/**
* 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();
}
/**
* Returns an atomized content for any node kind. The atomized value can be an attribute value or
* XML content.
*
* @param pre pre value
* @return atomized value
*/
public final byte[] atom(final int pre) {
switch (kind(pre)) {
case TEXT:
case COMM:
return text(pre, true);
case ATTR:
return text(pre, false);
case PI:
byte[] txt = text(pre, true);
final int i = indexOf(txt, ' ');
return i == -1 ? EMPTY : substring(txt, i + 1);
default:
// create atomized text node
TokenBuilder tb = null;
byte[] t = EMPTY;
int p = pre;
final int s = p + size(p, kind(p));
while (p != s) {
final int k = kind(p);
if (k == TEXT) {
txt = text(p, true);
if (t == EMPTY) {
t = txt;
} else {
if (tb == null) tb = new TokenBuilder(t);
tb.add(txt);
}
}
p += attSize(p, k);
}
return tb == null ? t : tb.finish();
}
}
/**
* 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;
}
@Override
public byte[] info() {
final TokenBuilder tb = new TokenBuilder(LI_STRUCTURE).add(SORTED_LIST).add(NL);
final IndexStats stats = new IndexStats(data);
for (int m = 1; m < size; ++m) {
final int oc = len[m];
if (stats.adding(oc)) stats.add(key(m));
}
stats.print(tb);
return tb.finish();
}
@Override
public synchronized byte[] info(final MainOptions options) {
final TokenBuilder tb = new TokenBuilder();
final long l = inX.length() + inY.length() + inZ.length();
tb.add(LI_NAMES).add(data.meta.ftinclude).add(NL);
tb.add(LI_SIZE + Performance.format(l, true) + NL);
final IndexStats stats = new IndexStats(options.get(MainOptions.MAXSTAT));
addOccs(stats);
stats.print(tb);
return tb.finish();
}
/**
* 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 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;
}
@Override
public byte[] info(final MainOptions options) {
final TokenBuilder tb = new TokenBuilder();
tb.add(LI_STRUCTURE).add(HASH).add(NL);
tb.add(LI_NAMES).add(data.meta.names(type)).add(NL);
final IndexStats stats = new IndexStats(options.get(MainOptions.MAXSTAT));
final int s = values.size();
for (int p = 1; p <= s; p++) {
final int oc = lenList.get(p);
if (oc > 0 && stats.adding(oc)) stats.add(values.key(p), oc);
}
stats.print(tb);
return tb.finish();
}
/**
* Returns a regular expression pattern.
*
* @param pattern input pattern
* @param modifier modifier item
* @param ctx query context
* @return pattern modifier
* @throws QueryException query exception
*/
private Pattern pattern(final Expr pattern, final Expr modifier, final QueryContext ctx)
throws QueryException {
final byte[] pat = checkStr(pattern, ctx);
final byte[] mod = modifier != null ? checkStr(modifier, ctx) : null;
final TokenBuilder tb = new TokenBuilder(pat);
if (mod != null) tb.add(0).add(mod);
final byte[] key = tb.finish();
Pattern p = patterns.get(key);
if (p == null) {
p = RegExParser.parse(pat, mod, ctx.sc.xquery3(), info);
patterns.add(key, p);
}
return p;
}
/**
* 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 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;
}
