static tree
parse_item (string s, int &i, int item_indent) {
int text_indent= item_indent, n= N(s);
while (i<n && is_blank (s[i])) i++;
if (is_new_item (s, i)) {
text_indent++;
while (s[i] == '-') i++;
}
while (i<n && is_spacing (s[i])) {
if (s[i] == ' ')
text_indent++;
i++;
}
tree r (CONCAT);
r << compound ("item");
string line= "";
string tmp= parse_line (s, i, text_indent);
if (tmp != "")
line << tmp;
while (can_parse_line (s, i, text_indent)) {
tmp= parse_line (s, i, text_indent);
if (tmp != "")
line << "\n" << tmp;
}
r << coqdoc_to_tree (line);
return r;
}
static int
parse_indent (string s, int i) {
int indent= 0, n= N(s);
while (i<n && is_spacing (s[i])) i++, indent++;
if (i<n && s[i] != '\n') return indent;
else return -1;
}
static bool
is_list_begining (string s, int i) {
int n= N(s);
if (!(i<n && s[i] == '\n')) return false;
i++;
while (i<n && is_spacing (s[i])) i++;
return is_new_item (s, i);
}
static array<string>
split_command (string s) {
int start= 0, i=0, n= N(s);
array<string> r;
while (i<n) {
if (s[i] == '.' && i+1<n && is_spacing (s[i+1])) {
r << s(start, ++i);
while (i<n && is_spacing (s[i])) i++;
start= i;
}
else if (start_comment (s, i))
parse_comment (s, i);
else
i++;
}
if (start < n)
r << s(start, n);
return r;
}
static bool
is_star_rule (string s, int i) {
int n= N(s), stars=0, blanks= 0;
i++;
while (i<n) {
if (s[i] == '*') stars++;
else if (is_spacing (s[i])) blanks++;
else break;
i++;
}
return (i >= n || (i>0 && test (s, i-1, "*)"))) && stars > blanks;
}
static bool
end_vernac_command (string s, int i) {
int n= N(s);
if (!(i<n && s[i] == '.')) return false;
i++;
while (i<n) {
if (is_spacing (s[i]))
i++;
else if (start_comment (s, i))
parse_comment (s, i);
else
break;
}
return i >= n || s[i] == '\n';
}
static sv_<ss_> split_string_to_clean_ui_pieces(const ss_ &s, size_t piece_len)
{
sv_<ss_> result;
size_t start_from = 0;
for(;;){
if(start_from >= s.size())
break;
size_t start_from_instead = start_from;
for(size_t i=start_from; i<s.size(); i++){
if((s[i] < 'a' || s[i] > 'z') && (s[i] < 'A' || s[i] > 'Z') &&
(s[i] < '0' || s[i] > '9')){
start_from_instead = i + 1;
continue;
}
break;
}
if(start_from_instead >= s.size()){
result.push_back(s.substr(start_from));
break;
}
if(start_from_instead >= s.size() - piece_len){
result.push_back(s.substr(start_from_instead));
break;
}
size_t end_at = start_from_instead + piece_len;
if(s.size() > end_at + 2 && end_at >= 3){
if(is_spacing(s[end_at-2]) && !is_spacing(s[end_at-1]) && !is_spacing(s[end_at]) && (is_spacing(s[end_at+1]) || is_spacing(s[end_at+2]))){
end_at = end_at - 1;
} else if(is_spacing(s[end_at-3]) && !is_spacing(s[end_at-2]) && !is_spacing(s[end_at-1]) && !is_spacing(s[end_at]) && (is_spacing(s[end_at+1]) || is_spacing(s[end_at+2]))){
end_at = end_at - 2;
}
}
result.push_back(s.substr(start_from_instead, end_at - start_from_instead));
start_from = end_at;
}
return result;
}
static tree
coqdoc_to_tree (string s) {
bool newline= true;
int i=0, n= N(s);
tree coqdoc (DOCUMENT), line (CONCAT);
if (starts (s, "(**")) {
line << "(**";
i+= 3;
}
while (i < n) {
if (test (s, i, "[[\n")) {
add_line (line, coqdoc);
tree vernac=
vernac_to_tree (parse_delimited (s, i, "[[\n", "\n]]", false));
coqdoc << compound ("coqdoc-vernac", vernac);
newline= true;
}
else if (s[i] == '[')
line << compound ("coqdoc-coq",
from_verbatim (parse_delimited (s, i, "[", "]", false)));
else if (newline && (test (s, i, "**** ") || test (s, i, "*** ") ||
test (s, i, "** ") || test (s, i, "* "))) {
string header= "section";
if (test (s, i, "** ")) header= "subsection";
if (test (s, i, "*** ")) header= "subsubsection";
if (test (s, i, "**** ")) header= "paragraph";
while (i<n && s[i] == '*') i++;
while (i<n && is_spacing (s[i])) i++;
int start= i;
while (i<n && (s[i] != '\n' && !test (s, i, "*)"))) i++;
line << compound (header, coqdoc_to_tree (s (start, i)));
}
else if (newline && is_defining_pretty_printing (s, i)) {
string str= parse_delimited (s, i, "(*", "*)", false);
parse_pretty_printing_definition (str);
}
else if (newline && is_removing_pretty_printing (s, i)) {
string str= parse_delimited (s, i, "(*", "*)", false);
parse_pretty_printing_removal (str);
}
else if (test (s, i, "%%")) {
line << "%";
newline= false;
i+= 2;
}
else if (test (s, i, "$$")) {
line << "$";
newline= false;
i+= 2;
}
else if (test (s, i, "##")) {
line << "#";
newline= false;
i+= 2;
}
else if (s[i] == '#' || s[i] == '%' || s[i] == '$') {
newline= false;
char delim= s[i];
string ext= unescape_coqdoc (parse_delimited (s, i, delim));
tree tm;
if (delim == '#')
tm= compound ("coqdoc-html", generic_to_tree (ext, "html-snippet"));
else if (delim == '$')
tm= compound ("coqdoc-latex",
generic_to_tree ("$"*ext*"$", "latex-snippet"));
else if (delim == '%')
tm= compound ("coqdoc-latex", generic_to_tree (ext, "latex-snippet"));
if (is_multi_paragraph (tm)) {
add_line (line, coqdoc);
coqdoc << tm;
}
else
line << tm;
}
else if (is_list_begining (s, i)) {
tree list= parse_list (s, i);
add_line (line, coqdoc);
coqdoc << list;
newline= true;
}
else if (test (s, i, "\n<<")) {
add_line (line, coqdoc);
string parsed= parse_delimited (s, i, "\n<<", "\n>>", false);
if (N(parsed) > 0 && parsed[0] == '\n')
parsed= parsed(1, N(parsed));
tree verb= verbatim_to_tree (parsed, false, "SourceCode");
if (is_atomic (verb))
verb= document (verb);
coqdoc << compound ("coqdoc-verbatim", verb);
newline= true;
}
else if (test (s, i, "<<")) {
string parsed= parse_delimited (s, i, "<<", ">>", false);
tree verb= verbatim_to_tree (parsed, true, "SourceCode");
line << compound ("coqdoc-verbatim", verb);
}
else if (s[i] == '_' && (i == 0 || !start_ident(s[i-1]))) {
line << coqdoc_parse_emphasis (s, i);
newline= false;
}
else if (test (s, i, "----")) {
i+= 4;
add_line (line, coqdoc);
coqdoc << compound ("hrule");
while (i<n && s[i] == '-') i++;
newline= true;
}
else if (s[i] == '\n') {
add_line (line, coqdoc);
i++;
if (is_whiteline (s, i)) {
coqdoc << "";
do
skip_whiteline (s, i);
while (is_whiteline (s, i));
i--;
}
newline= true;
}
else if (s[i] == '<') {
line << "<less>";
i++;
newline= false;
}
else if (s[i] == '>') {
line << "<gtr>";
i++;
newline= false;
}
else {
if (!is_spacing (s[i]))
newline= false;
int start= i;
decode_from_utf8 (s, i);
line << from_verbatim (s(start, i));
}
}
if (N(line) > 0)
add_line (line, coqdoc);
if (N(coqdoc) == 0)
return "";
else if (N(coqdoc) == 1)
return coqdoc[0];
else
return coqdoc;
}
static void
skip_whiteline (string s, int &i) {
int n= N(s);
while (i<n && is_spacing (s[i])) i++;
i++;
}
static bool
is_whiteline (string s, int i) {
int n= N(s);
while (i<n && is_spacing (s[i])) i++;
return s[i] == '\n';
}
static bool is_spacing(const char * str, t_size len) {
for(t_size walk = 0; walk < len; ++walk) if (!is_spacing(str[walk])) return false;
return true;
}
