// Should only be called when we actually have the start of an array
// Otherwise it is an internal error
bool JSON::parse_json_array() {
NOT_PRODUCT(const char* prev_pos);
int c;
mark_pos();
// Check that we are not called in error
if (expect_any("[", "array start character", INTERNAL_ERROR) <= 0) {
return false;
}
if (!callback(JSON_ARRAY_BEGIN, NULL, level++)) {
return false;
}
for (;;) {
mark_pos();
c = skip_to_token();
if (c == 0) {
error(SYNTAX_ERROR, "EOS when expecting a json value or array end");
return false;
} else if (c < 0) {
return false;
} else if (c == ']') {
// We got here from either empty array "[]" or ending comma "[1,]"
next();
break;
}
mark_pos();
NOT_PRODUCT(prev_pos = pos);
if (parse_json_value() == false) {
return false;
}
assert(pos > prev_pos, "parsing stalled");
c = skip_to_token();
mark_pos();
if (expect_any(",]", "value separator or array end") <= 0) {
return false;
}
if (c == ']') {
break;
}
}
assert(c == ']', "array parsing ended without array end token (']')");
return callback(JSON_ARRAY_END, NULL, --level);
}
//-------------------------------------------------------------------
// Process one entry "@default VAR VALUE"
//-------------------------------------------------------------------
static const char* get_default(sc_param *p, const char *ptr, int isScript)
{
ptr = skip_to_token(ptr);
if (p)
{
int type = MENUITEM_INT|MENUITEM_SCRIPT_PARAM;
int range = 0;
if (strncmp(ptr, "true", 4) == 0)
{
p->val = 1;
type = MENUITEM_BOOL|MENUITEM_SCRIPT_PARAM;
range = MENU_MINMAX(1,0); // Force boolean data type in Lua (ToDo: this is clunky, needs fixing)
}
else if (strncmp(ptr, "false", 5) == 0)
{
p->val = 0;
type = MENUITEM_BOOL|MENUITEM_SCRIPT_PARAM;
range = MENU_MINMAX(1,0); // Force boolean data type in Lua (ToDo: this is clunky, needs fixing)
}
else
{
p->val = strtol(ptr, NULL, 0);
}
p->old_val = p->val;
if (isScript) // Loading from script file (rather than saved param set file)
{
p->def_val = p->val;
p->range = range;
p->range_type = type;
}
}
return skip_token(ptr);
}
int htmlParse::skip_ignore_part(stToken *tok, stTagProc *curr_tag)
{
stTagProc *this_tag;
int prev_tok_val=0;
int tmp=0;
//fprintf(log,"--->IGNORE :%d :%d\n", tok->line_num, tok->offset);
tmp = tok->line_num;
while(htmlLex::get_token_mem(tok) != TOK_EOF) {
if (tok->tok_val == CMNT_BEGIN) htmlLex::skip_to_cmnt_end(tok);
else if (tok->tok_val == BTAG_BEGIN && curr_tag->pair==Opt_PAIR) {
tok->offset --;
break;
}
else if (tok->tok_val == ETAG_BEGIN) {
if (curr_tag->pair==Must_PAIR) {
skip_to_two_token(tok, STRING, TAG_END);
if (tok->tok_val == STRING) {
this_tag = htmlTagEntity::tag_idx(tok->tok_str);
//fprintf(log,"%s<--:%d :%d\n", tok->tok_str, tok->line_num, tok->offset );
if (curr_tag==this_tag) {
skip_to_token(tok, TAG_END);
break;
}
}
}
else { // Pot_PAIR
skip_to_token(tok, TAG_END);
break;
}
}
else if (tok->tok_val == STRING) {
this_tag = htmlTagEntity::tag_idx(tok->tok_str);
}
prev_tok_val = tok->tok_val;
} //
//fprintf(log,"<--- IGNORE :%d :%d [%d]\n", tok->line_num, tok->offset, tok->line_num - tmp);
return 0;
}
bool JSON::parse_json_value() {
int c;
c = skip_to_token();
if (c == -1) {
return false;
}
// Must start with object or array
if (level == 0) {
switch (c) {
case '{':
if (parse_json_object() == false) {
return false;
}
c = skip_to_token();
if (c > 0) {
mark_pos();
error(SYNTAX_ERROR, "Only one top level object/array is allowed.");
return false;
} else if (c < 0) {
return false;
}
return true;
case '[':
if (parse_json_array() == false) {
return false;
}
c = skip_to_token();
if (c > 0) {
mark_pos();
error(SYNTAX_ERROR, "Only one top level object/array is allowed.");
return false;
} else if (c < 0) {
return false;
}
return true;
case 0:
error(SYNTAX_ERROR, "EOS was encountered before any json declarations");
return false;
default:
error(SYNTAX_ERROR, "Json must start with an object or an array.");
return false;
}
} else { // level > 0
switch (c) {
case '{':
return parse_json_object();
case '[':
return parse_json_array();
case '"':
return parse_json_string();
case '-': case '0':
case '1': case '2': case '3':
case '4': case '5': case '6':
case '7': case '8': case '9':
return parse_json_number();
case 't':
return parse_json_symbol("true", JSON_TRUE);
case 'f':
return parse_json_symbol("false", JSON_FALSE);
case 'n':
return parse_json_symbol("null", JSON_NULL);
case 0:
error(SYNTAX_ERROR, "EOS was encountered when expecting a json value.");
return false;
default:
error(SYNTAX_ERROR, "Could not parse as a json value (did you forget to quote your strings?).");
return false;
}
}
}
// Should only be called when we actually have the start of an object
// Otherwise it is an internal error
bool JSON::parse_json_object() {
NOT_PRODUCT(const char* prev_pos);
int c;
mark_pos();
// Check that we are not called in error
if (expect_any("{", "object start", INTERNAL_ERROR) <= 0) {
return false;
}
if (!callback(JSON_OBJECT_BEGIN, NULL, level++)) {
return false;
}
for (;;) {
mark_pos();
c = skip_to_token();
if (c == 0) {
error(SYNTAX_ERROR, "EOS when expecting an object key or object end");
return false;
} else if (c < 0) {
return false;
} else if (c == '}') {
// We got here from either empty object "{}" or ending comma "{a:1,}"
next();
break;
}
NOT_PRODUCT(prev_pos = pos);
if (parse_json_key() == false) {
return false;
}
assert(pos > prev_pos, "parsing stalled");
skip_to_token();
mark_pos();
if (expect_any(":", "object key-value separator") <= 0) {
return false;
}
skip_to_token();
mark_pos();
NOT_PRODUCT(prev_pos = pos);
if (parse_json_value() == false) {
return false;
}
assert(pos > prev_pos, "parsing stalled");
c = skip_to_token();
mark_pos();
if (expect_any(",}", "value separator or object end") <= 0) {
return false;
}
if (c == '}') {
break;
}
}
assert(c == '}', "array parsing ended without object end token ('}')");
return callback(JSON_OBJECT_END, NULL, --level);
}
/**
HTML 문서를 parsing 하면서 text로 변환.
*/
int htmlParse::parse(stToken *tok, char *text, int maxlen)
{
int tok_val=0, prev_tok_val=0, prev_ch=0;
int is_in_tag=0;
uint4 tmp=0;
int not_moved=0;
char *start=text;
stTagProc *curr_tag=0;
//stTagProc **prev_tag=0;
stEntityProc *ent_proc;
int tag_depth=0;
int is_in_BODY_tag=0;
int is_in_A_tag=0;
int is_in_PRE_tag=0; // 2003.11.8
int hlink_chars = 0;
//int newline_delayed=0; // Optional Pair && Newline tag가 시작한 경우 TRUE
//int dont_newline=0; // TRUE means "dont write newline"
int pair_mark=0;
int words_in_line=0;
int b_tag_first_in_line=0; // 라인 첫머리에 B tag가 나온 경우
int glossary_marked=0;
int is_glossary=0;
tok->offset = 0;
tok->line_num=1;
*text = 0;
maxlen -= 20;
while(tok->offset < tok->src_len && (int)((uint4)text-(uint4)start) < maxlen) {
tmp = tok->offset;
htmlLex::get_token_mem(tok);
#ifdef BUG
if (tok->tok_len>=20) {
printf("long tok: line=%d, tok_val=%d, CH=%c\n",
tok->line_num, tok->tok_val, tok->src_mem[tok->offset-1]);
fflush(stdout);
printf("tok=%s\n", tok->tok_str);
fflush(stdout);
}
#endif
#ifdef DEB
/*
if (tok->tok_val==0 || tok->line_num==1) {
printf("line=%d, tok_val=%d, CH=%c\n",
tok->line_num, tok->tok_val, tok->src_mem[tok->offset-1]);
printf("offset=%d, text-start=%d\n", tok->offset, text-start);
}
*/
#endif
if (text > start) {
prev_ch = *(text-1);
if (prev_ch=='\n') {
words_in_line = 0;
b_tag_first_in_line = 0;
glossary_marked = 0;
}
}
if (tok->tok_val == TOK_EOF) break;
if (tok->tok_val == URL_CMNT) {
//printf("%s", tok->tok_str);
continue;
}
if (tok->offset==tmp) {
#ifdef BUG
//printf("parse(): offset not moved: line=%d, CH=%d(%c)\n",
// tok->line_num, tok->src_mem[tok->offset-1], tok->src_mem[tok->offset-1]);
#endif
tok->offset++;
if (++not_moved >= 2) break;
else continue;
}
tok_val = tok->tok_val;
if (tok_val == BTAG_BEGIN) {
skip_to_two_token(tok, STRING, TAG_END);
if (tok->tok_val != STRING) { // must be a HTML tag
skip_to_token(tok, TAG_END);
continue;
}
//prn_tabs();
//fprintf(log,"%s-->:%d :%d\n", tok->tok_str, tok->line_num, tok->offset);
curr_tag = htmlTagEntity::tag_idx(tok->tok_str);
if (curr_tag == NULL) {
skip_to_token(tok, TAG_END);
continue;
}
tag_depth++;
#ifdef MAIN_TEXT_ONLY
if (curr_tag->pair != Not_PAIR && is_in_BODY_tag && curr_tag!=htmlTagEntity::A_tag)
push_tag(curr_tag, tag_depth, (int)((uint4)text-(uint4)start));
if (curr_tag==htmlTagEntity::BODY_tag) {
is_in_BODY_tag = 1;
}
#endif
if (curr_tag==htmlTagEntity::A_tag) {
is_in_A_tag = 1;
}
else if (curr_tag==htmlTagEntity::PRE_tag) {
is_in_PRE_tag = 1;
}
else if(curr_tag==htmlTagEntity::B_tag||curr_tag==htmlTagEntity::DT_tag) {
if (words_in_line==0)
b_tag_first_in_line = 1;
}
else if (curr_tag==htmlTagEntity::TITLE_tag) {
get_hyperlink_title(tok->src_mem + tok->offset , htmlTitle, HTMLTITLE_LEN);
#ifdef DEB
PRN("tok->scr_mem=%X, offset=%d, title=%X\n",
tok->src_mem, tok->offset, htmlTitle);
PRN("title: %s (%d)\n", htmlTitle, strlen(htmlTitle) );
#endif
is_glossary = is_glossay_mode(htmlTitle);
}
else if (curr_tag==htmlTagEntity::BR_tag) {
*text++ = '\n';
}
if (curr_tag->proc == 0) { // 0
if (curr_tag->pair != Not_PAIR)
skip_ignore_part(tok, curr_tag);
else
skip_to_token(tok, TAG_END);
if (curr_tag->newline) {
if (prev_ch != '\n') *text++ = '\n';
}
continue;
}
if (curr_tag==htmlTagEntity::TITLE_tag) {
if (prev_ch != '\n') *text++ = '\n';
*text++ = '[';
*text++ = '[';
}
else if (curr_tag->proc == 1) {
}
else if (!pair_mark && curr_tag->proc == 2) {
pair_mark = 2;
*text++ = '<';
}
else if (!pair_mark && curr_tag->proc == 3) {
pair_mark = 3;
*text++ = '\'';
}
else if (!pair_mark && curr_tag->proc == 4) {
pair_mark = 4;
*text++ = '[';
}
else if (curr_tag->proc == 5) {
if (prev_ch != '\n') *text++ = '\n';
*text++ = '*';
*text++ = ' ';
words_in_line = 0;
}
else if (curr_tag->proc == 6) {
if (prev_ch != '\n') *text++ = '\n';
*text++ = '\n';
}
skip_to_token(tok, TAG_END);
if (!pair_mark && prev_ch != ' ') *text++ = ' ';
//if (!ISSPACE(*text)) *text++ = ' ';
}// BTAG_BEGIN '<'
else if (tok_val == ETAG_BEGIN) {
skip_to_two_token(tok, STRING, TAG_END);
if (tok->tok_val != STRING) { // must be a HTML tag
skip_to_token(tok, TAG_END);
continue;
}
curr_tag = htmlTagEntity::tag_idx(tok->tok_str);
if (curr_tag == NULL) {
skip_to_token(tok, TAG_END);
continue;
}
#ifdef MAIN_TEXT_ONLY
tag_depth--;
if (curr_tag==htmlTagEntity::BODY_tag) {
is_in_BODY_tag = 0;
}
if (curr_tag->pair != Not_PAIR && is_in_BODY_tag && curr_tag!=htmlTagEntity::A_tag)
end_tag(curr_tag, (int)((uint4)text-(uint4)start));
#endif
if (curr_tag==htmlTagEntity::A_tag) {
is_in_A_tag = 0;
#ifdef MARK_A_LINK
*text++ = '}';
#endif
}
else if (curr_tag==htmlTagEntity::PRE_tag) {
is_in_PRE_tag = 0;
}
else if(curr_tag==htmlTagEntity::DT_tag ||
(curr_tag==htmlTagEntity::B_tag && (b_tag_first_in_line && is_glossary)))
{
if (glossary_marked==0) {
*text++ = ' ';
*text++ = ':';
//*text++ = ' ';
b_tag_first_in_line = 0;
glossary_marked = 1;
}
}
else if (curr_tag==htmlTagEntity::TITLE_tag) {
*text++ = ']';
*text++ = ']';
*text++ = '\n';
}
else if (curr_tag->proc == 1) {
}
else if (pair_mark==2 && curr_tag->proc == 2) {
pair_mark = 0;
*text++ = '>';
}
else if (pair_mark==3 && curr_tag->proc == 3) {
pair_mark = 0;
*text++ = '\'';
}
else if (pair_mark==4 && curr_tag->proc == 4) {
pair_mark = 0;
*text++ = ']';
}
else if (curr_tag->proc == 5) {
}
else if (curr_tag->proc == 6) {
if (prev_ch != '\n') *text++ = '\n';
*text++ = '\n';
}
if (curr_tag->newline) {
*text++ = '\n';
}
skip_to_token(tok, TAG_END);
}// ETAG_BEGIN '</'
else if (tok_val == STAG_END) {
skip_to_token(tok, TAG_END);
is_in_tag = 0;
tag_depth--;
}
else if (tok_val == TAG_EXC) {
tmp = skip_to_token(tok, TAG_END);
//fprintf(log,"<- > skipped %d\n", tmp);
}
else if (tok_val == CMNT_BEGIN) {
//fprintf(log,"Cmnt -->: %d :%d\n", tok->line_num, tok->offset);
htmlLex::skip_to_cmnt_end(tok);
//fprintf(log,"Cmnt <--: %d :%d\n", tok->line_num, tok->offset);
}
else if (tok_val == ENTITY_STR) {
if (prev_ch !=' ') *text++ = ' '; //2002.12.2
ent_proc = htmlTagEntity::entity_idx(tok->tok_str);
if (ent_proc && ent_proc->conv[0]) {
#ifdef BUG
//prn_ent_proc(ent_proc);
#endif
strcpy(text, ent_proc->conv);
text += strlen(ent_proc->conv);
}
else {
//fprintf(log,"ignored entity = %s\n", tok->tok_str);
}
}
else if (tok_val == ENTITY_NUM) {
if (prev_ch !=' ') *text++ = ' '; //2002.12.2
ent_proc = htmlTagEntity::entity_id_idx(tok->tok_realval);
if (ent_proc && ent_proc->conv[0]) {
#ifdef BUG
//prn_ent_proc(ent_proc);
#endif
strcpy(text, ent_proc->conv);
text += strlen(ent_proc->conv);
}
else {
if (tok->tok_realval < 0x80) {
// 2005.7.19
*text = (char)tok->tok_realval ;
text++;
}
}
}
else {
char *t = text;
if (is_in_PRE_tag || tok->tok_len>1 || tok_val == STRING || tok_val == NUMBER) {
//if (prev_tok_val == TAG_END && *text != ' ') *text++ = ' ';
// 2003. 3.20
//if (prev_tok_val == TAG_END && isalnum(prev_ch) )
// *text++ = ' ';
words_in_line++;
strcpy(text, tok->tok_str);
text += tok->tok_len;
}
// 2002.10.16 한글은 space로 전환
else if (tok->tok_val == FR_STR) {
if (do_prn_hangul) {
if (prev_ch != ' ') *text++ = ' ';
strcpy(text, tok->tok_str);
text += tok->tok_len;
words_in_line++;
}
else {
if (prev_ch != ' ') *text++ = ' ';
}
}
else if (tok->tok_len==1) {
if (curr_tag==htmlTagEntity::PRE_tag) {
#ifdef ODD_CHAR
if ( (tok_val & 0xF0) != 0x90)
*text++ = tok_val;
else if (tok_val==0x92) {
if (prev_tok_val != 0x92)
*text++ = '\'';
}
#else
*text++ = tok_val;
#endif
}
//else if (tok_val == '\n' || tok_val == '\r') {
else if (ISSPACE(tok_val)) {
if (prev_ch != ' ') *text++ = ' ';
}
else {
if (prev_ch==':' && tok_val==':') { }
else *text++ = tok_val;
}
}
if (is_in_A_tag)
hlink_chars += (int)(text - t);
}
prev_tok_val = tok->tok_val;
}// while(1)
*text = 0;
if ((int)(text-start) >= maxlen-5) {
PRN("parse(): too far !! maxlen=%d, %d\n", maxlen, text-start);
}
if ((int)(text-start) < maxlen-5)
memset(text, 0, 4);
#ifdef DEB
PRN("text=%X, start=%X, text=%d, start=%d\n",
text, start, (int)text % 10000, (int)start % 10000);
PRN("start[0]=%d text[0]=%d\n", start[0], text[0]);
#endif
return ((int)text-(int)start);
}
