string
xml_html_parser::transcode (string s2) {
s= parse_string (s2);
string encoding;
if (test (s, "<?")) {
s += 2;
string target= parse_name ();
skip_space ();
if (target == "xml") {
// since html==true implies we can accept horribly broken HTML, the
// presence of an XML prolog is not enough to clear the flag.
/* html= false; */
while (s && !test (s, "?>")) {
string attname= parse_name ();
skip_space ();
if (!test (s, "=")) break;
s += 1;
skip_space ();
string val;
if (test (s, "\"")) {
s += 1;
val= parse_until ("\"");
skip_space ();
}
else if (test (s, "'")) {
s += 1;
val= parse_until ("'");
skip_space ();
}
if (attname == "encoding") {
encoding= upcase_all (val);
break;
}
}
}
}
if (N(encoding) != 0) {
// cout << "encoding was specified\n" ;
string s3= convert (s2, encoding, "UTF-8");
if (N(s3) == 0)
/* conversion from specified charset failed, do nothing (and pray) */ ;
else return s3;
}
else {
// cout << "guess encoding\n" ;
if (check_encoding (s2, "UTF-8"))
/* input encoding seems to be utf-8, do nothing */ ;
else {
string s3= convert (s2, "ISO-8859-1", "UTF-8");
if (N(s3) != 0) return s3;
}
}
return s2;
}
string
xml_html_parser::parse_quoted () {
if (test (s, "\42")) {
s += 1;
return parse_until ("\42");
}
if (test (s, "'")) {
s += 1;
return parse_until ("'");
}
return "";
}
ss_inst *init_insts(stream_t *stream) {
char *buf;
char type;
long IP, new_IP;
new_IP = IP = getpos(stream);
ss_inst *head, *ptr;
ss_inst *indexes[4];
head = ptr = malloc(sizeof(ss_inst));
while (listench(stream)) {
int len = get_parsable_length(stream);
buf = calloc(len + 1, sizeof(char));
getstr(stream, len, buf);
if (buf[len - 1] == '(' && *buf == '?') {
type = '(';
indexes[1] = parse_until(stream, ')');
} else if (*buf == '[') {
type = '[';
indexes[1] = parse_until(stream, ']');
} else if (*buf == '{') {
type = '{';
indexes[1] = parse_until(stream, '}');
} else if (*buf == '?') {
type = '?';
char iden = buf[len - 1];
if (iden == '(') {
indexes[1] = parse_until(stream, ')');/* {expr} */
} else if (strchr(RETN, iden)) {
move_stream(stream, 1);
new_IP += get_parsable_length(stream);
move_stream(stream, -1);
}
if (listench(stream) == '[') type = 'w';
move_stream(stream, new_IP - IP - 1);
// TODO conseq & alt and loop body
} else {
type = 0;
}
indexes[0] = init_inst(type, buf);
if (type) {
for (int i = 1; i < indexes[0]->branch_no; i ++)
indexes[0]->indexes[i] = indexes[i];
move_stream(stream, new_IP - IP - 1);
}
ptr->indexes[0] = indexes[0];
ptr = ptr->indexes[0];
}
ss_inst *retn = head->indexes[0];
free_inst(head);
return retn;
}
tree
xml_html_parser::parse_pi () {
s += 2;
string name= parse_name ();
skip_space ();
return tuple ("pi", name, parse_until ("?>"));
}
tree
xml_html_parser::parse_closing () {
s += 2;
string name= parse_name ();
(void) parse_until (">");
return tuple ("end", name);
}
static void parse_argv(const char **s, const char *argv[], size_t maxarg) {
for (size_t i = 0; i < maxarg; i++) {
skip_spaces(s);
if (**s == '"' || **s == '\'')
argv[i] = parse_delimited_text(s);
else
argv[i] = parse_until(s, " \t\n");
}
}
static inline
#endif
PsycParseIndexRC
psyc_parse_index (PsycParseIndexState *state, PsycString *idx)
{
ParseRC ret;
if (state->cursor >= state->buffer.length)
return PSYC_PARSE_INDEX_END;
state->startc = state->cursor;
switch (state->part) {
case PSYC_INDEX_PART_START:
case PSYC_INDEX_PART_TYPE:
idx->length = 0;
idx->data = NULL;
switch (state->buffer.data[state->cursor]) {
case '#':
state->part = PSYC_INDEX_PART_LIST;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_INSUFFICIENT);
goto PSYC_INDEX_PART_LIST;
case '.':
state->part = PSYC_INDEX_PART_DICT;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_INSUFFICIENT);
goto PSYC_INDEX_PART_STRUCT;
case '{':
state->part = PSYC_INDEX_PART_DICT;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_INSUFFICIENT);
goto PSYC_INDEX_PART_DICT_LENGTH;
default:
return PSYC_PARSE_INDEX_ERROR_TYPE;
}
case PSYC_INDEX_PART_LIST:
PSYC_INDEX_PART_LIST:
switch (parse_length((ParseState*)state, &idx->length)) {
case PARSE_SUCCESS: // list index is complete
state->part = PSYC_INDEX_PART_TYPE;
return PSYC_PARSE_INDEX_LIST;
case PARSE_INSUFFICIENT: // list index at the end of buffer
return PSYC_PARSE_INDEX_LIST_LAST;
case PARSE_ERROR: // no index
return PSYC_PARSE_INDEX_ERROR_LIST;
default: // should not be reached
return PSYC_PARSE_INDEX_ERROR;
}
case PSYC_INDEX_PART_STRUCT:
PSYC_INDEX_PART_STRUCT:
switch (parse_keyword((ParseState*)state, idx)) {
case PARSE_SUCCESS: // end of keyword
state->part = PSYC_INDEX_PART_TYPE;
return PSYC_PARSE_INDEX_STRUCT;
case PARSE_INSUFFICIENT: // end of buffer
return PSYC_PARSE_INDEX_STRUCT_LAST;
case PARSE_ERROR: // no keyword
return PSYC_PARSE_INDEX_ERROR_STRUCT;
default: // should not be reached
return PSYC_PARSE_INDEX_ERROR;
}
case PSYC_INDEX_PART_DICT_LENGTH:
PSYC_INDEX_PART_DICT_LENGTH:
switch (parse_length((ParseState*)state, &state->elemlen)) {
case PARSE_SUCCESS: // length is complete
state->elemlen_found = 1;
state->elem_parsed = 0;
idx->length = state->elemlen;
idx->data = NULL;
if (state->buffer.data[state->cursor] != ' ')
return PSYC_PARSE_INDEX_ERROR_DICT_LENGTH;
state->part = PSYC_INDEX_PART_DICT;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_INSUFFICIENT);
break;
case PARSE_INSUFFICIENT: // length is incomplete
return PSYC_PARSE_DICT_INSUFFICIENT;
case PARSE_ERROR: // no length
state->part = PSYC_INDEX_PART_DICT;
break;
default: // should not be reached
return PSYC_PARSE_INDEX_ERROR;
}
// fall thru
case PSYC_INDEX_PART_DICT:
if (state->elemlen_found) {
switch (parse_binary((ParseState*)state, state->elemlen, idx,
&state->elem_parsed)) {
case PARSE_SUCCESS:
if (idx->length == state->elem_parsed)
ret = PSYC_PARSE_INDEX_DICT;
else
ret = PSYC_PARSE_INDEX_DICT_END;
break;
case PARSE_INCOMPLETE:
if (idx->length == state->elem_parsed)
ret = PSYC_PARSE_INDEX_DICT_START;
else
ret = PSYC_PARSE_INDEX_DICT_CONT;
break;
default: // should not be reached
return PSYC_PARSE_INDEX_ERROR_DICT;
}
} else {
switch (parse_until((ParseState*)state, '}', idx)) {
case PARSE_SUCCESS:
ret = PSYC_PARSE_INDEX_DICT;
break;
case PARSE_INSUFFICIENT:
return PSYC_PARSE_INDEX_INSUFFICIENT;
default: // should not be reached
return PSYC_PARSE_INDEX_ERROR_DICT;
}
}
state->part = PSYC_INDEX_PART_TYPE;
state->cursor++;
return ret;
}
return PSYC_PARSE_INDEX_ERROR; // should not be reached
}
/**
* Parse dictionary.
*
* dict = [ type ] *dict-item
* dict-item = "{" ( dict-key / length SP OCTET) "}"
* ( type [ SP dict-value ] / [ length ] [ ":" type ] [ SP *OCTET ] )
* dict-key = %x00-7C / %x7E-FF ; any byte except "{"
* dict-value = %x00-7A / %x7C-FF ; any byte except "}"
*/
PsycParseDictRC
psyc_parse_dict (PsycParseDictState *state, PsycString *type, PsycString *elem)
{
ParseRC ret;
if (state->cursor >= state->buffer.length)
return PSYC_PARSE_DICT_END;
state->startc = state->cursor;
switch (state->part) {
case PSYC_DICT_PART_START:
type->length = elem->length = 0;
type->data = elem->data = NULL;
state->part = PSYC_DICT_PART_TYPE;
// fall thru
case PSYC_DICT_PART_TYPE:
switch (parse_keyword((ParseState*)state, type)) {
case PARSE_SUCCESS: // end of keyword
state->part = PSYC_DICT_PART_KEY_START;
return PSYC_PARSE_DICT_TYPE;
case PARSE_INSUFFICIENT: // end of buffer
return PSYC_PARSE_DICT_END;
case PARSE_ERROR: // no keyword
state->part = PSYC_DICT_PART_KEY_START;
break;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
// fall thru
case PSYC_DICT_PART_KEY_START:
if (state->buffer.data[state->cursor] != '{')
return PSYC_PARSE_DICT_ERROR_KEY_START;
type->length = elem->length = 0;
type->data = elem->data = NULL;
state->elem_parsed = 0;
state->elemlen_found = 0;
state->part = PSYC_DICT_PART_KEY_LENGTH;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_DICT_INSUFFICIENT);
// fall thru
case PSYC_DICT_PART_KEY_LENGTH:
switch (parse_length((ParseState*)state, &state->elemlen)) {
case PARSE_SUCCESS: // length is complete
state->elemlen_found = 1;
state->elem_parsed = 0;
elem->length = state->elemlen;
elem->data = NULL;
if (state->buffer.data[state->cursor] != ' ')
return PSYC_PARSE_DICT_ERROR_KEY_LENGTH;
state->part = PSYC_DICT_PART_KEY;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_INSUFFICIENT);
break;
case PARSE_INSUFFICIENT: // length is incomplete
return PSYC_PARSE_DICT_INSUFFICIENT;
case PARSE_ERROR: // no length
state->part = PSYC_DICT_PART_KEY;
break;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
// fall thru
case PSYC_DICT_PART_KEY:
if (state->elemlen_found) {
switch (parse_binary((ParseState*)state, state->elemlen, elem,
&state->elem_parsed)) {
case PARSE_SUCCESS:
if (elem->length == state->elem_parsed)
ret = PSYC_PARSE_DICT_KEY;
else
ret = PSYC_PARSE_DICT_KEY_END;
break;
case PARSE_INCOMPLETE:
if (elem->length == state->elem_parsed)
ret = PSYC_PARSE_DICT_KEY_START;
else
ret = PSYC_PARSE_DICT_KEY_CONT;
break;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
} else {
switch (parse_until((ParseState*)state, '}', elem)) {
case PARSE_SUCCESS:
ret = PSYC_PARSE_DICT_KEY;
break;
case PARSE_INSUFFICIENT:
return PSYC_PARSE_DICT_INSUFFICIENT;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
}
state->part = PSYC_DICT_PART_VALUE_START;
state->startc = state->cursor;
return ret;
case PSYC_DICT_PART_VALUE_START:
switch (state->buffer.data[state->cursor] != '}')
return PSYC_PARSE_DICT_ERROR_VALUE_START;
type->length = elem->length = 0;
type->data = elem->data = NULL;
state->elem_parsed = 0;
state->elemlen_found = 0;
state->part = PSYC_DICT_PART_VALUE_TYPE;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_DICT_VALUE_LAST);
// fall thru
case PSYC_DICT_PART_VALUE_TYPE:
if (state->buffer.data[state->cursor] == '=') {
ADVANCE_CURSOR_OR_RETURN(PSYC_PARSE_INSUFFICIENT);
switch (parse_keyword((ParseState*)state, type)) {
case PARSE_SUCCESS:
switch (state->buffer.data[state->cursor]) {
case ':':
state->part = PSYC_DICT_PART_VALUE_LENGTH;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_DICT_VALUE_LAST);
break;
case ' ':
state->part = PSYC_DICT_PART_VALUE;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_DICT_VALUE_LAST);
goto PSYC_DICT_PART_VALUE;
case '{':
state->part = PSYC_DICT_PART_KEY_START;
return PSYC_PARSE_DICT_VALUE;
break;
default:
return PSYC_PARSE_DICT_ERROR_VALUE_TYPE;
}
break;
case PARSE_INSUFFICIENT: // end of buffer
return PSYC_PARSE_DICT_VALUE_LAST;
case PARSE_ERROR:
return PSYC_PARSE_DICT_ERROR_VALUE_TYPE;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
}
// fall thru
case PSYC_DICT_PART_VALUE_LENGTH:
switch (parse_length((ParseState*)state, &state->elemlen)) {
case PARSE_SUCCESS: // length is complete
state->elemlen_found = 1;
state->elem_parsed = 0;
elem->length = state->elemlen;
elem->data = NULL;
break;
case PARSE_INSUFFICIENT: // length is incomplete
return PSYC_PARSE_DICT_INSUFFICIENT;
case PARSE_ERROR: // no length
break;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
switch (state->buffer.data[state->cursor]) {
case ' ':
state->part = PSYC_DICT_PART_VALUE;
ADVANCE_STARTC_OR_RETURN(PSYC_PARSE_DICT_VALUE_LAST);
break;
case '{':
state->part = PSYC_DICT_PART_KEY_START;
return PSYC_PARSE_DICT_VALUE;
default:
return PSYC_PARSE_DICT_ERROR_VALUE_LENGTH;
}
// fall thru
case PSYC_DICT_PART_VALUE:
PSYC_DICT_PART_VALUE:
if (state->elemlen_found) {
switch (parse_binary((ParseState*)state, state->elemlen, elem,
&state->elem_parsed)) {
case PARSE_SUCCESS:
if (elem->length == state->elem_parsed)
ret = PSYC_PARSE_DICT_VALUE;
else
ret = PSYC_PARSE_DICT_VALUE_END;
break;
case PARSE_INCOMPLETE:
if (elem->length == state->elem_parsed)
ret = PSYC_PARSE_DICT_VALUE_START;
else
ret = PSYC_PARSE_DICT_VALUE_CONT;
break;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
} else {
switch (parse_until((ParseState*)state, '{', elem)) {
case PARSE_SUCCESS:
ret = PSYC_PARSE_DICT_VALUE;
break;
case PARSE_INSUFFICIENT:
return PSYC_PARSE_DICT_VALUE_LAST;
default: // should not be reached
return PSYC_PARSE_DICT_ERROR;
}
}
state->part = PSYC_DICT_PART_KEY_START;
return ret;
}
return PSYC_PARSE_DICT_ERROR; // should not be reached
}
tree
xml_html_parser::parse_notation () {
s += 10;
return tuple ("notation", parse_until (">"));
}
tree
xml_html_parser::parse_attlist () {
s += 9;
return tuple ("attlist", parse_until (">"));
}
tree
xml_html_parser::parse_element () {
s += 9;
return tuple ("element", parse_until (">"));
}
tree
xml_html_parser::parse_cdata () {
s += 9;
return tuple ("cdata", parse_until ("]]>"));
}
tree
xml_html_parser::parse_comment () {
s += 4;
return tuple ("comment", parse_until ("-->"));
}
static char *parse_filename(const char **s) {
skip_spaces(s);
if (**s == '"' || **s == '\'')
return parse_delimited_text(s);
return parse_until(s, "\n");
}
static char *parse_shellcmd(const char **s) {
skip_spaces(s);
return parse_until(s, "\n");
}