static string
parse_identifier (string s, int &i) {
int n= N(s), start= i;
if (i<n && start_ident (s[i])) {
decode_from_utf8 (s, i);
while (i<n && continue_ident (s[i]))
decode_from_utf8 (s, i);
}
return as_string (from_verbatim (s (start, i)));
}
int
get_utf8_code (string c) {
string uc= cork_to_utf8 (c);
int pos= 0;
int code= decode_from_utf8 (uc, pos);
if (pos == N(uc)) return code;
else return -1;
}
int
str_length (string s, bool utf8) {
if (utf8) {
int i=0, r=0;
while (i < N(s)) {
decode_from_utf8 (s, i);
r++;
}
return r;
}
else return N(s);
}
void
goto_next_char (string s, int &i, bool utf8) {
if (utf8) decode_from_utf8 (s, i);
else if (i < N(s)) {
if (s[i] == '<') {
i++;
while (i < N(s) && s[i] != '>') i++;
if (i < N(s)) i++;
}
else i++;
}
}
string
get_unicode_range (string c) {
string uc= cork_to_utf8 (c);
int pos= 0;
int code= decode_from_utf8 (uc, pos);
string range= "";
if (code <= 0x7f) range= "ascii";
else if (code >= 0x80 && code <= 0x37f) range= "latin";
else if (code >= 0x380 && code <= 0x3ff) range= "greek";
else if (code >= 0x400 && code <= 0x4ff) range= "cyrillic";
else if (code >= 0x3000 && code <= 0x303f) range= "cjk";
else if (code >= 0x4e00 && code <= 0x9fcc) range= "cjk";
else if (code >= 0xff00 && code <= 0xffef) range= "cjk";
else if (code >= 0xac00 && code <= 0xd7af) range= "hangul";
else if (code >= 0x2000 && code <= 0x23ff) range= "mathsymbols";
else if (code >= 0x2900 && code <= 0x2e7f) range= "mathextra";
else if (code >= 0x1d400 && code <= 0x1d7ff) range= "mathletters";
if (pos == N(uc)) return range;
return "";
}
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;
}