// ---
bool cmp_ignore_spaces( const char* s1, const char* s2 ) {
if ( !s1 || !*s1 )
return !s2 || !*s2;
else if ( !s2 || !*s2 )
return false;
ignore_spaces( s1 );
ignore_spaces( s2 );
while( *s1 && (*s1 == *s2) ) {
ignore_spaces( ++s1 );
ignore_spaces( ++s2 );
}
return (*s1 == 0) && (*s2 == 0);
}
// ---
bool cmp_ignore_spaces_no_comment( const char* s1, const char* s2 ) {
if ( !s1 || !*s1 )
return !s2 || !*s2;
else if ( !s2 || !*s2 )
return false;
ignore_spaces( s1 );
ignore_spaces( s2 );
while( *s1 && (*s1 == *s2) && !is_comment(s1) && !is_comment(s2) ) {
ignore_spaces( ++s1 );
ignore_spaces( ++s2 );
}
return ((*s1 == 0) || (*s1 == '/')) && ((*s2 == 0) || (*s2 == '/'));
}
void parse_program_list(){
ignore_spaces();
parse_program_call();
programs++;
ignore_spaces();
switch(peek_token()){
case '|':
shift_token();
parse_program_list();
case EOS:
break;
default:
break;
}
}
// ---
bool is_comment( const char* s ) {
if ( !s || !*s )
return false;
ignore_spaces( s );
return *s && (*s == '/') && *(s+1) && (*(s+1) == '/' );
}
/* Analyzes a string for the presence of a meta refresh type url.
*
* This function receives the value of the content attribute of a meta tag and
* parses it in order to identify if a url is going to be present. This is the
* format of such tag:
*
* <meta http-equiv="refresh" content="5; URL=http://www.google.com">
*
* Using a regular expression library would be the most obvious way to implement
* this functionality, but introducing such a dependency is undesirable. We
* opted instead to parse programmaticly since the expression is simple enough.
*
* For reference, this is the spec on the meta http refresh tag:
* http://dev.w3.org/html5/spec/Overview.html#attr-meta-http-equiv-refresh
*
* If the value has no content after the expression, we know we are at the start
* of the URL. Otherwise we are past the start of the URL.
*
*
* Returns:
*
* This functions returns one of the following values:
* META_REDIRECT_TYPE_NONE - A url was not identified in the input string.
* META_REDIRECT_TYPE_URL_START - The input string ends exactly at the start
* of the url.
* META_REDIRECT_TYPE_URL - The input string ends somewhere in the middle or
* the end of the url.
*
* A few examples:
* "5"
* Returns META_REDIRECT_TYPE_NONE since we don't expect a url to follow.
*
* "5; URL = "
* The function returns META_REDIRECT_TYPE_URL_START since we expect a url to
* follow.
*
* "5; URL = http://www.google.com/?"
* Returns META_REDIRECT_TYPE_URL since the input value terminates in the
* middle or end of a url.
*
*
* Caveats: We are only recording up to 256 characters of attribute values, so
* our analysis is limited to that. This shouldn't be an issue in practice
* though as it would be unexpected for the part of the string that we are
* matching to be so long.
*/
enum meta_redirect_type_enum meta_redirect_type(const char *value) {
if (value == NULL)
return META_REDIRECT_TYPE_NONE;
/* Match while [ \t\r\n0-9]* */
value = ignore_spaces_or_digits(value);
/* Verify that we got a semi-colon character */
if (*value != ';')
return META_REDIRECT_TYPE_NONE;
value++;
/* Match while [ \t\r\n]* */
value = ignore_spaces(value);
/* Validate that we have 'URL' */
if (strncasecmp(value, "url", strlen("url")) != 0)
return META_REDIRECT_TYPE_NONE;
value += strlen("url");
/* Match while [ \t\r\n]* */
value = ignore_spaces(value);
if (*value != '=')
return META_REDIRECT_TYPE_NONE;
value++;
/* Match while [ \t\r\n]* */
value = ignore_spaces(value);
/* The HTML5 spec allows for the url to be quoted, so we skip a single or
* double quote if we find one.
*/
if (*value == '"' || *value == '\'')
value++;
if (*value == '\0')
return META_REDIRECT_TYPE_URL_START;
else
return META_REDIRECT_TYPE_URL;
}
void parse_program_call(){
set_parse_err(programs >= MAX_PROGRAMS,ETOOMANY);
parse_name();
ignore_spaces();
argument_number = 1;
parse_argument_list();
program_call[programs]
.arguments[argument_number] = NULL;
}
// ---
void uncomment( char* s ) {
ignore_spaces( (const char*&)s );
char* b = s;
if ( (*((WORD*)s) != MAKEWORD('/','/')) )
return;
s += 2;
if ( *s == '!' ) {
s++;
if ( *s == ' ' )
s++;
}
::memmove( b, s, ::lstrlen(s)+1 );
}
void parse_argument_list(){
set_parse_err(argument_number >= MAX_ARGS,EMANYARGS);
ignore_spaces();
if(reserved(peek_token()))
return;
char * arg = malloc(MAX_ARGLEN+1);
program_call[programs]
.arguments[argument_number] = arg;
char c; int i;
for(i = 0, c = next_token();
i < MAX_ARGLEN && !reserved(c);
i++, c = next_token() ){
arg[i] = c;
}
set_parse_err(i >= MAX_ARGLEN,EARGLONG);
arg[i] = '\0';
argument_number++;
ignore_spaces();
switch(c = peek_token()){
case EOS:
case '|':
break;
default:
if(!reserved(c)){
parse_argument_list();
}else{
set_parse_err(true,EINVCHAR);
}
break;
}
}
// ---
bool is_empty( const char* s ) {
if ( !s || !*s )
return true;
ignore_spaces( s );
return !s || !*s;
}
