void TextParser::setState(unsigned code, bool bSet)
{
if (bSet) {
if ((m_state & code) == code)
return;
m_state |= code;
} else {
if ((m_state & code) == 0)
return;
m_state &= ~code;
}
QString tag;
switch (code) {
case 1:
tag = "b";
break;
case 2:
tag = "i";
break;
case 4:
tag = "u";
break;
default:
return;
}
if (bSet) {
push_tag(tag);
} else {
pop_tag(tag);
}
}
hcerr_t start_element (xml_writer *writer,
char *element_name,
char **attribute_names,
char **attribute_values,
int n_attributes){
int i;
hc_simple_xml_writer_t *simple_writer = (hc_simple_xml_writer_t*) writer;
hc_tag_stack_t *top = 0;
require_ok(pretty_print(simple_writer, TRUE));
require_ok(push_tag (&top, simple_writer->tag_stack, element_name));
simple_writer->tag_stack = top;
require_ok(hc_write(simple_writer, "<"));
require_ok(hc_write(simple_writer, element_name));
/*fprintf(stdout, "%d attributes\n", n_attributes); */
for (i = 0; i < n_attributes; i++){
require_ok(hc_write(simple_writer, " "));
require_ok(hc_write(simple_writer, attribute_names[i]));
require_ok(hc_write(simple_writer, "=\""));
require_ok(hc_write(simple_writer, attribute_values[i]));
require_ok(hc_write(simple_writer, "\""));
}
require_ok(hc_write(simple_writer, ">"));
return HCERR_OK;
}
static int
mime_multipart_related_output_fn(const char* buf, int32_t size, void *stream_closure)
{
MimeMultipartRelated *relobj = (MimeMultipartRelated *) stream_closure;
char* ptr;
int32_t delta;
int status;
while (size > 0) {
if (relobj->curtag_length > 0) {
ptr = PL_strnchr(buf, '>', size);
if (!ptr) {
return push_tag(relobj, buf, size);
}
delta = ptr - buf + 1;
status = push_tag(relobj, buf, delta);
if (status < 0) return status;
status = flush_tag(relobj);
if (status < 0) return status;
buf += delta;
size -= delta;
}
ptr = PL_strnchr(buf, '<', size);
if (ptr && ptr - buf >= size) ptr = 0;
if (!ptr) {
return real_write(relobj, buf, size);
}
delta = ptr - buf;
status = real_write(relobj, buf, delta);
if (status < 0) return status;
buf += delta;
size -= delta;
PR_ASSERT(relobj->curtag_length == 0);
status = push_tag(relobj, buf, 1);
if (status < 0) return status;
PR_ASSERT(relobj->curtag_length == 1);
buf++;
size--;
}
return 0;
}
void TextParser::put_style()
{
if (!m_bChanged)
return;
m_bChanged = false;
QString style;
if (!color.isEmpty())
style = color;
if (!face.isEmpty()) {
if (!style.isEmpty())
style += ";";
style += face;
}
if (!size.isEmpty()) {
if (!style.isEmpty())
style += ";";
style += size;
}
QString tag("span style=\"");
tag += style;
tag += "\"";
pop_tag(tag);
push_tag(tag);
}
static int
flush_tag(MimeMultipartRelated* relobj)
{
int length = relobj->curtag_length;
char* buf;
int status;
if (relobj->curtag == NULL || length == 0) return 0;
status = push_tag(relobj, "", 1); /* Push on a trailing NULL. */
if (status < 0) return status;
buf = relobj->curtag;
PR_ASSERT(*buf == '<' && buf[length - 1] == '>');
while (*buf) {
char c;
char* absolute;
char* part_url;
char* ptr = buf;
char *ptr2;
char quoteDelimiter = '\0';
while (*ptr && *ptr != '=') ptr++;
if (*ptr == '=') {
/* Ignore = and leading space. */
/* Safe, because there's a '>' at the end! */
do {ptr++;} while (IS_SPACE(*ptr));
if (*ptr == '"' || *ptr == '\'') {
quoteDelimiter = *ptr;
/* Take up the quote and leading space here as well. */
/* Safe because there's a '>' at the end */
do {ptr++;} while (IS_SPACE(*ptr));
}
}
status = real_write(relobj, buf, ptr - buf);
if (status < 0) return status;
buf = ptr;
if (!*buf) break;
if (quoteDelimiter)
{
ptr = PL_strnchr(buf, quoteDelimiter, length - (buf - relobj->curtag));
} else {
for (ptr = buf; *ptr ; ptr++) {
if (*ptr == '>' || IS_SPACE(*ptr)) break;
}
PR_ASSERT(*ptr);
}
if (!ptr || !*ptr) break;
while(buf < ptr)
{
/* ### mwelch For each word in the value string, see if
the word is a cid: URL. If so, attempt to
substitute the appropriate mailbox part URL in
its place. */
ptr2=buf; /* walk from the left end rightward */
while((ptr2<ptr) && (!IS_SPACE(*ptr2)))
ptr2++;
/* Compare the beginning of the word with "cid:". Yuck. */
if (((ptr2 - buf) > 4) &&
((buf[0]=='c' || buf[0]=='C') &&
(buf[1]=='i' || buf[1]=='I') &&
(buf[2]=='d' || buf[2]=='D') &&
buf[3]==':'))
{
// Make sure it's lowercase, otherwise it won't be found in the hash table
buf[0] = 'c'; buf[1] = 'i'; buf[2] = 'd';
/* Null terminate the word so we can... */
c = *ptr2;
*ptr2 = '\0';
/* Construct a URL out of the word. */
absolute = MakeAbsoluteURL(relobj->base_url, buf);
/* See if we have a mailbox part URL
corresponding to this cid. */
part_url = nullptr;
MimeHashValue * value = nullptr;
if (absolute)
{
value = (MimeHashValue *)PL_HashTableLookup(relobj->hash, buf);
part_url = value ? value->m_url : nullptr;
PR_FREEIF(absolute);
}
/*If we found a mailbox part URL, write that out instead.*/
if (part_url && accept_related_part(relobj, value->m_obj))
{
status = real_write(relobj, part_url, strlen(part_url));
if (status < 0) return status;
buf = ptr2; /* skip over the cid: URL we substituted */
/* don't show that object as attachment */
if (value->m_obj)
value->m_obj->dontShowAsAttachment = true;
}
/* Restore the character that we nulled. */
*ptr2 = c;
}
/* rhp - if we get here, we should still check against the hash table! */
else
{
char holder = *ptr2;
char *realout;
*ptr2 = '\0';
/* Construct a URL out of the word. */
absolute = MakeAbsoluteURL(relobj->base_url, buf);
/* See if we have a mailbox part URL
corresponding to this cid. */
MimeHashValue * value;
if (absolute)
value = (MimeHashValue *)PL_HashTableLookup(relobj->hash, absolute);
else
value = (MimeHashValue *)PL_HashTableLookup(relobj->hash, buf);
realout = value ? value->m_url : nullptr;
*ptr2 = holder;
PR_FREEIF(absolute);
if (realout && accept_related_part(relobj, value->m_obj))
{
status = real_write(relobj, realout, strlen(realout));
if (status < 0) return status;
buf = ptr2; /* skip over the cid: URL we substituted */
/* don't show that object as attachment */
if (value->m_obj)
value->m_obj->dontShowAsAttachment = true;
}
}
/* rhp - if we get here, we should still check against the hash table! */
/* Advance to the beginning of the next word, or to
the end of the value string. */
while((ptr2<ptr) && (IS_SPACE(*ptr2)))
ptr2++;
/* Write whatever original text remains after
cid: URL substitution. */
status = real_write(relobj, buf, ptr2-buf);
if (status < 0) return status;
buf = ptr2;
}
}
if (buf && *buf) {
status = real_write(relobj, buf, strlen(buf));
if (status < 0) return status;
}
relobj->curtag_length = 0;
return 0;
}
/**
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);
}
void xml_tag_start(const char *tag)
{
push_tag(tag);
xprintf("<%s", tag);
}
