/**
* Consumes an Nmtoken. [7]
*
* @throws IOException I/O exception
*/
private void nmtoken() throws IOException {
final TokenBuilder name = new TokenBuilder();
int c;
while (isChar(c = nextChar())) name.add(c);
prev(1);
if (name.isEmpty()) error(INVNAME);
}
/**
* Adds some characters to the entity.
*
* @param ent token builder
* @throws IOException I/O exception
*/
private void completeRef(final TokenBuilder ent) throws IOException {
int ch = consume();
while (ent.size() < 10 && ch >= ' ' && ch != ';') {
ent.add(ch);
ch = consume();
}
}
/**
* Scans an external ID.
*
* @param f full flag
* @param r root flag
* @return id
* @throws IOException I/O exception
*/
private byte[] externalID(final boolean f, final boolean r) throws IOException {
byte[] cont = null;
final boolean pub = consume(PUBLIC);
if (pub || consume(SYSTEM)) {
checkS();
if (pub) {
pubidLit();
if (f) checkS();
}
final int qu = consume(); // [11]
if (qu == '\'' || qu == '"') {
int ch;
final TokenBuilder tok = new TokenBuilder();
while ((ch = nextChar()) != qu) tok.add(ch);
if (!f) return null;
final String name = string(tok.finish());
if (!dtd && r) return cont;
final XMLInput tin = input;
try {
final IO file = input.io().merge(name);
cont = file.read();
} catch (final IOException ex) {
Util.debug(ex);
// skip unknown DTDs/entities
cont = new byte[] {'?'};
}
input = new XMLInput(new IOContent(cont, name));
if (consume(XDECL)) {
check(XML);
s();
if (version()) checkS();
s();
if (encoding() == null) error(TEXTENC);
ch = nextChar();
if (s(ch)) ch = nextChar();
if (ch != '?') error(WRONGCHAR, '?', ch);
ch = nextChar();
if (ch != '>') error(WRONGCHAR, '>', ch);
cont = Arrays.copyOfRange(cont, input.pos(), cont.length);
}
s();
if (r) {
extSubsetDecl();
if (!consume((char) 0)) error(INVEND);
}
input = tin;
} else {
if (f) error(SCANQUOTE, (char) qu);
prev(1);
}
}
return cont;
}
/**
* Scans CDATA.
*
* @throws IOException I/O exception
*/
private void cDATA() throws IOException {
int ch;
while (true) {
while ((ch = nextChar()) != ']') token.add(ch);
if (consume(']')) {
if (consume('>')) return;
prev(1);
}
token.add(ch);
}
}
/**
* Consumes an XML name. [5]
*
* @param f force parsing
* @return name
* @throws IOException I/O exception
*/
private byte[] name(final boolean f) throws IOException {
final TokenBuilder name = new TokenBuilder();
int c = consume();
if (!isStartChar(c)) {
if (f) error(INVNAME);
prev(1);
return null;
}
do name.add(c);
while (isChar(c = nextChar()));
prev(1);
return name.finish();
}
/**
* Scans an entity value. [9]
*
* @param p pe reference flag
* @return value
* @throws IOException I/O exception
*/
private byte[] entityValue(final boolean p) throws IOException {
final int qu = consume();
if (qu != '\'' && qu != '"') {
prev(1);
return null;
}
TokenBuilder tok = new TokenBuilder();
int ch;
while ((ch = nextChar()) != qu) {
if (ch == '&') tok.add(ref(false));
else if (ch == '%') {
if (!p) error(INVPE);
tok.add(peRef());
} else {
tok.add(ch);
}
}
final XMLInput tmp = input;
input = new XMLInput(new IOContent(tok.finish()));
tok = new TokenBuilder();
while ((ch = consume()) != 0) {
if (ch == '&') tok.add(ref(false));
else tok.add(ch);
}
input = tmp;
return tok.finish();
}
/**
* Scans a processing instruction.
*
* @throws IOException I/O exception
*/
private void pi() throws IOException {
final byte[] tok = name(true);
if (eq(lc(tok), XML)) error(PIRES);
token.add(tok);
int ch = nextChar();
if (ch != '?' && !ws(ch)) error(PITEXT);
do {
while (ch != '?') {
token.add(ch);
ch = nextChar();
}
if ((ch = consume()) == '>') return;
token.add('?');
} while (true);
}
/**
* Scans an attribute value. [10]
*
* @param ch current character
* @throws IOException I/O exception
*/
private void attValue(final int ch) throws IOException {
boolean wrong = false;
int c = ch;
do {
if (c == 0) error(ATTCLOSE, (char) c);
wrong |= c == '\'' || c == '"';
if (c == '<') error(wrong ? ATTCLOSE : ATTCHAR, (char) c);
if (c == 0x0A) c = ' ';
if (c == '&') {
// verify...
final byte[] r = ref(true);
if (r.length == 1) token.add(r);
else if (!input.add(r, false)) error(RECENT);
} else {
token.add(c);
}
} while ((c = consume()) != quote);
}
/**
* Scans a comment.
*
* @throws IOException I/O exception
*/
private void comment() throws IOException {
do {
final int ch = nextChar();
if (ch == '-' && consume('-')) {
check('>');
return;
}
token.add(ch);
} while (true);
}
/**
* Scans XML text.
*
* @param ch current character
* @throws IOException I/O exception
*/
private void content(final int ch) throws IOException {
type = Type.TEXT;
boolean f = true;
int c = ch;
while (c != 0) {
if (c != '<') {
if (c == '&') {
// scan entity
final byte[] r = ref(true);
if (r.length == 1) token.add(r);
else if (!input.add(r, false)) error(RECENT);
} else {
if (c == ']') {
// ']]>' not allowed in content
if (consume() == ']') {
if (consume() == '>') error(CONTCDATA);
prev(1);
}
prev(1);
}
// add character to cached content
token.add(c);
}
} else {
if (!f && !isCDATA()) {
text = false;
prev(1);
if (chop) token.trim();
return;
}
cDATA();
}
c = consume();
f = false;
}
// end of file
if (!fragment) {
if (!ws(token.finish())) error(AFTERROOT);
type = Type.EOF;
}
}
/**
* Scans an XML tag.
*
* @param ch current character
* @throws IOException I/O exception
*/
private void scanTAG(final int ch) throws IOException {
int c = ch;
// scan tag end...
if (c == '>') {
type = Type.R_BR;
state = State.CONTENT;
} else if (c == '=') {
// scan equal sign...
type = Type.EQ;
} else if (c == '\'' || c == '"') {
// scan quote...
type = Type.QUOTE;
state = State.QUOTE;
quote = c;
} else if (c == '/') {
// scan empty tag end...
type = Type.CLOSE_R_BR;
if ((c = nextChar()) == '>') {
state = State.CONTENT;
} else {
token.add(c);
error(CLOSING);
}
} else if (s(c)) {
// scan whitespace...
type = Type.WS;
} else if (isStartChar(c)) {
// scan tag name...
type = state == State.ATT ? Type.ATTNAME : Type.ELEMNAME;
do token.add(c);
while (isChar(c = nextChar()));
prev(1);
state = State.ATT;
} else {
// undefined character...
error(CHARACTER, (char) c);
}
}
/**
* Scans a document encoding.
*
* @return encoding
* @throws IOException I/O exception
*/
private String encoding() throws IOException {
if (!consume(ENCOD)) {
if (fragment) error(TEXTENC);
return null;
}
s();
check('=');
s();
final TokenBuilder enc = new TokenBuilder();
final int d = qu();
int ch = nextChar();
if (letter(ch) && ch != '_') {
while (letterOrDigit(ch) || ch == '.' || ch == '-') {
enc.add(ch);
ch = nextChar();
}
prev(1);
}
check((char) d);
if (enc.isEmpty()) error(DECLENCODE, enc);
final String e = string(enc.finish());
input.encoding(e);
return e;
}
/**
* Scans a reference. [67]
*
* @param f dissolve entities
* @return entity
* @throws IOException I/O exception
*/
private byte[] ref(final boolean f) throws IOException {
// scans numeric entities
if (consume('#')) { // [66]
final TokenBuilder ent = new TokenBuilder();
int b = 10;
int ch = nextChar();
ent.add(ch);
if (ch == 'x') {
b = 16;
ent.add(ch = nextChar());
}
int n = 0;
do {
final boolean m = ch >= '0' && ch <= '9';
final boolean h = b == 16 && (ch >= 'a' && ch <= 'f' || ch >= 'A' && ch <= 'F');
if (!m && !h) {
completeRef(ent);
return QUESTION;
}
n *= b;
n += ch & 15;
if (!m) n += 9;
ent.add(ch = nextChar());
} while (ch != ';');
if (!valid(n)) return QUESTION;
ent.reset();
ent.add(n);
return ent.finish();
}
// scans predefined entities [68]
final byte[] name = name(false);
if (!consume(';')) return QUESTION;
if (!f) return concat(AMPER, name, SEMI);
byte[] en = ents.get(name);
if (en == null) {
// unknown entity: try HTML entities (lazy initialization)
if (HTMLENTS.size() == 0) {
for (int s = 0; s < HTMLENTITIES.length; s += 2) {
HTMLENTS.add(token(HTMLENTITIES[s]), token(HTMLENTITIES[s + 1]));
}
}
en = HTMLENTS.get(name);
}
return en == null ? QUESTION : en;
}
/**
* Reads and interprets the next token from the input stream.
*
* @return true if the document scanning has been completed
* @throws IOException I/O exception
*/
boolean more() throws IOException {
// gets next character from the input stream
token.reset();
final int ch = consume();
if (ch == 0) {
type = Type.EOF;
return false;
}
// checks the scanner state
switch (state) {
case CONTENT:
scanCONTENT(ch);
break;
case TAG:
case ATT:
scanTAG(ch);
break;
case QUOTE:
scanATTVALUE(ch);
}
return true;
}
