_ = header_line_start;
return boost::indeterminate;
}
else
{
return false;
}
case header_line_start:
if (input == '\r')
{
state_ = expecting_newline_3;
return boost::indeterminate;
}
else if (!req.headers.empty() && (input == ' ' || input == '\t'))
{
state_ = header_lws;
return boost::indeterminate;
}
else if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
req.headers.push_back(header());
req.headers.back().name.push_back(input);
state_ = header_name;
return boost::indeterminate;
}
case header_lws:
if (input == '\r')
{
state_ = expecting_newline_2;
return boost::indeterminate;
}
else if (input == ' ' || input == '\t')
{
return boost::indeterminate;
}
else if (is_ctl(input))
{
return false;
}
else
{
state_ = header_value;
req.headers.back().value.push_back(input);
return boost::indeterminate;
}
case header_name:
if (input == ':')
{
state_ = space_before_header_value;
return boost::indeterminate;
}
else if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
req.headers.back().name.push_back(input);
return boost::indeterminate;
}
case space_before_header_value:
if (input == ' ')
{
state_ = header_value;
return boost::indeterminate;
}
else
{
return false;
}
case header_value:
if (input == '\r')
{
state_ = expecting_newline_2;
return boost::indeterminate;
}
else if (is_ctl(input))
{
return false;
}
else
{
req.headers.back().value.push_back(input);
return boost::indeterminate;
}
case expecting_newline_2:
if (input == '\n')
{
state_ = header_line_start;
return boost::indeterminate;
}
else
{
return false;
}
case expecting_newline_3:
return (input == '\n');
default:
return false;
}
}
Type type_check_expr(int oper,Type t1,Type t2)
{
switch(oper)
{
case T_EQU:
case T_NEQ:
if( (is_char(t1) && is_char(t2) ) ||
(is_string(t1) && is_string(t2) ) ||
(is_numeric(t1) && is_numeric(t2)) ||
(is_boolean(t1) && is_boolean(t2)) )
{
return BOOLEAN_TYPE;
}
break;
case T_LES:
case T_LEQ:
case T_GTR:
case T_GTE:
if( (is_char(t1) && is_char(t2) ) ||
(is_string(t1) && is_string(t2) ) ||
(is_numeric(t1) && is_numeric(t2)) )
{
return BOOLEAN_TYPE;
}
break;
case T_ADD:
case T_SUB:
case T_MUL:
if( is_numeric(t1) && is_numeric(t2) )
{
if(t1>t2) return t1;
else return t2;
}
break;
case T_NEG:
if( is_numeric(t1) )
{
return t1;
}
break;
case T_RDIV:
if( is_numeric(t1) && is_numeric(t2) )
{
return REAL_TYPE;
}
break;
case T_MOD:
case T_IDIV:
if( is_integer(t1) && is_integer(t2) )
{
return INTEGER_TYPE;
}
break;
case T_OR:
case T_AND:
if( is_boolean(t1) && is_boolean(t2) )
{
return BOOLEAN_TYPE;
}
break;
case T_NOT:
if( is_boolean(t1) )
{
return BOOLEAN_TYPE;
}
break;
}
fatal_error("invalid types in expression\n");
return INTEGER_TYPE;
}
boost::tribool request_parser::consume(request& req, char input)
{
switch (state_)
{
case method_start:
if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
state_ = method;
req.method.push_back(input);
return boost::indeterminate;
}
case method:
if (input == ' ')
{
state_ = uri;
return boost::indeterminate;
}
else if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
req.method.push_back(input);
return boost::indeterminate;
}
case uri_start:
if (is_ctl(input))
{
return false;
}
else
{
state_ = uri;
req.uri.push_back(input);
return boost::indeterminate;
}
case uri:
if (input == ' ')
{
state_ = http_version_h;
return boost::indeterminate;
}
else if (is_ctl(input))
{
return false;
}
else
{
req.uri.push_back(input);
return boost::indeterminate;
}
case http_version_h:
if (input == 'H')
{
state_ = http_version_t_1;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_t_1:
if (input == 'T')
{
state_ = http_version_t_2;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_t_2:
if (input == 'T')
{
state_ = http_version_p;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_p:
if (input == 'P')
{
state_ = http_version_slash;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_slash:
if (input == '/')
{
req.http_version_major = 0;
req.http_version_minor = 0;
state_ = http_version_major_start;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_major_start:
if (is_digit(input))
{
req.http_version_major = req.http_version_major * 10 + input - '0';
state_ = http_version_major;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_major:
if (input == '.')
{
state_ = http_version_minor_start;
return boost::indeterminate;
}
else if (is_digit(input))
{
req.http_version_major = req.http_version_major * 10 + input - '0';
return boost::indeterminate;
}
else
{
return false;
}
case http_version_minor_start:
if (is_digit(input))
{
req.http_version_minor = req.http_version_minor * 10 + input - '0';
state_ = http_version_minor;
return boost::indeterminate;
}
else
{
return false;
}
case http_version_minor:
if (input == '\r')
{
state_ = expecting_newline_1;
return boost::indeterminate;
}
else if (is_digit(input))
{
req.http_version_minor = req.http_version_minor * 10 + input - '0';
return boost::indeterminate;
}
else
{
return false;
}
case expecting_newline_1:
if (input == '\n')
{
state_ = header_line_start;
return boost::indeterminate;
}
else
{
return false;
}
case header_line_start:
if (input == '\r')
{
state_ = expecting_newline_3;
return boost::indeterminate;
}
else if (!req.headers.empty() && (input == ' ' || input == '\t'))
{
state_ = header_lws;
return boost::indeterminate;
}
else if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
req.headers.push_back(header());
req.headers.back().name.push_back(input);
state_ = header_name;
return boost::indeterminate;
}
case header_lws:
if (input == '\r')
{
state_ = expecting_newline_2;
return boost::indeterminate;
}
else if (input == ' ' || input == '\t')
{
return boost::indeterminate;
}
else if (is_ctl(input))
{
return false;
}
else
{
state_ = header_value;
req.headers.back().value.push_back(input);
return boost::indeterminate;
}
case header_name:
if (input == ':')
{
state_ = space_before_header_value;
return boost::indeterminate;
}
else if (!is_char(input) || is_ctl(input) || is_tspecial(input))
{
return false;
}
else
{
req.headers.back().name.push_back(input);
return boost::indeterminate;
}
case space_before_header_value:
if (input == ' ')
{
state_ = header_value;
return boost::indeterminate;
}
else
{
return false;
}
case header_value:
if (input == '\r')
{
state_ = expecting_newline_2;
return boost::indeterminate;
}
else if (is_ctl(input))
{
return false;
}
else
{
req.headers.back().value.push_back(input);
return boost::indeterminate;
}
case expecting_newline_2:
if (input == '\n')
{
state_ = header_line_start;
std::vector<header>::const_iterator cit = req.headers.begin();
for (; cit != req.headers.end(); ++cit)
{
std::string n = (*cit).name;
std::transform(n.begin(), n.end(), n.begin(), ::toupper);
if (n == UPPER_CONTENT_LENGTH)
{
try
{
cl_ = boost::lexical_cast<int>((*cit).value);
}
catch (const boost::bad_lexical_cast& e)
{
return false;
}
}
else if (n == UPPER_CONTENT_TYPE)
{
std::string m = (*cit).value;
std::transform(m.begin(), m.end(), m.begin(), ::toupper);
if (m != UPPER_MIME_TYPE)
{
return false;
}
}
}
return boost::indeterminate;
}
else
{
return false;
}
case expecting_newline_3:
{
if (0 == cl_)
{
return (input == '\n');
}
else
{
state_ = post_data;
return boost::indeterminate;
}
}
case post_data:
{
req.post_data.push_back(input);
if (0 == --cl_)
return true;
else
return ((cl_ < 0) ? false : boost::indeterminate);
}
default:
return false;
}
}
STATIC void mp_lexer_next_token_into(mp_lexer_t *lex, mp_token_t *tok, bool first_token) {
// skip white space and comments
bool had_physical_newline = false;
while (!is_end(lex)) {
if (is_physical_newline(lex)) {
had_physical_newline = true;
next_char(lex);
} else if (is_whitespace(lex)) {
next_char(lex);
} else if (is_char(lex, '#')) {
next_char(lex);
while (!is_end(lex) && !is_physical_newline(lex)) {
next_char(lex);
}
// had_physical_newline will be set on next loop
} else if (is_char(lex, '\\')) {
// backslash (outside string literals) must appear just before a physical newline
next_char(lex);
if (!is_physical_newline(lex)) {
// SyntaxError: unexpected character after line continuation character
tok->src_line = lex->line;
tok->src_column = lex->column;
tok->kind = MP_TOKEN_BAD_LINE_CONTINUATION;
vstr_reset(&lex->vstr);
tok->str = vstr_str(&lex->vstr);
tok->len = 0;
return;
} else {
next_char(lex);
}
} else {
break;
}
}
// set token source information
tok->src_line = lex->line;
tok->src_column = lex->column;
// start new token text
vstr_reset(&lex->vstr);
if (first_token && lex->line == 1 && lex->column != 1) {
// check that the first token is in the first column
// if first token is not on first line, we get a physical newline and
// this check is done as part of normal indent/dedent checking below
// (done to get equivalence with CPython)
tok->kind = MP_TOKEN_INDENT;
} else if (lex->emit_dent < 0) {
tok->kind = MP_TOKEN_DEDENT;
lex->emit_dent += 1;
} else if (lex->emit_dent > 0) {
tok->kind = MP_TOKEN_INDENT;
lex->emit_dent -= 1;
} else if (had_physical_newline && lex->nested_bracket_level == 0) {
tok->kind = MP_TOKEN_NEWLINE;
uint num_spaces = lex->column - 1;
lex->emit_dent = 0;
if (num_spaces == indent_top(lex)) {
} else if (num_spaces > indent_top(lex)) {
indent_push(lex, num_spaces);
lex->emit_dent += 1;
} else {
while (num_spaces < indent_top(lex)) {
indent_pop(lex);
lex->emit_dent -= 1;
}
if (num_spaces != indent_top(lex)) {
tok->kind = MP_TOKEN_DEDENT_MISMATCH;
}
}
} else if (is_end(lex)) {
if (indent_top(lex) > 0) {
tok->kind = MP_TOKEN_NEWLINE;
lex->emit_dent = 0;
while (indent_top(lex) > 0) {
indent_pop(lex);
lex->emit_dent -= 1;
}
} else {
tok->kind = MP_TOKEN_END;
}
} else if (is_char_or(lex, '\'', '\"')
|| (is_char_or3(lex, 'r', 'u', 'b') && is_char_following_or(lex, '\'', '\"'))
|| ((is_char_and(lex, 'r', 'b') || is_char_and(lex, 'b', 'r')) && is_char_following_following_or(lex, '\'', '\"'))) {
// a string or bytes literal
// parse type codes
bool is_raw = false;
bool is_bytes = false;
if (is_char(lex, 'u')) {
next_char(lex);
} else if (is_char(lex, 'b')) {
is_bytes = true;
next_char(lex);
if (is_char(lex, 'r')) {
is_raw = true;
next_char(lex);
}
} else if (is_char(lex, 'r')) {
is_raw = true;
next_char(lex);
if (is_char(lex, 'b')) {
is_bytes = true;
next_char(lex);
}
}
// set token kind
if (is_bytes) {
tok->kind = MP_TOKEN_BYTES;
} else {
tok->kind = MP_TOKEN_STRING;
}
// get first quoting character
char quote_char = '\'';
if (is_char(lex, '\"')) {
quote_char = '\"';
}
next_char(lex);
// work out if it's a single or triple quoted literal
int num_quotes;
if (is_char_and(lex, quote_char, quote_char)) {
// triple quotes
next_char(lex);
next_char(lex);
num_quotes = 3;
} else {
// single quotes
num_quotes = 1;
}
// parse the literal
int n_closing = 0;
while (!is_end(lex) && (num_quotes > 1 || !is_char(lex, '\n')) && n_closing < num_quotes) {
if (is_char(lex, quote_char)) {
n_closing += 1;
vstr_add_char(&lex->vstr, CUR_CHAR(lex));
} else {
n_closing = 0;
if (is_char(lex, '\\')) {
next_char(lex);
unichar c = CUR_CHAR(lex);
if (is_raw) {
// raw strings allow escaping of quotes, but the backslash is also emitted
vstr_add_char(&lex->vstr, '\\');
} else {
switch (c) {
case MP_LEXER_CHAR_EOF: break; // TODO a proper error message?
case '\n': c = MP_LEXER_CHAR_EOF; break; // TODO check this works correctly (we are supposed to ignore it
case '\\': break;
case '\'': break;
case '"': break;
case 'a': c = 0x07; break;
case 'b': c = 0x08; break;
case 't': c = 0x09; break;
case 'n': c = 0x0a; break;
case 'v': c = 0x0b; break;
case 'f': c = 0x0c; break;
case 'r': c = 0x0d; break;
case 'u':
case 'U':
if (is_bytes) {
// b'\u1234' == b'\\u1234'
vstr_add_char(&lex->vstr, '\\');
break;
}
// Otherwise fall through.
case 'x':
{
uint num = 0;
if (!get_hex(lex, (c == 'x' ? 2 : c == 'u' ? 4 : 8), &num)) {
// TODO error message
assert(0);
}
c = num;
break;
}
case 'N':
// Supporting '\N{LATIN SMALL LETTER A}' == 'a' would require keeping the
// entire Unicode name table in the core. As of Unicode 6.3.0, that's nearly
// 3MB of text; even gzip-compressed and with minimal structure, it'll take
// roughly half a meg of storage. This form of Unicode escape may be added
// later on, but it's definitely not a priority right now. -- CJA 20140607
assert(!"Unicode name escapes not supported");
break;
default:
if (c >= '0' && c <= '7') {
// Octal sequence, 1-3 chars
int digits = 3;
int num = c - '0';
while (is_following_odigit(lex) && --digits != 0) {
next_char(lex);
num = num * 8 + (CUR_CHAR(lex) - '0');
}
c = num;
} else {
// unrecognised escape character; CPython lets this through verbatim as '\' and then the character
vstr_add_char(&lex->vstr, '\\');
}
break;
}
}
if (c != MP_LEXER_CHAR_EOF) {
if (c < 0x110000 && !is_bytes) {
vstr_add_char(&lex->vstr, c);
} else if (c < 0x100 && is_bytes) {
vstr_add_byte(&lex->vstr, c);
} else {
assert(!"TODO: Throw an error, invalid escape code probably");
}
}
} else {
vstr_add_char(&lex->vstr, CUR_CHAR(lex));
}
}
next_char(lex);
}
// check we got the required end quotes
if (n_closing < num_quotes) {
tok->kind = MP_TOKEN_LONELY_STRING_OPEN;
}
// cut off the end quotes from the token text
vstr_cut_tail_bytes(&lex->vstr, n_closing);
} else if (is_head_of_identifier(lex)) {
tok->kind = MP_TOKEN_NAME;
// get first char
vstr_add_char(&lex->vstr, CUR_CHAR(lex));
next_char(lex);
// get tail chars
while (!is_end(lex) && is_tail_of_identifier(lex)) {
vstr_add_char(&lex->vstr, CUR_CHAR(lex));
next_char(lex);
}
} else if (is_digit(lex) || (is_char(lex, '.') && is_following_digit(lex))) {
tok->kind = MP_TOKEN_NUMBER;
// get first char
vstr_add_char(&lex->vstr, CUR_CHAR(lex));
next_char(lex);
// get tail chars
while (!is_end(lex)) {
if (is_char_or(lex, 'e', 'E')) {
vstr_add_char(&lex->vstr, 'e');
next_char(lex);
if (is_char(lex, '+') || is_char(lex, '-')) {
vstr_add_char(&lex->vstr, CUR_CHAR(lex));
next_char(lex);
}
} else if (is_letter(lex) || is_digit(lex) || is_char_or(lex, '_', '.')) {
vstr_add_char(&lex->vstr, CUR_CHAR(lex));
next_char(lex);
} else {
break;
}
}
} else if (is_char(lex, '.')) {
// special handling for . and ... operators, because .. is not a valid operator
// get first char
vstr_add_char(&lex->vstr, '.');
next_char(lex);
if (is_char_and(lex, '.', '.')) {
vstr_add_char(&lex->vstr, '.');
vstr_add_char(&lex->vstr, '.');
next_char(lex);
next_char(lex);
tok->kind = MP_TOKEN_ELLIPSIS;
} else {
tok->kind = MP_TOKEN_DEL_PERIOD;
}
} else {
// search for encoded delimiter or operator
const char *t = tok_enc;
uint tok_enc_index = 0;
for (; *t != 0 && !is_char(lex, *t); t += 1) {
if (*t == 'e' || *t == 'c') {
t += 1;
} else if (*t == 'E') {
tok_enc_index -= 1;
t += 1;
}
tok_enc_index += 1;
}
next_char(lex);
if (*t == 0) {
// didn't match any delimiter or operator characters
tok->kind = MP_TOKEN_INVALID;
} else {
// matched a delimiter or operator character
// get the maximum characters for a valid token
t += 1;
uint t_index = tok_enc_index;
for (;;) {
for (; *t == 'e'; t += 1) {
t += 1;
t_index += 1;
if (is_char(lex, *t)) {
next_char(lex);
tok_enc_index = t_index;
break;
}
}
if (*t == 'E') {
t += 1;
if (is_char(lex, *t)) {
next_char(lex);
tok_enc_index = t_index;
} else {
tok->kind = MP_TOKEN_INVALID;
goto tok_enc_no_match;
}
break;
}
if (*t == 'c') {
t += 1;
t_index += 1;
if (is_char(lex, *t)) {
next_char(lex);
tok_enc_index = t_index;
t += 1;
} else {
break;
}
} else {
break;
}
}
// set token kind
tok->kind = tok_enc_kind[tok_enc_index];
tok_enc_no_match:
// compute bracket level for implicit line joining
if (tok->kind == MP_TOKEN_DEL_PAREN_OPEN || tok->kind == MP_TOKEN_DEL_BRACKET_OPEN || tok->kind == MP_TOKEN_DEL_BRACE_OPEN) {
lex->nested_bracket_level += 1;
} else if (tok->kind == MP_TOKEN_DEL_PAREN_CLOSE || tok->kind == MP_TOKEN_DEL_BRACKET_CLOSE || tok->kind == MP_TOKEN_DEL_BRACE_CLOSE) {
lex->nested_bracket_level -= 1;
}
}
}
// point token text to vstr buffer
tok->str = vstr_str(&lex->vstr);
tok->len = vstr_len(&lex->vstr);
// check for keywords
if (tok->kind == MP_TOKEN_NAME) {
// We check for __debug__ here and convert it to its value. This is so
// the parser gives a syntax error on, eg, x.__debug__. Otherwise, we
// need to check for this special token in many places in the compiler.
// TODO improve speed of these string comparisons
//for (int i = 0; tok_kw[i] != NULL; i++) {
for (int i = 0; i < MP_ARRAY_SIZE(tok_kw); i++) {
if (str_strn_equal(tok_kw[i], tok->str, tok->len)) {
if (i == MP_ARRAY_SIZE(tok_kw) - 1) {
// tok_kw[MP_ARRAY_SIZE(tok_kw) - 1] == "__debug__"
tok->kind = (mp_optimise_value == 0 ? MP_TOKEN_KW_TRUE : MP_TOKEN_KW_FALSE);
} else {
tok->kind = MP_TOKEN_KW_FALSE + i;
}
break;
}
}
}
}
