private void Add_br(Xop_ctx ctx, Xop_root_tkn root, int bgn_pos) {
ctx.Subs_add(root, ctx.Tkn_mkr().Xnde(bgn_pos, bgn_pos).Tag_(Xop_xnde_tag_.Tag_br));
}
public int Process_pre(
Xop_ctx ctx,
Xop_tkn_mkr tkn_mkr,
Xop_root_tkn root,
byte[] src,
int src_len,
int bgn_pos,
int cur_pos,
int txt_pos) {
Dd_clear(ctx);
Btrie_slim_mgr tblw_ws_trie = ctx.App().Utl_trie_tblw_ws();
Object o = tblw_ws_trie.Match_bgn(src, txt_pos, src_len);
if (o != null) { // tblw_ws found
Xop_tblw_ws_itm ws_itm = (Xop_tblw_ws_itm) o;
byte tblw_type = ws_itm.Tblw_type();
switch (tblw_type) {
case Xop_tblw_ws_itm.Type_nl: // \n\s
if (cur_mode == Mode_pre) { // already in pre; just process "\n\s"
ctx.Subs_add(root, tkn_mkr.NewLine(bgn_pos, bgn_pos, Xop_nl_tkn.Tid_char, 1));
prv_nl_pos =
bgn_pos; // NOTE: must update prv_nl_pos; PAGE:en.w:Preferred_number DATE:2014-06-24
return txt_pos;
}
break;
case Xop_tblw_ws_itm.Type_xnde:
int nxt_pos = tblw_ws_trie.Match_pos();
if (nxt_pos < src_len) { // bounds check
switch (src[
nxt_pos]) { // check that next char is "end" of xnde name; guard against false
// matches like "<trk" PAGE:de.v:Via_Jutlandica/Gpx DATE:2014-11-29
case Byte_ascii.Space:
case Byte_ascii.Nl:
case Byte_ascii.Tab: // whitespace
case Byte_ascii.Slash:
case Byte_ascii.Gt: // end node
case Byte_ascii.Quote:
case Byte_ascii.Apos: // quotes
if (bgn_pos != Xop_parser_.Doc_bgn_bos)
ctx.Para().Process_nl(ctx, root, src, bgn_pos, cur_pos);
return ctx.Xnde().Make_tkn(ctx, tkn_mkr, root, src, src_len, txt_pos, txt_pos + 1);
}
}
break;
default:
{
int tblw_rv =
ctx.Tblw()
.Make_tkn_bgn(
ctx,
tkn_mkr,
root,
src,
src_len,
bgn_pos,
txt_pos + ws_itm.Hook_len(),
false,
tblw_type,
Xop_tblw_wkr.Called_from_pre,
-1,
-1);
if (tblw_rv
!= -1) // \n\s| is valid tblw tkn and processed; otherwise process pre-code below;
// EX:w:Wikipedia:WikiProject_History/CategoryExample; DATE:2014-04-14
return tblw_rv;
break;
}
}
}
// NOTE: pre lxr emulates MW for "\n\s" by (1) calling Process nl for "\n"; (2) anticipating
// next line by setting prv_ws_bgn
// EX: "\na\n b\n"; note that "\n " is cur
if (bgn_pos != Xop_parser_.Doc_bgn_bos) // if bos, then don't close 1st para
Process_nl(
ctx,
root,
src,
bgn_pos,
bgn_pos + 1); // note that tkn is \n\s; so, bgn_pos -> bgn_pos + 1 is \n ...
if (cur_mode == Mode_pre) // in pre_mode
ctx.Subs_add(
root,
tkn_mkr.Space(
root, cur_pos,
txt_pos)); // cur_pos to start after \s; do not capture "\s" in "\n\s"; (not sure why
// not before \s)
prv_ws_bgn = txt_pos - cur_pos + 1;
return txt_pos;
}
private void Prv_para_new(Xop_ctx ctx, Xop_root_tkn root, int prv_nl_pos, int para_pos) {
this.prv_nl_pos = prv_nl_pos;
prv_para = ctx.Tkn_mkr().Para(para_pos);
ctx.Subs_add(root, prv_para);
}
public void Process_nl(
Xop_ctx ctx,
Xop_root_tkn root,
byte[] src,
int bgn_pos,
int cur_pos) { // REF.MW:Parser.php|doBlockLevels
Dd_clear(ctx);
if (block_is_bgn_xnde || block_is_end_xnde) {
para_stack = Para_stack_none; // MW: $paragraphStack = false;
Prv_para_end(); // MW: $output .= $this->closeParagraph()
if (block_is_bgn_blockquote
&& !block_is_end_blockquote) // MW: if ( $preOpenMatch and !$preCloseMatch )
in_blockquote = true; // MW: $this->mInPre = true;
else in_blockquote = false; // XO: turn off blockquote else following para / nl won't work;
// w:Snappy_(software); DATE:2014-04-25
in_block = !block_is_end_xnde; // MW: $inBlockElem = !$closematch;
} else if (!in_block && !in_blockquote) { // MW: elseif ( !$inBlockElem && !$this->mInPre ) {
boolean line_is_ws = Line_is_ws(src, bgn_pos);
if (prv_ws_bgn > 0
&& (cur_mode == Mode_pre
|| !line_is_ws)) { // MW: if ( ' ' == substr( $t, 0, 1 ) and ( $this->mLastSection ===
// 'pre' || trim( $t ) != '' ) ) {
if (cur_mode != Mode_pre) { // MW: if ( $this->mLastSection !== 'pre' ) {
para_stack = Para_stack_none; // MW: $paragraphStack = false;
prv_para.Space_bgn_(
prv_ws_bgn
- 1); // -1 to ignore 1st "\s" in "\n\s"; note that prv_ws_bgn only includes
// spaces, so BOS doesn't matter; DATE:2014-04-14
Prv_para_end();
Prv_para_bgn(Xop_para_tkn.Tid_pre); // MW: $output .= $this->closeParagraph() . '<pre>';
cur_mode = Mode_pre; // MW: $this->mLastSection = 'pre';
} else { // already in pre
if (line_is_ws) { // line is entirely ws
int next_char_pos = prv_nl_pos + 2; // "\n\s".length
if (next_char_pos < src.length // bounds check
&& src[next_char_pos] == Byte_ascii.Nl // is "\n\s\n"; i.e.: "\n" only
) {
ctx.Subs_add(
root,
ctx.Tkn_mkr()
.Bry_raw(
bgn_pos,
bgn_pos,
Byte_ascii
.Nl_bry)); // add a "\n" tkn; note that adding a NewLine tkn doesn't
// work, b/c Xoh_html_wtr has code to remove consecutive
// \n; PAGE:en.w:Preferred_numbers DATE:2014-06-24
prv_nl_pos = bgn_pos;
}
}
}
prv_ws_bgn = 0; // MW: $t = substr( $t, 1 );
} else {
if (bgn_pos - prv_nl_pos == 1
|| line_is_ws) { // line is blank ("b" for blank) MW: if ( trim( $t ) === '' ) {
if (para_stack
!= Para_stack_none) { // "b1"; stack has "<p>" or "</p><p>"; output "<br/>"; MW: if (
// $paragraphStack ) {
Para_stack_end(cur_pos);
Add_br(ctx, root, bgn_pos); // MW: $output .= $paragraphStack . '<br />';
para_stack = Para_stack_none; // MW: $paragraphStack = false;
cur_mode = Mode_para; // MW: $this->mLastSection = 'p';
} else { // stack is empty
if (cur_mode
!= Mode_para) { // "b2"; cur is '' or <pre> MW: if ( $this->mLastSection !==
// 'p' ) {
Prv_para_end(); // MW: $output .= $this->closeParagraph();
cur_mode = Mode_none; // MW: $this->mLastSection = '';
para_stack = Para_stack_bgn; // put <p> on stack MW: $paragraphStack = '<p>';
} else // "b3"; cur is p
para_stack =
Para_stack_mid; // put </p><p> on stack MW: $paragraphStack = '</p><p>';
}
} else { // line has text ("t" for text); NOTE: tkn already added before \n, so must change
// prv_para; EX: "a\n" -> this code is called for "\n" but "a" already processed
if (para_stack != Para_stack_none) { // "t1" MW: if ( $paragraphStack ) {
Para_stack_end(cur_pos); // MW: $output .= $paragraphStack;
para_stack = Para_stack_none; // MW: $paragraphStack = false;
cur_mode = Mode_para; // MW: $this->mLastSection = 'p';
} else if (cur_mode
!= Mode_para) { // "t2"; cur is '' or <pre> MW: elseif ( $this->mLastSection
// !== 'p' ) {
Prv_para_end();
Prv_para_bgn(
Xop_para_tkn
.Tid_para); // MW: $output .= $this->closeParagraph() . '<p>';
cur_mode = Mode_para; // MW: $this->mLastSection = 'p';
} else {
} // "t3"
}
}
}
if (in_blockquote
&& prv_ws_bgn
> 0) // handle blockquote separate; EX: <blockquote>\n\sa\n</blockquote>; note that "\s"
// needs to be added literally; MW doesn't have this logic specifically, since it
// assumes all characters go into $output, whereas XO, sets aside the "\s" in
// "\n\s" separately
prv_para.Space_bgn_(prv_ws_bgn);
prv_ws_bgn =
0; // nl encountered and processed; always prv_ws_bgn set to 0, else ws from one line will
// carry over to next
// in_blockquote = false;
block_is_bgn_xnde = block_is_end_xnde = false;
// if ( $preCloseMatch && $this->mInPre )
// $this->mInPre = false;
// prv_ws_bgn = false;
Prv_para_new(ctx, root, bgn_pos, cur_pos); // add a prv_para placeholder
if (para_stack == Para_stack_none) // "x1" MW: if ( $paragraphStack === false ) {
if (prv_para != null)
prv_para
.Nl_bgn_y_(); // add nl; note that "$t" has already been processed; MW: $output .= $t .
// "\n";
}